Design, synthesis and biological evaluation of novel oseltamivir derivatives as potent neuraminidase inhibitors

Article abstract of DOI:10.1016/j.bmcl.2017.11.003

Neuraminidase (NA) is one of the particular potential targets for novel antiviral therapy. In this work, a series of neuraminidase inhibitors with the cyclohexene scaffold were studied based upon the combination of 3D-QSAR, molecular docking, and molecular dynamics techniques. The results indicate that the built 3D-QSAR models yield reliable statistical information: the correlation coefficient (r²) and cross-validation coefficient (q²) of CoMFA (comparative molecular field analysis) are 0.992 and 0.819; the r² and q² of CoMSIA (comparative molecular similarity analysis) are 0.992 and 0.863, respectively. Molecular docking and MD simulations were conducted to confirm the detailed binding mode of enzyme-inhibitor system. The new NA inhibitors had been designed, synthesized, and their inhibitory activities against group-1 neuraminidase were determined. One agent displayed excellent neuraminidase inhibition, with IC₅₀ value of 39.6 μM against NA, while IC₅₀ value for oseltamivir is 61.1 μM. This compound may be further investigated for the treatment of infection by the new type influenza virus.

Full text of DOI:10.1016/j.bmcl.2017.11.003

Accepted Manuscript
Design, synthesis and biological evaluation of novel oseltamivir derivatives as
potent neuraminidase inhibitors
Zhen Wang, Li Ping Cheng, Xing Hua Zhang, Wan Pang, Liang Li, Jin Long
Zhao
PII:
S0960-894X(17)31079-X
https://doi.org/10.1016/j.bmcl.2017.11.003
BMCL 25409
DOI:
Reference:
To appear in:
Bioorganic & Medicinal Chemistry Letters
Received Date:
Revised Date:
Accepted Date:
6 August 2017
27 October 2017
2 November 2017
Please cite this article as: Wang, Z., Cheng, L.P., Zhang, X.H., Pang, W., Li, L., Zhao, J.L., Design, synthesis and
biological evaluation of novel oseltamivir derivatives as potent neuraminidase inhibitors, Bioorganic & Medicinal
Chemistry Letters (2017), doi: https://doi.org/10.1016/j.bmcl.2017.11.003
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com
Design, synthesis and biological evaluation of novel oseltamivir derivatives
as potent neuraminidase inhibitors
a,
Zhen Wang, Li Ping Cheng *, Xing Hua Zhang, Wan Pang, Liang Li, Jin Long Zhao
a
School of Chemical and Environmental Engineering, Shanghai Institute of Technology, Shanghai 201418, China
A RTICLE INFO
ABSTRACT
Article history:
Neuraminidase (NA) is one of the particular potential targets for novel antiviral therapy. In this work,
A series of neuraminidase inhibitors with the cyclohexene scaffold were studied based upon the
combination of 3D-QSAR, molecular docking, and molecular dynamics techniques. The results
indicate that the built 3D-QSAR models yield reliable statistical information: the correlation
Received 6 Augest 2017
Revised 26 September 2017
Accepted
2
2
Available online
coefficient (r ) and cross-validation coefficient (q ) of CoMFA (comparative molecular field analysis)
2 2
are 0.992 and 0.819; the r and q of CoMSIA (comparative molecular similarity analysis) are 0.992
and 0.863, respectively. Molecular docking and MD simulations were conducted to confirm the
detailed binding mode of enzyme-inhibitor system. The new NA inhibitors had been designed,
synthesized, and their inhibitory activities against group-1 neuraminidase were determined. One
agent displayed excellent neuraminidase inhibition, with IC50 value of 39.6 μM against NA, while IC50
value for oseltamivir is 61.1 μM. This compound may be further investigated for the treatment of
infection by the new type influenza virus.
Keywords:
Neuraminidase inhibitors
3D-QSAR
Molecular docking
Molecular dynamics simulations
Oseltamivir derivatives
2
009 Elsevier Ltd. All rights reserved
1
4-15
As worldwide epidemic strains, influenza can cause acute
respiratory diseases and annual influenza pandemic. The swine-
antiviral drugs in Japan.
Nevertheless, variant strains that are
against these drugs are also continuously been discovered.
origin H1N1 influenza virus unleashed a severe public panic in
1
Recent papers have reported some progresses in NA and its
inhibitors in theoretical and experimental studies. For instance,
Chen et al. discovered a series of crenatoside analogues as
2
009. It is reported that highly pathogenic influenza virus H5N1
2
,3
can lead to an alarming life threatening (60% mortality), and
recent studies have raised a major concern on possible
transmission of avian influenza virus H5N1 to humans. A lethal
avian influenza H7N9 has evolved to influence human since
16
novel influenza neuraminidase inhibitors, identified by
17
biological evaluation. Hoffmann et al. developed a platform for
determining the inhibition profile of NAIs in the N1 background.
2
013, endemic to China.
18
In addition, Xie et al. had designed and synthesized two series
Neuraminidase (NA),
playing a dominant role in the viral life cycle. The phylogenetic
tree divides all neuraminidase subtypes into two groups, there
a
membrane-bound glycoprotein,
of oseltamivir derivatives. Of these, two compounds obtained by
introduction of biphenyl substituents into the C-5 NH position
2
exhibited comparable or even better inhibitory activities relative
to the OS carboxylate against N1.
4
5
exists a large cavity (termed 150-cavity) in group-1 NAs,
adjacent to the active site of proteins, whereas that is not found
6
In this work, 65 oseltamivir analogs were selected as
in group-2 NAs. Neuraminidase inhibitors (NAIs) had been
18,19
dataset.
Some studies were employed using computational
regarded as the most potent agents to cure influenza disease. Up
to now, four NA inhibitors have been developed as efficacious
treatment of influenza infections: oseltamivir (1), zanamivir (2),
methods including 3D-QSAR, molecular docking and MD
simulations. The results of CoMFA and CoMSIA supply insights
7
into key structural factors required for the neuraminidase
peramivir (3), and laninamivir (4) (Figure 1). Among these,
20,21
inhibition.
Molecular docking and MD simulations were
orally administered oseltamivir (OS) has been posed to treat
broader cases since its approval in 1999, while recently, several
reported oseltamivir-resistant strains have seriously limited the
performed to validate the 3D-QSAR models and quantify the
22-27
interactions of NA-ligand.
biphenyl substituents introduced into C5-NH
28-30
Previous studies showed that the
8
-12
2
of ligand could
In terms of
drug’s clinical application.
Zanamivir is primarily
improve compound’s biological activity.
administered by inhalation via a nebulizer or intravenous
1
3
substituent’s structure, the chemical bonds between the biphenyl
groups can be freely rotated, this increases the opportunity to
interact with neighbouring amino acid residues. Accordingly,
channels. Peramivir has been authorized in the Republic of
Korea and China, and laninamivir have to date been approved as
*
Corresponding authors. Tel.: +86 21 60873250; fax: +86 21 60 877231 (L.C.).
E-mail address: chengliping@sit.edu.cn (L. Cheng).

some new oseltamivir derivatives were designed by introduction
of biphenyl substitutes into the C5 NH position based on the
built 3D-QSAR contour maps. Finally, the new designed
molecules were further confirmed by experimental synthesis and
biological evaluation.
predicted activity of the best optimal CoMFA and CoMSIA
model is described, the residual values of drugs between the
predicted and actual pIC50 exceeded 1 logarithm unit is treated as
2
32
outliers (compounds 56 and 60). Actually, when the classical
2
internal validation parameter q is greater than 0.5, a model is
33,34
reassuring and predictive.
Determined by leave-one-out
validation, the optimal principal components of CoMFA and
CoMSIA are 8 and 9, respectively. The F-test values are 700.919
and 607.026 for CoMFA and CoMSIA, respectively, a model is
statistically significant with high F-test value. The allocation
proportion of steric and electrostatic fields is 0.450 and 0.550,
respectively. For the chosen CoMSIA model, the relative
contribution of steric, electrostatic, hydrophobic, H-bond donor
and acceptor fields are 0.129, 0.277, 0.286, 0.199 and 0.106,
respectively. The paramount contribution in the model is
hydrophobic interaction.
Figure 1. Structures of the approved NA inhibitors 1-4.
The 65 oseltamivir derivatives in this paper randomly
divided into 55 training set compounds (84.6%) for model
generation and 10 test set compounds (15.4%) for model
validation. The chemical structures, active values against
neuraminidase subtype
Information (Table S1). To build more reliable and statistical
D-QSAR models, we corrected the active data originating in
1
are listed in Supplementary
Figure 2. Superimposition of the training set and compound 25 used as a
template for alignment. Molecules are colored in white for common C,
cyan for H, red for O, blue for N, yellow for S, green for F, Cl & Br,
respectively. The common substructure is cyclohexene (shown in bold),
and other important substituents involving A, B, H-bond formation
substituent and hydrophobic groups, shown in dashed orange frame.
3
two independent documents to eliminate errors. The specific
method is as follows: the IC50 active values were converted into
the corresponding pIC50 (-logIC50) values, then the pIC50 values
were transformed into superior pIC50 values (pIC50c), used as
dependent variables to build molecular models. pIC50c values
3
1
were obtained according to the following formula :
pIC50c=pIC50/pIC50(OS-C)
(1)
Where pIC50, pIC50(OS-C) represent the experimental
bioactive data of the data set compounds and oseltamivir
carboxylate (OS-C), respectively.
Table 1. Statistical results and relative contributions
of the models
a
Figure 3. Scatter plot of experimental pIC50c versus activity predicted
by the optimal CoMFA (a) and CoMSIA (b) model.
Methods
2
CoMFA
CoMSIA
q
Noc
0.819
8
0.863
9
2
r
0.992
0.013
700.919
0.992
0.014
607.026
For convenient analysis, N-substituted neuraminidase
inhibitors were separated into four regions: H-bond formation in
C1 position; C3- ethylpropoxy; C4-acylamino; hydrophobic
group in C5 position. The 3D contour maps of CoMFA and
CoMSIA models are explained in Figure 4~6. From these, we
can interpret the key structural features required for inhibitory
activity.
SEE
F-test
Relative
contributions
Steric
Electrostatic
H-donor
0.450
0.550
—
0.129
0.280
0.199
0.106
0.286
Figure 4a and 4b represent the steric contour maps of CoMFA
and CoMSIA, respectively. Steric favored in green and the
yellow body indicated negative spatial factor. The default value
of field favored and disfavored proportion are 80% and 20%
level contribution. It could be observed that the inhibitory
activity of compound 09 is lower than that of compound 07 from
Supplementary Information (Table S1), because more potent
compound 07 has larger functional groups at the C5 region, the
same case with compounds 08 and 10. A large green contour
may be favorable to cover the benzene groups, for example,
molecules 25, 26, 29 and 31 show excellent inhibitory activities
H-acceptor
Hydrophobic
—
—
a
pIC50c as the
independent variable.
In this study, the ligand-based alignment rule was used. The
alignment result is shown in Figure 2. We performed regression
analysis using partial least squares (PLS) to evaluate and judge
merits or demerits of the model with some statistical parameters,
including r , q , standard error of estimate (SEE) and F-test
values. The best optimal value of CoMFA and CoMSIA models
are listed in Table 1. From this, we observed both models show
satisfactory predictive capability, with q , r , and SEE are 0.819,
.992, 0.013 for CoMFA model, and 0.863, 0.992, 0.014 for
CoMSIA model. In Figure 3, the linear fit between actual and
2
2
2
2
with such substituents at this position. However, the C5-NH
position with too large group would have adverse effect on
2
0

bioactivity, such as compounds 25 and 30, which indicates that
the appropriate bulky groups are beneficial to inhibitory activity.
In addition, if the C3 position is replaced with -NH group,
2
molecules would possess weaker inhibitory ability even there
exists bulky groups, such as compounds 49 and 53-55. From
Figure 4b, it could be inspected that the CoMSIA contour maps
hold similar modes with that of CoMFA.
Figure 6. Hydrophobic contours of CoMSIA model. The yellow shows
the favored hydrophobic area and the white color shows the disfavored
hydrophobic area.
Docking results show that the highest total-score of
Figure 4. Steric contour maps of CoMFA (a) and CoMSIA (b). The
green color showsthe favored steric area and the yellow color shows
disfavored steric area.
compound 25 is 11.35, the total-score of docking studies are
shown in Table S2 (Supplementary Information). Based on the
contour maps obtained from 3D-QSAR models and the structural
features of the most active molecule 25, some new compounds
From Figure 5, the red (electronegative groups favored) and
blue (electropositive favored) contours also represent default
level contributions. Figure 5a shows that there are three small-
sized red contours near the –COO-, a medium-sized red contour
around the C3-methoxypentane, suggesting that the
electronegative groups are favored to bioactivity. One piece of
medium-sized red contour at the biphenyl group indicates the
importance of electronegative substituents in imparting better
biological activity. Therefore, compound 30 acts decreased
inhibitory activity compared to compound 29, and it is thus
feasible to introduce advisable electronegative groups to the C5-
3
-7 were designed by introduction of various substituents into
the biphenyl moiety at the C5-NH position. The docking scores
2
and the predicted pIC50 and pIC50c values of 3-7 based on
CoMFA and CoMSIA models are listed in Table S3
Supplementary Information). Table S3 shows that the predicted
(
and docking values of all designed molecules are equally high
with those of oseltamivir, suggesting their good inhibitory
activity and potent binding affinity with receptor. Among these
new compounds, 3 has the highest docking score (11.37) and
inhibitory activity predicted by both CoMFA and CoMSIA
models, it was subsequently synthesized and evaluated for NA
inhibition. Compound 4 was designed based on the molecular
structure of oseltamivir, it dose not possess the highest predictive
value.
2
NH substituent to obtain high active molecules. One could
observe that the compounds with too strong electronegative
groups would lead to a decreased biological activity. For
example, the inhibitory capacity of compound 23 is inferior to
that of compound 22, because the strong electronegativity of F
atom slightly inactivates the biphenyl-correlated drugs. Figure 5b
represents the CoMSIA electrostatic contour maps. Compared
with the CoMFA contour maps, there is only one small-sized red
contour around –COO-. A medium-sized blue contour surrounds
To better understand the discrepancy of activity between
compounds oseltamivir, 3 and 4 we compared their 3D binding
modes with NA derived from docking simulation. As shown in
Figure 7a, although oseltamivir (in purple) could occupy the
active site of NA, it is powerless to bond well with the 150-
cavity. Compound 3 (in green) had the ability to perfectly dock
into the interaction site of NA, and the biphenyl moiety of 3
appropriately embedded within the 150-cavity, which is the
crucial binding site of group-1 NA. Nevertheless, compound 4
–
NHAc, showing that the electropositive substituents are
beneficial to compounds inhibitory activity.
(in cyan) was out of the active site and the corresponding
biphenyl moiety could not occupy the 150-cavity. In addition,
from previous study, we found that binding interaction of the
biphenyl group with enzyme was hydrophobic. Figure 7b clearly
shows that most amino acids near this group are hydrophobic,
such as Trp178, Val149, and Arg224. The side chains of Arg292,
Arg371, and Glu276 also contribute significantly to the
hydrophobicity of the 150-cavity.
Figure 5. Electrostatic contour maps of CoMFA (a) and CoMSIA (b).
Electrostatic fields: favored (red) and disfavored (blue).
As shown in Figure 6, the yellow pieces denote that
hydrophobic groups are beneficial to inhibitors biological
activity and the white regions reveal that the hydrophilic
substituents may be tolerated. The C5-NH
is close to NA active pocket. One large yellow contour
surrounding C5-NH suggests that the existence of hydrophobic
2
group of oseltamivir
2
1
R group is advantageous to bioactivity. For example, the high
bioactivity of 25 appears to be partly dependent upon the
hydrophobic biphenyl group at this site. The change of
hydrophobic and hydrophilic circumstance around C3-
ethylpropoxy greatly influenced compounds inhibitory activity,
maintaining the original structure of C3-substituents is
reasonable.
Figure 7. The binding models of compounds oseltamivir (purple), 3
(
green) and 4 (cyan) with N1 (PDB 2HU0, a) and the key residues that
may form potential interactions with compound 3 (b).

The difference in inhibitory activity for 1, 3 and 4, may be
also interpreted by the number of hydrogen bonds formed by
inhibitor and enzyme. Figure 8a illustrates hydrogen bonding
interactions between compound oseltamivir and key residues
are shown in Scheme 1. The neuraminidase inhibitory assay
and virus inhibition results are listed in Table 2 and Table 3,
respectively. Experimentally, as shown in Table 2, compound 3
exerts good potency, with IC50 value of 39.6 µM against NA,
which is lower than that of oseltamivir (IC50 = 61.1 µM). From
Table 3, it can be observed that the inhibitory activity for 3 (IC50
= 1.72 µM) is almost equal to the oseltamivir (IC50 = 0.579
µM), the cytotoxicity for the compound 3 is 373 µM, while the
toxicity value for oseltamivir is > 500 µM. Thus, the selective
index (SI = CC50/IC50) for compound 3 is impressive, indicating
that the biphenyl group probably formed some special
interactions with the N1 enzyme. The inhibition curve and
toxicity test curve of compound 3 against H1N1 virus is shown
in Figure S1 (Supplementary Information).
Molecular dynamics (20 ns) for six inhibitors were
undertaken to confirm the reliability of docking results, . The
system stability and overall convergence of simulations were
monitored in terms of root-mean-square deviation (RMSD) of
NA-inhibitor backbone atoms (C, Cα, N, and O). The average
RMSD value of each step versus time of three representative
inhibitors is shown in Fig. S2 (Supplementary Information). It
can be seen that all complexes except NA- compound 4 are
stable after 4 ns, and fluctuations in the average RMSD values
during the whole simulation are between 2.0 Å and 4.0 Å,
indicating the other complexes keep an approximate invariable
state. Figure 9 depicts a superposition of two structures for NA-
compound 3, and there seems to be no significant difference
between initial structure (in green) and the lowest energy
structure (in pink) except for a slight drift and rotation. The MD
simulation results are in good agreement with molecular docking
results.
(
Arg152, Arg292, and Arg371) in the active site. A total of four
hydrogen bonds were formed. Figure 8b illustrates hydrogen
bonding interactions between compound 3 and key residues
(
Arg152, Arg292, Arg371, and Tyr347) in the active site. A total
of six hydrogen bonds were formed. The hydrogen bond
distances captured are 1.75 Å (Arg152-HN-H…O-CCH -NH),
.01 Å (Arg292-HN-H1…OHCO), 2.24 Å (Arg292-HN-H2…
3
2
OHCO), 1.91 Å (Arg371-HN-H1…OHCO), 2.54 Å (Arg371-
HN-H … O-COH), and 2.50 Å (Tyr347-HC-H … OHCO),
respectively. Figure 8c illustrates hydrogen bonding interactions
between compound 4 and key residues (Arg152, Tyr347) in the
active site, only two hydrogen bonds were formed. Figure 8d, 8e,
8
4
f show that compounds 5, 6, 7 can bind with the receptor for 3,
, 4 hydrogen-bonding, respectively. Thus, it is clearly that the
number of hydrogen bonds is correlated to NAIs activity.
Figure 8. Docking results of the compounds 1 (a), 3 (b), 4 (c), 5 (d), 6 (e)
and 7 (f) in the study. The binding site surrounds the ligand. Hydrogen
bonds are indicated as yellow dashed lines, with unit distance of Å.
Molecular surface is shown as cyan.
Table 2. IC50 ± SD (μM) value of compounds 3 and 1 against NA
Figure 9. View of superimposed backbone atoms of the initial structure
(
green) of the MD simulations and the lowest energy structure (pink) for
3
39.6 ± 6.5
1 (oseltamivir)
61.1 ± 6.3
compound 3-2HU0 complex. Compound 3 represented as carbon-chain
in green for the initial complex and carbon-chain in pink for the lowest
energy complex.
IC50 (μM)
Table 3. IC50 ± SD (µM) value of compounds 3 and 1 against
A/WSN/33 H1N1
To quantify the docking results, the binding free energy
calculations for the six complexes of the last 2 ns trajectory were
performed and are listed in Table S4 (Supplementary
Information). MM/GBSA (Molecular Mechanics/Generalized
3
1 (oseltamivir)
0.579 ± 0.082
>500
IC50 (μM)
CC50(μM)
1.72 ± 0.35
373 ± 19.50
Born
Surface
Area)
and
MM/PBSA
(Molecular
Mechanics/Poisson Boltzmann Surface Area) were applied in
free energy calculations. As listed in Table S4, the binding free
energies of 1, 3-7 via MM/GBSA were predicted to be -16.27, -
2
3.21, -12.14, -18.85, -14.95 and -12.25 kcal/mol, respectively,
and via MM/PBSA were predicted to be -15.29, -22.36, -13.35, -
7.23, -14.31 and -13.92 kcal/mol, respectively. It can be
1
detected that all binding free energies of 3-7 are more negative
than the observed free energy of -10.60 ~ -10.36 kcal/mol for
3
5
oseltamivir, implying that the introduced chemical structures
increase the binding affinities of ligands coupling with subtype
N1. Among the six compounds, 3 should possess the highest
activity on account of its most negative binding free energy. The
calculated binding free energies are in good agreement with the
Scheme 1. Synthesis of compounds 3. Regents and conditions: (a)
o
o
Aldehyde, EtOH, 30 C; NaBH
C, then acetic acid.
3
CN, 30 C; (b) MeOH, H
2
O, NaOH, 55
o
We conducted experimental synthesis to study the predictive
compound 3 severely. The major structures and synthetic route
3
D-QSAR predicted results.

In this paper, a series of N-substituted oseltamivir derivatives
were studied by computer-aided drug design processes, such as
12. Gupta R K.; Nguyen-Van-Tam J S. New. Engl. J. Med. 2006, 354,
1423-1424.
3
D-QSAR, molecular docking and molecular dynamics
1
3. Hayden, F. G.; Osterhaus, A. D.; Treanor, J. J.; Fleming, D. M.; Aoki,
simulations. The built 3D-QSAR models exhibit excellent
internal predictive activity, which can be extrapolated to design
new and more robust inhibitors. Docking results show that the
main residues in the active pocket 2HU0 are hydrophobic, and
Arg152, Arg292, Arg371 are key residues. Hydrogen bond and
hydrophobic interactions play key roles in binding of inhibitors
and NA. MD simulations were carried out for designed potential
compounds to verify the docking results. A novel oseltamivir
derivative was further synthesized in good yield with an
optimized strategy. The compound were then screened for its
bioactivity against neuraminidase and H1N1 virus in liquid
media and fluorescence detection. Compound 3 shows even
better inhibitory activity relative to the compound 1 (oseltamivir).
In conclusion, the N-substituted oseltamivir analogs discussed in
this paper provided neoteric viewpoint on enhancing drugs
inhibitory activity.
F. Y.; Nicholson, K. G.; et al. New. Engl. J. Med. 1997, 337, 874-880.
14. Chachra, R.; Rizzo, R. C. J. Chem. Theory. Comput. 2008, 4, 1526-
1540.
1
1
5. Birnkrant D, Cox E. New. Engl. J. Med. 2009, 361, 2204-2207.
6. Chen, B. L.; Wang, Y. J.; Guo, H.; Zeng, G. Y. Eur. J. Med.
Chem. 2016, 109, 199-205.
1
1
7. Hoffmann, A.; Schade, D.; Kirchmair, J.; Clement, B.; Sauerbrei, A.;
Schmidtke, M. J. Virol. Methods. 2016, 237, 192-199.
8. Xie, Y.; Xu, D.; Huang, B.; Ma, X., Qi, W.; Shi, F., ... Xu, W. J. Med.
Chem. 2014, 57, 8445-8458.
1
2
9. Li Y.; Zhou B.; Wang R. J. Mol. Graph. Model. 2009, 28, 203-219.
0. Romero-Parra J.; Chung H.; Tapia R A, et al. Eur. J. Pharm. Sci. 2017,
1
01:1.
Acknowledgments
2
2
1. Lu P.; Wei X.; Zhang R. Eur. J. Med. Chem. 2010, 45, 3413-3419.
2. Myint K Z.; Xie X Q. Int. J. Mol. Sci. 2010, 11, 3846-3866.
Thanks for financial support given by Natural Science
Foundation of Shanghai (No. 15ZR1440400), National Youth
Fund Project of China (No. 21602135), Collaborative Innovation
Fund (XTCX2016-14), Shanghai Municipal Education
Commission (Plateau Discipline Construction Program), and the
National Natural Science Foundation of China (No. 21472126).
Authors are also thankful to Jinan Huawei Pharmaceutical Com.
Ltd. and Institut Pasteur of Shanghai for their help in
neuraminidase inhibitory assay and virus inhibitory assay.
23. Sun, C.; Zhang, X.; Huang, H.; Zhou, P. Bioorg. Med. Chem. 2007, 14,
8574-8581.
2
4. Yi, X.; Guo, Z.; Chu, F. M. Bioorg. Med. Chem. 2003, 11, 1465-1474.
25. Shu, M.; Lin, Z.; Zhang, Y.; Wu, Y.; Mei, H.; Jiang, Y. J. Mol. Model.
2011, 17, 587-592.
Supplementary data
26. Zhu Y L.; Beroza P.; Artis D R. J. Chem. Inf. Model. 2014, 54, 462-
4
69.
Materials and Methods, Table S1, Table S2, Table S3, Table
S4, Figure S1, Figure S2, 1H NMR and 13C NMR spectra,
HRMS spectra for the new drug associated with this paper.
27. Mohan, S.; McAtamney, S.; Haselhorst, T.; von Itzstein, M.; Pinto, B.
M. J. Med. Chem. 2010, 53, 7377-7391.
References and notes
28. Albohy, A.; Mohan, S.; Zheng, R. B.; Pinto, B. M.; Cairo, C.
W. Bioorg. Med. Chem. 2011, 19, 2817-2822.
1
.
Mooney, C. A.; Johnson, S. A.; P, T. H.; Quarles, v. U. L.; de Haan, C.
A.; Moret, E. E.; et al. J. Med. Chem. 2014, 57, 3154-3160.
2
3
3
3
3
3
9. Collins, P. J.; Haire, L. F.; Lin, Y. P.; Liu, J.; Russell, R. J.; Walker, P.
A.; ... Gamblin, S. J. Nature. 2008, 453, 1258-1261.
2
.
Niikura, M.; Bance, N.; Mohan, S.; Pinto, B. M. Antivir. Res. 2011, 90,
1
60-163.
0. Jones, B. A.; Lessler, J.; Bianco, S.; Kaufman, J. H. Plos One. 2015,
10, e0137482.
3
4
5
6
7
.
.
.
.
.
Greenway K T.; Legresley E B.; Pinto B M. Plos One. 2013, 8,
e59873.
1. Selvaraj, C.; Tripathi, S. K.; Reddy, K. K.; Singh, S. K. Cur. Trends.
Biotechnol. Pharm. 2011, 5, 1104-1109.
Schade, D.; Kotthaus, J.; Riebling, L.; Kotthaus, J.; Müller-Fielitz, H.;
Raasch, W.; ... Clement, B. J. Med. Chem. 2014, 57, 759-769.
2. Nguyen, H. T.; Nguyen, T.; Mishin, V. P.; Sleeman, K.; Balish, A.;
Mallipeddi, P. L.; Kumar, G.; White, S. W.; Webb, T. R. Curr. Top.
Med. Chem. 2014, 14, 1875-1889.
Jones, J.; ... Nguyen, D. T. Emerg. Infect. Dis. 2013, 19, 1963-1971.
3. Roy P P.; Leonard J T.; Roy K. Chemometr. Intell. Lab. 2008, 90, 31-
Cheng, L. P.; Huang, X. Y.; Wang, Z.; Kai, Z. P.; Wu, F. H. Monatsh.
Chem. 2014, 145, 9-22.
42.
4. Gagic, Z.; Nikolic, K., Ivkovic, B.; Filipic, S.; Agbaba, D. J. Taiwan.
Inst. Chem. E. 2015, 59, 33–44.
Adabala, P. J. P.; LeGresley, E. B.; Bance, N.; Niikura, M.; Pinto, B.
M. J. Org. Chem. 2013, 78, 10867-10877.
3
3
5. Park J W, Jo W H. Eur. J. Med. Chem. 2010, 45, 536-541.
8
9
.
.
Sun J.; Mei H. Chemometr. Intell. Lab. 2015, 146, 485-493.
6. Fu, L.; Bi, Y.; Wu, Y.; Zhang, S.; Qi, J.; Li, Y.; ... Gao, G. F. J. Med.
Chem. 2016, 59, 6303-6012.
Wang, Z.; Cheng, L.; Kai, Z.; Wu, F.; Liu, Z.; Cai, M. Bioorg. Med.
Chem. Lett. 2014, 24, 3869-3876.
3
3
3
7. Russell, R. J.; Haire, L. F.; Stevens, D. J.; Collins, P. J.; Lin, Y. P.;
1
1
0. Gao, L.; Zu, M.; Wu, S.; Liu, A. L.; Du, G. H. Bioorg. Med. Chem.
Lett. 2011, 21, 5964-5970.
Blackburn, G. M.; ... Skehel, J. J. Nature. 2006, 443, 45-49.
8. Chen, B. L.; Wang, Y. J.; Guo, H.; Zeng, G. Y. Eur. J. Med. Chem.
2016, 109, 199–205.
1. Mckimmbreschkin, J.; Trivedi, T.; Hampson, A.; Hay, A.; Klimov, A.;
&
Tashiro, M.; et al. Antimicrob. Agents. Chemother. 2003, 47, 2264-
9. Liu, A. L.; Wang, H. D.; Lee, M. Y.; Wang, Y. T.; Du, G. H. Bioorg.
Med. Chem. 2008, 16, 7141-7147.
2
272.

Graphical abstract
Compound 3 (a), IC50 = 39.6 µM, more effective in
inhibiting subtype N1 than compound 1 (b) (with
IC50 value = 61.1 µM); docking results of the
potent molecule 3 and 1 binding amino acid
residues with hydrogen bonds.
4
0.