Structure-Based Design of Macrocyclic Coagulation Factor VIIa Inhibitors

Article abstract of DOI:10.1021/acs.jmedchem.5b00788

On the basis of a crystal structure of a phenylpyrrolidine lead and subsequent molecular modeling results, we designed and synthesized a novel series of macrocyclic FVIIa inhibitors. The optimal 16-membered macrocycle was 60-fold more potent than an acycl

Full text of DOI:10.1021/acs.jmedchem.5b00788

Subscriber access provided by University of Otago Library
Article
Structure-Based Design of Macrocyclic Coagulation Factor VIIa Inhibitors
E. Scott Priestley, Daniel L. Cheney, Indawati DeLucca, Anzhi Wei,
Joseph M. Luettgen, Alan R. Rendina, Pancras C Wong, and Ruth Wexler
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.5b00788 • Publication Date (Web): 07 Jul 2015
Downloaded from http://pubs.acs.org on July 11, 2015
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 free 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 accessible to all
readers and 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.
Journal of Medicinal Chemistry 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 40
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
StructureꢀBased Design of Macrocyclic Coagulation
Factor VIIa Inhibitors
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
E. Scott Priestley,* Daniel L. Cheney, Indawati DeLucca, Anzhi Wei, Joseph M. Luettgen, Alan R.
Rendina, Pancras C. Wong, and Ruth R. Wexler
BristolꢀMyers Squibb Research & Development, 311 PenningtonꢀRocky Hill Road, Pennington, NJ
08534.
RECEIVED DATE (to be automatically inserted after your manuscript is accepted if required
according to the journal that you are submitting your paper to)
Running title: Macrocyclic FVIIa Inhibitors.
* To whom correspondence should be addressed. Phone: (609) 818ꢀ5756. Fax: (609) 818ꢀ3331. Eꢀmail:
scott.priestley@bms.com.
ABSTRACT. Based on a crystal structure of a phenylpyrrolidine lead and subsequent molecular
modeling results, we designed and synthesized a novel series of macrocyclic FVIIa inhibitors. The
optimal 16ꢀmembered macrocycle was 60ꢀfold more potent than an acyclic analog. Further potency
optimization by incorporation of P1´ alkyl sulfone and P2 methyl groups provided a macrocycle with
TF/FVIIa Ki = 1.6 nM, excellent selectivity against a panel of seven serine proteases, and FVIIꢀdeficient
prothrombin time EC2x = 1.2 ꢀM. Discovery of this potent, selective macrocyclic scaffold opens new
possibilities for the development of orally bioavailable FVIIa inhibitors.
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 40
1
2
3
Introduction
4
5
6
7
8
9
The serine protease coagulation factor VIIa (FVIIa) plays a critical role in triggering the coagulation
cascade. Upon vessel injury or atherosclerotic plaque rupture, the membraneꢀbound glycoprotein tissue
factor (TF) is exposed to circulating blood containing FVIIa. The two proteins associate to form a
catalytically competent complex, TF/FVIIa, which activates its substrates, Factors IX and X, leading to
rapid local thrombin generation and thrombus formation.1
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Interest in the TF/FVIIa complex as an antithrombotic drug discovery target was sparked both by the
determination of its crystal structure2 and by in vivo studies demonstrating the antithrombotic activity of
active site inhibited FVIIa.3, 4 Subsequent studies of small molecule FVIIa inhibitors in preclinical
thrombosis and bleeding models show conclusively that inhibition of FVIIa affords excellent
antithrombotic efficacy along with low bleeding liability.5ꢀ12 Moreover, clinical trials with recombinant
nematode anticoagulant protein c2, a biological agent which binds to Factor X and inhibits TFꢀFVIIa,
demonstrate the antithrombotic efficacy and safety of this mechanism in humans.13ꢀ15 Despite this
excellent target validation, and extensive efforts to develop orally active FVIIa inhibitors from multiple
chemotypes, including 2ꢀarylbenzimidazoles,16ꢀ22 phenylglycines,23ꢀ26 pyrazinone and pyridone
34
36
peptidomimetics,27ꢀ31 tetrahydroquinolines,32 fluoropyridines,33,
amidinophenylureas,35,
biarylacids,37ꢀ41 αꢀketoacids,42 and phenylimidazoles,43 amongst others, there are no reports of orally
active compounds progressing into clinical trials.44 These small molecule FVIIa inhibitors typically
contain a cationic amidine group, which contributes significantly to their inhibitory potency by forming
strong, bivalent hydrogen bonds with Asp189 at the bottom of the S1 pocket of FVIIa.45 However, due
to their strongly basic pKa and high polar surface area which leads to poor intrinsic membrane
permeability, amidines are well known to have poor oral absorption and typically require a prodrug
strategy.10, 17, 21, 24 Alternatively, several groups have attempted to replace the amidine group in FVIIa
inhibitors with less basic, more permeable alternatives,19, 20, 22, 25, 38, 39, 41 with limited success to date,
although a recent fragment screening effort has identified promising neutral S1 binders.46 This strategy
2
ACS Paragon Plus Environment

Page 3 of 40
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
has proven successful with inhibitors of other coagulation proteases, particularly coagulation factor
Xa,47ꢀ49 where highly optimized, potent scaffolds have enabled replacement of amidines with more
permeable functional groups. Against this background, the discovery of novel FVIIa inhibitor scaffolds
with high intrinsic potency is an important research goal, as it may allow replacement of the
benzamidine group with a less basic, more permeable moiety to achieve oral bioavailability.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
When faced with challenging protease targets, medicinal chemists now frequently employ
macrocyclization as a strategy to improve the properties of their lead compounds. Macrocycles have
been shown to possess improved potency,50 selectivity,51, 52 and pharmaceutical and pharmacokinetic
properties relative to acyclic precursors.53ꢀ56 Moreover, many macrocyclic natural products or their
derivatives are in clinical use, including erythromycin, vancomycin, cyclosporin, and epothilone B, and
numerous syntheticallyꢀderived macrocycles targeting HCV NS3 protease are in late stage clinical trials
or have recently received regulatory approval.57 In this report, we disclose our initial discovery of a
series of rationally designed macrocycles with potent and selective inhibitory activity against FVIIa.
Design
During the course of our program to discover orally active FVIIa inhibitors, we investigated a series of
phenylglycinamides containing pyrrolidine as a conformational constraint.26 Compound 1, containing
an aminoisoquinoline P1 and an unsubstituted phenyl P2 group (Figure 1A), was found to be a potent
TF/FVIIa inhibitor (TF/FVIIa Ki = 8.0 nM, FVIIꢀdeficient prothrombin time12 EC2x = 12 ꢀM). A crystal
structure with FVIIa was subsequently obtained (Figure 1B), revealing the basic 1ꢀaminoisoquinoline
group to be bound as expected in the S1 pocket, forming a salt bridge with the Asp189 side chain and a
hydrogen bond with a conserved water molecule above the Tyr228 aryl ring. The central amide carbonyl
oxygen is engaged in hydrogen bonds with the protonated form of the His57 imidazole ring and the
Ser195 hyrdroxyl group. The P1´ and P2 phenyl rings have undergone hydrophobic collapse and are
engaged in favorable van der Waals interactions in S2, specifically with the side chains of His57,
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 40
1
2
3
4
5
6
7
Trp215 and Thr99. The acetanilide NH engages in hydrogen bonding with the structural water bridging
the Thr98 carbonyl and Asp60 side chain, while the acetanilide methyl group forms van der Waals
contacts with the 97ꢀ98 backbone. The P1´ sulfone isopropyl group forms a close contact with the
8
9
Cys42Sγ (closest C – S contact = 3.8 Å). Importantly, the proximity of the acetanilide methyl and P2
phenyl groups suggested that these positions might be linked to generate macrocyclic inhibitors (Figure
1C), which we hoped would have increased potency and selectivity and would provide a novel template
for optimization to an orally bioavailable inhibitor.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
A)
B)
C)
X
n
P2
NH2
O
NH
P2
NH2
N
H
N
N
O
NH
N
H
O
P1
P1
′
N
H
O
P1
N
compd
X
n
2
2
1
0
2
3
4
5
O
O
O
S
CH2
CH2
CH2
P1
′
1
Figure 1. (A) Pyrrolidine phenylglycinamide lead 1. (B) Crystal structure of 1 bound to FVIIa at 2.6 Å
resolution.58,59 The proximity of the P1´ acetanilide and P2 phenyl groups suggested that
macrocyclization was feasible, as indicated by the doubleꢀheaded yellow arrow. (C) Initial macrocycle
synthetic targets.
A series of macrocyclic structures containing various linkers joining the para position of the P2
phenyl with the meta position of the P1´ phenyl were initially evaluated using a simulated annealing
protocol implemented in the molecular dynamics software package Discover_3.60 Structures which
maintained key van der Waals and hydrogen bonding interactions were then subjected to conformational
analysis to estimate the ligand strain induced by binding. In order to ensure an acceptable level of
accuracy and reliability, we elected to assess the ligand strain energy using quantum chemical methods
4
ACS Paragon Plus Environment

Page 5 of 40
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
(for details, see the Experimental section). Of the 15 linker structures evaluated, butanamide appeared
most promising in that key intermolecular interactions were maintained while incurring little binding
strain. Figure 2 shows that the calculated global quantum chemical energy minimum for a macrocycle
containing the butanamide linker closely matches the crystallographically determined FVIIa bound
conformation of 1. Longer 4ꢀoxybutanamide and pentanamide linkers, while also engaging in the key
interactions, exhibited higher degrees of binding strain and greater flexibility, each possessing a
relatively large number of lower energy conformations. Other linkers either did not maintain key
intermolecular interactions or showed higher binding strain. Based on these results, we targeted the
synthesis of macrocycles 2ꢀ4, containing the 4ꢀoxybutanamide, pentanamide, and butanamide linkers,
respectively. We also targeted 5, containing a shorter propionamide linker, in order to generate a
homologous series of 15ꢀ to 17ꢀmembered macrocycles (Figure 1C). We retained the aminoisoquinoline
group of 1 in our targeted compounds, as it is less basic than benzamidine61 and offers better prospects
for achieving oral bioavailability.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Overlay of the calculated global quantum chemical energy minimum of macrocycle 4
(orange), containing the butanamide linker, on the crystal structure of 1 (white) bound to FVIIa.59
Chemistry
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 40
1
2
3
4
5
6
7
8
As shown in Scheme 1, the general synthetic route for preparing macrocycles 2-5 involves 3 key
steps: 1) preparation of appropriately functionalized boronic acids, 2) Petasis reaction with a P1 aniline
intermediate (9) and glyoxylic acid to form the phenylglycine core (10),62 and 3) macrocyclization to
form the upper amide bond. Specifically, for macrocycle 2, 4ꢀ(4ꢀbromophenylphenoxy)butyric acid 6a
was coupled with bis(neopentyl glycolato)diboron,63 followed by hydrolysis to the free boronic acid 7a.
The carboxylate was alkylated with benzyl bromide to give ester 8a. Petasis reaction of 8a with bisꢀBoc
protected 1ꢀaminoisoquinoline intermediate 926 and glyoxylic acid afforded the key phenylglycine
intermediate 10a in 68% yield. Subsequent amide coupling to mꢀnitrobenzylamine provided the fully
protected macrocyclization precursor 11a. Deprotection of the benzyl ester and reduction of the nitro
group was accomplished by hydrogenation, followed by macrolactamization with EDC/HOAt and final
Boc removal with hydrogen chloride, which gave macrocycle 2 in 25 % yield over three steps. In a
similar fashion, commercially available acids 6b, 6c, and 7d were elaborated to macrocycles 3, 4, and 5,
respectively. Macrocycle 4 was separated by chiral HPLC into its enantiomers (R)-4 and (S)-4. Acyclic
compounds 15 and 16 were prepared by Petasis reaction and amide formation as shown in Scheme 1.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 1. Synthesis of macrocycles 2ꢀ5a
6
ACS Paragon Plus Environment

Page 7 of 40
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
(a) bis(neopentyl glycolato)diboron, PdCl2(dppf), KOAc, DMSO, 80 °C; (b) 2M NaOH; (c) 6M
HCl(aq); (d) benzyl bromide, KHCO3, DMF, 60
°C; (e) glyoxylic acid monohydrate, 1,2ꢀdichloroethane,
100 C, ꢀwave; (f) mꢀnitrobenzylamine·HCl, EDC, HOAt, DIEA, CH2Cl2, DMF; (g) H2, 10% Pd/C,
°
MeOH, (h) EDC, HOAt, DIEA, CH2Cl2, DMF; (i) 4N HCl/dioxane/EtOAc (2, 5) or 50% TFA/CH2Cl2
(3, 4); (j) chiral HPLC separation (Chiralcel OD, 30% (1:1 MeOH:EtOH)/heptane); (k) phenylboronic
acid, glyoxylic acid monohydrate, acetonitrile, DMF, 100 °C, ꢀwave; (l) benzylamine, BOP, DIEA; (m)
mꢀnitrobenzylamine, BOP, DIEA; (n) Zn, NH4Cl, MeOH, THF; (o) AcCl, DIEA, acetonitrile; (p) TFA,
CH2Cl2.
Initial attempts to prepare sulfonylꢀsubstituted macrocycles 23ꢀ25 and 31 (Scheme 2) via
macrolactamization at the anilide bond were unsuccessful presumably due to the poor nucleophilicity of
the electronꢀdeficient paraꢀsulfonylaniline. A modified route, in which the anilide bond was formed
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 40
1
2
3
4
5
6
7
8
early in the synthesis, followed by macrolactamization at the phenylglycine amide, successfully
delivered the target compounds (Scheme 2). Beginning with 2ꢀfluoroꢀ5ꢀnitrobenzonitrile 17, substitution
with ethanethiol followed by oxidation gave ethylsulfone 18a. The nitro group of 18a was reduced to
give aniline 19a by hydrogenation and then reacted with the acid chloride prepared from 6c in the
presence of DMAP to give anilide 20a. Anilide 20a was cross coupled with bis(neopentyl
glycolato)diboron and the resulting boronic ester was isolated as the diethanolamine adduct,64 and then
hydrolyzed to boronic acid 21a during purification. Petasis reaction of 21a with aminoisoquinoline 9
and glyoxylic acid gave phenylglycine intermediate 22a. The latent benzylamine in 22a was revealed by
hydrogenation of the nitrile. Subsequent BOPꢀmediated macrolactamization, Boc deprotection, and
chiral HPLC separation provided (R)-23 in 20% yield over 4 steps. (R)-24 and (rac)-25 were prepared in
similar fashion from 17. In the case of benzamidine macrocycle (R)-31, the nitrile group in precursor
18a was reduced and Boc protected early in the synthesis to give benzylamine 26, which was elaborated
to boronic acid 28. Petasis reaction of 28 with p-aminobenzonitrile and glyoxylic acid gave
phenylglycine intermediate 29. Subsequent Boc deprotection, BOPꢀmediated macrolactamization, and
chiral separation afforded benzonitrile (R)-30. This benzonitrile was converted to the amidoxime by
treatment with hydroxylamine, followed by acetylation and hydrogenation to provide benzamidine (R)-
31.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 2. Synthesis of macrocycles 23-25 and 31a
8
ACS Paragon Plus Environment

Page 9 of 40
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
(a) RSH, Et3N, DMF; (b) mꢀCPBA, CH2Cl2; (c) H2, 10% Pd/C, MeOH; (d) 6c, (COCl)2, DMF,
CH2Cl2, then DMAP, Et3N, 1,2ꢀDCE, reflux; (e) bis(neopentyl glycolato)diboron, PdCl2(dppf), KOAc,
DMSO, 80 C; (f) diethanolamine (g) 9, glyoxylic acid monohydrate, DCE, ꢀwave, 100 C; (h) H2, 10%
°
°
Pd/C, 1M HCl(aq), MeOH; (i) BOP, DIEA, DMF; (j) 50% TFA, CH2Cl2; (k) chiral HPLC (Chiralcel ODꢀ
H, 30% (1:1 MeOH:EtOH) in heptane); (l) H2, 10% Pd/C, 1M HCl(aq), MeOH, 9 days; (m) BOCꢀON,
Et3N, THF; (n) pꢀaminobenzonitrile, glyoxylic acid monohydrate, CH3CN, DMF, ꢀwave, 100 °C; (o)
4N HCl/dioxane; (p) BOP, DMAP, CH2Cl2, DMF, 40 °C; (q) chiral HPLC (Chiralcel OD, 30% (1:1
MeOH:EtOH) in heptane); (r) NH2OH, DMSO; (s) Ac2O, CH2Cl2, then H2, 10% Pd/C, MeOH.
Methylated macrocycle (R)-37 was prepared as described in Scheme 3. Methyl butꢀ3ꢀenoate 32 was
hydroborated with 9ꢀBBN, coupled with 5ꢀbromoꢀ2ꢀiodotoluene using PdCl2(dppf),65 and then
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 40
1
2
3
4
saponified to give carboxylate 33. Subsequent amide formation with 19a provided 34, which was
advanced to (R)-37 using a sequence analogous to that which was used to prepare (R)-23 from 20a.
5
6
7
8
9
Scheme 3. Synthesis of macrocycle 37a
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
(a) 9ꢀBBN; (b) 5ꢀbromoꢀ2ꢀiodotoluene, PdCl2(dppf), NaOMe, THF, 70
C; (d) (COCl)2, DMF, CH2Cl2; (e) 19a, toluene, reflux; (f) bis(neopentyl glycolato)diboron,
PdCl2(dppf), KOAc, DMSO, 80 C; (g) NaOH (aq); (h) 9, glyoxylic acid monohydrate, CH3CN, DMF,
ꢀwave, 100 C; (i) H2, 10% Pd/C, 1M HCl(aq), MeOH; (j) BOP, DIEA, DMF; (k) 50% TFA, CH2Cl2; (l)
°C; (c) NaOH, THF, MeOH,
80
°
°
°
chiral SFC (Chiralpak AS, 75% CO2/25% MeOH/0.1% diethylamine, 100 bar, 40 °C).
Results and Discussion
Based on the results of the molecular modeling experiments, macrocycles 2-5 were prepared and
assayed for inhibition of TF/FVIIa (Table I). Consistent with the predictions from the modeling data, the
16ꢀmembered macrocycle 4 was at least 5ꢀfold more potent than either 17ꢀmembered macrocycles 2 and
3 or 15ꢀmembered macrocycle 5. Comparison of macrocycle 4 with acyclic analog 15 indicated that
incorporation of the optimized linker improved potency 60ꢀfold. Comparison of 4 and 15 with acyclic
compound 16 containing an acetamide group (representing the portion of the linker which interacts with
S2) shows that the acetamide moiety contributes 5ꢀfold to potency. Thus, the embedded amide is a key
element of the macrocyclic linker of 4, and the conformational constraint imposed by macrocyclization
10
ACS Paragon Plus Environment

Page 11 of 40
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
presumably contributes an additional 11ꢀfold potency increase relative to 16. Racemic macrocycle 4 was
separated into its enantiomers (R)-4 and (S)-4 by chiral HPLC, and all of the inhibitory activity was
found to reside in the Rꢀenantiomer (Table I, see below for stereochemical assignment).
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table I. Initial Linker Optimization of Phenylglycine Macrocycles
compd
La
M
TF/FVIIa
Ki (nM) b
ꢀOCH2CH2CH2ꢀ ꢀC(=O)NHꢀ
ꢀCH2CH2CH2CH2ꢀ ꢀC(=O)NHꢀ
4600
6100
2
3
ꢀCH2CH2CH2ꢀ
ꢀCH2CH2CH2ꢀ
ꢀCH2CH2CH2ꢀ
ꢀCH2CH2ꢀ
H
ꢀC(=O)NHꢀ
ꢀC(=O)NHꢀ
ꢀC(=O)NHꢀ
ꢀC(=O)NHꢀ
H
920
4
400
(R)-4
(S)-4
5
>18000
4600
54000
10000
15
H
CH3C(=O)NHꢀ
16
a
atom connectivity is P2 phenylꢀLꢀMꢀP1´ phenyl as the linker atoms are
b
shown left to right. TF/FVIIa Ki values were obtained from recombinant
human enzyme and were averaged from two experiments.12
We crystallized FVIIa with racemic macrocycle 4 and determined the structure of the resulting
complex at 2.06 Å resolution. In the crystal structure, the active site of FVIIa was occupied by the Rꢀ
enantiomer of 4 (Figure 3A), consistent with the preferred stereochemistry previously observed for the
pyrrolidine phenylglycinamide series.26 The stereochemistry of the active enantiomer of 4 (Table I), and
the other macrocycles containing this linker was therefore assigned as R. An overlay of (R)-4 with
phenylpyrrolidine 1 indicates that the fiveꢀatom butanamide linker provides an essentially optimal
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 40
1
2
3
4
5
6
7
8
geometric constraint (Figure 3B), in full agreement with the molecular modeling data (Figure 2).
Macrocycle (R)-4 thus engages in very similar intermolecular interactions as those described for 1 in
Figure 1. Key interactions with Asp 189 in S1 are maintained and the linker anilide NH engages in a
hydrogen bond with the aforementioned conserved water molecule in S2 (Figure 3A).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(A)
(B)
Figure 3. (A) Crystallographic structure of (R)-4 in FVIIa at 2.06 Å resolution.59,66 (B) Overlay of 1 and
(R)-4 bound to FVIIa.67
Based on structure activity relationships established in the phenylpyrrolidine series,26 in which a P1´
alkyl sulfone can improve potency >20ꢀfold relative to hydrogen, and the excellent structural overlay of
1 with (R)-4 shown in Figure 3B, we incorporated a P1´ alkyl sulfone group to improve the potency of
the macrocycles (Scheme 2). As shown in Table II, introduction of an ethylsulfone to give (R)-23
improved TF/FVIIa potency 50ꢀfold relative to (R)-4 and improved activity in the FVIIꢀdeficient
prothrombin time assay by 16ꢀfold, demonstrating promising activity in plasma.12 Additionally, (R)-23
has significantly improved selectivity against a panel of seven related serine proteases, particularly vs.
activated protein C (APC) and tissue kallikreinꢀ1, with 160ꢀfold and 12ꢀfold selectivity, respectively.
Isopropylsulfone (R)-24 is slightly more potent than (R)-23, while tertꢀbutyl sulfone (rac)-25 is about 2ꢀ
12
ACS Paragon Plus Environment

Page 13 of 40
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
fold less potent than (R)-23, considering that the former is racemic. Incorporation of a highly basic
benzamidine P1 group afforded (R)-31, which has 11ꢀfold improved potency relative to
aminoisoquinoline (R)-23, but reduced selectivity (58ꢀ and 83ꢀfold selectivity vs. APC and trypsin,
respectively). Incorporation of a methyl group in (R)-37 as described in Scheme 3, in order to fill the
small, hydrophobic S2 pocket in FVIIa, improved potency 5ꢀfold relative to (R)-23, and increased
selectivity vs. the serine protease panel (APC and tissue kallikreinꢀ1 selectivity for (R)-37 is >500ꢀfold
and 62ꢀfold, respectively). Importantly, (R)-37 has 5ꢀfold improved TF/FVIIa potency and 10ꢀfold
improved FVIIꢀdeficient prothrombin time EC2x relative to the original phenylpyrrolidine lead, 1. This
high level of enzyme inhibitory potency and activity in plasma is promising for further optimization. It
is also important to note that (R)-37 exists as a 1:1 mixture of two conformational isomers at room
temperature (likely atropisomers due to hindered rotation of the unsymmetrically substituted P2 phenyl
group). Control of the conformational dynamics of this scaffold may enable additional potency and
selectivity improvements.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table II. Potent Macrocyclic Sulfones a
NH2
N
Y
AIQ =
O
NH
X
NH
H
P1
*
N
N
BZD = H2N
H
O
compd P1
X
H
Y
TF/FVIIa FIXa
FXa Thrombin APC Trypsin Plasma Tissue FVII def. b
Ki (nM)
Ki
Ki
Ki
Ki
Ki
Kall. Ki Kall. Ki PT EC2x
(nM) (nM)
(nM)
(nM) (nM)
(nM)
(nM)
(ꢀM)
AIQ
H
H
400
8.0
>35000 >9000 >13000 2800 >6000 >6000
18
26
(R)-4
AIQ ꢀSO2Et
>35000 3200
>35000 3300
5800
4000
1300 3600
1800 3000
4600
3600
100
1.6
(R)-23
AIQ ꢀSO2ꢀiꢀPr
H
5.2
100
1.4
(R)-24
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 40
1
2
3
4
5
6
7
8
9
AIQ ꢀSO2ꢀtꢀBu H
30
0.70
1.6
>35000 4800 >13000 14000 >6000 >6000
22
8.6
0.42
1.2
(rac)-25
(R)-31
BZD ꢀSO2Et
H
34000 5500
16000 1000
180
41
58
50
1900
100
AIQ ꢀSO2Et Me
6700
830
2700
1700
(R)-37
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
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 room
temperature (25 °C), except tissue kallikreinꢀ1, which was run at 37 °C. See reference 12 for a detailed description
of the enzyme assay protocols. APC = activated protein C. Tissue Kall. = tissue kallikreinꢀ1. b see reference 12 for a
detailed description of the FVIIꢀdeficient prothrombin time assay.
Conclusions
Based on a crystal structure of phenylpyrrolidine lead 1 and subsequent molecular modeling, we
designed a novel series of macrocyclic FVIIa inhibitors. The optimal 16ꢀmembered macrocycle 4 was
60ꢀfold more potent than the corresponding acyclic analog 15. Further potency optimization by
incorporation of P1´ ethyl sulfone and P2 methyl substituents provided (R)-37, with TF/FVIIa Ki = 1.6
nM, excellent selectivity against a panel of seven serine proteases, and FVIIꢀdeficient prothrombin time
EC2x = 1.2 ꢀM. The macrocyclic FVIIa inhibitor chemotype exemplified by (R)-37 provides a novel,
potent, and selective scaffold for further optimization towards an orally bioavailable drug candidate. The
results of our efforts to achieve this goal will be reported in due course.
Experimental Section
Crystallization, data collection and structure refinement. Recombinant fullꢀlength FVIIa was
purchased from Novo Nordisk. The GLA domain (residues 1–44) was removed by Cathepsin G
digestion. The final protein concentration was approximately 25 mg/ml. Single crystals suitable for xꢀ
ray analysis were obtained by vapor diffusion. The crystals were grown at 4° C with 20% PEG 6000, 20
mM CaCl2, 0.1M MES (pH 6.0) in the reservoir.
Data for the FVIIa crystals in complex with ligands were collected at 17ꢀID [IMCA Collaborative
Access Team (IMCAꢀCAT)] located at the Advanced Photon Source (APS) or in the laboratory with a
14
ACS Paragon Plus Environment

Page 15 of 40
Journal of Medicinal Chemistry
1
2
3
Rigaku FRꢀE superbright generator, Rigaku MicroMax confocal optics, and a Rigaku Saturn 92
4
5
detector.
6
7
8
9
Image data were processed with the program HKL.68 The CNX program (Accelrys) was used for
crystallographic refinement.69 The atomic coordinates of human FVIIa (1DAN)2 were used as a search
model. Molecular replacement and rigid body refinement were performed with 8.0 to 3.5 Å resolution
data. The FVIIa coordinates were subjected to a few cycles of atomic positional refinements. Simulated
annealing (at a maximum temperature of 3000 K) was followed by Bꢀfactor refinement. Examination of
the resulting FoꢀFc electron density map revealed the density for the compound and allowed
unambiguous fitting of the compound. The inhibitors were built with the program QUANTA
(Accelrys). Peaks in the difference electron density map which were greater than 3σ and which were less
than 3.5 Å away from the protein were built in as solvent molecules. No major adjustments to the
protein model were needed during the course of the refinements. The structure was reꢀrefined prior to
deposition using BUSTER (GlobalPhasing, Ltd), GRADE (GlobalPhasing, Ltd), for the ligand restraint
files, RHOFIT (GlobalPhasing, Ltd), for ligand placement and COOT70 for electron density map fitting.
Coordinates and structure amplitudes have been deposited in the PDB with accession codes of 4ZXX
and 4ZXY.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Molecular Modeling. Idea structures were initially evaluated using a simulated annealing protocol
implemented in the Discover_3 molecular dynamics package60 and utilizing the CFF98 forcefield.71ꢀ75
Typically, a model of the idea structure bound in FVIIa was subjected to molecular dynamics at 310 K
(gas phase with a distance dependent dielectric), while allowing several contact amino acids and side
chains in the active site to relax. Snapshots were taken at 10 picosecond intervals, and then minimized.
Idea structures which maintained key van der Waals and hydrogen bonding interactions were then
subjected to an extended conformational analysis to assess the ligand strain requirement upon binding.
Initial conformation generation was performed using the mixed Monte Carloꢀlow mode search algorithm
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 40
1
2
3
4
5
6
7
8
in the MacroModel software package using the OPLSꢀAAꢀ2001 forcefield with the GBSA solvation
model.76 To avoid unrealistic intramolecular electrostatic interactions, the amidine was neutralized in
this and subsequent steps in the conformational analysis. Conformations within 5 kcal of the observed
global minimum77, 78 were further minimized with the quantum chemistry software package Jaguar using
B3LYP/6ꢀ31G**.79 Final single point energies were computed using the local MP2 method
implemented in Jaguar with the pseudospectral ccꢀpVTZ(ꢀf)++ basis set. In general, the observed global
minimum was taken as the lowest energy structure. Strain of the bound conformation was estimated in
two ways. Method 1: The annealed idea structure was removed from the protein and the equilibrium
structure was obtained at B3LYP/ 6ꢀ31G** in which only angles and bonds were allowed to equilibrate.
The strain was assessed by comparing the LMP2ꢀ/ ccꢀPVTZ energies of this partially relaxed pose and
the observed global minimum. Method 2: Identical to Method 1 except the idea structure was allowed to
fully relax to its nearest local minimum. We sometimes observed that although this structure was
conformationally similar to the one obtained in Method 1, it was significantly lower in energy and so
provided a better measure of strain. See Supporting Information Figure S1 for a summary of the linkers
evaluated.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Biological Assays. The TF/FVIIa, FIXa, FXa, thrombin, activated protein C, trypsin, plasma
kallikrein, and tissue kallikreinꢀ1 enzyme assays and the FVIIꢀdeficient prothrombin time assay were
performed according to published protocols.12
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. Flash chromatography was carried out on
ISCO CombiFlash 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. Purity of all final
16
ACS Paragon Plus Environment

Page 17 of 40
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
compounds was determined to be ≥ 95% by analytical HPLC using the following conditions:
Phenomenex Luna C18 column (4.6 x 75 mm) eluted at 4 mL/min with an 8 min gradient from 100%
solvent A to 100% solvent B (solvent A: 10% methanol, 89.9% water, 0.1% H3PO4; solvent B: 90%
methanol, 9.9% water, 0.1% H3PO4), monitoring uv absorbance at 220 nM.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4-(4-Boronophenoxy)-butyric acid (7a).
A
resealable tube was charged with 4ꢀ(4ꢀ
bromophenoxy)butyric acid 6a, (259 mg, 1.0 mmol), bis(neopentyl glycolato)diboron (249 mg, 1.1
mmol), potassium acetate (245 mg, 2.5 mmol), and DMSO (2 mL). The resulting orange suspension was
deoxygenated by sparging with nitrogen gas. Dichloro[1,1’ꢀbis(diphenylphosphino)ferrocene]
palladium(II) dichloromethane adduct (30 mg, 0.041 mmol) was added, and the tube was sealed tightly
and heated at 80°C overnight. Hydrochloric acid (1N) was added, and the mixture was extracted with
EtOAc (2x), washed with water (2x) and brine (1x), and dried (MgSO ). The organic layer was
4
concentrated in vacuo and the residue purified by flash chromatography (0 to 15% MeOH in DCM) to
give the 2,2ꢀdimethylꢀ1,3ꢀpropanediol boronic ester of 7a. This material was dissolved in diethyl ether
and washed with NaOH (2 N, 2x). The aqueous layers were washed with diethyl ether, combined, and
acidified to pH 4 with hydrochloric acid (6 N). The resulting solid precipitate was collected by filtration
1
to afford 7a (210 mg, 94%) as a beige solid. H NMR (400 MHz, CD OD) δ 2.06 (m, 2H), 2.48 (t, J =
3
7.5 Hz, 2H), 4.02 (t, J = 6.6 Hz, 2H), 6.88 (br s, 2H), 7.62 (br d, 2H).
4-(3-Benzyloxycarbonyl-propoxy)-phenylboronic acid (8a). A solution of 7a (200 mg, 0.89 mmol),
potassium bicarbonate (313 mg, 3.1 mmol), and benzyl bromide (0.163 mL, 1.4 mmol) in DMF (2 mL)
was heated at 60°C for 8 h. The reaction mixture was concentrated in vacuo and the residue was
partitioned between EtOAc and hydrochloric acid (1 N). The aqueous layer was extracted with EtOAc
(2x) and then the combined organics were washed with water (3x) and brine, dried (MgSO ), and
4
concentrated in vacuo. The residue was purified by flash chromatography (30 to 100% EtOAc in
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 40
1
2
3
1
hexane) to afford 8a (86 mg, 28%) as a white solid. H NMR (400 MHz, CD OD) δ 2.08 (m, 2H), 2.56
3
4
5
6
7
8
9
(t, J = 7.0 Hz, 2H), 4.00 (m, 2H), 5.12 (s, 2H), 6.84 (2 x d, 2H), 7.32 (m, 5H), 7.7ꢀ7.5 (2 x d, 2H).
4-{4-[(1-Di-tert-butoxycarbonylamino-isoquinolin-6-ylamino)-carboxy-methyl]-phenoxy}-
butyric acid benzyl ester (10a). A solution of 8a (82 mg, 0.26 mmol), 9 (72 mg, 0.20 mmol),26 and
glyoxylic acid monohydrate (22 mg, 0.24 mmol) in DCE (1 mL) was heated at 100°C for 10 min in a
microwave reactor. This solution was purified by flash chromatography (0 to 15% MeOH in DCM) to
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
give 10a contaminated with 8a (110 mg, 2.5:1 10a/8a, 68% yield based on content of 10a) as a yellow
1
oil. MS (ESI) m/z 686.3 (M+H)+. H NMR (400MHz, CD OD) δ 1.21ꢀ1.30 (m, 18H), 1.97ꢀ2.10 (m,
3
2H), 2.46ꢀ2.55 (m, 2H), 3.88ꢀ4.02 (m, 2H), 5.09 (s, 2H), 5.17 (s, 1H), 6.68 (d, J = 1.8 Hz, 1H), 6.85 (d, J
= 8.3 Hz, 2H), 7.21ꢀ7.34 (m, 7H), 7.41 (d, J = 6.2 Hz, 1H), 7.46 (d, J = 8.8 Hz, 2H), 7.63 (d, J = 9.2 Hz,
1H), 8.01 (d, J = 5.7 Hz, 1H).
4-{4-[(1-Di-tert-butoxycarbonylamino-isoquinolin-6-ylamino)-(3-nitro-benzylcarbamoyl)-
methyl]-phenoxy}3-butyric acid benzyl ester (11a). A solution of 10a (110 mg, 0.16 mmol), 3ꢀ
nitrobenzylamine hydrochloride (36 mg, 0.19 mmol), DIEA (0.084 mL, 0.48 mmol), HOAt (22 mg, 0.16
mmol), and EDC (62 mg, 0.32 mmol) in a mixture of DCM (2 mL) and DMF (0.5 mL) was stirred at rt
overnight. The reaction mixture was concentrated in vacuo and the residue was triturated with water and
then purified by flash chromatography (0 to 10% MeOH in DCM) to give 11a (88 mg, 67%). MS (ESI)
m/z 820.3 (M+H)+. 1H NMR (400MHz, CD OD) δ 1.21ꢀ1.30 (m, 18H), 1.23ꢀ1.24 (m, 1H), 2.08 (quin, J
3
= 6.6 Hz, 2H), 2.56 (t, J = 7.3 Hz, 2H), 4.00 (t, J = 5.9 Hz, 2H), 4.37ꢀ4.61 (m, 1H), 5.01ꢀ5.18 (m, 3H),
6.69 (s, 1H), 6.91 (d, J = 8.8 Hz, 2H), 7.25ꢀ7.35 (m, 7H), 7.39 (d, J = 6.6 Hz, 2H), 7.46 (d, J = 8.3 Hz,
2H), 7.50ꢀ7.58 (m, 1H), 7.61ꢀ7.70 (m, 1H), 7.97 (s, 1H), 7.99ꢀ8.03 (m, 2H), 8.06 (d, J = 6.2 Hz, 1H).
6,10
2-(1-Amino-isoquinolin-6-ylamino)-16-oxa-4,11-diaza-tricyclo[15.2.2.1 ]docosa-
1(20),6,8,10(22),17(21),18-hexaene-3,12-dione trifluoroacetic acid salt (2). A solution of 11a (88 mg,
0.11 mmol) in MeOH was hydrogenated (55 psi) over 10% palladium on carbon (36 mg) for 3 h. The
18
ACS Paragon Plus Environment

Page 19 of 40
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
reaction mixture was filtered and concentrated in vacuo to give a yellow glass (73 mg). A portion of this
material (63 mg) in a mixture of DCM (10 mL) and DMF (0.5 mL) was treated with DIEA (0.047 mL,
0.27 mmol), HOAt (12 mg, 0.088 mmol), and EDC (35 mg, 0.18 mmol) and the mixture was stirred at rt
overnight. The reaction mixture was concentrated in vacuo. A solution of the residue combined with the
product from 2 mg and 4 mg scale pilot reactions in EtOAc (0.5 mL) and hydrogen chloride in dioxane
(1 mL, 4N) was stirred at rt for 3 h. The reaction mixture was concentrated in vacuo and the residue was
purified by reverse phase HPLC to give 2 (15 mg, 25%) as a white solid. MS (ESI) m/z 482.3 (M+H)+.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1H NMR (400 MHz, CD OD) δ 2.03 ꢀ 2.19 (m, 2H), 2.44 ꢀ 2.58 (m, 2H), 3.95 (dd, J = 15.38, 3.52 Hz,
3
1H), 4.03ꢀ4.19 (m, 2H), 4.73 (dd, J = 15.38, 8.35 Hz, 1H), 5.09 (s, 1H), 5.50 (s, 1H), 6.70 (s, 1H), 6.82
(d, J = 7.03 Hz, 1H), 6.93 (d, J = 8.35 Hz, 3H), 7.10 ꢀ 7.20 (m, 3H), 7.29 (d, J = 7.03 Hz, 1H), 7.46 (d, J
= 7.03 Hz, 2H), 7.52 (d, J = 7.47 Hz, 1H), 8.05 (d, J = 9.23 Hz, 1H), 8.49 (dd, J = 7.91, 3.95 Hz, 1H).
2-(Ethylsulfonyl)-5-nitrobenzonitrile (18a). Ethanethiol (2.8 mL, 38 mmol) was added to a solution
of 2ꢀfluoroꢀ5ꢀnitrobenzonitrile 17 (5.00 g, 30.1 mmol) and triethylamine (9.3 mL, 67 mmol) in DMF
(100 mL). The reaction mixture was stirred for 1 h and then poured into water (500 mL). The resulting
precipitate was isolated by filtration, dissolved in DCM, washed with water and brine, dried (MgSO ),
4
and concentrated under reduced pressure. The residue (6.14 g) was dissolved in DCM (100 mL), cooled
to 0°C, and treated with mꢀCPBA (16.0 g, 71 mmol) in one portion. The reaction mixture was allowed to
stir at rt overnight, and then was extracted with sodium bicarbonate solution (saturated), sodium bisulfite
solution (10%), and brine. The organic layer was dried (MgSO4) and concentrated under reduced
1
pressure to afford 18a (5.6 g, 80%) as a pale yellow solid. H NMR (400 MHz, CDCl ) δ 1.02 (s, 6H),
3
1.97 (m, 2H), 2.36 (t, J = 7.5 Hz, 2H), 2.68 (t, J = 7.7 Hz, 2H), 3.76 (s, 4H), 7.18 (d, J = 7.9 Hz, 2H),
7.72 (d, J = 7.5 Hz, 2H).
5-Amino-2-(ethylsulfonyl)benzonitrile (19a). A solution of 18a (0.554 g, 2.31 mmol) in MeOH (60
mL) was hydrogenated (60 psi) over 10% palladium on carbon (99 mg) for 3 h at rt. The reaction
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 40
1
2
3
mixture was filtered and concentrated under reduced pressure to give 19a (464 mg, 96%) as an offꢀwhite
4
5
1
solid. H NMR (400 MHz, CDCl ) δ 1.30 (t, J = 7.25 Hz, 3H), 3.30 (q, J = 7.47 Hz, 2H), 4.62 (s, 2H),
3
6
7
8
6.89 (dd, J = 8.79, 2.64 Hz, 1H), 7.04 (d, J = 2.20 Hz, 1H), 7.83 (d, J = 8.79 Hz, 1H).
9
4-(4-Bromophenyl)-N-(3-cyano-4-(ethylsulfonyl)phenyl)butanamide (20a). Oxalyl chloride (0.175
mL, 2.0 mmol) was added slowly dropwise to a solution of 4ꢀ(4ꢀbromophenyl)butyric acid (243 mg, 1.0
mmol) in DCM (3 mL) and DMF (2 drops). The reaction was stirred for 1 h at rt and then concentrated
under reduced pressure. The residue was coevaporated with toluene and then chloroform to give the acid
chloride as a crude brown oil. A solution of the crude acid chloride (197 mg, 0.75 mmol) and 19a (106
mg, 0.50 mmol) in DCM (1 mL) was treated with triethylamine (0.140 mL, 1.0 mmol) and DMAP (10
mg, 0.08 mmol). The reaction mixture was stirred for 14 h at rt. DCE (1 mL) was added, and the
reaction mixture was heated to reflux for 30 h. The reaction was cooled to rt, diluted with DCM, and
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
washed with 1 N HCl, dried (MgSO ), and concentrated under reduced pressure. The residue was
4
purified by silica gel chromatography (EtOAc/hexanes) and then by reverse phase HPLC to give 20a (83
1
mg, 38%). MS (ESI) m/z 435.2, 437.2 (M+H)+. H NMR (400MHz, CDCl ) δ 1.32 (t, J = 7.5 Hz, 3H),
3
2.06 (quin, J = 7.4 Hz, 2H), 2.42 (t, J = 7.3 Hz, 2H), 2.68 (t, J = 7.5 Hz, 2H), 3.35 (q, J = 7.5 Hz, 2H),
7.08 (d, J = 7.9 Hz, 2H), 7.41 (d, J = 8.3 Hz, 2H), 7.67 (s, 1H), 7.88 (dd, J = 8.8, 1.8 Hz, 1H), 8.04 (d, J
= 8.8 Hz, 1H), 8.16 (d, J = 2.2 Hz, 1H).
4-(4-(3-Cyano-4-(ethylsulfonyl)phenylamino)-4-oxobutyl)phenylboronic acid (21a).
A
flask
containing 20a (83 mg, 0.19 mmol), bis(neopentyl glycolato)diboron (47.6 mg, 0.211 mmol), potassium
acetate (83 mg, 0.84 mmol), and dichloro[1,1’ꢀbis(diphenylphosphino)ferrocene] palladium(II)
dichloromethane adduct (4.4 mg, 0.0060 mmol) was purged with argon. DMSO (1 mL) was added, and
the reaction mixture was degassed with three cycles of vacuum followed by argon backfill. The reaction
mixture was heated for 2 h at 80 °C, cooled to rt, and diluted with water (100 mL). The aqueous solution
was extracted with diethyl ether (3x 25 mL), and the organic layers were dried (MgSO ), and
4
20
ACS Paragon Plus Environment

Page 21 of 40
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
concentrated under reduced pressure. The residue was dissolved in a mixture of diethyl ether (1 mL),
DCM (~0.1 mL), and EtOAc (~0.1 mL). Diethanolamine (22 mg, 0.21 mmol) in isopropanol (0.5 mL)
was added, and the reaction mixture was stirred overnight at rt. The reaction mixture was concentrated,
and the residue was purified by reverse phase HPLC (under the standard acidic conditions) to give 21a
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
(44 mg, 57%) as a clear oil. MS (ESI) m/z 425.4 (M+CH OHꢀH O+H)+. H NMR (400MHz,
3
2
tetrahydrofuranꢀd8) δ 1.12ꢀ1.26 (m, 3H), 2.01 (quin, J = 7.3 Hz, 2H), 2.27ꢀ2.37 (m, 2H), 2.69 (t, J = 7.5
Hz, 2H), 3.24ꢀ3.39 (m, 2H), 3.74 (s, 3H), 7.09ꢀ7.19 (m, 2H), 7.67 (d, J = 7.5 Hz, 1H), 7.86ꢀ8.06 (m,
2H), 8.18ꢀ8.39 (m, 1H), 9.57ꢀ9.70 (m, 1H).
2-(1-Di-tert-butoxycarbonylaminoisoquinolin-6-ylamino)-2-(4-(4-(3-cyano-4-
(ethylsulfonyl)phenylamino)-4-oxobutyl)phenyl)acetic acid (22a). Using a procedure analogous to
that which was used to prepare 10a, 21a (43.6 mg, 0.109 mmol) was reacted with 9 and glyoxylic acid
monohydrate to afford 22a (42.8 mg, 52%) as a yellow solid. MS (ESI) m/z 772.3 (M+H)+. H NMR
1
(400MHz, DMSOꢀd6) δ 1.14 (t, J = 7.3 Hz, 3H), 1.31 (s, 18H), 1.85ꢀ1.97 (m, 2H), 2.41 (t, J = 7.5 Hz,
2H), 2.63 (t, J = 7.5 Hz, 2H), 3.39 (q, J = 7.5 Hz, 2H), 5.26 (s, 1H), 6.72 (s, 1H), 7.25 (d, J = 7.9 Hz,
2H), 7.33ꢀ7.43 (m, 2H), 7.46ꢀ7.54 (m, 3H), 8.02 (s, 2H), 8.07 (d, J = 5.7 Hz, 1H), 8.33 (s, 1H).
(R)-2-(1-Amino-isoquinolin-6-ylamino)-7-ethanesulfonyl-4,11-diaza-
6,10
tricyclo[14.2.2.1 ]henicosa-1(19),6,8,10(21),16(20),17-hexaene-3,12-dione ((R)-23). A solution of
22a (17 mg, 0.022 mmol) in a mixture of methanol (5 mL) and 1M hydrochloric acid (0.050 mL) was
hydrogenated (60 psi) over 10% palladium on carbon (7 mg) for 17 h. The reaction mixture was filtered
and concentrated under reduced pressure. A solution of the residue, DIEA (0.010 mL, 0.057 mmol), and
BOP (5.3 mg, 0.012 mmol) in DMF (1.0 mL) was stirred at rt for 1 h. The reaction mixture was
concentrated in vacuo and purified by reverse phase HPLC. The residue was dissolved in 50%
TFA/DCM (1.5 mL) with 2 drops of water added and stirred for 1.25 h at rt. The solvent was evaporated
under a stream of nitrogen and the residue was purified by reverse phase HPLC and then by chiral
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 40
1
2
3
HPLC to give (S)-23 (peak 1, 2.6 mg, 21%) and (R)-23 (peak 2, 2.4 mg, 20%). The chromatography
4
5
conditions were the following: Chiralcel ODꢀH column (2.5 cm ID x 25 cm L, Chiral Technologies,
6
7
Inc.), 30% (1:1 ethanol/methanol)/70% heptane as eluent, 15 mL/min flow rate, and uv detection at 254
8
9
1
nm. (R)-23 analytical data: MS (ESI) m/z 558.3 (M+H)+. H NMR (400 MHz, CD OD) δ 1.24 (t, J =
3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
7.51 Hz, 3H), 2.02ꢀ2.16 (m, 1H), 2.27ꢀ2.48 (m, 3H), 2.55ꢀ2.69 (m, 1H), 2.87ꢀ3.00 (m, 1H), 3.32ꢀ3.47
(m, 2H), 4.21 (d, J = 16.84 Hz, 1H), 5.05 (d, J = 16.84 Hz, 1H), 5.11 (s, 1H), 6.58 (d, J = 2.20 Hz, 1H),
6.64 (d, J = 1.46 Hz, 1H), 6.71 (d, J = 6.22 Hz, 1H), 6.91 (dd, J = 8.42, 2.20 Hz, 1H), 7.01 (dd, J = 8.97,
2.38 Hz, 1H), 7.05ꢀ7.16 (m, 2H), 7.40 (d, J = 7.69 Hz, 1H), 7.48 (d, J = 5.86 Hz, 1H), 7.59 (dd, J = 7.87,
1.65 Hz, 1H), 7.77 (d, J = 8.42 Hz, 1H), 7.83 (d, J = 9.15 Hz, 1H). Chiral analytical HPLC retention
times: peak 1, 8.65 min; peak 2, 10.08 min using the following chromatography conditions: Chiralcel
OD column (4.6 mm ID x 250 mm L, Chiral Technologies, Inc.), 30% (1:1 ethanol/methanol)/70%
heptane as eluent, 1 mL/min flow rate, and uv detection at 254 nm.
tert-Butyl 5-amino-2-(ethylsulfonyl)benzylcarbamate (26). A solution of 18a (2.0 g, 8.32 mmol) in
MeOH (100 mL) and hydrochloric acid (1 N, 20 mL) was hydrogenated (60 psi) over 20% Pd(OH)
2
(380 mg) for three days. The reaction mixture was filtered and hydrogenated twice more for three days
each time over fresh catalyst. The reaction mixture was filtered and then concentrated in vacuo to give a
white solid (2.15 g) after trituration with ethyl acetate and ether. 1.0 g of the solid was dissolved in THF
(25 mL) and triethylamine (1 mL) and treated with 2ꢀ(tertꢀbutoxycarbonyloxyimino)ꢀ2ꢀ
phenylacetonitrile (0.905 g, 3.67 mmol). The reaction mixture was stirred overnight at rt. The reaction
mixture was concentrated in vacuo and the residue was extracted twice with DCM and saturated sodium
bicarbonate. The combined organics were extracted with brine, dried, and concentrated in vacuo. The
residue was purified by silica gel chromatography (gradient from 0 to 50% ethyl acetate in hexanes) to
give 26 (1.07 g, 88%) as a clear oil. MS (ESI) m/z 315.12 (M+H)+. 1H NMR (400MHz, CDCl ) δ 1.21ꢀ
3
22
ACS Paragon Plus Environment

Page 23 of 40
Journal of Medicinal Chemistry
1
2
3
4
1.31 (m, 3H), 3.11 (q, J = 7.5 Hz, 2H), 4.43 (d, J = 6.6 Hz, 2H), 5.55 (br. s., 1H), 6.62 (dd, J = 8.6, 2.4
Hz, 1H), 6.81 (s, 1H), 7.70 (d, J = 8.8 Hz, 1H).
5
6
7
8
9
tert-Butyl 5-(4-(4-bromophenyl)butanamido)-2-(ethylsulfonyl)benzylcarbamate (27). Using a
procedure analogous to that which was used to prepare 20a, except that pyridine was used in place of
triethylamine, 26 (0.314 g, 1.00 mmol) was coupled to 4ꢀ(4ꢀbromophenyl)butyric acid to give 27 (0.540
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
g, 100%) as a white foam. 1H NMR (400 MHz, CDCl ) δ ppm 1.23ꢀ1.31 (m, 3H) 1.41 (s, 9H) 1.97ꢀ2.08
3
(m, 2H) 2.37 (t, J = 7.47 Hz, 2H) 2.59ꢀ2.70 (m, 2H) 3.11ꢀ3.23 (m, 2H) 4.12 (q, J = 7.03 Hz, 1H) 4.50 (d,
J = 6.15 Hz, 2H) 5.60 (t, J = 5.93 Hz, 1H) 7.06 (d, J = 8.35 Hz, 2H) 7.39 (d, J = 8.35 Hz, 2H) 7.45ꢀ7.50
(m, 1H) 7.83ꢀ7.95 (m, 2H).
4-(4-(3-((tert-Butoxycarbonylamino)methyl)-4-(ethylsulfonyl)phenylamino)-4-
oxobutyl)phenylboronic acid (28). Using a procedure analogous to that which was used to prepare 21a,
27 (0.541 g, 1.0 mmol) was coupled to bis(neopentyl glycolato)diboron (0.249 g, 1.1 mmol) and then
hydrolyzed to the free boronic acid to give 28 (334 mg, 82%) as a peach colored foam. MS (ESI) m/z
505.03 (M+H)+. 1H NMR (400MHz, tetrahydrofuranꢀd8) δ 1.17 (t, J = 7.3 Hz, 3H), 1.41 (s, 9H), 2.00 (t,
J = 7.5 Hz, 2H), 2.27ꢀ2.35 (m, 2H), 2.68 (t, J = 7.3 Hz, 2H), 3.20 (q, J = 7.3 Hz, 2H), 4.53 (d, J = 6.2
Hz, 2H), 6.56 (br. s., 1H), 7.05 (s, 1H), 7.16 (d, J = 7.9 Hz, 2H), 7.63 (s, 1H), 7.70 (d, J = 7.9 Hz, 1H),
7.78ꢀ7.83 (m, 1H), 7.93 (d, J = 9.2 Hz, 1H).
2-(4-(4-(3-((tert-Butoxycarbonylamino)methyl)-4-(ethylsulfonyl)phenylamino)-4-
oxobutyl)phenyl)-2-(4-cyanophenylamino)acetic acid (29). Using a procedure analogous to that which
was used to prepare 10a, 28 (0.101 g, 0.20 mmol) was reacted with 4ꢀaminobenzonitrile (0.024 g, 0.2
mmol) and glyoxylic acid monohydrate (0.018 g, 0.2 mmol) to give, after purification by reverse phase
HPLC, 29 (78 mg, 62%) as a white solid. MS (ESI) m/z 634.9 (M+H)+. 1H NMR (400MHz, CD3OD) δ
1.11ꢀ1.27 (m, 3H), 1.44 (s, 9H), 1.92ꢀ2.04 (m, 2H), 2.40 (t, J = 7.3 Hz, 2H), 2.66 (t, J = 7.5 Hz, 2H),
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 40
1
2
3
4
3.20ꢀ3.31 (m, 2H), 4.57 (s, 2H), 6.66 (d, J = 9.2 Hz, 2H), 7.20 (d, J = 7.9 Hz, 2H), 7.35 (d, J = 9.2 Hz,
2H), 7.42 (d, J = 7.9 Hz, 2H), 7.75 (dd, J = 8.8, 2.2 Hz, 1H), 7.80ꢀ7.87 (m, 2H).
5
6
7
4-((R)-7-Ethanesulfonyl-3,12-dioxo-4,11-diaza-tricyclo[14.2.2.16,10]henicosa-
8
9
1(19),6,8,10(21),16(20),17-hexaen-2-ylamino)-benzonitrile ((R)-30). Hydrogen chloride (4N solution
in dioxane, 1 mL, 4 mmol) was added to a solution of 29 (78 mg, 0.12 mmol) in ethyl acetate (1 mL).
The reaction mixture was stirred at rt overnight and then concentrated to a yellow solid (65 mg). This
material was combined with additional intermediate (135 mg), dissolved in DMF (7 mL), and added
dropwise over 0.5 h to a solution of BOP (331 mg, 0.74 mmol) and DMAP (226 mg, 1.85 mmol) in
DCM (75 mL) at 40 °C. The reaction mixture was heated for an additional 0.5 h and then concentrated
in vacuo. The residual solid was purified by reverse phase HPLC to give 81 mg of white solid. This was
combined with 31 mg from a previous run, dissolved in MeOH, and purified by chiral HPLC to give (S)-
30 (peak 1, 48 mg, 21%) and (R)-30 (peak 2, 40 mg, 18%). The preparative chromatography conditions
were the following: Chiralcel OD column (5 cm ID x 50 cm L, 20 micron, Chiral Technologies, Inc.),
30% (1:1 ethanol/methanol)/70% heptane as eluent, 50 mL/min flow rate, and UV detection at 254 nm.
(R)-30 analytical data: MS (ESI) m/z 517.1 (M+H)+. 1H NMR (400MHz, CD3OD) δ 1.24 (t, J = 7.5 Hz,
3H), 1.87 (dt, J = 6.8, 3.2 Hz, 2H), 2.03ꢀ2.14 (m, 1H), 2.27ꢀ2.46 (m, 4H), 2.56ꢀ2.68 (m, 1H), 2.93 (td, J
= 7.0, 4.8 Hz, 1H), 3.39 (td, J = 14.7, 7.0 Hz, 2H), 3.68ꢀ3.76 (m, 2H), 4.15 (d, J = 17.1 Hz, 1H), 5.01ꢀ
5.09 (m, 2H), 6.62 (d, J = 1.8 Hz, 1H), 6.65ꢀ6.70 (m, 2H), 6.89 (dd, J = 8.3, 1.8 Hz, 1H), 6.98ꢀ7.03 (m,
1H), 7.07ꢀ7.13 (m, 1H), 7.34ꢀ7.40 (m, 3H), 7.54 (dd, J = 7.7, 2.0 Hz, 1H), 7.76 (d, J = 8.3 Hz, 1H).
Chiral analytical HPLC retention times: peak 1, 6.95 min; peak 2, 8.17 min using the following
chromatography conditions: Chiralcel OD column (4.6 mm ID x 250 mm L, 10 micron, Chiral
Technologies, Inc.), 30% (1:1 ethanol/methanol)/70% heptane as eluent, 1 mL/min flow rate, and UV
detection at 254 nm.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
24
ACS Paragon Plus Environment

Page 25 of 40
Journal of Medicinal Chemistry
1
2
3
6,10
4-((R)-7-Ethanesulfonyl-3,12-dioxo-4,11-diaza-tricyclo[14.2.2.1 ]henicosa-
4
5
6
7
8
9
1(19),6,8,10(21),16(20),17-hexaen-2-ylamino)-benzamidine trifluoroacetic acid salt ((R)-31). (R)-30
(30 mg, 0.058 mmol) was dissolved in a solution of hydroxylamine in DMSO (3M, 0.400 mL). The
reaction mixture was heated at 70 °C for 3 h. An additional aliquot of hydroxylamine in DMSO (3M,
0.300 mL) was added, and heating was continued for 2 h. The reaction mixture was diluted with ethyl
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
acetate, extracted with water and brine, dried (MgSO ), and concentrated in vacuo. Acetic anhydride
4
(0.030 mL, 0.32 mmol) was added to a solution of the residue dissolved in DCM (3 mL). After 30 min at
rt, the reaction mixture was concentrated in vacuo. The residue was dissolved in MeOH and
hydrogenated (50 psi) over 10% palladium on carbon (37 mg) for 2.5 h. The reaction mixture was
filtered and concentrated in vacuo. The residue was purified by preparative reverse phase HPLC to give
1
(R)-31 (7.3 mg, 19%). MS (ESI) m/z 534.0 (M+H)+. H NMR (400 MHz, CD OD) δ 1.25 (t, J = 7.25
3
Hz, 3H) 2.00ꢀ2.16 (m, 1H) 2.27ꢀ2.49 (m, 3H) 2.54ꢀ2.69 (m, 1H) 2.87ꢀ3.00 (m, 1H) 3.32ꢀ3.51 (m, 2H)
4.16 (dd, J = 16.70, 5.27 Hz, 1H) 5.00ꢀ5.14 (m, 2H) 6.63 (d, J = 1.76 Hz, 1H) 6.76 (d, J = 8.79 Hz, 2H)
6.90 (dd, J = 8.57, 1.98 Hz, 1H) 7.02 (dd, J = 7.91, 2.20 Hz, 1H) 7.11 (dd, J = 7.91, 1.32 Hz, 1H) 7.39
(dd, J = 7.91, 1.76 Hz, 1H) 7.52ꢀ7.62 (m, 3H) 7.76 (d, J = 8.35 Hz, 1H) 8.91 (t, J = 5.71 Hz, 1H).
4-(4-Bromo-2-methylphenyl)butanoic acid (33). 9ꢀBorabicyclo[3.3.1]nonane (10 mL, 5 mmol, 0.5N
in THF) was treated with methyl butꢀ3ꢀenoate 32 (0.5 g, 5 mmol) dropwise at rt under argon. The
reaction mixture was stirred for 3 h at rt. A resealable tube was charged with 5ꢀbromoꢀ2ꢀiodotoluene
(1.48 g,
5
mmol), sodium methoxide (853 mg, 15.8 mmol), and dichloro[1,1’ꢀ
bis(diphenylphosphino)ferrocene] palladium(II) dichloromethane adduct (110 mg, 0.15 mmol) in THF
(15 mL). After the hydroboration reaction was complete, it was added to the resealable tube, and the
combined reaction mixture was heated to 70 °C for 4 h, then at rt for several days. The reaction mixture
was diluted with water and extracted with EtOAc (2x). The combined organics were extracted with
water and brine, dried (MgSO ), and concentrated in vacuo. The residue was purified by silica gel
4
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 40
1
2
3
4
5
6
7
chromatography (EtOAc/hexane) to give 513 mg of ester. This material was dissolved in THF (3 mL),
MeOH (1.5 mL) and 1 M NaOH (3 mL) and heated for 1 h at 80 °C. Most of the solvent was removed in
vacuo, 1N HCl (4 mL) was added, and the mixture was extracted with EtOAc (3x). The combined
8
9
organic layers were washed with brine, dried (MgSO ), and concentrated in vacuo. The residue was
4
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
purified by reverse phase HPLC to give 33 (287 mg, 22%) as a white solid. 1H NMR (400 MHz, CDCl )
3
δ 1.82ꢀ1.96 (m, 2H), 2.28 (s, 3H), 2.42 (t, J = 7.25 Hz, 2H), 2.57ꢀ2.65 (m, 2H), 6.99 (d, J = 7.91 Hz,
1H), 7.23ꢀ7.26 (m, 1H), 7.27ꢀ7.31 (m, 1H).
4-(4-Bromo-2-methylphenyl)-N-(3-cyano-4-(ethylsulfonyl)phenyl)butanamide
(34).
Oxalyl
chloride (0.200 mL, 2.29 mmol) was added dropwise to a solution of 33 (287 mg, 1.12 mmol) in DCM
(4 mL) and DMF (1 drop). The reaction mixture was stirred for 4 h, and then concentrated in vacuo. The
residue was coevaporated with toluene and then dissolved in toluene (10 mL). 19a (227 mg, 1.08 mmol)
was added, and the reaction mixture was heated to reflux for 2 h. The reaction mixture was concentrated
in vacuo, and the residue was purified by silica gel chromatography (EtOAc/hexane) to give 34 (373 mg,
77%) as an offꢀwhite solid. MS (ESI) m/z 449.1, 451.1 (M+H)+. 1H NMR (400MHz, CDCl3) δ 1.28ꢀ1.35
(m, 3H), 1.95ꢀ2.06 (m, 2H), 2.30 (s, 3H), 2.46 (t, J = 7.3 Hz, 2H), 2.62ꢀ2.70 (m, 2H), 3.35 (q, J = 7.5
Hz, 2H), 7.00 (d, J = 8.3 Hz, 1H), 7.24 (dd, J = 8.1, 2.0 Hz, 1H), 7.28 (s, 1H), 7.78 (s, 1H), 7.90 (dd, J =
8.8, 2.2 Hz, 1H), 8.01ꢀ8.05 (m, 1H), 8.15 (d, J = 2.2 Hz, 1H).
4-(4-(3-Cyano-4-(ethylsulfonyl)phenylamino)-4-oxobutyl)-3-methylphenylboronic acid (35).
Using a procedure analogous to that which was used to prepare 7a, 34 (448 mg, 1.00 mmol) was reacted
with bis(neopentyl glycolato)diboron to afford the crude boronic acid. The crude boronic ester (212 mg,
0.44 mmol) was reacted with aqueous NaOH and then purified by silica gel chromatography
(MeOH/DCM) to afford 35 (113 mg, 55%). 1H NMR (400 MHz, THFꢀd ) δ 1.21 (t, 3H), 1.94ꢀ2.04 (m,
8
2H), 2.40 (t, J = 7.25 Hz, 2H), 2.70 (t, J = 7.47 Hz, 2H), 3.31 (q, J = 7.47 Hz, 2H), 7.01 (s, 2H), 7.08 (d,
26
ACS Paragon Plus Environment

Page 27 of 40
Journal of Medicinal Chemistry
1
2
3
4
J = 7.47 Hz, 1H), 7.53 (d, J = 7.47 Hz, 1H), 7.56 (s, 1H), 7.93 (dd, 1H), 7.99 (d, 1H), 1.00 (d, J = 2.20
Hz, 1H), 9.66 (s, 1H).
5
6
7
8
9
2-(1-Di-tert-butoxycarbonylaminoisoquinolin-6-ylamino)-2-(4-(4-(3-cyano-4-
(ethylsulfonyl)phenylamino)-4-oxobutyl)-3-methylphenyl)acetic acid (36). Using a procedure
analogous to that which was used to prepare 10a, 35 (113 mg, 0.273 mmol) was reacted with 9 and
glyoxylic acid monohydrate to afford 36 (154 mg, 72%) as an offꢀwhite solid. MS (ESI) m/z 786.3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
(M+H)+. H NMR (400MHz, DMSOꢀd6) δ 1.13 (t, J = 7.3 Hz, 3H), 1.30 (s, 18H), 1.74ꢀ1.90 (m, 2H),
2.28 (s, 3H), 2.45 (t, J = 7.0 Hz, 2H), 2.55ꢀ2.64 (m, 2H), 3.39 (q, J = 7.3 Hz, 2H), 5.19 (d, J = 7.5 Hz,
1H), 6.70 (s, 1H), 7.17 (d, J = 7.9 Hz, 2H), 7.27ꢀ7.36 (m, 3H), 7.40 (d, J = 5.7 Hz, 1H), 7.49 (d, J = 9.2
Hz, 1H), 8.03 (d, J = 0.9 Hz, 2H), 8.06 (d, J = 5.7 Hz, 1H), 8.32 (s, 1H), 10.68 (s, 1H).
(R)-2-(1-Amino-isoquinolin-6-ylamino)-7-ethanesulfonyl-20-methyl-4,11-diaza-
6,10
tricyclo[14.2.2.1 ]henicosa-1(19),6,8,10(21),16(20),17-hexaene-3,12-dione ((R)-37). Using
a
procedure analogous to that which was used to prepare 23, 36 (154 mg, 0.263 mmol) was hydrogenated,
cyclized, and deprotected to afford racemic 37 (22.6 mg, 12%) as an offꢀwhite amorphous solid. NMR
and analytical HPLC are consistent with a 1:1 mixture of atropisomers. MS (ESI) m/z 572.1 (M+H)+. 1H
NMR (400 MHz, CD OD) δ 1.19ꢀ1.27 (m, 6H), 1.86ꢀ1.98 (m, 2H), 2.12ꢀ2.21 (m, 1H), 2.28 (t, J = 8.97
3
Hz, 1H), 2.35ꢀ2.49 (m, 4H), 2.65ꢀ2.75 (m, 1H), 2.87ꢀ2.97 (m, 1H), 3.08ꢀ3.19 (m, 2H), 3.34ꢀ3.45 (m,
4H), 4.10ꢀ4.25 (m, 3H), 5.09 (dd, J = 17.03, 5.68 Hz, 1H), 5.13 (s, 1H), 5.14 (s, 1H), 6.64ꢀ6.71 (m, 4H),
6.80ꢀ6.94 (m, 5H), 6.97 (s, 1H), 7.05 (d, J = 7.69 Hz, 1H), 7.09ꢀ7.17 (m, 2H), 7.29 (d, J = 6.96 Hz, 2H),
7.33ꢀ7.38 (m, 1H), 7.41ꢀ7.46 (m, 1H), 7.47 (s, 1H), 7.76 (d, J = 8.42 Hz, 2H), 8.01 (d, J = 4.03 Hz, 1H),
8.02ꢀ8.05 (m, 1H), 8.89 (t, J = 5.86 Hz, 1H), 8.95 (t, J = 6.04 Hz, 1H).
A solution of racemic 37 (0.100 g, 0.146 mmol) in methanol was purified by chiral supercritical fluid
chromatography to give (S)-37 (peak 1, 28 mg, 34%) and (R)-37 (peak 2, 29 mg, 35%). The
chromatography conditions were the following: Chiralpak AS column (3.0 cm ID x 25 cm L, 10 micron,
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 40
1
2
3
Chiral Technologies, Inc.), 75% CO /25% methanol/0.1% diethylamine as eluent, 100 bar, 40 °C, 65
2
4
5
mL/min flow rate, and UV detection at 220 nm. (R)-37 analytical data: MS (ESI) m/z 572.10 (M+H)+.
6
7
8
1H NMR (400 MHz, CD OD) δ 1.21ꢀ1.33 (m, 3H) 1.87ꢀ2.00 (m, 1H) 2.26 (s, 1H) 2.37ꢀ2.54 (m, 5H)
3
9
2.66ꢀ2.76 (m, 1H) 2.89ꢀ3.00 (m, 1H) 3.08ꢀ3.18 (m, 1H) 3.33ꢀ3.51 (m, 3H) 4.23 (t, J = 17.36 Hz, 1H)
4.99ꢀ5.12 (m, 2H) 6.55ꢀ6.62 (m, 1H) 6.64ꢀ6.69 (m, 1H) 6.69ꢀ6.76 (m, 1H) 6.85ꢀ6.95 (m, 1H) 6.97ꢀ7.10
(m, 2H) 7.30ꢀ7.36 (m, 2H) 7.38ꢀ7.51 (m, 2H) 7.77 (d, J = 8.79 Hz, 1H) 7.81ꢀ7.88 (m, 1H), mixture of
two atropisomers. Chiral analytical HPLC retention times: peak 1, 20.00 min; peak 2, 24.99 min using
the following chromatography conditions: Whelkꢀ01 (R,R) column (4.6 mm ID x 250 mm L, 10
micron.), 40% (1:1 ethanol/methanol)/60% heptane/0.1% diethylamine as eluent, 2 mL/min flow rate,
and UV detection at 264 nm.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Supporting Information. Analytical data and selected experimental procedures for compounds 3ꢀ5,
15, 16, (R)-24, and (rac)-25, and crystallographic refinement statistics. This material is available free of
charge via the Internet at http://pubs.acs.org.
PDB ID Codes. 4ZXX and 4ZXY
Abbreviations Used. TF, tissue factor; FVIIa, coagulation factor VIIa; FIXa, coagulation factor IXa;
FXa, coagulation factor Xa; APC, activated protein C; tissue kall., tissue kallikreinꢀ1; AIQ,
aminoisoquinoline; BZD, benzamidine; EC2x, effective concentration that doubles clotting time.
Acknowledgement. The authors thank Peter Glunz, Dean Wacker, and Don Pinto for critical review
of this manuscript, We thank Steven Sheriff for helping to prepare the coordinates and data of
compounds 1 and (R)-4 and depositing them in the PDB, Yanfeng Zhang for preparation of compounds
15 and 16, Jeff Bozarth, Randi Brown, Tara Peterson, and Mojgan Abousleiman for obtaining the
enzyme inhibition data, and Frank Barbera and Yiming Wu for obtaining the clotting assay data.
References
28
ACS Paragon Plus Environment

Page 29 of 40
Journal of Medicinal Chemistry
1
2
3
1.
Mackman, N.; Tilley, R. E.; Key, N. S. Role of the extrinsic pathway of blood coagulation in
4
5
6
7
hemostasis and thrombosis. Arterioscler. Thromb. Vasc. Biol. 2007, 27, 1687ꢀ1693.
2. Banner, D. W.; D'Arcy, A.; Chene, C.; Winkler, F. K.; Guha, A.; Konigsberg, W. H.; Nemerson,
8
9
Y.; Kirchhofer, D. The crystal structure of the complex of blood coagulation factor VIIa with soluble
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
tissue factor. Nature 1996, 380, 41ꢀ46.
3.
without hemorrhagic risks by inhibiting tissue factor pathway. Haemostasis 1996, 26, Suppl. 1, 76ꢀ82.
4. Golino, P.; Ragni, M.; Cirillo, P.; D'Andrea, D.; Scognamiglio, A.; Ravera, A.; Buono, C.;
Harker, L. A.; Hanson, S. R.; Wilcox, J. N.; Kelly, A. B. Antithrombotic and antilesion benefits
Ezban, M.; Corcione, N.; Vigorito, F.; Condorelli, M.; Chiariello, M. Antithrombotic effects of
recombinant human, active siteꢀblocked factor VIIa in a rabbit model of recurrent arterial thrombosis.
Circ. Res. 1998, 82, 39ꢀ46.
5.
Suleymanov, O. D.; Szalony, J. A.; Salyers, A. K.; Lachance, R. M.; Parlow, J. J.; South, M. S.;
Wood, R. S.; Nicholson, N. S. Pharmacological interruption of acute thrombus formation with minimal
hemorrhagic complications by a small molecule tissue factor/factor VIIa inhibitor: Comparison to factor
Xa and thrombin inhibition in a nonhuman primate thrombosis model. J. Pharmacol. Exp. Ther. 2003,
306, 1115ꢀ1121.
6.
Szalony, J. A.; Suleymanov, O. D.; Salyers, A. K.; PanzerꢀKnodle, S. G.; Blom, J. D.; LaChance,
R. M.; Case, B. L.; Parlow, J. J.; South, M. S.; Wood, R. S.; Nicholson, N. S. Administration of a small
molecule tissue factor/factor VIIa inhibitor in a nonꢀhuman primate thrombosis model of venous
thrombosis: effects on thrombus formation and bleeding time. Thromb. Res. 2003, 112, 167ꢀ174.
7.
Salyers, A. K.; Szalony, J. A.; Suleymanov, O. D.; Parlow, J. J.; Wood, R. S.; South, M. S.;
Nicholson, N. S. Assessment of bleeding propensity in nonꢀhuman primates by combination of selective
tissue factor/VIIa inhibition and aspirin compared to warfarin and aspirin treatment. Pharmacology
2004, 70, 100ꢀ106.
29
ACS Paragon Plus Environment