From Naproxen Repurposing to Naproxen Analogues and Their Antiviral Activity against Influenza A Virus

Article abstract of DOI:10.1021/acs.jmedchem.8b00557

The nucleoprotein (NP) of influenza A virus (IAV) required for IAV replication is a promising target for new antivirals. We previously identified by in silico screening naproxen being a dual inhibitor of NP and cyclooxygenase COX2, thus combining antivira

Full text of DOI:10.1021/acs.jmedchem.8b00557

Subscriber access provided by University of Winnipeg Library
Article
From naproxen repurposing to naproxen analogues
and their antiviral activity against Influenza A virus.
Sébastien Dilly, Aurelien Fotso Fotso, Nathalie Lejal, Gloria Zedda, Mohamad Chebbo, Fryad
Rahman, Simon Companys, Hélène Charlotte Bertrand, Jasmina Vidic, Magali Noiray, Marie-
Christine Alessi, Bogdan Tarus, Stephane Quideau, Beatrice Riteau, and Anny SLAMA-SCHWOK
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b00557 • Publication Date (Web): 20 Jul 2018
Downloaded from http://pubs.acs.org on July 22, 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 68
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
From naproxen repurposing to naproxen analogues
and their antiviral activity against Influenza A virus.
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
1
2
,
3
4
2
Sébastien Dilly ‡, Aurélien Fotso Fotso ‡ Nathalie Lejal ‡, Gloria Zedda ‡, Mohamad Chebbo ,
2
4
4
3
5
Fryad Rahman , Simon Companys , Hélène C. Bertrand , Jasmina Vidic , Magali Noiray , Marieꢀ
2
3
4
2
1,3
Christine Alessi , Bogdan Tarus , Stéphane Quideau , Béatrice Riteau , Anny SlamaꢀSchwok *
1
2
ꢀ Gustave Roussy Institute, Paris Saclay University, UMR8200 CNRS, Villejuif, France.
ꢀ Aix Marseille University, INSERM, INRA, NORT, UMR 1260/1062, Marseille, France;
3ꢀ Paris Saclay University, UR 892, INRA, Jouy en Josas, France;
4
5
ꢀ Bordeaux University, ISM (CNRSꢀUMR 5255), Talence, France;
ꢀ Paris Sud University, Paris Saclay University, UMS IPSIT ꢀ Intermol, ChâtenayꢀMalabry,
France;
KEYWORDS: Antiviral, monomeric nucleoprotein, RNAꢀprotein and proteinꢀprotein
interactions, in vivo mice model of intranasal IAV infection, naproxen.
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 68
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: The nucleoprotein (NP) of Influenza A virus (IAV) required for IAV replication, is
a promising target for new antivirals. We previously identified by in silico screening naproxen
being a dual inhibitor of NP and cyclooxygenase COX2, thus combining antiviral and antiꢀ
inflammatory effects. However, the recently shown strong COX2 antiviral potential makes
COX2 inhibition undesirable. Here we designed and synthesized two new series of naproxen
analogues called derivatives 2, 3 and 4 targeting highly conserved residues of the RNA binding
groove, stabilizing NP monomer without inhibiting COX2. Derivative 2 presented improved
antiviral effects in infected cells compared to that of naproxen, and afforded a total protection of
mice against a lethal viral challenge. Derivative 4 also protected infected cells challenged with
circulating 2009ꢀpandemic and oseltamivirꢀresistant H1N1 virus. This improved antiviral effect
likely results from derivatives 2 and 4 inhibiting NPꢀRNA and NPꢀpolymerase acidic subunit PA
Nꢀterminal interactions.
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
2

Page 3 of 68
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Introduction:
Influenza A viruses (IAV) are a societal and economic burden with yet unpredictable bursts.
The growing resistance of circulating strains to present amantadineꢀderivatives and the low (1ꢀ
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
%) resistance to neuraminidase inhibitors urged the development of additional therapeutic
options. Various hypotheses have been pursued to overcome resistance. Recent antiviral
strategies have focused on internal viral proteins, in particular on the ribonucleoprotein complex
1
ꢀ6
7
,
as detailed below and the nonꢀstructural proteinꢀ1 NSꢀ1 being a virulence factor . Additional
omics studies have identified host cellular factors used by the virus, as several MAP kinases for
which targeted inhibitors are available drugs. It is likely that these studies will lead to
repurposing of many drugs in the near future.
New therapeutic perspectives have been opened up by targeting conserved viral proteins, as
protein of the ribonucleoprotein complex involved in transcription and replication of the virus.
8
, 9
The crystal structures of the nucleoprotein (NP) assembled in a trimeric form
have
highlighted the selfꢀassembly of NP monomers via its tailꢀloop and hence NP activity could be
10
inhibited by targeting the crucial E339ꢀR416 salt bridge for NP trimer stabilization . High
throughput studies have identified nucleozin as a potent inhibitor of the NP that binds at Y52,
1
1, 12
13
N309
, a site shared with antiviral proteins of the MX1 families . Disrupting the interaction
5, 14
of NP and viral RNA was achieved by naproxen that stabilized the NP in a monomeric form
.
Compound RK424 binds in a small pocket close to the dimer interface and the nuclear expert
signalꢀ3 NES3, at a phosphorylation site nearby the naproxen binding site. The residues
15
interacting with RK424 deduced from docking studies were R162, S165, L264, and Y487 .
Additional proteinꢀprotein interactions could be targeted by small molecules inhibitors based on
2
earlier structures of isolated domains of the polymerase subunits and recent structures of the
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 68
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
4, 16ꢀ18ꢀ19
trimeric polymerase of Influenza A, B and C
. Still, the interactions between NP and the
polymerase remain to be characterized to serve as a structural basis for drug design.
In our earlier studies, we have been developing new inhibitors targeting the nucleoprotein by
disrupting the interactions between NP and viral RNA. For this, we screened in silico for
molecules that would stack on Y148, the only aromatic residue in the putative RNA binding
groove of NP, stabilized by hydrophobic interactions with the Cꢀterminal, in particular F489, and
by electrostatic interactions with nearby arginine residues as R355, R361 or R152. Our first hit
was naproxen, a known drug for its antiꢀinflammatory effect, being a cyclooxygenase (COX2)
inhibitor. We showed that naproxen combined antiviral and antiꢀinflammatory effects by
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
targeting both NP and COX2 . Of note, naproxen also did not promote resistance after eight cell
5
passages when oseltamivir already generated resistance to treatment after 4ꢀ5 cell passages . The
present work describes the design, in silico docking, chemical synthesis, cellular and in vivo
antiviral effects of a new series of naproxen analogues. We demonstrate that some of these
derivatives not only impeded NPꢀRNA interactions, but also NPꢀpolymerase association.
Results
1
- Modeling naproxen analogues by fragment-based drug design:
Derivatives of naproxen were designed to additionally target specific arginine residues of the
RNA binding groove of NP besides R361 recognized by naproxen (Figure 1A). To improve their
affinity and specificity for NP, addition of an aliphatic or aromatic carboxylate fragment to
naproxen led to the first two derivatives 1 and 2, respectively (Figure 1B, C). The additional
ACS Paragon Plus Environment
4

Page 5 of 68
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
carboxylate fragment aimed at reaching the residues R152/ R150 in addition to R361, while
keeping stacking interactions of Y148 and the Cꢀterminal F489 with the aromatic naphthalene
core of naproxen. The stability of naproxen derivatives in their binding pocket were investigated
through MD simulations of the complex with NP (Table 1). We rapidly discovered that the
aliphatic carboxylate fragment of derivative 1 was not optimal in in vitro and cellular assays
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
5
, 6, 20
(vide infra). Our previous simulations
with derivative 2 showed that Y148 made transient
πꢀπ interactions with both the phenyl and naphthalene rings of this derivative leading to some
destabilization within the binding site. As Y148 is the only aromatic residue in the RNA binding
8
groove suggested to be important for NP interaction with RNA, we aimed at a better
recognition of this residue by designing new naproxen derivatives. We extended the structure of
derivative 2 with an aminophenol fragment at either the meta position (derivative 3, Figure 1D)
or the ortho position (derivative 4, Figure 1E) of the 1,3ꢀbenzenedicarboxylate unit. The
derivative 3 was designed to donate a hydrogen bond to the hydroxyl group of Y148, and the
derivative 4 was designed to target the ring of Y148 through a hydrophobic interaction with the
aminophenol ring.
The stabilities of the new derivatives 3 and 4 within their initial putative binding pockets on
NP were tested through molecular dynamics (MD) simulations in explicit water and compared to
that of derivative 2. Table 1 and Supplementary Figure 1 compare the interaction energies of the
ligands in complex with the nucleoprotein. We expected the following order of stability of the
NPꢀligand complexes: derivative 4 > derivative 3 > derivative 2 by the free binding energy
estimation (Table 1). Docking and MD simulations showed that the binding site of derivative 4
(green) was distant from those of other compounds (Figure 1F). This difference occurred after 30
ns of simulation (Supplementary Figure 1). The carboxylate groups of derivative 4 form tighter
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 68
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
interactions with new arginine residues R150, R152, R156 and R195 (Figure 1E and Table 1) as
compared with those involved in binding to derivative 2 involving mainly R152 and R355
(Figure 1C). R152 and R156 were also engaged in πꢀcation interactions with the naphtalene
moiety, thus strongly stabilizing this part of derivative 4. As expected, the additional
aminophenol group of derivative 4 recognized Y148 and R150 through an aromatic and cationꢀπ
interactions.
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
In contrast with derivative 4, derivative 3 bound at a NP site close to the binding site of
derivatives 1 and 2 (Figure 1D). However, a stabilization of the aminophenol group by an
stacking with Y148 and a πꢀcation interaction with R361 induced a rotation of 180° as compared
to the orientation observed in naproxen and derivatives 1 and 2 (Figure 1AꢀC). Derivative 3
forms only a salt bridge with R361 while one carboxylate interacting with the backbone and the
sideꢀchain of N149 through hydrogen bonds, thus explaining its lower affinity than that of
derivative 4.
Besides interactions of the naproxen analogue 4 within its binding site, structural changes of
6
the protein are observed upon binding, as previously reported for derivative 2 binding to NP .
Clear changes in the oligomerization loop folding back upon its own RNA binding groove, the
cavity around E339 where the oligomerization loop protrudes from the next monomer within NP
trimer shrinks upon derivative 4 binding to NP, all these changes are consistent with a
destabilization of NP trimer and stabilization of NP monomer by derivative 4. Figure 2A shows
that the induced fit observed upon derivative 4 binding to NP resembles the structural changes
observed in R416A (Figure 2B), a prototype of monomeric NP with very weak RNA binding
ACS Paragon Plus Environment
6

Page 7 of 68
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
14
ability . Moreover, the cavity receiving the oligomerization loop (around E339) shrinks by
derivative 4 binding to NP in a similar way as that observed in R416A (Figure 2C and D,
respectively).
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
- Chemical synthesis
Synthesis of derivative 4 (12) (see Scheme 1). Derivative 4 (12) was synthesized in ten steps as
depicted in Scheme 1. Commercially available 2ꢀnitroꢀmetaꢀxylene (1) was brominated in 93%
21
yield using molecular bromine and a catalytic amount of powdered iron . The resulting bromo
derivative 2 was reduced with iron in the presence of acetic acid to furnish 2,6ꢀdimethylꢀ4ꢀ
2
2
bromoaniline (3) in 90% yield . This aromatic amine was converted into the corresponding
iodo derivative 4 by a Sandmeyer reaction in 60% yield. Oxidation of the two methyl
substituents of 4 using potassium permanganate (KMnO ) in a refluxing 1:1 tertꢀbutanol (tꢀ
4
BuOH)/H2O solvent mixture led to the unknown isophthalic acid derivative 5 in 65% yield.
Esterification of both carboxylic groups of 5 could not be accomplished by simply using MeOH
in the presence of concentrated H SO . Only a monoester was obtained, so the second carboxylic
2
4
acid group was esterified using MeI in the presence of K CO . The diester 6 was then coupled
2
3
with commercially available 6ꢀmethoxyꢀ2ꢀnaphthaleneboronic acid (7) under Suzuki conditions
23
to afford the naphthylisophtalate derivative 8 in 47% yield. Coupling of 8 with paraꢀ
2
4
acetamidophenol 9, easily prepared from paraꢀaminophenol , was accomplished under
25
Ullmannꢀtype conditions , but unfortunately the desired diaryl ether 10 was isolated in only
0% yield. Nevertheless, derivative 4 (12) was finally obtained by complete hydrolysis of 10 in
1
two steps; the hydrolysis of the amide group was performed using 1 N HCl in 1,4ꢀdioxane to
ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 68
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
afford 11, which was then saponified using aqueous potassium hydroxide (KOH) in refluxing
THF to furnish 12 in 50% yield over the two steps (Scheme 1, see the Supporting Information for
details).
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
Synthesis of derivative 3 (20) (see Scheme 2). Derivative 3 (20) was synthesized in seven steps
starting from commercially available 4ꢀbromoꢀ2,6ꢀdimethylaniline (13), as shown in Scheme 2.
This aromatic amine was quantitatively converted into to the corresponding iodo derivative 14
via a Sandmeyer reaction. Oxidation of its two methyl substituents using potassium
permanganate (KMnO ) in a refluxing 1:1 tertꢀbutanol (tꢀBuOH)/H2O solvent mixture led to the
4
2
6
known isophthalic acid derivative 15 , which was classically esterified using MeOH in the
presence of concentrated H SO . Coupling between the resulting 5ꢀbromoꢀ2ꢀiodoisophthalate 16
2
4
and the commercially available 6ꢀmethoxyꢀ2ꢀnaphthaleneboronic acid (7) under Suzuki
23
conditions afforded the 2ꢀnaphthylisophtalate derivative 17 in 40% yield. Compound 17 was
25
then coupled with paraꢀacetamidophenol 9 under Ullmannꢀtype conditions , but again the
desired diaryl ether 18 could be isolated in only 10% yield. Derivative 3 (20) was finally
obtained by first hydrolyzing the amide group of 18 in quantitative yield using 1 N HCl in 1,4ꢀ
dioxane, and the resulting free aniline derivative 19 was saponified using aqueous KOH in
refluxing THF to furnish 20 in 50% yield, as for derivative 4 (12) (Scheme 2, see the Supporting
Information for details).
3
- Differential stabilization of monomeric NP by the naproxen analogues
NP contains an oligomerization loop involved in selfꢀassociation and together with two other
flexible loops promotes interactions with RNA of great importance for NP function (see Figure
ACS Paragon Plus Environment
8

Page 9 of 68
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2
). Therefore, we anticipated that a compound stabilizing the monomeric NP would potentially
disturb intermolecular interactions, as NPꢀNP and NPꢀRNA interactions. We used dynamic light
scattering for monitoring the thermal stability of NP by the apparent protein size as a function of
temperature. The experimental conditions used for preparation of recombinant NP have been
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
20
shown to produce NP in a monomeric form (> 80%) . This experiment estimated whether
naproxen or the naproxen analogues stabilized monomeric NP against thermal denaturation. The
first derivatives of the melting curves considered as the apparent melting temperatures (Tm) of
the NPꢀligand complexes were used for ranking the relative affinity of compounds to bind NP at
1
/1 NP/ compound ratio (Supplementary Figure 2). While naproxen did not stabilize NP,
derivatives 2 and 4 stabilized NP (Table 2). The melting of derivative 3ꢀNP complex showed a
biphasic melting profile, thus, the formed complexes were less stable than those formed between
NP and derivatives 2 and 4. Such ranking of naproxen analogues regarding their ability to bind
NP (derivative 4 ≥ derivative 2 >> naproxen) was similar from the modeling studies (Figure 1).
The biphasic behavior of derivative 3 could be explained by the aminophenol fragment binding
tended at destabilizing the naproxen backbone that was not well buried into NP RNA binding
groove.
The above models suggested that derivatives 2 and 4 binding to NP likely involved residues of
27
NP also proposed to be important for polymerase assembly and function as Cꢀterminal F489 .
Naproxen and derivative 2 binding to NP reduced the flexibility of loop 2 (residues 200ꢀ214) and
that of a loop close to the Cꢀterminal (residues 429ꢀ436) and involved R152 located very close to
2
0
R150. Since loop 2 residues R204, W207 and R208 and R150 are required for vRNA synthesis
28
29
and polymerase function , we anticipated that the new naproxen derivatives, in particular
derivative 4 that interact with R150 would not only impede NPꢀRNA interactions but also
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 68
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
interfere with polymerase function. We have recently identified a complex between NP and the
30
Nꢀterminal of polymerase acidic subunit PA, away from the endonuclease active site of PA .
We investigated whether the naproxen analogues may disrupt the NPꢀPA(1ꢀ27) peptide complex.
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
We have previously described a pullꢀdown assay based on complex formation between proteins
30
from lysates of IAVꢀinfected cells with the PA (1ꢀ27) attached to magnetic beads . A specific
complex was observed between PA(1ꢀ27) and recombinant NP or NP from the lysate of
influenza virus infected cells as revealed by Western blot analysis using a monoclonal antibody
against NP (Figure 3A). We then compared naproxen and derivatives for their ability to impede
this interaction. Both naproxen and derivative 1 were inefficient in disrupting this interaction. In
contrast, derivatives 2, 3 and 4 efficiently inhibited this interaction as detected by western blots.
Of note, NP was likely bound to nucleic acids in WB assay since the cell lysates were not treated
for digestion of nucleic acids.
4
- SPR experiments
We performed surface plasmon resonance experiments for a quantitative comparison of the
relative affinities of naproxen and derivative 2, as they differ in their ability for NPꢀPA(1ꢀ27)
complex disruption (Figure 3A). In these experiments, the biotinylated PA(1ꢀ27) peptide was
captured onto the chip surface and NP alone or NP with the naproxen or derivative 2 were
injected onto the chip. We also tested NPꢀDNA complexes in the presence of the naproxen
analogues. We were unable to replace DNA by RNA as the RNAꢀcomplexes stick to the
reference surface to a great extent. When NP alone was injected to the chip, a complex NPꢀPA(1ꢀ
27) was evidenced.
ACS Paragon Plus Environment
10

Page 11 of 68
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Addition of derivative 2 (100 nM) to NP largely reduced the signal of the complex (Figure 2Bꢀ
a). We reproduced the same experiment while replacing derivative 2 by naproxen. Naproxen
perturbed NPꢀPA(1ꢀ27) complex less efficiently than derivative 2 did and was unable of NPꢀ
DNAꢀPA(1ꢀ27) complex disruption, even at a concentration of 50 µM (Figure 3Bꢀb,c). In
contrast, the NPꢀ PA(1ꢀ27) complex formed in the presence of DNA was disrupted by derivative
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
in doseꢀdependent manner, with an IC50 of about 30 nM (Figure 3Bꢀd). These data are in good
agreement with the western blots (Figure 3A).
5
- Cytotoxicity tests:
Before testing the antiviral activities of derivatives 2, 3 and 4, we performed cell viability tests
by MTT and trypan blue tests to verify their cytotoxicity (Figure 4 and Table 2). The naproxen
analogues in their salt form were very soluble in buffer. The cytotoxic concentration CC of
50
derivatives 2, 3 and 4 were tenfold higher than that of naproxen, being 13 mM and 1.4 mM,
respectively (Figure 4).
6- Comparison of the antiviral effects of the naproxen analogues and naproxen in IAV-
infected MDCK and A549 cells.
The protection of infected MDCK cells against a viral IAV challenge was tested at two
ꢀ3
multiplicities of infection, MOI= 10 after 48 hours and MOI = 2 after 18 hours reflecting
conditions of multiple cycles of viral replication, with mild and higher viral load. At low MOI,
we determined viral titers whereas at higher MOIs, we measured NP levels by IF, corrected for
ACS Paragon Plus Environment
11

Journal of Medicinal Chemistry
Page 12 of 68
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
cell viability assessed by DAPI measurements. We used both A/WSN/33 (H1N1) (Table 2:
MDCK cells) and A/Udorn/72 (H3N2) strains (Table 4, MDCK cells). The latter strain was also
applied on A549 cells as summarized in Table 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
ꢀAntiviral effect against A/WSN/33 (H1N1)
At MOI = 2, the doseꢀ response curve of derivative 2 was clearly shifted toward lower
concentrations as compared to that of naproxen (Figures 4A,B). The antiviral effect of derivative
2
was similar for the two viral strains, giving IC about 5 times lower than that for naproxen
50
(Tables 2 and 3). Consequently, the selectivity index (SI) of derivative 2 being 684 increased by
~50 fold as compared to a SI value of 14 for naproxen.
ꢀ3
At MOI 10 , derivative 1 achieved a partial protection of the infected cells and was therefore
discarded for further investigation. The IC values were about one order of magnitude lower for
50
derivatives 2 and 4 than for naproxen, yielding selectivity indexes of 5200 and 9230,
respectively, compared to 87 for naproxen (Table 2).
-Antiviral effect against A/Udorn/72 (H3N2):
Tables 2, 3 and 4 showed that the derivatives 2 and 4 presented very similar antiviral effects
against H3N2 and H1N1 strains, protecting both MDCK and A549 cells.
ꢀAntiviral effect against the 2009 pandemic Influenza strain A/Netherlands/602/09
Figure 5A, B and Table 5 summarize the protection afforded by the derivatives 2, 3 and 4 against
a viral challenge with the pandemic 2009 Influenza A strain. 10 µM of derivatives 2 and 4
significantly protected the MDCK cells while the derivative 3 only slightly protected the cells.
Quantification of the IF results presented in Figure 5C, yielded the following order of efficacy:
ACS Paragon Plus Environment
12

Page 13 of 68
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
derivative 4 > derivative 2 > derivative 3. The selectivity index of derivative 4 reached 17000 as
compared to 2600 and 619 for derivatives 2 and 3, respectively.
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
-
Antiviral effect against the oseltamivir-resistant strain
We generated oseltamivirꢀresistant A/PR/8/34 (H1N1) virus obtained after 4 cell passages as
described in the Experimental section. This virus was used to infect MDCK cells and
representative IF data are shown in Figure 6AꢀB; quantification is shown in the insert of panel B.
Treatment with increasing concentrations of derivatives 2 and 3 afforded a better protection of
the cells, in contrast with 2µM of derivative 4 being already protective. Oseltamivir could not
protect these cells against this oseltamivir ꢀresistant strain as shown in panel C, while oseltamivir
had a clear antiviral effect in oseltamivir ꢀsensitive A/PR/8/34 (H1N1). Table 6 summarizes the
antiviral effects of derivatives 2, 3 and 4. Note that derivative 4 is equally efficient in the
pandemic and oseltamivir ꢀresistant Influenza A strains within experimental error.
7
- Loss of the anti-COX2 activity of naproxen by affording more selective antiviral
compounds.
Naproxen is an antiꢀinflammatory drug which analgesic and antiꢀinflammatory effects are
dependent on the extent and duration of cycloꢀoxygenaseꢀ2 (COX2) inhibition in the spinal cord
and inflammatory sites. However, the simultaneous inhibition of COX2 in the vasculature
increases the risk of hypertension and heart failure while lowering doses cannot avoid the
3
1
cardiovascular risk of naproxen in highꢀrisk CV patients . Therefore, it is desirable for an
increased safety of naproxen analogues to avoid COX2 inhibition. Moreover, the infection with
influenza A viruses (IAV) provokes activation of cellular defense mechanisms contributing to
ACS Paragon Plus Environment
13

Journal of Medicinal Chemistry
Page 14 of 68
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
the innate immune and inflammatory response. In this process the cyclooxygenaseꢀ2 (COX2)
plays an important role in the induction of prostaglandinꢀdependent inflammation. While it has
been reported that COX2 is induced upon IAV infection, inhibition of COX2 expression by IAV
3
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
viruses strongly suggests that COX2 inhibition is undesirable to fight against IAV .
We tested whether COX2 levels were increased in IAV infected cells compared to uninfected
cells and whether naproxen and derivative 2 would reduce COX2 levels. Figure 7 shows an
increase of COX2 induced by IAV. This increased COX2 level was significantly reduced by
addition of 300 µM naproxen, a concentration at which the viral titer is reduced by two orders of
5
magnitude in infected cells . In contrast, derivatives 2 or 4 (100 µM) did not inhibit COX2.
For the below in vivo studies, we used derivative 2 as a prototype of the naproxen analogues and
compared its effect with naproxen treatment of IAVꢀinfected mice.
8
- Derivative 2 protected mice more efficiently than naproxen against a viral challenge.
The antiviral activity of derivative 2 was then evaluated in a mice model of Influenza A/PR/8/34
H1N1) infection (500 pfu/mouse), infected at day 0. The mice were treated by intraperitoneally
administration of naproxen or derivative 2 or oseltamivir or saline buffer shortly after infection
5ꢀ10 minutes), and once a day during the next two days, at days 1 and 2. The protection
(
(
afforded by derivative 2 administered at 1 or 8 mg/mouse/day (50 or 400 mg/kg, respectively)
was compared with naproxen treatment used at 1mg/day. The survival curves showed a doseꢀ
dependent protection when the mice were treated with derivative 2 (Figure 8A). The survival
rates at day 12 postꢀinoculation reached 70 % when treated with 1 mg/day of derivative 2 while
only 20% of the mice survived after naproxen treatment. A complete protection was achieved by
ACS Paragon Plus Environment
14

Page 15 of 68
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
treating the mice with derivative 2 at 8 mg/day or with oseltamivir. Consistently, when treated
with only 1 mg of derivative 2, the mice could recover up to 95% of their initial weight while
naproxen could only stabilize their weight loss to 85% of their initial weight (Figure 8B).
Amazingly, the mice completely recovered from a mild (5%) weight loss upon treatment with
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
8
mg/day of derivative 2. The mice treated with derivative 2 (8mg) were more protected against
weight loss than with oseltamivir treatment during the first days of infection, while all mice
survived under both treatments.
To further corroborate the observed antiviral effect deduced from the survival and body weight
loss curves, we determined the viral titers in the BAL at the third day postꢀinoculation. In
7
infected vehicleꢀtreated mice, the viral titer reached 2 x 10 pfu/ml. In mice treated with naproxen
or derivative 2 (1 mg/day), the viral titer was reduced by 2 and 6 folds. Higher derivative 2
concentration (8 mg/day) decreased the viral titer by two orders of magnitude (120 folds). In the
same conditions, oseltamivir reduced the viral titer by 260 folds, within experimental error a
similar protection than that of derivative 2.
Discussion:
NP plays a pleiotropic role in the RNP function, stabilizing both cRNA and vRNA, disrupting
RNA secondary structures facilitating transcription, supporting elongation of nascent transcripts
during replication. NP is a highly abundant and conserved viral protein, with no equivalent
protein in the host cell transcription/ replication machinery. When combined, these arguments
strongly support that NP is a good target for new antivirals. From these essential roles of NP in
33
viral functions, it could be anticipated from an inhibitor of NP to interfere with RNP function .
ACS Paragon Plus Environment
15

Journal of Medicinal Chemistry
Page 16 of 68
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
-
Stabilizing both NP C-terminal and nearby residues of the RNA binding groove for targeting
both NP and the polymerase ie PA N-terminal. The interaction modes of the NP with polymerase
subunits still require insight at the molecular level. The strength of molecular modeling and
proteinꢀprotein docking is to suggest models based on very solid crystal structures of each NP
and polymerase component to be validated in experimental studies. Interactions of NP with PB1
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
and PB2 , NP and the Cꢀterminus of PA were observed. Electron microscopy showed a
1
, 4, 11
rearrangement of the polymerase, in particular in the PA subunit
in these NPꢀPA
30
interactions, a phenomenon confirmed by our docking studies . We also identified a complex
between NP and PA Nꢀterminal by docking, further supported by experimental tests in vitro and
3
0
in the lysate of infected cells. This complex could be a target for antiviral drug .
We went here one step further in targeting not only the NPꢀRNA interactions as demonstrated for
naproxen but also proteinꢀprotein interaction. To do so, we extended naproxen by fragmentꢀ
6
based drug design , to stabilizing Y148 and the Cꢀterminal aromatic residues F489, F488 by
stacking interactions with their aromatic backbone and reaching nearby RNA binding groove
residues R152 and R355 in addition to R361 recognized by naproxen. Of note, mutation of all
34
these residues was found to be lethal, since they could not be rescued . Preliminary attempts to
extend the methoxy group of naproxen were destabilizing compared to naproxen binding, likely
because of the presence of a “bump” formed by Y487, F338 from one side, F488 from the other
side that would not favor binding of short polar fragments (data not shown). This hydrophobic
region was proposed to stabilize the biphenyl group of RK424 (via Y487 or F488) bearing a
naproxenꢀlike quinoline core, a compound proposed to bind NP at a closeꢀby site from the
1
5
binding site of naproxen demonstrated by point mutations . Because of NP flexibility, the
binding of derivative 2 induced changes in the binding pocket of the tail loop, affecting in
ACS Paragon Plus Environment
16

Page 17 of 68
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
particular E339, Y487 and S165, and reduced the flexibility of loop 1 (residues 73–90) and loop
6
2
(residues 200–214) and a Cꢀterminal linker (residues 428ꢀ438) . As S165 is a residue that can
be phosphorylated, such induced fit by derivative 2 or derivative 4 binding may affect NP postꢀ
1
4, 35
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
translational modifications and likely other proteinꢀNP interactions
. Because the RNA
binding groove and Cꢀterminal residues of NP stabilized by derivatives 2, 3 and 4 are
participating to NPꢀPA – Nꢀterminal interactions, we anticipated that these naproxen analogues
would disrupt the complex. We identified this complex in the supernatant of infected cells; it was
disrupted by derivative 2, 3 and 4 while naproxen and derivative 1 were inefficient in this
competing binding, in agreement with our expected interaction site for NPꢀPA Nꢀterminal
complex. Additionally, derivative 1 had a biphasic behavior in competition studies, suggesting
more than one binding site. The SPR data confirmed and quantified the specific disruption of the
complex with PA(1ꢀ27) by derivative 2 with an IC50 of 30 nM. The tenꢀfold decrease in IC50
value of derivative 2 for NP was consistent with a more efficient competition for disrupting RNA
5
, 6
binding to NP than that observed for naproxen
.
Conclusions:
-
Suitable predicted drug-like profile of the naproxen analogues. The new naproxen analogues
present a relatively low molecular weight, a good solubility in aqueous solution, properties of
many drugꢀlike compounds. To assess more precisely the predicted pharmacokinetics and drugꢀ
like profile of the naproxen derivatives, we used different relevant drugꢀlike filters (Table 7).
3
6
First, the bioavailability radar calculated by the free web tool SwissADME indicated that the
two negative charges of the carboxylate moieties of derivatives 2, 3 and 4 does not really impact
ACS Paragon Plus Environment
17

Journal of Medicinal Chemistry
Page 18 of 68
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
the bioavailability profile of the compounds in comparison with naproxen, with 5 out of 6
physicoꢀchemical parameters important for the bioavailability are still present the favorable pink
3
7
region. We also applied the Lipinski’s rules and a bioavailability score developed by Abbot
3
8
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
laboratories . According to the first filter, no violation is observed for the different derivatives,
indicating that they have high predicted probability to be an oral drug. The second filter shows
for all compounds having a probability of 56 % and at least 10% oral bioavailability in rat or
measurable Cacoꢀ2 permeability. Then, we used the GI absorption predictor calculated by the
BOILEDꢀEgg model, a model for passive human gastrointestinal that indicates a high score of
3
9
absorption for all compounds . Finally, we employed a Pꢀglycoprotein (Pꢀgp) substrate filter to
4
0
evaluate the capacity of the compounds to undergo efflux through biological membrane . None
of the compounds were predicted to be substrates of the Pꢀgp protein. In conclusion, the results
of these drugꢀlike filters indicate that the structures of the naproxen derivatives, especially the
presence of two carboxylic acids, are not critical for the predicted pharmacokinetics profile of
the compounds. Another possible drawback is the presence of an aniline moiety in the structure
of the derivative 4. Indeed, this chemical group is known to present liver cytotoxicity. To
overcome this possibility, the amino group could be removed without affecting the activity of the
compound since it does not participate in the molecular recognition by the NP (Figure 1E).
Indeed, the recognition is mainly driven by the phenyl moiety involved in πꢀπ stacking
interactions with Y148 and cationꢀπ interactions with R150.
Improved antiviral efficacy of the naproxen analogues. The higher affinity of derivatives 2, 3 and
4
for NP than that of naproxen was at the expense of a lower affinity for COX2. Although it is
desirable to reduce airway inflammation induced by Influenza A virus, naproxen inhibition of
COX2 could be associated with an increased cardiovascular risk. Therefore, the lack of
ACS Paragon Plus Environment
18

Page 19 of 68
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
significant inhibition of COX2 levels by derivative 2 might increase the cardiovascular safety, in
addition with antiviral properties.
Accordingly, when compared to naproxen, derivative 2 presented more efficiently protected cells
against a viral challenge with two viral stains, H1N1 (Influenza A/ WSN/33 and A/PR8/34) and
H3N2 (Influenza A/Udorn/72). A 5 to tenꢀ fold decrease in the relative efficacy in the antiviral
effect combined with a higher CC50 value makes derivative 2 an interesting antiviral lead
candidate (Tables 2 & 3). This increase of efficacy was also observed in in vivo experiments:
derivative 2 could totally protect mice from a viral challenge. This correlated with a two orders
of magnitude decrease of the pulmonary viral titer.
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
Moreover, derivative 4 was equally protecting MDCK cells against viral challenges with
pandemic A/Netherlands/602/09 strain, A/WSN/33 and oseltamivirꢀ resistant A/PR/8/34 (H1N1)
ꢀ3
virus, with IC50 being 1.0 ± 0.2 µM, 1.3 ± 0.2 µM, 0.7 ± 0.1 µM, respectively at MOI 10 with
SI values in the range 10000 ꢀ 20000. We emphasize that derivative 4 was more potent than
derivative 2, for which the following IC50 were obtained 8.8 ± 2.5 µM, 2.9 ± 0.4µM, 2.0 ± 0.5 µM
and 5 ± 3 µM for A/Netherlands/602/09 strain, A/WSN/33, A/Udorn/72 (H3N2) and oseltamivirꢀ
resistant A/PR/8/34 (H1N1), respectively.
-Interest in combining antiviral and anti-inflammatory effects in severe cases of IAV infection.
Previous combination regimen has shown that a biꢀtherapy is more effective with drugs
2
, 3
addressed to two targets as for example HA and NA proteins than two drugs targeting a single
protein. Recently, a clinical trial enrolling 217 patients with influenza A (H3N2) showed the
improved efficacy of a combined regimen of naproxen, oseltamivir and clarithromycin than
ACS Paragon Plus Environment
19

Journal of Medicinal Chemistry
Page 20 of 68
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
oseltamivir alone in reducing both 30ꢀ and 90ꢀday patient mortality and length of their hospital
4
1
stay .
Our present results suggest that the improved efficacy of the naproxen analogues may result from
disrupting the interactions involving two viral proteins, more efficient than inhibition of a single
viral target by disrupting NPꢀRNA interactions. While a biꢀtherapy directed against two viral
targets might be appropriate, a (bi or triꢀ) therapy combining antiviral and antiꢀinflammatory
action may afford a better protection, in particular in severed case of IAV infection or in patients
with allergic background (asthma) or immunocompromised status. Such therapies could combine
molecules such as derivatives 2 or 4 along with antagonists of receptors promoting the harmful
lung inflammation during influenza, such as the formylꢀpeptide receptorꢀ2 or with inhibitors of
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
2ꢀ45
NADPH oxidases 2 and 4
. Further work along these lines is in progress in our laboratories.
Experimental section
5
Molecular modelling: Naproxen was selected through virtual screening as described before .
The 3D crystal structure of NP monomeric form (PDB ID: 2IQH), was used as target throughout
the virtual screening in a volume centered on Y148. The MD simulations were carried out
5
46
following a protocol detailed elsewhere . The program NAMD with the CHARMM27 force
4
7
field was employed to perform the MD simulations. The NPꢀnaproxen analogues complex was
48
centered in a cubic cell of preꢀequilibrated TIP3P water model . The system was
electrostatically neutralized and set to an ionic strength of 0.15 M by adding sodium and chloride
ions. The electrostatic interactions were calculated using the particle mesh Ewald summation
4
9
algorithm . The equations of motion were iterated using the velocity Verlet integrator with a
ACS Paragon Plus Environment
20

Page 21 of 68
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
time step of 2 fs. An initial step of system energy minimization was followed by heating to 300
K. Molecular dynamics simulation in NVE and NPT ensemble were further employed to
equilibrate the system and for production run. Trajectories of 40 ns were produced for the NPꢀ
derivative 2, 3 and 4 complexes. For each trajectory, the displacement of the ligands was studied
by RMSD calculation. The structures with a close position (RMSD < 1 Å) were then grouped in
the same cluster. The representative structure (ie with the smallest average RMSD from all other
structures of the cluster) of the largest cluster of each complex was selected and then used for an
estimation of the free binding energy using the software Discovery Studio version 2016.
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
Estimation of the cavity in which the oligomerization loop inserts: the generation of the cavities
for the NPꢀderivative 4 complex, NP alone and the mutated R416A NP was carried out using the
5
0
cavity detection algorithm implemented in Discovery Studio version 2016 .
Drug-like properties of the naproxen analogues: the evaluation of six physicochemical
properties (lipophilicity, size, polarity, solubility, flexibility and saturation) for a rapid appraisal
3
6
of drugꢀlikeness was assessed according to .
Chemical synthesis:
1
13
Solvents were freshly distilled before use. H and C NMR spectra were recorded on a Bruker
Avance 300 using solvent residuals as internal references. Melting points (MP, uncorrected)
were determined on a Büchi Melting Point Bꢀ540. TLC analysis was carried out on silica gel
(Merck 60Fꢀ254) with visualization at 254 and 365 nm. Preparative flash chromatography was
carried out with Merck silica gel (Si 60, 40ꢀ63 µm). IR spectra were recorded with a FTꢀIR
PerkinElmer Spectrum 100 spectrometer. Mass spectrometry low resolution data (LRMS) and
high resolution data (HRMS) were obtained from the Centre d’Etudes Structurales et d’Analyses
ACS Paragon Plus Environment
21

Journal of Medicinal Chemistry
Page 22 of 68
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
des Molécules Organiques (CESAMO), Université de Bordeaux. The HPLC analysis was
performed with the following equipment and experimental conditions: HPLC Thermo Separation
Products Spectra SYSTEM (ThermoFisher), Software: ChromQuest, Isocratic elution: 50%
acetonitrile / 50% water + 0,1% formic acid by volume, Flow: 1,2 mL/min, Column: Pursuit 5
C18 250 x 4,6 mm, Pump: Spectra SYSTEM P1000XR, Detector: Spectra SYSTEM UV6000LP.
All naproxen derivatives had a purity ≥ 95%. The purity percentages of the two newly
synthesized and tested compounds have been determined by HPLC analysis to be 97% and 95%.
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
13
The details of the HPLC and high resolution H and C NMR data are provided in the
Supporting Information.
The naproxen analogues are protected by a patent WO2012143141 (PCT/EP2012/001720 and
5
1
US patent US14112987, grant US9783482B2, application US20180028478A1) .
Reagents: Reagents and chemicals for chemical synthesis were purchased from SigmaꢀAldrich.
NP primary monoclonal antibody (sc101352) and monoclonal antibody antiꢀCOX2 scꢀ376861
were purchased from Santa Cruz Biotechnology. Oligonucleotides: Biotinylated (5’ end) DNA
fragments were purchased from Dharmacon (ThermoꢀFischer, France) with HPLC purification;
the following sequences were used: Flu1: 5’ TTT GTT ACA CAC ACA CAC GCT GTG 3’.
The peptide PA(1ꢀ27) and a scramble peptide were purchased from Gencust (Luxemburg) at a
purification level over 95%. HPLCꢀpurified oligonucleotides were purchased from Eurofins and
MGW.
Peptide
Sequence
Length
ACS Paragon Plus Environment
22

Page 23 of 68
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
PA(1ꢀ27)
Scramble
Biot GSGS MED FVR QCF NPM IVE LAE KAM KEY GED
31
36
Biotꢀ GSGS MAI EKI SEI QRG SVF VDR LNV VET ISP DRR GQ
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
The Ni magnetic beads (Pure proteome) were purchased from Millipore. Naproxen, NaCl, Tris
buffer, guanidinium chloride, NaOH were from Sigma, MEM (minimal essential medium) media
6
from Lonza (Cologne, Germany). Derivative 2 was synthesized as previously described .
Virus Madin–Darby canine kidney (MDCK) cells were obtained from the American Type
Culture Collection. The influenza viruses A/PR/8/34 (H1N1), A/Udorn/72 (H3N2) were grown
in 8 days old embryonated hens’eggs and titered in MDCK cells. A/WSN/33 (H1N1) was
produced by reverse genetics: the 12 plasmids from a reverse genetics system were used as
described previously (32) to generate a recombinant virus of the influenza A/WSN/33 strain.
Briefly, 293T cells were transfected with eight plasmids encoding the individual vRNA segments
of influenza A/WSN/33 (H1N1) virus from a truncated human polymerase I promoter and four
expression plasmids encoding the corresponding subunits of the viral polymerase and the
nucleoprotein. The generated wild type (WT) virus was amplified on MadinꢀDarby canine
kidney cells and titrated by plaque assays. MDCK cells were cultured in minimal essential
medium (MEM, Sigma) containing 0.2% NaHCO (Sigma), MEM amino acids (Gibco), MEM
3
vitamins (Gibco), PSG, and 5% Newborn Calf Serum (ICN Biochem).
The Pandemic A/Netherlands/602/09 strain was a gift from Dr GF Rimmelzwaan, Research
Center for Emerging Infections and Zoonoses, University of Veterinary Medicine, Hannover
Bünteweg 17, 30559 Hannover, Germany.
ACS Paragon Plus Environment
23

Journal of Medicinal Chemistry
Page 24 of 68
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
6
–
Production of oseltamivirꢀresistant PR8 virus: In a 6ꢀwell plate, 1.5 × 10 MDCK cells were
added per well in DMEM 10% FBS medium. After 24 hours of incubation, the cells were
ꢀ3
infected with PR8 virus at MOI of 10 in the presence of 500 ꢁM oseltamivir in DMEM without
FBS. After 24 hours of incubation, the supernatant was taken and the virus was titrated. Then
this virus was used to repeat the same experience. After 4 cell passages, the virus became
resistant to oseltamivir.
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
0
Protein expression and purification of wt NP: Protein expression was performed as described
with modifications of the purification protocol by affinity chromatography using Ni magnetic
beads (Pure proteome, Millipore) and performed as recommended by the manufacturer, followed
by extensive dialysis at 50 mM NaCl, 20 mM Tris buffer at pH= 7.5. In these conditions, NP is
over >80% monomer. After purification, the protein concentration was determined by the
ꢀ1
ꢀ1
extinction coefficient ε = 56200 Μ .cm at 280 nm.
Dynamic light scattering: The measurements were performed on a Malvern Zetasizer nanoS
apparatus thermostated at 20°C. The size distribution was calibrated with latex particles of 65
and 200 nm radii. The scattering intensity data were processed using the instrumental software to
obtain the hydrodynamic diameter (Dh) and the size distribution of scatterers in each sample. A
total of 10 scans with an overall duration of 5 min were obtained for each sample and time point.
All data were analyzed in triplicate. Melting curves were obtained by recording the mean size as
a function of temperature in the range 20 to 70 to 80°C, using a temperature gradient of 1°C/
minute The Tm value was obtained from the maximum of the first derivative curve. The protein
concentrations usually were in the range of 5 to 15 µM. The samples for dynamic light scattering
were prepared in 20 mM Tris buffer at pH = 7.4 with 50 mM NaCl.
ACS Paragon Plus Environment
24

Page 25 of 68
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Surface plasmon resonance experiments: The binding kinetics were performed on a Biacore
T100 apparatus using streptavidin coated sensorꢀchips (SA, GE Heathcare). Capture of the
biotinylated peptide on the streptavidin coated sensorchip was carried out in running buffer. The
SPR kinetic measurements were performed in 20 mM TrisꢀHCl buffer, pH=7.4 at 25 °C,
containing 300 mM NaCl and 0.005% P20. Regeneration was performed by injecting guanidium
chloride during 30 sec, followed by two 30 seconds pulses of NaOH 10 mM. The peptides stock
solutions at a concentration of 250 µM were dissolved in water containing 1% DMF. The
peptides diluted in the running buffer at a concentration of 2.5 nM were injected for attachment
to the surface using the manufacturer protocol usually at a density of about 1500 RU. The NP
protein was injected at concentrations of 12 to 2500 nM in the presence or absence of naproxen
analogues, up to 50µM in some experiments. Measurements were conducted at 25°C and
samples were injected at 30 µl/min flow rate. The scramble and PA(1ꢀ27) peptides were attached
to the surface of the same chip at similar RU. To normalize SPR signal intensities, we probed
known interactions between a peptide P1 from Nꢀterminal of PB2 and a peptide P4 from PB1 Cꢀ
terminal 34 of similar size than that of PA(1ꢀ27) as a positive control. The signal that
corresponded to the P1ꢀP4 complex observed at saturation when P4 (MW 4185 Da) was injected
on immobilized P1 in the running buffer containing 300 mM NaCl and 0.005% P20 was 25 RU.
The signal of NP is expected to be proportional to the ratio of the molecular weights of NP
versus P4. NP adopts mainly a trimeric/ tetrameric form with an apparent MW of 197 Kda. Thus,
the signal at saturation for NPꢀPA(1ꢀ27) complex formation is expected to be 25 x 197 170/ 4185
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
=
1175 RU using P1ꢀP4 interaction as a reference. When 1.6 µM NP was injected over
immobilized PA(1ꢀ27) surface, the SPR signal corresponding to the NP ꢀ PA(1ꢀ27) complex
formation was found to be 1200 RU, in good accordance with a calculated Rmax.
ACS Paragon Plus Environment
25

Journal of Medicinal Chemistry
Page 26 of 68
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
Cell culture and infection: MadinꢀDarby canine kidney cells (MDCK) were purchased from the
American Type Culture Collection (Manasses, VA) and grown in MEM medium supplemented
with 10% fetal calf serum (FCS), 2 mM glutamine, and antibiotics. Cells were cultured at 37°C
in a 5% CO2 incubator. Before infection, cells were washed with FCSꢀfree medium and
incubated with A/WSN/1933 (H1N1) virus at multiplicity of infection of 2. Infected cells were
then incubated at 37°C until collection at 18 hours postꢀinfection. For WB, collected cell pellets
were homogenized in PBS containing antiꢀproteases (Roche), 1:1 volume. Both pellets and
supernatants were frozen at ꢀ80°C until WB probing. For measurements of NP levels by
immunofluorescence, the cells were labeled with a monoclonal antibody (sc101352).
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
Cell viability assay by MTT assay: A549 or MDCK cells (3.2x10 cells / well) were cultured in
1
2ꢀwell plates at 37 °C for 1 day in MEM without Fetal Calf Serum. Serial dilutions of naproxen
(1µM – 10 mM) were added to the cells that were further incubated at 37°C for 24, 48 hours.
Same culture conditions were used for the antiviral tests and toxicity measurements. MTT
analysis was performed based on the standard method. The absorbance was measured at 550 nm
with subtraction of the blank value at 650 nm using a Plate Reader (PerkinElmer). The presence
of naproxen or its derivatives in the supernatants did not affect the readings within experimental
error. The data at 24 h were best fitted by a doseꢀresponse curve The toxicity data (2 µM – 30
mM) were fitted with a doseꢀresponse curve (Origin, v8.5 software) to determine the CC .
5
0
5
Cell viability by Trypan blue dye exclusion method: MDCK cell cultures (3.2 x 10 cells /well)
were grown in 12ꢀwell tissue culture plates. After a 24 h period of incubation, the same
procedure for the cytotoxicity assay described above was followed. After 48 h the supernatant
was removed, cells were washed and trypsinized, and Trypan blue dye aqueous solution was
added to a cell suspension. Viable and dead cells were counted with a phase contrast microscope.
ACS Paragon Plus Environment
26

Page 27 of 68
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Pulldown assays: Infected MDCK cells were harvested and resuspended in iceꢀcold PBS buffer
supplemented with a complete protease inhibitor cocktail (Roche). After addition of NP40
surfactant (0.5%), the cells were lysed by freezing at ꢀ80°C. For pulldown assays streptavidinꢀ
coated magnetic beads, SAꢀMBs, (Dynabeads Mꢀ280 Streptavidin, Invitrogen) were incubated
with 250µM PA(1ꢀ27) in PBS for 30 min at room temperature. PA(1ꢀ27)ꢀSAꢀMBs were
separated with a magnet and washed in PBS. The beads were resuspended in the lysates of
infected cells for 18 hours at 4°C. The beads were extensively washed with PBS buffer, boiled in
Laemmli sample buffer, and analyzed by SDSꢀPAGE or immunoblotting. Negative controls were
performed by incubating PA(1ꢀ27)ꢀSAꢀMBs with lysates of nonꢀinfected MDCK cells, or
incubated beads carrying no peptide (SAꢀMB), or carrying a biotinylated scramble peptide
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
(PScrꢀSAꢀMB), with lysis of infected cells.
Immunoblotting: Proteins were eluted from SAꢀMB in Laemmli buffer at 95°C and subjected to
SDSꢀPAGE. Then, proteins were transferred to nitrocellulose membranes. The membranes were
incubated in a blocking solution (PBS–0.1% Tween 20 with 5% nonfat milk) for 3 h. The blots
were rinsed with PBS–0.1% (vol/vol) Tween 20 and incubated overnight at 4°C with primary
antiꢀNP monoclonal antibody (Santa Cruz Biotech). The membranes were rinsed and incubated
with the antiꢀmouse horseradish peroxidase (HRP)ꢀconjugated secondary antibody.
Immunodetection was performed using the Clarity Western ECL substrate (BioꢀRad).
6
Antiviral Effect: MDCK or A549 cells (0.32 x 10 cells / well) were cultured in 5% CO for 24
2
hours to 80% confluence as described above without Fetal Calf Serum. Serumꢀstarved A549
ꢀ3
cells or MDCK cells were infected with A/WSN/33 virus at 10 MOI (multi cycle growth assay)
and naproxen or naproxen analogues (0.1 ꢀ 500 ꢁM) or vehicle were added to the cells
immediately after virus adsorption. Some experiments were also performed at higher multiplicity
ACS Paragon Plus Environment
27

Journal of Medicinal Chemistry
Page 28 of 68
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
of infection MOI = 2 during 18 hours or by replacing Influenza/A/WSN/33 (H1N1) by
A/Udorn/72 (H3N2). Viral titers of the clarified cell supernatants were determined by plaque
assay using crystal violet readꢀout at different times as indicated. The viral load was also
measured by the level of viral NP (MOI =2) by immunofluorescence using an NP antibody. The
experiments were carried out in triplicate and repeated at least twice for confirmation. The IC50
was determined by fitting the data to a dose –response curve (Origin version 8.5). In addition, the
same method used to evaluate cell viability as described above was followed to determine the
CC50 by MTT tests.
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
6
MDCK cells (0.32 x 10 cells / well) were cultured in 5% CO for 24 hours to 80% confluence
2
without Fetal Calf Serum. Serumꢀstarved MDCK cells were infected with the 2009 pandemic
ꢀ3
strain or Oseltamivirꢀresistant PR8 virus at MOI 10 during 48 hours. Immediately after
infection, the cells were treated or not with the naproxen analogues and used in the concentration
range of 2ꢀ200 µM. After fixation and treatment with triton, the cells were labeled with a
monoclonal NP antibody overnight and labeled with a secondary red antibody. The viral load
was assessed by the level of viral NP by immunofluorescence. The experiments were carried out
in triplicate and repeated at least twice for confirmation. The NP and DAPI levels were
quantified using Image J by determining the raw intensity in each cell subtracted by the raw
intensity of the background near the cell. The results were averaged on 10ꢀ20 cells. The NP
levels in treated cells by derivatives 2, 3 or 4 compared to untreated cells were tested for
statistical relevance by ANOVA tests. All statistical analyses were performed using GraphPad
Prism.
Determination of COX2 levels by IF: A549 cells were washed twice in DMEM without phenol
red and the derivatives were added to the cells and incubated for 30 minutes. Then, viral
ACS Paragon Plus Environment
28

Page 29 of 68
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
8
inoculum at 5x 10 pfu/ml was added to the cells (MOI 2). Alternatively, the cells were activated
by 100 µM γꢀATP for one hour in the presence or absence of the derivatives. The cell
supernatants were removed, the cells fixed with 3% PFA in PBS for 15 minutes, washed twice
and permeabilized with 0.1% triton in PBS + 3% BSA, incubated 15 minutes, rinsed twice and
labeled with the primary antibody anti COX2 (ab15191) overnight, then washed twice and
labeled with Alexa 647 secondary antibody.
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
In vivo experiments: Five to sixꢀweekꢀold female C57BL/6J mice (Charles Rivers) were
anesthetized with Ketamine/ Xylazine (43/5 mg/kg) and inoculated intranasally with 20 ꢁl of a
solution containing A/PR/8/34 virus (500 PFU). Five to ten minutes after virus inoculation, the
mice were treated once a day at days 0, 1 and 2. Five mice groups, 10 mice in each group, were
treated intraperitoneally: serum, naproxen 1mg/day; derivative 2 at doses of 1, 4 and 8 mg/day.
Survival rates and loss of body weight were scored daily. Twenty percent weight loss was used
as the end point in the mortality curve following the European Ethics committee
recommendations. At the end of the experiment, mice were sacrificed by cervical dislocation.
For assessing virus replication, bronchoꢀalveolar lavage (BAL) was harvested from the sacrificed
mice, and infectious virus titers were determined by plaque assay. Tamiflu obtained from a
pharmacy (0.2 mg dissolved in PBS) was used as a positive control. The protocol was approved
by the committee of animal experiments of the Faculty of Marseille la Timone (number:
G130555). All animal experiments were also carried out under the authority of a license issued
by “la direction des services Vétérinaires” (accreditation 693881479). All statistical analyses
were performed using GraphPad Prism. To assess viral replication, we used the MannꢀWhitney
test. The survival curves were analyzed with the Rank test (MantelꢀCox). The analysis of
ACS Paragon Plus Environment
29