9-Aminoacridine-based agents impair the bovine viral diarrhea virus (BVDV) replication targeting the RNA-dependent RNA polymerase (RdRp)

Article abstract of DOI:10.1016/j.bmc.2018.01.001

Bovine viral diarrhea virus (BVDV) infection is still a plague that causes important livestock pandemics. Despite the availability of vaccines against BVDV, and the implementation of massive eradication or control programs, this virus still constitutes a

Full text of DOI:10.1016/j.bmc.2018.01.001

Accepted Manuscript
9-Aminoacridine-based agents impair the bovine viral diarrhea virus (BVDV)
replication targeting the RNA-dependent RNA polymerase (RdRp)
Roberta Loddo, Valeria Francesconi, Erik Laurini, Silvia Boccardo, Suzana
Aulic, Maurizio Fermeglia, Sabrina Pricl, Michele Tonelli
PII:
S0968-0896(17)31921-1
https://doi.org/10.1016/j.bmc.2018.01.001
BMC 14155
DOI:
Reference:
Bioorganic & Medicinal Chemistry
To appear in:
Received Date:
Revised Date:
Accepted Date:
28 September 2017
5 December 2017
3 January 2018
Please cite this article as: Loddo, R., Francesconi, V., Laurini, E., Boccardo, S., Aulic, S., Fermeglia, M., Pricl, S.,
Tonelli, M., 9-Aminoacridine-based agents impair the bovine viral diarrhea virus (BVDV) replication targeting the
RNA-dependent RNA polymerase (RdRp), Bioorganic & Medicinal Chemistry (2018), doi: https://doi.org/10.1016/
j.bmc.2018.01.001
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

9-Aminoacridine-based agents impair the bovine viral diarrhea virus (BVDV) replication
targeting the RNA-dependent RNA polymerase (RdRp)
Roberta Loddo,a,* Valeria Francesconi,b Erik Laurini,c Silvia Boccardo,c Suzana Aulic,c Maurizio
Fermeglia,c Sabrina Pricl,c,§ Michele Tonellib,*,§
aDepartment of Biomedical Sciences and Technology, University of Cagliari, Cittadella
Universitaria, 09042 Monserrato (CA), Italy
bDepartment of Pharmacy, University of Genoa, Viale Benedetto XV n. 3, 16132, Genoa, Italy
cMolecular Simulation Engineering (MOSE) Laboratory, Department of Chemical, Environmental,
and Raw Materials Engineering, University of Trieste, Piazzale Europa 1, 34127 Trieste, Italy
*Corresponding authors. Tel.: +39 070 6754202; fax: +39 070 6754210 (R. Loddo); tel.: +39 010
3538378; fax: +39 010 3538399 (M. Tonelli). E-mail addresses: rloddo@unica.it (R. Loddo),
tonelli@difar.unige.it (M. Tonelli).
§Senior co-authors.
Abstract
Bovine viral diarrhea virus (BVDV) infection is still a plague that causes important livestock
pandemics. Despite the availability of vaccines against BVDV, and the implementation of massive
eradication or control programs, this virus still constitutes a serious agronomic burden. Therefore,
the alternative approach to combat Pestivirus infections, based on the development of antiviral
agents that specifically inhibit the replication of these viruses, is of preeminent actuality and
importance.
Capitalizing from a long-standing experience in antiviral drug design and development, in this work
we present and characterize a series of small molecules based on the 9-aminoacridine scaffold that
exhibit potent anti-BVDV activity coupled with low cytotoxicity. The relevant viral protein target –
the RNA-dependent RNA polymerase – the binding mode, and the mechanism of action of these
new antivirals have been determined by a combination of in vitro (i.e., enzymatic inhibition,
isothermal titration calorimetry and site-directed mutagenesis assays) and computational
experiments. The overall results obtained confirm that these acridine-based derivatives are
promising compounds in the treatment of BVDV infections and, based on the reported structure-
activity relationship, can be selected as a starting point for the design of a new generation of
improved, safe and selective anti-BVDV agents.

Keywords: 9-aminoacridine derivatives; anti-BVDV activity; RNA polymerase inhibition;
isothermal titration calorimetry; computational studies.
1. Introduction
The Bovine Viral Diarrhea - Mucosal Disease (BVD-MD) is a highly contagious infectious disease
that affects cattle. The causative agent is a single-stranded RNA virus (bovine viral diarrhea virus or
BVDV) which, together with the classical swine fever (CSFV) virus and the border disease virus
(BDV) of sheep, belongs to the Pestivirus genus (Flaviviridae family). Pestiviruses are responsible
for a range of clinical manifestations including abortion, teratogenesis, respiratory problems,
chronic wasting disease, immune system dysfunction and predisposition to secondary viral and
bacterial infections. BVDV, the prototype of this genus, is a major pathogen of cattle and can also
establish a persistent infection in animals that remain viremic throughout life and serve as
continuous virus reservoirs.1 Infection by the ruminant pestiviruses often results in mild clinical
signs, allowing the disease to spread undetected through the infection of the susceptible fetus and
the creation of additional persistently infected (PI) animals.2 At the same time the role of such
animals in disease transmission has led to the current recommendation of identification and culling,
thus the early detection of infections with sensitive and specific methods is critical to the success of
control plans.3
From the economical standpoint, in the United States BVDV infection alone leads to substantial
costs to the private-public sector (from 10 to 40 million USD for million calves) through decreased
reproductive performance as direct losses, and increased control efforts as indirect losses.2 The
direct losses of BVDV infection, such as reduced milk yield, respiratory disorders, congenital
defects, growth retardation, extended calving intervals, reduced first service conception, and
increased mortality of animals due to immunosuppression can justify the implementation of
programs to prevent or mitigate the disease.4 CSFV also represents a cause for major economic
losses, especially in those countries with industrialized pig breeding.5 Even if the access to vaccines
against BVDV and CSFV,6 and the implementation of massive eradication or control programs3b
are exploited strategies, both viruses still constitute a serious agronomic burden. Although not likely
suited to treat large herds, antiviral chemotherapy may provide specific and selective inhibitors of
Pestiviruses replication, which might contribute substantially to the control of the viral outbreaks.7
Yet, while in vitro research has been promising (e.g., several potent anti-BVDV agents have been
developed as RNA polymerase-inhibitors8-12), the paucity of efficacy studies in animals has
hindered the commercial development of effective antiviral drugs against Pestiviruses.7

In the field of small molecule drug development, the acridine nucleus is an important scaffold
associated with a broad spectrum of biological properties including antimicrobial, antiviral,
antimalarial, antitrypanosomic, leishmanicidal, anti-prion and, in particular, anticancer activities.13
Easy to synthesize, the acridine tricyclic system represents a valuable chemical space in virtue to
the presence of ten substitution sites and ten fusion points on its structure that led to the
identification of a huge number of functionalized derivatives for medicinal purposes. Most of the
biological effects of an acridine-based compound are related to its planar structure14 responsible for
intercalation between DNA base pairs, inducing the disruption of the cellular machinery.15
Substances of clinical importance, such as amsacrine, as well as several drugs used for the treatment
of parasitic diseases (e.g., quinacrine), have the ability to bind to DNA by intercalation as principle
mechanism of action,15,16 but they also demonstated to hit specific factors involved in replication
and/or transcription pathways of viral life cycle. Several research groups have focused on acridine
derivatives for cancer therapy, owing to their wide range of biological properties and to their ability
to interact with different targets.17,18 Concerning the antiviral activity of small molecules, based on
acridine scaffold, initial studies have been directed towards DNA viruses (i.e., the Herpesviridae
herpes symplex virus (HSV) and human cytomegalovirus (HCMV)), relying on the assumption that
virus inhibition occurred by targeting host cellular enzymes involved in DNA replication.19 Most of
antiviral research efforts have been oriented to study acridones,18 as a subclass of acridines, since
the 9(10H)-acridone basic structure is present in a large number of natural products and synthetic
compounds, providing a wide range of antimicrobial properties.19,20 The interest of acridones as
antivirals was next extended to single stranded RNA viruses belonging to the Flaviviridae family,19
which represents a serious hazard for public health. A work proved that some intercalating drugs,
including quinacrine, were capable of inhibiting hepatitis C virus IRES-mediated translation by
impairing the binding of the HCV Internal ribosome entry sites (IRES) to the 40S/80S ribosome.21
IRESs are required to start translation of some viral and host cell mRNAs and may be characterized
by complex secondary and tertiary structures forming regions of double stranded RNA.22 Likewise,
quinacrine and 9-aminoacridine (aminacrine) intercalate into the RNA architecture and block IRES-
mediated mRNA translation of encephalomyocarditis virus (EMCV) and poliovirus, not interfering
with IRESs present in the host cell transcripts, that are featured by less-complex structures.22 In
another study, quinacrine was highly effective (in a dose-dependent manner) by blocking IRES
binding to cellular factors such as the Polypyrimidine-Tract Binding Protein (PTB) in enterovirus
71 (EV71) and some Coxsackie viruses infected cell cultures.23 In addition, quinacrine and some
other structurally-related intercalating antimalarial drugs, were suggested to have some anti-human
immunodeficiency virus (HIV) effects.24 Interestingly, quinacrine and chlorpromazine were

reported as inhibitors of tomato bushy stunt virus (TBSV) of plants, used as a model of positive
stranded RNA virus for studying viral RNA replication and recombination. TBSV was observed to
co-opt some host factors that also affect prion disease propagation, such as protein chaperones and
ribosomal proteins.13b In fact, during our research activities some quinacrine-related analogues
elicited neuroprotective activity with higher efficacy than the parent drug in an in vitro model of
prion-related cell toxicity.25 Accumulating evidence revealed that quinacrine and analogues could
be effective tools to study similarities in replication strategies of various positive-sense single-
stranded RNA viruses.13b
In the last years, our group has reported different chemotypes endowed with antiviral activity
against BVDV RNA-dependent RNA polymerase (RdRp), as arylazoenamines,26,27 4-[(tert-
aminoalkyl)amino]arylazo compounds28 and particularly some 2-phenylbenzimidazole
derivatives.29,30 Subsequently, 9-amino-6-chloro-2-methoxyacridine derivatives, structurally related
to quinacrine and acranil, have also been assayed in vitro for cytotoxicity and antiviral activity
against BVDV and several other RNA and DNA viruses.31 In our findings, the prototypes
quinacrine and acranil failed to affect BVDV replication, while the most interesting analogues
exhibited EC50 values in the range 0.1-31 µM, comparing favorably with the reference compounds
ribavirin (EC50= 8.0 µM) and NM108 (2’-C-methylguanosine, EC50= 1.7 µM).31
2. Results and discussion
2.1. Chemistry
With the aim of further improving the activity of these promising anti-Pestivirus agents, in the
present work three prototypes - namely compounds AVR15, AVR17 and AVR26 - have been
selected to explore some chemical additional features on the 9-aminoacridine side chain, as shown
in Figure 1.
Together with the three lead compounds AVR15, AVR17, and AVR26, all 18 novel 9-
aminoacridine derivatives were synthesized and tested for anti-BVDV activity in cell-based assays.
Next, to identify which viral protein along the replicative cycle of BVDV virus was targeted by
these molecular series, a reasoned selection of derivatives endowed with a good selectivity index
(SI) profile was assayed for BVDV NS5B RNA-dependent RNA-polymerase (RdRp) inhibition.
Accordingly, a dose-dependent efficacy in the low micromolar range was verified, confirming
RdRp as the molecular target. Finally, isothermal titration calorimetry and computational studies
were performed to disclose the thermodynamics and structural signatures of the binding mode of
this class of antiviral compounds onto the viral RdRp.

Figure 1. Structures of the novel 9-aminoacridine-based derivatives investigated in the present
work.

AVR-15, AVR-17, AVR-2631 and 6-chloro-2-methoxy-9-[(pyridin-4’-yl)]aminoacridine (15)32
were synthesized according to the literature.
The novel compounds (1-14, 16-18) were yielded by reacting the 6,9-dichloro-2-methoxyacridine,
previously dissolved in phenol or DMF, with the proper amine in a molar ratio 1:1 or 1:2 for
primary or secondary amines, respectively (Scheme1).
Scheme 1. Reagents and conditions: (a) ratio 1:1 in phenol, 90°C, 5h; (b) ratio 1:2 in DMF, 140°C,
3h.
The intermediates c-e33-35 and h,33 i,34 k,35 required for the synthesis of the related 9-aminoacridine
derivatives, were prepared according to the cited references.
The new 1-aminopiperazines (f and g) were obtained by nitrosating the proper 1-substituted
piperazines with sodium nitrite and hydrochloric acid, and then reducing the corresponding N-
nitrosopiperazine (a, b) with Zn dust in acetic acid (Scheme 2). Starting from the 1-(3-
methoxyphenyl)piperazine the nitrosating reaction moved far from the expected product towards
decomposition; thus, 1-amino-4-(3-methoxyphenyl)piperazine (j) was obtained by direct amination
of 1-(3-methoxyphenyl)piperazine with hydroxylamine-O-sulfonic acid in basic medium (Scheme
3).

Scheme 2. Reagents and conditions: (a) NaNO2, 1N HCl (pH=5-6), 70°C, 1h; (b) Zinc dust,
AcOH/H2O ratio 1:1, 50°C, 3h.
Scheme 3. Reagents and conditions: (a) Hydroxylamine-O-sulfonic acid, NaOH pellets, H2O, 60°C,
15 min; then r.t., 24h.
2.2. Anti-BVDV activity
The 9-aminocridine derivatives considered in this work were evaluated for antiviral activity against
a panel of RNA and DNA viruses (Table 1S). Sixteen out of twenty-one compounds (76%) proved
to selectively inhibit BVDV replication (Table 1), since among the other virus strains only CVB-5
virus was occasionally susceptible to two compounds (11 and 12), which shared a significantly
lower efficacy (mean EC50 = 31 µM, Table 1). The most active compounds exhibited a potency
profile against BVDV ranging from 0.8 to 11.5 µM (Table 1), the 4-(2’-hydoxyethyl)piperazinyl (2,
EC50= 0.8 µM) and (6’-methoxy)pyridin-3-yl moieties (14, EC50= 1.2 µM) being the most efficient
substitutions on the 9-aminoacridine nucleus to impair BVDV replication. Interestingly, the most
promising compounds reached (1, 2, 6, 11 and 13) or surpassed (AVR15, 3, AVR17 and 14) the
potency and selectivity index of the licensed drug ribavirin (EC50= 8.0 µM, SI>12.5), a broad
spectrum antiviral agent that is not tailored to treat BVDV infections. Finally, 63% of the active
compounds (10/16) elicit additional interest for their general low toxicity (CC50 > 76 µM) against
three animal cell lines used as hosts to support the replication of viruses under investigation (see:
mean CC50 in Table 1). These results concur with a high safety profile of our compounds series.
Table 1. Antiviral activity against BVDV and cytotoxicity of the three prototypes (AVR) and of new 9-
aminocridine derivatives 1-18.a
BVDV
MDBK
SId
CC50/EC50
>33.3
>8.7
11.5
>13.3
-
1
BHK-21
VERO76
mean CC50
Cpd
EC50 (µM)b CC50 (µM)c
CC50 (µM)e CC50 (µM)f
for host cells
3.0 0.02
11.5 1.2
0.80 0.03
7.5 0.4
>100
>100
>100
9.2
>100
>100
68
32
10
>47.3
>100
16.5
>100
>100
43.7
AVR15
>100
19.7
>100
>100
30
>100
20.6
>100
>100
33
1
2
3
4
5
6
68 2.3
2.9 0.18
46
15.9
42.5
49.8
46.1

35 1.7
11.9 1.0
6.0 0.2
34 2.1
38 1.6
10.3 0.1
>100
8.0 0.4
1.2 0.5
63 2.4
31 0.2
20.2 2.0
59 0.9
75 1.7
54
43
>100
34
>100
100
>100
80
1.5
3.6
>16.7
1
>2.6
9.7
-
10
33.3
1.6
>3.2
>5.0
1.7
70
38
66
32
>100
42
63.3
37.7
>100
40.3
7
8
>100
45
>100
>100
>100
95
>100
>100
>100
>100
>100
>100
>100
>100
AVR17
9
96
>98.7
>100
>100
>91.7
>76
>100
>100
>100
>100
>88
10
11g
12g
13
>100
>100
>100
88
40
14
>100
>100
>100
>100
75
>100
>100
>100
>100
89
15
AVR26
16
17
1
18
>100
>100
>100
>100
2’-C-metilguanosine 1.70 0.08
Ribavirin 8.0 1.0
>100
>100
>58.8
>12.5
aNone of the compounds 1-18 proved active against any of the other investigated viruses (EC50> 100 µM or > CC50;
data not shown). bCompound concentration (µM) required to achieve 50% protection of MDBK cells from the BVDV-
c
induced cytopathogenicity, as determined by the MTT method. Compound concentration (µM) required to reduce the
viability of mock-infected MDBK cells by 50%, as determined by the MTT method. dSI= selectivity index. eCompound
concentration (µΜ) required to reduce the viability of mock-infected BHK (Hamster normal kidney fibroblast)
f
monolayers by 50%, as determined by the MTT method. Compound concentration (µM) required to reduce the
viability of mock-infected VERO 76 (Monkey normal kidney) monolayers by 50%. gActive against CVB-5: 11, EC50=
39 µM; 12, EC50= 23 µM (reference cpd. pleconaril: EC50= 0.005 µM). Values shown are the mean of two different
determinations, each performed in duplicate; SD <10%.
2.3. BVDV RdRp inhibition assays
Since the RNA-dependent RNA-polymerase plays a critical role in viral replication and, as such, it
constitutes one of the most promising drug targets for the development of new anti-BVDV agents,
we decided to investigate whether our newly synthesized 9-aminoacridine derivatives were able to
act as inhibitor of this viral enzyme. To the purpose, a test set of 6 compounds (i.e., AVR15, 2, 6,
AVR17, 14, and AVR26) was selected and assayed, observing the related EC50 values span three
order of magnitudes (i.e., EC50 = 0.80 µM for 2 and EC50 = 31 µM for AVR26, see Table 1). Two
further compounds, 4 and 12, were also tested as negative control. The aggregate results of the in
vitro assays are listed in Table 2, while all corresponding dose-response curves are shown in Figure
2.
Table 2. In vitro BVDV RdRp inhibition by selected new 9-aminoacridine derivatives. All experiments were
run in triplicate. Errors on IC50 are within 15%.

Cpd
AVR17
12
Cpd
AVR15
2
IC50 (µM)
IC50 (µM)
1.2
4.3
0.62
n.d.
n.d.
0.81
38
4
6
14
0.88
AVR26
Figure 2. Dose-response curves for the test set of 6 new 9-aminoacridine derivatives AVR15 (A), 2 (B),
6 (C), AVR17 (D), 14 (E), and AVR26 (F) as obtained from in vitro enzyme assay with the BVDV RNA-
dependent RNA-polymerase.
Four out of six compounds demonstrated potent inhibition of the BVDV RdRp, with biochemical
IC50 ranging from 0.62 µM (2) to 4.3 µM (AVR17) (Table 2). Moreover, immediate structure-
activity considerations suggest a critical role for the –NH group bridging the aminoacridine scaffold
and the piperazine/pyridine moieties for biological activity, whereas compound AVR26, in which
the piperazine ring is directly connected to the tricyclic structure, is a weaker RdRp inhibitor (see
next section for detailed discussion). Overall, these data support the hypothesis that this new series
of 9-aminoacridine derivatives target the BVDV replication key enzyme RNA-dependent RNA-
polymerase.
2.4. Binding thermodynamics of compounds to the BVDV RdRp by Isothermal Titration Calorimetry
(ITC) and molecular modeling
In order to investigate the binding affinities of all title compounds for the viral RNA-dependent
RNA-polymerase in detail, thermodynamic studies were performed using isothermal titration

calorimetry (ITC). The averaged dissociation constant (Kd), binding free energy (∆Gb), enthalpy
(∆Hb) and entropy terms (-T∆Sb) resulting from three independent ITC measurements on each
compound are listed in Table 3, while Figure 3 displays some representative ITC results and fitting
curves for the 6 compounds test set in Table 2 and Figure 2.
Table 3. ITC determined thermodynamic data of all synthesized 9-aminoacridine derivatives binding to the
BVDV RdRp. ∆Gb = ∆Hb –T∆Sb. ∆Gb = RT ln Kd. n = number of binding site. All experiments were run in
triplicate. Errors on ∆Hb are within 5%.
∆Gb (kcal/mol)
-8.24
∆Hb (kcal/mol)
-11.26
-T∆Sb (kcal/mol)
n (-)
0.92
1.10
Cpd
Kd (µM)
0.91
3.02
6.96
AVR15
9.1
0.57
4.8
38
-6.88
-13.84
1
2
3
4
5
6
7
8
-8.52
-7.26
-6.03
-5.99
-8.22
-6.33
-6.91
-8.20
-6.58
-6.38
-6.48
-5.45
-7.84
-8.38
-6.15
-6.21
-6.36
-6.02
-5.89
-14.07
-12.46
-10.43
-10.37
-13.52
-11.96
-13.15
-11.94
-10.21
-10.15
-10.09
-8.87
5.55
5.20
4.40
4.38
5.30
5.63
6.24
3.74
3.63
3.77
3.61
3.42
3.69
4.01
3.72
5.21
5.40
4.98
4.89
1.03
0.98
1.00
1.04
0.95
0.99
0.93
1.03
1.01
0.98
0.96
0.90
0.92
1.05
0.89
1.08
1.11
0.99
0.91
41
0.95
23
8.6
0.98
15
AVR17
9
21
10
18
11
44
12
1.8
0.72
31
-11.53
-12.39
-9.87
13
14
15
28
-11.42
-11.76
-11.00
-10.78
AVR26
16
22
39
17
48
18

Figure 3. Representative ITC binding isotherms for AVR15 (A), 2 (B), 6 (C), AVR17 (D), 14 (E), and
AVR26 (F) titrations into BVDV RdRp solutions. Inserts: ITC raw data.
For all the compounds, the interaction with the viral protein resulted in an exothermic ligand-
binding event, with Kd values in the range 0.57 - 48 µM. Generally, the RdRp-ligand binding is
driven by the action of stabilizing hydrogen bonds and cationic/π interactions (∆Hb < 0), while
entropy slightly disfavors complex formation (-T∆Sb > 0). The conformational entropy change is
commonly unfavorable in protein-ligand binding event, as the binding process involves the loss of
configurational degrees of freedom for both the drug molecule and the protein, resulting in an
ultimate negative change in conformational entropy. The stoichiometry of the binding process, n, is
always very close to 1, thereby confirming the 1:1 character of each binding event. Of note, the Kd
values for the 6 compounds shown in Figure 3 are in excellent agreement with the relevant IC50
values obtained from the in vitro RdRp inhibition assays (compare the corresponding values in
Tables 2 and 3); this further supports the notion that the viral polymerase is the target protein for
this series of compounds.
To understand whether the antiviral activity of the present derivatives is dependent on their RdRp
binding, we next examined the relationship between the Kd values derived by the ITC experiments
and the EC50 values determined in the cell-based assays. As shown in Figure 4, the resulting EC50
vs. Kd plot shows a highly positive correlation (R2 = 0.90), suggestive of a strong connection
between the binding of the new 9-aminoacridine derivatives to their target protein and their potency
against BVDV. Therefore, we can conclude that the antiviral activity of all new derivatives is
largely attributed to their capacity of inhibiting the viral RNA-dependent RNA polymerase.

Figure 4. Correlation between BVDV RdRp binding activity (Kd, Table 2) and antiviral activity (EC50, Table
1) of all new 9-aminoacridine derivatives (R2 = 0.90).
Molecular modeling of the title compounds in complex with the BVDV RdRp revealed the most
important features underlying protein/ligand binding. Docking followed by free energy of binding
scoring in the framework of the MM/PBSA methodologies36 resulted in calculated ∆Gb,comp values
in excellent agreement (R2 = 0.85) with the corresponding ITC-derived experimental data (∆Gb), as
illustrated in Figure 5A. Analogous results were found for the enthalpic component (∆Hb,comp and
∆Hb, Figure 5B, R2 = 0.89), while for the entropic term MM/PBSA achieved relatively satisfactory
performance (Figure 5C, R2 = 0.54). This is not surprising given the fact that the normal-mode
entropies employed in MM/PBSA37 omit important contributions to the true ligand-binding entropy.
In particular, it considers only the local stiffness of the actual binding conformation and does not
include any information whether the receptor or the ligand may have many different conformations
or whether the number of conformations changes upon binding. Also, it is well known that both
∆Hb,comp and –T∆Sb,comp are generally overestimated in MM/PBSA calculations;38 yet, 1) a parallel
shift is generally observed between experimental and computed ∆H and -T∆S so that, ultimately,
their difference (i.e., ∆Gb,comp) is indeed close to the relevant experimental value (∆Gb), and, most
importantly, 2) the trend exhibited by these two quantities with respect to their experimental
counterparts is fully and correctly predicted, as it is the present case. Accordingly, on the basis of all
the considerations developed above, the binding site and the relevant poses of the title compounds
in the retrieved binding pocket of the BVDV RdRp can be reliably assumed for further analysis. By
virtue of the agreement between simulated and experimental binding thermodynamics, the
following discussion will refer to the ITC-derived thermodynamics parameters (Table 3), which
constitute a true added value.

Figure 5. Correlation between computational and ITC-derived (Table 3) free energy of binding (A, R2 =
0.85), enthalpy (B, R2 = 0.89) and entropy (C, R2 = 0.54) for all new 9-aminoacridine derivatives in complex
with the BVDV RdRp.
Starting such analysis with the most potent compound 2, the inspection of the corresponding
equilibrated molecular dynamics (MD) trajectory reveals that this molecule is engaged in a
bifurcated hydrogen bond (HB) between the –NH group bridging the tricyclic scaffold and the
piperazine moiety, and the side chains of E265 (average dynamic length (ADL) = 2.03 0.04 Å)
and R285 (ADL = 1.93 0.02 Å), as shown in Figure 6A. Another permanent HB is detected
between the methoxy substituent on the acridine skeleton and the ammonium group of K263 (ADL
= 1.95 0.01 Å). Lastly, a third HB between the hydroxyethyl group of 2 and the positively
charged side chain of K525 (ADL = 2.10 0.04 Å) further stabilize the protein/ligand complex (see
Figure 6A). This ensemble of positive intermolecular interactions results into a very high affinity of
the compound for the protein (Kd = 0.57 µM), dominated by the corresponding enthalpic
component (∆Hb = -14.07 kca/mol). In binding to the viral polymerase, compound AVR15 is able
to preserve the first two fundamental anchoring HBs performed by 2 (ADL = 1.91 0.01 Å and
1.90 0.02 Å, respectively). This results again in a quite favorable, enthalpically-driven binding
affinity (Kd = 0.91 µM, ∆Hb = -11.26 kcal/mol), coupled with a lower entropy penalty (-T∆Sb =
3.02 kcal/mol) arising from the replacement of the hydroxyethyl substituent by a small, freely
rotating –CH3 group.

Figure 6. Equilibrated snapshots of compounds 2 (top left, 6A), 3 (top right, 6B), and 8 (bottom, 6C), in
complex with the BVDV RdRp. In all panels, the protein is portrayed as a transparent orange ribbon,
together with its van der Waals surface. Compounds are shown as atom-colored sticks-and-balls (C, gray; N,
blue; O, red; Cl, green). Hydrogen atoms, water molecules, ions and counterions are omitted for clarity.
The mini-series of analogous compounds 3, 4 and 5, differing from AVR15 by substitution of the
methyl group on the piperazine ring with a phenyl moiety and by the relative position of the Cl
atom on the latter, is quite interesting. Indeed, 4 and 5 establish the same, yet substantially less
optimized interactions described for AVR15 with the protein – resulting in considerably lower
protein affinity (Kd = 38 and 41 µM, ∆Hb = -10.43 and -10.37 kcal/mol for 4 and 5, respectively).
On the contrary, compound 3, besides the same two stabilizing HBs, is engaged in a favorable π-
cation interaction between the ortho-chlorine substituted phenyl ring and the positively charged side
chain of R529 (Figure 6B). This reflects in a overall lower Kd value (4.8 µM) with respect to the
two isomers 4 and 5, with an exothermic ∆Hb value of -12.46 kcal/mol leading to spontaneous and
stable protein binding. In summary, the position of the Cl atom onto the phenyl substituent of the
piperazine ring contributes in orienting the molecule within the binding pocket in that only when
the ortho position is occupied the π-cation interaction can be effectively in place.
The next small subset of analogues 6, 7 and 8 offers another, interesting perspective. Also in this
case, only 6 is able to perform the same (and slightly more efficient), interactions described for
compound 3 within the polymerase binding site, and this endows the compound with very good
binding thermodynamics (Kd = 0.95 µM, ∆Hb = -13.52 kcal/mol). Compound 8, featuring the –
OCH3 group at the para position of the phenyl ring, although not able to perform the π-cation
interaction described for 3, is involved in a further HB interaction between its OCH3 substituent and
the same residue R529 (ADL = 2.32 0.06 Å, Figure 6C). This results in an interesting protein
binding affinity (Kd = 8.6 µM and ∆Hb = -13.15 kcal/mol). Finally, compound 7 fails to exploit the

m-OCH3 group in engaging a stabilizing interaction, leading to a Kd value of 23 µM (∆Hb = -11.96
kcal/mol).
Compound AVR17 revealed a very good affinity for the BVDV RdRp (Kd = 0.98 µM, ∆Hb = -
11.94 kcal/mol), due to the persistence of the same HB scheme described for compound 2 in which,
however, the 3rd HB is replaced by same type of interaction between the pyridine nitrogen of
AVR17 and the side chain of R267 (ADL = 2.02 0.01 Å). Varying the position of –N atom on the
pyridine moiety to ortho (9) and para (15) positions, respectively, results in a progressive
decrement of protein binding (Kd = 15 and 31 µM, and ∆Hb = -10.21 and -9.87 kcal/mol for 9 and
15, respectively) due to the absence of the directional HB with R267. Placing a Cl atom onto the
pyridine ring of AVR17 is detrimental for compound 12, featuring the halogen atom in the ortho
position (Kb = 44 µM, ∆Hb = -8.87 kcal/mol) which leads to a sterical hindrance, while the p-Cl
substitution does not influence the binding significantly (compound 13, Kd = 1.8 µM, DHb = -11.53
kcal/mol). Shifting the relative positions of the pyridine –N and the Cl substituent does not result in
any improvement in protein affinity (11, Kd = 18 µM, ∆Hb = -10.09 kcal/mol; 10, Kd = 21 µM, ∆Hb
= -10.15 kcal/mol), thereby confirming the importance of the meta position of the nitrogen atom on
the aromatic ring. The combination of the favorable meta position of the pyridine –N and the
substitution with a para OCH3 group (compound 14) leads to a very good RdRp binder (Kd = 0.72
µM, ∆Ηβ = -12.39 kcal/mol). The reasons for this notable protein affinity resides in the fulfillment
of the same HB network described for AVR17 with the additional HB between the acceptor p-
OCH3 of the compound and the donor backbone –NH group between E265 and K266. (ADL = 2.43
0.05 Å).
Finally, the results obtained for the AVR26-based series of compounds testify the pivotal role of
the–NH group bridging the acridine scaffold and the piperazine/pyridine moieties for biological
activity. Indeed, all these compounds (AVR26, 16-18) are endowed with substantially lower RdRp
binding affinities (Kd = 28, 22, 39 and 48 µM), mainly ascribable by the absence of the fundamental
bifurcated HB described for the two previous derivative series. Thus, taking AVR-26 as a proof-of-
concept, this molecules is engaged in two sub-optimized HB, the former between the –OCH3 group
on the tricyclic system and K263, while the second intermolecular interaction is seen between one
of the piperazine nitrogen and the side chain of R285.
2.5 Binding site validation via site-directed mutagenesis and ITC
The computational results reported above allowed to define the principal intermolecular interactions
driving the complexation between the BVDV RdRp and the present classes of molecules. To
validate these indications, we next performed the in vitro mutagenesis study of the K263 protein

residue, which is mainly involved in ligand/protein recognition. According to the modeling results
(Figure 6), this amino acid exerts an exclusive role in the binding mode of all compounds;
therefore, K was mutated into alanine (A), causing a significant modification of the drug binding
configuration and of the related intermolecular contacts. Then, the in vitro enzymatic inhibition
assay and the experimental affinity of the most active compound 2 towards the K263A mutant
RdRp were concurrently evaluated. These results are shown in Figure 7.
Figure 7. (A) Dose-response curve and (B) binding isotherm for compound 2 as obtained from in vitro
enzyme assay and isothermal titration calorimetry (ITC) towards the BVDV K263A mutant RNA-dependent
RNA-polymerase.
Compared to the wild-type enzyme (IC50 = 1.2 µM, Table 2), compound 2 was considerably less
effective in inhibiting the K263A variant of the BVDV RdRp (Figure 7A), showing an IC50 value of
48 µM. Contextually, the affinity of the mutant protein isoform for the compound was also
dramatically reduced (Figure 7B), as the relevant Kd value increased from 0.57 µM for the native
protein to 25 µM for the K263A mutant one. In aggregate, and in addition to those reported in the
previous section, these data confirm the nature of the target enzyme and the validity of the proposed
binding site and mode of the present compounds on the BVDV RdRp.
3. Conclusions
This report describes the synthesis of a novel series of antiviral agents characterized by a 9-
aminoacridine scaffold, responsible for an RNA-dependent RNA polymerase inhibition mechanism.
These new compounds, structurally related to three previously studied prototypes (AVR15,
AVR17, and AVR26), confirmed their ability to selectively hit BVDV replication. Most
compounds exhibited EC50 values in the range 0.8-11.5 µM, with the corresponding selectivity
index comparing favorably with the reference drug ribavirin. Interestingly, most of the active
compounds provided high CC50 values against three animal host cells, thus pointing to a valuable
safety profile for these 9-aminoacridine-containing derivatives.

In order to assess the potential molecular target, the most promising compounds were tested against
the BVDV RdRp, confirming enzymatic inhibition with sub-micromolar or low micromolar
potencies. Binding affinities of the entire molecular series for the viral polymerase were directly
determined by ITC measurements, yielding Kd values in the range 0.57-48 µM. Furthermore,
computational studies disclosed the most probable binding mode of this class of antiviral
compounds on the RdRp. Notably, the antiviral activity against BVDV in cell-based assays (EC50),
the Kd values on the purified RdRp enzyme, and the in silico estimated binding affinity (∆G) shared
excellent comparable SAR trends. Finally, in vitro site-directed mutagenesis reaction and
subsequent ITC measurements allowed us to confirm the binding site of this class of compounds,
providing strong argumentation that these 9-aminoacridine derivatives exert their mechanism by
efficiently weakening the virus replication machinery through a direct inhibitory effect on the
BVDV RNA polymerase.
These novel compounds combine good antiviral activity with low cost and easily accessible
chemical synthesis, that are relevant to relieve the heavy economic burden produced by BVDV
infections of livestock and, more in general, by viruses belonging to Flaviviridae family. Therefore,
our results prompt to further explore the structure-activity relationships of this class of compounds
designing new improved anti-BVDV agents, capable of targeting the RNA-dependent RNA
polymerase.
4. Experimental section
4.1. General
Chemicals, solvents and reagents used for the syntheses were purchased from Sigma-Aldrich or
Alfa Aesar and were used without any further purification. Column chromatography (CC): neutral
1
alumina (Al2O3), activity 1 or silica gel 60 (SiO2) (Merck). Mps: Büchi apparatus, uncorrected. H
NMR and 13C NMR spectra were recorded on a Varian Gemini-200 instrument at 200 and 50 MHz,
respectively; DMSO-d6; δ in ppm rel. to Me4Si as internal standard. J in Hz. Elemental analyses
were performed on a Carlo Erba EA-1110 CHNS instrument in the Microanalysis Laboratory of the
Department of Pharmacy of Genoa University.
When required, the compounds were converted into the corresponding hydrochloride salts with 1N
HCl ethanolic solution, as an additional purification step to perform elemental analysis. Results of
elemental analyses indicated that the purity of all compounds was ≥95%.
4.2. General procedure for the synthesis of compounds 1-14

A mixture of 6,9-dichloro-2-methoxyacridine (1.8 mmol), the appropriate amine (1.8 mmol) and
phenol (1.13 g) was heated at 90 °C for 5 h. After cooling, the mixture was treated with 6 M NaOH
till strong alkalinity and extracted with Et2O. After removing the solvent, the residue was purified
by CC.
In the case of 12 the reaction mixture was rinsed up with water and washed thoroughly with H2O,
obtaining the final compound as hydrochloride salt.
Compounds 2-8 were obtained by washing the reaction mixture in the order, with a boiling solution
of 2N NaOH and then with water in order to remove the excess of phenol, affording a solid residue
that was chromatographed on silica gel, eluting with the indicated solvent.
4.2.1. 6-Chloro-9-[2-(4’-hydroxypiperidin-1-yl)ethyl]amino-2-methoxyacridine (1)
1
Yield 42%. Mp 208-210°C (CH2Cl2). CC(SiO2/CH2Cl2). H-NMR (200 MHz, DMSO-d6): 8.20-
7.96 (m, 2 arom. H), 7.44 (dd, J= 9.8 Hz, 2.2 Hz, 1 arom. H), 7.38-7.23 (m, 3 arom. H), 6.24 (br. s,
NH, exchanges with D2O), 3.99 (s, 3H, OCH3), 3.92-3.74 (m, 1H, HO-CH piperidine and 2H
HNCH2CH2-piperidine), 3.00-2.81 (m, 2H, piperidine), 2.68 (t, J= 7.0 Hz, 2H, HNCH2CH2-
piperidine), 2.44-2.22 (m, 2H, piperidine), 2.13-1.85 (m, 2H, piperidine and 1H, OH, exchanges
with D2O, superimposed), 1.83-1.63 (m, 2H, piperidine). 13C-NMR (50 MHz, DMSO-d6): 154.72,
150.15, 147.62, 133.03, 130.24, 126.56, 126.05, 123.81, 122.27, 116.83, 114.44, 99.85, 65.80,
57.14, 55.11, 50.36, 46.24, 34.04. Anal. calcd for C21H24ClN3O2: C 65.36, H 6.27, N 10.89; found:
C 65.60, H 6.18, N 10.78.
4.2.2. 6-Chloro-9-{[4-(2’-hydroxyethyl)piperazin-1-yl]amino}-2-methoxyacridine (2)
1
Yield 23.5%. CC(SiO2/CH2Cl2+10%MeOH). H-NMR (200 MHz, DMSO-d6): 8.31-7.24 (m, 6
arom. H), 4.44 (br s, NH, exchanges with D2O), 3.79 (s, OCH3), 3.51 (t, J = 6.0, 2H, C(2)), 3.38
(pseudo s, 4H, piperazine), 3.11-2.20 (m, 2H, C(1), 4H, piperazine and OH, exchanges with D2O,
superimposed). 13C-NMR (50 MHz, DMSO-d6): 153.71, 151.71, 141.11, 134.30, 131.72, 120.32,
116.55, 113.11, 106.51, 59.79, 58.27, 55.08, 54.89, 53.63, 52.04. Dihydrochloride salt: Mp 258-
260°C (dec.). Anal. calcd for C20H23ClN4O2·2HCl·2H2O: C 48.45, H 5.90, N 11.30; found: C
48.14, H 5.88, N 11.67.
4.2.3. 6-Chloro-9-{[4-(2’-chlorophenyl)piperazin-1-yl]amino}-2-methoxyacridine (3)
Yield 57%. Mp 173-175°C (Et2O/n-hexane). CC(SiO2/CH2Cl2+10%MeOH). 1H-NMR (200 MHz,
DMSO-d6): 8.18 (d. J = 9.0 Hz, 1 arom. H); 7.80-7.64 (m, 2 arom. H); 7.42-6.82 (m, 7 arom. H);
3.78 (s, 3H, OCH3); 3.37 (pseudo s, 4H, piperazine); 3.05 (pseudo s, 4H, piperazine); 2.80 (br s,

NH, exchanges with D2O). 13C-NMR (50 MHz, DMSO-d6): 151.90, 148.32, 134.20, 129.96,
127.66, 127.26, 123.58, 120.46, 115.13, 64.52, 55.05, 54.11, 49.76. Dihydrochloride salt: Mp 271-
272°C. Anal. calcd for C24H22Cl2N4O·2HCl: C 54.77, H 4.60, N 10.65; found: C 54.83, H 4.61, N
10.60.
4.2.4. 6-Chloro-9-{[4-(3’-chlorophenyl)piperazin-1-yl]amino}-2-methoxyacridine (4)
Yield 52.3%. CC(SiO2/CH2Cl2+10%MeOH). 1H-NMR (200 MHz, DMSO-d6): 8.16 (d, J = 8.8 Hz,
1 arom. H); 7.65 (s, 1 arom. H); 7.30-6.72 (m, 8 arom. H); 3.74 (s, 3H, OCH3); 3.40 (pseudo s, 4H,
piperazine); 3.07 (pseudo s, 4H, piperazine); 2.76 (br s, NH, exchanges with D2O). 13C-NMR (50
MHz, DMSO-d6): 151.58, 147.86, 134.28, 133.45, 130.04, 122.60, 117.75, 114.15, 113.21, 64.38,
54.98, 53.66, 46.56. Dihydrochloride salt: Mp 269-270°C. Anal. calcd for C24H22Cl2N4O·2HCl: C
54.77, H 4.60, N 10.65; found: C 54.66, H 4.61, N 10.75.
4.2.5. 6-Chloro-9-{[4-(4’-chlorophenyl)piperazin-1-yl]amino}-2-methoxyacridine (5)
1
Yield 45%. Mp 185.5-186.3°C (Et2O/hexane). CC(SiO2/CH2Cl2+10%MeOH). H-NMR (200 MHz,
DMSO-d6): 8.14 (d. J = 9.6 Hz, 1 arom. H); 7.64 (s, 1 arom. H); 7.30-6.72 (m, 8 arom. H); 3.74 (s,
3H, OCH3); 3.33 (pseudo s, 4H, piperazine); 3.11 (pseudo s, 4H, piperazine); 2.68 (br s, NH,
exchanges with D2O). 13C-NMR (50 MHz, DMSO-d6): 149.12, 147.74, 135.31, 133.24, 128.21,
122.02, 116.38, 64.50, 54.96, 53.63, 46.96. Anal. calcd for C24H22Cl2N4O: C 63.58, H 4.89, N
12.36; found: C 63.69, H 4.88, N 12.28.
4.2.6. 6-Chloro-9-{[4-(2’-methoxyphenyl)piperazin-1-yl]amino}-2-methoxyacridine (6)
1
Yield 65%. CC(SiO2/CH2Cl2+10%MeOH). H-NMR (200 MHz, DMSO-d6): 8.17 (d. J = 9.6 Hz, 1
arom. H); 7.70 (s, 1 arom. H); 7.20-6.73 (m, 8 arom. H); 3.77 (s, 6H, 2OCH3); 3.38 (pseudo s, 4H,
piperazine); 3.10 (pseudo s, 4H, piperazine); 2.88 (br s, NH, exchanges with D2O). 13C-NMR (50
MHz, DMSO-d6): 153.69, 151.59, 134.35, 131.73, 126.08, 120.47, 113.75, 111.63, 106.47, 64.52,
54.96, 53.92, 48.78. Dihydrochloride salt: Mp 225-226°C. Anal. calcd for C25H25ClN4O2·2HCl: C
57.54, H 5.22, N 10.74; found: C 57.34, H 5.52, N 11.08.
4.2.7. 6-Chloro-9-{[4-(3’-methoxyphenyl)piperazin-1-yl]amino}-2-methoxyacridine (7)
1
Yield 32.6%. Mp 98-101°C (Et2O). CC(SiO2/CH2Cl2+10%MeOH). H-NMR (200 MHz, DMSO-
d6): 8.17 (d. J = 9.2 Hz, 1 arom. H); 7.71 (s, 1 arom. H); 7.20-6.77 (m, 8 arom. H); 3.75 (s, 3H,
OCH3); 3.70 (s, 3H, OCH3); 3.45 (br s, NH, exchanges with D2O); 3.38 (pseudo s, 4H, piperazine);
3.12 (pseudo s, 4H, piperazine). 13C-NMR (50 MHz, DMSO-d6): 153.74, 151.76, 134.38, 131.74,

126.81, 120.12, 113.87, 111.14, 106.47, 64.48, 55.05, 53.65, 48.30. Anal. calcd for C25H25ClN4O2:
C 66.88, H 5.61, N 12.48; found: C 66.86, H 5.36, N 12.21.
4.2.8. 6-Chloro-9-{[4-(4’-methoxyphenyl)piperazin-1-yl]amino}-2-methoxyacridine (8)
1
Yield 63.5%. Mp 186.5-188.8°C (Et2O). CC(SiO2/CH2Cl2+10%MeOH). H-NMR (200 MHz,
DMSO-d6): 8.20 (d. J = 9.6 Hz, 1 arom. H); 7.71 (s, 1 arom. H); 7.30-6.91 (m, 4 arom. H); 6.52 (d,
J = 7.8 Hz, 2 arom. H); 6.39 (d, J = 8.0 Hz, 2 arom. H); 3.77 (s, 3H, OCH3); 3.74 (s, 3H, OCH3);
3.37 (pseudo s, 4H, piperazine); 3.05 (pseudo s, 4H, piperazine and NH, exchanges with D2O,
superimposed). 13C-NMR (50 MHz, DMSO-d6): 153.85, 151.71, 134.68, 129.25, 120.08, 113.91,
107.57, 103.92, 64.29, 54.99, 53.72, 47.07. Dihydrochloride salt: Mp 243-244°C. Anal. calcd for
C25H25ClN4O2·2HCl: C 57.54, H 5.22, N 10.74; found: C 57.33, H 5.38, N 10.91.
4.2.9. 6-Chloro-2-methoxy-9-[(pyridin-2-yl)amino]acridine (9)
1
Yield 53%. Mp 212-213°C. CC(Al2O3/CH2Cl2+2%DEA). H-NMR (200 MHz, DMSO-d6): 9.38
(br. s, NH, exchanges with D2O); 8.28-7.86 (m, 4 arom. H); 7.77-7.27 (m, 4 arom. H); 6.83 (d, J =
8.0, 2 arom. H); 3.80 (s, 3H, OCH3). 13C-NMR (50 MHz, DMSO-d6): 154.80, 147.90, 144.02,
134.90, 129.18, 125.23, 119.72, 115.02, 108.30, 101.05, 54.98. Anal. calcd for C19H14ClN3O: C
67.96, H 4.20, N 12.51; found: C 68.21, H 4.56, N 12.49.
4.2.10. 6-Chloro-9-{[(4’-chloro)pyridin-2-yl]amino}-2-methoxyacridine (10)
1
Yield 80%. Mp 257-259°C (Et2O). H-NMR (200 MHz, DMSO-d6): 8.20-7.76 (m, 4 arom. H);
7.64-7.20 (m, 3 arom. H); 7.00-6.73 (m, 2 arom. H); 3.79 (s, 3H, OCH3); 3.38 (br. s, NH, exchanges
with D2O). 13C-NMR (50 MHz, DMSO-d6): 156.02, 148.88, 143.22, 133.65, 128.97, 126.56,
125.10, 121.80, 114.95, 109.11, 100.03, 55.08. Anal. calcd for C19H13Cl2N3O: C 61.64, H 3.54, N
11.35; found: C 61.45, H 3.65, N 10.99.
4.2.11. 6-Chloro-9-{[(5’-chloro)pyridin-2-yl]amino}-2-methoxyacridine (11)
Yield 40%. Mp 203-204°C (CH2Cl2). CC(SiO2/CH2Cl2). 1H-NMR (200MHz, DMSO-d6): 8.10-7.78
(m, 4 arom. H), 7.60-7.23 (m, 3 arom. H), 6.98-6.69 (m, 2 arom. H), 3.83 (s, 3H, OCH3), 3.35 (NH,
exchanges with D2O). 13C-NMR (50 MHz, DMSO-d6): 154.70, 146.04, 144.98, 133.89, 129.17,
126.76, 125.10, 119.94, 114.43, 111.24, 106.56, 99.85, 55.10. Anal. calcd for C19H13Cl2N3O: C
61.64, H 3.54, N 11.35; found: C 61.72, H 3.84, N 10.99.
4.2.12. 6-Chloro-9-{[(2’-chloro)pyridin-3-yl]amino}-2-methoxyacridine hydrochloride (12)

1
Yield 38%. Mp 286-290°C (Et2O). H-NMR (200 MHz, DMSO-d6): 8.36-8.24 (m, 1 arom. H);
7.96-7.34 (m, 4 arom. H); 7.30-7.20 (m, 1 arom. H); 3.66 (s, 3H, OCH3 and 1H, NH, exchanges
with D2O). 13C-NMR (50 MHz, DMSO-d6): 156.15, 148.76, 141.21, 137.41, 133.98, 128.84,
125.56, 123.67, 121.17, 107.93, 99.82, 54.45. Anal. calcd for C19H13Cl2N3O·HCl·H2O: C 53.73, H
3.80, N 9.89; found: C 53.86, H 4.01, N 9.49.
4.2.13. 6-Chloro-9-{[(6’-chloro)pyridin-3-yl]amino}-2-methoxyacridine (13)
Yield 34.94%. Mp 158-160°C (Et2O). CC(SiO2/Et2O). 1H-NMR (200 MHz, DMSO-d6): 8.23-6.82
(m, 8 arom. H); 6.76 (d, J = 8.0, 1 arom. H); 3.70 (s, 1H, NH, exchanges with D2O); 3.36 (s, 3H,
OCH3). 13C-NMR (50 MHz, DMSO-d6): 156.26, 151.59, 141.53, 135.59, 133.45, 129.19, 127.53,
124.22, 121.73, 115.43, 99.65, 55.11. Anal. calcd for C19H13Cl2N3O: C 61.64, H 3.54, N 11.35;
found: C 61.66, H 3.82, N 11.44.
4.2.14. 6-Chloro-9-{[(6’-methoxy)pyridin-3-yl]amino}-2-methoxyacridine (14)
1
Yield 15%. Mp 195-197°C (Et2O). CC(CH2Cl2/DEA2%). H-NMR (200MHz, DMSO-d6): 8.28-
6.58 (m, 9 arom. H). 4.15-3.39 (m, 7H, 2OCH3 and 1H, NH, exchanges with D2O). 13C-NMR (50
MHz, DMSO-d6): 158.47, 154.31, 135.55, 134.22, 130.09, 127.21, 123.15, 118.92, 110.23, 54.72,
52.68. Anal. calcd for C20H16ClN3O2: C 65.67, H 4.41, N 11.49; found: C 65.61, H 4.50, N 10.99.
4.3. General procedure for the synthesis of compounds 16-18
A solution of 6,9-dichloro-2-methoxyacridine (3.5 mmol) and cyclic secondary amine (7.0 mmol)
in DMF (5 mL) was heated at 140 °C for 3 h. The mixture was taken up with water, alkalinized
with 2 M NaOH and extracted with CH2Cl2. The solvent was evaporated and the oily residue was
purified by CC. The isolated compounds were crystallized from the indicated solvents.
4.3.1. 6-Chloro-2-methoxy-9-[4-(2’-methoxyphenyl)piperazin-1-yl]acridine (16)
1
Yield 27%. Mp 166-168°C (CH2Cl2). CC(SiO2/CH2Cl2). H-NMR (200 MHz, DMSO-d6): 8.43 (d,
J = 9.8 Hz, 1 arom. H); 8.20-7.93 (m, 2 arom. H); 7.70-7.42 (m, 3 arom. H); 7.16-6.83 (m, 4 arom.
H); 3.96 (s, 3H, OCH3); 3.84 (s, 3H, OCH3); 3.70 (pseudo s, 4H, piperazine); 3.29 (pseudo s, 4H,
piperazine). 13C-NMR (50 MHz, DMSO-d6): 157.99, 151.80, 146.33, 138.94, 133.95, 122.13,
120.27, 118.67, 116.60, 113.75, 64.62, 54.23, 49.01, 48.84. Anal. calcd for C25H24ClN3O2: C
69.20, H 5.57, N 9.68; found: C 69.21, H 5.49, N 9.31.
4.3.2. 6-Chloro-2-methoxy-9-[4-(3’-methoxyphenyl)piperazin-1-yl]acridine (17)

1
Yield 24%. Mp 163-165°C (Et2O). CC(SiO2/Et2O). H-NMR (200 MHz, DMSO-d6): 8.38 (d. J =
9.8 Hz, 1 arom. H); 8.16-7.95 (m, 2 arom. H); 7.57-7.48 (m, 4 arom. H); 7.19 (t, J = 7.6 Hz, 1 arom.
H); 6.74-6.57 (m, 2 arom. H); 6.50-6.42 (m, 1 arom. H); 3.96 (s, 3H, OCH3); 3.76 (s, 3H, OCH3);
3.48 (pseudo s, 4H, piperazine); 3.35 (pseudo s, 4H, piperazine). 13C-NMR (50 MHz, DMSO-d6):
158.82, 152.53, 145.60, 138.94, 133.99, 120.11, 118.65, 116.82, 113.32, 64.47, 54.78, 54.04, 48.56.
Anal. calcd for C25H24ClN3O2: C 69.20, H 5.57, N 9.68; found: C 69.21, H 5.49, N 9.31.
4.3.3. 6-Chloro-2-methoxy-9-[4-(4’-metoxyphenyl)piperazin-1-yl)acridine (18)
1
Yield 45%. Mp 172.5-174°C (Et2O). CC(SiO2/Et2O). H-NMR (200MHz, DMSO-d6): 8.42 (d. J =
10.0 Hz, 1 arom. H); 8.16-8.00 (m, 2 arom. H); 7.60-7.46 (m, 3 arom. H); 7.05 (d, J = 8.8 Hz, 2
arom. H); 6.89 (d, J = 8.8 Hz, 2 arom. H); 3.96 (s, 3H, OCH3); 3.73 (s, 3H, OCH3); 3.56 (pseudo s,
4H, piperazine); 3.35 (pseudo s, 4H, piperazine). 13C-NMR (50 MHz, DMSO-d6): 159.35, 151.97,
146.05, 139.08, 133.85, 119.63, 115.66, 113.24, 64.52, 54.54, 53.84, 48.49. Anal. calcd for
C25H24ClN3O2: C 69.20, H 5.57, N 9.68; found: C 69.39, H 5.72, N 9.36.
4.4. General procedure for the synthesis of compounds a, b
To the proper 1-phenylpiperazine hydrochloride (a, b) (3.77 mmol) in H2O (8 mL), 1N HCl was
added dropwise until the pH was 5-6. The solution was cooled to 0°C and a solution of NaNO2
(3.77 mmol) in water (5 mL) was added over a period of 30 min, maintaining a pH of 5-6 by
dropwise addition of 4N HCl. Then the reaction mixture was stirred at 70°C for 1h.
On cooling to room temperature, the solution was neutralized with 2 M NaOH. The precipitate (a)
was collected by filtration and crystallized from dry Et2O.
In the case of compound b the neutral aqueous solution was extracted with CHCl3. After drying
(Na2SO4) the solvent was evaporated obtaining the title compound.
4.4.1. 4-(2-Chlorophenyl)-1-nitroso-piperazine (a)
Yield 63%. Mp 72-73°C (Et2O). 1H-NMR (200MHz, DMSO-d6): δ 7.46 (d, J = 7.8 Hz, 1 arom. H),
7.37-7.00 (m, 3 arom. H), 4.50 (pseudo s, 2H, piperazine), 4.08 (pseudo s, 2H, piperazine), 3.34
(pseudo s, 2H, piperazine), 3.09 (pseudo s, 2H, piperazine). Anal. calcd for C10H12ClN3O: C 53.22,
H 5.36, N 18.62; found: C 53.13, H 5.10, N 18.67.
4.4.2. 4-(3-Chlorophenyl)-1-nitroso-piperazine (b)
1
Yield 83.5%. Oil. H-NMR (200MHz, DMSO-d6): δ 7.35-7.12 (m, 1 arom. H), 7.00-6.75 (m, 3
arom. H), 4.44 (pseudo s, 2H, piperazine), 3.99 (pseudo s, 2H, piperazine), 3.46 (pseudo s, 2H,

piperazine), 3.23 (pseudo s, 2H, piperazine). Anal. calcd for C10H12ClN3O: C 53.22, H 5.36, N
18.62; found: C 53.10, H 5.18, N 18.71.
4.5. General procedure for the synthesis of compounds f, g
Zinc dust (5 mmol) was added over a period of 20 min to the proper nitroso derivative (1.85 mmol)
previously suspended in 5 mL of 50% (v/v) aqueous AcOH. The mixture was heated at 50°C for 3h
and then filtered. The solution was cooled and basified with 6N NaOH and extracted with CHCl3.
The organic layer was washed with H2O, dried (Na2SO4) and then evaporated. The crude residue
was purified by CC (SiO2/CHCl3+10%MeOH).
4.5.1. 1-Amino-4-(2-chlorophenyl)piperazine hydrochloride (f)
1
+
Yield 58.8%. Mp 197-198°C (EtOH/Et2O). H-NMR (200MHz, DMSO-d6): δ 9.64 (br s, NH3,
exchange with D2O), 7.44 (d, J = 8.0 Hz, 1 arom. H), 7.39-7.00 (m, 3 arom. H), 3.13 (pseudo s, 8H,
piperazine). Anal. calcd for C10H14ClN3·HCl: C 48.40, H 6.09, N 16.93; found: C 48.07, H 5.89, N
16.90.
4.5.2. 1-Amino-4-(3-chlorophenyl)piperazine hydrochloride (g)
1
+
Yield 27.5%. Mp 229-230°C (EtOH/Et2O). H-NMR (200MHz, DMSO-d6): δ 9.56 (br s, NH3,
exchange with D2O), 7.23 (t, J = 7.4 Hz, 1 arom. H), 7.12-6.79 (m, 3 arom. H), 3.74 (s, 3H, OCH3),
3.41 (pseudo s, 4H, piperazine), 3.15 (pseudo s, 4H, piperazine). Anal calcd for C10H14ClN3·HCl: C
48.40, H 6.09, N 16.93; found: C 48.20, H 6.01, N 17.02.
4.6. Synthesis of 1-amino-4-(3-Methoxyphenyl)piperazine hydrochloride (j)
A mixture of 1-(3-methoxyphenyl)piperazine (3.3 mmol) in 25 mL of H2O and NaOH pellets (1.32
g) was heated at 60° for 20 min. Maintaining the temperature at 60°C, hydroxylamine-O-sulphonic
acid (5 mmol) was slowly added over 15 min. Then the mixture was stirred at 25°C for 24h. The
solution was basified with 6N NaOH, and extracted with CHCl3. After evaporation of the solvent,
the solid residue was purified with CC (SiO2/CHCl3+10%MeOH).
1
+
Yield 40%. Mp 175-177°C (EtOH/Et2O). H-NMR (200MHz, DMSO-d6): δ 9.20 (br s, NH3,
exchange with D2O), 7.16 (t, J = 7.0 Hz, 1 arom. H), 6.71-6.54 (m, 2 arom. H), 6.47 (d, J = 8.0 Hz,
1 arom. H), 3.74 (s, 3H, OCH3), 3.31 (pseudo s, 4H, piperazine), 3.14 (pseudo s, 4H, piperazine).
Anal. calcd for C11H17N3O·HCl: C 54.21, H 7.44, N 17.24; found: C 54.30, H 7.18, N 17.19.
4.7. Biological tests

4.7.1. Compounds
The test and control compounds were dissolved in DMSO at 100 mM and then diluted in culture
medium. The final DMSO concentration did not exceed 1% (commonly 0.1%) which did not affect
the biological assay results.
As reference inhibitors were used: Pleconaril for CVB-5 and Sb-1 only; NM108 (2’-C-
methylguanosine), NM176 (2’-C-ethynylcytidine) and Ribavirin for the other ssRNA+ and dsRNA
viruses; NM299 (6-azauridine), Ribavirin and M5255 (mycophenolic acid) for the ssRNA- viruses;
M5255 and ACG (acyclovir) for DNA viruses.
ACG, M5255, NM108, NM299, Pleconaril and Ribavirin were purchased from Sigma-Aldrich;
NM176 was purchased from Idenix Pharmaceuticals.
4.7.2. Cells and viruses
Cell lines were purchased from American Type Culture Collection (ATCC). The absence of
mycoplasma contamination was checked periodically by the Hoechst staining method. Cell lines
supporting the multiplication of RNA and DNA viruses were the following: Madin-Darby Bovine
Kidney (MDBK) [ATCC CCL 22 (NBL-1) Bos Taurus]; Baby Hamster Kidney (BHK-21) [ATCC
CCL 10 (C-13) Mesocricetus auratus] and Monkey kidney (Vero-76) [ATCC CRL 1587
Cercopithecus Aethiops]. Viruses were purchased from American Type Culture Collection (ATCC)
except Yellow Fever Virus (YFV). Viruses representative of positive-sense single-strand RNA
(ssRNA+) group used were: i) Flaviviridae family: YFV [strain 17-D vaccine (Stamaril Pasteur
J07B01)], Bovine Viral Diarrhea Virus (BVDV) [strain NADL (ATCC VR-534)]; ii)
Picornaviridae family: Human Coxsackievirus type B5 (CVB-5) strain (ATCC® VR-
1036AS/HO™) and Human Poliovirus type-1 Sabin (Sb-1) [strain Chat (ATCC VR-1562)]. Virus
representative of a negative-sense single-strand RNA (ssRNA-) group used were: Vesicular
Stomatitis Virus (VSV) [strain Indiana Lab (ATCC VR-158)] and Human Respiratory Syncytial
Virus (RSV) [strain A2 (ATCC VR-1540)]. A virus representative of a double strand RNA
(dsRNA) group used was: Reovirus type-1 [strain 3651 (SV-12, simian virus 12) (ATCC VR-214)].
Viruses representatives of DNA group used were: i) Poxviridae family: Vaccinia Virus (VV) [strain
Elstree-Lister Vaccine (ATCC VR-1549)]; ii) Herpesviridae family: Human Herpesvirus 1 (HSV-1)
[strain KOS (ATCC VR-1493)].
4.7.3. Antiviral assays
Compounds activity against YFV and Reo-1 was based on inhibition of virus-induced
cytopathogenicity in BHK-21 cells acutely infected with a m.o.i. of 0.01. Compounds activity

against BVDV was based on inhibition of virus-induced cytopathogenicity in MDBK cells acutely
infected with a m.o.i. of 0.01. Briefly, BHK and MDBK cells were seeded in 96-well plates at a
density of 5x104 and 3x104 cells/well, respectively, and were allowed to form confluent monolayers
by incubating overnight in growth medium at 37 °C in a humidified CO2 (5%) atmosphere. Cell
monolayers were then infected with 50 µL of a proper virus dilution in maintenance medium
(MEM-E with L-glutamine, supplemented with 0.5% inactivated FBS, 1 mM sodium pyruvate and
25 µg/mL kanamycin) to give a m.o.i of 0.01. After 2 hrs, 50 mL of maintenance medium, without
or with serial dilutions (from 100 µM, ratio 1:5) of test compounds, were added. After a 3-4 days
incubation at 37 °C, cell viability was determined by the MTT method.
Compounds activity against CVB-5, Sb-1, VSV, VV, HSV-1 and RSV was determined by plaque
reduction assays in infected Vero-76 cell monolayers. To this end, Vero-76 cells were seeded in 24-
well plates at a density of 2x105 cells/well and were allowed to form confluent monolayers by
incubating overnight in growth medium (Dulbecco's Modified Eagle Medium (D-MEM) with L-
glutamine and 4,5 µg/mL D-glucose, supplemented with 10% FBS and 25 µg/mL kanamycin) at 37
°C in a humidified CO2 (5%) atmosphere. Then, monolayers were infected for 2 hrs with 250 µL of
proper virus dilutions to give 50-100 PFU/well. Following removal of unadsorbed virus, 500 µL of
maintenance medium (D-MEM medium with L-glutamine and 4,5 µg/mL D-glucose supplemented
with 1% inactivated FBS and 0.75% methyl-cellulose), without or with serial dilutions (from 100
µSV), 3 (CVB-5, VV and HSV-1) or 5 days (RSV) and then fixed with PBS containing 50%
ethanol and 0.8% crystal violet, washed and air-dried. Plaques were then counted and EC50 (50%
effective concentration) was calculated by linear regression technique. The cytotoxicity of test
compounds was determined in parallel on the same 24-well plate used for the EC50 determination.
Linear regression analysis: viral and cell growth at each drug concentration was expressed as
percentage of untreated controls and concentrations resulting in 50% (EC50 and CC50) growth
inhibition were determined by linear regression analysis.
4.7.4. Cytotoxicity assays
Cytotoxicity assays were run in parallel with antiviral assays. MDBK and BHK cells were seeded at
an initial density of 6x105 and 1x106 cells/mL in 96-well plates, respectively, in culture medium
(Minimum Essential Medium with Earle’s salts (MEM-E) with L-glutamine, supplemented with
10% horse serum and 1 mM sodium pyruvate (for MDBK cells) or with 10% fetal bovine serum
(FBS) (for BHK cells), 25 µg/mL kanamycin). Cell cultures were then incubated at 37 °C in a
humidified, 5% CO2 atmosphere in the absence or presence of serial dilutions (from 100 µM, ratio
1:5) of test compounds. Cell viability was determined after 48-96 hrs at 37 °C by the MTT

method55. Vero-76 cells were seeded at an initial density of 4x105 cells/mL in 24-well plates, in
culture medium (Dulbecco’s Modified Eagle Medium (D-MEM) with L-glutamine, supplemented
with fetal bovine serum (FBS), 25 µg/mL kanamycin). Cell cultures were then incubated at 37 °C in
a humidified, 5% CO2 atmosphere in the absence or presence of serial dilutions of test compounds.
Cell viability was determined after 48-96 hrs at 37 °C by the Crystal violet staining method.
4.7.5. Expression and purification of the BVDV-NS5B∆24 polymerase
The BVDV-NS5B∆24 polymerase was expressed and purified following the protocols reported in
details in our previous work.39,40 Briefly, the expression plasmid encoding the His-tagged C-
terminal24-aminoacid-deleted BVDV-NS5B was introduced into the Escherichia coli strain
Rosetta™2(DE3)pLysS (Novagene) by chemical transformation. Transformant bacteria were
cultured overnight at 30 °C in 5 mL of lysogeny broth (LB) supplemented with 25 mg/mL
kanamycin and 30 mg/mL chloramphenicol. Cultures were then diluted into 1 L of LB medium
containing the same quantities of the two antibiotics, and incubated at 30 °C until an A600 of 0.6–
0.7 was reached. The culture was then induced overnight with 1 mM isopropyl-β-D-
thiogalactopyranoside, after which cells were harvested by centrifugation and stored at 80 °C until
purification.
For protein purification, cell pellets were thawed and quickly lysed by the addition of 10 mL of
CelLytic B (Sigma). Insoluble material was removed by centrifugation at 11,000 rpm for 60 min at
4 °C. The soluble extract was applied to a 5-mL column of nickel-nitrilotriacetic acid-agarose
(Qiagen) that had been equilibrated with the lysis buffer (50 mM NaH2PO4, 300 mM NaCl, 10 mM
imidazole, pH 8.0). After extensive column washing with wash-buffer (50 mM NaH2PO4, 300 mM
NaCl, 20 mM imidazole, pH 8.0), the protein was eluted stepwise with the elution buffer containing
increasing concentration of imidazole (50 mM NaH2PO4, 300 mM NaCl, 50-250 mM imidazole, pH
8.0). The polypeptide composition of each column fraction was monitored via Coomassie-stained
SDS-PAGE analysis. Fractions enriched in pure 6xHis-tagged NS5B protein, recovered in the 130-
250 mM imidazole eluates, were pooled and dialyzed against a buffer containing 25 mM Tris-HCl,
pH 7.5, 2.5 mM MgCl2, 1 mM dithiothreitol, and 50%glycerol. Protein concentration was
determined by the micro-Bradford method (Bio-Rad) using Bovine Serum Albumin (BSA) as
standard. Following dialysis, the purified 6xHis-tagged BVDVNS5B∆24 protein was divided into
aliquots and stored at -80 °C.
4.7.6. RNA-dependent RNA polymerase inhibition assay

Enzyme assays were performed in 96-well plates using 10 µg/mL poly(rC) (GE Healthcare) as
template, 0.1 µg/mL oligo(rG)12 (Invitrogen) as primer, and 80 µM GTP (Invitrogen) as substrate,
in a 20 µL reaction mixture containing 20 mM Tris/HCl, pH 7.0, 1 mM dithiothreitol, 25 mM NaCl,
20 U/mL RNasin (Promega), 5 mM MgCl2, 5% DMSO, 5%glycerol, and 500 ng of the purified
protein.39,40 After enzyme/drug pre-incubation for 30 min at room temperature, reactions were
started by the addition of GTP. One microliter of three-fold serial dilutions of the test compounds in
0.5% DMSO were added, and the samples were incubated for 120 min at 37 °C. DMSO alone and
the nucleotide analog 30-deoxyguanosine-50-triphosphate (30-dGTP) (tebu-bio) were used as
negative and positive controls, respectively. Reactions were stopped by adding 2 µL of 200 mM
EDTA. 138 µL of the PicoGreen Quantitation Reagent (Molecular Probes) diluted 1:345 in TE
(Tris/EDTA) buffer were added to each sample, followed by incubation for 5 min at room
temperature in the dark. After excitation at 480 nm, fluorescence was measured at 520 nm in a
fluorescence microplate reader (Infinite F200, Tecan). Effective fluorescence was calculated by
subtracting the mean fluorescence of the blank samples and by converting it into % of activity.
Percent of residual activity was then plotted vs. compound concentrations. Dose–response curves
were fit with GraphPad Prism (v. 6.0 for Mac OS X, GraphPad Software, La Jolla California USA)
to obtain the drug concentration providing 50% inhibition (IC50).
4.7.7. Site-directed mutagenesis.
The plasmid encoding the His-tagged C-terminal 24-aminoacid-deleted BVDV-NS5B was used
as the parental clone for all subsequent manipulations. Site-directed mutagenesis was carried out by
using a QuickChange mutagenesis kit (Stratagene). The expression of the BVDV NS5B mutants
was carried out in the Escherichia coli strain Rosetta™2(DE3)pLysS (Novagene) as described
above.
4.7.8. Determination of the compound/RdRp binding thermodynamics by Isothermal Titration
Calorimetry (ITC)
Isothermal titration calorimetry (ITC) experiments were performed with a MicroCal PEAQ-ITC
calorimeter (Malvern, UK) at 37 °C (cell volume = 280 µL). Thermodynamics of all
BDVD/compound complex formation was investigated in 20 mM Tris/HCl, 25 mM NaCl (pH 7.0
buffered solutions. Specifically, a solution of BVDV RdRp (20 µM, sample cell) was titrated with
19 step-by-step injections of 2 µL volume of each compound (200 µM, syringe). Solutions and
buffer were degassed for 30 min at room temperature under stirring at 750 rpm prior to each
experiment. All experiments were run in triplicate. Data acquisition and analysis were performed
using inbuilt Origin 7.0 to determine the binding stoichiometry (n) and thermodynamic parameters

(∆Gb, ∆Hb, T∆Sb, and Kd) of the binding reaction. Graphs were prepared with GraphPad Prism (v.
6.0 for Mac OS X, GraphPad Software, La Jolla California USA).
4.9. Computational methods
All simulations were carried out using the Pmemd modules of Amber 1641 running on our own
CPU/GPU calculation cluster. Molecular graphics images were produced using the UCSF Chimera
package (v.1.10).42 Chimera is developed by the Resource for Biocomputing, Visualization, and
Informatics at the University of California, San Francisco (supported by NIGMS P41-GM103311).
All other graphs were obtained using GraphPad Prism (v. 6.0 v. 6 for Mac OsX, GraphPad
Software, La Jolla California USA).
The structures of all compounds were build and geometrically optimized using the GAFF force
field;43 next, the molecules were docked into the BVDV RdRp binding pocket using the optimized
structure of the RdRp of BVDV taken from our previous work (PDB entry code 1S48).27,39,44 All
docking experiments were performed with Autodock 4.2.6/Autodock Tools 1.4.645 on a win64
platform. The resulting docked conformations were clustered and visualized; then, the structure of
each resulting complex characterized by the lowest Autodock interaction energy in the prevailing
cluster was selected for further modeling. Each ligand/polymerase complex was further refined in
Amber 16 using the quenched molecular dynamics (QMD) method as previously described (see, for
instance,39,40,46,47 and reference therein). Next, the best energy configuration of each complex
resulting from QMD was subsequently solvated by a cubic box of TIP3P water molecules48
extending at least 10 Å in each direction from the solute. The system was neutralized and the
solution ionic strength was adjusted to 20 mM NaCl by adding the proper amounts of Na+ and Cl-
ions. Each solvated system was relaxed (500 steps of steepest descent followed by 500 other
conjugate-gradient minimization steps) and then gradually heated to the target temperature of 37 °C
in intervals of 50 ps of constant volume-constant temperature (NVT) molecular dynamics (MD)
simulations (Verlet integration method, time step = 1.0 fs). The Langevin thermostat was used to
control temperature. During this phase of MD, the protein was restrained with a force constant of
2.0 kcal/(mol Å), and all simulations were carried out with periodic boundary conditions.
Subsequently, the density of the system was equilibrated via MD runs in the isothermal-isobaric
(NPT) ensemble, with a time step of 1 fs. All restraints on the protein atoms were then removed,
and each system was further equilibrated using NPT MD runs at 37°C. Three equilibration steps
were performed (4 ns each, time step = 2.0 fs). System stability was monitored by the fluctuations
of the root-mean-square-deviation (rmsd) of the simulated position of the backbone atoms of the
viral polymerase with respect to those of the initial protein model (data not shown). The

equilibration phase was followed by a data production run consisting of 50 ns of MD simulations in
the NVT ensemble. Data collection was performed on over the last 20 ns of each equilibrated MD
trajectory were considered for statistical data collections. 1000 trajectory snapshots were analyzed
for each compound/RdRp complex. In all above simulations, the GAFF force field was used for the
small molecule compounds while the ff14SB force field49 was adopted for the protein.
The computational free energy of binding ∆Gb,comp between the compounds and the BVDV RdRp,
its major components (∆Hb,comp and –T∆Sb,comp), and the relevant Kd,comp values were estimated by
resorting to the well-validated Molecular Mechanics/Poisson-Boltzmann Surface Area (MM/PBSA)
approach36 implemented in Amber 16.
Acknowledgments
M.T. acknowledges the University of Genoa for financial support (PRA2014, 100006-2014-TM-
PRA_001 to M.T.). S.P and E.L. gratefully acknowledge the Associazione Italiana per la Ricerca
sul Cancro (AIRC) for the generous financial support (grant IG 2015 Id.17413 to S.P.). MT thanks
Dr. M. Anzaldi for performing NMR spectra and Mr. O. Gagliardo for performing elemental
analysis.
Author contributions: M.T. and S.P. conceived the study; M.T. and V.F. designed and synthesized
all compounds, providing their structural characterization; R.L. performed cell-based assays; E.L.
and S.P. designed biochemical and computational research and performed biochemical assays; E.L.,
S.B. and M.F. performed computational studies; M.T., E.L., S.P., S.B. and M.F. analyzed data;
M.T. and S.P. wrote the paper.
Supplementary data
Cytotoxicity and antiviral activities of the compounds series (1-18) and reference compounds
against the investigated panel of RNA and DNA viruses are depicted in Tables 1S.
Conflict of Interest
All authors declare no conflicts of interest.
References
1. O’Connor AM, Sorden SD, Apley MD. Association between the existence of calves persistently
infected with bovine viral diarrhea virus and commingling on pen morbidity in feedlot cattle.
Am J Vet Res. 2005;66:2130-2134.