Structure-Based Optimization of N-Substituted Oseltamivir Derivatives as Potent Anti-Influenza A Virus Agents with Significantly Improved Potency against Oseltamivir-Resistant N1-H274Y Variant

Article abstract of DOI:10.1021/acs.jmedchem.8b01065

Due to the emergence of highly pathogenic and oseltamivir-resistant influenza viruses, there is an urgent need to develop new anti-influenza agents. Herein, five subseries of oseltamivir derivatives were designed and synthesized to improve their activity toward drug-resistant viral strains by further exploiting the 150-cavity in the neuraminidases (NAs). The bioassay results showed that compound 21h exhibited antiviral activities similar to or better than those of oseltamivir carboxylate (OSC) against H5N1, H5N2, H5N6, and H5N8. Besides, 21h was 5- to 86-fold more potent than OSC toward N1, N8, and N1-H274Y mutant NAs in the inhibitory assays. Computational studies provided a plausible rationale for the high potency of 21h against group-1 and N1-H274Y NAs. In addition, 21h demonstrated acceptable oral bioavailability, low acute toxicity, potent antiviral activity in vivo, and high metabolic stability. Overall, the above excellent profiles make 21h a promising drug candidate for the treatment of influenza virus infection.

Full text of DOI:10.1021/acs.jmedchem.8b01065

Subscriber access provided by Kaohsiung Medical University
Article
Structure-Based Optimization of N-Substituted Oseltamivir
Derivatives as Potent Anti-influenza A Virus Agents with Significantly
Improved Potency against Oseltamivir-Resistant N1-H274Y Variant
Jian Zhang, Natarajan Arul Murugan, Ye Tian, Chiara Bertagnin, Zengjun Fang, Dongwei Kang,
Xiujie Kong, Haiyong Jia, Zhuosen Sun, Ruifang Jia, Ping Gao, Vasanthanathan Poongavanam,
Arianna Loregian, Wenfang Xu, Xiuli Ma, Xiao Ding, bing huang, Peng Zhan, and Xinyong Liu
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b01065 • Publication Date (Web): 26 Oct 2018
Downloaded from http://pubs.acs.org on October 26, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 92
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 92
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
Structure-Based Optimization of N-Substituted Oseltamivir Derivatives as Potent
Anti-influenza A Virus Agents with Significantly Improved Potency against
Oseltamivir-Resistant N1-H274Y Variant
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
†
&
†,@

†, #
Jian Zhang, Natarajan Arul Murugan, Ye Tian, Chiara Bertagnin, Zengjun Fang,
†
†
†
†
†
†
Dongwei Kang, Xiujie Kong, Haiyong Jia, Zhuosen Sun, Ruifang Jia, Ping Gao,
|,◊

†
‡
Vasanthanathan Poongavanam, Arianna Loregian, Wenfang Xu, Xiuli Ma, Xiao
†
‡,
†,
†,
Ding, Bing Huang, * Peng Zhan, * and Xinyong Liu *
†
Department of Medicinal Chemistry, Key Laboratory of Chemical Biology (Ministry of
Education), School of Pharmaceutical Sciences, Shandong University, 44 West Culture
Road, 250012 Jinan, Shandong, PR China.
&
Division of Theoretical Chemistry and Biology, School of Biotechnology, KTH Royal
Institute of Technology, S-106 91 Stockholm, Sweden.
@
Key Laboratory for Experimental Teratology of Ministry of Education, Key Laboratory
of Infection and Immunity of Shandong Province, Department of Immunology, School of
Basic Medical Sciences, Shandong University, 44 West Culture Road, 250012 Jinan,
Shandong, P. R. China.

Department of Molecular Medicine, University of Padova, via Gabelli 63, 35121
Padova, Italy.
‡
Institute of Poultry Science, Shandong Academy of Agricultural Sciences, 1, Jiaoxiao
Road, 250023 Jinan, Shandong, P. R. China.
|
Department of Physics, Chemistry, and Pharmacy, University of Southern Denmark,
DK-5230 Odense M. Denmark.
◊
◊
Present Address: Biomedicinskt Centrum (BMC), Department of Chemistry, Uppsala
University, Husargatan 3, 75237 Uppsala, Sweden.
1
ACS Paragon Plus Environment

Page 3 of 92
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
#
The Second Hospital of Shandong University, No. 247 Beiyuan Avenue, Jinan250033,
China.
ABSTRACT: Due to the emergence of highly pathogenic and oseltamivir-resistant
influenza viruses, there is an urgent need to develop new anti-influenza agents. Herein,
five sub-series of oseltamivir derivatives were designed and synthesized to improve their
activity towards drug-resistant viral strains by further exploiting the 150-cavity in the
neuraminidases (NAs). The bioassay results showed that compound 21h exhibited
antiviral activities similar to or better than those of oseltamivir carboxylate (OSC) against
H5N1, H5N2, H5N6 and H5N8. Besides, 21h was 5- to 86-fold more potent than OSC
toward N1, N8, and N1-H274Y mutant NAs in the inhibitory assays. Computational
studies provided a plausible rationale for the high potency of 21h against group-1 and
N1-H274Y NAs. In addition, 21h demonstrated acceptable oral bioavailability, low acute
toxicity and potent antiviral activity in vivo, and high metabolic stability. Overall, above
excellent profiles make 21h a promising drug candidate for the treatment of influenza
virus infection.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 92
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
INTRODUCTION
Despite the availability of vaccines and antiviral drugs, influenza A virus frequently
1
causes severe respiratory disease. The worldwide spread of swine-origin 2009 pandemic
H1N1 (H1N1pdm09) and avian influenza virus (AIV), especially the H5 subtypes of
highly pathogenic avian influenza (HPAI), have raised public concern globally due to
their high morbidity and mortality rates.2,3 For example, H5N1 subtype of AIV is a
deadly disease in birds, mammals, and humans, with a 63% mortality rate in humans.4
Although the H5N1 virus has not yet acquired the ability to bind efficiently to the human
receptor, it might eventually overcome the species barrier and achieve human-to-human
transmission.5,6
Influenza virus neuraminidase (NA) is an important surface antigenic glycoprotein,
which facilitates viral shedding by cleaving terminal sialic acid residues between another
essential surface antigenic glycoproteins hemagglutinin (HA) and host-cell, and is thus an
attractive target for the prophylaxis and therapy of influenza.7,8 Several neuraminidase
9
inhibitors (NAIs), including oseltamivir (1a), oseltamivir carboxylate (OSC, 1b),
1
0
11
12
zanamivir (ZA, 2), laninamivir octanoate (3), and peramivir (4) (Figure 1), have
been approved for clinical use. Among them, oseltamivir (1a), the first orally available
NAI, was marketed in 2000, and has been stockpiled by many countries as a precaution
against a new pandemic, primarily because of its convenient oral availability.13
3
ACS Paragon Plus Environment

Page 5 of 92
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
O
O
HO
NH2
O
O
HO
O
R
OH
OH
OR2
H
HN
HN
N
H
O
H N
O
O
2
NH
NH OR1
HN
H N
2
HN
O
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
zanamivir (2), R1 = R2 = H
oseltamivir (1a), R = Et
oseltamivir carboxylate (OSC, 1b), R = H laninamivir octanoate (3), R1 = Me, R2 = C7H15CO peramivir (4)
Figure 1. Chemical structures of clinically used neuraminidase inhibitors.
Despite the success of oseltamivir, resistance is a growing concern. Oseltamivir
resistance is generally conferred via a single amino acid mutation (H274Y) in strains
possessing the N1 subtype of NA (exemplified by H1N1pdm09-H274Y and
H5N1-H274Y). Since 2007, viruses containing H274Y have rapidly become predominant
among human seasonal H1N1 and avian influenza H5N1 isolates.14-16 In NA inhibition
assays, NAs with H274Y mutant were more resistant to oseltamivir than the wild-type
NAs by a factor of more than 100,17,18 leading to clinical failure of oral oseltamivir.15,19,20
Therefore, discovery of new drugs active against oseltamivir-resistant strains, especially
those with H274Y, is becoming increasingly urgent.
So far, based on phylogenetic relationships and X-ray crystallographic studies of
NAs, the nine NAs of influenza A viruses have been well characterized and can be
divided into two distinct families: group-1, containing N1, N4, N5 and N8 subtypes, and
group-2, containing N2, N3, N6, N7 and N9 subtypes.21,22 In group-1 NAs, a flexible
150-loop, consisting of residues 147-152, usually adopts an open conformation, whereas
21
in group-2 NAs, this loop is always closed. The open 150-loop can form in an open
large cavity (150-cavity) adjacent to the active site in group-1 NAs, but this cavity is not
present in group-2 NAs. Surprisingly, although the 2009 H1N1 neuraminidase (09N1)
belongs to group-1 NAs, crystal structure studies showed it does not have the 150-cavity
4
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 92
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
23
(
Figure 2). Even so, recent computational studies of the 09N1 subtype suggest that the
1
50-loop has remarkable mobility and may open to a greater extent in solution than is
observed in the crystal.24,25
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
Structure-based design is typically focused on the optimization of a chemical series
to increase affinity for a mutable target protein, to probe biologically relevant chemical
space, and to improve the potency against resistant mutations, as exemplified by our
recent studies.26,27 In this context, the flexibility of the 150-loop and also the opening of
the 150-cavity near the active site of group-1 enzymes can be regarded as important
targets for further drug modification studies. We speculated that modifications of
oseltamivir that enable higher-affinity binding at the amino acids forming the 150-cavity
could yield novel NA inhibitors that are not sensitive to common mutations of NA,
including H274Y.
N1
09N1
N2
HO
O
active site
active site
active site
H2N
O
O
HN
1
50-cavity
open 150-loop
closed 150-loop
closed 150-loop
Figure 2. Comparison of the crystal structures of representative group-1 NAs (N1, PDB ID: 2HU0),
group-2 NAs (N2, PDB ID: 4GZP), and 09N1 (PDB ID: 3TI6).
Close examination of the crystal structure of oseltamivir carboxylate bound to the
N1 subtype indicated that the C-5 amino group of oseltamivir carboxylate could serve as
a potential modification site in the design of group-1-specific inhibitors (Figure 2). On
this basis, many investigators have attempted to chemically modify oseltamivir in order
to overcome resistance-conferring mutations of NAs.28 Our own previous research
focused on exploiting the 150-cavity led to the identification of a series of N-substituted
5
ACS Paragon Plus Environment

Page 7 of 92
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
oseltamivir derivatives with low nanomolar N1-selective inhibitory activity against NAs
1
7
from three H5N1 viruses. It is noteworthy that, compared with OSC, the two most
potent compounds 5 and 6 (group-1-specific NA inhibitors in Figure 3A) showed about
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
-9 times greater inhibitory potency against three H5N1 NA subtypes. However,
compound 5 did not show improved activity against the H5N1-H274Y mutant as
compared to OSC (5, IC = 1.16 μM; OSC, IC = 2.1 μM). But, interestingly, 6 showed
50
50
a 12-fold increase in activity against this mutant relative to OSC (IC50 = 0.16 μM),
supporting the idea that addition of substituents to the oseltamivir core could result in
stronger interactions with the 150-cavity, restoring affinity for the H274Y mutant.17
HO
O
A
B
N
O
O
H
HN
1
50-cavity
5
IC50 (5) = 0.0019 M (N1, H5N1)
0
1
.58 M (N2, H9N2)
.16 M (N1, H5N1-H274Y)
HO
O
N
O
O
H
HN
S
6
IC50 (6) = 0.0021 M (N1, H5N1)
.21 M (N2, H9N2)
0.16 M (N1, H5N1-H274Y)
0
Figure 3. (A) Structures of our previously reported group-1-specific influenza NA inhibitors 5 and 6;
(
B) Key residues that may interact with 5.17
Our previous structure-activity relationships (SARs) study and molecular modeling
indicated that some of the highest-affinity 150-cavity binders interact primarily through
1
7
hydrophobic interactions. As shown in Figure 3B, biphenyl or (thiophen-2-yl)phenyl
6
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 92
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
groups in compounds 5 and 6 might develop additional interactions with hydrophobic
amino acid residues, such as W438, V116, G147, Q136, and R152 in N1 of H5N1. Thus,
the SAR of the C5-NH -substituted benzyl group of compounds 5 and 6, which occupy
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
the 150-cavity of N1, needs to be further explored to improve the potency of these
compounds against N1-H274Y mutants, while retaining their outstanding potency against
other group-1 NAs.
In this work, therefore, we investigated a series of N-substituted benzenes
incorporating various hydrophobic groups directly connected to the aromatic moiety, as
well as some benzophenone and phenylsulfonyl derivatives (Figure 4). Specifically, we
designed five new sub-series (11, 17, 21, 25 and 28, and 32) of C5-NH -substituted
2
oseltamivir derivatives, synthesized them, and then biologically tested them in cell-based
assays that address the cytopathic effect (CPE) of influenza virus infection, as well an in
an enzymatic assay. Furthermore, we conducted preliminary evaluation of druggability,
including oral bioavailability, acute toxicity in vivo and metabolic stability, as well as
computational modeling studies, of some promising compounds.
Ph
,
,
,
O
O
O
O
O
OH
Ph
,
,
,
O
OH
S
S
S
S
O
O
N
H
,
,
,
optimization
NH
S
Ar
S
O
O
N
S
R
Ph
H
,
NH
S
Ar =
,
S
,
S
5
6
S
,
S
O
S
O
Ph
C
Ph,
O
Figure 4. Previously discovered N-substituted oseltamivir derivatives 5 and 6 and design approach to
7
ACS Paragon Plus Environment

Page 9 of 92
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
new derivatives. Pink benzene ring represents the privileged moiety; brown aromatic ring represents
the active moiety of the lead compounds in our previous work; blue part represents our present
structural optimization.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
CHEMISTRY
The new N-substituted oseltamivir derivatives (11a-d, 17a-f, 21a-h, 25a-d, 28, and
32a-b) were synthesized as illustrated in Schemes 1-6.
Scheme 1. Preparation of Compounds 11a-da
O
R
7a-c
O
O
HO
O
O
oseltamivir
phosphate
Br
O
ii
N
O
O
iii
N
O
O
H
H
HN
10a-d
HN
R
R
i
B
11a-d
HO
OH
8
9
7a, 10a, 11a, R =
7b, 10b, 11b, R =
7c, 10c, 11c, R =
10d, 11d, R =
a
(
i) H O, THF, K CO , Pd(PPh ) , N , 60℃; (ii) NaBH CN, EtOH, MeOH, r.t.; (iii) NaOH, MeOH,
2 2 3 3 4 2 3
H O, r.t., then HCl (2 mol/L).
2
As shown in Scheme 1, 4-benzylbenzaldehyde (9) was synthesized from
4-formylphenyl)boronic acid (8) via Suzuki reaction.29 Then, following a previously
(
reported procedure,17 the commercially available para-substituted benzaldehydes 7a-c
and newly prepared 9 were reacted with commercial oseltamivir phosphate in the
presence of NaBH CN to afford the key intermediates 10a-d. Finally, target compounds
3
1
1a-d were prepared by direct hydrolysis of the intermediates 10a-d.
Scheme 2. Preparation of Compounds 17a-fa
8
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 92
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
O
O
R
1
2a-c
O
O
O
HO
O
O
Br
oseltamivir
phosphate
O R
14a-b
ii
N
O
O
iii
N
O
O
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
i
O
R
H
O
R
H
OH
HN
HN
1
3a-b
O
16a-f
17a-f
O
15
1
2a, 16a, 17a, (R = 4-methyl); 12b, 16b, 17b, (R = 3-methyl); 12c, 16c, 17c, (R = 4-isopropyl);
13a, para-position, 13b, meta-position; 14a, 16d, 17d, (R = 4-(pentan-3-yl));
14b, 16e, 17e, (R = 3-(pentan-3-yl));
16f, 17f, (R = 4-phenyl)
a
(
i) K CO , KI, DMF, 100℃. (ii) NaBH CN, EtOH, MeOH, r.t.; (iii) NaOH, MeOH, H O, r.t., then
2
3
3
2
HCl (2 mol/L).
Scheme 2 illustrates the synthesis of compounds 17a-f. Commercially available 3-
or 4-hydroxybenzaldehydes (13a,b) were coupled with 3-bromopentane in DMF in the
presence of K CO and KI to obtain the aromatic aldehydes 14a-b. Then, target
2
3
compounds 17a-f were prepared from 14a-b and other commercially available
arylaldehydes 12a-c and 15 using a similar procedure to that described above.
Scheme 3. Preparation of Compounds 21a-ha
O
O
HO
O
O
O
oseltamivir
phosphate
R
SH
i
ii
N
H
O
O
iii
N
O
O
S
S
H
HN
0a-h
HN
F
S
R
R
R
1
8a-b
19a-h
2
21a-h
1
8a, para-position; 18b, meta-position; 19a, 20a, 21a, (R = 4-isobutyl);
19b, 20b, 21b, (R = 4-sec-butyl);
19e, 20e, 21e, (R = 3-cyclopentyl);
19h, 20h, 21h, (R = 4-phenyl)
19c, 20c, 21c, (R = 3-sec-butyl);
19f, 20f, 21f, (R = 4-cyclohexyl);
19d, 20d, 21d, (R = 4-cyclopentyl);
19g, 20g, 21g, (R = 4-thiophen-2-yl);
a
(i) DMF, K CO , 100℃. (ii) NaBH CN, EtOH, MeOH, r.t.; (iii) NaOH, MeOH, H O, r.t., then HCl
2
3
3
2
(
2 mol/L).
As shown in Scheme 3, treatment of fluorinated aromatic aldehydes 18a-b with
alkyl mercaptan, naphthenic mercaptan, thiophenol or thiophene-2-thiol gave
9
ACS Paragon Plus Environment

Page 11 of 92
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
sulfur-containing aromatic aldehydes 19a-h, which were used for the preparation of
compounds 21a-h in the same manner as described above.
Scheme 4. Preparation of Compounds 25a-da
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
O
O
HO
O
O
O
HO
OH
B
R
oseltamivir
phosphate
N
H
O
O
iii
N
O
O
ii
i
R
R
H
Br
R
HN
4a-d
HN
22a,b
23a-d
2
25a-d
R =
S
,
,
S
S
a
(
i) DMSO, K CO , Pd(PPh ) , N , 100℃. (ii) NaBH CN, EtOH, MeOH, r.t.; (iii) NaOH, MeOH,
2 3 3 4 2 3
H O, r.t., then HCl (2 mol/L).
2
Scheme 5. Preparation of Compound 28a
O
O
O
HO
O
O
Br
oseltamivir
phosphate
S
N
O
O
iii
N
O
O
i
ii
H
H
HN
HN
S
S
B
HO
OH
S
8
27
28
26
a
(i) DMSO, K CO , Pd(PPh ) , N , 100℃. (ii) NaBH CN, EtOH, MeOH, rt; (iii) NaOH, MeOH, H O,
2
3
3 4
2
3
2
r.t., then HCl (2 mol/L).
As shown in Scheme 4, aromatic aldehydes 23a-d were obtained via Suzuki reaction
of the 3- or 4-bromobenzaldehyde (22a or 22b) with the corresponding arylboronic acids,
(
4-(methylthio)phenyl)boronic
acid,
acid.
benzo[b]thiophen-2-ylboronic
acid,
and
5,
benzo[b]thiophen-3-ylboronic
As
(26)
illustrated
was
in
Scheme
4
4
-(benzo[b]thiophen-6-yl)benzaldehyde
synthesized
from
-formylphenyl)boronic acid (8) and 5-bromobenzo[b]thiophene. The target compounds
25a-d and 28 were obtained from the aromatic aldehydes 23a-d and 26 (Schemes 4 and
) using a similar procedure to that described above.
5
1
0
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 92
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
Scheme 6 illustrates the conversion of methylarenes into bromobenzyls by treatment
with N-bromosuccinimide (NBS) in the presence of a catalytic amount of benzoyl
peroxide (BPO), followed by hydrocarbylation and saponification reactions to give the
target compounds 32a-b.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 6. Preparation of Compounds 32a-ba
Br
O
O
HO
O
oseltamivir
phosphate
X
i
ii
N
O
O
iii
N
O
O
X
H
H
HN
HN
X
X
31a-b
32a-b
29a-b
30a-b
a
X = CO or SO ; (i) NBS, BPO, CCl , reflux; (ii) CH CN, K CO , reflux; (iii) NaOH, MeOH, H O,
2
4
3
2
3
2
r.t., then HCl (2 mol/L).
RESULTS AND DISCUSSION
Anti-Influenza-Virus Activity in Cell Culture
In order to investigate the efficacy of the synthesized compounds against influenza
virus infection, we employed cell-based assays (chicken embryo fibroblasts, CEFs) that
address the CPE of AIV infection using A/goose/Guangdong/SH7/2013 (H5N1) and
A/goose/Jiangsu/1306/2014 (H5N8) as representatives of group-1 NAs-containing
influenza
strains,
and
A/Chicken/Hebei/LZF/2014
(H5N2)
and
A/duck/Guangdong/674/2014 (H5N6) as representatives of group-2 NAs-containing
influenza strains. OSC, ZA, and ribavirin (Rib) were run in parallel as control drugs. The
assays were conducted in triplicate and repeated three times. The values of EC50
(anti-avian influenza A virus potency) and CC50 (cytotoxicity) of the synthesized
compounds are summarized in Table 1. Notably, all the tested compounds exhibited
1
1
ACS Paragon Plus Environment

Page 13 of 92
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
antiviral activity in CEFs. Furthermore, they showed no appreciable cytotoxicity at the
highest tested concentration (CC > 200 μM) in CEFs.
5
0
Table 1. Anti-Influenza-Virus Activity and Cytotoxicity of Oseltamivir Derivatives
in CEF Cells
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
HO
O
N
O
O
R
H
2
HN
4
3
b
EC50 values (μM) against influenza virus
Compounda
R
CC50c
H5N1d
H5N2e
H5N6f
H5N8g
Group-1
Group-2
Group-2
Group-1
>200h
1
1a
11b
1c
1d
7a
17b
7c
7d
7e
7f
1a
1b
21c
1d
1e
1f
21g
1h
4-Et
2.67±0.56
1.10±0.39
5.39±1.48
5.83±1.79
1.27±0.35
31.88±6.23
19.42±0.97
10.61±1.65
51.09±0.22
10.71±1.18
18.09±2.53
6.37±2.35
31.08±4.66
26.63±8.38
24.16±9.12
38.0±7.23
0.25±0.03
0.39±0.08
2.99±0.25
3.97±0.65
6.71±0.1
1.97±0.26
1.24±0.46
3.39±0.62
1.13±0.32
5.65±0.85
3.42±0.79
17.98±5.19
4.92±1.69
17.56±3.71
0.32±0.09
12.96±3.80
1.83±0.85
4.72±1.76
14.82±6.07
6.50±1.36
1.75±0.79
1.21±0.40
0.52±0.09
3.93±0.51
2.48±0.92
6.79±1.24
0.59±0.09
4.73±1.04
>100
4-CH(Me)
4-CH(Me)
2
3
>200
>200
>200
>200
>200
>200
>200
>200
>200
>200
1
1
4-CH
2
Ph
0.25±0.08
0.78±0.07
1.33±0.06
1.37±0.11
1.10±0.15
2.68±0.03
0.28±0.07
0.74±0.14
1.15±0.11
1.29±0.23
5.28±0.92
0.87±0.37
0.68±0.21
0.14±0.03
0.09±0.01
1
4-OMe
3-OMe
1
4-OCH(Me)
2
16.87±3.61
17.9±0.91
3.97±0.96
0.20±0.06
2.59±0.94
1
4-OCH(Et)
3-OCH(Et)
4-OPh
2
1
2
1
2
4-isobutylthio
4-sec-butylthio
3-sec-butylthio
4-cyclopentylthio
3-cyclopentylthio
4-cyclohexylthio
4-thiophen-2-ylthio
4-phenylthio
2
80.13±15.88 >200
10.78±4.46
4.27±1.85
5.75±1.78
0.32±0.09
0.17±0.027
0.056±0.03
>200
>200
>200
>200
>200
>200
2
2
2
2
1
2
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 92
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
2
5a
25b
5c
5d
28
2a
2b
4-(methylthio)phenyl
3-(methylthio)phenyl
4-benzo[b]thiophen-2-yl
4-benzo[b]thiophen-3-yl
4-benzo[b]thiophen-5-yl
4-PhCO
4.10±0.87
35.74±7.32
21.34±3.60
8.97±2.68
16.40±6.48
7.59±1.21
>100
0.51±0.13
0.81±0.13
0.61±0.08
0.64±0.17
0.68±0.21
0.33±0.11
1.46±0.14
0.10±0.01
0.26±0.02
15.81±1.71
2.57±1.09
23.87±8.35
8.04±2.81
3.81±1.06
2.47±0.69
1.16±0.14
1.73±0.87
0.44±0.22
0.31±0.14
5.06±1.11
0.64±0.21
1.31±0.32
2.14±0.24
0.22±0.08
0.31±0.09
1.13±0.45
70.28±8.8
1.99±0.36
5.76±0.48
11.33±0.58
>200
>200
>200
>200
>200
>200
>200
>200
>200
>200
2
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
3
4-PhSO
---
2
OSC
ZA
0.53±0.06
0.51±0.01
11.27±2.56
---
Rib
---
a
Compounds 11a-32b were checked against pan-assay interference compounds (PAINS), and none
was flagged.30,31
b
EC : concentration of a compound required to achieve 50% protection of CEF cultures against
5
0
influenza virus-induced cytotoxicity, presented as the mean ± standard deviation (SD) and determined
by the CCK-8 method.
c
CC : concentration required to reduce the viability of mock-infected cell cultures by 50%, as
5
0
determined by the CCK-8 method.
d
A/goose/Guangdong/SH7/2013. eA/Chicken/Hebei/LZF/2014. fA/duck/Guangdong/674/2014.
gA/goose/Jiangsu/1306/2014.
h
Highest tested concentration.
First, we focused our attention on novel C5-NH -substituted oseltamivir derivatives
2
(
11a-c) in which the phenyl or thienyl group of the 4-benzyl moiety of 5 or 6 was
replaced with relatively small alkyl groups containing 2 to 4 carbon atoms to explore the
effect of steric bulk at this position. We also added a methylene bridge between the two
benzenes of the lead compound 5 to give 11d, aiming to introduce better flexibility to
adapt to the cavity in NAs.
1
3
ACS Paragon Plus Environment

Page 15 of 92
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Among the synthesized compounds, 11b (R = 4-CH(Me) ) exhibited the highest
2
potency against H5N1 (EC = 1.10 μM); it was less potent than OSC (EC = 0.53 μM)
5
0
50
or ZA (EC = 0.51 μM), but its activity was about 10 times higher than that of Rib (EC
50
5
0
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
=
11.27 μM). Replacing the isopropyl (11b) group with ethyl (11a), tert-butyl (11c) or
benzyl (11d) slightly reduced the antiviral activity (11a, EC = 2.67 μM; 11c, EC =
5
0
50
5
.39 μM; 11d, EC = 5.83 μM), but these compounds were still more potent than Rib.
50
The order of potency against H5N1 of compounds with various substituents at the
R-position in the benzene ring was: 4-CH(Me) > 4-Et > 4-CH(Me) > 4-CH Ph.
2
3
2
Interestingly, in contrast to H5N1, 11d with the benzyl substituent turned out to be
the most potent inhibitor in this sub-series against H5N2, H5N6 and H5N8 strains, with
good EC values of 0.25, 1.13 and 0.59 μM, respectively. Although its activities against
5
0
H5N2 and H5N6 were similar to those of OSC (H5N2, EC = 0.10 μM; H5N6, EC =
5
0
50
0.44 μM) and ZA (H5N2, EC = 0.26 μM; H5N6, EC = 0.31 μM), it showed higher
5
0
50
potency against H5N8 as compared to OSC (EC = 1.99 μM) or ZA (EC = 5.76 μM).
5
0
50
In contrast, 11a-c exhibited weaker potency against the H5N2 and H5N6 strains
compared to OSC and ZA. As for H5N8, 11a-c showed weaker potency than OSC, but
comparable efficacy to ZA. Thus, it appears that a moderately flexible aromatic group
(
4-CH Ph in 11d) at the R position increases the potency toward H5N2, H5N6 and H5N8
2
strains.
To extend the SAR information, we inserted an oxygen atom between the alkyl
group (or benzene) and benzyl group (compounds 17a-f). However, 17b-f displayed
reduced anti-H5N1 activity (EC = 10.61-51.09 μM) compared to the above sub-series.
5
0
Moreover, the meta-substituted compounds (17b, R = 3-OMe, EC =31.88 μM; 17e, R =
5
0
1
4
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 92
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
3
-OCH(Et) , EC = 51.09 μM) showed considerably decreased potency against H5N1
2
50
compared with their para-substituted counterparts (17a, R = 4-OMe, EC =1.27 μM;
5
0
17d, R = 4-OCH(Et) , EC =10.61 μM).
2
50
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
Notably, 17f (R = 4-OPh), bearing an aromatic phenoxy group, was the most potent
compound among 17a-f against H5N2, H5N6, and H5N8 strains (H5N2, EC = 0.28
5
0
μM; H5N6, EC50 =0.32 μM; H5N8, EC50 =0.20 μM). Moreover, 17f displayed
comparable activity to OSC and ZA against H5N2 and H5N6, and showed greater
potency than OSC and ZA against H5N8. Compared to 11d, the replacement of carbon
by oxygen led to a marked increase of potency of 17f against H5N6 and H5N8.
Inspired by those results, we next employed a bioisosterism strategy to design
alkylsulfide and arylsulfide derivatives, with the aim of improving the antiviral potency,
including against the N1-H274Y mutant. Specifically, we replaced the oxygen of 17f
with a sulfur atom, which has a larger radius and different bond angles, to afford
compound 21h as a derivative of lead 5. Similarly, a sulfur atom was inserted between
the two aryls of compound 6 to give 21g. More diverse alkyl groups were also used in
order to extend the SAR study.
However, compounds 21a-f with alkyl sulfide or cycloalkyl sulfide groups were less
potent than OSC and ZA against H5N1, with EC values from 6.37 to 38.0 μM. Among
5
0
them, only 21b (EC = 6.37 μM) with a 4-sec-butylthio group was more potent than Rib.
50
But encouragingly, compounds 21g and 21h exhibited considerably enhanced anti-H5N1
activities (21g, EC = 0.25 μM; 21h, EC = 0.39 μM), which were even higher than
5
0
50
OSC and ZA. On the other hand, compounds 21d and 21f almost completely lost activity,
although they have a hydrophobic 4-cycloalkyl moiety similar in size to the substituents
1
5
ACS Paragon Plus Environment

Page 17 of 92
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
of 21g and 21h. This result confirms that the aromatic moieties play a critical role in the
interaction with the binding pocket. It is also noteworthy that 21h is about 13 or 27 times
more potent than the C- and O- counterparts 11d and 17f, respectively. The order of
potency seems to be in accordance with that of bond length: S-C (~180 pm) > C-C (~150
pm) > O-C (~140 pm). Thus, the larger size of 4-(phenylthio)benzyl, which may better
occupy the 150-cavity in group-1 NAs, is essential for strong antiviral potency against
H5N1.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
As for inhibitory activity against H5N2 and H5N6, 21a-f also showed lower activity
than OSC and ZA, with EC values ranging from 0.68 to 5.28 μM and 1.75 to 14.82 μM,
50
respectively. Importantly, 21g and 21h exhibited excellent potency against H5N2, with
EC values of 0.14 and 0.09 μM, respectively, which are similar to those of OSC and ZA
5
0
and more than 100 times higher than that of Rib. In addition, 21h (EC = 0.52 μM)
5
0
displayed comparable activity to OSC and ZA against H5N6, though 21g was slightly
less potent (EC = 1.21 μM) than OSC and ZA against H5N6.
5
0
Against H5N8, which is relatively insensitive to all three of the control drugs, 21a-e
showed only moderate to low activity (EC = 2.59-80.13 μM), while compound 21f
50
(
EC = 0.32 μM) with the 4-cyclohexyl group was about five times more potent than
50
OSC. Further, 21g and 21h displayed superior activity against H5N8, with EC values of
5
0
0
.17 to 0.056 μM, being far more potent than OSC or ZA.
Inspired by the importance of the sulfur atom in this series, we designed five more
sulfur-containing compounds based on the structures of 5 and 6.
Introduction of a 4-methylthiophenyl group into R-position yielded 25a (R =
4-(4-methylthiophenyl)) and 25b (R = 3-(4-methylthiophenyl)), respectively. In addition,
1
6
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 92
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
we replaced the 4-phenyl group of compound 5 with three different fused
4
-benzo[b]thiophenyls. Generally, the resulting five compounds (25a-d and 28) were
much less potent than OSC against H5N1, H5N2 and H5N6 strains. However, against
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
H5N8, most of them (25a, EC = 0.64 μM; 25b, EC = 1.31 μM; 25d, EC = 0.22 μM;
5
0
50
50
and 28, EC = 0.31 μM) were more potent than all of the control drugs.
5
0
To further explore the SAR of the linker between the two aromatic rings, we also
examined the use of carbonyl and sulfonyl (32a-b) to replace the sulfur atom in 21h.
However, both compounds, especially 32b, showed greatly reduced antiviral activity
toward all four strains.
In general, compounds bearing the aromatic phenyl group displayed greater
activities towards H5N2, H5N6, and H5N8, as compared with the alkyl derivatives.
However, the anti-H5N1 activities of the compounds were more structure-sensitive.
Among all the compounds, 21h showed excellent activity against the whole viral panel
(
H5N1, H5N2, H5N6, and H5N8). Its potency is comparable to or slightly better than that
of the best control drug, OSC, against H5N1, H5N2 and H5N6. Even more remarkably,
1h potently inhibited H5N8, being 35 and 103 times more active than OSC and ZA,
2
respectively. Thus, we obtained a very promising candidate for further evaluation.
Then, we tested compounds 17a, 17d, 21f, 21g, 21h and 25c against influenza virus
A/PR/8/34 (H1N1), A/Wisconsin/67/05 (H3N2), and B/Lee/40 strains by plaque
reduction assays (PRA) in MDCK cells. OSC and ZA were used as positive controls for
inhibition. We also tested the cytotoxicity of the compounds, as well as OSC and ZA as a
reference, in MDCK cells by MTT assays. No cytotoxic effect was observed at the tested
concentrations for any of the compounds. The values of EC (anti-influenza potency) of
5
0
1
7
ACS Paragon Plus Environment

Page 19 of 92
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
the selected oseltamivir derivatives and the positive control drugs are summarized in
Table 2.
In the anti-influenza activities evaluation against H1N1 and H3N2 strains, 17a, 17d,
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
1f, 21g, 21h and 25c turned out to be potent inhibitors against H1N1 with EC values
50
from 0.041 to 17.07 μM. Among them, 21g and 21h (21g, EC = 0.04 μM; 21h, EC =
5
0
50
0
.06 μM) were the two most potent than other tested compounds, nearly equal to that of
OSC and ZA. But all the tested compounds exhibited weaker potency against the H3N2
and FluB strains compared to OSC and ZA, which can be easily explained by the
structural features of N2 and FluB NA. The closed 150-loop and the lack of 150-cavity in
NA of H3N2 and FluB (Figure 2) indicate that the cavity would not well accommodate
the selected oseltamivir derivatives bearing large groups at C-5-NH2 position. These
results agreed with the data of the enzymatic assay against N1 and N2.
Table 2. Anti-Influenza-Virus Activity and Cytotoxicity of Oseltamivir Derivatives
in MDCK Cells
EC50a values (μM) against influenza
CC50b
virus
Compound
R
H1N1c
Group-1
3.7 ± 0.8
13.0 ± 0.5
17.0 ± 3.0
H3N2d
Group-2
>100
FluBe
>250f
1
7a
17d
1f
4-OMe
>100
>100
>100
4-OCH(Et)2
>100
>250
>250
2
4-cyclohexylthio
>100
1
8
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 92
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
2
1g
4-thiophen-2-ylthio
0.04 ± 0.02
0.06 ± 0.02
7.2 ± 2.9
79 ± 5
71 ± 3
8.5 ± 1.1
19.0 ± 2.0
>100
>250
>250
>250
>250
>250
21h
25c
4-phenylthio
4-benzo[b]thiophen-2-yl
>100
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
OSC
ZA
---
---
0.041 ± 0.003
0.038 ± 0.005
1.2 ± 0.1
0.3 ± 0.1
2.9 ± 0.3
0.9 ± 0.2
a
EC : concentration of compound required to achieve 50% protection of MDCK cell
5
0
cultures against influenza virus-plaque formation, presented as the mean ± standard
deviation (SD) and determined by the PRA method.
b
CC : concentration required to reduce the viability of mock-infected cell cultures by
5
0
5
0%, as determined by the MTT method.
c
d
e
f
A/PR/8/34. A/Wisconsin/67/05. B/Lee/40. Highest tested concentration.
Inhibition of Influenza Neuraminidases
Eleven representative compounds were selected for enzymatic assay against
influenza neuraminidases (H5N1, H5N2, H5N6, H5N8 and H5N1-H274Y) to validate
their binding target. OSC and ZA were also run in parallel as control drugs. The results
are shown in Table 3. OSC showed very good inhibitory activity against the four
wild-type NAs, with IC values of 26.63, 4.85, 15.25 and 8.94 nM, respectively. ZA also
5
0
displayed excellent potency against all the wild-type NAs; it still showed an IC of 22.8
5
0
nM even against the least sensitive NA, i.e., H5N6. Of course, OSC and ZA did not
equally inhibit these NA subtypes, but showed some selectivity, in accordance with the
reported data 17,32-36 and these findings served to verify the reliability of our assay.
N1 (H5N1) and N8 (H5N8) belong to the group-1 NAs, and most of the inhibitors
1
9
ACS Paragon Plus Environment

Page 21 of 92
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
showed relatively similar potency against both subtypes. In particular, 11a, 11d, 17a, 17f,
2
1g and 21h proved to be very potent inhibitors of N1 and N8 with most of the IC s at or
50
below the two-digital nanomolar range. Notably, the best compound 21h showed
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
excellent inhibitory activity with IC values of 0.96 and 1.89 nM against N1 and N8,
5
0
respectively, being 27 and 4 times more potent than OSC, and 7 and 3 times more potent
than ZA, respectively.
We also performed enzymatic assay against N2 (H5N2) and N6 (H5N6). As
expected, all of the compounds tested were significantly less active against the N2 and
N6 subtypes compared to the N1 subtype, because of the closed 150-loop and absent
150-cavity in N2 and N6. The poor inhibitory activities suggested that N2 and N6 do not
well accommodate compounds bearing a large group at the C5-NH position. In other
2
words, the conformations of these oseltamivir derivatives might be distorted in N2 and
N6, resulting in unfavorable interactions with the binding pockets of these subtypes.
These results further support our hypothesis and design strategy.
In agreement with previous reports,17,28 OSC showed >100-fold weaker inhibitory
activity against N1-H274Y (A/Anhui/1/2005 (H5N1-H274Y), IC = 2824 nM) compared
50
to that against wild-type N1 (A/goose/Guangdong/SH7/2013 (H5N1), IC = 26.63 nM),
5
0
whereas ZA was equally effective against N1-H274Y and wild-type N1 (A/Anhui/1/2005
(
H5N1-H274Y), IC = 7.77 nM; A/goose/Guangdong/SH7/2013 (H5N1), IC = 7.67
5
0
5
0
nM).
Apart from 21g and 21h, all the tested compounds showed weak inhibitory activity
against N1-H274Y mutation, with IC50 values of more than 363 nM. But, notably,
compounds 21g and 21h showed outstanding inhibition of N1-H274Y (IC = 36.34 and
5
0
2
0
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 92
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
3
2.81 nM, respectively), being better than OSC and comparable to ZA in potency. Thus,
we have successfully overcome a major disadvantage of OSC and the lead compounds 5
and 6, i.e., their weak inhibitory activity against the most important clinical mutant.
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
1
ACS Paragon Plus Environment

Page 23 of 92
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 3. Neuraminidase (NA) Inhibition of Oseltamivir Derivatives in Chemiluminescence-based Assay
HO
O
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
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
N
O
O
R
H
2
HN
4
3
a
NA-Inhibitory Activity, IC (nM)
5
0
Compd
R
H5N1b
Group-1
63.89±8.75
H5N2c
H5N6d
Group-2
H5N8e
Group-1
41.78±8.37
H5N1-H274Yf
Group-1
Group-2
1
1a
4-Et
965.83±42.03
7250.67±563.99
727.23±86.95
1
046.03±78.9
7
1
1d
4-CH Ph
63.87±1.24
20.45±1.61
6890.33±554.39
18105.33±1261.62
10850±245.76
56.36±3.39
131.63±20.47
195.07±8.13
1143±124.21
41.96±5.36
915.67±104.16
3147.33±193.75
30026.67±3423.76
>100000
2
1
7a
4-OMe
3-OMe
1649±123.41
112.87±184.
1
1
1
7b
7d
266.33±27.10
1050.37±134.75
37.31±0.43
7
9
4-OCH(Et)2
4-OPh
1243±120.58
595.67±104.
3228±583.94
1
1
2
7f
1f
4250.33±113.47
2099.33±131.25
363.87±12.15
>100000
0
4
4-cyclohexylthio
2023.67±232.73
398.5±66.39
951.63±79.47
2
2
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 92
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
2
1g
1h
5a
4-thiophen-2-ylthio
3.83±0.38
0.96±0.15
1154.5±65.76
896.9±74.104
275±24.16
5681±494.14
3.47±0.34
1.89±0.23
36.34±5.62
32.81±2.11
2
4-phenylthio
3065±217.73
3045±59.174
2
4-(methylthio)phenyl
1795.33±318.63
592.6±61.31
82186.67±10941.01
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
4
-benzo[b]thiophen-2-y
2
5c
1179.33±188.99
541.27±51.06
4814.67±975.02
466.2±70.29
27003.33±1909.62
l
OSC
ZA
---
---
26.63±5.12
7.77±0.5
4.85±0.46
15.25±1.73
22.80±1.08
8.94±0.11
7.43±0.62
2824.67±56.05
7.67±1.0
10.04±0.74
a
Concentration required to reduce NA activity to 50% of control NA activity (IC ). Values are the mean of three experiments, presented as the mean ±
5
0
b
c
d
e
f
standard deviation (SD). A/goose/Guangdong/SH7/2013. A/Chicken/Hebei/LZF/2014. A/duck/Guangdong/674/2014. A/goose/Jiangsu/1306/2014.
A/Anhui/1/2005.
2
3
ACS Paragon Plus Environment

Page 25 of 92
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
We also tested compounds 21g and 21h in an enzymatic assay against NA of
H1N1pdm09 (09N1), as shown in Table 4. OSC and ZA were quite effective against
09N1, with IC values of 5.20 and 5.28 nM, respectively, in accordance with reported
5
0
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
data.28,36 Surprisingly, 21g and 21h displayed robust activity against 09N1, with IC50
values of 7.05 and 5.57 nM, respectively, which were similar to those of the control
drugs. The excellent inhibition to 09N1 by 21g and 21h can be easily explained on the
basis of previous reports24 showing that 09N1 exists preferentially with an open
1
50-cavity in solution, though it crystallizes without the 150-cavity that is characteristic
of group-1 NAs.
Table 4. 09N1 (H1N1pdm09)-Inhibitory Activities of Compounds 21g and 21ha
Compounds
21g
21h
OSC
ZA
IC (nM): 09N1
5
0
7
.05±0.51
5.57±0.22
5.20±0.26
5.28±0.79
(
H1N1pdm09)b
a
b
Values are the mean of three independent experiments. A/California/04/2009.
In conclusion, like the lead compounds 5 and 6, the representative derivatives
synthesized in this work displayed high selectivity for N1 and N8 over N2 and N6.
Notably, 21g and 21h were superior to OSC and ZA against N1 and N8. Furthermore,
against the NA of H5N1-H274Y, they displayed comparable activity to ZA, exhibiting
1
7
higher potency than leads 5 and 6, and were 77 and 86 times more potent than OSC. It
is interesting that they also exhibited significant inhibitory activity toward 09N1.
It is noteworthy that, relative to the NAs (N2, N6, and N8)-inhibitory activities,
some compounds showed unexpectedly improved anti-influenza activities in cell-based
assay against H5N2, H5N6, and H5N8. In other words, although the enzyme-inhibitory
(N2, N6, and N8) activities of some compounds were worse or slightly better than that of
2
4
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 92
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
OSC and ZA, the anti-influenza activities of them were comparable or far greater than
that of OSC and ZA. Specifically, 11d, 17f, and 21g-h showed remarkably elevated
anti-influenza activities in comparison with their anti-NAs activities against H5N2 and
H5N6 viruses (e.g. 21h and OSC towards N2-NA/H5N2, mean IC values of 896.9 nM
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
(
21h) and 4.85 nM (OSC) in enzymatic assay versus EC values of 90 nM (21h) and 100
50
nM (OSC) in cell-based assay), A similar situation was found in the other two sets (for
7f, and 21h towards N6-NA/H5N6 strain; for 11d, 17f, 21f-h, 25d and 28 towards
1
N8-NA/H5N8 strain).
These findings suggest that, in the cellular environment, some mechanism or effects
(such as inhibition of other targets, protein-protein interactions, or asymmetric
intracellular localization of compounds33,37) other than NA inhibition may be involved in
the anti-influenza virus (H5N2, H5N6, and H5N8) action. There may also be other
uncertainties, for example, host range restriction of influenza virus replication in
CEFs.38-41
In Vivo Anti-influenza Virus Activity in Specific Pathogen Free (SPF) Chicken
Embryonated Egg
In further antiviral research, we tested anti-influenza virus activity of compound 21h
with an chicken embryonated eggs (10-11 days) model which was regarded to have no
4
2
conflict with ethical and legal aspects of animal protection. We infected the chicken
embryonated eggs with H5N2 virus strain and treated them with injecting different
concentrations of 21h. Also, OSC was selected as control drug in parallel. At 48 h and 72
h post infection and treatment, we recorded the number of the survival and death to
evaluate the anti-influenza virus activity of the compounds. As shown in Table 5
2
5
ACS Paragon Plus Environment

Page 27 of 92
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(
Experiment and figure data were shown in Supporting Information), all embryos were
survived within only infected normal saline (survived/dead = 5/0), and in other groups all
were dead after infection without any treatment within 48 h (H5N2: survived/dead = 0/5),
suggesting the test procedure was standard. After 48 h, all embryos were survived in
curing with 21h in concentration of 10 mM, while 3 embryos were survived in curing
with OSC in the same concentration. By decreasing the concentration of 21h and OSC,
the number of death was raised in both experiments, but the results clearly showed that
the ability of 21h to protect chicken embryos was stronger than that of OSC at the same
concentration.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Table 5. Survival number of chicken embryos after inoculation of specific pathogen
free (SPF) embryonated eggs with H5N2 and administration of 21h or OSC
Therapeutic
Chick Embryos Survived/Dead Number
(survival rate)
Compounds
Concentration
(
mM) a
48 h
72 h
1
0
5/0 (100%)
4/1 (80%)
3/2 (60%)
1/4 (0%)
0/5 (0%)
3/2 (60%)
1/4 (20%)
0/5 (0%)
1/4 (0%)
0/5 (0%)
5/0 (100%)
3/2 (60%)
1/4 (20%)
0/5 (0%)
0/5 (0%)
0/5 (0%)
0/5 (0%)
0/5 (0%)
0/5 (0%)
0/5 (0%)
0/5 (0%)
5/0 (100%)
2
.5
2
1h
0.625
0.156
0.039
10
2.5
OSC
0.625
0.156
0.039
---
Normal aline control
2
6
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 92
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
virus (H5N2) control
---
0/5 (0%)
0/5 (0%)
a
The virus solution was mixed with an equal volume of tested compounds and then incubated for
1
h before inoculation.
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
Molecular Modeling
To further investigate the binding mode of the most promising compound 21h with
respect to that of the reference compound OSC in the 150-cavity of various NA
structures, we built an open 150-loop conformation of the N8, 09N1, and N1-H247Y
structures for molecular docking and MD-based free energy calculations.
Molecular dynamics analysis shows that all the protein-ligand complexes are stable
at finite temperature and 1 atm pressure. The RMSD plots (see Supporting Information,
Figure S1) for 21h and OSC indicate that these ligands are localized in the binding site
during the entire course of molecular dynamics (MD) simulation. The RMSF was also
computed for each complex (see Supporting Information, Figure S2) and the general
behavior seen for all neuraminidases is that RMSF was significantly reduced for 21h as
compared with OSC. Such overall low flexibility of 21h is due to the fact that 21h can
bind effectively to both the 150-cavity and active site as compared to OSC, which binds
only at the active site.
To further understand the key interactions between ligand and protein, we conducted
MM-GBSA (Molecular Mechanics-Generalized Born combined with solvent accessible
Surface Area) binding affinity calculations. A summary of the MM-GBSA output is
provided in the Supporting Information ST1. As can be seen in the cases of subtypes N1,
0
9N1 and N8, compound 21h showed lower binding free energy (-23.81 to -38.37
kcal/mol) than OSC (-16.05 to -26.48 kcal/mol).
2
7
ACS Paragon Plus Environment

Page 29 of 92
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
When we examine the various contributing energy terms, it is evident that the
increase in van der Waals (vdW) interaction between 21h and NAs predominates over
electrostatic interaction. As can be seen in Figure 5, compound 21h shows key
interactions in both the active site and the 150-cavity, as we expected (a schematic
representation of the protein-ligand interaction of 21h is provided in Supporting
information, Figure S3). These key interactions include the formation of a salt-bridge
between the carboxylate group of 21h and Arg residue, as observed in the X-ray crystal
structures of NAs bound to OSC. In addition, 21h forms a strong hydrogen-bonding
network between the amino-linker and anionic residues (Asp70 in N1 and N1-H247Y,
Asp71 in 09N1, Glu38 in N8), which is not observed in OSC. Furthermore, the
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
-(phenylthio)benzyl group of 21h is well adapted to the 150-cavity of N1, N8, 09N1 and
N1-H247Y, in which the 150-cavity is surrounded mostly by polar and non-polar
residues. Overall, the increased binding affinity of 21h is primarily due to the better
shape complementarity with the 150-cavity, as reflected by the van der Waals interaction
energies from MM-GBSA calculation. However, electrostatic interactions do play a role
in binding of the common OSC fragment in the active site of NAs.
2
8
ACS Paragon Plus Environment