System-oriented optimization of multi-target 2,6-diaminopurine derivatives: Easily accessible broad-spectrum antivirals active against flaviviruses, influenza virus and SARS-CoV-2

European Journal of Medicinal Chemistry 224 (2021) 1 13683

Contents lists available at ScienceDirect

European Journal of Medicinal Chemistry

journal homepage: h ttp://www.elsevier.com/locate/ejmech

System-oriented optimization of multi-target 2,6-diaminopurine

derivatives: Easily accessible broad-spectrum antivirals active against

ﬂaviviruses, inﬂuenza virus and SARS-CoV-2

,

Ilaria Vicenti ^{a 1}, Maria Grazia Martina ^{b 1}, Adele Boccuto ^a, Marta De Angelis ^c,

Giorgia Giavarini ^b, Filippo Dragoni ^a, Serena Marchi ^d, Claudia Maria Trombetta ^d,

Emmanuele Crespan ^e, Giovanni Maga ^e, Cecilia Eydoux ^f, Etienne Decroly ^f,

Emanuele Montomoli ^d^g, Lucia Nencioni ^c, Maurizio Zazzi ^a, Marco Radi ^b

^aDepartment of Medical Biotechnologies, University of Siena, Siena, Italy

,

, *

b

ꢀ

Dipartimento di Scienze degli Alimenti e del Farmaco, Universita degli Studi di Parma, Viale delle Scienze, 27/A, 43124, Parma, Italy

^cDepartment of Public Health and Infectious Diseases, Laboratory Afﬁliated to Istituto Pasteur Italia-Fondazione Cenci Bolognetti, Sapienza University of

Rome, Piazzale Aldo Moro 5, 00185, Rome, Italy

^dDepartment of Molecular and Developmental Medicine, University of Siena, Siena, Italy

^eIstituto di Genetica Molecolare, IGM-CNR “Luigi Luca Cavalli-Sforza”, Via Abbiategrasso 207, 27100, Pavia, Italy

f

ꢁ

AFMB, CNRS, Universite Aix-Marseille, UMR 7257, Marseille, France

^gVisMederi s.r.l, Strada del Petriccio e Belriguardo 35, 53100 Siena, Italy

a r t i c l e i n f o

a b s t r a c t

Article history:

The worldwide circulation of different viruses coupled with the increased frequency and diversity of new

outbreaks, strongly highlight the need for new antiviral drugs to quickly react against potential pandemic

pathogens. Broad-spectrum antiviral agents (BSAAs) represent the ideal option for a prompt response

against multiple viruses, new and re-emerging. Starting from previously identiﬁed anti-ﬂavivirus hits, we

report herein the identiﬁcation of promising BSAAs by submitting the multi-target 2,6-diaminopurine

chemotype to a system-oriented optimization based on phenotypic screening on cell cultures infected

with different viruses. Among the synthesized compounds, 6i showed low micromolar potency against

Received 21 May 2021

Received in revised form

16 June 2021

Accepted 1 July 2021

Available online 5 July 2021

Keywords:

Dengue, Zika, West Nile and Inﬂuenza A viruses (IC₅₀¼ 0.5e5.3

mM) with high selectivity index. Inter-

estingly, 6i also inhibited SARS-CoV-2 replication in different cell lines, with higher potency on Calu-

Diaminopurine

Broad-spectrum antivirals

Dengue

Zika

West nile virus

Inﬂuenza

3 cells that better mimic the SARS-CoV-2 infection in vivo (IC₅₀¼ 0.5 M, SI ¼ 240). The multi-target

m

effect of 6i on ﬂavivirus replication was also analyzed in whole cell studies (in vitro selection and

immunoﬂuorescence) and against isolated host/viral targets.

SARS-CoV-2

1. Introduction

prevented with vaccines [1,2]. In 2018, the Global Virome Project

estimated that our planet harbours around 1.6 million yet-to-be-

Viruses make up roughly two-thirds of all newly discovered

human pathogens d far more than either bacteria or fungi. Despite

much social visibility and alertness to the potential impact of

widespread epidemics of devastating viral infections, a signiﬁcant

number of viral pathogens remains without effective treatment or

cure, whilst only a few major viral infections (except HIV) can be

discovered potentially zoonotic viruses, and less than 1% of these

has been identiﬁed to date [3]. In addition, old viruses previously

conﬁned in speciﬁc and isolated areas are re-emerging and rapidly

spreading as a consequence of globalization and climate changes,

but also those viruses that were thought to be eradicated with

vaccination campaigns are reappearing due to public misinforma-

tion, no-vax campaigns and geopolitical instability in speciﬁc ter-

ritories. As human activity takes people into new environments or

down to anti-scientiﬁc pathways, the risk of zoonotic virus spillover

into humans increases and, along with that, the possibility of new

epidemic outbreaks in the coming decades. The ongoing severe

ꢀ

* Corresponding author. Dipartimento di Farmacia, Universita degli Studi di

Parma, Viale delle Scienze, 27/A, 43124, Parma, Italy.

E-mail address: marco.radi@unipr.it (M. Radi).

These authors equally contributed to the present work.

1

https://doi.org/10.1016/j.ejmech.2021.113683

I. Vicenti, M.G. Martina, A. Boccuto et al.

European Journal of Medicinal Chemistry 224 (2021) 113683

Abbreviations

RSA

radical scavenging activity

CPE

cytopathic effect

BSAA

DENV

ZIKV

broad-spectrum antiviral agent

Dengue virus

Zika virus

RdRp

RNA-dependent RNA-polymerase

PR8 H1N1 inﬂuenza A Puerto Rico/8/34H1N1 virus

p2009H1N1 pandemic A/California/04/09H1N1 virus

Ulster H7N1 avian A/parrot/Ulster/73H7N1

WNV

West Nile virus

COVID-19 coronavirus disease 2019

ICW

MTT

In Cell Western assay

CCR5

SAR

CeC chemokine receptor type 5

structureꢀactivity relationship

phenotypic drug discovery

target-based drug discovery

secondary yield reduction assay

direct yield reduction assay

plaque reduction assay

[3-(4,5-dimethylthiazol-2-yl)-2,5-

diphenyltetrazolium bromide] assay

Hemagglutinin

PDD

TDD

SYRA

DYRA

PRA

IA

HA

FLU

CC₅₀

IC₅₀

SI

inﬂuenza

50% cytotoxic concentration

50% inhibitory concentration

Selectivity Index

immunodetection assay

coronavirus disease 2019 (COVID-19) pandemic caused by the new

severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) is

just one of the possible outcome, but emerging widespread viruses

with moderate mortality rate (e.g. Flaviviruses) may also represent

a serious epidemic threat [4]. In fact, viruses are subject to

continuous evolution and new variants may become more virulent

and/or acquire different tissue tropism, as recently documented for

Zika virus (ZIKV) isolates causing microcephaly and other congen-

ital anomalies in fetus as well as neurologic disorders in adults. As

an example, a single casual mutation (S139 N) in ZIKV has been

recently linked to an increased infectivity in vitro, to more severe

microcephaly and higher mortality rates in neonatal mice [5]. The

National Institute of Allergy and Infectious Diseases (NIAID) has

therefore classiﬁed Dengue virus (DENV), West Nile virus (WNV),

and ZIKV as categories A and B emerging infectious pathogens,

accounting for 17% of infections worldwide and for which no drugs

or universally protective vaccines are currently available. The

epidemic potential of these ﬂaviviruses is well represented by the

increasing number of autochthonous vector-borne ﬂavivirus in-

fections in the last decade [6,7]. An additional problem with ﬂavi-

virus infections is represented by coinfections, not only by different

viruses in the same family (e.g. DENV/ZIKV [8,9]), but also by

different virus families (e.g. DENV/SARS-CoV-2 [10,1 1] or DENV/

inﬂuenza [12]), which creates immune response disorders, confu-

sion in diagnosis and delays in treatment, leading to additive or

synergistic morbidities [13,14]. These diagnostic and therapeutic

issues are even more serious in the case of newly emerging

potentially pandemic viruses or in case of emerging variants with

higher mortality rate since the time needed to develop a speciﬁc

vaccine or drug is not compatible with a prompt response. There-

fore, development of pan-viral drugs (broad-spectrum antiviral

agents; BSAAs) capable of inhibiting the replication of multiple

viruses belonging to different families, is an ambitious but attrac-

tive option in preparedness for outbreaks of previously unknown

viral diseases and also for the treatment of widespread viral in-

fections where an effective drug is not available yet. However, to

date, the rarely used and mildly effective Ribavirin (RBV) is the only

BSAA approved for the treatment of infections caused by different

virus families. Other nucleoside antivirals, despite acting on the

highly conserved viral transcription process, are approved for the

treatment of infections caused by a single virus family. On the other

hand, a number of experimental non-nucleoside BSAAs with

mechanism of action (MOA) that is often complementary to that of

nucleoside inhibitors are also under clinical development for the

treatment of multiple virus families, even if sometime their MOA is

not fully understood due to their effect on multiple targets (e.g.

Nitazoxanide, [71] Arbidol [72]) (Fig. 1).

To come up with new antivirals needed to ﬁght the growing

circulation of different viruses, the antiviral drug discovery com-

munity has been gradually revising its original “one-drug, one vi-

rus” dogma, based on targeting speciﬁc viral proteins with a single

drug, passing through different paradigm shifts: i) the “one-drug,

multiple viruses” paradigm, based on the design of molecules tar-

geting viral proteins highly conserved among different viruses (e.g.

RNA/DNA polymerases); ii) the “host-targeting antivirals” para-

digm, based on targeting host-proteins (e.g. CCR5 with Maraviroc)

used by many viruses for their replication [15]; iii) the “hypothesis-

free drug repositioning” approach, based on testing safe in man

compounds as antiviral candidates, independently from their

original indication, for an accelerated approval. The latter approach

is being strongly pursued to quickly discover new antivirals for

emerging diseases, despite no clinical success has been reported in

the last two decades [16]. A recent addition to the ﬁeld is the “one-

drug, multiple-targets” paradigm (polypharmacology), which rep-

resents a more realistic approach that takes into account the high

number of on- and off-targets interacting with a drug candidate

and the presence of compensatory pathways that may limit the

effectiveness of highly speciﬁc drugs [1 7,18]. In fact, it is not un-

common that looking back at old antivirals (e.g. Ribavirin; Fig. 1)

with this new awareness, new pharmacological targets or mecha-

nisms are found involved in determining the observed therapeutic

effect and, possibly, the pan-antiviral activity [19]. Partial inhibition

of multiple targets with a single multi-target drug seems more

effective than full inhibition of a single target, reducing on- and off-

target-related toxicity and attrition rate in clinical development

[1 7]. In line with this paradigm shift in antiviral drug discovery, our

research group has successfully reported the development of multi-

target antivirals active on multiple viruses and related diseases

in vitro [20e23].

As a continuation of our previous work in this ﬁeld, we describe

herein a phenotypic-based study focused at the optimization of 2,6-

diaminopurine multi-target anti-ﬂavivirus agents (e.g. 1e3, Fig. 1),

[20,22] to identify new easily accessible BSAAs as suitable candi-

dates for the treatment of multiple viral infections. Compound 6i

was identiﬁed as the most promising BSAA able to reduce the

replication of viruses belonging to different families (DENV, ZIKV,

WNV, Inﬂuenza A, SARS-CoV-2) at low micromolar concentration

while the reference BSAA Ribavirin does not show signiﬁcant ac-

tivity against Inﬂuenza A virus and SARS-CoV-2 [34,36]. Despite the

full characterization of the antiviral MOA of 6i against all viruses

investigated is not the scope of the present work, we also con-

ducted a series of biochemical studies (in whole cell and against

2

I. Vicenti, M.G. Martina, A. Boccuto et al.

European Journal of Medicinal Chemistry 224 (2021) 113683

Fig. 1. Selection of approved broad-spectrum antiviral agents (BSAAs), experimental BSAAs, 2,6-diaminopurine-based antivirals 1e3 and target BSAAs reported in this work.

isolated targets) to conﬁrm the original multi-target hypothesis for

the inhibition of Flavivirus replication.

by hydrolysis of the corresponding methyl esters 6d,f. In addition,

we also prepared the two “inverted” derivatives of compounds 6c

and 6i where C2 and C6 substituents were exchanged: starting

from 4, the phenoxyethylamino chain was initially introduced in C6

(compound 7) followed by C2 nucleophilic substitution with the

opportune substituted aniline to give the desired “inverted” com-

pounds 8a,b (Scheme 1).

2. Results and discussion

By exploiting the complementary advantages of target-based

drug discovery (TDD) and phenotypic drug discovery (PDD), a

ﬁrst generation of purine-based multi-target antivirals (e.g. com-

pounds 1e2, Fig. 1) was previously reported by us [20]. These

compounds proved to inhibit DENV replication (the only affected

virus), blocking the interaction between the recombinant viral

proteins NS5/NS3 and inhibiting, at the same time, the activity of

host c-Src/Fyn kinases. The following target-based optimization

based on the allosteric cavity B of NS5 led to a second generation of

derivatives (e.g. compound 3, Fig. 1) with a wider spectrum of ac-

tivity, blocking the replication of all four DENV serotypes and ZIKV

[22]. Unfortunately, this targeted optimization did not allow to

improve the antiviral potency of the second generation derivatives

as we expected from the modeling predictions. A summary of

antiviral potency for compounds 1e3 is reported in Fig. 1. Here we

reasoned that a more efﬁcient approach to optimize multi-target

antivirals could be based on phenotypic considerations resulting

from functional screening of focused derivatives since: i) the anti-

viral effect in a cellular context may be the result of a synergic

activity on multiple targets; ii) unexpected targets or novel modes

of action can be responsible for the antiviral effect. To further

explore the antiviral potential of the 2,6-diaminopurine chemotype

we planned to build a new series of derivatives by combining the C2

substituents of the hit compounds 1e3 with different substituted

anilines in C6 (Scheme 1). Compounds 6a-k were thus obtained by

applying the fast and efﬁcient two-step microwave-assisted pro-

tocol previously developed, starting from commercially available

2,6-dichloropurine (4) [20]. Compounds 6l,m were then obtained

Compounds 6a-m and 8a,b where then analyzed for their anti-

DENV and anti-ZIKV activity in a Huh7 hepatoma cell-based sec-

ondary yield reduction assay (SYRA) [24] which quantiﬁes infec-

tious culture supernatants by immunodetection (Table 1). All non-

toxic compounds tested showed promising antiviral activity against

both DENV and ZIKV. Within the 3⁰-F and 3⁰-CF₃series (compounds

6a-c and 6g-i, respectively) the selectivity index (SI) with both vi-

ruses increased with the length of the C2 side chain. The 3⁰-CO₂Me

series (compounds 6d-f) was characterized by a similar potency on

ZIKV and, surprisingly, the 2-phenoxyethylamino derivative 6f did

not show any activity on DENV. Compared to the 3⁰-CO₂Me series,

the 3⁰-CO₂H series (compounds 6l-m) showed a reduced potency

and opposite speciﬁcity since the 2-phenoxyethylamino derivative

6m was active against both DENV and ZIKV (compared to 6f) while

the 2-methylamino derivative 6l was active only against ZIKV

(compared to 6d). Also the 3⁰,4⁰-OH derivative 6j was active only

against ZIKV while the trimethoxy derivative 6k was quite cyto-

toxic. It was also interesting to note that the two “inverted” com-

pounds 8a,b, analogues of the active compounds 6c and 6i, showed

high cytotoxicity and were not evaluated for antiviral potency. Out

of these 15 tested compounds, 6i showed the most potent antiviral

activity against both ZIKV (SI 182) and DENV (SI 77) with IC₅₀values

in the low micromolar range, comparable to or slightly better than

those of the reference drugs Sofosbuvir and Ribavirin. Based on

these results and on those reported by us for the previous gener-

ations of 2,6-diaminopurine antivirals,

a few rough SAR

3

I. Vicenti, M.G. Martina, A. Boccuto et al.

European Journal of Medicinal Chemistry 224 (2021) 113683

a

Scheme 1. Reagents and conditions: i.) amines, n-BuOH, Et₃N, 70e100 ^ꢁC, 10e50 min,

m

W; ii.) amines, n-BuOH, TFA, 150e170 ^ꢁC, 40e70 min,

m

W; iii.) 6d or 6f, LiOH, THF/MeOH, r.t.,

24 h (for 6l,m).

Table 1

DENV-2/ZIKV/WNV replication inhibitory effect.

DENV-2

ZIKV

WNV

Cpd.

CC^a₅₀

(

m

M)

IC^b₅₀

(

m

M)

SI^c

IC₅₀

(

m

M)

SI

IC₅₀

(m

M)

SI

6a

6b

6c

6d

6e

6f

84.0

9.3 4.2

29.8 0.3

12.0 0.2

18.0 12.2

17.5 7.3

NA^d

3.5 4.0

(4.9 2.6)^f

4.5 0.2

9

5

17

11

e

14.7 5.7

8.9 0.9

2.9 0.9

8.0 1.5

6.2 2.5

8.2 1.5

ND

6

ND

e

136.0

200.0

100.0

200

15

68

25

32

12

e

6g

57

41

11

(NA)^f

ND

6h

6i

50.0

10.6 0.5

5

e

(1.9 2.9)^f

(1.5 1.3)^g

ND

26

33

e

200.0

2.6 0.3

(NA)^f

77

-

222

e

1.1 0.4

(0.8 0.3)^f

(0.5 0.32)^g

25.0 7.1

ND

182

250

400

8

(NA)^f

e

(0.9 0.04)^g

NA

(3.9 0.67)^g

ND

51

e

6j

6k

200.0

7.5

ND

-

e

ND

e

6l

6m

8a

200.0

8.0

NA

28.2 3.9

ND

e

33.5 7.6

29.0 4.8

ND

6

7

ND

e

7

-

e

8b

6.0

ND

-

ND

e

ND

e

Sofosbuvir

200.0

4.7 0.7

(4.6 1.4)^f

(3.8 1.1)^g

4.0 0.6

(7.6 1.2)^f

(4.1 0.6)^g

42

43

53

50

26

49

3.2 0.7

(2.7 0.5)^f

(2.0 1.1)^g

4.4 0.6

(3.8 0.6)^f

(2.2 0.6)^g

62

74

100

45

53

91

ND

e

(1.7 0.5)^f

(1.2 0.3)^g

ND

117

166

e

Ribavirin

200.0

(9.5 4.0)^f

(6.7 0.6)^g

21

30

^aCC₅₀: half-maximal cytotoxic concentration ( M); ^bIC₅₀: half-maximal inhibitory concentration calculated through SYRA; each value is the mean of three experiments

m

standard deviation (SD); ^cSI: selectivity index (SI ¼ CC₅₀/IC₅₀). ^dNA: not active; ^eND: not determined; ^fIC₅₀: half-maximal inhibitory concentration calculated through DYRA;

^gIC₅₀: half-maximal inhibitory concentration calculated through PRA. DMSO was used as control in all cytotoxicity and antiviral assays.

considerations can be drawn: i) substituted anilines in C2 lead to

cytotoxic compounds (e.g. 8a,b and 12c-e ref. [22]); ii) cyclic sec-

ondary amines in C2 lead to inactive or poorly active antivirals (e.g.

12f ref. [22], 16a-c ref. [20]); iii) increasing the length of the C2

substituent (primary amines) increase the selectivity index and the

antiviral spectrum; iv) different functional groups on the C6 aniline

moiety are well tolerated for the antiviral activity (see Table 1), with

a few exceptions (e.g. 6k).

yield reduction assay (DYRA) data can be used to obtain preliminary

information about early versus late antiviral activity of drug can-

didates [24e26]. Compound 6i was thus analyzed through DYRA

against ZIKV and DENV replication. As shown in Table 1, 6i

conﬁrmed the antiviral potency seen in SYRA against ZIKV but it did

not show any antiviral activity against DENV in DYRA. These results

suggest that 6i exerts the antiviral activity against DENV not at early

stages of viral replication and probably at a later stage than

Sofosbuvir, which showed the same potency in SYRA and DYRA, as

As previously reported, comparison between SYRA and direct

4

I. Vicenti, M.G. Martina, A. Boccuto et al.

European Journal of Medicinal Chemistry 224 (2021) 113683

expected.

decolorization method in comparison with 11a, 6i and with a

The anti-DENV efﬁcacy of 6i was ﬁnally conﬁrmed by plaque

potent antioxidant molecule (ascorbic acid) used as reference. [33].

reduction assay (PRA), giving an IC₅₀of 0.90 0.04

mM (in agree-

As reported in Fig. 2, compound 10a showed a clear

ment with SYRA) that is four-fold the IC₅₀and SI of both Sofosbuvir

and Ribavirin. Curiously, 6i showed low micromolar potency in

SYRA/PRA but also in DYRA against ZIKV, suggesting an effect on

replication of this virus at an earlier stage compared to DENV. The

antiviral activity of the most potent DENV inhibitors 6h and 6i was

also assessed against WNV through DYRA and PRA in comparison

with Sofosbuvir/Ribavirin. All compounds were active in the low

micromolar range but, again, 6i proved to inhibit WNV replication

in PRA and not in DYRA, matching the result obtained with DENV.

Overall, this phenotypic optimization led to the identiﬁcation of 6i

as the most interesting broad-spectrum anti-ﬂavivirus candidate,

whose potency and SI was generally superior to that of Sofosbuvir

and Ribavirin.

In the attempt to further investigate and expand the multi-

target propensity of the 2,6-diaminopurine class of molecules

(generally referred to as privileged structure), we aimed at intro-

ducing small structural modiﬁcations that may confer additional

antioxidant properties to 6i and closely related analogues. The in-

duction of oxidative stress is in fact recognized as a common trigger

of exacerbation in diseases caused by different viruses, including

ﬂavivirus, coronavirus and, particularly, Inﬂuenza virus [27]. A few

literature reports show that opportune functionalization of the

purine scaffold can lead to derivatives also endowed with antioxi-

dant properties by scavenging reactive oxygen species (ROS)

[28e30]. Among the many possible functionalization options, we

were intrigued by a recent work from the Feringa group, describing

the development of photoswitchable purine derivatives bearing an

azobenzene moiety in C6 [31]. The synthesis of these compounds

pass through a 6-phenylhydrazino intermediate that is easily con-

verted into the desired azobenzene derivative by O₂-mediated

oxidation: although the hydrazine motif is generally considered a

toxicophore moiety, we thought that opportunely substituted C6-

phenylhydrazino derivatives may counteract the virus-induced

oxidative stress converting themselves into stable azo-derivatives.

This choice was also supported by the low cytotoxicity of 6-

phenylhydrazino purines previously reported as anti-malaria

agents [32]. The synthesis of our target compounds 10a-e fol-

lowed the same synthetic sequence used for the synthesis of 6a-m:

concentration-dependent radical scavenging activity that was not

observed for 11a and 6i. In addition, the stability of 11a towards

oxidative stress was further assessed by reaction with H₂O₂in

argon atmosphere: while compound 10a was quickly converted

into 11a, the latter compound and 6i were stable for days in the

presence of 40 mM H₂O₂in ethanol (data not shown).

To further explore the spectrum of antiviral activity of the most

promising inhibitors herein reported, compounds 6c, 6i, 10a were

also evaluated against a segmented negative RNA virus (Inﬂuenza A

Puerto Rico/8/34H1N1, PR8). MDCK and A549 cell lines were used

to evaluate compounds’ cytotoxicity and antiviral potency. The

BSAA Ribavirin (RBV) is reported in Table 3 as reference. [34]. The

selected compounds did not display cytotoxicity on both cell lines

and, by comparing the CC₅₀of Ribavirin, the three compounds

showed higher CC₅₀values. Regarding the antiviral effect, 6c, 6i and

10a inhibited the replication of Inﬂuenza A virus at low micromolar

concentration. In particular, 6i and 10a showed higher SI and IC₅₀

with respect to 6c, and a stronger anti-inﬂuenza activity in com-

parison with RBV. Since compound 10a inhibited replication when

added after the infection and due to its antioxidants activity, we

hypothesize its role in maintaining reducing conditions into the

cells. In fact it has been demonstrated that inﬂuenza virus induces

redox changes in terms of ROS production and glutathione deple-

tion, to activate redox-sensitive pathways useful for viral replica-

tion [35]. Further studies are in progress to evaluate this hypothesis.

To get further insights on the effect of compound 6i on Inﬂuenza

virus replication, it was added at different times: i) 2 h before

infection (PRE); ii) for 1 h during viral adsorption (DUR); iii) after

viral infection for 24 h (POST) or iv) before, during and after

infection for the following 24 h (PDP). Supernatants from infected

cells were recovered at 24 h from infection and used to infect a

fresh cell monolayer. The expression of viral protein (hemaggluti-

nin; HA) was quantiﬁed as percentage of relative ﬂuorescence units

(RFU) analyzed by ICW, as described in Methods Section. As shown

in Fig. 3A, the percentage of HA was similarly reduced (50% inhi-

bition) when 6i was added during (DUR) or post infection (POST).

On the contrary, pretreatment alone (PRE) was not quite effective.

Next, speciﬁc steps of infection were evaluated by using a time-of-

addition assay. In particular, compound 6i was added: i) immedi-

ately after infection for 2, 4 and 6 h on A549 cell monolayers and

then maintaining cells in fresh medium without the compound

until 24 h from infection; ii) after 2, 4 and 6 h from infection and

maintaining the compound until 24 h from infection. Supernatants

were recovered and used to infect fresh monolayers of MDCK cells.

As shown in Fig. 3B, the percentage of HA expression was mostly

reduced when 6i was added within the ﬁrst 6 h from infection.

Indeed, if the compound was added after 6 h, we did not ﬁnd a

signiﬁcant inhibition, suggesting that the early phases of infection

were impaired by 6i addition.

microwave-assisted C6 nucleophilic substitution on

4 with

substituted phenylhydrazines afforded intermediates 9a-e that

were ﬁnally converted into 10a-e by reaction with phenoxyethyl-

amine at high temperature for 90 min (Scheme 2). The latter

compounds were slowly converted into stable aza-derivatives 11a-

d by simply stirring their methanolic solution at room temperature

under air for 96 h. It was interesting to note that, under the same

oxidative conditions, the 3-CF₃derivative 11e tend to decompose

and the corresponding azo-derivative 11e was not isolated.

Compounds 10a-e and 11a-d were then submitted for biological

evaluation against ﬂaviviruses. Unfortunately, most of the com-

pounds displayed high cytotoxicity in Huh7 cells and were not

tested for inhibition of virus replication. Only 3⁰-F derivatives 10a

and 11a, which showed acceptable cytotoxic effect, were analyzed

for their anti-ﬂavivirus potency (Table 2): both compounds dis-

played low micromolar activity against DENV, ZIKV and WNV with

the highest SI against the latter virus. Interestingly, the target

compound 10a and the corresponding oxidized derivative 11a

showed the same antiviral activity, supporting the idea that 10a

may exert a synergic antiviral/antioxidant activity in infected cells

and then being converted into a stable metabolite (11a) that still

displays antiviral activity (dual action). The potential antioxidant

effect of compound 10a was then experimentally evaluated in

terms of ‘radical scavenging activity’ (RSA) using a ABTS radical

The expression of hemagglutinin (HA) was analyzed by ICW,

using LI-COR Image Studio Software, as described in Methods. The

percentage of relative ﬂuorescence units (RFU) was calculated in

comparison to untreated-infected cells (PR8 H1N1) (considered

100%). Data represent the mean

S.D. of 3 replicates (A) or 4

replicates (B) obtained from two experiments performed. Statistical

signiﬁcance of the data vs PR8 H1N1 was deﬁned as *P < 0.05 and

**P < 0.01.

To evaluate whether 6i was effective against other inﬂuenza

strains, we performed similar treatments on A549 cells infected

with three different strains: the human PR8 H1N1, the human

pandemic 2009H1N1 or the avian Ulster H7N1 as described in

Methods. In the ﬁrst set of experiments, cell monolayers were

5

I. Vicenti, M.G. Martina, A. Boccuto et al.

European Journal of Medicinal Chemistry 224 (2021) 113683

a

Scheme 2. Reagents and conditions: i.) amines, n-BuOH, Et₃N, 70 ^ꢁC, 20e40 min,

m

W; ii.) phenoxyethylamine, n-BuOH, TFA, 170 ^ꢁC, 120 min,

mW; iii.) MeOH, air, r.t., stirring, 96 h.

Table 2

Fig. 4A, the addition of 6i inhibited viral proteins expression at all

the conditions performed, even though the treatment with the

compound after infection (POST) and before, during and after

infection (PDP) was the most effective. Then, A549 cells were

infected with the three strains and treated with 6i at different times

from infection, as described before. Supernatants of treated cells

were recovered after 24 h infection and used to infect fresh MDCK

monolayers. As shown in Fig. 4B, the treatment with 6i was effec-

tive when added in the ﬁrst 6 h from infection, thus suggesting an

impairment in the early phases of virus life-cycle. Nevertheless, we

cannot exclude that the compound partly affects host cell mem-

brane [as indicated by about 20e26% inhibition during the pre-

treatment (PRE) of cell monolayers] or the viral attach/entry

[28e45% inhibition during the viral adsorption phase (DUR)]

(Figs. 3A and 4A). Further studies are needed to ascertain those

steps of viral infection.

The expression of FLU proteins was analyzed by ICW, using LI-

COR Image Studio Software. The percentage of relative ﬂuores-

cence units (RFU) was calculated in comparison to untreated-

infected cells (considered 100%). Values are the mean S.D. of 2

replicates (A) or 3 replicates (B) obtained from one of the two ex-

periments performed. Statistical signiﬁcance of the data vs

untreated-infected cells was deﬁned as *P < 0.05, ** P < 0.01 and

***P < 0.001.

DENV-2/ZIKV/WNV replication inhibitory effect.

DENV-2

ZIKV

WNV

M) SI IC₅₀

M) SI^cIC₅₀

(

m

(

m

M)

SI

b

Cpd.

Huh7

CC₅₀

IC₅₀

(

m

a

(mM)

10a

10b

10c

10d

10e

11a

20.0

8.5

5.0

8.0

7.5

3.4 0.9

ND^e

ND

6

-

2.4 0.3

ND

8

(1.1 1.5)^d18

e

7

ND

e

18.5

4.0 1.02

5

2.8 0.8

(0.6 0.2)^d28

(0.6 0.4)^f30

11b

11c

11d

12

7.5

8.0

ND

4.7 0.7

-

ND

e

ND

e

Sofosbuvir 200.0

42 3.2 0.7 62 (1.7 0.5)^d117

(1.2 0.3)^f166

a

CC₅₀: half-maximal cytotoxic concentration (

IC₅₀: half-maximal inhibitory concentration calculated through SYRA, each

m

M).

b

value is the mean of three experiments standard deviation (SD).

c

SI: selectivity index (SI ¼ CC₅₀/IC₅₀).

d

IC₅₀: half-maximal inhibitory concentration calculated through DYRA.

e

ND: not determined.

IC₅₀: half-maximal inhibitory concentration calculated through PRA. DMSO was

f

used as control in all cytotoxicity and antiviral assays.

Finally, the BSAA 6i was selected as the most interesting

candidate to be evaluated against SARS-CoV-2: it was noteworthy

to ﬁnd that 6i inhibited the replication of this new coronavirus in

the low micromolar range in all human cell lines evaluated

(Table 3). Compound 6i showed submicromolar potency in SARS-

CoV-2-infected Calu-3 lung cell line and a substantial reduction of

efﬁcacy in the monkey kidney VERO E6 cell line as observed for

Remdesivir. [37,38] Compound 6i also showed low micromolar

anti-SARS-CoV-2 potency in liver Huh7 and intestinal Caco-2 cells,

while Remdesivir showed low nanomolar potency on both cell lines

in agreement with literature data [37,39]. The activity of 6i against

SARS-CoV-2 was examined in depth in the pulmonary Calu-3 cell

line, which better mimics the SARS-CoV-2 infection in vivo. Further

experiments were performed incubating the cells and the virus

with scalar concentration of 6i at two different conditions: 6i was

(i) incubated with cells and removed after virus-cell adsorption

(ENTRY-DYRA) or simply added after viral adsorption as previously

Fig. 2. Radical scavenging activity of compounds 10a, 11a and 6i expressed as % in-

hibition of the ABTS absorbance at 734 nm. The potent antioxidant ascorbic acid was

used as a positive control.

performed (DYRA). The IC₅₀obtained in ENTRY-DYRA (1.9 0.4

mM)

and DYRA (0.5 0.1 M) were quite similar and the difference

m

observed (3.8-fold) excludes a possible mechanism of action as

entry inhibitor. On the other hands, using the entry inhibitor

Camostat (CMT) as reference compound, [73] its antiviral potency

differs of more than 20-fold when tested with ENTRY DYRA

infected with the three strains and the expression of viral proteins,

stained with an anti-FLU antibody capable of recognizing more

viral proteins belonging to different strains, was observed directly

on the infected monolayers at 24 h p.i. by ICW assay. As shown in

(0.04 0.01 mM) or DYRA (0.82 0.32 mM) protocol. Probably, the 6i

6

I. Vicenti, M.G. Martina, A. Boccuto et al.

European Journal of Medicinal Chemistry 224 (2021) 113683

Table 3

Inﬂuenza A virus and SARS-CoV-2 replication inhibitory effect.

Inﬂuenza A Puerto Rico 8/34/H1N1

SARS-CoV-2^d

Cpd MDCK

A549

Huh7

Caco-2

Vero E6

Calu-3

a

b

CC₅₀

M)

IC₅₀

M)

SI^c

CC₅₀IC₅₀

SI

CC₅₀IC₅₀

(

m

M)

SI

CC₅₀IC₅₀

(m

M)

SI

CC₅₀

IC₅₀

(

mM)

SI CC₅₀IC₅₀

(m

M)

SI

(

m

(

m

(

mM)

(m

M)

(

m

M)

(

m

M)

(

m

M)

(m

M)

6c

6i

77.0

14.0

5.3

5.5

58.0 20.8

2.8

11

ND^e

200

ND

2.7 0.1

e

ND ND

81.3 5.9 1.8

e

ND

e

ND

e

14.5 106.0 9.9

74

14

200.0 53.4 1.1

4

120.0 0.5 0.1

240

(1.9 0.4)^f

10a 136.0 6.3

21.6 52.4 14.9

3.5

ND

e

ND

e

ND

e

ND

e

RBV (>50) (9.5)³⁴(>5.3) (>50) (37.4)³⁴(>1.3) ND

(>400) (109.5)36

e

RMD ND

CMT ND

ND

e

ND

e

77

0.004 0.005

(0.002) [37]

ND

19250 80.0 0.01 0.006

(0.38) [39]

8000 166.0 6.0 1.1

(26.9) [38]

28 97.0 0.11 0.04

(0.28) [37]

882

ND

ND ND

ND

ND 200.0 0.82 0.32

243

(0.04 0.01)^f

a

CC₅₀: half-maximal cytotoxic concentration (mM).

b

c

d

e

f

IC₅₀: half-maximal inhibitory concentration, each value is the mean of three experiments standard deviation (SD) at 0.01 MOI.

SI: selectivity index (SI ¼ CC₅₀/IC₅₀).

IC₅₀: half-maximal inhibitory concentration calculated through DYRA.

ND: Not determined.

IC₅₀calculated incubating with the ENTRY-DYRA protocol. Literature antiviral data for Remdesivir (RMD) and Ribavirin (RBV) are reported in parenthesis. CMT: Camostat.

DMSO was used as control in all cytotoxicity and antiviral assays.

Fig. 3. Compound 6i interferes with early steps of inﬂuenza virus replication. A549 cells were infected with PR8 and treated or not with 6i (20 mM) at different phases of the virus

life-cycle: the compound was added for 2 h before (PRE); during viral adsorption for 1 h (DUR); immediately after viral adsorption for 24 h (POST); before, during and for the

following 24 h (PDP). The supernatants were recovered and used to infect fresh A549 monolayers for the following 24 h (A); compound 6i was immediately added after viral

adsorption for 2 (0e2 h), 4 (0e4 h) or 6 (0e6 h) hours. Then cells were maintained up to 24 h without 6i in fresh medium plus 2% FBS; compound 6i was added after 2 (2e24 h), 4

(4e24 h) or 6 (6e24 h) hours infection and maintained until 24 h from infection. The supernatants were recovered and used to infect fresh MDCK monolayers for 24 h (B).

DYRA IC₅₀is slightly lower with respect to that observed in ENTRY-

DYRA, because the compound is never removed from the cells and

it's available for 72 h of incubation. Considering that DYRA mea-

sures only the SARS-CoV-2 N protein reduction, we also investi-

gated the role of 6i in the inhibition of viral replication. To assess

the viral infectivity, we collected the supernatants derived from

DYRA and we titred them infecting in quadruplicate VERO E6 cell

line. The titer of each supernatant expressed as 50% tissue infec-

tious culture dose (TCID₅₀) was determined by Reed and Muench

method and compared to the control virus titers. At the concen-

Overall, phenotypic screening of the synthesized purine de-

rivatives against different RNA viruses and different cell lines,

clearly identiﬁed compound 6i as the most promising BSAA that

can be quickly produced by two simple synthetic steps in only

60 min. Although 6i is just a promising lead compound, its cheap

and easy synthetic accessibility is particularly relevant in compar-

ison with the many synthetic steps that are needed to synthesize

the reference nucleoside antivirals Sofosbuvir [40] and Remdesivir

[41].

Considering the high anti-DENV and anti-ZIKV activity of 6i in

Huh7 (SI 222 and 400 respectively in PRA), in vitro selection ex-

periments were performed to investigate the 6i genetic barrier.

tration of 20

mM and 0.8 mM, 6i was able to reduce the viral repli-

cation of 91.9 5.1% and 83.5 5.1% respectively.

7

I. Vicenti, M.G. Martina, A. Boccuto et al.

European Journal of Medicinal Chemistry 224 (2021) 113683

Fig. 4. Compound 6i is effective against different inﬂuenza A virus strains. A549 cells were infected with inﬂuenza A virus strains as described in Methods and treated or not with 6i

(20 M) at different phases of the virus life-cycle: A) compound 6i was added for 2 h before (PRE); during viral adsorption for 1 h (DUR); immediately after viral adsorption for 24 h

m

(POST); before, during and for the following 24 h (PDP). Infected treated cells were ﬁxed and stained with anti-FLU antibodies as already described; B) compound 6i was

immediately added after viral adsorption for 2 (0e2 h), 4 (0e4 h) or 6 (0e6 h) hours. Then cells were maintained without the compound up to 24 h in fresh medium plus 2% FBS; the

compound was added after 2 (2e24 h), 4 (4e24 h) or 6 (6e24 h) hours infection and maintained until 24 h from infection. The supernatants were recovered and used to infect fresh

MDCK monolayers for 24 h.

ZIKV and DENV inputs (0.05 MOI), each in duplicate, were used to

infect Huh-7 cells in the presence of increasing concentration of 6i,

starting from 15 mM of 6i up to 120 mM. Uninfected cells treated

with 6i were used as negative control to discriminate between

cytopathic effect (CPE) derived from 6i cytotoxicity and physio-

logical cell mortality. 6i pressure delayed viral growth with respect

to the no-drug virus control (VC). Indeed, around 80% CPE was

transiently or permanently observed in the absence of drug pres-

sure, suggesting that changes in the viral genome driven by virus

adaptation to the cell line used or stochastically also occur.

Even if at different timing, both ZIKV and DENV escaped from 6i

drug pressure, selecting mutation mixtures which were then lost at

subsequent drug increment steps. When considering mutation

proﬁles including pure mutations, the two viral strains emerging at

reached at 5e9 days post infection (dpi) with 15e120

m

M 6i while

120

were not quantiﬁable by IA. The same occurred with the virus

carrying the mixed aa proﬁle selected in ZIKV NS5 at 60 M 6i

(D245D/A combined with D788D/E). On the basis of these obser-

vations, we can hypothesize that these proﬁles selected by 6i

impair the viral ﬁtness resulting in lack of CPE at subsequent

infection rounds. In this context, CPE is probably caused not only by

mM (I600V in DENV NS3) and at 15 mM (T860P in ZIKV NS5)

the VC was consistently collected at 2e4 dpi (Table4). In vitro se-

lection experiments were stopped after a mean of 23.8 1.5 dpi

when the drug pressure was 133-fold and 240-fold higher than 6i

IC₅₀against DENV and ZIKV wild type, respectively. Aminoacidic

(aa) variations in NS5 and NS3 regions selected under 6i pressure

are indicated in Table 4. It must be noted that some mutations were

m

8

I. Vicenti, M.G. Martina, A. Boccuto et al.

European Journal of Medicinal Chemistry 224 (2021) 113683

Table 4

In vitro ZIKV and DENV resistance selection to 6i.

replicating virus but also by the accumulation of viral transcripts

and non-replication competent viruses. The mixed and pure mu-

tations selected under 6i drug pressure were distributed in

different NS5 domains for ZIKV and DENV.

comparison to our previously published work [24] investigating the

genetic barrier of sofosbuvir against WNV with the same meth-

odology, 6i seems not to have a high genetic barrier. Indeed the

selected mutations were lost in the subsequent culture passages

with increasing drug concentration. In addition, the time of viral

breakthrough in the presence of 6i was not signiﬁcantly increased

with respect to VC. However, the I600V substitution for DENV NS3

and the D788D/G/E alone or coupled with the T860P substitution

for ZIKV NS5 reduced dramatically the viral ﬁtness. Indeed,

mutated proﬁles were not able to replicate in subsequent cultures

in the presence of 6i and reverted to wild type when grown in the

absence of 6i (data not shown). While the mutations emerging

from this study cannot explain the mechanism of action of 6i in

detail, their location supports the targeting of the replication ma-

chinery and the interaction between NS3 and NS5 as per our

original design [20]. To further corroborate the interaction between

6i and DENV/ZIKV NS5, we evaluated the inhibition of both MTase

and polymerase enzymatic activity in cell free experiments. Com-

pounds 6c, 10a and 11a were tested as comparison together with

the lead compound 6i (Table 5). Increasing concentrations of in-

hibitor were incubated with the puriﬁed MTase or polymerase and

the respective enzymatic activity was determined as previously

described [50,51]. The IC₅₀of compounds, deduced from Hill slope

curve-ﬁtting, are presented in Table 5. The results show that the

compounds barely inhibit the MTase activity. Conversely we

observed that 6c and 6i inhibit the polymerase of DENV and ZIKV at

a micromolar concentrations slightly higher than the correspond-

ing IC_50sin cell, further suggesting a multi-target mechanism for

the inhibition of DENV/ZIKV replication.

With the exception of DENV E427 E/K and ZIKV D788D/G/E, all

mutations selected in NS5 occurred at conserved sites among

different ﬂaviviruses [42]. The majority of NS5 aa variations

selected (D788D/G, D788D/E, T792 T/P, M806 M/I, T860P for ZIKV

and E427 E/K for DENV) were located in the polymerase region

(RdRp); none of them were located in the catalytic site.43 The E427

is located between the G and F motifs of DENV RdRp: these two

motifs interact with the nascent RNA template. D788, T792, M806

and T860 are located in the ZIKV priming loop which is involved in

the correct positioning of nucleotides during polymerization.

Interestingly, the T860P mutation emerged in ZIKV is located in the

allosteric RdRp cavity B: it has been observed that mutated residues

located in this cavity near T860 (W859A, I863A in DENV) signiﬁ-

cantly reduced viral replication and this could explain the ﬁtness

impairment of strains carrying this mutational proﬁle [44]. All the

remaining mixed mutations (K61 K/E, E74 E/K, P154 P/T, N241 N/H

and T244 T/P) emerged in DENV were located in the N terminal

region of NS5 in the methyltransferase domain (MTase) which in-

teracts with the C-terminal domain of NS3 involved in the un-

winding of double stranded RNA intermediate (ATPase/helicase

activity) and in viral RNA 5⁰-capping reaction (triphosphatase) [45].

Under 6i drug pressure, no pure or mixed mutations were selected

in ZIKV NS3. The mixed R54 R/P substitution selected in DENV NS3

is located in the N terminal domain near the catalytic site (His53,

Asp77 and Ser138) of the trypsin-like serine protease [43]. The

mixed substitution M267 M/I is located in the C terminal sub-

domain I of NS3; the conserved motifs in subdomains I and II are

key regions for binding to RNA to generate the ATP binding cleft

[46]. It's important to underline that, the only pure aa variation

In vitro resistance selection experiments showed that DENV and

ZIKV escape 6i pressure by selecting a number of mutations in the

NS5 and NS3 proteins that may impair their activity and/or the

formation of the replication complex. Cell free experiment on iso-

lated NS5 conﬁrmed a moderate inhibitory effect of 6i, mostly on

the polymerase activity. To further investigate the effect of 6i on the

formation of the DENV replication complex in whole-cell context,

we conducted an immunoﬂuorescence analysis to follow the

contemporary intracellular localization of NS5 and NS3 proteins at

I600V acquired in the presence of 120

the subdomain 3 of the helicase region of the NS3 protein and

between two -helices (559e607 residues) 9 and 10 [47]. Inter-

mM 6i by DENV is located in

a

estingly, the subdomain III (NS3 483e618 aa) contains the inter-

action site with NS5 (320e368 aa) of the RdRP domain [48,49]. In

9

I. Vicenti, M.G. Martina, A. Boccuto et al.

European Journal of Medicinal Chemistry 224 (2021) 113683

Table 5

IC₅₀values of compounds 6c, 6i, 10a and 11a against DENV/ZIKV polymerases and MTases.

Cpds.

IC₅₀

(

m

M)^a

IC₅₀(m

M)^a

D2pol^b

D2 NS5^c

ZIKV NS5^d

DENV NS5-MTase

ZIKV NS5-MTase

6c

6i

10a

11a

3'dATP control

Sinefungin

31.2 7.8

20.8 5.7

>100

0.27 0.03

e

>100

18.9 3.1

12.8 2.2

105 24.3

>100

NA

43.7 12.0

153.2 38.2

>100

0.32 0.05

e

112.0 51.5

NA

226.1 107.4

e

1

0.2

e

(0,63 0,04)52

(1,18 0,05)53

a

inhibition concentration 50%.

D2pol: polymerase domain of dengue 2 virus.

D2 NS5: full length DENV protein.

b

c

d

ZIKV NS5: full length ZIKV protein. Literature data for Sinefungin are reported in parenthesis.

8, 12, 16 and 24 h post infection with and without the presence of 6i

against selected host kinases that are known from the literature to

be involved in the replication of ﬂaviviruses, Inﬂuenza viruses and

Coronaviruses (Table 6) [58e60]. The above compounds were

tested at ﬁxed concentration corresponding to their median anti-

or Sofosbuvir using non-toxic concentration (50 and 100

mM,

respectively). Preliminary confocal microscopy analysis revealed

that compound 6i signiﬁcantly reduced the expression of NS3

(p ꢂ 0.05 at 12 h) and NS5 (p ꢂ 0.0001 at 16 h) proteins (Fig. 5,

panels B and C). On the other hands, treatment with Sofosbuvir

signiﬁcantly reduced NS3 and NS5 proteins expression starting

from 12 h (p ꢂ 0.05 and p ꢂ 0.01, respectively) (Fig. 5, panels B and

C). Next, the localization of both NS5 and NS3 in untreated control

(VC) was analyzed at different time points: although ﬂavivirus RNA

replication occurs in the membrane-bound multi-protein assembly

(the replication complex) localized in the cytoplasm, NS5 nuclear

accumulation is a key process in ﬂavivirus replication [54e56]

while the contemporary NS3 localization has been less investigated

[57]. Results from our experiments showed that initially (8 h) the

NS3 protein had predominantly nuclear localization and moved to

the cytoplasm at 24 h.

viral IC₅₀(5

mM) in comparison with the hit 3. Despite potent

nanomolar kinase inhibitors with no antiviral activity have been

reported among 2,6-diaminopurine derivatives, [61] our com-

pounds showed a moderate inhibitory effect on the selected ki-

nases at micromolar concentration: in particular, compound 6i

reduced the activity of Src, Abl and PI3Ka by roughly 50% at 5 mM.

The get a wider picture on the kinase speciﬁcity of 6i, it was also

tested against an additional panel of 15 host kinases selected

among those known from the literature to be involved in the

replication of different viruses (Table S3, Supporting Information)

[74]. However, no signiﬁcant inhibition of these additional targets

was found testing 6i at 5

mM concentration, making Src, Abl and

PI3K the only kinases (among the few selected) that are mildly

a

On the other hand, NS5 protein initially (8 h) had predominantly

cytoplasmic localization and translocation to the nucleus began at

16 h to be completed within 24 h. This analysis on the untreated

control showed a sort of concerted nucleocytoplasmic shuttling of

NS5 and NS3 proteins in opposite directions during infection: the

NS3:NS5 colocalization peak, which could be ideally associated

with the formation of the replication complex, occurred at 16 h,

mainly at nuclear and perinuclear level (Fig. 5, panel D, merge line).

Looking at the localization of NS5:NS3 in treated cells, sofosbuvir

seems to abolish the formation of the replication complex: NS5

translocation into the nucleus was reduced or absent while, at the

same time, NS3 protein did not translocate into the cytoplasm

(Fig. 5, panel D, Sofosbuvir column). In cells treated with 6i,

NS3:NS5 colocalization was signiﬁcantly reduced in treated cells

(Fig. 5, panel D, 6i column). The NS3:NS5 colocalization reduction

observed in 6i treated cells was conﬁrmed and quantiﬁed by

colocalization analysis, with the greatest reduction observed at 16 h

(MCC NS3/NS5 ¼ 0.404 0.106 and MCC NS5/NS3 ¼ 0.397 0.101;

Costes P ¼ 100%) compared to the untreated control (MCC NS3/

NS5 ¼ 0.726 0.135 and MCC NS5/NS3 ¼ 0.752 0.125; Costes

P ¼ 100%) (p ꢂ 0.0001) (Fig. 6).

inhibited by our lead compound. This moderate host kinases inhi-

bition together with the observed effect on NS5/NS3 may therefore

account for the anti-ﬂaviviral effect of these multi-target inhibitors.

As expected from the literature, [58e60] the effect of 6i on the

above kinases could also play a partial role in inhibiting the repli-

cation of Inﬂuenza A and SARS-CoV-2 but the full understanding of

its MOA is a complex endeavour and major efforts will be required

to solve this challenge in the future. It is also important to

remember that the precise knowledge of a drug's target is not an

essential requirement in drug discovery and around 10e20% of

approved drugs have unknown target or MOA [62,63]. Especially in

the discovery of new BSAAs for emerging diseases, the right

compromise between target identiﬁcation and efﬁcacy in multiple

functional assays should probably drive the rapid development of

new and efﬁcient candidates for preclinical studies.

3. Conclusion

The development of new BSAAs represents the ideal option to

counteract the spreading of different viruses (new and re-

emerging), to deal with viral co-infections and in preparedness

for new pandemics. Considering also the fact that the majority of

the worlds’ population lives in underprivileged regions, efforts

should be placed on the development of cheap and easy to be

produced drugs. In our previous study on the identiﬁcation of

multi-target anti-ﬂavivirus agents, we reported the design and

identiﬁcation of the hit compounds 1 and 2, as the most interesting

multi-target compound endowed with anti-Src/Fyn activity,

micromolar activity against DENV-2 (the only affected virus) and

inhibition of the interaction between recombinant NS3 and NS5

proteins in an AlphaScreen assay at high micromolar concentration

[20]. The following target-based optimization study led to the

It was interesting to observe that, although 6i and Sofosbuvir

showed a similar effect on the reduction of NS5/NS3 expression

over time, they induced a completely different effect on the local-

ization of the same proteins, which reﬂects the different mecha-

nism of action of the two compounds and indirectly support the

difference seen in the DYRA/SYRA assays and the mutational proﬁle

that is observed under 6i pressure (not compatible with that of a

nucleoside inhibitor).

Finally, to verify the multi-target effect of this class of com-

pounds, the most interesting 2,6-diaminopurine derivatives syn-

thesized in this work (6c, 6i, 10a and 11a) were also evaluated

10

I. Vicenti, M.G. Martina, A. Boccuto et al.

European Journal of Medicinal Chemistry 224 (2021) 113683

Fig. 5. Immunoﬂuorescence analysis. DENV-2 infected cells were treated with Sofosbuvir (100 mM) or 6i compound (50 mM) and stained at different time points with DENV NS3

antibody (detected using AlexaFluor® 488 labeled secondary antibody) and DENV NS5 antibody (detected using AlexaFluor® 568 labeled secondary antibody), and examined by

confocal microscopy. A) Cells stained with DAPI and counted at 40ꢃ magniﬁcation in 5 random ﬁelds per well (* p ꢂ 0.05). B) Expression of the viral NS3 protein in cells counted at

40ꢃ magniﬁcation in 5 random ﬁelds per well (* p ꢂ 0.05; ** p ꢂ 0.01; **** p ꢂ 0.0001). C) Expression of the viral NS5 protein in cells counted at 40ꢃ magniﬁcation in 5 random

ﬁelds per well (** p ꢂ 0.01; *** p ꢂ 0.001; **** p ꢂ 0.0001). D) Individual antibody stained as well as merged images in untreated control, cells treated with Sofosbuvir, and cells

treated with 6i compound. Each experiment was repeated at least two times.

identiﬁcation of compound 3 as a promising derivative active

against all four DENV serotypes and ZIKV but possessing a potency

CoV-2 proving to inhibit its replication at low micromolar and

submicromolar concentrations in all human cell lines evaluated.

We also conducted a series of biochemical studies in the attempt

to conﬁrm the multi-target effect of this class of compounds on

viral and host targets previously employed for the design of the

antiviral hits 1 and 2. The original rationale that led us to the

development of the anti-Flaviviral 2,6-diaminopurines was based

on the contemporary inhibition of Src/Fyn host kinases and

blockage of viral NS3:NS5 interaction. In vitro selection experi-

ments were thus performed on DENV and ZIKV under 6i pressure:

even if compound 6i seems not to have a high genetic barrier to

develop resistance it selected few mutations in important func-

tional sites of both NS5 and NS3 which dramatically reduced the

viral ﬁtness. Interestingly, mutations on the NS5 allosteric site

involved in NS3:NS5 interaction were selected both on DENV

(I600V) and ZIKV (T860P), in line with the original rationale.

Immunoﬂuorescence experiments on DENV-infected cells also

showed a signiﬁcant reduction of NS5/NS3 colocalization after

treatment with 6i. These results are in line with the in vitro selec-

tion experiments and further indicate, in a whole cell context, that

the anti-DENV effect of 6i is directly connected with a negative

modulation of the replication complex as originally suggested by

AlphaScreen experiments on recombinant NS3 and NS5 proteins

[20]. In addition, it was also possible to highlight that 6i and

Sofosbuvir act on the replication complex with two different MAO,

thus suggesting a possible synergy between nucleoside antivirals

(IC₅₀DENV-2 ¼ 10

m

M; IC₅₀ZIKV ¼ 16.7

mM) comparable to that of

compounds 1 and 2 [22]. Considering the sub-optimal results of

this target-based optimization and the advantages of a system-

oriented investigation in deciphering the functional efﬁcacy of

multi-target antivirals, we have reported here a phenotypic hit-to-

lead optimization that has allowed us to identify compound 6i as a

promising BSAA active against all tested viruses in the sub-

micromolar to low micromolar range (IC₅₀¼ 0.5e5.3

mM).

Starting from preliminary SAR consideration on previously

published 2,6-diaminopurine antivirals, we developed a ﬁrst set of

molecules (6a-m; 8a,b) that showed improved antiviral potency

against selected ﬂaviviruses (DENV-2, ZIKV and WNV). We then

explored the possibility to functionalize the 2,6-diaminopurine

scaffold with a ROS scavenging moiety to relieve the cellular

oxidative stress generated by viral infections. Within this second

set of molecules (10a-e; 11a-d), only two derivatives (10a and 11a)

showed interesting antiviral results against selected ﬂaviviruses.

The three most promising compounds identiﬁed from the

phenotypic screening against ﬂaviviruses (6c, 6i and 10a) were

then evaluated against Inﬂuenza A (human and avian viruses) thus

identifying 6i as the most potent and promising candidate, which

retained activity on different Inﬂuenza strains (the human PR8

H1N1 and the pandemic 2009H1N1 or the avian Ulster H7N1).

Compound 6i was also tested against the novel coronavirus SARS-

11

I. Vicenti, M.G. Martina, A. Boccuto et al.

European Journal of Medicinal Chemistry 224 (2021) 113683

Fig. 6. Immunoﬂuorescence analysis. DENV-2 infected cells were treated with 6i (50

mM) and stained at different time points with DENV NS3 antibody (detected using AlexaFluor®

488 labeled secondary antibody) and DENV NS5 antibody (detected using AlexaFluor® 568 labeled secondary antibody), and examined by confocal microscopy. NS3:NS5 stained

cells merged images and white dots colocalization (WDC) images in untreated control (above) and treated with 6i compound (below) are shown as indicated. Each experiment was

repeated at least two times.

advantages of combining multi-target TBDD with the efﬁcacy of a

PDD approach for the identiﬁcation and optimization of new

promising BSAAs.

Table 6

Inhibitory effect of BSAAs 6c, 6i, 10a and 11a against selected kinases.

Residual enzymatic activity (%)^b

Overall, the identiﬁed lead compound 6i represents a very

interesting BSAA capable of inhibiting the replication of DENV-2,

ZIKV, WNV, Inﬂuenza A virus and SARS-CoV-2 with high selec-

tivity index. By comparison with the reference antiviral drugs used

in this work (Sofosbuvir, Ribavirin, Remdesivir and Camostat),

compound 6i is very easy/cheap to be produced (just two synthetic

steps from commercial materials) and displays at low micromolar

and submicromolar antiviral efﬁcacy in all infected cell lines while

presenting a unique broad-spectrum proﬁle that is not shown by

any of these drugs. Further studies on this class of 2,6-

diaminopurine antivirals are currently ongoing to evaluate the

ADME proﬁle and suitable formulations for in vivo studies, while

additional functionalization in N9 are being introduced to fully

explore the biologically relevant chemical space around this che-

motype. Additional results will be published in due course.

Cpd.^a

Src

Abl

Fyn

PI3K

a

CDK9/cT1

6c

6i

10a

11a

3

45.0 20.0

47.0 20.0

84.0 1.0

NA

54.0 4.0

70.0 10.0

67.0 2.0

NA

NA^c

NA

89.0 2.0

45.0 6.9

69.0 1.0

66.0 0.2

77.0 3.2

78.0% 0.3

ND

87.0 14.0

NA

88.0 10.0

NA

59.0 9.4

65.0 5.6

NA

a

Each compound was tested at ﬁxed 5 mM concentration. Values for reference

compound 3 reported from the literature.

b

Values are the mean SD of two independent experiments.

NA: residual enzymatic activity >98%.

c

and our multi-target molecules.

Finally, the multi-target effect of this class of 2,6-diaminopurine

antivirals was also supported by the moderate inhibitory activity of

6i against ZIKV/DENV polymerase and against Src, Abl and PI3K

a

out of a panel of 20 selected kinases. While a direct correlation

between the inhibition of these host kinases and the broad-

spectrum antiviral activity of 6i has not been demonstrated at the

4. Experimental section

4.1. Chemistry

moment, Src, Abl and PI3K

a are validated and well recognized

targets for the development or repurposing of antivirals against

ﬂavivirus, Inﬂuenza A virus and SARS-CoV-2.[74] In summary, the

moderate inhibition of multiple isolated targets (and possibly

others still unknown) herein observed, well describes the

General. All commercially available chemicals were purchased

from Merck or Fluorochem and, unless otherwise noted, used

without any previous puriﬁcation. Solvents used for work-up and

puriﬁcation procedures were of technical grade. TLC was carried

12

I. Vicenti, M.G. Martina, A. Boccuto et al.

European Journal of Medicinal Chemistry 224 (2021) 113683

out using Merck TLC plates (silica gel on Al foils, SUPELCO Analyt-

ical). Where indicated, products were puriﬁed by silica gel ﬂash

chromatography on columns packed with Merck Geduran Si 60

100 ^ꢁC. Yield: 25%. MS (ESI) [MþH]^þ: 304.12 m/z. ¹H NMR

(DMSO‑d₆, 400 MHz): d 3.87 (s, 3H), 7.51 (m, 1H), 7.66 (m, 1H), 8.14

(m, 1H), 8.33 (s, 1H), 8.54 (s, 1H), 10.42 (s, 1H), 13.36 (s, 1H).

2-chloro-N-(3-(triﬂuoromethyl)phenyl)-9H-purin-6-amine

(5c): 1st Irradiation cycle: 15 min, 100 ^ꢁC. Yield: 96%. MS (ESI)

(40e63

AVANCE 300 MHz and BRUKER AVANCE 400 MHz spectrometers.

Chemical shifts ( scale) are reported in parts per million relative to

m

m). ¹H and ¹³C NMR spectra were recorded on BRUKER

d

[MþH]^þ: 314.23 m/z. ¹H NMR (DMSO‑d₆, 400 MHz):

d 7.42 (d, 1H,

TMS. ¹H NMR spectra are reported in this order: multiplicity and

number of protons; signals were characterized as: s (singlet),

d (doublet), dd (doublet of doublets), ddd (doublet of doublet of

doublets), t (triplet), m (multiplet), bs (broad signal). Low resolu-

tion mass spectrometry measurements were performed on quat-

tromicro API tandem mass spectrometer (Waters, Milford, MA,

USA) equipped with an external APCI or ESI ion source. ESI-mass

spectra are reported in the form of (m/z). Melting points were

taken using a Gallenkamp melting point apparatus and were un-

corrected. The FDA-approved anti-HCV drugs sofosbuvir (MCE®

cat. HY-15005), remdesivir (MCE® cat. HY-104077) and ribavirin

(MCE® cat. HY-B0434) used as reference compounds, were sup-

plied as powder and dissolved in 100% dimethyl sulfoxide (DMSO).

Camostat mesylate (MCE® cat. HY-13512), an inhibitor of TMPRSS2

showing antiviral activity against SARS-CoV-2 was dissolved in

water and used as reference compound in ENTRY-DYRA assay. All

ﬁnal compounds were screened for PAINS using the Free ADME-Tox

Filtering Tool (FAF-Drugs4) program (http://fafdrugs4.mti.univ-

paris-diderot.fr/) [70]: as expected, the azo-derivatives 11a-

d were ﬂagged as potential PAINS. Elemental analyses were per-

formed on a ThermoQuest (Italia) FlashEA 1112 Elemental Analyzer.

All ﬁnal compounds were >95% pure as determined by elemental

analysis. data for C, H, and N (within 0.4% of the theoretical values;

Table S4).

J ¼ 8 Hz), 7.60 (t,1H, J ¼ 8 Hz), 8.15 (d, 1H, J ¼ 8 Hz), 8.36 (s, 1H), 8.39

(s, 1H), 10.54 (s, 1H), 13.34 (s, 1H).

4-((2-chloro-9H-purin-6-yl)amino) benzene-1,2-diol (5d): 1st

Irradiation cycle: 20 min, 70 ^ꢁC; 2nd Irradiation cycle: 15 min,

120 ^ꢁC. Yield: 29%. ¹H NMR (DMSO‑d₆, 400 MHz):

d

6.71 (d, 1H,

J ¼ 8 Hz), 7.02 (d, 1H, J ¼ 8 Hz), 7.22 (s, 1H), 8.33 (s, 1H), 9.90 (s, 1H).

2-chloro-N-(3,4,5-trimethoxyphenyl) -9H-purin-6-amine

(5e): 1st Irradiation cycle: 20 min, 70 ^ꢁC. Yield: 70%. ¹H NMR

(DMSO‑d₆, 400 MHz): 3.65 (s, 3H), 3.78 (s, 3H), 7.40 (s, 2H), 8.30 (s,

d

1H), 10.12 (s, 1H), 13.33 (s, 1H).

General Procedure for the Synthesis of Compound 6a-m: In a

microwave tube 5a-e (0.19 mmol) and the proper amine

(0.570 mmol) were suspended in n-BuOH (3 mL). TFA (0.19 mmol)

was added and the tube was heated under microwave irradiation at

150e170 ^ꢁC for 40e70 min (max

mW power input: 300 W; ramp

time: 1 min; power max: off; maximum pressure: 260 psi). The

reaction mixture was concentrated under vacuum, dissolved with

EtOAc and washed with NaHCO₃. The combined organic phases

were washed with brine, dried over anhydrous Na₂SO₄, and evap-

orated to dryness. The crude material was puriﬁed by silica gel ﬂash

chromatography, using the proper eluent: 6a, 6c, 6d, 6f, 6i: CH₂Cl₂/

Formic Acid/MeOH 96:3:1; 6b, 6e: CH₂Cl₂/Formic Acid/MeOH

92:3:5; 6g: CH₂Cl₂/MeOH 96:4; 6h CH₂Cl₂/MeOH 9:1 plus 1% of

formic acid; 6j CH₂Cl₂/Formic Acid/MeOH 95:3:2e96:1:3; 6k

CH₂Cl₂/MeOH/Formic Acid 96:3:1.

Microwave Irradiation Experiments. Microwave reactions

were conducted using a CEM Discover Synthesis Unit (CEM Corp.,

Matthews, NC). The machine consists of a continuous focused mi-

crowave power delivery system with an operator-selectable power

output from 0 to 300 W. The temperature inside the reaction vessel

was monitored using a calibrated infrared temperature control

mounted under the reaction vessel. All experiments were per-

formed using a stirring option whereby the reaction mixtures were

stirred by means of a rotating magnetic plate located below the

ﬂoor of the microwave cavity and a Teﬂon-coated magnetic stir bar

in the vessel.

N6-(3-ﬂuorophenyl)-N2-methyl-9H-purine-2,6-diamine

(6a): 1st Irradiation cycle: 10 min, 170 ^ꢁC; 2nd Irradiation cycl_þe:

30 min, 150 ^ꢁC. Yield: 20%. Mp: 266e267 ^ꢁC. MS (ESI) [MþH] :

259.17 m/z. ¹H NMR (DMSO‑d₆, 400 MHz):

d

2.83 (d, 3H, J ¼ 2 Hz),

6.74 (s, 1H), 6.77 (t, 1H, J ¼ 4 Hz), 7.29 (q, 1H, J ¼ 15.2, 8 Hz), 7.76 (d,

1H, J ¼ 8.4 Hz), 7.88 (s, 1H), 8.17 (m, 1H), 9.70 (s, 1H), 12.64 (bs, 1H).

¹³C NMR (DMSO‑d₆, 100 MHz):

d 28.89, 107.03, 108.44, 115.94,

130.24, 137.25, 142.79, 151.85, 153.34, 159.84, 161.43, 163.48. Anal.

(C₁₂H₁₁FN₆) C, H, N.

2-((6-((3-ﬂuorophenyl)amino)-9H-purin-2-yl)amino)ethan-

1-ol (6b): 1st Irradiation cycle: 10 min, 170 ^ꢁC; 2_þnd Irradiation cy-

cle: 30 min, 150 ^ꢁC. Yield: 27%. MS (ESI) [MþH] : 289.12 m/z. ¹H

General Procedure for the Synthesis of intermediate 5a-e: In a

microwave tube 2,6-dichloro-9H-purine 4 (100 mg; 0.53 mmol)

and the proper amine (2.645 mmol) were suspended in n-BuOH

NMR (DMSO‑d₆, 400 MHz): d 3.36 (m, 2H), 3.56 (m, 2H), 6.59 (s,1H),

(3 mL). NEt₃(258

tube was heated at 70e120 ^ꢁC until the starting material was

consumed (max W power input: 100 W; ramp time: 1 min; power

m

L; 1.852 mmol) was added, except for 5e, and the

6.75 (m, 1H), 7.28 (q, 1H J ¼ 8 Hz), 7.78 (d, 1H, J ¼ 4 Hz), 7.84 (s, 1H),

8.10 (s,1H), 9.62 (s,1H),12.50 (s,1H). ¹³C NMR (DMSO‑d₆, 100 MHz):

m

d 44.49, 60.55, 106.65, 106.92, 108.13, 115.86, 130.15, 130.25, 142.71,

max: off; maximum pressure: 260 psi). At the end of irradiation, the

solid obtained was isolated by ﬁltration over a Buchner funnel after

the addiction of cold n-hexane and ethyl acetate. In a few cases (5b,

5d), at the end of irradiation, the reaction mixture was concen-

trated under vacuum, dissolved with ethyl acetate, and washed

with NaHCO₃. The combined organic were washed with brine,

dried over anhydrous Na₂SO₄, ﬁltered and concentrated. The crude

material was puriﬁed by silica gel chromatography, using the

proper eluent: 5b: CH₂Cl₂/MeOH 99:1e96:4; 5d: CH₂Cl₂/Formic

Acid/MeOH 93:3:4.

142.83, 159.51, 161.43, 163.81. Anal. (C₁₃H₁₃FN₆O) C, H, N.

N6-(3-ﬂuorophenyl)-N2-(2-phenoxyethyl)-9H-purine-2,6-

diamine (6c): 1st Irradiation cycle: 10 min, 170 ^ꢁC; 2nd Irradiation

cycle: 30 min, 150 ^ꢁC. Yield: 28%. Mp: 217e219 ^ꢁC. MS (ESI)

[MþH]^þ: 365.20 m/z. ¹H NMR (DMSO‑d₆, 400 MHz):

d 3.67 (q, 2H,

J ¼ 6.32 Hz), 4.14 (t, 2H, J ¼ 6.32 Hz), 6.75 (t, 1H, J ¼ 10.84 Hz), 6.94

(m, 4H), 7.27 (m, 3H), 7.77 (d, 1H, J ¼ 10.12 Hz), 7.87 (s, 1H), 8.13 (s,

1H), 9.64 (s,1H),12.54 (s,1H). ¹³C NMR (DMSO‑d₆,100 MHz):

d 41.11,

66.50, 106.77, 107.03, 108.48, 114.86, 115.96, 120.98, 129.97, 137.23,

142.61, 142.72, 152.02, 158.97, 159.22, 161.43, 163.51, 163.81. Anal.

(C₁₉H₁₇FN₆O) C, H, N.

2-chloro-N-(3-ﬂuorophenyl)-9H-purin-6-amine (5a): 1st

Irradiation cycle: 20 min, 100 ^ꢁC; 2nd Irradiation cycle: 35 min,

120 ^ꢁC. Yield: 67%. MS (ESI) [MþH]^þ: 264.14 m/z. ¹H NMR

Methyl 3-((2-(methylamino)-9H-purin-6 yl)amino)benzoate

(6d): 1st Irradiation cycle: 10 min, 170 ^ꢁC; 2nd Irradiation cycl_þe:

40 min, 150 ^ꢁC. Yield: 55%. Mp: 255e257 ^ꢁC. MS (ESI) [MþH] :

(DMSO‑d₆, 400 MHz):

d 5.37 (s, 1H), 6.29 (m, 1H), 6.99 (dd, 1H,

J ¼ 15.4, 7.9 Hz), 7.39 (d, 1H, J ¼ 7.4 Hz), 7.88 (d, 1H, J ¼ 12.1 Hz), 8.34

299.29 m/z. ¹H NMR (DMSO‑d₆, 400 MHz):

d

2.86 (d, 3H, J ¼ 4.4 Hz),

(s, 1H), 10.44 (s, 1H).

3.84 (s, 3H), 6.49 (q, 1H, J ¼ 8 Hz), 7.41 (m, 1H), 7.55 (d, 1H, J ¼ 8 Hz),

7.84 (s, 1H), 8.13 (s, 1H), 8.22 (s, 1H), 9.66 (s, 1H), 12.42 (bs, 1H). ¹³

NMR (DMSO‑d₆, 100 MHz): 28.99, 52.51, 120.97, 122.61, 124.75,

Methyl 3-((2-chloro-9H-purin-6-yl)amino)benzoate (5b): 1st

Irradiation cycle: 20 min, 70 ^ꢁC; 2nd Irradiation cycle: 15 min,

C

d

13

I. Vicenti, M.G. Martina, A. Boccuto et al.

European Journal of Medicinal Chemistry 224 (2021) 113683

129.09, 130.28, 141.40, 152.01, 153.63, 160.06, 160.15, 163.52, 166.94.

Anal. (C₁₄H₁₄N₆O₂) C, H, N.

114.02, 115.06, 121.01, 121.73, 129.98, 133.01, 136.69, 151.90, 152.97,

158.21, 158.88, 163.49. Anal. (C₂₂H₂₄N₆O₄) C, H, N.

Methyl 3-((2-((2-hydroxyethyl)amino)-9H-purin-6-yl)amino)

benzoate (6e): 1st Irradiation cycle: 10 min, 170 ^ꢁC; 2nd Irradiation

cycle: 30 min, 150 ^ꢁC. Yield: 45%. Mp: 250e252 ^ꢁC. MS (ESI)

General Procedures for the Synthesis of compounds 6l-6m:

Compound 6d (45 mg; 0.151 mmol) or 6f (33 mg; 0.082 mmol)

were dissolved in a 1:1 mixture of THF/MeOH (1.53 mL), treated

with 2 N aqueous solution of LiOH (0.30 mL) was added and the

resulting mixtures were stirred for 24 h at room temperature. At

the end of reaction, the solvent was evaporated under vacuum. The

crude material was puriﬁed by silica gel ﬂash chromatography

using the proper eluent: 6l: CH₂Cl₂/NEt_3/MeOH 82:3:15; 6m:

CH₂Cl₂/Formic Acid/MeOH 95:3:2.

[MþH]^þ: 329.36 m/z. ¹H NMR (DMSO‑d₆, 300 MHz):

d 3.41 (m, 2H),

3.56 (d, 2H, J ¼ 3 Hz), 3.86 (s, 3H), 6.35 (t, 1H, J ¼ 3 Hz), 7.42 (t, 1H,

J ¼ 9 Hz), 7.56 (d, 1H, J ¼ 6 Hz), 7.84 (s, 1H), 8.25 (s, 1H), 8.71 (s, 1H),

9.65 (s, 1H), 12.45 (bs, 1H). ¹³C NMR (DMSO‑d₆, 75 MHz):

d 44.59,

52.53, 60.61, 120.97, 122.65, 124.87, 129.12, 130.27, 137.10, 141.33,

152.10, 159.51, 163.57, 166.91. Anal. (C₁₅H₁₆N₆O₃) C, H, N.

Methyl

3-((2-((2-phenoxyethyl)amino)-9H-purin-6-yl)

3-((2-(methylamino)-9H-purin-6-yl)amino)benzoic acid (6l):

amino)benzoate (6f): 1st Irradiation cycle: 10 min, 170 ^ꢁC; 2nd

Yield: 16%. Mp: 211e213 ^ꢁC. MS (ESI) [MþH]^þ: 285.26 m/z. ¹H NMR

Irradiation cycle: 30 min, 150 ^ꢁC. Yield: 32%. Mp: 209e211 ^ꢁC. MS

(DMSO‑d₆, 300 MHz):

d

2.85 (d, 3H, J ¼ 4.5 Hz), 6.42 (d,1H, J ¼ 3 Hz),

(ESI) [MþH]^þ: 405.36 m/z. ¹H NMR (DMSO‑d₆, 400 MHz):

d

3.71 (d,

7.38 (t, 1H, J ¼ 8.1 Hz), 7.53 (d, 1H, J ¼ 7.2 Hz), 7.82 (s, 1H), 8.24 (m,

2H, J ¼ 5.6 Hz), 3.84 (s, 3H), 4.14 (t, 2H, J ¼ 4 Hz), 6.71 (t, 1H,

J ¼ 4 Hz), 6.92 (m, 3H), 7.26 (t, 2H, J ¼ 8 Hz), 7.39 (t, 1H, J ¼ 8 Hz),

7.55 (d, 1H, J ¼ 7.6 Hz), 7.85 (s, 1H), 8.26 (s, 1H), 8.73 (s, 1H), 9.70 (s,

1H), 8.65 (s, 1H), 9.58 (s, 1H), 12.38 (bs, 1H). ¹³C NMR (DMSO‑d₆,

100 MHz):

d 28.70, 113.88, 121.41, 122.89, 124.52, 128.85, 131.61,

136.80, 141.18, 152.24, 158.44, 160.18, 168.06. Anal. (C₁₃H₁₂N₆O₂) C,

H, N.

1H), 12.44 (s, 1H). ¹³C NMR (DMSO‑d₆, 100 MHz):

d 41.21, 52.52,

66.72, 114.87, 120.95, 121.06, 122.73, 124.94, 129.12, 129.95, 130.30,

137.09, 141.27, 152.30, 159.03, 159.34, 163.51, 166.89. Anal.

(C₂₁H₂₀N₆O₃) C, H, N.

N2-methyl-N6-(3-(triﬂuoromethyl)phenyl)-9H-purine-2,6-

diamine (6g): 1st Irradiation cycle: 10 min, 170 ^ꢁC; 2nd Irradiation

cycle: 20 min, 170 ^ꢁC. Yield: 40%. ¹H NMR (DMSO‑d₆, 400 MHz):

3-((2-((2-phenoxyethyl)amino)-9H-purin-6-yl)amino)ben- _þ

zoic acid (6m): Yield: 91%. Mp: 209e211 ^ꢁC. MS (ESI) [MþH] :

391.46 m/z. ¹H NMR (DMSO‑d₆, 300 MHz):

d

3.70 (q, 2H, J ¼ 4.2,

8.7 Hz), 4.14 (t, 2H, J ¼ 4.2 Hz), 6.68 (t, 1H, J ¼ 3.6 Hz), 6.92 (m, 2H);

7.27 (t,1H, J ¼ 7.2 Hz), 7.38 (t,1H, J ¼ 8 Hz), 7.55 (d,1H, J ¼ 8 Hz), 7.85

(s, 1H), 8.27 (s, 1H), 8.62 (s, 1H), 9.63 (s, 1H) 12.43 (s, 1H). ¹³C NMR

d

2.83 (d, 3H, J ¼ 4.8), 6.59 (d, 1H, J ¼ 4.4), 7.28 (d, 1H, J ¼ 7.6 Hz),

7.50 (t, 1H, J ¼ 8 Hz), 7.86 (s, 1H), 8.26 (s, 1H); 8.67 (s,1H), 9.79 (s,

1H), 12.44 (s, 1H). ¹³C NMR (DMSO‑d₆, 100 MHz):

28.90, 113.91,

(DMSO‑d₆, 100 MHz): d 41.20, 66.78, 114.88, 120.96, 123.08, 124.65,

128.99, 129.97, 131.48, 136.98, 141.00, 159.02, 159.35, 168.00. Anal.

(C₂₀H₁₈N₆O) C, H, N.

d

116.32, 117.96, 123.53, 123.62, 126.23, 129.78, 136.98, 141.86, 152.07,

153.45, 160.08. Anal. (C₁₃H₁₁F₃N₆) C, H, N.

2-((6-((3-(triﬂuoromethyl)phenyl)amino)-9H-purin-2-yl)

amino)ethan-1-ol (6h): 1st Irradiation cycle: 10 min, 170 ^ꢁC; 2nd

Irradiation cycle: 20 min, 150 ^ꢁC. Yield: 35%. ¹H NMR (DMSO‑d₆,

Synthesis of 2-chloro-N-(2-phenoxyethyl)-9H-purin-6-amine

(7): In a microwave tube 2,6-dichloro-9H-purine 4 (100 mg;

0.53 mmol) and 2-phenoxyethylamine (1.06 mmol) were sus-

pended in n-BuOH (3 mL). NEt₃(258

mL; 1.852 mmol) was added

and the tube was heated at 70 ^ꢁC for 30 min (max

mW power input:

400 MHz):

d

3.39 (t, 2H, J ¼ 8 Hz), 3.57 (t, 2H, J ¼ 4 Hz), 6.41 (t, 1H,

100 W; ramp time: 1 min; power max: off; maximum pressure: 260

psi). At the end of irradiation, the solid obtained was isolated by

ﬁltration over a Buchner funnel and washed with MeOH and EtOAc.

J ¼ 5.2 Hz), 7.28 (d, 1H, J ¼ 8 Hz), 7.50 (t, 1H, J ¼ 8 Hz), 7.87 (s, 1H),

8.15 (s, 1H), 8.32 (s, 1H), 8.49 (s, 1H), 12.44 (bs, 1H). ¹³C NMR

(DMSO‑d₆, 100 MHz):

123.64, 126.17, 129.44, 129.76, 129.84, 137.13, 141.75, 152.01, 159.44.

Anal. (C₁₄H₁₃F₃N₆O) C, H, N.

d

31.16, 60.48, 116.27, 118.05, 118.07, 123.46,

Yield: 85%. ¹H NMR (DMSO‑d₆, 300 MHz):

d

3.21 (t, 2H, J ¼ 3 Hz),

4.17 (t, 2H, J ¼ 6 Hz), 6.96 (m, 3H), 7.30 (m, 2H), 8.15 (s, 1H), 8.30 (s,

1H), 9.00 (bs, 1H).

N

[2]-(2-phenoxyethyl)-N [6]-(3-(triﬂuoromethyl)phenyl)-

General Procedures for the Synthesis of compounds 8a,b: In a

microwave tube 7 (50 mg; 0.173 mmol) and 3-ﬂuoroaniline or 3-

triﬂuoromethylaniline (0.865 mmol) were suspended in n-BuOH

(2 mL). TFA (0.259 mmol) was added and the tube was heated

under microwave irradiation for two consecutive cycles: 1st Irra-

diation cycle: 10 min, 170 ^ꢁC; 2nd Irradiation cycle: 40 min, 150 ^ꢁC

9H-purine-2,6-diamine (6i): 1st Irradiation cycle: 10 min, 170 ^ꢁC;

2nd Irradiation cycle: 50 min, 150 ^ꢁC;. Yield: 56%. Mp: 180e181 ^ꢁC.

MS (ESI) [MþH]^þ: 415.21 m/z. ¹H NMR (DMSO‑d₆, 400 MHz):

d 3.68

(q, 2H, J ¼ 6 Hz), 4.12 (t, 2H, J ¼ 6 Hz), 6.78 (t, 1H, J ¼ 6 Hz), 6.90 (m,

3H), 7.26 (t, 3H, J ¼ 7.6 Hz), 7.48 (t, 1H, J ¼ 8 Hz), 7.87 (s, 1H), 8.29 (s,

1H), 8.48 (s, 1H), 9.81 (s, 1H), 12.48 (bs, 1H). ¹³C NMR (DMSO‑d₆,

(max mW power input: 300 W; ramp time: 1 min; power max: off;

100 MHz):

d

41.16, 66.59, 114.83, 116.35, 118.19, 120.97, 123.73,

maximum pressure: 260 psi). The reaction mixture was concen-

trated under vacuum, diluted with ethyl acetate, and washed with

NaHCO₃. The combined organic were washed with NH₄Cl, brine,

dried over anhydrous Na₂SO₄, ﬁltered and concentrated. The crude

material was puriﬁed by silica gel ﬂash chromatography using the

proper eluent: 8a: CH₂Cl₂/Formic Acid/MeOH 95:3:2; 8b: CH₂Cl₂/

Formic Acid/MeOH 96.5:3:0.5.

126.17, 129.50, 129.81, 129.86, 129.97, 141.71, 159.00, 159.30. Anal.

(C₂₀H₁₇F₃N₆O) C, H, N.

4-((2-((2-phenoxyethyl)amino)-9H-purin-6-yl)amino)ben-

zene-1,2-diol (6j): 1st Irradiation cycle: 10 min, 170 ^ꢁC; 2nd Irra-

diation cycle: 60 min, 150 ^ꢁC. Yield: 39%. Mp: 170e172 ^ꢁC. MS (ESI)

[MþH]^þ: 379.39 m/z. ¹H NMR (DMSO‑d₆, 300 MHz):

d 3.63 (q, 2H,

J ¼ 6 Hz), 4.11 (t, 2H, J ¼ 6 Hz), 6.44 (t, 1H, J ¼ 6 Hz), 6.61 (d, 1H,

J ¼ 9 Hz), 6.95 (m, 3H), 7.26 (m, 3H), 7.77 (s, 1H), 8.20 (s, 1H), 8.93 (s,

N2-(3-ﬂuorophenyl)-N6-(2-phenoxyethyl)-9H-purine-2,6-

diamine (8a): Yield: 33%. Mp: 208e210 ^ꢁC. MS (ESI) [MþH]^þ:

1H), 12.42 (bs, 1H). ¹³C NMR (DMSO‑d₆, 100 MHz):

d

41.10, 66.67,

365.38 m/z. ¹H NMR (DMSO‑d₆, 400 MHz):

d 3.89 (s, 2H), 4.20 (m,

109.87, 112.38, 114.88, 115.48, 120.94, 129.97, 132.68, 141.06, 145.18,

159.02, 159.47, 163.91. Anal. (C₁₉H₁₈N₆O₃) C, H, N.

2H), 6.25 (m, 1H), 6.93 (m, 3H), 7.20 (m, 1H), 7.30 (m, 2H), 7.48 (d,

1H, J ¼ 1.2 Hz), 7.63 (s, 1H), 7.87 (s, 1H), 7.95 (d, 1H, J ¼ 13.2 Hz), 9.15

N2-(2-phenoxyethyl)-N6-(3,4,5-trimethoxyphenyl)-9H-pu-

rine-2,6-diamine (6k): 1st Irradiation cycle: 30 min, 170 ^ꢁC; 2nd

Irradiation cycle: 30 min, 170 ^ꢁC. Yield: 52%. MS (ESI) [MþH]^þ:

(s, 1H), 12.54 (s, 1H). ¹³C NMR (DMSO‑d₆, 100 MHz):

d 40.66, 66.31,

104.65, 104.91, 106.34, 106.55, 114.20, 114.87, 121.06, 129.98, 130.07,

137.47, 144.10, 144.22, 156.26, 158.89, 161.71, 163.49, 164.08. Anal.

(C₁₉H₁₇FN₆O) C, H, N.

437,47 m/z. ¹H NMR (DMSO‑d₆, 400 MHz):

d 3.62 (s, 3H), 3.72 (m,

2H), 3.78 (s, 6H), 4.12 (t, 2H, J ¼ 6.4 Hz), 6.93 (m, 3H), 7.27 (m, 2H),

7.43 (s, 1H), 7.89 (s, 1H), 8.14 (s, 1H), 9.27 (s, 1H), 12.63 (m, 1H). ¹³

NMR (DMSO‑d₆, 100 MHz): 41.08, 56.20, 60.55, 66.45, 98.22,

N [6]-(2-phenoxyethyl)-N [2]-(3 (triﬂuoromethyl) phenyl)-

C

9H-purine-2,6-diamine (8b): Yield: 42%. Mp: 211e213 ^ꢁC. MS (ESI)

d

[MþH]^þ: 415.39 m/z. ¹H NMR (DMSO‑d₆, 400 MHz):

d 3.89 (s, 2H),

14

I. Vicenti, M.G. Martina, A. Boccuto et al.

European Journal of Medicinal Chemistry 224 (2021) 113683

4.20 (m, 2H), 6.91 (m, 3H), 7.15 (d, 1H, J ¼ 8 Hz), 7.27 (m, 2H), 7.41 (t,

1H, J ¼ 8 Hz), 7.68 (s, 1H), 7.88 (s, 1H), 7.97 (s, 1H), 8.42 (s, 1H), 9.30

9H-purin-2-amine

(10b):

Yield:

42%.

MS

(ESI)

[MþH]^þ¼ 398,30 m/z. ¹H NMR (DMSO‑d₆, 300 MHz):

d 3.50 (d, 2H,

(s, 1H), 12.53 (bs, 1H). ¹³C NMR (DMSO‑d₆, 100 MHz):

d

40.66, 66.39,

J ¼ 6), 3.92 (m, 2H), 6.31 (m, 3H), 6.51 (m, 1H), 6.87 (m, 3H), 7.24 (q,

2H, J ¼ 6, 11.2 Hz), 7.76 (s, 1H), 8.35 (s, 1H), 9.30 (s, 1H), 12.36 (s, 1H).

114.23, 114.85, 116.41, 121.06, 121.81, 123.64, 126.35, 129.52, 129.69,

129.83, 129.98, 143.06,156.21, 158.91, 163.50. Anal. (C₂₀H₁₇F₃N₆O) C,

H, N.

General Procedures for the Synthesis of intermediates 9a-e:

In a microwave tube 2,6-dichloropurine (100 mg; 0,529 mmol) and

the proper phenylhydrazine (2645 mmol) were suspended in n-

¹³C NMR (CDCl₃þCD₃OD, 75 MHz):

d 41.24, 66.42, 92.89, 94.99,

95.25, 114.75, 114.94, 120.87, 129.91, 153.97, 155.27, 158.98, 159.28,

162.44, 162.60, 163.50, 164.84, 165.00. Anal. (C₁₉H₁₇F₂N₇O) C, H, N.

6-(2-(3,5-bis(triﬂuoromethyl)phenyl) hydrazinyl)-N-(2- phe-

noxyethyl)-9H-purin-2-amine (10c): Yield: 38%. MS (ESI)

BuOH (3 mL). NEt₃(221

was heated under microwave irradiation at 70 ^ꢁC for 20e40 min

(max W power input: 100 W; ramp time: 1 min; power max: off;

mL; 1587 mmol) was added and the tube

[MþH]^þ¼ 498.29 m/z. ¹H NMR (DMSO‑d₆, 400 MHz):

d 3.58 (q, 2H,

J ¼ 8 Hz), 3.97 (t, 2H, J ¼ 8 Hz), 6.18 (t, 1H, J ¼ 4 Hz); 6.89 (m, 3H),

7.19 (s,1H), 7.24 (t, 2H, J ¼ 4 Hz), 7.32 (s, 1H), 7.73 (s, 1H), 8.14 (s, 1H),

m

maximum pressure: 260 psi). When the reaction ended, the

mixture was concentrated under vacuum, diluted with ethyl ace-

tate, and washed with NH₄Cl. The combined organic were washed

with NH₄Cl, brine, dried over anhydrous Na₂SO₄, ﬁltered and

concentrated under vacuum. The crude material was puriﬁed by

silica gel ﬂash chromatography using the proper eluent: 9a: CH₂Cl₂/

MeOH 96:4; 9b-d: hexane/acetone 1:1; 9e: CH₂Cl₂/MeOH 97:3.

(E)-2-chloro-6-(2-(3-ﬂuorophenyl)hydrazinyl)-9H-purine

(9a): 1st Irradiation cycle: 10 min, 70 ^ꢁC; 2nd Irradiation cycle:

8.54 (s, 1H). ¹³C NMR (DMSO‑d₆, 100 MHz):

d 40.91, 66.23, 106.69,

111.65, 114.62, 114.64, 120.85, 122.64, 125.36, 128.07, 129.87, 131.11,

131.43, 137.14, 137.76, 158.88, 163.55. Anal. (C₂₁H₁₇F₆N₇O) C, H, N.

N-(2-phenoxyethyl)-6-(2-(4-(triﬂuoromethyl)phenyl)

hydrazinyl)-9H-purin-2- amine (10d): Yield: 41%. MS (ESI)

[MþH]^þ¼ 430.36 m/z. ¹H NMR (DMSO‑d₆, 300 MHz):

d 3.76 (q, 2H,

J ¼ 6 Hz), 4.15 (t, 2H, J ¼ 6 Hz), 6.89 (t, 2H, J ¼ 9 Hz), 6.96 (d, 2H,

J ¼ 9 Hz), 7.26 (m, 2H), 7.37 (bs, 1H), 7.68 (s, 1H), 8.03 (s, 1H), 8.05 (s,

1H), 8.25 (s, 1H), 8.28 (s, 1H), 8.37 (s, 1H). ¹³C NMR (DMSO‑d₆,

10 min, 70 ^ꢁC. Yield: 60%. ¹H NMR (DMSO‑d₆, 400 MHz):

d

6.58 (m,

75 MHz): d 49.12, 66.38, 111.87, 114.73, 114.92, 120.83, 121.00,

2H), 6.30 (m, 1H), 6.18 (m, 1H), 7.83 (s, 1H), 8.14 (s, 1H), 9.56 (s, 1H),

12,41 (m, 1H).

124.33, 126.54, 127.37, 129.87, 129.96, 154.56, 158.95, 159.47, 159.88.

Anal. (C₂₀H₁₈F₃N₇O) C, H, N.

(E)-2-chloro-6-(2-(3,5-diﬂuorophenyl)

hydrazinyl)-9H-

N-(2-phenoxyethyl)-6-(2-(3-

(triﬂuoromethyl)phenyl)

(10e): MS (ESI)

3.50 (m, 2H);

purine (9b): 1st Irradiation cycle: 10 min, 70 ^ꢁC; 2nd Irradiation

hydrazinyl)-9H-purin-2-

amine

cycle: 10 min, 70 ^ꢁC; 3rd Irradiation cycle: 20 min, 70 ^ꢁC. Yield: 37%.

[MþH]^þ¼ 430.15 m/z. ¹H NMR (DMSO, 400 MHz)

d

MS (ESI) [MþH]^þ: 297.13 m/z. ¹H NMR (DMSO‑d₆, 400 MHz):

d

6.47

3.89 (m, 2H), 6.46 (bs, 1H), 6.90 (t, 1H, J ¼ 8 Hz), 6.99 (m, 4H), 7.28

(m, 3H), 8.23 (s, 1H), 8.72 (s, 1H), 9.99 (s, 1H), 12.88 (m, 1H).

(m, 4H), 7.76 (s, 1H), 8.21 (s, 1H), 9.28 (bs, 1H), 12.34 (bs, 1H). ¹³

NMR (DMSO‑d₆, 100 M Hz) 41.10, 66.30, 114.81, 114.93, 120.94,

C

(E)-6-(2-(3,5-bis(triﬂuoromethyl)phenyl)

hydrazinyl)-2-

d

chloro-9H-purine (9c): 1st Irradiation cycle: 10 min, 70 ^ꢁC; 2nd

Irradiation cycle: 10 min, 70 ^ꢁC. Yield: 73%. MS (ESI) [MþH]^þ:

121.00, 122.88, 125.56, 129.97, 130.62, 130.73, 131.05, 131.54, 131.55,

143.09, 158.97, 163.55. Anal. (C₂₀H₁₈F₃N₇O) C, H, N.

397.06 m/z. ¹H NMR (DMSO‑d₆, 400 MHz):

d

7.28 (m, 2H), 7.42 (m,

General procedures for synthesis of compounds 11a-d:

Compounds 10b-e (20 mg each) were dissolved in MeOH and

stirred at room temperature under air atmosphere for 96 h. The

corresponding oxidized compounds 11a-d were obtained as pure

products in quantitative yields.

2H), 8.28 (s, 1H); 9.11 (m, 1H), 10.02 (m, 1H), 10.25 (m, 1H), 12.99 (m,

1H).

(E)-2-chloro-6-(2-(4-(triﬂuoromethyl)phenyl) hydrazinyl)-

9H-purine (9d): 1st Irradiation cycle: 10 min, 70 ^ꢁC; 2nd Irradia-

tion cycle: 10 min, 70 ^ꢁC. Yield: 68%. MS (ESI) [MþH]^þ¼ 331.20 m/z.

6-((3-ﬂuorophenyl)diazenyl)-N-(2-phenoxyethyl)-9H-purin-

¹H NMR (DMSO‑d₆, 400 MHz):

d

6.86 (d, 2H, J ¼ 8 Hz), 7.48 (d, 2H,

2-amine (11a): MS (ESI) [MþH]^þ: 378.28 m/z. ¹H NMR (CDCl₃,

J ¼ 7.6 Hz), 8.23 (s, 1H), 8.79 (s, 1H), 10.09 (s, 1H), 12.9 (m, 1H).

2-chloro-6-(2-(3-(triﬂuoromethyl)phenyl) hydrazinyl)-9H-

purine (9e): 1st Irradiation cycle: 10 min, 70 ^ꢁC; 2nd Irradiation

cycle: 10 min, 70 ^ꢁC. Yield: 54%. MS (ESI) [MþH]^þ: 328.69 m/z. ¹H

400 MHz): d 4.00 (m, 2H), 4.21 (m, 2H), 6.11 (bs, 1H), 6.94 (m, 3H),

7.27 (m, 3H), 7.50 (q, 1H, J ¼ 8 Hz), 7.73 (s, 1H), 7.89 (d, 1H, J ¼ 4 Hz),

8.24 (s, 1H). ¹³C NMR (CDCl₃þCD₃OD, 100 MHz):

d 41.22, 66.45,

108.88, 109.18, 114.46, 119.95, 120.93, 121.34, 129.45, 130.47, 130.58,

153.87, 158.58, 159.78, 161.50, 164.81. Anal. (C₁₉H₁₆FN₇O) C, H, N.

6-((3,5-diﬂuorophenyl)diazenyl)-N-(2-phenoxyethyl)-9H-

purin-2-amine (11b): [MþH]^þ: 396.22 m/z. ¹H NMR

NMR (DMSO‑d₆, 400 MHz):

d 7.14 (m, 4H), 8.24 (m, 1H), 8.63 (m,

1H), 10.00 (m, 1H), 12.97 (m, 1H).

General procedures for the Synthesis of compounds 10a-e: In

a microwave tube 9a-e and the proper amine (2.5 equivalent) were

suspended in n-BuOH (2 mL). TFA (1 equivalent) was added and the

tube was heated at 170 ^ꢁC for 4 consecutive cycles of 30 min each

(CDCl₃þCD₃OD, 400 MHz):

d

3.93 (t, 2H, J ¼ 4 Hz), 4.19 (t, 2H,

J ¼ 4 Hz), 6.91 (m, 3H), 7.02 (q, 1H, J ¼ 4 Hz), 7.25 (t, 2H, J ¼ 8 Hz),

7.28 (s, 1H), 7.70 (d, 2H, J ¼ 4 Hz), 8.12 (s, 1H), 13.65 (s, 1H). ¹³C NMR

(max

mW power input: 300 W; ramp time: 1 min; power max: off;

(CDCl₃þCD₃OD, 100 MHz):

d 41.23, 66.41, 106.91, 107.18, 107.72,

maximum pressure: 300 psi). The reaction mixture was concen-

trated under vacuum. Products were puriﬁed by silica gel ﬂash

chromatography, using the proper eluent: 10a: hexane/acetone 3:7

plus 1% of formic acid; 10b: hexane/acetone 3:7; 10c: hexane/

acetone 4:6 plus 1% of formic acid; 10d: hexane/acetone 2:8 plus 1%

of formic acid; 10e: CHCl₃/MeOH 94:4 plus 1% of formic acid.

6-(2-(3-ﬂuorophenyl) hydrazinyl)-N-(2-phenoxyethyl)-9H-

purin-2-amine (10a): Yield: 46%. MS (ESI) [MþH]^þ¼ 380.29 m/z.

107.98, 108.21, 114.45, 120.96, 129.46, 154.23, 158.58, 159.71, 162.02,

164.38. Anal. (C₁₉H₁₅F₂N₇O) C, H, N.

6-((3,5-bis(triﬂuoromethyl)phenyl)diazenyl)-N-(2-

phenoxyethyl)-9H-purin-2-amine (11c): MS (ESI) [MþH]^þ:

496.34 m/z. ¹H NMR (CDCl₃, 400 MHz):

d 4.02 (m, 2H), 4.23 (m, 2H),

6.13 (bs, 1H), 6.95 (m, 3H), 7.28 (m, 2H), 8.08 (s, 1H), 8.26 (s, 1H),

8.56 (s, 1H). ¹³C NMR (CDCl₃þCD₃OD, 100 MHz):

d 41.51, 66.42,

114.45, 121.09, 121.38, 123.76, 124.09, 125.98, 129.53, 132.58, 132.92,

133.26, 133.60, 152.48, 158.50, 159.93. Anal. (C₂₁H₁₅F₆N₇O) C, H, N.

¹H NMR (DMSO‑d₆, 300 MHz):

d 3.50 (s, 2H), 3.91 (s, 2H), 6.43 (m,

3H), 6.58 (d, 1H, J ¼ 9 Hz), 6.90 (m, 3H), 7.13 (q, 1H, J ¼ 9 Hz), 7.25 (t,

N-(2-phenoxyethyl)-6-((4-

(triﬂuoromethyl)phenyl)dia-

2H, J ¼ 9 Hz), 7.77 (s, 1H), 8.06 (s, 1H), 9.17 (s, 1H). ¹³C NMR

zenyl)-9H-purin-2-amine (11d): MS (ESI) [MþH]^þ: 428.15 m/z. ¹H

(DMSO‑d₆, 100 MHz):

d

40.97, 66.42, 99.15, 104.41, 108.69, 114.76,

NMR (CDCl₃, 400 MHz):

d

4.01 (t, 2H, J ¼ 4 Hz), 4.22 (t, 2H, J ¼ 4 Hz),

114.86, 120.84, 120.93, 121.23, 129.88, 130.56, 130.70, 153.10, 155.45,

158.98, 159.47, 162.42, 164.84. Anal. (C₁₉H₁₈FN₇O) C, H, N.

6-(2-(3,5-diﬂuorophenyl) hydrazinyl)-N-(2-phenoxyethyl)-

6.09 (s, 1H), 6.94 (m, 3H), 7.27 (m, 2H), 7.81 (d, 2H, J ¼ 8 Hz), 8.15 (d,

2H, J ¼ 8 Hz), 8.24 (s, 1H), 13.65 (bs, 1H) . ¹³C NMR (CDCl₃þCD₃OD,

100 MHz):

d 45.12, 70.29, 118.35, 118.39, 124.84, 126.19, 127.89,

15

I. Vicenti, M.G. Martina, A. Boccuto et al.

European Journal of Medicinal Chemistry 224 (2021) 113683

128.90, 130.42, 133.35, 137.92, 138.24, 158.23, 162.54, 162.59, 163.67.

Anal. (C₂₀H₁₆F₃N₇O) C, H, N.

then infected eggs were maintained overnight at 4 ^ꢁC. Subse-

quently, the allantoic ﬂuid was collected and clariﬁed by centrifu-

gation (2500ꢃg for 30 min). The recovered virus was used for the

infection of MDCK and A549 cells. Allantoic ﬂuid from uninfected

eggs was used as reference for mock infection [68].

4.2. ABTS radical assay

ABTS assay was performed according to the protocol reported

on reference 33. The stock solution was prepared by mixing ABTS

(2, 2 azobis-(3-ethylbenzothiozoline-6-sulphonic acid), dissolved

in water to a 7 mM concentration, with 2.45 mM potassium per-

sulfate and allowing the mixture to stand in the dark at room

temperature for 12e16 h before use. The latter solution, oppor-

tunely diluted in ethanol (1 : 100, v/v), gave an absorbance of 0.70

( 0.02) at 734 nm and was used as control solution. 3.0 mL of the

4.3.2. Drugs and cytotoxicity assay

Cytotoxicity of investigational compounds was determined by

CellTiter-Glo 2.0 Luminescent Cell Viability Assay (Promega) ac-

cording to the manufacturer's protocol in the cells used to subse-

quently test ﬂaviviruses and SARS-COV-2 (Vero E6, Huh7 A549

Caco-2 and Calu-3). The luminescence values obtained from cells

treated with investigational compounds or DMSO control were

measured through the GloMax® Discover Multimode Microplate

Reader (Promega) and elaborated with the GraphPad PRISM soft-

ware version 6.01 (La Jolla) to calculate the half-maximal cytotoxic

concentration (CC₅₀). The cytotoxicity assay of candidate com-

pounds was evaluated on cell lines used for inﬂuenza virus infec-

latter control solution were added to 30

pound (Ascorbic Acid) or selected compounds dissolved in DMSO,

at different concentrations (10 M, 50 M, 0.1 mM, 0.2 mM, 0.5 mM,

mL of the reference com-

m

1 mM). The absorbance at 734 nm was read exactly after 10 min of

incubation time. The scavenging activity was estimated based on

the percentage of ABTS radicals scavenged by the following

formula:

tion

by

the

MTT

[3-(4,5-dimethylthiazol-2-yl)-2,5-

diphenyltetrazolium bromide] assay [69]. Brieﬂy, cells were

seeded in 96-well plates at a density of 2 ꢃ 10⁴cells/well in 100

mL

%

¼ [(A₀ꢀ A_s)/A₀] ꢃ 100

of complete RPMI without phenol red for 24 h at 37 ^ꢁC. Subse-

quently, cell monolayers were treated or not with increasing con-

scavenging

where A₀is absorption of control solution (run in each assay), A_Sis

absorption in the presence of reference compound or sample.

centrations (0.5, 2.5, 5, 10 and 20

37 ^ꢁC. After 24 h, 10

L of MTT solution (5 mg/mL) were added to

each well for 3e4 h at 37 ^ꢁC. Afterwards, each sample was acidiﬁed

by adding 0.1 N HCl in isopropanol (100 L/well) for 30 min in slow

mM) of compounds for 24 h at

m

4.3. Biology

m

agitation to ensure that all formazan crystals were dissolved.

Absorbance of samples was read at 570 nm, using an automatic

plate reader (Multiskan EX, Ascent Software, Thermo Fisher Sci-

entiﬁc). Untreated cells were used as control. The 50% cytotoxic

concentration (CC₅₀), deﬁned as the compound's concentration

required to reduce cell viability by 50%, was calculated using the

GraphPad, Prism 6 Software.

4.3.1. Cells and viruses

The H/PF/2013 ZIKV strain belonging to the Asian lineage, the

New Guinea C DENV serotype 2 strain and WNV lineage 1 (Italy/

2009) strain were kindly provided by the Istituto Superiore di

ꢀ

Sanita (Rome, Italy) while the SARS-CoV-2 strain, belonging to

lineage B1 was kindly provided by the Department of Biomedical

and Clinical Sciences Luigi Sacco, University of Milan. [64] Vero E6

(ATCC catalogue no. CRL-1586), Huh7 (kindly provided by Istituto

Toscano Tumori, Core Research Laboratory, Siena, Italy), A549

(ATCC catalogue no. CCL-185), Caco-2 (ATCC catalogue HTB-37) and

Calu-3 (ATCC catalogue no. HTB-55), MDCK (ATCC catalogue no.

CCL-34) were used to determine the antiviral activity of candidate

compounds. Cell lines were maintained in high glucose Dulbecco's

4.3.3. Antiviral assays

To determine the antiviral activity of candidate compounds

against WNV, DENV and ZIKV, a DYRA based on infection of cells in

presence of serial drug dilutions was performed as previously

described [24,66]. Brieﬂy, 7000 Huh7 cells per well were infected

with 50 TCID₅₀of DENV or ZIKV and 25 TCID₅₀of WNV. Viral

adsorption was performed in 96-well plates for 1 h at 37 ^ꢁC with 5%

CO₂and after virus removal, serial dilutions of each tested com-

pound were added to the cells and incubated at 37 ^ꢁC with 5% CO₂.

After 48 h (WNV) and 72 h (DENV and ZIKV) of incubation, the

antiviral activity was measured on cell monolayer by immunode-

tection assay (IA). ZIKV DENV and WNV harvested supernatants,

were used to infect pre-seeded Huh7 cells in SYRA or PRA as pre-

viously described [24,67]. Yield reduction protocol was adapted to

evaluate the antiviral activity of candidate compounds against

SARS-COV-2 in Vero E6, Caco-2, Calu-3 and Huh7 using the MOI of

0.01. Viral adsorption was performed in 96-well plates for 2 h at

37 ^ꢁC with 5% CO₂and after virus removal, serial dilutions of each

tested compound were added to the cells and incubated 48 h at

37 ^ꢁC with 5% CO₂(DYRA). To discriminate the activity of drug as

entry inhibitor, serial dilutions of reference compound and 6i were

added onto the Calu-3 cell and removed after cell-virus adsorption

(ENTRY-DYRA). The CPE was used as read out in Vero E6 and

measured by CellTiter-Glo 2.0 Luminescent Cell Viability Assay

(Promega) according to the manufacturer's protocol. The IA pro-

tocol was adapted for the detection of SARS nucleocapsid protein in

infected Caco-2 cells. Brieﬂy, ﬁxation and permeabilization were

performed as previously described and cell lines were incubated for

1 h with a monoclonal SARS Nucleocapsid Protein Antibody (Novus,

cat. AP201054) diluted 1:1000 in blocking buffer (PBS containing

Modiﬁed Eagle's Medium with sodium pyruvate and L-glutamine

(DMEM; Euroclone), Minimum Essential Medium Eagle (EMEM;

Euroclone) or Roswell Park Memorial Institute medium (RPMI-

1640, Sigma) supplemented with 10% Fetal Bovine Serum (FBS;

Euroclone) and 1% Penicillin/Streptomycin (Pen/Strep, Euroclone).

The same medium was used but with a lower FBS concentration for

the viral propagation and drug susceptibility testing (1e2%) and for

in vitro selection experiments (3%). Cells were incubated at 37 ^ꢁC in

a humidiﬁed incubator supplemented with 5% CO₂. All the viral

stocks were titrated by plaque reduction assay (PRA) as previously

described [65]. Brieﬂy, conﬂuent Vero E6 (WNV, DENV, SARS-CoV-

2) and A549 (ZIKV) in 6-well format were infected with three

tenfold dilutions of viral stock and after 1 h adsorption at 37 ^ꢁC, the

cells were washed with PBS and 0.75% Sea Plaque Agarose (Lonza)

was added to each well. After 3 (WNV), 5 (ZIKV, SARS-COV-2) and

10 (DENV) days incubation at 37 ^ꢁC, the monolayers were ﬁxed with

10% formaldehyde (Carlo Erba Chemicals) and stained with 0.1%

crystal violet (Carlo Erba Chemicals). After 3 h incubation, the agar

overlay was removed by water washing and PFU were counted.

Allantoic cavities of 11-day-old embryonated chicken eggs were

used to grow inﬂuenza viruses A/Puerto Rico/8/34H1N1 (PR8

H1N1), pandemic A/California/04/09H1N1 (p2009H1N1) and avian

A/parrot/Ulster/73H7N1 (Ulster H7N1). Viral suspension was

inoculated in the allantoic cavity and incubated for 48 h at 37 ^ꢁC,

16

I. Vicenti, M.G. Martina, A. Boccuto et al.

European Journal of Medicinal Chemistry 224 (2021) 113683

1% bovine serum albumine (BSA) and 0.1% Tween 20). After

washing, monolayers were incubated for 1 h with a polyclonal HRP-

coupled anti-mouse IgG secondary antibody (Novus Bio NB7570)

diluted 1: 5000 in blocking buffer. After cell washing, the 3,3⁰,5,5⁰-

Tetramethylbenzidine substrate (Sigma Aldrich) was added to each

well and the reaction was stopped with one volume of 0.5 M sul-

furic acid. Absorbance was measured at 450 nm optical density

(OD450) using the Absorbance Module of the GloMax® Discover

Multimode Microplate Reader (Promega). The half-maximal

inhibitory concentration (IC₅₀) was calculated through a non-

linear regression analysis of the dose-response curves generated

with GraphPad PRISM software version 6.01. In each test, sofosbuvir

and remdesivir were used as reference compounds against ﬂavi-

viruses and SARS-CoV-2 respectively. Infected and uninfected cells

without drugs were used to calculate the 100% and 0% of viral

replication, respectively. Selectivity Index (SI) was calculated as

ratio between CC₅₀and IC_50.

The anti-inﬂuenza activity of candidate compounds was evalu-

ated by Hemagglutination assay (HAU) in the supernatants of

infected cells recovered after 24 h from infection [69]. Brieﬂy,

MDCK and A549 cells seeded in 12-well plates at a density of

2.5 ꢃ 10⁵cells/well, were infected with inﬂuenza virus A/Puerto

Rico/8/34H1N1 and incubated for 1 h at 37 ^ꢁC. After viral adsorp-

tion, cells monolayers were washed with PBS and then, fresh me-

dium supplemented with 2% FBS with different concentrations (0.5,

different phases of virus life-cycle: the compound was added for 2 h

before (PRE); during viral adsorption for 1 h (DUR); immediately

after viral adsorption for 24 h (POST); before, during and for the

following 24 h (PDP). After the conditions of pretreatment and

during, cells were washed with PBS, and successively maintained in

fresh medium for the following 24 h. In another set of experiments,

i) the compound was immediately added after viral adsorption for 2

(0e2 h), 4 (0e4 h) or 6 (0e6 h) hours. Then cells were washed with

PBS and maintained up to 24 h without 6i in fresh medium plus 2%

FBS; ii) the compound was added after 2 (2e24 h), 4 (4e24 h) or 6

(6e24 h) hours infection and maintained until 24 h from infection.

The supernatants were recovered and used to infect fresh MDCK

monolayers for 24 h. Untreated infected cells were used as control.

The expression of viral proteins was analyzed directly on the

infected cell monolayers or on other monolayers infected for 24 h

with the recovered supernatants. Statistical signiﬁcance of the data

was analyzed by means of a one-way ANOVA Multiple comparisons

test and *P values of <0.05 were considered signiﬁcant. Data were

expressed as mean the standard deviation (SD).

4.3.6. In vitro selection experiments

The in vitro selection experiments were performed as previously

described [67]. Huh7 cells at 70% conﬂuence were infected in

duplicate with ZIKV and DENV at 0.005 multiplicity of infection

(MOI). After 1 h adsorption at 37 ^ꢁC with 5% CO₂, the virus inoculum

was removed and cells were treated with an initial concentration of

2.5, 5, 10 and 20 mM) of selected compounds was added for 24 h at

37 ^ꢁC. Each compound solved in DMSO was diluted in RPMI (the

highest concentration of DMSO added in the culture medium was

equal to 0.2% of ﬁnal volume). As reference, infected cells were

treated with DMSO alone. The IC₅₀of each compound was calcu-

lated using the GraphPad, Prism 6 Software. The SI was calculated

as described above.

15 mM 6i, corresponding to 17-fold and 30-fold IC₅₀(as observed by

PRA) against DENV and ZIKV, respectively. Cell cultures were

monitored every 24 h and when 80% viral CPE was observed, cells

and supernatant were freeze-thawed, cleared by centrifugation and

viral stocks were used to re-infect fresh pre-seeded Huh7 cells with

a twofold increased 6i concentration (30, 60 and 120 mM). All viral

stocks collected during in vitro selection experiments were

analyzed by population sequencing to detect emergent mutations

in the NS5 and NS3 regions. Wild type and mutant viruses were

titrated by IA.

4.3.4. In cell western assay

This is an innovative cell-based technique that allows a sensitive

quantiﬁcation of inﬂuenza virus proteins expressed on cell mono-

layers [69]. Brieﬂy, permissive cells were seeded in 96-well plates at

a density of 2 ꢃ 10⁴cells/well for 24 h at 37 ^ꢁC in an atmosphere of

5% CO₂. Conﬂuent cell monolayers were infected with different

inﬂuenza A virus strains or with supernatants recovered from

infected cells for 1 h with PBS and then fresh medium supple-

mented with 2% FBS was added for 24 h at 37 ^ꢁC. Subsequently, cell

monolayers were ﬁxed with 3.7% formaldehyde for 20 min, per-

meabilized with 0.1% Triton X-100 and incubated with Odyssey

blocking buffer (LI-COR Bioscience, Lincoln, NE) for 60 min at room

temperature. The cells were stained at 4 ^ꢁC overnight with primary

antibodies against inﬂuenza Hemagglutinin (HA) or anti-FLU. After

incubation, three washes with Phosphate Buffered Saline (PBS) plus

0.1% Tween 20 were performed and then the cells were stained

with a mixture of ﬂuorochrome-conjugated secondary antibodies

(ﬂuorescence emission at 800 nm), properly diluted in Odyssey

blocking buffer and ﬂuorochrome-conjugated Cell Tag (ﬂuores-

cence emission at 700 nm), for 1 h at room temperature. Cell Tag

was used as control of the integrity of cell monolayer. Subsequently,

three washes with PBS plus 0.1% Tween 20 were performed and

plates were analyzed by the Odyssey infrared imaging system (LI-

COR). Integrated intensities of ﬂuorescence were determined by the

LI-COR Image Studio software and the relative ﬂuorescence units

(RFU) were expressed as percentage compared to untreated infec-

ted cells (100%).

4.3.7. Viral RNA ampliﬁcation and sequencing

Total RNA was extracted in duplicate from 150 mL of viral stocks

derived from in vitro selection experiments, using the ZR Viral RNA

Kit (Zymo Research) according to the manufacturer's protocol. Ten

microliters of extracted ZIKV RNA were denatured at 70 ^ꢁC for 5 min

and then added to a mixture including 664 mM dNTPs, 6 mL of 5X

ImProm-II TM Reaction Buffer, 50 ng Hexanucleotides, 1.5 mM

MgCl₂, 20U RNasin® Plus RNase Inhibitor and 1U of ImProm-II™

Reverse Transcriptase (Promega) in a ﬁnal volume of 30 mL. Reverse

transcription reactions were run in Eppendorf Mastercycler Nexus

(Eppendorf) with an initial step at 25 ^ꢁC for 5 min, followed by a

reverse transcription step at 37 ^ꢁC for 30 min and a ﬁnal step at

80 ^ꢁC for 5 min cDNA was used as the template for PCR ampliﬁ-

cation of the whole NS5 and NS3 genes, using the Q5 Hot Start

High-Fidelity DNA Polymerase (NEB) protocol.

To design primers with a high degree of conservation, the ZIKV

and DENV alignment available at the NCBI website was used

(https://www.ncbi.nlm.nih.gov/genomes/VirusVariation);

the

primer sequences and coordinates, as referred to the H/PF/2013

ZIKV strain (Gen Bank Accession Number KJ776791) and DENV

DENV-2 New Guinea C (Gen Bank Accession Number KM204118.1)

are indicated in Table S1 (Supporting information). Bidirectional

DNA sequencing was performed using the BrilliantDye TM Termi-

nator Kit v1.1 (Nimagen) with 8 and 6 different primers spanning

the whole NS5 and NS3 regions of ZIKV and DENV, respectively

(Table S1; Supporting Information). The sequencing reactions were

treated with the X-Terminator®Puriﬁcation kit (AppliedBiosys-

tems) in a 96-well plate as suggested by the manufacturer, resolved

4.3.5. Time-of-addition assay

A549 cells were seeded in 12-well plates at a density of

2.5 ꢃ 10⁵cells/ml for 24 h at 37 ^ꢁC, infected with different inﬂuenza

A virus strains as described before and treated or not with 6i at

17

I. Vicenti, M.G. Martina, A. Boccuto et al.

European Journal of Medicinal Chemistry 224 (2021) 113683

by capillary electrophoresis on the 3130 XL Genetic Analyzer

(Applied Biosystems) and analyzed with the DNAStar Lasergene

7.1.0 SeqMan Pro module.

to manufacturer's instructions, using 10e50 ng of enzyme. The

nature of the substrates and their concentration and the presence

of NP-40 or BSA are reported in Table S2 (Supporting information).

All reactions were performed in 10

protein low-binding tubes. Reactions were stopped by adding 5

of phosphoric acid 0.8%. For protein kinases, aliquots (10 L) were

m

L at 30 ^ꢁC for 10 min using

4.3.8. Immunoﬂuorescence

mL

DENV NS3 and NS5 protein expression and colocalization were

visualized by confocal microscopy analysis. Huh-7 cells were

seeded 1 ꢃ 10⁴cells/well onto 1 cm round glass coverslips, infected

m

then transfer into a P30 Filtermat (PerkinElmer), washed ﬁve times

with 75 mM phosphoric acid and once with acetone for 5 min. For

PIK3CA, aliquots (10 mL) were then transfer into a GF/C Filtermat A

with DENV-2 (MOI 0.05), treated with Sofosbuvir (100

mM) or

compound 6i (50

m

M), and ﬁxed in 4% paraformaldehyde/PBS at

(PerkinElmer), washed ﬁve times with 75 mM phosphoric acid and

once with distilled water for 5 min. The ﬁlter was dried and

transferred to a sealable plastic bag and 4 mL scintillation cocktail

was added. Spotted reactions were read in a scintillation counter

(Trilux, PerkinElmer).

different time-points (8-12-16-24 h). After permeabilization (0,5%

Triton-X 100, 7 min), non-speciﬁc binding sites were blocked in 3%

BSA for 1 h and cells were incubated overnight with primary

antibody (anti-NS3 and anti-NS5, diluted in PBS containing 0.5%

BSA þ 0.5% Triton-X 100). After washing (PBS containing 0.5%

BSA þ 0.5% Triton-X 100 þ 0.05% Tween 20), cells were incubated

for 1 h with AlexaFluor® 488- and 568-labeled secondary antibody

diluted in PBS with 0.5% BSA þ 0.5% Triton-X 100. Cells were stained

with 4⁰,6-diamidino-2-phenylindole (DAPI, 10 min). Coverslips

were mounted in Fluoromount Aqueous Mounting Medium and

pictures of stained cells were taken through a confocal microscope

(Zeiss LSM500; Carl Zeiss, Germany). Reagents for immunoﬂuo-

rescence (PBS, Triton-X 100, Tween 20, anti-NS3 and anti-NS5 an-

tibodies, and Fluoromount) were purchased from Sigma-Aldrich

(St. Louis, MO, USA). BSA was purchased from Gibco by Life Tech-

nologies (Carlsbad, CA, USA) and secondary antibodies were from

Invitrogen (Carlsbad, CA, USA). The number of cells was determined

by the count of cells stained with DAPI, and randomly counted (5

ﬁelds) at 40ꢃ original magniﬁcation. NS3 and NS5 expressing cells

were counted and data were reported as percentage of positive

cells/well. Images were pre-processed using ImageJ software (U. S.

National Institutes of Health, Bethesda, MD, USA). The colocaliza-

tion of NS3 and NS5 proteins was represented as white dots, using

the Colocalization plug-in. The quantitative colocalization analyses

were performed using the JACoP plug-in to determine Manders'

colocalization coefﬁcients (MCCs) and Costes' randomization

(Costes P) values. Costes P values ꢄ 95% were considered as a

colocalization signal. Statistical analyses were performed using

Prism 6 software (Graphpad Software, La Jolla, CA, USA). Differ-

ences between treated and untreated control were estimated using

2-way ANOVA followed by Dunnett's multiple comparisons test.

4.3.11. In vitro polymerase and MTase assays

The polymerase assays were performed as previously described

[51]. Brieﬂy, the compounds were added in wells (2 mL, 5% DMSO

ﬁnal concentration), and the enzyme was distributed in wells. Re-

actions were started by the addition of the nucleotide mix (100 mM

ATP) and were incubated at 30 C for 10 min before stopping by

EDTA (100 mM) addition. Reactions were revealed by addition of

Picogreen® ﬂuorescent reagent (Molecular probes), diluted to 1/

800 in TE buffer and the ﬂuorescence signal was then read at

480 nm (excitation) and 530 nm (emission) using a TecanSaﬁre2.

IC₅₀was determined using the equation: % of active enzyme ¼ 100/

(1þ(I)2/IC50), where I is the concentration of inhibitor and 100% of

activity is the ﬂuorescence intensity without inhibitor. The MTase

activity was determined by ﬁlter binding assay as previously

described [50]. The MTase activity was assayed by incubating the

MTase in presence of increasing inhibitor concentration together

with 7MeGpppAC4 and [3H] AdoMet. The assay was performed in

20

m

L samples containing 40 mM TriseHCl pH 7.5, 5 mM DTT, 10

m

M

AdoMet (0.2e2

m

Ci [3H] AdoMet), 1 M of MTase and 1

m

7MeGpppAC4 [52]. Reactions were incubated at 30 ^ꢁC for 30 min

and stopped by 20-fold dilution in ice-cold 100 mM AdoHcy solu-

tion. Samples were then transferred onto DEAE membrane (DEAE

Filtermat; Wallac) by a Filtermat Harvester (Packard Instruments)

washed with 0.01 M ammonium formate (pH 8.0), water and

ethanol, and the radioactivity transferred onto RNA was measured

using a Wallac 1450 MicroBeta Trilux Liquid Scintillation Counter.

DENV3 N7-MTase activity was performed as described above but

Values were expressed as the mean

considered statistically signiﬁcant.

SD and p < 0.05 was

using of 0.5

m

M GpppA2⁰OMe-RNA74 as substrate. The different

protein used for polymerase and MTase assays were produces and

puriﬁed as preciously describes [51,53].

4.3.9. Statistical analysis

Statistical analyses were performed using Prism 6 software

(Graphpad Software, La Jolla, CA, USA). Differences between treated

and untreated control were estimated using 2-way ANOVA fol-

lowed by Dunnett's multiple comparisons test. Values were

expressed as the mean SD and p < 0.05 was considered statisti-

cally signiﬁcant.

Author contributions

The manuscript was written by M.R. through contributions of all

authors, who have given approval to the ﬁnal version of the

manuscript.

4.3.10. In vitro kinase inhibition assay

PIK3CA substrate. 1.1 mg of phosphatidylinositol (PI) (Sigma,

Darmstadt, Germany) and 3.3 mg of 2-oleoyl-1-palmitoyl-sn-glyc-

Notes

M. Z. reports consultancy for ViiV Healthcare, Gilead Sciences

and Janssen-Cilag, and grants for his institution from ViiV Health-

care and Gilead outside the submitted work. All other authors

declare no competing ﬁnancial interest.

ero-3-phospho-L-serine (PS) (Sigma, Darmstadt, Germany) were

dissolved in 9:1 chloroform/methanol solution in TPX Poly-

methylpentene microtubes (Diagenode e Liege, Belgium). The lipid

mixture was dried, resuspended in distilled water and sonicated for

13 cycles (1 min ON and 30 s OFF) by the Bioruptor® Plus sonication

device (Diagenode e Liege, Belgium). The PI:3 PS lipid solution was

ﬁnally transferred into low-binding tubes and stored at ꢀ20 ^ꢁC.

PI:3 PS stock solution was than prepared to a ﬁnal concentration of

Declaration of competing interest

The authors declare that they have no known competing

ﬁnancial interests or personal relationships that could have

appeared to inﬂuence the work reported in this paper.

750

mM in TrisHCl 50 mM, EGTA 1.25 mM.

Kinase reactions. All kinase reactions were performed according

18

I. Vicenti, M.G. Martina, A. Boccuto et al.

European Journal of Medicinal Chemistry 224 (2021) 113683

J. Neyts, P. Leyssen, G. Costantino, M. Radi, Discovery of multi-target antivirals

acting on both the dengue virus NS5-NS3 interaction and the host Src/Fyn

kinases, J. Med. Chem. 58 (2015) 4964e4975.

Acknowledgment

This work was supported by the Ministero dell'Istruzione, del-

l'Universita della Ricerca Italiano (MIUR), PRIN 2017 project (cod.

2017BMK8JR) “ORIGINALE CHEMIAE in Antiviral Strategy - Origin

and Modernization of Multi-Component Chemistry as a Source of

Innovative Broad Spectrum Antiviral Strategy” to M.R., M.Z., L.N., by

Tuscany region, project Tuscany Antiviral Research Network

-TUSCAVIR.NET (Bando Ricerca Salute 435, 2018) (to M.Z.) and by

Sapienza University of Rome grants (L.N.). E.C. was supported by

AIRC: MFAG Id.18811. G.M. was supported by Fondazione Banca del

Monte di Lombardia. Francesca Preziuso is kindly acknowledged for

the support on the ABTS assay. We also thank Priscila Sutto-Ortiz

for excellent technical assistance.

[21] S. Tassini, L. Sun, K. Lanko, E. Crespan, E. Langron, F. Falchi, M. Kissova,

J.I. Armijos-Rivera, L. Delang, C. Mirabelli, J. Neyts, M. Pieroni, A. Cavalli,

G. Costantino, G. Maga, P. Vergani, P. Leyssen, M. Radi, Discovery of multi-

target agents active as broad-spectrum antivirals and correctors of cystic

ﬁ brosis transmembrane conductance regulator (CFTR) for associated pulmo-

nary diseases, J. Med. Chem. 60 (2017) 1400e1416.

[22] S.J.F. Kaptein, P. Vincetti, E. Crespan, J.I. Armijos-Rivera, G. Costantino,

G. Maga, J. Neyts, M. Radi, Identi ﬁ cation of broad-spectrum dengue/Zika virus

replication inhibitors by functionalization of quinoline and 2,6-diaminopurine

scaffolds, ChemMedChem 13 (2018) 1371e1376.

[23] S. Tassini, E. Langron, L. Delang, C. Mirabelli, K. Lanko, E. Crespan, M. Kissova,

ꢀ

G. Tagliavini, G. Fonto, S. Bertoni, S. Palese, C. Giorgio, F. Ravanetti,

L. Ragionieri, C. Zamperini, A. Mancini, E. Dreassi, G. Maga, P. Vergani, J. Neyts,

M. Radi, Multi-target CFTR modulators endowed with multiple bene ﬁ cial side

effects for cystic ﬁ brosis patients: toward a simpli ﬁ ed therapeutic approach,

J. Med. Chem. 62 (2019) 10833e10847.

[24] I. Vicenti, F. Dragoni, A. Giannini, F. Giammarino, M. Spinicci, F. Saladini,

A. Boccuto, M. Zazzi, Development of a cell-based immunodetection assay for

simultaneous screening of antiviral compounds inhibiting Zika and dengue

virus replication, SLAS Discov. 25 (2020) 506e514.

Appendix A. Supplementary data

Supplementary data to this article can be found online at

https://doi.org/10.1016/j.ejmech.2021.1 13683.

ꢁ

[25] M.C. Puertas, M.J. Buzon, M. Ballestero, P. Van Den Eede, B. Clotet, J.G. Prado,

J. Martinez-Picado, Novel two- round phenotypic assay for protease inhibitor

susceptibility testing of recombinant and primary HIV-1 isolates, J. Clin.

Microbiol. 50 (2012) 3909e3916.

References

[26] F. Saladini, A. Giannini, A. Boccuto, I. Vicenti, M. Zazzi, Agreement between an

in-house replication competent and a reference replication defective recom-

binant virus assay for measuring phenotypic resistance to HIV-1 protease,

reverse Transcriptase, and integrase inhibitors, J. Clin. Lab. Anal. 32 (2017),

e22206.

[27] P. Checconi, M. De Angelis, M.E. Marcocci, A. Fraternale, M. Magnani,

A.T. Palamara, L. Nencioni, Redox-modulating agents in the treatment of viral

infections, Int. J. Mol. Sci. 21 (2020) 4084.

[28] L.F.B. Duarte, R.L. Oliveira, K.C. Rodrigues, G.T. Voss, B. Godoi, R.F. Schumacher,

G. Perin, E.A. Wilhelm, C. Luchese, D. Alves, Organoselenium compounds from

purines: synthesis of 6-arylselanylpurines with antioxidant and anticholin-

esterase activities and memory improvement effect, Bioorg. Med. Chem. 25

(2017) 6718e6723.

[29] A. Brathe, G. Andresen, L.-L. Gundersen, K.E. Malterud, F. Rise, Antioxidant

activity of synthetic cytokinin analogues: 6-alkynyl- and 6-alkenylpurines as

novel 15-lipoxygenase inhibitors, Bioorg. Med. Chem. 10 (2002) 1581e1586.

[30] E.N. Djuidje, V. Dissette, B. Bino, S. Benetti, J. Balzarini, S. Liekens,

S. Manfredini, S. Vertuani, A. Baldisserotto, A multitarget approach toward the

development of 8- substituted purines for photoprotection and prevention of

[1] E. De Clercq, G. Li, Approved antiviral drugs over the past 50 years, Clin.

Microbiol. Rev. 29 (2016) 695e747.

[2] J.A. Symons, Deval, Innovation and trends in the development and approval of

antiviral medicines: 1987e2017 and beyond, J. Antiviral Res. 155 (2018)

76e88.

[3] D. Carroll, P. Daszak, N.D. Wolfe, G.F. Gao, C.M. Morel, S. Morzaria, A. Pablos-

ꢁ

Mendez, O. Tomori, J.A.K. Mazet, The global Virome project, Science 359

̀

(2018) 872e874.

[4] T.C. Pierson, M.S. Diamond, The continued threat of emerging ﬂ aviviruses, Nat.

Microbiol. 5 (2020) 796e812.

[5] L. Yuan, et al., A single mutation in the prM protein of Zika virus contributes to

fetal microcephaly, Science 358 (2017) 933e936.

[6] D. Tomasello, P. Schlagenhauf, Chikungunya and dengue autochthonous cases

in Europe, 2007-2012. Travel , Med. Infect. Dis. 11 (2013) 274e284.

[7] C.M. Gossner, E. Ducheyne, F. Schaffner, Increased risk for autochthonous

vector-borne infections transmitted by Aedes albopictus in continental

Europe, Euro Surveill. 23 (24) (2018) 1800268.

ꢁ

[8] A.J. Rodriguez-Morales, W.E. Villamil-Gomez, C. Franco-Paredes, The arboviral

[31] D. Kolarski, W. Szymanski, B.L. Feringa, Two-step, one-pot synthesis of visible-

light-responsive 6-azopurines, Org. Lett. 19 (2017) 5090e5093.

[32] K. Too, D.M. Brown, E. Bongard, V. Yardley, L. Vivas, D. Loakesa, Anti-malarial

activity of N6-modi ﬁ ed purine analogues, Bioorg. Med. Chem. 15 (2007)

5551e5562.

burden of disease caused by co-circulation and co-infection of dengue, chi-

kungunya and Zika in the Americas, Trav. Med. Infect. Dis. 14 (2016) 177e179.

[9] O. Olawoyin, C. Coinfection Kribs, Altered vector infectivity, and antibody-

dependent enhancement: the dengueezika interplay, Bull. Math. Biol. 82

(2020) 1e13.

[33] R. Re, N. Pellegrini, A. Proteggente, A. Pannala, M. Yang, C. Rice-Evans, Anti-

oxidant activity applying an improved ABTS radical cation decolorization

assay, Free Radic. Biol. Med. 26 (1999) 1231e1237.

[34] J. Park, G. Avila-Perez, A. Nogales, P. Blanco-Lobo, J. de la Torre, L. Martínez-

Sobrido, Identi ﬁ cation and characterization of novel compounds with broad-

spectrum antiviral activity against in ﬂ uenza A and B viruses, J. Virol. 94 (2020)

e02149-19.

[10] C. Lorenz, T.S. Azevedo, F. Chiaravalloti-Neto, COVID-19 and dengue fever: a

dangerous combination for the health system in Brazil, Trav. Med. Infect. Dis.

35 (2020) 101659.

[11] H. Harapan, M. Ryan, B. Yohan, R.S. Abidin, F. Nainu, A. Rakib, I. Jahan,

T.B. Emran, I. Ullah, K. Panta, K. Dhama, R.T. Sasmono, Covid-19 and dengue:

double punches for dengue-endemic countries in Asia, Rev. Med. Virol. 31 (2)

(2021), e2161.

ꢁ

[35] D. Amatore, R. Sgarbanti, K. Aquilano, S. Baldelli, D. Limongi, L. Civitelli,

L. Nencioni, E. Garaci, M.R. Ciriolo, A.T. Palamara, In ﬂ uenza virus replication in

lung epithelial cells depends on redox-sensitive pathways activated by NOX4-

derived ROS, Cell Microbiol. 17 (2015) 131e145.

[12] M.A. Schmid, K.N. Gonzalez, S. Shah, J. Pena, M. Mack, L.B. Talarico, F.P. Polack,

E. Harris, In ﬂ uenza and dengue virus co-infection impairs monocyte recruit-

ment to the lung, increases dengue virus titers, and exacerbates pneumonia,

Eur. J. Immunol. 47 (2017) 527e539.

[36] M. Wang, R. Cao, L. Zhang, X. Yang, J. Liu, M. Xu, Z. Shi, Z. Hu, W. Zhong,

G. Xiao, Remdesivir and chloroquine effectively inhibit the recently emerged

novel coronavirus (2019-NCoV) in vitro, Cell Res. 30 (2020) 269e271.

€

[13] S. Hansen, S.-K. Hotop, O. Faye, O. Ndiaye, S. Bohlken-Fascher, R. Pesso^a,

F. Hufert, C. Stahl-Hennig, R. Frank, C.-P. Czerny, J. Schmidt-Chanasit,

€

S.S. Sanabani, A.A. Sall, M. Niedrig, M. Bronstrup, H.-J. Fritz, A.A. El Wahed,

€

[37] A. Pruijssers, A. George, A. Schafer, S. Leist, L. Gralinksi, K. Dinnon, B. Yount,

Diagnosing Zika virus infection against a background of other ﬂ aviviruses:

studies in high resolution serological analysis, Sci. Rep. 9 (2019) 3648.

[14] Henrina, J.; Santosa Putra, I. C.; Lawrensia, S.; Handoyono, Q. F.; Cahyadi, A.

Coronavirus disease of 2019: a mimicker of dengue infection? Sn Compr. Clin.

Med. 10.1007/s42399-020-00364-3.

[15] N. Kumar, S. Sharma, R. Kumar, B.N. Tripathi, S. Barua, H. Ly, B.T. Roused, Host-

directed antiviral therapy, Clin. Microl. Rev. 33 (3) (2020) e00168-19.

[16] A. Edwards, What are the odds of ﬁ nding a COVID-19 drug from a lab

repurposing screen? J. Chem. Inf. Model. 60 (12) (2020) 5727e5729, in press.

[17] Designing multi-target drugs, in: J.R. Morphy (Ed.), RSC Drug Discovery Series,

RSC Publ, Cambridge, 2012.

M. Agostini, L. Stevens, J. Chappell, X. Lu, T. Hughes, K. Gully, D. Martinez,

A. Brown, R. Graham, J. Perry, V. Du Pont, J. Pitts, B. Ma, D. Babusis,

E. Murakami, J. Feng, J. Bilello, D. Porter, T. Cihlar, R. Baric, M. Denison,

T. Sheahan, Remdesivir inhibits SARS-CoV-2 in human lung cells and chimeric

SARS-CoV expressing the SARS-CoV-2 RNA polymerase in mice, Cell Rep. 32

(2020) 107940.

[38] K. Choy, A. Wong, P. Kaewpreedee, S. Sia, D. Chen, K. Hui, D. Chu, M. Chan,

P. Cheung, X. Huang, M. Peiris, H. Yen, Remdesivir, lopinavir, emetine, and

homoharringtonine inhibit SARS-CoV-2 replication in vitro, Antivir. Res. 178

(2020) 104786.

[39] S. De Meyer, D. Bojkova, J. Cinatl, E. Van Damme, C. Buyck, M. Van Loock,

B. Woodfall, S. Ciesek, Lack of antiviral activity of darunavir against SARS-CoV-

2, Int. J. Infect. Dis. 97 (2020) 7e10.

[40] M.J. So ﬁ a, D. Bao, W. Chang, J. Du, D. Nagarathnam, S. Rachakonda, P.G. Reddy,

B.S. Ross, P. Wang, H.-R. Zhang, S. Bansal, C. Espiritu, M. Keilman, A.M. Lam,

[18] M. Bizzarri, A. Giuliani, N. Monti, R. Verna, A. Pensotti, A. Cucina, Rediscovery

of natural compounds acting via multitarget recognition and noncanonical

pharmacodynamical actions, Drug Discov. Today 25 (2020) 920e927.

[19] K. Nystrom, J. Waldenstrom, K.-W. Tang, M. Lagging, Ribavirin: pharmacology,

multiple modes of action and possible future perspectives, Future Virol. 14

(2019) 153e160.

H.M. Micolochick Steuer, C. Niu, M.J. Otto, P.A. Furman, Discovery of a

Deoxy-2 ’- -ﬂ uoro-20- -C-methyluridine nucleotide prodrug (PSI-7977) for

the treatment of hepatitis C virus, J. Med. Chem. 53 (2010) 7202e7218.

b -D-20-

a

b

[20] P. Vincetti, F. Caporuscio, S. Kaptein, A. Gioiello, V. Mancino, Y. Suzuki,

N. Yamamoto, E. Crespan, A. Lossani, G. Maga, G. Rastelli, D. Castagnolo,

19

I. Vicenti, M.G. Martina, A. Boccuto et al.

European Journal of Medicinal Chemistry 224 (2021) 113683

ꢁ

~

[41] D. Siegel, H.C. Hui, E. Doerf ﬂ er, M.O. Clarke, K. Chun, L. Zhang, S. Neville,

E. Carra, W. Lew, B. Ross, Q. Wang, L. Wolfe, R. Jordan, V. Soloveva, J. Knox,

J. Perry, M. Perron, K.M. Stray, O. Barauskas, J.Y. Feng, Y. Xu, G. Lee,

A.L. Rheingold, A.S. Ray, R. Bannister, R. Strickley, S. Swaminathan, W.A. Lee,

S. Bavari, T. Cihlar, M.K. Lo, T.K. Warren, R.L. Mackman, Discovery and syn-

B. Gutierrez Castaneda, Caveolin-1 in lipid rafts interacts with dengue virus

NS3 during polyprotein processing and replication in HMEC-1 cells, PloS One

9 (2014), e90704.

[58] N. Kumar, S. Sharma, R. Kumar, B.N. Tripathi, S. Barua, H. Ly, B.T. Rouse, Host-

directed antiviral therapy, Clin. Microbiol. Rev. 33 (3) (2020) e00168-19.

[59] R. Meineke, G. Rimmelzwaan, H. Elbahesh, In ﬂ uenza virus infections and

cellular kinases, Viruses 11 (2) (2019) 171.

[60] E. Weisberg, A. Parent, P.L. Yang, M. Sattler, Q. Liu, Q. Liu, J. Wang, C. Meng,

S.J. Buhrlage, N. Gray, J.D. Grif ﬁ n, Repurposing of kinase inhibitors for treat-

ment of COVID-19, Pharm. Res. (N. Y.) 37 (9) (2020) 167.

[61] H. Huang, J. Ma, J. Shi, L. Meng, H. Jiang, J. Ding, H. Liu, Discovery of novel

purine derivatives with potent and selective inhibitory activity against c-src

tyrosine kinase, Bioorg. Med. Chem. 18 (2010) 4615e4624.

[62] R.L. Davis, Mechanism of action and target identi ﬁ cation: a matter of timing in

drug discovery, iScience 23 (9) (2020) 101487.

thesis of

a phosphoramidate prodrug of a pyrrolo[2,1-f][triazin-4-amino]

adenine C-nucleoside (GS-5734) for the treatment of ebola and emerging

viruses, J. Med. Chem. 60 (2017) 1648e1661.

[42] A. Dubankova, E. Boura, Structure of the yellow fever NS5 protein reveals

conserved drug targets shared among ﬂ aviviruses, Antivir. Res. 169 (2019)

104536.

[43] J. Wu, W. Liu, P. Gong, A structural overview of RNA-dependent RNA poly-

merases from the ﬂ aviviridae family, Int. J. Mol. Sci. 16 (2015) 12943e12957.

[44] G. Zou, Y.-L. Chen, H. Dong, C.C. Lim, L.J. Yap, Y.H. Yau, S.G. Shochat, J. Lescar,

P.-Y. Shi, Functional analysis of two cavities in ﬂ avivirus NS5 polymerase,

J. Biol. Chem. 286 (2011) 14362e14372.

[63] J.G. Moffat, F. Vincent, J.A. Lee, J. Eder, M. Prunotto, Opportunities and chal-

lenges in phenotypic drug discovery: an industry perspective, Nat. Rev. Drug

Discov. 16 (8) (2017) 531e543.

[45] M.Y.F. Tay, S.G. Vasudevan, The transactions of NS3 and NS5 in ﬂ aviviral RNA

replication, Adv. Exp. Med. Biol. 1062 (2018) 147e163.

[46] C. Sarto, S.B. Kaufman, D.A. Estrin, M. Arrar, Nucleotide-dependent dynamics

of the dengue NS3 helicase, Biochim. Biophys. Acta Protein Proteonomics

1868 (2020) 140441.

[47] J. Wu, A.K. Bera, R.J. Kuhn, J.L. Smith, Structure of the ﬂ avivirus helicase: im-

plications for catalytic activity, protein interactions, and proteolytic process-

ing, J. Virol. 79 (2005) 10268e10277.

[48] W.G. Saw, A. Pan, M.S. Subramanian Manimekalai, A. Grüber, G. Grüber,

Structure and ﬂ exibility of non-structural proteins 3 and -5 of dengue- and

Zika viruses in solution, Prog. Biophys. Mol. Biol. 143 (2019) 67e77.

[49] M.Y.F. Tay, W.G. Saw, Y. Zhao, K.W.K. Chan, D. Singh, Y. Chong, J.K. Forwood,

E.E. Ooi, G. Grüber, J. Lescar, D. Luo, S.G. Vasudevan, The C-terminal 50 amino

acid residues of dengue NS3 protein are important for NS3-NS5 interaction

and viral replication, J. Biol. Chem. 290 (2015) 2379e2394.

[50] K. Barral, C. Sallamand, C. Petzold, B. Coutard, A. Collet, Y. Thillier,

J. Zimmermann, J.-J. Vasseur, B. Canard, J. Rohayem, F. Debart, E. Decroly,

Development of speci ﬁ c dengue virus 2⁰-O- and N7-methyltransferase assays

for antiviral drug screening, Antivir. Res. 99 (2013) 292e300.

[64] A. Lai, A. Bergna, S. Caucci, N. Clementi, I. Vicenti, F. Dragoni, A.M. Cattelan,

S. Menzo, A. Pan, A. Callegaro, A. Tagliabracci, A. Caruso, F. Caccuri,

S. Ronchiadin, C. Balotta, M. Zazzi, E. Vaccher, M. Clementi, M. Galli,

G. Zehender, Molecular tracing of SARS-CoV-2 in Italy in the ﬁ rst three

months of the epidemic, Viruses 12 (8) (2020) 798.

[65] I. Vicenti, A. Boccuto, A. Giannini, F. Dragoni, F. Saladini, M. Zazzi, Comparative

analysis of different cell systems for Zika virus (ZIKV) propagation and eval-

uation of anti-ZIKV compounds in vitro, Virus Res. 244 (2018) 64e70.

[66] F. Dragoni, A. Boccuto, F. Picarazzi, A. Giannini, F. Giammarino, F. Saladini,

M. Mori, E. Mastrangelo, M. Zazzi, I. Vicenti, Evaluation of sofosbuvir activity

and resistance pro ﬁ le against West Nile virus in vitro, Antivir. Res. 175 (2020)

104708.

[67] A. Di Sotto, P. Checconi, I. Celestino, M. Locatelli, S. Carissimi, M. De Angelis,

V. Rossi, D. Limongi, C. Toniolo, L. Martinoli, S. Di Giacomo, A.T. Palamara,

L. Nencioni, Antiviral and antioxidant activity of a hydroalcoholic extract from

humulus lupulus L, Oxid. Med. Cell. Longev. 2018 (2018) 1e14.

[68] P. Checconi, R. Sgarbanti, I. Celestino, D. Limongi, D. Amatore, A. Iuvara,

A. Alimonti, E. Garaci, A.T. Palamara, L. Nencioni, The environmental pollutant

cadmium promotes in ﬂ uenza virus replication in MDCK cells by altering their

redox state, Int. J. Mol. Sci. 14 (2013) 4148e4162.

[51] F. Benmansour, C. Eydoux, G. Querat, X. de Lamballerie, B. Canard, K. Alvarez,

J.-C. Guillemot, K. Barral, Novel 2-phenyl-5-[(E)-2-(Thiophen-2-Yl)Ethenyl]-

1,3,4-Oxadiazole

and

3-phenyl-5-[(E)-2-(Thiophen-2-Yl)Ethenyl]-1,2,4-

Oxadiazole derivatives as dengue virus inhibitors targeting NS5 polymerase,

Eur. J. Med. Chem. 109 (2016) 146e156.

[69] B.M. Bizzarri, A. Fanelli, D. Piccinino, M. De Angelis, C. Dolfa, A.T. Palamara,

L. Nencioni, C. Zippilli, M. Crucianelli, R. Saladino, Synthesis of stilbene and

chalcone inhibitors of in ﬂ uenza A virus by sba -15 supported hoveyda-grubbs

metathesis, Catalysts 9 (2019) 983.

[70] D. Lagorce, L. Bouslama, J. Becot, M.A. Miteva, B.O. Villoutreix, FAF-Drugs4:

free ADME-tox ﬁ ltering computations for chemical biology and early stages

drug discovery, Bioinformatics 33 (2017) 3658e3660.

[52] B. Selisko, F.F. Peyrane, B. Canard, K. Alvarez, E. Decroly, Biochemical char-

acterization of the (Nucleoside-2 ’O)-Methyltransferase activity of dengue vi-

rus protein NS5 using puri ﬁ ed capped RNA oligonucleotides 7MeGpppACn

and GpppACn, J. Gen. Virol. 91 (1) (2010), 112e12.1.

ꢀ

[53] B. Coutard, K. Barral, J. Lichiere, B. Selisko, B. Martin, W. Aouadi,

M.O. Lombardia, F. Debart, J.-J. Vasseur, J.C. Guillemot, B. Canard, E. Decroly,

Zika virus methyltransferase: structure and functions for drug design per-

spectives, J. Virol. 91 (5) (2017) e02202-16, e02202-e02216.

[71] A.V. Stachulski, J. Taujanskas, S.L. Pate, R.K.R. Rajoli, G. Aljayyoussi,

S.H. Pennington, S.A. Ward, W.D. Hong, G.A. Biagini, A. Owen, G.L. Nixon,

S.C. Leung, P.M. O ’Neill, Therapeutic potential of nitazoxanide: an appropriate

choice for repurposing versus SARS-CoV-2? ACS Infect. Dis. 7 (6) (2021)

1317e1331.

[54] H. Hannemann, P.-Y. Sung, H.-C. Chiu, A. Yousuf, J. Bird, S.P. Lim,

A.D. Davidson, Serotype-speci ﬁ c differences in dengue virus non-structural

protein 5 nuclear localization, J. Biol. Chem. 288 (2013) 22621e22635.

[55] W. Ji, G. Luo, Zika virus NS5 nuclear accumulation is protective of protein

degradation and is required for viral RNA replication, Virology 541 (2020)

124e135.

ꢁ

[72] J. Blaising, S.J. Polyak, E.-I. Pecheur, Arbidol as a broad-spectrum antiviral: an

update, Antivir. Res. 107 (2014) 84e94.

[73] S. Xiu, A. Dick, H. Ju, S. Mirzaie, F. Abdi, S. Cocklin, P. Zhan, X. Liu, Inhibitors of

SARS-CoV-2 entry: current and future opportunities, J. Med. Chem. 63 (2020)

12256e12274.

[74] R. Raghuvanshi, S.B. Bharate, Recent developments in the use of kinase in-

hibitors for management of viral infections, J. Med. Chem. (2021) (acs.jmed-

chem.0c01467). In press.

[56] A.J. Lopez-Denman, A. Russo, K.M. Wagstaff, P.A. White, D.A. Jans,

J.M. Mackenzie, Nucleocytoplasmic shuttling of the West Nile virus RNA

-dependent RNA polymerase NS5 is critical to infection, Cell Microbiol. 20

(2018), e12848.

ꢁ

[57] J. García Cordero, M. Leon Juarez, J.A. Gonzalez-Y-Merchand, L. Cedillo Barron,

20

Article Doi

System-oriented optimization of multi-target 2,6-diaminopurine derivatives: Easily accessible broad-spectrum antivirals active against flaviviruses, influenza virus and SARS-CoV-2

DOI: 10.1016/j.ejmech.2021.113683

Source and publish data:

Authors:

Article abstract of DOI:10.1016/j.ejmech.2021.113683

Full text of DOI:10.1016/j.ejmech.2021.113683

Products guided by the article

R&D Labs maybe for 190654-85-4

Relevant to this article

Hot Product