Structure-Based Design of Macrocyclic Factor XIa Inhibitors: Discovery of the Macrocyclic Amide Linker

Article abstract of DOI:10.1021/acs.jmedchem.6b01460

A novel series of macrocyclic FXIa inhibitors was designed based on our lead acyclic phenyl imidazole chemotype. Our initial macrocycles, which were double-digit nanomolar FXIa inhibitors, were further optimized with assistance from utilization of structure-based drug design and ligand bound X-ray crystal structures. This effort resulted in the discovery of a macrocyclic amide linker which was found to form a key hydrogen bond with the carbonyl of Leu41 in the FXIa active site, resulting in potent FXIa inhibitors. The macrocyclic FXIa series, exemplified by compound 16, had a FXIa K_i = 0.16 nM with potent anticoagulant activity in an in vitro clotting assay (aPTT EC_1.5x = 0.27 μM) and excellent selectivity against the relevant blood coagulation enzymes.

Full text of DOI:10.1021/acs.jmedchem.6b01460

Subscriber access provided by UNIV OF CALIFORNIA SAN DIEGO LIBRARIES
Article
Structure-Based Design of Macrocyclic Factor XIa
Inhibitors: Discovery of the Macrocyclic Amide Linker.
James R. Corte, Tianan Fang, Honey Osuna, Donald J. P. Pinto, Karen A Rossi, Joseph E Myers,
Steven Sheriff, Zhen Lou, Joanna J Zheng, Timothy W. Harper, Jeffrery M. Bozarth, Yiming
Wu, Joseph M Luettgen, Dietmar A Seiffert, Carl P. Decicco, Ruth R. Wexler, and Mimi L Quan
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.6b01460 • Publication Date (Web): 13 Jan 2017
Downloaded from http://pubs.acs.org on January 15, 2017
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 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
StructureꢀBased Design of Macrocyclic Factor XIa
Inhibitors: Discovery of the Macrocyclic Amide
Linker.
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
∗
James R. Corte , Tianan Fang, Honey Osuna, Donald J. P. Pinto, Karen A. Rossi, Joseph E.
Myers Jr., Steven Sheriff, Zhen Lou, Joanna J. Zheng, Timothy W. Harper, Jeffrey M. Bozarth,
Yiming Wu, Joseph M. Luettgen, Dietmar A. Seiffert, Carl P. Decicco, Ruth R. Wexler and Mimi
L. Quan
BristolꢀMyers Squibb Company, Research and Development, P.O. Box 5400, Princeton, New
Jersey 08543, United States
KEYWORDS. Factor XIa; FXIa; Factor XIa Inhibitors; FXIa Inhibitors; Thrombosis; Activated
Partial Thromboplastin Time; aPTT; Macrocycle; Anticoagulant; Antithrombotic.
ABSTRACT. A novel series of macrocyclic FXIa inhibitors was designed based on our lead
acyclic phenyl imidazole chemotype. Our initial macrocycles, which were doubleꢀdigit
nanomolar FXIa inhibitors, were further optimized with assistance from utilization of structureꢀ
based drug design and ligand bound Xꢀray crystal structures. This effort resulted in the
discovery of a macrocyclic amide linker which was found to form a key hydrogen bond with the
carbonyl of Leu41 in the FXIa active site resulting in potent FXIa inhibitors. The macrocyclic
∗
Corresponding author. Tel.: +1ꢀ609ꢀ466ꢀ5030; fax: +1ꢀ609ꢀ818ꢀ3331; eꢀmail: james.corte@bms.com
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 55
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
FXIa series, exemplified by compound 16, had a FXIa Ki = 0.16 nM with potent anticoagulant
activity in an in vitro clotting assay (aPTT EC1.5x = 0.27 ꢀM) and excellent selectivity against the
relevant blood coagulation enzymes.
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
INTRODUCTION
Cardiovascular diseases remain the leading cause of death worldwide, accounting for
1
1
7.3 million deaths in 2013. Two thirds of these deaths were attributed to ischemic heart disease
1
and ischemic stroke which arises when an occlusive clot or thrombus forms in the circulation.
Antithrombotic therapies, such as anticoagulant or antiplatelet agents, have been developed to
prevent the formation of thrombi but they are also accompanied by an increase in the risk of
bleeding. Achieving an optimal balance between preventing an ischemic event and minimizing
the risk of bleeding is central to a successful antithrombotic therapy. Recently approved novel
oral anticoagulants (NOACs), such as direct Factor Xa inhibitors (rivaroxaban, apixaban, and
edoxaban) and direct thrombin inhibitor (dabigatran), are highly effective for the prevention and
treatment of cardiovascular events such as deep vein thrombosis (DVT), pulmonary embolism,
2
and stroke from nonꢀvalvular atrial fibrillation. These NOACs exhibited a favorable benefitꢀ
risk profile in the indications mentioned above, and in fact, displayed a significant reduction in
2
intracranial bleeding when compared to warfarin in the stroke prevention studies. However,
NOACs have not shown a favorable benefitꢀrisk profile in preventing venous thromboembolism
3
4
(
VTE) in medically ill patients or in secondary prevention of acute coronary syndrome (ACS)
5
where an increased risk of bleeding has been observed. Anticoagulants that are efficacious and
show a reduced bleeding risk have the potential to address the unmet medical needs of patients.
Trypsinꢀlike serine protease Factor XIa (FXIa), the activated form of the zymogen Factor
6
XI (FXI), is an important enzyme in the intrinsic pathway of the blood coagulation cascade.
ACS Paragon Plus Environment
2

Page 3 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
FXIa plays a critical role in the amplification of thrombin generation which leads to the growth
6
and maturation of thrombi. Humans with a genetic deficiency in FXI (Hemophilia C) display a
minor bleeding tendency which is in stark contrast to the abnormal, spontaneous and often times
severe bleeding that is seen in humans with a genetic deficiency in either Factor VIII
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
(Hemophilia A) or Factor IX (Hemophilia B). Epidemiological studies suggest that a FXI
deficiency may offer protection against pathologic thrombosis. Individuals with a severe FXI
8
deficiency have a lower risk for ischemic stroke and DVT. In addition, individuals with
9
elevated levels of FXI are at a higher risk of acute myocardial infarction and DVT. Since FXIa
appears to contribute to thrombosis risk, but has a limited role in hemostasis, inhibition of either
1
0
FXI or FXIa could provide an effective anticoagulant therapy with a safer bleeding profile.
1
1
12
Both neutralizing antibodies and antisense oligonucleotides (ASO) which specifically
target FXI have been shown be effective in a number of animal thrombosis models with minimal
effects on bleeding time. In a Phase II clinical trial in total knee arthroplasty, an antisense
oligonucleotide IONISꢀFXIRx (formerly ISISꢀFXIRx) was superior to enoxaparin for the
13
prevention of DVT and “appeared to be safe with respect to the risk of bleeding.” In addition to
targeting FXI, the discovery of active site and allosteric inhibitors of FXIa have also garnered
1
4ꢀ24
considerable attention.
The active site FXIa inhibitors, which are divided into irreversible
1
7
and reversible inhibitors, are much more potent than the current allosteric FXIa inhibitors.
1
4
Active site irreversible FXIa inhibitors, represented by the previously published βꢀlactam and
1
5
αꢀketothiazole peptidomimetics, were found to be efficacious in both venous and arterial
1
6
thrombosis animal models with little to no effect on provoked bleeding. We have previously
disclosed several potent, reversible FXIa inhibitors from a number of structurally diverse
1
8
19
20,21
chemotypes which include the tetrahydroquinoline, indole, phenyl imidazole,
phenyl
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 55
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
22
23
pyridine, and the phenylalanine derived diamide series. Importantly, several of these
inhibitors exhibited potent antithrombotic efficacy in rabbit thrombosis models with no
1
8b,21b
detectable effects on cuticle bleeding.
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
Figure 1. ImidazoleꢀBased FXIa Inhibitors.
Extensive SAR in the phenyl imidazole series led to the discovery of 1, a singleꢀdigit
nanomolar FXIa inhibitor (FXIa Ki = 5.8 nM) with good in vitro anticoagulant activity as
measured by an activated partial thromboplastin time (aPTT) clotting assay (EC1.5x = 5.3 ꢀM).
To address the need for additional improvements in both FXIa binding affinity and aPTT
potency, we sought to preorganize the bioactive conformation of 1 via the formation of a
2
1a
macrocyclic ring. Macrocyclic scaffolds have been recognized as an important structural class in
drug research and development demonstrating improved properties such as binding affinity,
selectivity, and physicochemical properties when compared to the corresponding acyclic
25
precursor. A recent analysis by Giordanetto and Kihlberg revealed that there are currently 68
marketed macrocyclic drugs and 35 macrocycles that are in various stages of clinical
2
6
development. Macrocyclization strategies were successful in discovering potent inhibitors for
27
28
other blood coagulation proteases such as thrombin and Factor VIIa. Herein we report on the
discovery of a novel series of potent and selective macrocyclic FXIa inhibitors.
DESIGN OF MACROCYCLIC FXIa INHIBITORS
ACS Paragon Plus Environment
4

Page 5 of 55
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
Figure 2. Xꢀray crystal structure of acyclic phenyl imidazole 1 in Factor XIa. The red spheres depict water molecules and the dotted lines depict
hydrogen bonds. The red arrows depict the potential for macrocyclization between the P1 prime and P2 prime groups. Ethylene diol is an artifact
of the flashꢀcooling procedure.
The Xꢀray crystal structure of phenyl imidazole 1 bound to FXIa (Figure 2) shows the
21b
ligand occuping the S1, S1 prime, and the S2 prime binding pockets. The pꢀchloroꢀ
phenyltetrazole moiety fills the S1 pocket. The benzyl moiety occupies the lipophilic S1 prime
pocket with an edgeꢀon hydrophobic interaction between the phenyl ring and the disulfide bridge
Cys58ꢀCys42. The methyl Nꢀphenyl carbamate occupies the S2 prime pocket with the
carbamate forming several hydrogen bonds either directly with the enzyme or through water
molecules. Owing to the close spatial proximity of the P1 prime benzyl and the P2 prime phenyl
groups, we envisioned that these two groups could be linked in a favorable geometry to form a
macrocyclic ring. We anticipated that preorganization of the bioactive conformation of 1, via
this macrocyclization strategy, would lead to desired improvements in FXIa potency. In
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 55
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
addition, the macrocyclic linker would allow further exploration of the region in the FXIa active
site between the S1 prime and S2 prime pockets.
P1 Prime
( )n
N N
N N
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
N
O
N
O
X
NHCO Me
N
N
2
NHCO Me
N
2
N
H
H
HN
1
HN
Cl P2 prime
Cl
Cl
Cl
A
X = CH , O
2
n = 0-3
Figure 3. Macrocyclization strategy is to link the P1 prime to the ortho position of the P2 prime phenyl to form a ring.
One of our macrocyclization strategies, represented by structure A (Figure 3), removes
the P1 prime phenyl and connects the P1 prime to the ortho position of the P2 prime phenyl via
an alkyl or ether linker of varying lengths. We envisioned that the macrocyclic alkyl/ether linker
would provide the required conformational flexibility to efficiently interact with the FXIa
binding pocket. Molecular modeling was employed to help guide our initial choice of ring size.
A series of medium size and macrocyclic ring structures, ranging in size from 11ꢀ to 14ꢀ
membered rings, were evaluated. Truncated macrocycles, in which the P1 amide was removed,
were subjected to conformational analysis using MacroModel (Schrodinger, LLC). The
truncated macrocycles were used because when the entire molecule was subjected to
conformational analysis, the majority of the conformations that resulted showed a hydrophobic
collapse between the P1 and P2 prime regions of the molecule and did not represent realistic
bioactive conformations. The conformers were overlayed on the crystal structure of acyclic
phenyl imidazole 1, superimposing the imidazole scaffolds and determining the ring sizes that
may best fit into the active site. The macrocycle rings were then spliced into the crystal structure
of 1 and minimized using Impact (Schrodinger, LLC) in the presence of the protein and some
crystallographic waters and allowing 5 Å around the protein to flex. After minimization, ring
ACS Paragon Plus Environment
6

Page 7 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
strain was evaluated using Jaguar (Schrodinger, LLC) DFT (B3YLP/6ꢀ31G**). The analysis
showed that the 11ꢀmembered ring was too strained and could not adopt a reasonable bioactive
conformation. The 14ꢀmembered ring was able to maintain conformations similar to the acyclic
compound, but the additional methylene groups in the linker did not increase interactions with
the protein. The 12ꢀ and 13ꢀmembered macrocyclic rings had conformations that were similar to
the acyclic crystallographic conformation, without any ring strain or unnecessary methylene
groups in the linker and therefore were recommended for synthesis (Please see supporting
information for a more detailed conformational analysis).
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
CHEMISTRY
Scheme 1. Synthesis of RingꢀClosing Metathesis Precursors for Alkylꢀ and EtherꢀLinked
a
Macrocycles.
a
Reagents and conditions: (a) KHCO , DMF, 0 °C; (b) NH OAc, xylene, reflux or toluene at 160 °C in microwave,
3
4
4
6ꢀ56% over two steps, then chiral preparatory supercritical fluid chromatography; (c) NaH, SEMꢀCl, THF, 0 °C to
rt, 40ꢀ51%; (d) pentꢀ4ꢀenyl boronic acid or hexꢀ5ꢀenyl boronic acid, Pd(dppf)Cl ⋅CH Cl , K CO , Ag O, THF,
2
2
2
2
3
2
reflux, 15ꢀ19%.
A variety of 12ꢀ and 13ꢀmembered macrocycles were prepared utilizing a ringꢀclosing
2
9
olefin metathesis (RCM) strategy. The preparation of RCM precursors for the alkylꢀ and etherꢀ
linked macrocycles is described in Scheme 1. Alkylation of the potassium salt of 1a with αꢀ
ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 55
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
bromoketone 2a gave the corresponding keto ester intermediate. Condensation of the keto ester
intermediate with NH OAc in xylene or toluene at elevated temperatures provided imidazole
4
3
0
3
a. The % enantiomeric excess (% ee) of 3a, which varied between 82 to 95% ee depending on
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 reaction, was improved to >99% ee by preparative supercritical fluid chromatography.
Imidazole 3a was deprotonated with NaH and then reacted with SEMꢀCl which yielded a
mixture of protected imidazole derivatives 4a and 5a. SuzukiꢀMiyaura coupling between 4a and
either pentꢀ4ꢀenyl boronic acid or hexꢀ5ꢀenyl boronic acid provided RCM precursors for the 12ꢀ
3
1
and 13ꢀmembered alkylꢀlinked macrocycles 4b and 4d, respectively. RCM precursors for the
2ꢀ and 13ꢀmembered etherꢀlinked macrocycles 4c and 4e were prepared following an analogous
1
sequence as described above for the preparation of 4a.
Scheme 2. Synthesis of RingꢀClosing Metathesis Precursors for Amideꢀ, Carbamate, and
a
AmineꢀLinked Macrocycles.
a
Reagents and conditions: (a) NH OH, CuI, Lꢀproline, K CO , DMSO, 85 °C, 59% or NaN , CuI, Lꢀproline,
4
2
3
3
NaOH, EtOH, 95 °C, 50%; (b) 4ꢀpentenoic acid or 3ꢀbutenoic acid, T P, Hunig’s base, EtOAc, ꢀ10 °C to rt, 75ꢀ87%;
3
(
c) prop‐2‐en‐1‐yl chloroformate, pyridine, DCM, 0 °C, 96%; (d) pentꢀ4ꢀenꢀ1ꢀamine, CuI, Lꢀproline, K CO ,
2
3
DMSO, 90 °C, 45%.
The synthesis of RCM precursors for the amideꢀ, carbamateꢀ, and amineꢀlinked
macrocycles is described in Scheme 2. Amination of bromide 4a with NH OH in the presence of
4
3
2
CuI and Lꢀproline afforded aniline 6a. Alternatively, amination of bromide 4a with NaN in the
3
3
3
presence of CuI and Lꢀproline, followed by in situ reduction of the azide, provided aniline 6b.
ACS Paragon Plus Environment
8

Page 9 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Aniline 6a was coupled with 4ꢀpentenoic acid or 3ꢀbutenoic acid using T P which gave amides
3
7a and 7b, respectively. Carbamate 7c was prepared by reacting aniline 6b with prop‐2‐en‐1‐yl
chloroformate. Bromide 4a was coupled with pentꢀ4ꢀenꢀ1ꢀamine using CuI which provided
substituted aniline 7d.
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
Scheme 3. Synthesis of 12ꢀ and 13ꢀMembered Alkylꢀ and EtherꢀLinked Macrocycles 2ꢀ12.
a
Reagents and conditions: (a) Second generation Grubbs II catalyst (40 mol%), pTsOH, DCM, reflux; (b)
preparatory HPLC to separate E/Zꢀalkene isomers: 52% for 8a:8b (3.6:1), 38% for 8c:8d (2.2:1), 63% for 8e:8f
1:1.7), 58% for 8g:8h (1:1.3); (c) 5M HCl (aq), MeOH, 75 °C, 75ꢀ94%; (d) 10% Pd/C, H , MeOH, 38ꢀ78% over
(
2
two steps; (e) 10a, Hunig’s base, DMF, 26ꢀ91%; (f) NCS, ACN, CHCl , with or without Hunig’s base, 65 °C, 31ꢀ
3
5
9%.
The synthesis of the 12ꢀ and 13ꢀmembered alkylꢀ and etherꢀlinked macrocycles 2ꢀ12 is
3
4
described in Scheme 3. Using a modified procedure described by Lovely, 4b was pretreated
with pꢀTsOH to form the imidazolium ion and then cyclized via ringꢀclosing metathesis using the
second generation Grubbs (II) catalyst in refluxing DCM to give the 12ꢀmembered alkylꢀlinked
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 55
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
macrocycle as a 3.6:1 mixture of Eꢀalkene 8a and Zꢀalkene 8b isomers. The E/Zꢀalkene isomers
were separated by reverse phase chromatography. The geometry of the alkene bond was
1
1
determined by Hꢀ Hꢀhomonuclear decoupling of the two sets of vicinal methylene protons
which revealed the J coupling constant between the double bond protons; the Eꢀalkene exhibiting
a J
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
coupling = ~15 Hz and the Zꢀalkene exhibiting a J coupling = ~10.5 Hz. Deprotection of both
Boc and SEM groups with aqueous 5 M HCl in MeOH at 75 °C afforded the unsaturated
macrocyclic amine 9a. Hydrogenation of the E/Zꢀalkene mixture 8a and 8b over palladium on
carbon, followed by deprotection as described above for 9a, yielded the saturated macrocyclic
amine 9b. The pꢀchlorophenyltetrazole cinnamide P1 was installed by coupling amines 9a and
9b with the activated carboxylic ester 10a which gave amides 2 and 3. Chlorination of 2 with
NCS at elevated temperature provided chloroꢀimidazole 4. In a similar fashion, RCM precursors
4cꢀ4e were converted to their corresponding 12ꢀ and 13ꢀmembered alkylꢀ and etherꢀlinked
macrocycles 5ꢀ12.
Scheme 4. Synthesis of 13ꢀMembered Amideꢀ, Carbamateꢀ, and AmineꢀLinked Macrocycles
a
13ꢀ20.
ACS Paragon Plus Environment
10

Page 11 of 55
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
Reagents and conditions: (a) Second generation Grubbs II catalyst (5 to 40 mol%), pTsOH, DCM, reflux; (b)
preparatory HPLC or normal phase chromatography to separate E/Zꢀalkene isomers: 76% for 11a:11b (1.4:1 E:Z),
74% for 11c:11d (4.7:1 E:Z), 89% for 11e:11f (1:4.2 E:Z); (c) 5M HCl (aq), MeOH, 75 °C; or 4M HCl in dioxane,
75 °C, 46ꢀ98%; (d) 10% Pd/C, H , MeOH or EtOAc; (e) 10a, Hunig’s base, DMF, 13ꢀ54%; (f) NCS, ACN, CHCl ,
2
3
6
5 °C, 41ꢀ44%.
Scheme 4 describes the synthesis of the 13ꢀmembered amideꢀ, carbamateꢀ, and amineꢀ
linked macrocycles 13ꢀ20 and Scheme 5 describes the synthesis of the 12ꢀmembered amideꢀ
linked macrocycles 22 and 23. These macrocycles were synthesized from RCM precursors 7aꢀ
7
d using an analogous sequence to that described for the preparation of macrocycles 2ꢀ4.
Scheme 5. Synthesis of 12ꢀMembered Saturated Amideꢀ and Unsaturated AmideꢀLinked
Macrocycles 22 and 23.
a
ACS Paragon Plus Environment
11

Journal of Medicinal Chemistry
Page 12 of 55
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
a
Reagents and conditions: (a) Second generation Grubbs II catalyst (40 mol%), pTsOH, DCM, reflux, 66% for 13a
(
2.2:1 E:Z); (b) 4M HCl in dioxane, 50 °C, 74ꢀ82%; (c) 10% Pd/C, H , MeOH, EtOAc; (d) 10a, Hunig’s base,
2
DMF, 27ꢀ37% over two steps.
RESULTS AND DISCUSSION
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
Our initial survey of macrocyclic FXIa inhibitors focused on 12ꢀ and 13ꢀmembered alkylꢀ
and etherꢀlinked macrocycles (Table 1). Even though we were particularly interested in the
saturated alkyl/ether linked macrocycles represented by structure A (Figure 3), we also assayed
the unsaturated linkers since they were available from the RCM chemistry. Macrocycle 2, the
1
2ꢀmembered ring possessing the Eꢀalkene, showed a significant loss in FXIa activity (FXIa Ki
3
5
= 636 nM) compared to acyclic phenyl imidazole 1 (FXIa Ki = 5.8 nM). Removing the
unsaturation from the macrocyclic linker gave the saturated alkyl linker 3 (FXIa Ki = 87 nM)
which improved FXIa activity by 7ꢀfold. We had previously discovered in the acyclic series that
chlorination of the imidazole scaffold would provide a 5 to 10ꢀfold increase in FXIa activity;
however, only a modest 2ꢀfold increase in FXIa activity was observed with chloroꢀimidazole 4
2
0
(
FXIa Ki = 42 nM). The 12ꢀmembered etherꢀlinked macrocycles 5 and 6 were much less potent
than their corresponding alkylꢀlinked derivatives 3 and 4. In contrast to the SAR trends in the
2ꢀmembered alkyl macrocycles, the 13ꢀmembered macrocycle 7, possessing the unsaturated
1
alkylꢀlinker, was more potent than the saturated alkyl macrocycle 8. In addition, the chloroꢀ
imidazole in the 13ꢀmembered alkyl linker 9 (FXIa Ki = 68 nM) resulted in an 8ꢀfold increase in
FXIa activity. The 13ꢀmembered etherꢀlinked macrocycles 11 and 12 were either slightly
improved or equipotent to their corresponding alkylꢀlinked derivatives 8 and 9. The most potent
macrocycles identified from our initial survey, analogs 4, 9, and 12, were approximately an order
of magnitude less potent than acyclic phenyl imidazole 1.
Table 1. 12ꢀ and 13ꢀMembered Alkylꢀ and EtherꢀLinked Macrocycles
ACS Paragon Plus Environment
12

Page 13 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
a
b
Entry
Alkene
--
n
--
1
1
1
1
1
2
2
2
2
2
2
Ring
Size
X
--
R
--
FXIa Ki
(nM)
aPTT EC1.5x
(ꢀM)
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
2
--
5.8
5.3
c
d
E-alkene
Saturated
Saturated
Saturated
Saturated
E-alkene
Saturated
Saturated
E-alkene
Saturated
Saturated
12
12
12
12
12
13
13
13
13
13
13
CH
CH
CH
O
2
2
2
H
H
Cl
H
Cl
H
H
Cl
H
H
Cl
636
NT
3
87
42
>40
27
4
c
d
5
420
NT
c
d
6
O
250
NT
7
CH
CH
CH
O
2
2
2
184
600
68
>40
d
8
NT
9
34
d
10
3,920
159
76
NT
d
1
1
1
2
O
NT
O
>40
a
o
Ki values were obtained from purified human enzyme at 37 C and were averaged from multiple determinations (n=2), as described in Ref. 18a.
aPTT (activated partial thromboplastin time) in vitro clotting assay was performed in human plasma. Ki values were determined at 25 °C. NT
= Not tested.
b
c
d
In order to better understand how these macrocycles bind in the active site of FXIa, an Xꢀ
ray crystal structure of the 13ꢀmembered macrocycle 9 bound to FXIa (1.6 Å resolution, Rꢀwork
was 0.160 and Rꢀfree was 0.176, Figure 4) was obtained. An overlay of macrocycle 9 with
acyclic phenyl imidazole 1 (Figure 5) revealed conformation changes in the macrocycle which
may explain the loss in FXIa activity. The benzylic carbon of the P2 prime phenyl, which is part
of the macrocyclic linker, is in close spatial proximity (3.2 Å) to the carbonyl of Leu41. In order
to minimize the steric clash between these two groups, the P2 prime phenyl rotates 28° which
ACS Paragon Plus Environment
13

Journal of Medicinal Chemistry
Page 14 of 55
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
increases the dihedral angle in the macrocycle compared to acyclic phenyl imidazole 1 (Figure
3
6
5B; dihedral angle of P2 prime phenyl was 53° for 9 and 25° for 1). The larger dihedral angle
of the P2 prime phenyl does not appear to affect the key hydrogen bonding interactions of 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
Figure 4. Xꢀray crystal structure of 13ꢀmembered macrocycle 9 bound to Factor XIa with omit mFoꢀDFc electron density contoured at 5 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 flashꢀcooling
procedure.
37
carbamate moiety but it pulls the macrocyclic linker out of the S1 prime pocket minimizing the
3
8
interaction with the lipophilic region. The Xꢀray crystal structure of 9 can explain the SAR
trends observed with the 12ꢀ and 13ꢀmembered etherꢀlinked macrocycles. The etherꢀlinked
macrocycles have an oxygen atom in close spatial proximity to the carbonyl of Leu41. For the
12ꢀmembered macrocycles 5 and 6, the electrostatic repulsion between the oxygen atom in the
macrocyclic linker and the carbonyl of Leu41 leads to a loss in FXIa activity. What is interesting
however, was the finding that the 13ꢀmembered etherꢀlinked macrocycles 11 and 12 did not
show a loss in FXIa activity compared to their alkylꢀlinked analogs 8 and 9. It is postulated that
ACS Paragon Plus Environment
14

Page 15 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
the additional methylene in the linker provides enough flexibility whereby the linker can adopt a
conformation which minimizes the electrostatic repulsion. However, when unsaturation was
present in the macrocyclic linker, as in 10, this conformational flexibility was lost and the
electrostatic repulsion could not be minimized which resulted in a weak FXIa inhibitor (FXIa Ki
= 3,920 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
Figure 5. (A) Overlay of Xꢀray crystal structures of acyclic phenyl imidazole 1 (yellow) and 13ꢀmembered macrocycle 9 (cyan) in Factor XIa.
(B) Overlay of Xꢀray crystal structures of acyclic phenyl imidazole 1 (yellow) and 13ꢀmembered macrocycle 9 (cyan) in Factor XIa looking into
the S2 prime pocket. The increased dihedral angle in macrocycle 9 pulled the macrocyclic linker out of the S1 prime pocket.
Owing to the close spatial proximity of the benzylic carbon of the macrocyclic linker in 9
to the carbonyl of Leu41, we envisioned that introduction of an amide NꢀH in the linker would
improve the FXIa activity via a hydrogen bonding interaction. Indeed, incorporation of the
amide moiety into the 13ꢀmembered macrocycle had a profound impact on both FXIa binding
affinity and potency in the aPTT clotting assay (Table 2). Specifically, the saturated amide
linker 13 (FXIa Ki = 3.2 nM; aPTT = 1.2 ꢀM) was a singleꢀdigit nanomolar FXIa inhibitor with
good activity in the aPTT clotting assay. Furthermore, chloroꢀimidazole 14 led to a 6 to 7ꢀfold
increase in FXIa activity (FXIa Ki = 0.47 nM) but with only a modest improvement in the aPTT
ACS Paragon Plus Environment
15

Journal of Medicinal Chemistry
Page 16 of 55
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
clotting activity; however in vitro human liver microsome (HLM) T1/2 was improved compared
to 13. Importantly, comparing macrocycle 14 with acyclic phenyl imidazole 1, the macrocyclic
analog improved FXIa
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
Table 2. 13ꢀMembered Amide, Carbamate, and AmineꢀLinked Macrocycles
a
b
c
Entry
Alkene
Saturated
Saturated
Z-alkene
E-alkene
E-alkene
Saturated
E-alkene
E-alkene
--
X
CH
CH
CH
CH
CH
O
Y
R
H
Cl
H
H
Cl
H
H
H
--
FXIa Ki
aPTT EC1.5x
HLM T1/2
(nM)
(ꢀM)
(min.)
1
1
3
4
2
2
2
2
2
C=O
C=O
C=O
C=O
C=O
C=O
C=O
3.2
1.2
34
0.47
5.2
0.92
12
86
e
15
NT
1
1
1
1
6
7
8
9
0.16
0.03
3.8
0.27
0.28
0.86
0.45
41
58
62
69
17
12
O
0.88
514
27
d
20
CH
2
CH
2
NT
2
1
--
--
21
a
o
Ki values were obtained from purified human enzyme at 37 C and were averaged from multiple determinations (n=2), as described in Ref. 18a.
b
d
c
aPTT (activated partial thromboplastin time) in vitro clotting assay was performed in human plasma. HLM = human liver microsome stability.
NT = Not tested
affinity by 10ꢀfold and aPTT potency by 5ꢀfold. Though the Zꢀalkene amide macrocycle 15 was
less potent than the saturated amide linker 13, the Eꢀalkene amide macrocycle 16 had a FXIa Ki
=
0.16 nM with potent anticoagulant activity in the in vitro clotting assay (aPTT EC1.5x = 0.27
ꢀM). The corresponding chloroꢀimidazole 17 further improved the FXIa affinity with a FXIa Ki
=
0.03 nM however the potency in the clotting assay was unchanged. Next, we examined other
ACS Paragon Plus Environment
16

Page 17 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
NH bond donors at the benzylic position of the macrocyclic linker. The saturated carbamate
linker 18 was equipotent to the saturated amide linker 13. Even though the unsaturated
carbamate linker 19 was 5 to 6ꢀfold less potent than the corresponding amide 16, it was still a
subꢀnanomolar FXIa inhibitor. The type of NH bond donor at the benzylic position was critical
for FXIa affinity, the amideꢀlinked macrocycle 16 was >1,000ꢀfold more potent than the amineꢀ
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
3
9
linked macrocycle 20. We went back to evaluate installation of the amide moiety in the acyclic
series to determine whether similar potency enhancements could be achieved. However, this
was not the case, as acyclic analog 21 resulted in a 160ꢀfold loss in FXIa activity compared to
macrocycle 16.
Figure 6. Xꢀray crystal structure of 13ꢀmembered unsaturated amideꢀlinked macrocycle 16 bound to Factor XIa with omit mFoꢀDFc electron
density contoured at 3.5 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 flashꢀcooling procedure.
An Xꢀray crystal structure of the 13ꢀmembered unsaturated amideꢀlinked macrocycle 16
bound to FXIa (2.1 Å resolution, Rꢀwork was 0.170 and Rꢀfree was 0.192 , Figure 6) was
obtained and it confirmed the key hydrogen bonding interaction between the amide NH in the
ACS Paragon Plus Environment
17

Journal of Medicinal Chemistry
Page 18 of 55
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
macrocyclic linker and the carbonyl of Leu41 (2.8 Å). The macrocyclic linker was critical in
orienting the amide NH in an ideal position to form a strong hydrogen bonding interaction. Even
though the dihedral angle of the P2 prime phenyl in 16 (∠ = 43°) is smaller than the alkylꢀlinked
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
macrocycle 9 (∠ = 53°), the macrocyclic linker is still pulled away from the S1 prime pocket
4
0
preventing van der Waals contact with the S1 prime disulfide bridge. As the result, the 20ꢀfold
increase in FXIa activity that was seen with incorporation of Eꢀalkene into the macrocyclic linker
16, compared to the saturated macrocyclic linker 13 (Figures S1 and S2 in supporting
information), was most likely due to conformation entropy.
Table 3. 12ꢀ and 13ꢀMembered Saturated and Unsaturated AmideꢀLinked Macrocycles
a
b
c
Entry
Alkene
Saturated
Saturated
E-alkene
E-alkene
n
0
1
0
1
Ring Size
FXIa Ki
aPTT EC1.5x
HLM T1/2
(nM)
(ꢀM)
(min.)
2
1
2
2
3
3
12
13
12
13
1.0
0.30
90
3.2
1.2
>40
0.27
34
d
264
0.16
NT
16
41
a
o
Ki values were obtained from purified human enzyme at 37 C and were averaged from multiple determinations (n=2), as described in Ref. 18a.
b
d
c
aPTT (activated partial thromboplastin time) in vitro clotting assay was performed in human plasma. HLM = human liver microsome stability.
NT = Not tested.
With the discovery of the macrocyclic amide linker in the 13ꢀmembered macrocycle we
shifted our focus to the 12ꢀmembered macrocycle. Both the saturated amide linker and the
unsaturated amide linker were incorporated in the 12ꢀmembered ring system (Table 3) and
compared to the 13ꢀmembered ring system. The 12ꢀmembered macrocycle 22 containing the
saturated amide linker resulted in a 3ꢀfold improvement in FXIa affinity and a 4ꢀfold
improvement in the aPTT clotting assay compared to the 13ꢀmembered macrocycle 13. The
ACS Paragon Plus Environment
18

Page 19 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
HLM stability for 22 (T = 90 min) was also improved compared to 13 (T = 34 min). Based
1
/2
1/2
on FXIa potency and HLM stability, the 12ꢀmembered ring system was preferred for the
saturated amide linker. In contrast, the 13ꢀmembered ring system was preferred for the
unsaturated amide linker as the 12ꢀmembered ring 23 resulted in a significant loss in potency.
Molecular modeling indicated that the loss in potency with the 12ꢀmembered macrocycle 23 was
due to ring strain which prevented the amide moiety from adopting the preferred orientation to
participate in the key hydrogen bond with the carbonyl of Leu41.
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
Pharmacokinetic Profiles. The pharmacokinetic profile for several macrocyclic compounds
41
were evaluated in dog, using an Nꢀinꢀone cassette dosing protocol (Table 4). The 12ꢀ
membered alkylꢀlinked macrocycle 3 exhibited a low clearance, a moderate halfꢀlife, and modest
oral bioavailability in dog (%F = 12). Introduction of the amide moiety into the 12ꢀmembered
macrocyclic linker 22 increased the clearance, shortened the halfꢀlife and significantly reduced
the oral exposure (3: AUC = 458 nM*h versus 22: AUC = 21 nM*h). The 13ꢀmembered etherꢀ
linked macrocycle 11 exhibited a moderate clearance and showed modest oral bioavailability
(%F = 17). Similar to what was observed for the 12ꢀmembered macrocycles, the 13ꢀmembered
amide linked macrocycles 13 and 16 showed high clearance, a short halfꢀlife, and were not orally
bioavailable. The amide moiety, which led to a significant improvement in terms of both FXIa
activity and potency in the aPTT clotting assay, was deleterious for oral bioavailability. The
poor oral bioavailability of the amideꢀlinked macrocycles was attributed to the low permeability
2
(
Cacoꢀ2 A to B < 15 nm/sec) arising from the high polar surface area (169 Å ).
Table 4. Pharmacokinetic Profile In Dog For Selected Macrocycles.
a
d
Entry
Cl
T
1/2
Vdss
(L/kg)
F
(%)
AUC
(nM*h)
Dose
iv/po
(mpk)
DLM
T
Dog Protein
Binding
(% bound)
(mL/h/kg)
(h)
3.2
1.0
1/2 (min)
b
e
3
2.6
15
0.7
0.8
12
6
458
0.21/0.42
0.20/0.23
>120
NT
c
22
21
>120
90.2
ACS Paragon Plus Environment
19

Journal of Medicinal Chemistry
Page 20 of 55
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
b
e
1
1
1
1
3
6
13
68
23
2.1
1.1
1.2
2.2
3.9
1.8
17
0
119
0.20/0.40
0.20/0.35
0.19/0.32
88
NT
b
0
>120
>120
93.4
91.5
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
c
1
2.5
a
b
Compounds were dosed in dog in an Nꢀinꢀ1 format. Vehicle for iv: 70% PEG 400; 20% water; 10% ethanol. Vehicle for po: 70% PEG 400;
0% TPGS; 10% ethanol. Vehicle for both iv and po: 70% PEG 400; 20% water; 10% ethanol. DLM = dog liver microsome stability. NT =
c
d
e
2
Not tested
Serine Protease Selectivity. The 13ꢀmembered unsaturated amideꢀlinked macrocycle 16 and the
1
2ꢀmembered saturated amideꢀlinked macrocycle 22 were tested against a panel of relevant
serine proteases including the blood coagulation proteases (Table 5). For comparison, acyclic
phenyl imidazole 1 is included. Similar to the acyclic phenyl imidazole 1, the macrocycles 16
and 22 showed >1,000ꢀfold selectivity against many of the relevant proteases, except for plasma
4
2
kallikrein (70 to 100ꢀfold).
Table 5. Human Serine Protease Selectivity Profile for 1, 16, and 22.
a
Human Enzyme Ki (nM)
1
16
22
Factor XIa
5.8
>13,400
>27,100
>13,300
>3,050
>11,500
>10,000
519
0.16
1,670
1.0
3,190
b
Factor VIIa
Factor IXa
Factor Xa
>27,100
>13,300
>3,050
>11,500
1,520
>27,100
>13,300
>3,050
>11,500
8,390
Factor XIIa
Thrombin
Trypsin
Plasma Kallikrein
Activated Protein C
Plasmin
16
70
>21,500
>15,200
>6,150
>15,100
>4,530
>21,500
>14,200
>6,150
5,050
>21,500
>15,200
>6,150
>15,100
3,940
TPA
Urokinase
Chymotrypsin
1,500
a
o
b
o
i i
K values in nM were obtained using human purified enzymes at 37 C. K values in nM were obtained using human purified enzymes at 25 C.
CONCLUSION
ACS Paragon Plus Environment
20

Page 21 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
A novel series of macrocyclic FXIa inhibitors was designed based on the Xꢀray crystal
structure of the acyclic phenyl imidazole 1. Our initial macrocycles, which were doubleꢀdigit
nanomolar FXIa inhibitors, were further optimized by structureꢀbased drug design. This effort
resulted in the discovery a macrocyclic amide linker which led to picomolar FXIa inhibitors with
excellent potency in the aPTT clotting assay and greater than a 1,000ꢀfold selectivity against the
relevant blood coagulation enzymes. As an example, the 13ꢀmembered unsaturated amideꢀ
linked macrocycle 16 is a 0.16 nM inhibitor of FXIa with potent anticoagulant activity in an in
vitro clotting assay (aPTT EC1.5x = 0.27 ꢀM). While the macrocyclic amide linker was critical
for FXIa potency, it also contributed to the poor pharmacokinetic profile. The optimization of
macrocyclic FXIa inhibitors to address the poor oral bioavailability will be reported in due
course.
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
EXPERIMENTAL SECTION
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. Normal phase
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 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 ꢁm C18, 3.0 x 150 mm); Gradient elution (0.5 mL/min) from 10ꢀ100%
Solvent B for 12 min and then 100% Solvent B for 3 min. Solvent A is (95% water, 5%
ACS Paragon Plus Environment
21

Journal of Medicinal Chemistry
Page 22 of 55
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
acetonitrile, 0.05% TFA) and Solvent B is (5% water, 95% acetonitrile, 0.05% TFA);
monitoring UV absorbance at 220 and 254 nm.
Methyl N‐[(11E,14S)‐14‐[(2E)‐3‐[5‐chloro‐2‐(1H‐1,2,3,4‐tetrazol‐1‐yl)phenyl]prop‐ 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
2,7
enamido]‐16,18‐diazatricyclo[13.2.1.0 ]octadeca‐1(17),2,4,6,11,15(18)‐hexaen‐5‐yl]-
carbamate, trifluoroacetic acid salt (2). A cloudy suspension of 9a (0.014 g, 0.043 mmol) and
2
2b
(
E)ꢀ2,5ꢀdioxopyrrolidinꢀ1ꢀyl 3ꢀ(5ꢀchloroꢀ2ꢀ(1Hꢀtetrazolꢀ1ꢀyl)phenyl)acrylate 10a (0.015 g,
0.043 mmol) in DMF (0.43 mL) and Hunig's base (0.037 mL, 0.214 mmol) was stirred at rt.
After 1.5 h, the reaction was partitioned between water and EtOAc and then the layers were
separated. The aqueous layer was extracted with EtOAc (1x). The organic layers were
combined and washed with sat. NaHCO , brine, dried over Na SO , filtered and concentrated to
3
2
4
give a pale, yellow foam weighing 0.032 g. Purification by reverse phase chromatography gave
after concentration and lyophilization 2 (0.0131 g, 45% over two steps) as a white solid. MS
+
+
1
(ESI) m/z: 559.1 (M+H) and 561.1 (M+2+H) . H NMR (500 MHz, CD OD, 50 °C) δ ppm
3
9.47 (s, 1H), 7.94ꢀ7.98 (m, 1H), 7.65ꢀ7.71 (m, 1H), 7.58 (d, J=8.25 Hz, 1H), 7.46 (br. s, 1H),
7.37ꢀ7.41 (m, 1H), 7.29ꢀ7.37 (m, 2H), 7.19 (d, J=15.4 Hz, 1H), 6.76 (d, J=15.4 Hz, 1H), 5.50ꢀ
5.60 (m, 1H), 5.09ꢀ5.19 (m, 1H), 5.02ꢀ5.09 (m, 1H), 3.75 (s, 3H), 2.77ꢀ2.86 (m, 1H), 2.40ꢀ2.60
(m, 3H), 1.95ꢀ2.05 (m, 1H), 1.85ꢀ1.95 (m, 1H), 1.49ꢀ1.62 (m, 1H), 1.25ꢀ1.38 (m, 1H). Note:
proton decoupling confirmed the Eꢀalkene geometry with a J coupling = 15.4 Hz.
Methyl N‐[(14S)‐14‐[(2E)‐3‐[5‐chloro‐2‐(1H‐1,2,3,4‐tetrazol‐1‐yl)phenyl]prop‐2‐ enamido]‐
2
,7
1
6,18‐diazatricyclo[13.2.1.0 ]octadeca‐1(17),2,4,6,15(18)‐pentaen‐5‐yl]-carbamate, 1 tri-
fluoroacetic acid salt (3). Using a procedure analogous to that which was used to prepare 2, 9b
(0.058 g, 0.10 mmol) was coupled to 10a to give 3 (0.044 g, 62%) as a white solid. MS (ESI)
ACS Paragon Plus Environment
22

Page 23 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
+
1
m/z: 561.1 (M+H) . H NMR (500 MHz, CD OD, 50 °C) δ ppm 9.47 (s, 1 H), 7.95 (d, J=2.2 Hz,
3
1
H), 7.67 (dd, J=8.8, 2.2 Hz, 1H), 7.57 (d, J=8.2 Hz, 1H), 7.46 (d, J=2.2 Hz, 1H), 7.41 ꢀ 7.44
(
m, 2 H), 7.37 (d, J=8.3 Hz, 1 H), 7.17 (d, J=15.9 Hz, 1 H), 6.77 (d, J=15.9 Hz, 1 H), 5.01 (dd,
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
J=9.1, 5.8 Hz, 1 H), 3.75 (s, 3 H), 2.47 ꢀ 2.56 (m, 1 H), 2.36 ꢀ 2.44 (m, 1 H), 2.17 ꢀ 2.25 (m, 1
H), 1.82 ꢀ 1.91 (m, 1 H), 1.31 ꢀ 1.56 (m, 4 H), 1.23ꢀ1.26 (m, 2 H), 0.87 ꢀ 0.98 (m, 1 H), 0.38ꢀ0.53
(m, 1 H).
tert‐Butyl N‐[(1S)‐1‐(4‐{2‐bromo‐4‐[(methoxycarbonyl)amino]phenyl}‐1H‐imidazol‐ 2‐
yl)but‐3‐en‐1‐yl]carbamate (3a). To a clear, colorless solution of 1a (3.89 g, 18.1 mmol) in
DMF (45.2 mL) was added KHCO (2.17 g, 21.7 mmol). The reaction was stirred at rt for 20
3
min and then the reaction was cooled to 0 °C. Next a clear, yellow solution of 2a (6.34 g, 18.1
mmol) in DMF (20 mL) was added dropwise. The reaction was stirred at 0 °C for 30 min and
then the reaction was warmed to rt and stirred overnight. The reaction was cooled to 0 °C,
poured into iceꢀcold water (200 mL) to give a white suspension and then extracted with EtOAc
(3x). The organic layers were combined, washed with water (1x), brine (1x), dried over Na SO ,
2
4
filtered and concentrated to give 2‐{2‐bromo‐4‐[(methoxycarbonyl)ꢀamino]phenyl}‐2‐oxoethyl
(2S)‐2‐{[(tert‐butoxy)carbonyl]amino}pent‐4‐enoate (10.26 g) as a clear, yellow oil which was
+
used in the next reaction without further purification. MS (ESI) m/z: 387.0 (MꢀC H O +2+H) .
5
8
2
To a clear, yellow solution of 2‐{2‐bromo‐4‐[(methoxycarbonyl)ꢀamino]phenyl}‐2‐
oxoethyl (2S)‐2‐{[(tert‐butoxy)carbonyl]amino}pent‐4‐enoate (8.77 g, 18.07 mmol) in xylene
(
181 mL) was added NH OAc (13.93 g, 181 mmol). The flask was equipped with a Deanꢀstark
4
o
trap and a reflux condenser. The reaction was warmed to 140 C. After 3 h, the reaction was
cooled to rt, diluted with EtOAc (400 mL), washed with 1.5 M K PO (2 x 150 mL), brine, dried
3
4
over Na SO , filtered, and concentrated to give a brown foam. Purification by normal phase
2
4
ACS Paragon Plus Environment
23

Journal of Medicinal Chemistry
Page 24 of 55
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
chromatography (gradient elution 0ꢀ60% EtOAc/Hex) gave 3a (4.61 g, 55%) as a pale, orange
solid. The % enantiomeric excess (% ee) of 3a varied (82 to 95% ee) between reactions. The %
ee was further improved to >99% ee by preparative supercritical fluid chromatography (SFC)
using the following conditions: Chiralcel ODꢀH (3 x 25 cm, Chiral Technologies, Inc.), 25%
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
MeOH in CO , 110 mL/min, 35 °C, and UV detection at 240 nm. The analytical SFC conditions
2
were the following: Chiralcel ODꢀH (0.46 x 25 cm, Chiral Technologies, Inc.), 25% MeOH in
CO , 3 mL/min, 35 °C, and UV detection at 215 nm, (R)ꢀ3a {[α] = +36.9 (c = 1.18, MeOH)}
2
D
was peak 1 at 3.0 min. and (S)ꢀ3a {[α] = ꢀ38.8 (c = 1.28, MeOH)} was peak 2 at 5.4 min. An
D
analytical sample was obtained by reverse phase chromatography which gave 3a⋅1TFA for
1
+
characterization. MS (ESI) m/z: 467.0 (M+2+H) . H NMR (500 MHz, CD OD) δ 7.99 (d,
3
J=1.65 Hz, 1H), 7.60 (s, 1H), 7.56 (dd, J=1.92, 8.52 Hz, 1H), 7.44 (d, J=8.25 Hz, 1H), 5.81 (tdd,
J=7.15, 10.04, 16.91 Hz, 1H), 5.12ꢀ5.20 (m, 2H), 4.91ꢀ4.96 (m, 1H), 3.77 (s, 3H), 2.64ꢀ2.78 (m,
1
3
2
1
2
H), 1.44 (s, 9H). C NMR (125 MHz, CD OD) δ 157.5, 156.1, 149.9, 143.5, 133.1, 133.0,
3
32.9, 124.2, 123.7, 122.9, 120.6, 118.8, 118.7, 81.6, 53.0, 49.6 (overlaps with CD OD), 38.5,
3
8.7.
Methyl N‐[(14S)‐17‐chloro‐14‐[(2E)‐3‐[5‐chloro‐2‐(1H‐1,2,3,4‐tetrazol‐1‐ yl)phenyl]prop‐2‐
2
,7
enamido]‐16,18‐diazatricyclo[13.2.1.0 ]octadeca‐1(17),2,4,6,15(18)‐pentaen‐5‐yl]-
carbamate, trifluoroacetic acid salt (4). To a vial containing a solution of 3 (0.012 g, 0.018
mmol) in ACN (0.30 mL)/CHCl (0.30 mL) was added NCS (2.85 mg, 0.021 mmol). The vial
3
was sealed with a teflonꢀcoated screw cap and the reaction was warmed to 65 °C. After 4 h,
additional NCS (2.85 mg, 0.021 mmol) was added. After another 1 h, the reaction was cooled to
rt and then the reaction was concentrated. Purification by reverse phase chromatography
ACS Paragon Plus Environment
24

Page 25 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
+
1
afforded 4 (0.0040 g, 31%) as a yellow solid. MS (ESI) m/z: 595.1 (M+H) . H NMR (500
MHz, CD OD, 50 °C) δ ppm 9.46 (s, 1 H), 7.97 (d, J=2.2 Hz, 1 H), 7.65 (dd, J=8.8, 2.2 Hz, 1
3
H), 7.55 (d, J=8.3 Hz, 1 H), 7.40 (d, J=2.2 Hz, 1 H), 7.36 (dd, J=8.3, 1.6 Hz, 1 H), 7.27 (d, J=8.2
Hz, 1 H), 7.13 (d, J=15.4 Hz, 1 H), 6.77 (d, J=15.9 Hz, 1 H), 4.85 (dd, J=10.4, 5.5 Hz, 1 H), 3.75
(s, 3 H), 2.43 ꢀ 2.52 (m, 1 H), 2.31 ꢀ 2.40 (m, 1 H), 2.05 ꢀ 2.15 (m, 1 H), 1.57 ꢀ 1.68 (m, 1 H),
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
1.10 ꢀ 1.53 (m, 6 H), 0.95 ꢀ 1.07 (m, 1 H), 0.26ꢀ0.38 (m, 1 H).
tert‐Butyl N‐[(1S)‐1‐(4‐{2‐bromo‐4‐[(methoxycarbonyl)amino]phenyl}‐1‐{[2‐ (trimethyl-
silyl)ethoxy]methyl}‐1H‐imidazol‐2‐yl)but‐3‐en‐1‐yl]carbamate (4a). To a cooled (0 °C)
suspension of NaH (0.343 g, 8.58 mmol) in THF (15 mL) was added dropwise a clear, yellow
solution of 3a (3.63 g, 7.80 mmol) in THF (15 mL). After 30 min, the reaction was allowed to
warm to rt. After 2.5 h the reaction was a thick, viscous mixture so DMF (6 mL) was added to
facilitate stirring. The resulting clear, dark yellow solution was cooled to 0 °C and then SEMꢀCl
(
1.38 mL, 7.80 mmol) was added dropwise. After 30 min, additional SEMꢀCl (0.20 equiv.) was
added. After another 30 min the reaction was quenched with sat. NH Cl and the reaction was
4
allowed to warm to rt. The reaction was extracted with EtOAc (2x). The organic layers were
combined and washed with sat. NaHCO , brine, dried over Na SO , filtered, and concentrated to
3
2
4
give a clear, thick viscous orange oil. Purification by normal phase chromatography (gradient 0ꢀ
5
0% EtOAc/Hex) gave 4a (2.35 g, 51%) as a white solid and 5a (0.757 g) as a clear, colorless
1
+
+
oil. For 4a: MS (ESI) m/z: 595.1 (M+H) and 597.2 (M+2+H) . H NMR (500 MHz, CD OD)
3
δ 7.87 (s, 1H), 7.66 (d, J=8.25 Hz, 1H), 7.55 (s, 1H), 7.41 (dd, J=2.20, 8.80 Hz, 1H), 5.75ꢀ5.86
(
m, 1H), 5.62 (d, J=11.00 Hz, 1H), 5.33 (d, J=11.00 Hz, 1H), 5.14 (d, J=17.05 Hz, 1H), 5.05 (d,
J=9.90 Hz, 1H), 4.96 (t, J=7.15 Hz, 1H), 3.75 (s, 3H), 3.60 (t, J=7.97 Hz, 2H), 2.60ꢀ2.74 (m,
ACS Paragon Plus Environment
25

Journal of Medicinal Chemistry
Page 26 of 55
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
13
2
H), 1.43 (s, 9H), 0.87ꢀ1.02 (m, 2H), 0.00 (s, 9H). C NMR (125 MHz, CD OD) δ 157.6,
3
156.3, 149.8, 140.6, 138.8, 135.4, 132.0, 130.3, 123.9, 122.8, 121.0, 118.7, 118.6, 80.7, 76.4,
+
6
7.4, 52.8, 47.8, 40.0, 28.9, 18.9, ꢀ1.2. For 5a: MS (ESI) m/z: 725.3 (M+H) and 727.2
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
1
+
(M+2+H) . H NMR (500 MHz, CD OD) δ 7.81 (d, J=8.25 Hz, 1H), 7.68 (s, 1H), 7.64 (d,
3
J=2.20 Hz, 1H), 7.35 (dd, J=2.20, 8.25 Hz, 1H), 5.75ꢀ5.86 (m, 1H), 5.63 (d, J=11.00 Hz, 1H),
5
.35 (d, J=11.00 Hz, 1H), 5.14 (d, J=17.05 Hz, 1H), 5.02ꢀ5.08 (m, 3H), 4.97 (t, J=7.42 Hz, 1H),
3.74 (s, 3H), 3.64ꢀ3.69 (m, 2H), 3.57ꢀ3.63 (m, 2H), 2.60ꢀ2.77 (m, 2H), 1.42 (s, 9H), 0.88ꢀ1.02
m, 4H), 0.02 (s, 9H), 0.00 (s, 9H).
(
tert‐Butyl N‐[(1S)‐1‐(4‐{4‐[(methoxycarbonyl)amino]‐2‐(pent‐4‐en‐1‐yl)phenyl}‐1‐{[2‐
trimethylsilyl)ethoxy]methyl}‐1H‐imidazol‐2‐yl)but‐3‐en‐1‐yl]carbamate (4b). To a flameꢀ
dried, thickꢀwalled vial was placed 4a (0.366 g, 0.62 mmol), pentꢀ4ꢀenylboronic acid (0.154 g,
.35 mmol), Ag O (0.712 g, 3.07 mmol), K CO (0.510 g, 3.69 mmol), and PdCl (dppf)⋅CH Cl
(
1
2
2
3
2
2
2
adduct (0.050 g, 0.061 mmol). The vial was purged with argon for several minutes and then
degassed THF (2.6 mL) was added. The vial was sealed with a teflonꢀcoated screw cap and the
black suspension was warmed to 80 °C. After 8 h the reaction was cooled to rt. The reaction
mixture was filtered and the filtrate was concentrated to give a brown foam. Purification by
normal phase chromatography (gradient elution 0ꢀ30% EtOAc/Hex) gave impure material. The
impure material was purified by reverse phase chromatography. The pure fractions were
combined, neutralized with sat. NaHCO and then concentrated to give a white solid. The white
3
,
solid was partitioned between water and EtOAc and the layers were separated. The aqueous
layer was extracted with EtOAc (1x). The organic layers were combined and washed with brine,
dried over Na SO , filtered and concentrated to afford the 4b (0.054 g, 15%) as a white foam.
2
4
ACS Paragon Plus Environment
26

Page 27 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
+
MS (ESI) m/z: 585.3 (M+H) . H NMR (400 MHz, CD OD) δ 7.27ꢀ7.36 (m, 3H), 7.11 (s, 1H),
3
5.67ꢀ5.84 (m, 2H), 5.64 (d, J=10.99 Hz, 1H), 5.31 (d, J=10.99 Hz, 1H), 5.12 (d, J=16.26 Hz,
1
H), 5.03 (d, J=9.67 Hz, 1H), 4.85ꢀ4.98 (m, 3H), 3.73 (s, 3H), 3.58 (t, J=8.13 Hz, 2H), 2.55ꢀ2.84
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
(
m, 4H), 1.94ꢀ2.05 (m, 2H), 1.49ꢀ1.62 (m, 2H), 1.42 (s, 9H), 0.84ꢀ1.03 (m, 2H), 0.00 (s, 9H).
tert‐Butyl N‐[(1S)‐1‐{4‐[2‐bromo‐4‐(9,9‐dimethyl‐3‐oxo‐2,6‐dioxa‐4‐aza‐9‐siladecan‐ 4‐
yl)phenyl]‐1‐{[2‐(trimethylsilyl)ethoxy]methyl}‐1H‐imidazol‐2‐yl}but‐3‐en‐1‐ yl]carbamate
(5a). Compound 5a was prepared as a byꢀproduct in the preparation of compound 4a. Please
see experimental procedure for compound 4a for preparation and characterization of 5a.
tert‐Butyl N‐[(1S)‐1‐(4‐{2‐amino‐4‐[(methoxycarbonyl)amino]phenyl}‐1‐{[2‐ (trimethyl-
silyl)ethoxy]methyl}‐1H‐imidazol‐2‐yl)but‐3‐en‐1‐yl]carbamate (6a). A thickꢀwalled vial
containing 4a (2.0 g, 3.36 mmol), CuI (0.128 g, 0.67 mmol), Lꢀproline (0.155 g, 1.34 mmol) and
K CO (1.39 g, 10.1 mmol) in DMSO (6.7 mL) was vacuumed and backꢀfilled with argon three
2
3
times. Then 28% aq. NH OH (0.61 mL, 4.37 mmol) was added. The vial was sealed with a
4
teflonꢀcoated screw cap and the reaction was warmed to 85 °C. After 20 h, the reaction was
cooled to rt, diluted with EtOAc, washed with water, brine, dried over Na SO , filtered and
2
4
concentrated. Purification by normal phase chromatography (gradient elution 0ꢀ40%
1
+
EtOAc/Hex) afforded 6a (1.05 g, 59%) as a yellow solid. MS (ESI) m/z: 532.5 (M+H) . H
NMR (500 MHz, CD OD) δ 7.29 (s, 1H), 7.27 (d, J=8.53 Hz, 1H), 6.95ꢀ6.98 (m, 1H), 6.73 (dd,
3
J=2.06, 8.39 Hz, 1H), 5.75ꢀ5.86 (m, 1H), 5.57 (d, J=11.00 Hz, 1H), 5.30 (d, J=11.00 Hz, 1H),
5
.13 (d, J=16.23 Hz, 1H), 5.04 (d, J=10.18 Hz, 1H), 4.92ꢀ4.97 (m, 1H), 3.72 (s, 3H), 3.58 (t,
J=8.12 Hz, 2H), 2.68ꢀ2.76 (m, 1H), 2.60ꢀ2.68 (m, 1H), 1.42 (s, 9H), 0.87ꢀ0.99 (m, 2H), 0.00 (s,
ACS Paragon Plus Environment
27

Journal of Medicinal Chemistry
Page 28 of 55
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
13
9
H). C NMR (100 MHz, CDCl ) δ 155.1, 153.9, 146.9, 145.6, 140.5, 137.7, 133.6, 127.7,
3
118.3, 115.1, 113.0, 108.0,106.0, 79.7, 74.9, 66.3, 52.2, 46.1, 39.4, 28.4, 17.8, ꢀ1.4.
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
tert‐Butyl N‐[(1S)‐1‐(4‐{4‐[(methoxycarbonyl)amino]‐2‐(pent‐4‐enamido)phenyl}‐ 1‐{[2‐
(
trimethylsilyl)ethoxy]methyl}‐1H‐imidazol‐2‐yl)but‐3‐en‐1‐yl]carbamate (7a). To a cooled
(
ꢀ10 °C) solution of 6a (1.34 g, 2.52 mmol), pentꢀ4ꢀenoic acid (0.26 mL, 2.52 mmol) and Hunig's
base (1.32 mL, 7.56 mmol) in EtOAc (72 mL) was added a solution of 50% 1ꢀ
propanephosphonic acid cyclic anhydride in EtOAc (2.97 mL, 5.04 mmol). Following the
addition, the reaction was allowed to warm to rt. After 2 h, the reaction was concentrated and
purified by normal phase chromatography (gradient elution 0ꢀ40% EtOAc/Hex) which gave 7a
1
+
(
1.33 g, 86%) as a yellow solid. MS (ESI) m/z: 614.2 (M+H) . H NMR (400 MHz, CD OD) δ
3
8.43 (d, J=2.20 Hz, 1H), 7.50ꢀ7.57 (m, 2H), 7.33 (d, J=7.70 Hz, 1H), 7.13 (d, J=8.24 Hz, 1H,
NH), 5.76ꢀ5.98 (m, 2H), 5.60 (d, J=10.44 Hz, 1H), 5.34 (d, J=10.99 Hz, 1H), 4.92ꢀ5.20 (m, 5H),
3
.73 (s, 3H), 3.60 (t, J=8.25 Hz, 2H), 2.66ꢀ2.85 (m, 2H), 2.54ꢀ2.62 (m, 2H), 2.44ꢀ2.54 (m, 2H),
1
3
1.43 (s, 9H), 0.87ꢀ1.02 (m, 2H), 0.00 (s, 9H). C NMR (100 MHz, CD OD) δ 173.5, 157.8,
3
1
56.7, 149.0, 140.4, 139.6, 138.4, 137.5, 135.5, 128.0, 118.7, 118.5, 117.8, 116.3, 115.5, 112.5,
80.7, 76.4, 67.5, 52.7, 47.9, 39.8, 38.9, 31.2, 29.0, 18.9, ꢀ1.1.
tert‐Butyl N‐[(11E,14S)‐5‐[(methoxycarbonyl)amino]‐16‐{[2‐(trimethylsilyl)ethoxy]-methyl}
2
,7
‐16,18‐diazatricyclo[13.2.1.0 ]octadeca‐1(17),2,4,6,11,15(18)‐hexaen‐14‐yl]carbamate (8a)
and tert‐Butyl N‐[(11Z,14S)‐5‐[(methoxycarbonyl)amino]‐16‐{[2‐(trimethylsilyl)ethoxy]-
2
,7
methyl}‐16,18‐diazatricyclo[13.2.1.0 ]octadeca‐1(17),2,4,6,11,15(18)‐hexaen‐14‐yl]-
carbamate (8b). (Flask 1): To a flameꢀdried flask was added Grubbs (II) (0.087 g, 0.10 mmol).
ACS Paragon Plus Environment
28

Page 29 of 55
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The flask was degassed with argon for several minutes and then degassed DCM (3 mL) was
added to give a clear, burgundy solution. (Flask 2): To a separate flameꢀdried RBF was added
4b (0.150 g, 0.26 mmol), pꢀtoluenesulfonic acid monohydrate (0.054 g, 0.28 mmol) and DCM
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
(
366 mL). The flask was equipped with a reflux condenser and the solution was degassed with
argon for 30 min. The reaction was heated to 40 °C. After 1 h, the solution of Grubbs (II) from
flask one was added dropwise to flask 2. After 1.5 h, the reaction was cooled to rt, washed with
sat. NaHCO , brine, dried over MgSO , filtered and concentrated to give a brown foam. HPLC
3
4
of the reaction mixture prior to purification indicated a 3.2:1 E:Z alkene ratio. The E/Zꢀalkene
isomers were separated by reverse phase chromatography. The pure fractions of each isomer
were combined, neutralized with sat. NaHCO and then concentrated to give a white solid. The
3
,
white solid was partitioned between water and EtOAc and the layers were separated. The
aqueous layer was extracted with EtOAc (1x). The organic layers were combined and washed
with brine, dried over Na SO , filtered and concentrated to afford Eꢀalkene 8a (0.058 g, 41%,
2
4
peak 1) as a clear, pale brown oil and Zꢀalkene 8b (0.016 g, 11%, peak 2) as a clear, pale brown
1
+
oil. For Eꢀalkene 8a: MS (ESI) m/z: 557.4 (M+H) . H NMR (500 MHz, CDCl , 50 °C) δ 7.20ꢀ
3
7
.25 (m, 3H), 6.85 (s, 1H), 6.60 (s, 1H), 5.97ꢀ6.08 (m, 1H), 5.49 (ddd, J=4.95, 10.17, 15.12 Hz,
H), 5.38 (d, J=11.00 Hz, 1H), 5.19ꢀ5.27 (m, 1H), 5.16 (d, J=10.45 Hz, 1H), 4.64ꢀ4.77 (m, 1H),
1
3.79 (s, 3H), 3.54ꢀ3.72 (m, 3H), 2.72ꢀ2.82 (m, 1H), 2.23ꢀ2.36 (m, 2H), 1.91ꢀ2.01 (m, 1H), 1.66ꢀ
1
1
1
.80 (m, 2H), 1.46 (s, 9H), 0.83ꢀ1.06 (m, 3H), 0.02 (s, 9H). Hꢀ Hꢀhomonuclear decoupling of
the two sets of vicinal methylene protons which revealed the J coupling constant between the
1
3
double bond protons J = 14.8ꢀ15.4 Hz. C NMR (125 MHz, CD OD) δ 157.4, 156.6, 148.1,
3
1
45.0, 139.9, 139.7, 136.4, 130.7, 129.1, 126.8, 121.5, 119.9, 117.0, 80.8, 76.1, 67.5, 52.6, 50.9,
9.7, 32.6, 32.1, 32.0, 28.9, 18.9, ꢀ1.2. For Zꢀalkene 8b: MS (ESI) m/z: 557.3 (M+H)+.
3
ACS Paragon Plus Environment
29