Hepatitis C Virus NS3/4A Protease Inhibitors Incorporating Flexible P2 Quinoxalines Target Drug Resistant Viral Variants

Article abstract of DOI:10.1021/acs.jmedchem.7b00426

A substrate envelope-guided design strategy is reported for improving the resistance profile of HCV NS3/4A protease inhibitors. Analogues of 5172-mcP1P3 were designed by incorporating diverse quinoxalines at the P2 position that predominantly interact with the invariant catalytic triad of the protease. Exploration of structure-activity relationships showed that inhibitors with small hydrophobic substituents at the 3-position of P2 quinoxaline maintain better potency against drug resistant variants, likely due to reduced interactions with residues in the S2 subsite. In contrast, inhibitors with larger groups at this position were highly susceptible to mutations at Arg155, Ala156, and Asp168. Excitingly, several inhibitors exhibited exceptional potency profiles with EC₅₀ values ≤5 nM against major drug resistant HCV variants. These findings support that inhibitors designed to interact with evolutionarily constrained regions of the protease, while avoiding interactions with residues not essential for substrate recognition, are less likely to be susceptible to drug resistance.

Full text of DOI:10.1021/acs.jmedchem.7b00426

Subscriber access provided by Binghamton University | Libraries
Article
Hepatitis C Virus NS3/4A Protease Inhibitors Incorporating
Flexible P2 Quinoxalines Target Drug Resistant Viral Variants
Ashley N. Matthew, Jacqueto Zephyr, Caitlin J. Hill, Muhammad Jahangir, Alicia Newton,
Christos J. Petropoulos, Wei Huang, Nese Kurt Yilmaz, Celia A. Schiffer, and Akbar Ali
J. Med. Chem., Just Accepted Manuscript • Publication Date (Web): 08 Jun 2017
Downloaded from http://pubs.acs.org on June 10, 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 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Hepatitis C Virus NS3/4A Protease Inhibitors
Incorporating Flexible P2 Quinoxalines Target Drug
Resistant Viral Variants
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
†
†
†,#
†,§
Ashley N. Matthew, Jacqueto Zephyr, Caitlin. J. Hill, Muhammad Jahangir, Alicia
‡
‡
‡
†
†,
Newton, Christos J. Petropoulos, Wei Huang, Nese KurtꢀYilmaz, Celia A. Schiffer *, and
†
,
Akbar Ali *
†
Department of Biochemistry and Molecular Pharmacology, University of Massachusetts
Medical School, Worcester, Massachusetts 01605, United States
‡
Monogram Biosciences, South San Francisco, California 94080, United States
Corresponding Authors
Akbar Ali: Phone: +1 508 856 8873; Fax: +1 508 856 6464; Eꢀmail: Akbar.Ali@umassmed.edu
Celia A. Schiffer: Phone: +1 508 856 8008; Fax: +1 508 856 6464; Eꢀmail:
Celia.Schiffer@umassmed.edu
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 61
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
ABSTRACT
A substrate envelopeꢀguided design strategy is reported for improving the resistance profile of
HCV NS3/4A protease inhibitors. Analogues of 5172ꢀmcP1P3 were designed by incorporating
diverse quinoxalines at the P2 position that predominantly interact with the invariant catalytic
triad of the protease. Exploration of structureꢀactivity relationships showed that inhibitors with
small hydrophobic substituents at the 3ꢀposition of P2 quinoxaline maintain better potency
against drug resistant variants, likely due to reduced interactions with residues in the S2 subsite.
In contrast, inhibitors with larger groups at this position were highly susceptible to mutations at
Arg155, Ala156 and Asp168. Excitingly, several inhibitors exhibited exceptional potency
profiles with EC values ≤ 5 nM against major drug resistant HCV variants. These findings
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
0
support that inhibitors designed to interact with evolutionarily constrained regions of the
protease, while avoiding interactions with residues not essential for substrate recognition, are less
likely to be susceptible to drug resistance.
INTRODUCTION
Hepatitis C virus (HCV) infects over 130 million people globally and is the leading cause of
1
chronic liver disease, cirrhosis, and hepatocellular carcinoma. HCV is known as a “silent killer”
as a majority of affected patients remain unaware of their infection, and over time the acute
2
infection progresses to chronic liver disease. The rate of cirrhosis is estimated to increase from
3
16% to 32% by the year 2020 due to the high number of untreated patients. Thus, there is an
urgent need to ensure that patients infected with HCV receive proper treatment. However, HCV
infection is difficult to treat, as the virus is genetically diverse with six known genotypes
4
(genotype 1–6), each of which is further subꢀdivided into numerous subtypes. Genotype 1
ACS Paragon Plus Environment
2

Page 3 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(
GT1) and genotype 3 (GT3) are the most prevalent accounting for 46% and 30% of global
4
,5
infections, respectively.
Therapeutic regimen and viral response are largely genotype
6
dependent with most treatments being efficacious only against GT1.
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
The recent advent of directꢀacting antivirals (DAAs) targeting essential viral proteins NS3/4A,
NS5A, and NS5B has remarkably improved therapeutic options and treatment outcomes for
6
,7
HCV infected patients. Four new allꢀoral combination treatments have been approved by the
8
9
US FDA: (1) sofosbuvir/ledipasvir, (2) ombitasvir/paritaprevir/ritonavir/dasabuvir, (3)
1
0
11
elbasvir/grazoprevir, and (4) sofosbuvir/velpatasvir. The DAAꢀbased therapies are highly
6,7
effective against GT1 with sustained virological response (SVR) rates greater than 90%.
However, most of the FDA approved treatments and those in clinical development are not
7
efficacious against other genotypes, especially GT3. Moreover, except for sofosbuvir, all
12
current DAAs are susceptible to drug resistance. Therefore, more robust DAAs need to be
developed with higher barriers to drug resistance and a broad spectrum of activity against
different HCV genotypes.
The HCV NS3/4A protease is a major therapeutic target for the development of panꢀgenotypic
13,14
15
16
HCV inhibitors.
The NS3/4A protease inhibitors (PIs) telaprevir and boceprevir were the
first DAAs approved for the treatment of HCV GT1 infection in combination therapy with
1
7,18
19
20
pegylatedꢀinterferon and ribavirin.
Three recently approved PIs, simeprevir, paritaprevir
2
1
and grazoprevir, (Figure 1) are integral components of various combination therapies currently
6,7,14
22
used as the standard of care for HCV infected patients.
Two other NS3/4A PIs, asunaprevir
2
3
and vaniprevir, have been approved in Japan. In addition, a number of next generation NS3/4A
2
4
25
PIs are in clinical development including glecaprevir and voxilaprevir (Figure 1).
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 61
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
All NS3/4A PIs share a common peptidomimetic scaffold and are either linear or macrocyclic;
14
the macrocycle is located either between P1–P3 or P2–P4 moieties. In addition, these inhibitors
contain a large heterocyclic moiety attached to the P2 proline, which significantly improves
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
26,27
inhibitor potency against wildꢀtype (WT) NS3/4A protease.
However, all NS3/4A PIs are
susceptible to drug resistance, especially due to single site mutations at protease residues
2
8,29
Arg155, Ala156 and Asp168.
Notably, D168A/V mutations are present in nearly all patients
1
2
who fail treatment with PIs. Moreover, natural polymorphisms at this position are responsible
3
0
for significantly reduced inhibitor potency against GT3. We previously determined the
molecular mechanisms of drug resistance due to single site mutations by solving highꢀresolution
3
1ꢀ34
crystal structures of PIs bound to WT and mutant proteases.
These crystal structures revealed
that the large heterocyclic P2 moieties of PIs bind outside the substrate binding region, defined
as the substrate envelope, and make extensive interactions with residues Arg155, Ala156 and
3
2,33
Asp168.
The inhibitor P2 moiety induces an extended S2 subsite by forcing the Arg155 side
31
chain to rotate nearly 180° relative to its conformation in substrate complexes. This altered
Arg155 conformation is stabilized by electrostatic interactions with Asp168, providing additional
hydrophobic surface that is critical for efficient inhibitor binding. Disruption of electrostatic
interactions between Arg155 and Asp168 due to mutations underlies drug resistance against
3
1ꢀ33,35
NS3/4A PIs.
Moreover, we have shown that structural differences at the P2 moiety largely
3
6
determine the resistance profile of these inhibitors.
Grazoprevir (MKꢀ5172, 1), one of the most potent HCV NS3/4A PIs, has a unique binding
mode where the P2 quinoxaline moiety interacts with residues of the catalytic triad, avoiding
3
2
direct interactions with Arg155 and Asp168 (Figure 2). As a result, 1 has an excellent potency
profile across different genotypes and relatively low susceptibility to drug resistance due to
ACS Paragon Plus Environment
4

Page 5 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
21,37
mutations at Arg155 and Asp168.
However, 1 is highly susceptible to mutations at Ala156,
mainly due to steric clashes of larger side chains with the P2–P4 macrocycle. We have shown
that the P1–P3 macrocyclic analogue 5172ꢀmcP1P3 (2) avoids this steric clash while still
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
38
maintaining the unique binding mode of 1 (Figure 2). Compound 2, though slightly less potent
than 1 against WT HCV, has an excellent potency profile with EC values in the single digit
5
0
nanomolar range against drug resistant variants including A156T. Similar to 1, the P2
quinoxaline moiety in 2 stacks against the catalytic residues His57 and Asp81 and largely avoids
3
8
direct interactions with residues around the S2 subsite. But unlike 1, the flexible P2 quinoxaline
moiety in 2 better accommodates mutations at Ala156, resulting in an overall improved
3
6,38
resistance profile.
Thus, the P1–P3 macrocyclic analogue 2 is a promising lead compound for
structureꢀactivity relationship (SAR) studies to further improve potency against drug resistant
variants and other genotypes.
The substrate envelope model provides a rational approach to design NS3/4A PIs with
improved resistance profiles by exploiting interactions with the protease residues essential for
39ꢀ41
function and avoiding direct contacts with residues that can mutate to confer drug resistance.
Another approach applied to design PIs with improved resistant profiles involves incorporation
of conformational flexibility that can allow the inhibitor to adapt to structural changes in the
3
5
protease active site due to mutations. Here, we describe a structureꢀguided strategy that
combines these two approaches and, together with our understanding of the mechanisms of drug
resistance, led to the design of NS3/4A PIs with exceptional potency profiles against major drug
resistant HCV variants. Based on the lead compound 2, a series of analogues were designed and
synthesized with diverse substituents at the 3ꢀposition of P2 quinoxaline moiety. Investigation of
SARs identified P2 quinoxaline derivatives that predominantly interact with the invariant
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 61
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
catalytic triad and avoid contacts with the S2 subsite residues. The results indicate that
combining the substrate envelope model with optimal conformational flexibility provides a
general strategy for the rational design of NS3/4A PIs with improved resistance profiles.
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
CHEMISTRY
The NS3/4A PIs with diverse P2 quinoxaline moieties were synthesized using the reaction
sequence outlined in Scheme 1. A Cs CO ꢀmediated S 2 reaction of 3ꢀsubstituted quinoxalinꢀ2ꢀ
2
3
N
ones 8a-g with the activated proline derivative 3 provided the key P2 intermediates 9a-g in 75–
90% yield. The alternate S Ar reaction between activated quinoxaline derivatives and Bocꢀ
N
protected hydroxyꢀproline resulted in lower yields, and purification of the resulting P2 acid
products was significantly more challenging. The 3ꢀsubstituted 7ꢀmethoxyꢀquinoxalinꢀ2ꢀones 8a-
b and 8d-e were prepared by condensation reactions of 4ꢀmethoxybenzeneꢀ1,2ꢀdiamine with the
corresponding ethyl glyoxylates (see Supporting Information for details). The 3ꢀchloroꢀ7ꢀ
2
1
methoxyquinoxalinꢀ2ꢀone 8c was prepared according to a reported method.
The P1–P3 macrocyclic PIs were assembled from the P2 intermediates 9a-g using a sequence
of deprotection and peptide coupling steps followed by the ringꢀclosing metathesis (RCM)
reaction (Method A). Removal of the Boc group in 9a-g using 4 N HCl provided the amine salts
10a-g, which were coupled with the amino acid 11 in the presence of HATU and DIEA to yield
the P2–P3 ester intermediates 12a-g. Hydrolysis of these esters with LiOH and reaction of the
4
2
43
resulting carboxylic acids 13a-g with the P1–P1’ fragments 14 and 15 under HATU/DIEA
coupling conditions provided the bisꢀolefin intermediates 16a-g and 17a-e. Finally, cyclization
of the bisꢀolefin intermediates was accomplished using a highly efficient RCM catalyst Zhan 1B,
and provided the inhibitors 18b-g and 19a-e in 45–80% yield. Interestingly, RCM reactions of
bisꢀolefins 17a-e bearing the 1ꢀmethylcyclopropylsulfonamide provided higher yield than the
ACS Paragon Plus Environment
6

Page 7 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
corresponding cyclopropylsulfonamide analogues 16a-g. Finally, removal of the Boc group and
reaction of the resulting amine salts 20a-g and 21a-e with the Nꢀ(cyclopentyloxycarbonyloxy)ꢀ
succinimide in the presence of DIEA afforded the inhibitors 22a-g and 23a-e with the Nꢀterminal
cyclopentyl P4 moiety.
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 subset of inhibitors was synthesized using an alternate reaction sequence that allowed lateꢀ
stage modification at both the P1’ and P4 positions as illustrated in Scheme 2 (Method B). The
P2–P3 acid intermediates 13a-d were reacted with the commercially available amine salt 24
under HATU/DIEA coupling conditions to afford the bisꢀolefin intermediates 25a-d. RCM
reaction in the presence of Zhan 1B catalyst provided the macrocyclic intermediates 26a-d in
75–90% yield, which was better than that obtained in the presence of the P1’ acylsulfonamide.
The P1–P3 macrocyclic core intermediates 26a-d can be modified in either direction after
removing the Cꢀ or Nꢀterminal protecting groups. Thus, hydrolysis of the Cꢀterminal ethyl ester
with LiOH provided the acids 27a-d, which were then reacted with either
cyclopropylsulfonamide 28 or 1ꢀmethylcyclopropylsulfonamide 29 in the presence of CDI and
DBU to afford the final inhibitors 18b-d and 19a-d. The Nꢀterminal tertꢀbutyl capping group
was replaced with the cyclopentyl moiety as described earlier to provide the target inhibitors
22a-d and 23a-d.
RESULTS AND DISCUSSION:
Our goal was to develop a structureꢀguided design strategy to improve the resistance profile of
39,40
HCV NS3/4A PIs based on the substrate envelope model.
Compound 2 is an attractive
scaffold for exploring this strategy due to the unique structural features: (1) the P2 quinoxaline
moiety that predominantly interacts with the highly conserved catalytic residues Asp81 and
ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 61
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
His57 and (2) the conformational flexibility that allows the inhibitor to efficiently accommodate
structural changes in the S2 subsite due to resistance mutations. Despite these promising
features, optimization of substituents at the P2 quinoxaline and the Nꢀterminal capping may be
key to discovering analogues with improved potency and resistance profiles. Therefore, efforts
were focused on exploration of SARs at the P2 quinoxaline moiety in 2, specifically substituting
the ethyl group at the 3ꢀposition that directly interacts with protease S2 subsite residues Arg155
and Ala156. The SAR strategy was based on insights from detailed structural analysis of 1 and 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
3
2,38
bound to wildꢀtype NS3/4A protease and drug resistant variants.
Based on these insights, we
hypothesized that small hydrophobic groups at the 3ꢀposition of the quinoxaline would be
preferred for retaining inhibitor potency against drug resistant variants, but larger groups that
make extensive interactions with Arg155, Ala156 and Asp168 would result in inhibitors highly
susceptible to mutations at these positions. To test this hypothesis, a series of inhibitors with
diverse substituents at the 3ꢀposition of P2 quinoxaline were designed and synthesized.
The potency and resistance profiles of NS3/4A PIs were assessed using biochemical and
replicon assays. The enzyme inhibition constants (Ki) were determined against wildꢀtype GT1a
NS3/4A protease, drugꢀresistant variant D168A, and GT3a NS3/4A protease (Table 1). The
cellular antiviral potencies (EC ) were determined using repliconꢀbased antiviral assays against
5
0
wildꢀtype HCV and drugꢀresistant variants R155K, A156T, D168A, and D168V (Table 2).
Grazoprevir (GZR, 1) was used as a control in all assays. The observed antiviral potencies are
generally higher than protease inhibitory potencies, likely because biochemical assays were
performed using the protease domain alone rather then the fullꢀlength NS3/4A.
Compound 1 showed subꢀnanomolar inhibitory potency against WT NS3/4A protease and
maintained nanomolar activity against drug resistant variant D168A and GT3a protease.
ACS Paragon Plus Environment
8

Page 9 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Similarly, in replicon assays 1 exhibited an excellent potency profile with subꢀnanomolar activity
against WT HCV (EC = 0.14 nM) and low nanomolar activity against drug resistant variants
50
3
6
R155K, D168A, and D168V. However, in line with previous reports, 1 was highly susceptible
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
to the A156T mutation (EC = 238 nM), losing over 1000ꢀfold potency against this variant.
5
0
Compared to 1, the P1–P3 macrocyclic analogue 2 exhibited lower inhibitory potency against
WT protease and the D168A variant. Also, the inhibitory activity of 2 against the GT3a protease
3
6
was considerably lower than that of 1. However, as we have previously shown, 2 displayed a
superior potency profile in replicon assays with subꢀnanomolar activity against WT HCV (EC50
=
0.33 nM) and maintained single digit nanomolar potency against all drugꢀresistant variants
tested. Notably, unlike 1, compound 2 maintained low nanomolar potency against the A156T
variant (EC = 9.65 nM). Thus, with an improved resistance profile compared to 1, the P1–P3
5
0
macrocyclic analogue 2 is an attractive lead compound for further optimization.
Modifications of P1’ and P4 Capping Groups. Initial SAR efforts to optimize lead
compound 2 focused on exploring changes at the P1’ position and Nꢀterminal capping group.
Recent SAR studies of diverse NS3/4A PIs indicate that replacement of the
cyclopropylsulfonamide moiety at the P1’ position with a slightly more hydrophobic 1ꢀ
4
3,44
methylcyclopropylsulfonamide improves inhibitor potency in replicon assays.
Moreover,
changes at the P4 position have been shown to significantly affect inhibitor potency against drug
resistant variants, as these groups bind in close proximity to the pivotal drug resistance site
45
Asp168. For carbamateꢀlinked P4 capping groups, generally bulky hydrophobic moieties are
preferred but the size of the group appears to be dependent on the heterocyclic moiety present at
3
5
the P2 position.
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 61
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
First, replacing the cyclopropylsulfonamide at the P1’ position in 2 with 1ꢀ
methylcyclopropylsulfonamide provided the analogue 19a. Compared to the parent compound 2,
19a showed slightly better Ki values against WT, D168A and GT3a proteases and exhibited
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
similar or slightly better antiviral potency against WT and drug resistant variants. Next, the tertꢀ
butyl P4 capping group in both 2 and 19a was replaced with a larger cyclopentyl moiety,
resulting in analogues 22a and 23a. Unlike the change at the P1’ position, the P4 cyclopentyl
modification provided mixed results. Compound 22a afforded a 2ꢀfold increase in potency than 2
in biochemical assays against WT protease and a slight improvement against the D168A variant,
but was equipotent to 2 against GT3a protease. Similarly, in replicon assays 22a exhibited 2ꢀfold
enhanced potency against WT HCV and D168A variant, but showed similar potency as 2 against
the R155K and D168V variants. Compound 23a, with a 1ꢀmethylcyclopropylsulfonamide moiety
at the P1’ position and a cyclopentyl group at the P4 position, exhibited potency profile largely
similar to 22a. Surprisingly, a slight loss in potency was observed against the A156T variant for
compounds with a cyclopentyl versus tertꢀbutyl capping group. Overall, these minor
modifications at the P1’ and Nꢀterminal capping regions of inhibitor 2 were tolerated and
provided analogues with improved potency profiles.
SAR Exploration of P2 Quinoxaline. Next, SARs at the P2 quinoxaline in compound 2 were
explored. Efforts mainly focused on replacing the 3ꢀposition ethyl group with diverse functional
groups with respect to size and electronic properties. Replacement of the ethyl group in 2 with a
smaller methyl group provided analogue 18b. As expected, reducing the size of the hydrophobic
group at this position resulted in improved potency profile. Compound 18b showed slightly
enhanced potency against drug resistant variants in biochemical and antiviral assays, with a
notable ~2ꢀfold improvement against the D168V variant (EC = 3.17 nM). The introduction of
5
0
ACS Paragon Plus Environment
10

Page 11 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
ꢀmethylcyclopropylsulfonamide moiety at the P1’ position afforded inhibitor 19b with protease
inhibitory activity comparable to the parent compound 18b. However, similar to the 3ꢀ
ethylquinoxaline analogue (19a), compound 19b demonstrated significant gain in potency in
replicon assays. In fact, compared to 2, 19b exhibited 2ꢀ to 6ꢀfold enhancement in potency
against drug resistant variants R155K (EC = 0.80 nM), A156T (EC = 1.57 nM), D168A (EC
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
5
0
50
50
=
2.37 nM), and D168V (EC = 1.6 nM). Replacement of the tertꢀbutyl P4 capping in 18b and
50
19b with a cyclopentyl group, providing 22b and 23b, resulted in an increase in WT and D168A
inhibitory activity as well as 2ꢀ to 3ꢀfold increase in WT replicon potency. Unlike the
corresponding 3ꢀethylquinoxaline analogues (22a and 23a), the 3ꢀmethyquinoxaline compounds
22b and 23b maintained the excellent potency profile observed for the corresponding tertꢀbutyl
analogues. Remarkably, with the exception of 18b (A156T EC = 5.95 nM), all compounds in
5
0
the 3ꢀmethylquinoxaline series display exceptional potency profiles with EC values below 5
5
0
nM against WT and clinically relevant drug resistant variants.
To gain insights into the excellent potency profile observed for the 3ꢀmethyquinoxaline series,
we determined the Xꢀray crystal structure of inhibitor 19b in complex with the WT NS3/4A
protease at a resolution of 1.8 Å (Figure 3, Table S3, PDB code: 5VOJ). The WTꢀ19b complex
structure was compared with the previously reported structures of compound 2 in complex with
3
8
WT protease and the A156T variant (PDB codes: 5EPN and 5EPY). The two WT structures
overlap very well, with only minor differences in the S1 and S2 subsites because of
modifications in the inhibitor structure. In the WTꢀ2 crystal structure, the 3ꢀethyl group at the P2
quinoxaline makes hydrophobic interactions with the hydrocarbon portion of the Arg155 sideꢀ
chain, while the methylene portion of this group interacts with the sideꢀchain of Ala156. The
smaller methyl group at this position in the WTꢀ19b structure maintains hydrophobic interactions
ACS Paragon Plus Environment
11

Journal of Medicinal Chemistry
Page 12 of 61
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
with Ala156, while minimizing chances of steric clash with a larger sideꢀchain, such as in
A156T.
Unlike inhibitor 1, the P1–P3 macrocyclic analogues retain potency against the A156T variant.
Comparison of the WTꢀ2 and A156Tꢀ2 (PDB code: 5EPY) structures shows subtle changes in
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
3
8
inhibitor interactions with the mutant protease. In the A156Tꢀ2 structure the P2 quinoxaline
largely maintains interactions with the catalytic residues, but the ethyl group is shifted away
from Arg155 side chain toward A156T. Moreover, to accommodate a larger Thr sideꢀchain, the
Asp168 side chain adopts another conformation, moving away from Arg155. These changes
underlie reduced inhibitor potency against the A156T variant, but unlike 1, inhibitor 2 is able to
better accommodate these changes due to a flexible P2 moiety. The 3ꢀmethylquinoxaline
analogues are more potent against the A156T variant than the corresponding 3ꢀethylquinoxaline
compounds likely due the reduced interactions of the smaller methyl group with the Thr sideꢀ
chain. Replacing the methyl group with hydrogen at the 3ꢀposition of quinoxaline would further
reduce interactions with the S2 subsite residues, but could result in a highly flexible P2 moiety,
likely destabilizing interactions with the catalytic residues. Thus, a small hydrophobic group at
the 3ꢀposition of P2 quinoxaline is preferred to maintain favorable interactions with Ala156 and
avoid steric clashes with the Thr sideꢀchain in the A156T variant.
The improved potency profile of 3ꢀmethyquinoxaline compounds led to exploration of
bioisosteric replacements of the 3ꢀmethyl group with varied size and electronic properties. To
that end, analogues 18c and 19c bearing the 3ꢀchloroꢀ7ꢀmethoxyquinoxaline at the P2 position
were prepared. The protease inhibitory potency profiles of these compounds were excellent and
showed improvement against WT, D168A and GT3a over 2. These potency gains were not only
maintained in replicon assays but were more significant, with the only exception of A156T
ACS Paragon Plus Environment
12

Page 13 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
variant. Both compounds 18c and 19c were more active than the corresponding 3ꢀ
methylquinoxaline analogues (18b and 19b) with EC50 values less than 1 nM against WT,
R155K and D168V and less than 2 nM against the D168A variant, but experienced about 3ꢀ to 6ꢀ
fold reduction in potency against the A156T variant. However, potency losses against the A156T
variant were largely reversed when the P4 tertꢀbutyl group in 18c and 19c was replaced with a
larger cyclopentyl moiety to afford 22c and 23c. Similar to the 3ꢀmethylquinoxaline compounds,
the 3ꢀchloroquinoxaline analogues displayed exceptional potency profiles with EC values of
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
50
less than 5 nM against all drug resistant variants including A156T. These results clearly
demonstrate that small hydrophobic groups with weak electronꢀdonating properties at the 3ꢀ
position of P2 quinoxaline can be replaced with weak electronꢀwithdrawing groups without
affecting the overall potency profile.
Next, a larger and strongly electronꢀwithdrawing trifluoromethyl moiety was explored at the 3ꢀ
position of P2 quinoxaline, leading to inhibitors 18d and 19d. This modification, however,
resulted in significant potency losses in both biochemical and replicon assays. Compound 18d
was about 2ꢀ to 4ꢀfold less active than 2 against WT protease and variants. Analogue 19d with
the 1ꢀmethylcyclopropylsulfonamide moiety at the P1’ position showed similar trends when
compared to the corresponding 19a. In line with biochemical data, both 18d and 19d suffered 2ꢀ
to 6ꢀfold decrease in replicon potency against WT and drug resistant variants, though 19d
maintained relatively good potency profile. In contrast to the results in previous series, the
introduction of the larger cyclopentyl P4 capping group, as in 22d and 23d, was detrimental to
replicon potency, particularly against the A156T variant. Moreover, compounds in the 3ꢀ
(trifluoromethyl)quinoxaline series were among the least active against the GT3a protease in
biochemical assays. These results indicate that strong electronꢀwithdrawing groups at the 3ꢀ
ACS Paragon Plus Environment
13

Journal of Medicinal Chemistry
Page 14 of 61
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
position of the P2 quinoxaline may be detrimental to potency. However, a recent SAR study
indicates that PIs incorporating the 3ꢀ(trifluoromethyl)quinoxaline can be optimized with
modifications at the 7ꢀposition of quinoxaline in combination with changes at the P1–P3
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
46
macrocycle and P4 capping group.
In the absence of a coꢀcrystal structure, the lower inhibitory potencies of compounds in the 3ꢀ
(trifluoromethyl)quinoxaline series against WT protease could not be explained by molecular
modeling, which suggested a similar binding conformation of the P2 quinoxaline in 18d as
observed for 2 (Figure 4A). Perhaps there are repulsive interactions between trifluoromethyl
moiety and the side chain of Asp168, and/or the strong electronꢀwithdrawing effect may weaken
the overall interactions of the P2 quinoxaline with the catalytic residues. Potency losses against
resistant variants may also result from the larger size of the trifluoromethyl moiety, which is
47
comparable to that of an ethyl group, though both have different topographical shapes.
To isolate the effects of larger size versus electronic properties on potency, inhibitors 18e and
19e with the larger isopropyl group at the 3ꢀposition of the P2 quinoxaline were designed and
evaluated. These compounds showed WT protease inhibitory activity similar to the
corresponding 3ꢀethylquinoxaline analogues (2 and 19a), but experienced 2ꢀ to 4ꢀfold reduced
activity against the D168A variant. A broader reduction in potency was observed for both 18e
and 19e in replicon assays against WT and drug resistant variants. The cyclopentyl P4 group in
analogues 22e and 23e slightly improved biochemical and replicon potency against WT and
D168A variants, but was largely unfavorable to replicon potency against R155K and A156T
variants. This trend is broadly similar to the results observed with the 3ꢀ
(trifluoromethyl)quinoxaline series, indicating that both electronic properties and size of the
group at the 3ꢀposition of P2 quinoxaline are important for maintaining potency against drug
ACS Paragon Plus Environment
14

Page 15 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
resistant variants. Modeling indicated that compared to 2 the P2 quinoxaline moiety in 18e has to
shift away from the catalytic triad in order to accommodate the larger isopropyl group thereby
weakening critical stacking interactions with His57 (Figure 4B). Overall, SAR data from the 3ꢀ
isopropylꢀ and 3ꢀ(trifluoromethyl)ꢀquinoxaline series supports the hypothesis that large
substituents at the 3ꢀposition of P2 quinoxaline have detrimental effect on inhibitor potency
against drug resistant variants.
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
These findings were further reinforced by the results obtained for the 3ꢀ(thiophenꢀ2ꢀ
yl)quinoxaline analogues 18f and 22f. Based on molecular modeling, the large thiophene moiety
in these compounds was expected to make extensive interactions with the residues Arg155 and
Ala156, resulting in improved potency against WT protease. However, mutations at these
positions as well as at Asp168 would cause significant potency losses, as these residues are
crucial for efficient inhibitor binding. As expected, compound 18f (a previously reported
1
9,48
NS3/4A PI incorrectly labeled as ABTꢀ450)
biochemical potency but was dramatically less active against the D168A variant, losing over
800ꢀfold potency. Similarly, in replicon assays analogue 18f showed considerably reduced
showed a 3ꢀfold enhancement in WT
1
potency against all drug resistant variants with losses ranging from 20ꢀ to 250ꢀfold compared to
WT (Table S1 and S2). The cyclopentyl P4 analogue 22f also experienced large potency losses
against the variants, albeit to a lesser extent than 18f. Thus inhibitors with large groups at the 3ꢀ
position of P2 quinoxaline are highly susceptible to mutations at residues Arg155, Ala156 and
Asp168, leading to poor resistance profiles.
The Xꢀray crystal structure of inhibitor 18f in complex with WT NS3/4A protease was
determined at a resolution of 1.9 Å, providing insights into the binding modes of P2 quinoxaline
with a larger thiophene substituent at the 3ꢀposition (Figure 5, Table S3, PDB code: 5VP9).
ACS Paragon Plus Environment
15

Journal of Medicinal Chemistry
Page 16 of 61
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
Comparison of the WTꢀ18f and WTꢀ2 crystal structures showed significant differences in the
interactions of quinoxaline moieties with the catalytic triad and S2 subsite residues. As predicted,
the quinoxaline moiety in WTꢀ18f structure is shifted toward the active site to accommodate the
larger thiophene substituent. The thiophene ring makes extensive interactions with residues in
the S2 subsite, including cationꢀπ interactions with Arg155, likely contributing to the improved
potency against the WT protease. As this Arg155 conformation is stabilized by electrostatic
interactions with Asp168, mutations at either residue would disrupt inhibitor binding by loss of
direct interactions as well as indirect structural effects. In addition, the A156T mutation would
result in a steric clash with the thiophene ring, as reflected in the antiviral data for this variant.
These biochemical and structural findings are in line with previous studies that show inhibitors
that are dependent on extensive interactions with the S2 subsite residues for potency are highly
susceptible to mutations at residues Arg155, Ala156 and Asp168.
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
As compounds 18f and 22f lacked the Cꢀ7 substituent at the P2 quinoxaline, analogues 18g
and 22g were prepared to investigate the effect of this group on inhibitor potency. Compared to
2
, analogue 18g experienced about 2ꢀfold decrease in biochemical potency and only minor loss
in replicon potency against WT and drug resistant variants. The P4 cyclopentyl analogue 22g
resulted in about 2ꢀfold reduced potency compared to the corresponding compound 22a. Thus
removal of the Cꢀ7 methoxy group has minimal effect on inhibitor potency. The slightly reduced
potency of 18g and 22g is likely due to the reduced hydrophobic interactions with the aromatic
ring of Tyr56 and the methylene portion of His57 of the catalytic triad. In contrast, the observed
potency losses against resistant variants for the 3ꢀ(thiophenꢀ2ꢀyl)quinoxaline compounds most
likely result from loss of interactions of the 2ꢀthiophene moiety with the S2 subsite residues of
the protease.
ACS Paragon Plus Environment
16

Page 17 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Effects of P2 Substituent Size and Flexibility. Taken together, our SAR results indicate that
resistance profiles of compound 2 and analogues are strongly influenced by the substituent at the
3ꢀposition of P2 quinoxaline and Nꢀterminal capping group. While all PIs showed reduced
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
potency against drug resistance variants in both enzyme inhibition and replicon assays, fold
potency losses varied significantly depending on the substituents at the 3ꢀposition of P2
quinoxaline. To evaluate susceptibility to the clinically important D168A variant, to which all
current NS3/4A PIs are susceptible, potencies were normalized to WT for PIs with the same P4
capping groups (Figure 6). Fold changes in Ki against the D168A protease variant for PIs with
the same P1’ and P4 capping groups largely trended with the size of the substituent at the 3ꢀ
position of P2 quinoxaline, with the exception of trifluoromethyl compounds. Losses in potency
were significantly higher for compounds with the larger 2ꢀthiophene substituent at the P2
quinoxaline. These results strongly support using the substrate envelope model to reduce direct
inhibitor interactions in the S2 subsite, thereby reducing inhibitor susceptibility to drug
resistance.
31ꢀ33,35
As we and others have shown,
the reduced potencies of NS3/4A PIs against drug
resistant variants R155K, A156T, and D168A/V mainly result from disruption of the electrostatic
interactions between Arg155 and Asp168. Compared to 1, compound 2 and most analogues
incorporating flexible P2 quinoxaline showed lower foldꢀchanges in potency against these
variants (Table S1 and S2). In these P1–P3 macrocyclic PIs the conformational flexibility of the
P2 allows this moiety to adapt to the structural changes caused by mutations at Arg155, Ala156
and Asp168, resulting in better resistance profiles. Potency losses were higher for compound 1
because constraint imposed by the macrocycle does not allow the P2 moiety to adapt to the
structural changes resulting from these mutations. Compound 1 and similar P2–P4 macrocyclic
ACS Paragon Plus Environment
17

Journal of Medicinal Chemistry
Page 18 of 61
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
PIs, such as voxilaprevir and glecaprevir, are likely to be more susceptible to mutations that
cause significant structural changes in the protease active site. However, the P1–P3 macrocyclic
compounds reported here, as well as those reported in patent literature that incorporate similar
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
49
flexible P2 quinoxaline moieties, are likely to be more effective against clinically relevant drug
resistant variants. More broadly, combining the substrate envelope model with optimal
conformational flexibility provides a rational approach to design NS3/4A PIs with improved
resistance profiles.
CONCLUSIONS
Drug resistance is a major problem across all DAA classes targeting HCV. As new therapies
are developed the potential for drug resistance must be minimized at the outset of inhibitor
design. The substrate envelope model provides a rational approach to design robust NS3/4A PIs
with improved resistance profiles. Our SAR findings support the hypothesis that reducing PI
interactions with residues in the S2 subsite leads to inhibitors with exceptional potency and
resistance profiles. Specifically, the P1–P3 macrocyclic inhibitors incorporating flexible P2
quinoxaline moieties bearing small hydrophobic groups at the 3ꢀposition maintain excellent
potency in both enzymatic and antiviral assays against drug resistant variants. While these
inhibitors protrude from the substrate envelope, they leverage interactions with the essential
catalytic triad residues and avoid direct contacts with residues that can mutate to confer
resistance. Moreover, conformational flexibility at the P2 moiety is essential to efficiently
accommodate structural changes due to mutations in the S2 pocket in order to avoid resistance.
These insights provide strategies for iterative rounds of inhibitor design with the paradigm that
designing inhibitors with flexible P2 quinoxalines, leveraging evolutionarily constrained areas in
ACS Paragon Plus Environment
18

Page 19 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
the protease active site and expanding into the substrate envelope may provide inhibitors that are
robust against drug resistant variants.
EXPERIMENTAL SECTION
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
General. All reactions were performed in ovenꢀdried round bottomed or modified Schlenk
flasks fitted with rubber septa under argon atmosphere, unless otherwise noted. All reagents and
solvents, including anhydrous solvents, were purchased from commercial sources and used as
received. Flash column chromatography was performed using silica gel (230–400 mesh, EMD
Millipore). Thinꢀlayer chromatography (TLC) was performed using silica gel (60 Fꢀ254) coated
aluminum plates (EMD Millipore), and spots were visualized by exposure to ultraviolet light
(UV), exposure to iodine adsorbed on silica gel, and/ or exposure to an acidic solution of pꢀ
1
13
anisaldehyde (anisaldehyde) followed by brief heating. H NMR and C NMR spectra were
acquired on Varian Mercury 400 MHz and Bruker Avance III HD 500 MHz NMR instruments.
Chemical shifts are reported in ppm (δ scale) with the residual solvent signal used as reference
and coupling constant (J) values are reported in hertz (Hz). Data are presented as follows:
chemical shift, multiplicity (s = singlet, d = doublet, dd = doublet of doublet, t = triplet, q =
quartet, m = multiplet, br s = broad singlet), coupling constant in Hz, and integration. Highꢀ
resolution mass spectra (HRMS) were recorded on a Thermo Scientific Orbitrap Velos Pro mass
spectrometer coupled with a Thermo Scientific Accela 1250 UPLC and an autosampler using
electrospray ionization (ESI) in the positive mode. The purity of final compounds was
determined by analytical HPLC and was found to be ≥95% pure. HPLC was performed on a
Waters Alliance 2690 system equipped with a Waters 2996 photodiode array detector and an
autosampler under the following conditions: column, Phenomenex Lunaꢀ2 RPꢀC18 (5 ꢁm, 4.6 ×
2
50 mm, 120 Å, Torrance, CA); solvent A, H O containing 0.1% formic acid (FA), solvent B,
2
ACS Paragon Plus Environment
19

Journal of Medicinal Chemistry
Page 20 of 61
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
CH CN containing 0.1% FA; gradient, 50% B to 100% B over 15 min followed by 100% B over
3
5
min; injection volume, 10 ꢁL; flow rate, 1 mL/ min. Retention times and purity data for each
target compound are provided in the experimental section.
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
Typical procedures for the synthesis of protease inhibitors using Method A:
1-(tert-Butyl) 2-methyl (2S,4R)-4-((3-ethyl-7-methoxyquinoxalin-2-yl)oxy)pyrrolidine-1,2-
dicarboxylate (9a). A solution of 3ꢀethylꢀ7ꢀmethoxyquinoxalinꢀ2ꢀone 8a (3.0 g, 14.7 mmol) in
anhydrous NMP (45 mL) was treated with Cs CO (7.40 g, 22.7 mmol). After stirring the
2
3
reaction mixture at room temperature for 15 min, proline derivative 3 (6.20 g, 13.3 mmol) was
added in one portion. The reaction mixture was heated to 55 °C, stirred for 4 h, and then another
portion of proline derivative 3 (0.48 g, 1.0 mmol) was added. The resulting reaction mixture was
stirred at 55 °C for an additional 2 h, cooled to room temperature, quenched with aqueous 1 N
HCl solution (150 mL), and extracted with EtOAc (300 mL). The organic fraction was washed
successively with saturated aqueous NaHCO and NaCl (150 mL each), dried (Na SO ), filtered,
3
2
4
and evaporated under reduced pressure. The residue was purified by flash column
chromatography using 15–30% EtOAc/hexanes as the eluent to provide 9a (5.50 g, 87%) as a
1
white foamy solid. H NMR (400 MHz, CDCl ) (mixture of rotamers, major rotamer) δ 7.85 (d, J
3
=
9.0 Hz, 1 H), 7.18 (m, 1H), 7.11 (d, J = 2.8 Hz, 1 H), 5.73 (br s, 1 H), 4.47 (t, J = 8.0 Hz, 1 H),
3.98–3.86 (m, 5 H), 3.78 (s, 3 H), 2.92 (q, J = 7.2 Hz, 2 H), 2.68–2.60 (m, 1 H), 2.43–2.36 (m, 1
13
H), 1.43 (s, 9 H), 1.31 (t, J = 7.2 Hz, 3 H) ppm; C NMR (100 MHz, CDCl ) δ 173.56, 160.59,
3
1
5
4
55.38, 154.02, 148.95, 141.26, 134.12, 129.07, 119.02, 106.11, 80.76, 73.81, 58.43, 55.93,
+
2.73, 52.40, 36.88, 28.47, 26.68, 11.97 ppm; HRMS (ESI) m/z: [M + H] calcd for C H N O ,
22 30
3
6
32.2129; found 432.2135.
ACS Paragon Plus Environment
20

Page 21 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
-(tert-Butyl)
2-methyl
(2S,4R)-4-((7-methoxy-3-(trifluoromethyl)quinoxalin-2-
yl)oxy)pyrrolidine-1,2-dicarboxylate (9d). The same procedure was used as described above
for compound 9a. 7ꢀmethoxyꢀ3ꢀ(trifluoromethyl)quinoxalinꢀ2(1H)ꢀone 8d (4.76 g, 19.5 mmol)
in NMP (65 mL) was treated with Cs CO (9.80 g, 30.0 mmol) and proline derivative 3 (9.0 g,
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
2
3
1
19.3 mmol) to provide 9d (6.50 g, 71%) as a pale yellow foamy solid. H NMR (500 MHz,
CDCl ) (mixture of rotamers, major rotamer) δ 7.77 (d, J = 9.0 Hz, 1 H), 7.48–7.43 (m, 2 H),
3
5
.76 (br s, 1 H), 4.50 (t, J = 8.0 Hz, 1 H), 3.97–3.91 (m, 5 H), 3.78 (s, 3 H), 2.69–2.64 (m, 1 H),
1
3
2.41–2.34 (m, 1 H), 1.42 (s, 9 H) ppm; C NMR (125 MHz, CDCl ) δ 173.43, 159.58, 153.98,
3
1
52.11, 138.39, 137.22, 127.99, 125.73, 120.70 (q, J = 273.4 Hz), 107.64, 80.69, 74.62, 58.27,
19
5
6.02, 52.32, 52.11, 36.70, 28.34 ppm; F NMR (470 MHz, CDCl ); −67.73 ppm; HRMS (ESI)
3
+
m/z: [M + H] calcd for C H F N O , 472.1690; found 472.1689.
21
25
3
3
6
Methyl
methoxyquinoxalin-2-yl)oxy)pyrrolidine-2-carboxylate (12a). A solution of ester 9a (4.80 g,
1.1 mmol) in anhydrous CH Cl (30 mL) was treated with a solution of 4 N HCl in 1,4ꢀdioxane
(2S,4R)-1-((S)-2-((tert-butoxycarbonyl)amino)non-8-enoyl)-4-((3-ethyl-7-
1
2
2
(
30 mL). After stirring the reaction mixture at room temperature for 3 h, solvents were
evaporated under reduced pressure, and the residue was dried under high vacuum. The pale
yellow solid was triturated with diethyl ether (3 × 30 mL) and dried under high vacuum to yield
the amine salt 10a (4.0 g, 98%) as an offꢀwhite powder.
A mixture of amine salt 10a (4.0 g, 10.9 mmol) and (S)ꢀ2ꢀ((tertꢀbutoxycarbonyl)amino)nonꢀ8ꢀ
enoic acid 11 (3.0 g, 11.1 mmol) in anhydrous DMF (60 mL) was treated with DIEA (7.30 mL,
4
4.2 mmol) and HATU (6.35 g, 16.7 mmol). The resulting reaction mixture was stirred at room
temperature for 4 h, then diluted with EtOAc (400 mL), and washed successively with aqueous
0
.5 N HCl, saturated aqueous NaHCO , and saturated aqueous NaCl (250 mL each). The organic
3
ACS Paragon Plus Environment
21

Journal of Medicinal Chemistry
Page 22 of 61
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
portion was dried (Na SO ), filtered, and evaporated under reduced pressure. The residue was
2
4
purified by flash chromatography using 20–30% EtOAc/hexanes as the eluent to provide 12a
1
(
5.50 g, 86%) as a white foamy solid. H NMR (400 MHz, CDCl ) (mixture of rotamers, major
3
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
rotamer) δ 7.86 (d, J = 8.8 Hz, 1 H), 7.20 (dd, J = 9.2, 2.8 Hz, 1 H), 7.12 (d, J = 2.8 Hz, 1 H),
5.87–5.75 (m, 2 H), 5.20 (d, J = 8.4 Hz, 1 H), 5.02–4.92 (m, 2 H), 4.73 (t, J = 8.4 Hz, 1 H), 4.38
(q, J = 7.2 Hz, 1 H), 4.17 (d, J = 12.0 Hz, 1 H), 4.06 (dd, J = 11.6, 4.4 Hz, 1 H), 3.94 (s, 3 H),
3.78 (s, 3 H), 2.90 (q, J = 7.6 Hz, 2 H), 2.69–2.64 (m, 1 H), 2.41–2.34 (m, 1 H), 2.05 (app q, J =
13
6.8 Hz, 2 H), 1.82–1.74 (m, 1 H), 1.63–1.56 (m, 1 H), 1.45–1.25 (m, 18 H) ppm; C NMR (100
MHz, CDCl ) δ 172.34, 171.96, 160.61, 155.61, 155.13, 148.95, 141.08, 139.18, 129.22, 119.08,
3
114.58, 106.14, 79.84, 74.48, 58.19, 55.91, 52.88, 52.67, 52.05, 35.16, 33.88, 32.88, 29.14,
+
28.96, 28.46, 26.52, 24.92, 11.86 ppm; HRMS (ESI) m/z: [M + H] calcd for C H N O
3
1
45
4
7
5
85.3283; found 585.3286.
Methyl (2S,4R)-1-((S)-2-((tert-butoxycarbonyl)amino)non-8-enoyl)-4-((7-methoxy-3-
trifluoromethyl)quinoxalin-2-yl)oxy)pyrrolidine-2-carboxylate (12d). The same procedure
(
was used as described above for compound 12a. Compound 9d (6.0 g, 12.7 mmol) was treated
with 4 N HCl (40 mL) to afford amine salt 10d (5.10 g, 12.5 mmol), which was coupled with
acid 11 (3.80 g, 14.0 mmol) using DIEA (9.25 mL, 56.0 mmol) and HATU (7.60 g, 20.0 mmol)
1
to provide 12d (6.40 g, 81%) as a pale yellow foamy solid. H NMR (500 MHz, CDCl ) (mixture
3
of rotamers, major rotamer) δ 7.78 (d, J = 9.0 Hz, 1 H), 7.48 (dd, J = 9.0, 2.5 Hz, 1 H), 7.44 (d, J
=
2.5 Hz, 1 H), 5.86 (br s, 1 H), 5.84–5.78 (m, 1 H), 5.18 (d, J = 9.0 Hz, 1 H), 5.01–4.92 (m, 2
H), 4.75 (t, J = 8.0 Hz, 1 H), 4.35 (q, J = 7.5 Hz, 1 H), 4.19 (d, J = 12.0 Hz, 1 H), 4.08 (dd, J =
11.5, 4.5 Hz, 1 H), 3.95 (s, 3 H), 3.78 (s, 3 H), 2.70–2.65 (m, 1 H), 2.41–2.35 (m, 1 H), 2.04 (app
13
q, J = 7.0 Hz, 2 H), 1.80–1.75 (m, 1 H), 1.60–1.54 (m, 1 H), 1.45–1.28 (m, 15 H) ppm; C NMR
ACS Paragon Plus Environment
22

Page 23 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(
125 MHz, CDCl ) δ 172.10, 171.60, 159.99, 155.37, 151.78, 138.98, 138.41, 136.93, 134.40 (q,
3
J = 36.3 Hz), 127.85, 125.66, 120.53 (q, J = 273.4 Hz), 114.33, 107.54, 79.58, 75.05, 57.83,
1
9
5
5.91, 52.44, 52.33, 51.75, 34.77, 33.65, 32.70, 28.91, 28.73, 28.18, 24.70 ppm; F NMR (470
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
+
MHz, CDCl ); −67.73 ppm; HRMS (ESI) m/z: [M + H] calcd for C H F N O , 625.2844;
3
30 40
3
4
7
found 625.2844.
tert-Butyl
((S)-1-((2S,4R)-2-(((1R,2S)-1-((cyclopropylsulfonyl)carbamoyl)-2-
vinylcyclopropyl)carbamoyl)-4-((3-ethyl-7-methoxyquinoxalin-2-yl)oxy)pyrrolidin-1-yl)-1-
oxonon-8-en-2-yl)carbamate (16a). A solution of ester 12a (5.86 g, 10.0 mmol) in THFꢀH O
2
mixture (1:1, 140 mL) was treated with LiOH.H O (1.40 g, 33.4 mmol). The resulting reaction
2
mixture was stirred at room temperature for 24 h. The reaction mixture was cooled to ~5 °C,
acidified to a pH of 2.0 by slow addition of aqueous 0.25 N HCl (~ 200 mL), and extracted with
EtOAc (2 × 400 mL). The organic portions were washed separately with saturated aqueous NaCl
(
200 ml), dried (Na SO ), filtered, and evaporated under reduced pressure. The gummy residue
2 4
was dissolved in CHCl (50 mL), concentrated under reduced pressure, and the residue was dried
3
under high vacuum overnight to yield the acid 13a (5.70 g, 100%) as a white foamy solid.
4
2
A mixture of acid 13a (2.10 g, 3.7 mmol) and amine salt 14 (1.20 g, 4.5 mmol) in anhydrous
DMF (35 mL) was treated with DIEA (2.43 mL, 14.7 mmol) and HATU (2.1 g, 5.5 mmol). The
resulting reaction mixture was stirred at room temperature for 2.5 h, then diluted with EtOAc
(300 mL) and washed successively with aqueous 0.5 N HCl, saturated aqueous NaHCO , and
3
saturated aqueous NaCl (200 mL each). The organic portion was dried (Na SO ), filtered, and
2
4
evaporated under reduced pressure. The residue was purified by flash chromatography using 50–
7
0% EtOAc/hexanes as the eluent to provide the bisꢀolefin compound 16a (2.50 g, 86%) as a
1
white solid. H NMR (400 MHz, CDCl ) δ 10.24 (s, 1 H), 7.84 (d, J = 8.8 Hz, 1 H), 7.18 (dd, J =
3
ACS Paragon Plus Environment
23

Journal of Medicinal Chemistry
Page 24 of 61
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
8
.8, 2.4 Hz, 1 H), 7.13 (d, J = 2.8 Hz, 1 H), 7.04 (s, 1 H), 5.91 (br s, 1 H), 5.85–5.73 (m, 2 H),
.32 (d, J = 8.4 Hz, 1 H), 5.27 (d, J = 17.2 Hz, 1 H), 5.14 (d, J = 11.2 Hz, 1 H), 5.01–4.90 (m, 2
5
H), 4.47 (t, J = 7.6 Hz, 1 H), 4.38–4.33 (m, 1 H), 4.20 (d, J = 11.6 Hz, 1 H), 4.02 (dd, J = 11.2,
.0 Hz, 1 H), 3.94 (s, 3 H), 2.96–2.84 (m, 3 H), 2.56–2.51 (m, 2 H), 2.11 (q, J = 8.8 Hz, 1 H),
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
1
3
2.05–1.99 (m, 3 H), 1.74–1.54 (m, 2 H), 1.47–1.10 (m, 21 H), 1.08–1.03 (m, 2 H) ppm; C
NMR (100 MHz, CDCl ) δ 174.09, 172.58, 168.69, 160.54, 155.89, 154.99, 148.88, 140.95,
3
139.07, 134.69, 132.71, 129.45, 119.02, 118.77, 114.67, 106.13, 80.0, 74.66, 60.61, 55.91, 53.42,
52.62, 41.83, 35.46, 34.47, 33.89, 32.40, 31.39, 28.98, 28.89, 28.47, 26.68, 25.47, 23.83, 11.85,
+
6.68, 6.26 ppm; HRMS (ESI) m/z: [M + H] calcd for C H N O S, 783.3746; found 783.3734.
3
9
55
6
9
tert-Butyl
((S)-1-((2S,4R)-4-((3-ethyl-7-methoxyquinoxalin-2-yl)oxy)-2-(((1R,2S)-1-(((1-
methylcyclopropyl)sulfonyl)carbamoyl)-2-vinylcyclopropyl)carbamoyl)pyrrolidin-1-yl)-1-
oxonon-8-en-2-yl)carbamate (17a). The same procedure was used as described above for
4
3
compound 16a. Acid 13a (1.50 g, 2.6 mmol) was coupled with amine salt 15 (0.90 g, 3.2
mmol) using DIEA (1.75 mL, 10.6 mmol) and HATU (1.50 g, 3.9 mmol) to provide the bisꢀ
1
olefin compound 17a (1.75 g, 84%) as a white solid. H NMR (400 MHz, CDCl ) δ 10.02 (s, 1
3
H), 7.84 (d, J = 9.2 Hz, 1 H), 7.19 (dd, J = 8.8, 2.8 Hz, 1 H), 7.13 (d, J = 2.8 Hz, 1 H), 7.06 (s, 1
H), 5.90 (br s, 1 H), 5.83–5.73 (m, 2 H), 5.37 (d, J = 9.2 Hz, 1 H), 5.27 (d, J = 17.2 Hz, 1 H),
5
.14 (d, J = 10.8 Hz, 1 H), 5.98 (dd, J = 17.2, 1.6 Hz, 1 H), 4.92 (dd, J = 10.4, 1.2 Hz, 1 H), 4.48
(t, J = 8.0 Hz, 1 H), 4.39–4.33 (m, 1 H), 4.16 (d, J = 12.0 Hz, 1 H), 4.02 (dd, J = 11.6, 4.0 Hz, 1
H), 3.94 (s, 3 H), 2.89 (q, J = 7.6 Hz, 2 H), 2.57–2.50 (m, 2 H), 2.12 (q, J = 8.8 Hz, 1 H), 2.05–
1.99 (m, 3 H), 1.75–1.58 (m, 4 H), 1.49 (s, 3 H), 1.45–1.18 (m, 19 H), 0.93–0.79 (m, 2 H) ppm;
1
3
C NMR (100 MHz, CDCl ) δ 173.79, 172.41, 167.51, 160.31, 155.71, 154.76, 148.63, 140.73,
3
1
38.85, 134.41, 132.60, 129.18, 118.80, 118.54, 114.41, 105.89, 79.74, 74.42, 60.36, 55.68,
ACS Paragon Plus Environment
24

Page 25 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
5
3.17, 52.43, 41.71, 36.56, 35.23, 34.22, 33.64, 32.19, 28.70, 28.67, 28.25, 26.43, 25.35, 23.49,
+
18.37, 14.27, 13.29, 11.64 ppm; HRMS (ESI) m/z: [M + H] calcd for C H N O S, 797.3902;
40 57
6
9
found 797.3887.
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
tert-Butyl ((2R,6S,13aS,14aR,16aS,Z)-14a-((cyclopropylsulfonyl)carbamoyl)-2-((3-ethyl-7-
methoxyquinoxalin-2-yl)oxy)-5,16-dioxo-1,2,3,5,6,7,8,9,10,11,13a,14,14a,15,16,16a-
hexadecahydrocyclopropa[e]pyrrolo[1,2-a][1,4]diazacyclopentadecin-6-yl)carbamate (2). A
degassed solution of bisꢀolefin 16a (1.40 g, 1.8 mmol) in 1,2ꢀDCE (300 mL) was heated to 50 °C
under argon, then Zhan 1b catalyst (0.150 g, 0.20 mmol) was added in two portions over 10 min.
The resulting reaction mixture was heated to 70 °C and stirred for 6 h. The reaction mixture was
cooled to room temperature and concentrated under reduced pressure. The residue was purified
by flash chromatography using 50–80% EtOAc/hexanes as the eluent to yield the P1–P3
1
macrocyclic product 2 (0.72 g, 53%) as an offꢀwhite solid. H NMR (400 MHz, CDCl ) δ 10.28
3
(
s, 1 H), 7.84 (d, J = 9.6 Hz, 1 H), 7.20–7.15 (m, 2 H), 6.91 (s, 1 H), 5.90 (br s, 1 H), 5.69 (q, J =
.8 Hz, 1 H), 5.14 (d, J = 7.6 Hz, 1 H), 4.96 (t, J = 9.2 Hz, 1 H), 4.59 (t, J = 7.6 Hz, 1 H), 4.49
d, J = 11.6 Hz, 1 H), 4.30–4.24 (m, 1 H), 4.02 (dd, J = 10.8, 4.0 Hz, 1 H), 3.94 (s, 3 H), 2.94–
2.85 (m, 3 H), 2.70–2.51 (m, 3 H), 2.31 (q, J = 8.8 Hz, 1 H), 1.93–1.64 (m, 2 H), 1.60–1.05 (m,
8
(
1
3
2
4 H), 0.95–0.89 (m, 1 H) ppm; C NMR (100 MHz, CDCl ) δ 177.15, 173.28, 168.02, 160.29,
3
155.00, 154.90, 148.66, 140.88, 136.31, 134.28, 128.90, 124.47, 118.82, 105.91, 79.84, 74.68,
59.45, 55.72, 53.08, 51.92, 44.57, 34.65, 32.81, 31.01, 29.70, 28.14, 27.11, 27.16, 26.31, 26.06,
+
2
2.16, 20.92, 11.56, 6.67, 6.12 ppm; HRMS (ESI) m/z: [M + H] calcd for C H N O S,
3
7
51
6
9
7
55.3433; found 755.3410. Anal. HPLC: t 14.23 min, purity 97%.
R
tert-Butyl ((2R,6S,13aS,14aR,16aS,Z)-2-((3-ethyl-7-methoxyquinoxalin-2-yl)oxy)-14a-(((1-
methylcyclopropyl)sulfonyl)carbamoyl)-5,16-dioxo-
ACS Paragon Plus Environment
25

Journal of Medicinal Chemistry
Page 26 of 61
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
1
,2,3,5,6,7,8,9,10,11,13a,14,14a,15,16,16a-hexadecahydrocyclopropa[e]pyrrolo[1,2-
a][1,4]diazacyclopentadecin-6-yl)carbamate (19a). The same procedure was used as described
above for compound 2. Bisꢀolefin 17a (1.45 g, 1.8 mmol) was treated with Zhan 1b catalyst
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
(0.150 g, 0.20 mmol) in 1,2ꢀDCE (300 mL) to afford the P1–P3 macrocyclic product 19a (1.0 g,
1
71%) as an offꢀwhite solid. H NMR (400 MHz, CDCl ) δ 10.16 (s, 1 H), 7.82 (d, J = 10.0 Hz, 1
3
H), 7.18–7.15 (m, 2 H), 6.94 (s, 1 H), 5.90 (br s, 1 H), 5.69 (q, J = 9.2 Hz, 1 H), 5.16 (d, J = 8.0
Hz, 1 H), 4.99 (t, J = 9.2 Hz, 1 H), 4.59 (t, J = 8.0 Hz, 1 H), 4.49 (d, J = 11.6 Hz, 1 H), 4.30–4.25
(m, 1 H), 4.04 (dd, J = 11.6, 4.0 Hz, 1 H), 3.94 (s, 3 H), 2.87 (q, J = 7.6 Hz, 2 H), 2.70–2.51 (m,
3
H), 2.33 (q, J = 8.0 Hz, 1 H), 1.92–1.68 (m, 4 H), 1.60–1.15 (m, 24 H), 0.85–0.78 (m, 2 H)
13
ppm; C NMR (100 MHz, CDCl ) δ 177.19, 173.24, 167.0, 160.23, 154.99, 154.88, 148.73,
3
140.84, 136.26, 134.25, 129.03, 124.89, 118.72, 105.92, 79.84, 74.67, 59.48, 55.72, 53.11, 51.92,
44.71, 36.43, 34.68, 32.80, 29.62, 28.14, 27.09, 26.38, 26.12, 22.19, 20.93, 18.17, 14.51, 12.50,
+
11.54 ppm; HRMS (ESI) m/z: [M + H] calcd for C H N O S, 769.3589; found 769.3561.
38 53
6
9
Anal. HPLC: t 15.01 min, purity 99%.
R
Cyclopentyl
((2R,6S,13aS,14aR,16aS,Z)-14a-((cyclopropylsulfonyl)carbamoyl)-2-((3-
ethyl-7-methoxyquinoxalin-2-yl)oxy)-5,16-dioxo-1,2,3,5,6,7,8,9,10,11,13a,14,14a,15,16,16a-
hexadecahydrocyclopropa[e]pyrrolo[1,2-a][1,4]diazacyclopentadecin-6-yl)carbamate (22a).
Compound 2 (0.40 g, 0.53 mmol) was treated with a solution of 4 N HCl in 1,4ꢀdioxane (10 mL).
The reaction mixture was stirred at room temperature for 3 h, then concentrated under reduced
pressure, and the residue was dried under high vacuum. The offꢀwhite solid was triturated with
diethyl ether (3 × 10 mL) and dried under high vacuum to yield the amine salt 20a (0.37 g,
1
00%) as a white powder.
ACS Paragon Plus Environment
26

Page 27 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
A solution of the above amine salt 20a (0.37 g, 0.53 mmol) in anhydrous CH CN (15 mL) was
3
treated with DIEA (0.35 mL, 2.1 mmol) and Nꢀ(cyclopentyloxycarbonyloxy)ꢀsuccinimide (0.15
g, 0.66 mmol). The reaction mixture was stirred at room temperature for 36 h, then concentrated
under reduced pressure and dried under high vacuum. The residue was purified by flash
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
chromatography using 50–90% EtOAc/hexanes as the eluent to provide the target compound 22a
1
(
0.32 g, 79%) as a white solid. H NMR (400 MHz, CDCl ) δ 10.29 (s, 1 H), 7.83 (d, J = 9.6 Hz,
3
1
H), 7.21–7.16 (m, 2 H), 6.94 (s, 1 H), 5.93 (br s, 1 H), 5.70 (q, J = 8.8 Hz, 1 H), 5.26 (d, J =
8.0 Hz, 1 H), 4.96 (t, J = 8.4 Hz, 1 H), 4.86–4.82 (m, 1 H), 4.60 (t, J = 7.6 Hz, 1 H), 4.45 (d, J =
1
1.2 Hz, 1 H), 4.34–4.28 (m, 1 H), 4.03 (dd, J = 11.2, 4.0 Hz, 1 H), 3.94 (s, 3 H), 2.93–2.85 (m,
H), 2.70–2.48 (m, 3 H), 2.30 (q, J = 8.8 Hz, 1 H), 1.93–1.23 (m, 23 H), 1.15–1.06 (m, 2 H),
3
1
3
0.96–0.88 (m, 1 H) ppm; C NMR (100 MHz, CDCl ) δ 177.18, 173.03, 168.04, 160.28, 155.65,
3
1
5
2
7
54.93, 148.78, 140.90, 136.27, 134.20, 128.92, 124.46, 118.80, 105.92, 77.87, 74.55, 59.47,
5.72, 53.01, 52.17, 44.54, 34.58, 32.72, 32.63, 32.59, 31.01, 29.70, 27.14, 27.05, 26.40, 26.05,
+
3.56, 22.16, 20.90, 11.61, 6.67, 6.12 ppm; HRMS (ESI) m/z: [M + H] calcd for C H N O S,
3
8
51
6
9
67.3433; found 767.3408. Anal. HPLC: t 14.50 min, purity 98%.
R
Cyclopentyl ((2R,6S,13aS,14aR,16aS,Z)-2-((3-ethyl-7-methoxyquinoxalin-2-yl)oxy)-14a-
((1-methylcyclopropyl)sulfonyl)carbamoyl)-5,16-dioxo-
,2,3,5,6,7,8,9,10,11,13a,14,14a,15,16,16a-hexadecahydrocyclopropa[e]pyrrolo[1,2-
(
1
a][1,4]diazacyclopentadecin-6-yl)carbamate (23a). The same procedure was used as described
above for compound 22a. Compound 19a (0.40 g, 0.52 mmol) was treated with 4 N HCl in 1,4ꢀ
dioxane (10 mL) to yield the amine salt 21a, which was treated with DIEA (0.35 mL, 2.1 mmol)
and Nꢀ(cyclopentyloxycarbonyloxy)ꢀsuccinimide (0.15 g, 0.66 mmol) to provide the target
1
compound 23a (0.30 g, 74%) as a white solid. H NMR (400 MHz, CDCl ) δ 10.17 (s, 1 H), 7.81
3
ACS Paragon Plus Environment
27

Journal of Medicinal Chemistry
Page 28 of 61
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
(
d, J = 9.6 Hz, 1 H), 7.21–7.16 (m, 2 H), 6.93 (s, 1 H), 5.92 (br s, 1 H), 5.70 (q, J = 9.2 Hz, 1 H),
5
4
2
.26 (d, J = 7.6 Hz, 1 H), 4.99 (t, J = 9.6 Hz, 1 H), 4.86–4.81 (m, 1 H), 4.59 (t, J = 7.6 Hz, 1 H),
.45 (d, J = 11.2 Hz, 1 H), 4.34–4.28 (m, 1 H), 4.04 (dd, J = 11.6, 4.0 Hz, 1 H), 3.94 (s, 3 H),
.87 (q, J = 7.2 Hz, 2 H), 2.70–2.48 (m, 3 H), 2.32 (q, J = 8.8 Hz, 1 H), 1.92–1.23 (m, 27 H),
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
3
0.85–0.78 (m, 2 H) ppm; C NMR (100 MHz, CDCl ) δ 177.21, 172.99, 166.98, 160.22, 155.63,
3
1
54.90, 148.84, 140.85, 136.22, 134.36, 129.05, 124.88, 118.70, 105.93, 77.86, 74.54, 59.51,
5
5.71, 53.05, 52.16, 44.70, 36.43, 34.61, 32.72, 32.64, 32.58, 29.63, 27.13, 27.06, 26.47, 26.12,
+
23.56, 22.18, 20.94, 18.17, 14.49, 12.50, 11.59 ppm; HRMS (ESI) m/z: [M + H] calcd for
C H N O S, 781.3589; found 781.3561. Anal. HPLC: t 15.25 min, purity 99%.
39
53
6
9
R
Typical procedures for the synthesis of protease inhibitors using Method B:
Ethyl (1R,2S)-1-((2S,4R)-1-((S)-2-((tert-butoxycarbonyl)amino)non-8-enoyl)-4-((7-
methoxy-3-(trifluoromethyl)quinoxalin-2-yl)oxy)pyrrolidine-2-carboxamido)-2-
vinylcyclopropane-1-carboxylate (25d). A solution of ester 12d (6.40 g, 10.25 mmol) in THFꢀ
H O (1:1 mixture, 140 mL) was treated with LiOH.H O (1.38 g, 32.0 mmol). The resulting
2
2
reaction mixture was stirred at room temperature for 24 h, then cooled to ~5 °C, acidified to a pH
of 2.0 by slow addition of aqueous 0.25 N HCl (~ 200 mL), and extracted with EtOAc (2 × 500
mL). The organic portions were washed separately with saturated aqueous NaCl (250 ml), dried
(Na SO ), filtered, and evaporated under reduced pressure. The gummy residue was dissolved in
2 4
CHCl (50 mL), concentrated under reduced pressure, and the residue was dried under high
3
vacuum to yield the acid 13d (6.12 g, 98%) as a pale yellow foamy solid.
A solution of acid 13d (6.12 g, 10.0 mmol) and amine salt 24 (2.50 g, 13.0 mmol) in
anhydrous CH Cl (100 mL) was treated with DIEA (9.10 mL, 55.0 mmol), HATU (5.30 g, 14.0
2
2
mmol) and DMAP (0.60 g, 4.9 mmol). The resulting reaction mixture was stirred at room
ACS Paragon Plus Environment
28

Page 29 of 61
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
temperature for 14 h, then diluted with EtOAc (500 mL), and washed successively with aqueous
1
.0 N HCl, saturated aqueous NaHCO , and saturated aqueous NaCl (250 mL each). The organic
3
portion was dried (Na SO ), filtered, and evaporated under reduced pressure. The residue was
2
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 flash chromatography using 25–35% EtOAc/hexanes as the eluent to provide the bisꢀ
1
olefin compound 25d (6.54 g, 87%) as a pale yellow foamy solid. H NMR (500 MHz, CDCl )
3
(
mixture of rotamers, major rotamer) δ 7.78 (d, J = 9.2 Hz, 1 H), 7.53 (br s, 1 H), 7.47 (dd, J =
.2, 2.8 Hz, 1 H), 7.43 (d, J = 2.4 Hz, 1 H), 5.88 (br s, 1 H), 5.81–5.70 (m, 2 H), 5.30 (dd, J =
16.8, 0.8 Hz, 1 H), 5.14–5.10 (m, 2 H), 5.01–4.89 (m, 2 H), 4.79 (dd, J = 14.0, 5.6 Hz, 1 H),
9
4.35–4.29 (m, 1 H), 4.21–4.08 (m, 3 H), 3.94 (s, 3 H), 2.90–2.82 (m, 1 H), 2.48–2.38 (m, 1 H),
2.16 (q, J = 9.0 Hz, 1 H), 2.04–1.98 (m, 2 H), 1.86 (dd, J = 8.0, 5.2 Hz, 1 H), 1.66–1.52 (m, 2 H),
1
3
1.46 (dd, J = 9.6, 5.6 Hz, 1 H), 1.43–1.21 (m, 19 H) ppm; C NMR (125 MHz, CDCl ) δ 173.02,
3
171.00, 169.87, 159.62, 155.52, 152.03, 138.92, 138.48, 137.16, 133.66, 128.02, 125.73, 120.72
(q, J = 273.6 Hz), 118.08, 114.52, 107.66, 79.98, 75.26, 61.40, 58.41, 56.02, 52.58, 52.43, 40.14,
1
9
3
3.89, 33.77, 32.76, 32.62, 28.97, 28.78, 28.31, 25.18, 23.11, 14.48 ppm; F NMR (470 MHz,
+
CDCl ); −67.77 ppm; HRMS (ESI) m/z: [M + H] calcd for C H F N O , 748.3528; found
3
37 49
3
5
8
748.3514.
Ethyl
(2R,6S,13aS,14aR,16aS,Z)-6-((tert-butoxycarbonyl)amino)-2-((7-methoxy-3-
(
trifluoromethyl)quinoxalin-2-yl)oxy)-5,16-dioxo-1,2,3,6,7,8,9,10,11,13a,14,15,16,16a-
tetradecahydrocyclopropa[e]pyrrolo[1,2-a][1,4]diazacyclopentadecine-14a(5H)-carboxylate
26d). A degassed solution of bisꢀolefin 25d (1.50 g, 2.0 mmol) in 1,2ꢀDCE (300 mL) was
(
heated to 50 °C under argon, then Zhan 1b catalyst (0.150 g, 0.20 mmol) was added in two
portions over 10 min. The resulting mixture was heated to 70 °C and stirred for 5 h. The reaction
mixture was cooled to room temperature and concentrated under reduced pressure. The residue
ACS Paragon Plus Environment
29