Design, synthesis, and preliminary in vitro and in silico antiviral activity of [4,7]phenantrolines and 1-oxo-1,4-dihydro-[4,7]phenantrolines against single-stranded positive-sense RNA genome viruses

Article abstract of DOI:10.1016/j.bmc.2007.01.005

Following the antiviral screening of a wide series of new angular and linear N-tricyclic systems both in silico and in vitro, the [4,7]phenantroline nucleus emerged as a new ring system endowed with activity against viruses containing single-stranded, positive-sense RNA genomes (ssRNA⁺). Here, we report our new pathway to the synthesis of this nucleus and of several related derivatives, as well as the results of both cell-based antiviral assays and molecular dynamics simulations. In the antiviral screening, several compounds (9 and 16-20) showed to be fairly active against BVDV, CVB-2, and Polio 1 (EC₅₀, 6-25 μM). According to molecular dynamics simulations, compounds (15) and (17) emerged for its potency against the HCV NS5B, with a calculated IC₅₀ of 11-12 μM.

Full text of DOI:10.1016/j.bmc.2007.01.005

Bioorganic & Medicinal Chemistry 15 (2007) 1914–1927
Design, synthesis, and preliminary in vitro and in silico antiviral
activity of [4,7]phenantrolines and 1-oxo-1,4-dihydro-
[4,7]phenantrolines against single-stranded
positive-sense RNA genome viruses
Antonio Carta,a,* Mario Loriga,a Giuseppe Paglietti,a Marco Ferrone,b
Maurizio Fermeglia,b Sabrina Pricl,b Tiziana Sanna,c Cristina Ibba,c
Paolo La Collac,* and Roberta Loddoc
a
`
Dipartimento Farmaco Chimico Tossicologico, Universita degli Studi di Sassari, Via Muroni 23/a, 07100 Sassari, Italy
b
`
Laboratorio MOSE, Dipartimento di Ingegneria Chimica, dell’Ambiente e delle Materie Prime, Universita degli Studi di Trieste,
Piazzale Europa 1, 34127 Trieste, Italy
cDipartimento di Scienze e Tecnologie Biomediche, Sezione di Microbiologia e Virologia Generale e Biotecnologie Microbiche,
`
Universita degli Studi di Cagliari, Cittadella Universitaria, S.S. 554, Km 4500, 09042 Monserrato (Ca), Italy
Received 22 May 2006; revised 4 January 2007; accepted 4 January 2007
Available online 9 January 2007
Abstract—Following the antiviral screening of a wide series of new angular and linear N-tricyclic systems both in silico and in vitro,
the [4,7]phenantroline nucleus emerged as a new ring system endowed with activity against viruses containing single-stranded, posi-
tive-sense RNA genomes (ssRNA+). Here, we report our new pathway to the synthesis of this nucleus and of several related deriv-
atives, as well as the results of both cell-based antiviral assays and molecular dynamics simulations. In the antiviral screening, several
compounds (9 and 16–20) showed to be fairly active against BVDV, CVB-2, and Polio 1 (EC50, 6–25 lM). According to molecular
dynamics simulations, compounds (15) and (17) emerged for its potency against the HCV NS5B, with a calculated IC50 of
11–12 lM.
ꢀ 2007 Elsevier Ltd. All rights reserved.
1. Introduction
170 million people worldwide are presently infected with
this virus.3 Most infections become persistent, and
about 60% of cases progress toward chronic liver dis-
ease, which in turn, can lead to development of cirrhosis,
hepatocellular carcinoma, and liver failure.4,5 Actually,
with the exception of YFV, no vaccine exists against
the various Flaviviridae members; therefore, new thera-
pies and preventative agents are strongly needed.
Viruses belonging to the Flaviviridae family have a single-
stranded, positive-sense RNA genome (ssRNA+)1 and
cause clinically significant diseases in humans and
animals. This virus family includes three genera, i.e.,
Pestiviruses (i.e., bovine viral diarrhea virus [BVDV]),
Flaviviruses (i.e., yellow fever [YFV], dengue, and West
Nile viruses), and Hepaciviruses (Hepatitis C virus
[HCV]).
Other important ssRNA+ viruses are those belonging to
the Picornaviridae family. These viruses cause a variety
of illnesses, including meningitis, cold, heart infection,
conjunctivitis, and hepatitis6. This family includes nine
genera, some of which comprise major human
pathogens, namely, Enterovirus (including Poliovirus,
Coxsackievirus, Echovirus), Rhinovirus (approximately
105 serotypes), and Hepatovirus (Hepatitis A virus
[HAV]). At present, no specific antiviral therapy is
available for the treatment of Picornaviridae infections.
HCV is a major cause of human hepatitis, globally.2 The
World Health Organization (WHO) estimates that over
Keywords: [4,7]Phenantrolines; ssRNA+; RNA-dependent RNA-poly-
merase; HCV NS5B; Pestiviruses; Flaviviruses; Enteroviruses; Molec-
ular dynamics simulations.
*
Corresponding authors. Tel.: +39 079228722; fax: +39 07922820;
e-mail addresses: acarta@uniss.it; placolla@unica.it
0968-0896/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2007.01.005

A. Carta et al. / Bioorg. Med. Chem. 15 (2007) 1914–1927
1915
As part of our antiviral research program, we have
recently synthesized and tested in vitro a series of new
angular and linear N-tricyclic systems against represen-
tatives of the above-cited ssRNA+ viruses. Further we
tested these compounds for activity against viruses rep-
resentative of an additional genus of ssRNA+ genomes
[Human immunodeficiency virus (HIV-1)], of double-
stranded RNA genomes (dsRNA) [Reovirus (Reo-1)],
and of single-stranded, negative-sense RNA genomes
(ssRNAꢀ) [respiratory syncytial virus (RSV) and vesicu-
lar stomatitis virus (VSV)]. Interestingly, [4,7]phenantr-
olines were found to be endowed with inhibitory
activity against members of the Flaviviridae and Picor-
naviridae families.
interactions between the latter and the HCV NS5B
were studied in detail, the relevant free energies of bind-
ing DGbind calculated, and the corresponding IC50
values predicted accordingly.
The [4,7]phenantroline nucleus is known since 19098 and
some of its derivatives have been subsequently de-
scribed.9 Here, we report a new pathway for the synthe-
sis of this nucleus and of related derivatives.
Substituents at position 1 (chloro, alkoxy, amino) were
chosen in order to verify whether they were able to
either improve or direct the activity against specific
viruses. 1-Oxo-[4,7]phenantroline derivatives were nec-
essary intermediates of the reaction, and they were also
tested for biological activity. In addition, we evaluated
the influence on the antiviral activity of chlorine atoms
at positions 5 and 6.
In the light of some similarities between a number of
our active compounds and the nonnucleoside proprie-
tary inhibitor ((2s)-2-[(2,4-dichloro-benzoyl)-(3-trifluo-
romethyl-benzyl)-amino]-3-phenyl-propionic
acid)7
(hereafter named Shire A, see Fig. 1) targeting the
HCV NS5B, we considered that this enzyme could also
be the target of title compounds. Thus, the in silico
2. Chemistry
The synthesis of key intermediates 6-amino-7,8-dichlo-
roquinoline (7) and diethyl 2-[(7,8-dichloroquinol-6-yla-
mino)-methylene]malonate (8) is outlined in Scheme 1.
The synthetic route of (7) was modified according to a
procedure previously described by some of us.10 The
mixture of regioisomers (3a and 3b) obtained by nitra-
tion of the acetanilide (2) at 2–5 ꢁC was hydrolyzed with
concentrated sulfuric acid, yielding the corresponding
mixture of anilines (4 and 5). Chromatographic separa-
tion afforded the desired aniline (4). The nitroquinoline
(6) was in turn obtained via Skraup synthesis starting
from (4).10 Finally, the known aminoquinoline (7) was
obtained in 60% yield by reduction of (6) with meth-
ylhydrazine in ethanol, heated in a sealed steel vessel
at 100 ꢁC for 65 h. The same reaction performed with
hydrazine hydrate (99%) in ethanol at 160 ꢁC (in sealed
O
OH
N
F
F
F
O
Cl
Cl
Figure 1. Chemical structure of ((2s)-2-[(2,4-dichloro-benzoyl)-(3-tri-
fluoromethyl-benzyl)-amino]-3- phenyl-propronic acid or Shire A), a
proprietary nonnucleoside inhibitor of HCV NS5B.7
H2N
Cl
(CH3CO)2O
O2N
Cl
1
H3CCONH
Cl
NO2
Cl
O2N
H2N
Cl
KNO3/H2SO4
Cl
3a
+
H2SO4
H2N
Cl
H3CCONH
Cl
Cl
NO2
Cl
Cl
2
(98%)
4
5
(55%)
(24%)
H3CCONH
Cl
3b
Skraup
CO2Et
CO2Et
EMME
CH3HNNH2
EtOH
NO2
Cl
NH2
Cl
NH
Dowtherm
150 ˚C
100 ˚C / 80 h
N
N
N
Cl
Cl
Cl
Cl
6
8
7
(58%)
(78%)
(60%)
Scheme 1. Synthetic pathway for key quinoline intermediates 7 and 8.

1916
A. Carta et al. / Bioorg. Med. Chem. 15 (2007) 1914–1927
steel vessel) for 10 h afforded the aminoquinoline (7) in
lower yield (50%).10 This new procedure allowed
improving the total yield of the five steps (19% against
11%) necessary to obtain compound (7). The second
key intermediate (8) was obtained in good yield (78%),
according to a described procedure,11 by reaction of
(7) with diethyl ethoxymethylenemalonate (EMME) in
Dowtherm (mixture of biphenyl and phenyl ether in
26.5:73.5 ratio) at 150 ꢁC.
the ester (11) at 250 ꢁC for 12 h resulted in compound
hydrolysis followed by decarboxylation, and yielded 12
in 81% yield. Acid hydrolysis of (11) gave the acid (13)
in 91% yield. Known derivative (14)9a was finally ob-
tained carrying out the dehalogenation of 12 (56% yield)
with hydrazine and palladium–charcoal, as described
above.
The synthesis of 1-substituted-[4,7]phenantrolines is
shown in Scheme 4. Ethylation with diethyl sulfate on
(12) occurred at the oxygen to give the ethoxy derivative
(16) (72% yield), whereas chlorination with POCl3 led to
the trichlorophenantroline (15) (72% yield). The latter
compound, after nucleophilic displacement by ammonia,
yielded the corresponding aminophenantroline (18) (85%
yield) that, after acetylation with acetic anhydride, gave
(20) in 78% yield. Finally, dehalogenation of (18) was car-
ried out as reported above producing (19) in 94% yield.
The syntheses of all target compounds are outlined in
Schemes 2–4. In general, [4,7]phenantroline and all its
new derivatives (9–20) were obtained in good yields
(51–94%). 5,6-Dichlorophenantroline (9), and its known
unsubstituted nucleus (10),8 was prepared submitting
the amine (7) to Skraup reaction, followed by dehalo-
genation with hydrazine and palladium–charcoal in eth-
anol heating in a sealed steel vessel at 100 ꢁC for 2.5 h
(yields of 86% and 54%, respectively) (Scheme 2).
Preparation of 1-oxo-1,4-dihydro-[4,7]phenantrolines is
summarized in Scheme 3. Intermediate (8) underwent
ring closure in Dowtherm at 250 ꢁC for 2 h, giving rise
to the expected 1-oxo-[4,7]phenantroline (11) (51%
yield), accompanied by its decarboxylation product
(12) (20% yield). Compound (12) was also obtained
alone carrying out the same reaction for a longer time
(15 h, 69% yield). Finally, we observed that heating
3. Virology and molecular modeling
Title compounds (9–20) were evaluated in vitro against
ssRNA+ viruses representative of the Flaviviridae and
Picornaviridae families. In particular, viruses representa-
tive of two of the three genera of the Flaviviridae family,
i.e., YFV (Flaviviruses) and BVDV (Pestiviruses), and of
one genus of the Picornaviridae family: CVB-2 and Polio
H2NNH2
NH2
N
N
Skraup
EtOH
N
Cl
N
Cl
Pd/C 10%
N
Cl
7
Cl
10
Reflux / 2.5 h
9
(54%)
(86%)
Scheme 2. Synthetic pathway for [4,7]-phenantroline nucleus and its 5,6-dichloroderivative.
CO2Et
O
CO2H
O
NH
NH
Cl
H2SO4
N
Cl
N
CO2Et
CO2Et
Cl
Cl
13
(91%)
11 (51%) i;
(0%) ii.
NH
Dowtherm/253˚C
Dowtherm/253˚C
i; ii
iii
N
Cl
Cl
O
O
8
H2NNH2
EtOH
NH
Cl
NH
Pd/C 10%
Reflux / 2.5 h
N
N
Cl
12
(20%) i;
(69%) ii;
(81%) iii.
14
(56%)
Scheme 3. Synthetic pathway for 1-oxo-[4,7]-phenantroline nucleus and some of its derivatives. (i) 2 h; (ii) 15 h; (iii) 13 h.

A. Carta et al. / Bioorg. Med. Chem. 15 (2007) 1914–1927
1917
CH3CH2O
(C2H5O)2SO2
CH3CH2O
O
H2NNH2
EtOH
N
N
NH
Cl
Pd/C 10%
Reflux / 2.5 h
N
Cl
N
N
Cl
Cl
12
17
(52%)
16
(72%)
POCl3
Cl
H2N
N
H2N
H2NNH2
EtOH
N
N
N
NH3/EtOH
16
+
Pd/C 10%
Reflux / 2.5 h
(5%)
N
Cl
Cl
N
Cl
15
(72%)
Cl
18
(85%)
19
(94%)
(CH3CO)2O
CH3CONH
N
N
Cl
Cl
(78%)
20
Scheme 4. Synthetic pathway for 1-substituted-[4,7]-phenantrolines.
1 (Enteroviruses) were tested. Compounds were also test-
ed against an additional ssRNA+ virus [Human immu-
nodeficiency virus (HIV-1)], and representatives of
dsRNA viruses [Reoviridae (Reo-1)] and ssRNAꢀ virus-
es [respiratory syncytial virus (RSV) and vesicular sto-
matitis virus (VSV)]. Due to the lack of efficient cell
culture systems for the multiplication of HCV, title
compounds were evaluated in silico against the most
probable HCV target protein, i.e., the RNA-dependent
RNA-polymerase. All compounds were modeled and
docked into the protein, following an ad hoc developed
procedure. The corresponding free energies of binding,
Table 1. Cytotoxicity and antiviral activity of [4,7]phenantrolines (9, 10, and 15–20) and 1-oxo-1,4-dihydro-[4,7]phenantrolines (11–14) against
ssRNA+ viruses
Compound
Pestiviruses
aCC50 eMDBK bEC50 BVDV aCC50 fBHK cEC50 YFV aCC50 gVero cEC50 CVB-2 cEC50 Sb-1 dCC50
hMT-4 HIV-1
Flaviviruses
Enteroviruses
Retroviruses
cEC50
9
10
11
12
13
14
15
16
17
18
19
20
>100
>100
>100
>100
>100
>100
54
16
17
>100
>100
25
>100
>100
>25
25
80
11
>80
>90
>80
>100
>100
>100
2
6
>80
>90
>80
85
24
61
>24
>61
35
16
90
80
32
47
>32
>47
>100
23
66
>100
>23
>66
>100
95
100
>100
>100
75
>100
>100
>100
>100
>100
>100
11
>54
25
>100
25
35
>100
>25
>35
6.4
>6.4
64
64
>100
56
>100
>56
6
80
40
5
17
14
24
63
P24
9
11
>100
18
>11
>100
>18
2
7.2
>7.2
90
50
>90
2
>90
21
50
13
>50
>13
32
22
a–dData represent mean values for three independent determinations. Variation among duplicate samples was less than 15%. Interesting antiviral
activity values have been highlighted in bold for convenience.
a Compound concentration (lM) required to reduce the viability of mock-infected cells by 50%, as determined by the MTT method, or the
confluence of the monolayer, as determined by methylene blue staining.
b Compound concentration (lM) required to achieve 50% protection from virus-induced cytopathogenicity, as determined by the MTT method.
c Compound concentration (lM) required to reduce the virus plaque number by 50%.
d Compound concentration (lM) required to reduce cell proliferation by 50%, as determined by the MTT method, under conditions allowing
untreated controls to undergo at least three consecutive rounds of multiplication.
e Madin Darby bovine kidney.
f Baby hamster kidney.
g Monkey kidney.
h CD4+ human T-cells.

1918
A. Carta et al. / Bioorg. Med. Chem. 15 (2007) 1914–1927
as well as the estimated IC50, were then calculated from
molecular dynamics (MD) simulations.
As outlined before, these in vitro tests were paralleled by
in silico experiments. In general, there are two key
requirements for the computer-aided structure-based
4. Results and discussion
[4,7]Phenantrolines (9, 10, and 15–20) and 1-oxo-[4,7]-
phenantrolines (11–14) were evaluated in parallel cell-
based assays for cytotoxicity and antiviral activity
(Table 1). Different compounds exhibited different cyto-
toxicity for different cell lines. Compounds 18 and 20
turned out to be the most cytotoxic for all the cell lines
used in this study. On the other hand, compound 9 was
slightly cytotoxic for Vero and MT-4, compound 11 for
BHK and MT-4, compounds 13, 16, and 17 for BHK,
whereas compound 15 was the most cytotoxic for MT-
4 cell lines.
As far as the antiviral activity is concerned, title phe-
nantrolines can be divided into derivatives devoid of
(10, 14, 17, and 19) or characterized by (9, 11–13, 15,
16, 18, and 20) the 5,6-dichloro substituent. With the
due exceptions, the former are specifically active against
the Pestivirus representative, whereas the latter are more
broadly active against Pesti and Enterovirus representa-
tives. None of title compounds, however, turned out ac-
tive against YFV, HIV-1 or representatives of dsRNA
(Reo) and ssRNAꢀ viruses (RSV or VSV) (data not
shown).
Among the phenantrolines devoid of the 5,6 dichloro
substituent (10, 14, 17, and 19), compounds 10 and 19
turned out to be selective inhibitors of BVDV in the
range 9–17 lM, whereas 14 was inactive and 17 was ac-
tive but not selective. The lack of the two chlorine atoms
correlates with a specific activity against BVDV, provid-
ed that position 1 is unsubstituted (10) or substituted
with amino (19) groups. In fact, when oxygen (14) is
present at this position, the antiviral activity is lost.
The reason why 17 is both potently active against
Enteroviruses and BVDV is not immediately evident
(Further studies are in progress).
Among 5,6-dichloro-phenantrolines (9, 11, 12, 13, 15,
16, 18, and 20), derivatives 9, 16, 18, and 20 inhibited
CVB-2, Polio 1, and BVDV at concentrations ranging
from 2 to 11, 6 to 21, and 16 to 25, respectively, even
if in a few cases some EC50 are close to CC50. On the
other hand, the remaining compounds were inactive
against the enteroviruses and only 11 and 12 showed
activity against BVDV.
SAR studies suggest that the presence of a 5,6-dichloro
substituent correlates with the broad spectrum activity
of phenantrolines 9, 16, 18, and 20, which is character-
ized by a higher potency and selectivity against Coxsack-
ie B-2 provided that position 1 is unsubstituted (9) or
substituted with alkoxy (16), amino (18) or acetylamino
(20) groups. In fact, when oxygen (1-oxo-[4,7]phenantro-
line derivatives 11–13) or chlorine (1-chloro-[4,7]phe-
nantroline 15) is present at this position, the antiviral
activity is lost. The reason why 11 and 12 maintain some
anti-BVDV activity is not immediately evident.
Figure 2. (a, top) Comparison between the co-crystallized conforma-
tion of inhibitor Shire A into the allosteric binding site of HCV NS5B;
(b, middle) the corresponding docked conformation, and (c, bottom)
binding mode of compound (9) in the same enzyme pocket. Models in
Figure 1b and c were obtained upon application of the computational
strategy adopted in this work. The inhibitor molecules are shown as a
stick model, and the atom color-coding is as follows: carbon, gray;
nitrogen, blue, oxygen, red; fluorine, cyan; and chlorine, green
Hydrogen atoms and water molecules are omitted for clarity.

A. Carta et al. / Bioorg. Med. Chem. 15 (2007) 1914–1927
1919
drug design methods: correctly generating conforma-
tions of docked ligands and accurately predicting the
binding affinity. Accordingly, as a preliminary test, we
modeled, docked, and calculated the free energy of bind-
ing (and the corresponding IC50 values) for the proprie-
tary inhibitors of HCV NS5B Shire A, for which both
the crystallographic structures of the relevant RdRp
complex and the corresponding IC50 value were avail-
able.7 This was basically done by removing the inhibitor
from the enzyme allosteric site, building a new molecu-
lar model for this compound, applying the conforma-
tional procedure described in details in the Section 6
for the phenantroline derivatives, and finally docking
it back into the NS5B surface pocket. The best docked
structure, which is the configuration with the lowest
docking energy in a prevailing cluster, was then
compared with the corresponding crystal structure.
Figure 2a and b shows a comparison between the
Figure 3. (a, top) Ribbon diagram of HCV NS5B/compound (9) complex structure as resulting from the applied docking/MD procedure. The protein
is colored according to its secondary structure: a-helices, purple; b-sheets, gold; turns and coils, light gray. The inhibitor (9) is represented as a CPK
model with carbons in gray, nitrogens in blue, and chlorines in green. (b, middle) Detail of compound (9) (in a stick representation) in the binding
pocket in the enzyme thumb subdomain. Color scheme as above. (c, bottom) A stereo view of the inhibitor-binding site of HCV NS5B/(9) complex.
The inhibitor molecule is shown as a stick model (color scheme as above). The side chains of all residues that form the primary binding pocket
interacting with compound (9) are shown as stick models, and the atom color-coding is as follows: L419, brown; R422, gold; M423, hot pink; L474,
olive green; H475, dark slate blue; S476, dark cyan; Y477, plum; I482, sea green; L497, kaki; R501, sky blue; W528, red. Hydrogen atoms and water
molecules are omitted for clarity.

1920
A. Carta et al. / Bioorg. Med. Chem. 15 (2007) 1914–1927
co-crystallized conformation of inhibitor Shire A into the
allosteric binding site of NS5B and the docked conforma-
tion obtained upon application of the computational
strategy adopted in this work. At a first glance it can be
seen that the agreement between the two structures is
excellent. Further, the root-mean-square deviation
(RMSD) between the docked configuration and the rele-
vant crystal structures of this test inhibitor is equal to
and hydrophobic interactions are exchanged with the
side chains of L419, M423, Y477, L474, H475, and
W528. In addition, the same charged side chain of
R422 contributes to screen the electronegative chlorine
atoms, whilst the positive partial charges on the edges
of the same aromatic ring bearing the halogen substitu-
ents are counterbalanced by the carbonyl oxygen atoms
of L497 and F472, thus yielding favorable electrostatic
stabilization to the complex.
˚
0.01 A. In the light of these blank-test results, and of
the fact that this computational procedure was already
successfully applied by us to predict the anti-HCV activ-
ity of other allosteric inhibitors of HCV NS5B,12 the con-
ceived modeling/docking procedure was applied for
predicting the binding mode of the phenantroline deriva-
tives in the HCV nonnucleoside binding site targeted by
Shire A. Figure 2c illustrates, as an example, compound
9 docked into the HCV NS5B allosteric binding site.
In analogy with the nonnucleoside inhibitor Shire A, one
of the heterocyclic rings and part of the unsubstituted
side of inhibitor 9 basically float above the protein sur-
face without making any strong interaction with the en-
zyme (see Fig. 2c). By inducing a rigid constraint to the
entire molecular structure, however, the third condensed
ring may contribute to a global conformation that is
favorable to the protein/small molecule interaction.
This phenantroline derivative occupies the central por-
tion of the polymerase extended allosteric cleft (see
Fig. 3a and b), and the detailed interactions between 9
and the enzyme are shown in Figure 3c. The bottom
of the binding pocket is formed by the main-chain atoms
and the side chains of R501, W528, and H475, whilst the
surrounding walls are formed by the side chains of
R422, M423, L474, S476, and Y477. Finally, residues
L497, L419, and I482 enclose the inhibitor from top
(see Fig. 3c). As results from the inspection of the rele-
vant HCV NS5B/9 complex MD trajectory, this inhibi-
tor makes a series of hydrophobic and van der Waals
interactions with the residues lining the allosteric bind-
ing pocket in the enzyme. However, one of the most
important interactions is a dual hydrogen bond that
the positively charged guanidinium group of R422
forms with one of the heterocyclic nitrogen atoms of
the inhibitor on one side (average dynamic length
The success of docking molecules into a target site is
ultimately dependent, however, on the accuracy of the
scoring function that ranks the compounds or, better,
how well the corresponding binding affinities can be pre-
dicted. Ligand binding is governed by kinetic and ther-
modynamic principles. Factors that contribute to
ligand binding include the hydrophobic effect, van der
Waals interactions, hydrogen bonding, other electrostat-
ic interactions, solvation, and entropic effects. If the
change in free energy associated with complex formation
is negative, the association will be favorable. Thus, once
a candidate ligand is constructed, its interaction energy
with its target protein is calculated and compared with
that for other existing ligands. Accordingly, the last
row of Table 2 reports the binding free energy DGbind
,
and the corresponding IC50 values, obtained from the
application of our molecular simulation procedure to
test compound Shire A. As seen in Table 2, the value cal-
culated is in outstanding agreement with the corre-
sponding experimental result. Taken together, all these
evidences constitute a proof of the validity of the overall
˚
(ADL) = 2.7 A), and with oxygen of the main-chain
˚
CO group of L474 on the other (ADL = 2.3 A). The
dichlorophenyl ring of 9 has extensive favorable interac-
tions with at least seven residues. Topical van der Waals
Table 2. Free energy components, total binding energy DGbind, and IC50 values resulting from molecular dynamics simulations on HCV NS5B in
complex with compounds (9–20)
Compound
DEvdW
DEele
DGPB
DGNP
ꢀTDS
10.1
9.1
DGbind
aIC50,calc
9
10
ꢀ46.5 (0.5)
ꢀ25.6 (0.3)
ꢀ50.2 (0.4)
ꢀ49.1 (0.4)
ꢀ49.3 (0.5)
ꢀ25.4 (0.3)
ꢀ47.2 (0.4)
ꢀ48.3 (0.5)
ꢀ27.2 (0.4)
ꢀ46.0 (0.4)
ꢀ28.9 (0.4)
ꢀ47.1 (0.5)
ꢀ66.8 (0.4)
ꢀ37.9 (0.5)
ꢀ30.3 (0.6)
ꢀ86.0 (1.0)
ꢀ62.3 (0.9)
ꢀ85.1 (0.8)
ꢀ31.5 (0.4)
ꢀ55.8 (0.5)
ꢀ86.3 (0.6)
ꢀ36.0 (0.5)
ꢀ77.9 (0.8)
ꢀ50.6 (0.3)
ꢀ80.2 (0.9)
ꢀ89.1 (0.6)
71.8 (0.8)
43.3 (0.6)
114.1 (1.1)
96.4 (0.9)
118.1 (1.1)
45.5 (0.4)
88.8 (0.5)
118.2 (0.5)
47.1 (0.3)
110.0 (0.8)
66.3 (0.5)
109.6 (0.6)
133.1 (0.5)
ꢀ3.9 (0.1)
ꢀ2.9 (0.1)
ꢀ3.5 (0.1)
ꢀ2.9 (0.1)
ꢀ3.1 (0.1)
ꢀ2.9 (0.1)
ꢀ3.2 (0.1)
ꢀ3.4 (0.1)
ꢀ3.2 (0.1)
ꢀ3.1 (0.1)
ꢀ2.9 (0.1)
ꢀ3.5 (0.1)
ꢀ4.4 (0.1)
ꢀ6.4
ꢀ6.4
ꢀ5.8
ꢀ6.3
ꢀ5.2
ꢀ5.4
ꢀ6.8
ꢀ6.0
ꢀ6.7
ꢀ5.9
ꢀ6.3
ꢀ6.3
ꢀ7.7
20.0
20.0
59.0
25.0
144.0
110.0
11.0
39.0
12.0
50.0
25.0
25.0
11
12
19.8
11.6
14.2
8.9
13
14
15
16
10.6
13.8
12.6
11.1
9.8
17
18
19
20
14.9
19.5
Shire A
2.3 (2.2)b
The values obtained for the proprietary inhibitor Shire A are also reported for comparison. All energy values are in kcal/mol. The calculated IC50
values are in lM. The values in parentheses represent the standard error of the mean. The values of the total free energy of binding and the
corresponding estimated IC50 values (last two columns) have been highlighted in bold for convenience.
a IC50 values are calculated from Eq. 413,14 (see Section 6 for details).
b Experimental value from Ref. 7.

A. Carta et al. / Bioorg. Med. Chem. 15 (2007) 1914–1927
1921
modeling procedure adopted, from model building to
5. Conclusions
protein docking and free energy calculations. Thus, we
applied the same strategy for the estimation of the affin-
ities of our set of phenantroline derivatives to HCV
NS5B, as a prediction of the enzyme–drug interaction.
The remaining rows of Table 2 list the results obtained
for all compounds (9–20).
In the light of the above-mentioned results, we conclude
that [4,7]phenantroline derivatives showed, in general,
to be endowed with good activity against viruses con-
taining a single-stranded positive-sense RNA genome
(ssRNA+) whereas 1-oxo-1,4-dihydro-[4,7]phenantro-
lines are less active or completely inactive. In particular
in cell-based assays the compounds 9, 16, and 17 the
most potent against Polio-1, CVB-2, and BVDV, respec-
tively. Furthermore, owing to their good ratio of activ-
ity/cytotoxicity they might represent new interesting
leads. Molecular modeling of the interactions between
this sets of compounds and their putative final target,
the HCV NS5B, confirmed both the trend in the activity
of these compound and, in particular, the choice of
compound (17) as a promising lead, worthy of further
development as a nonnucleoside inhibitor of HCV
RdRp.
From the inspection of this Table we can see that the ap-
plied computational strategy is able to yield absolute free
energies of binding, and hence IC50 values, which are
quite reasonable. Further insights into the forces involved
in the binding of these compounds to HCV NS5B can be
found by analyzing the contributions afforded by the sin-
gle components to DGbind, which are also listed in Table 2.
Comparing the van der Waals/nonpolar (DEvdW + DGNP
)
with the electrostatic (DEele + DGPB) contributions, we
find that the association between the 12 ligands and the
RdRp is mainly driven by more favorable nonpolar inter-
actions in the complex than in the solution, in harmony
with a proposed general scheme for noncovalent associa-
tion.15 However, as indicated for instance by the energy
components of compound (13), this driving force can be
considerably weakened when the polar groups on the
molecule do not find an adequate bonding pattern in
the protein compared to water. The free energy penalty
for this (DEele + DGPB) is least for compound (17) and
together with its substantial van der Waals contribution
and moderate entropic unfavorable term consequently
leads to one of the highest binding affinities in this set of
phenantroline inhibitors.
In conclusion, we can affirm that [4,7]phenantrolines,
owing to both the new synthetic pathway proposed
and the good antiviral activities recorded, can be further
developed through the introduction of alternative sub-
stituents on the position 1 and submitted to in vivo
experiments.
6. Experimental
6.1. Chemistry
Finally, we can observe a good agreement between the
trend exhibited by the IC50 values reported in Table 2
and the corresponding biological activity determined
for these compounds in BVDV infected cell lines (see
Table 1). Although we obviously cannot directly com-
pare the computed binding free energy (and hence the
corresponding IC50) with the EC50 values derived from
experiment, we can observe that, in most cases, the rank
of the inhibitors with respect to their activity toward
their putative targets, the RdRps of BVDV and HCV,
respectively, is maintained, although with some discrep-
ancies. This is quite reasonable if we consider that the
thumb domain of the RdRps where the allosteric bind-
ing site is located is one of the most diverse features
among the polymerase structures. Although there is
some structural similarity between the HCV and BVDV
thumb domains, the overall topology is rather different.
Further, only a few of the amino acids making up the
allosteric binding site are conserved in the two viral en-
zymes, as we inferred from our alignment of the two
RdRp primary sequences (data not shown). Notwith-
standing, it is important to remark here that, according
to our modeling considerations reported above, the
main interactions between inhibitors (9–20) and the
RdRp of HCV take place with residues R422 and
L474, which are conserved in the BVDV counterpart
(R560 and L601, respectively); further, other interac-
tions involve main-chain groups which do not depend
on the corresponding residue side chain type. This
undoubtedly contributes to justify the agreement
between the different data sets.
Melting points (mp) are uncorrected and were taken in
open capillaries in a Digital Electrothermal IA9100
melting point apparatus. H NMR spectra were record-
1
ed on a Varian XL-200 (200 MHz) instrument, using
TMS as internal standard. The chemical shift values
are reported in ppm (d) and coupling constants (J) in
Hertz (Hz). Signal multiplicities are represented by: s
(singlet), d (doublet), dd (double doublet), t (triplet), q
(quadruplet) and m (multiplet). MS spectra were record-
ed on a combined HP 5790 (GC)-HP 5970 (MS) appara-
tus. Column chromatography was performed using
70–230 mesh (Merck silica gel 60). Light petroleum re-
fers to the fraction with bp 40–60 ꢁC. The progress of
the reactions and the purity of the final compounds were
monitored by TLC using Merck F-254 commercial
plates. Analyses indicated by the symbols of the ele-
ments were within 0.4% of the theoretical values.
6.1.1. Preparation of the intermediates 6-amino-7,8-
dichloroquinoline (7) and diethyl 2-[(7,8-dichloroquinol-
6-ylamino)-methylene]-malonate (8). A suspension of
2,3-dichloroaniline (1) (50.0 g, 308.6 mmol) in acetic
anhydride (150 mL) was stirred at room temperature
overnight, the resulting precipitate was filtered off,
washed with water, and dried to afford 61.71 g (98%)
of the desired 2,3-dichloroacetanilide (2). To a solution
of 20 g (98.01 mmol) of anilide (2) in 100 mL of concen-
trated sulfuric acid, cooled at 2–5 ꢁC, a solution of
potassium nitrate (9.90 g, 9.79 mol) in 20 mL of concen-
trated sulfuric acid was slowly added under vigorous
stirring. After the addition was complete, the reaction

1922
A. Carta et al. / Bioorg. Med. Chem. 15 (2007) 1914–1927
mixture was stirred always below 5 ꢁC for an additional
4 h and poured into 500 g of crushed ice. The resulting
precipitate (mixture 3a and 3b) was filtered off and add-
ed under stirring with 150 mL of concentrated sulfuric
acid. The solution obtained was heated at 100–110 ꢁC
and the stirring was continued for an additional 2 h.
On cooling at room temperature, the reaction mixture
was added to 600 g of crushed ice and the crude precip-
itate was collected by filtration and purified by chroma-
tography on silica gel eluting with a 8:2 mixture of
diethyl ether–light petroleum to yield in sequence 2,3-
dichloro-6-nitroaniline (5) (4.87 g, 24%) and 2,3-dichlo-
ro-4-nitroaniline (4) (11.16 g, 55%). Six grams
(28.97 mmol) of the latter compound was added to
10 g (108.87 mmol) of glycerol and 15 g (54.1 mmol) of
arsenic pentoxide hydrate; the resulting mixture was
then cooled to 0 ꢁC and slowly added with 8 mL of con-
centrated sulfuric acid under vigorous mechanical stir-
ring. After complete addition, the temperature was
slowly raised to 150 ꢁC and the stirring continued for
an additional 3 h. On cooling at room temperature,
the reaction mixture was added to 250 g of crushed ice
and the crude precipitate was collected by filtration
and purified by chromatography on silica gel eluting
with a 8:2 mixture of diethyl ether–light petroleum to
yield 4.23 g (17.40 mmol, 58%) of 7,8-dichloro-6-nitro-
quinoline (6). A solution of 6 (3.0 g, 12.4 mmol) and
methyl hydrazine (3 mL) in 120 mL of ethanol was heat-
ed in a sealed steel vessel at 100 ꢁC for 65 h. The reaction
mixture was then cooled at room temperature and the
solvent was removed in vacuo. The crude solid was puri-
fied by chromatography on silica gel eluting with diethyl
ether to give 6-amino-7,8-dichloroquinoline (7) (1.58 g,
60%). Melting points, analytical, and spectroscopical
data of intermediates (2–7) have been previously report-
ed10. To a stirring suspension of the amine (7) (2.51 g,
11.8 mmol) in 40 g of Dowtherm, 3.30 g (15 mmol) of
EMME was added, then the temperature was raised to
150 ꢁC and the stirring continued for an additional
15 h. On cooling, the solution was taken up with
150 mL of hexane. The precipitate formed was filtered
off and washed with hexane to give:
tate was collected by filtration and crystallized by
ethanol to yield 3.06 g (86% of yield) of 9. Mp 265–
266 ꢁC, IR (Nujol): mmax 1650, 1570 cmꢀ1; UV (EtOH):
kmax 224, 258, 289, 373 nm; 1H NMR (CDCl3+
DMSO-d6): d 9.15 (dd, 2H, J = 4.6 and 1.6 Hz, H-3 +
H-8), 9.08 (d, 2H, J = 8.4 Hz, H-1 + H-10), 7.78 (dd,
2H, J = 4.6 and 1.6 Hz, H-2 + H-9); MS m/z 248/250/
252 (M+). Anal. C12H6Cl2N2 (C, H, N).
6.1.3. Preparation of [4,7]phenantroline (10). A suspension
of 9 (1.0 g, 4 mmol), 10% palladium–charcoal (0.20 g),
and 3 mL of hydrazine monohydrate (98%) in 20 mL of
ethanol was refluxed under stirring for 2.5 h. On cooling,
the catalyst was filtered off and the solvent and the excess
of hydrazine were removed in vacuo. The crude solid was
recrystallized by acetone to yield 0.38 g (54% yield) of 10.
Mp 217–218 ꢁC, IR (Nujol): mmax 1623, 1597 cmꢀ1; UV
(EtOH): kmax 232, 274 nm;1H NMR (DMSO-d6): d 9.71
(d, 2H, J = 5.4 Hz, H-3 + H-8), 9.26 (d, 2H, J = 8.6 Hz,
H-1 + H-10), 8.53 (s, 2H, H-5 + H-6), 8.10 (d, 2H,
J = 8.6 and 5.4 Hz, H-2 + H-9); MS M/Z 180 (M+). Anal.
C12H8N2 (C, H, N).
6.1.4. Preparation of 5,6-dichloro-1-oxo-1,4-dihydro-
[4,7]phenantroline-2-corboxylic acid ethyl ester (11) and
5,6-dichloro-1-oxo-1,4-dihydro-[4,7]phenantroline (12).
Intermediate (8) (3.0 g, 7.8 mmol) was suspended in
50 g of Dowtherm and heated at 250 ꢁC under stirring
for 2.5 h. On cooling, it was taken up with 200 mL of hex-
ane. The precipitate formed was filtered off and chro-
matographed on silica gel eluting with a 9:1 mixture of
diethyl ether–light petroleum to yield in sequence 11
(1.34 g, 51%) and 12 (0.40 g, 20%). Continuing the stir-
ring at 250 ꢁC for an additional 12.5 h, after dilution of
the Dowtherm solution, gave the sole compound (12) in
69% yield. Alternatively, by submitting the ester (11)
(1.0 g, 3.0 mmol) to the same reaction in 20 g of Dow-
therm at 250 ꢁC under stirring for 13 h, compound (12)
was obtained in 81% yield.
6.1.4.1. 5,6-Dichloro-1-oxo-1,4-dihydro-[4,7]phenantr-
oline-2-carboxylic acid ethyl ester (11). Mp 235–237 ꢁC
(acetone), IR (Nujol): mmax 3196, 1714, 1615 cmꢀ1; UV
(EtOH): kmax 211, 225, 253, 281, 342, 358 nm; 1H NMR
(DMSO-d6): d 12.16 (s, 1H, NH), 10.47 (d, 1H,
J = 8.8 Hz, H-8), 9.01 dd, 1H, J = 4.2 Hz, H-10), 8.39
(s, 1H, H-3), 7.79 (dd, 1H, J = 8.8 and 4.2 Hz, H-9),
4.28 (q, 2H, J = 7.0 Hz, CH2), 1.33 (t, 3H, J = 7.0 Hz,
CH3); MS m/z 336/338/340 (M+). Anal. C15H10Cl2N2O3
(C, H, N).
6.1.1.1. Diethyl 2-[(7,8-dichloroquinol-6-ylamino)-
methylene]-malonate (8). Yield 78%, mp 171–173 ꢁC (ace-
tone), IR (Nujol): mmax 3158, 1697, 1655, 1589 cmꢀ1; UV
(EtOH): kmax 211, 241, 295, 329, 351 nm; 1H NMR
(CDCl3): d 11.53 (d, 1H, J = 12.8 Hz, NH), 8.98 (dd,
1H, J = 4.4 and 1.6 Hz, H-2), 8.61 (d, 1H, J = 12.8 Hz,
CH=), 8.16 (dd, 1H, J = 8.2 e 1.6 Hz, H-4), 7.57 (s, 1H,
H-5), 7.51 (dd, 1H, J = 8.2 e 4.4 Hz, H-3), 4.36 (m, 4H,
2 CH2), 1.39 (m, 6H, 2 CH3); MS m/z 382/384/386 (M+).
Anal. C17H16Cl2N4O4 (C, H, N).
6.1.4.2. 5,6-Dichloro-1-oxo-1,4-dihydro-[4,7]phenantr-
oline (12). Mp > 300 ꢁC (ethanol), IR (Nujol): mmax 3467,
3389, 1624 cmꢀ1; UV (EtOH): kmax 226, 272, 344, 358 nm;
1H NMR (DMSO-d6): d 11.85 (s, 1H, NH), 10.66 (d, 1H,
J = 8.8 Hz, H-8), 9.00 (d, 1H, J = 4.2 Hz, H-10), 7.94 (d,
1H, J = 7.2 Hz, H-3), 7.80 (dd, 1H, J = 8.8 and 4.2 Hz,
H-9), 6.43 (d, 1H, J = 7.2 Hz, H-2); MS m/z 250/252/254
(M+). Anal. C12H6Cl2N2O (C, H, N).
6.1.2. Preparation of 5,6-dichloro-[4,7]phenantroline (9).
To a mixture of 7 (3.0 g, 14.1 mmol), glycerol (5.4 g,
58 mmol), and arsenic pentoxide hydrate (10.0 g
31 mmol), cooled at 0 ꢁC, 4 mL of concentrated sulfuric
acid was slowly added under vigorous mechanical stir-
ring. After addition, the temperature was slowly raised
to 150 ꢁC and the stirring continued for an additional
2 h. The mixture was then cooled to room temperature,
poured into 250 g of crushed ice, and the crude precipi-
6.1.5. Preparation of 5,6-dichloro-1-oxo-1,4-dihydro-
[4,7]phenantroline-2-carboxylic acid (13). The ester (11)
(0.5 g, 1.4 mmol) was dissolved in 50 mL of concentrated

A. Carta et al. / Bioorg. Med. Chem. 15 (2007) 1914–1927
1923
sulfuric acid and heated at 100–110 ꢁC under stirring for
2 h. On cooling, the solution was poured onto 50 g of
crushed ice. The crude precipitate formed was collected
by filtration and crystallized by ethanol to yield 0.41 g
(91% of yield) of 13. Mp > 300 ꢁC, IR (Nujol): mmax
3558, 3401, 1715, 1617 cmꢀ1; UV (EtOH): kmax 207,
H-2), 4.46 (q, 2H, J = 7.0 Hz, CH2), 1.38 (t, 3H,
J = 7.0 Hz, CH3); MS m/z 224 (M+). Anal. C14H12N2O
(C, H, N).
6.1.9. Preparation of 1,5,6-trichloro-[4,7]phenantroline
(15). A solution of 12 (1.50 g, 6 mmol) in 20 mL of phos-
phorus(V)oxychloride (99%) was refluxed under stirring
for 48 h. The reaction mixture was then cooled at room
temperature and poured into 200 g of crushed ice and
the resulting solution was made alkaline with 25% ammo-
nia solution. The crude solid obtained was purified by
chromatography on silica gel with a 9:1 mixture of chlo-
roform–methanol to give 15 (1.2 g, 72%). Mp 238–
240 ꢁC (ethanol), IR (Nujol): mmax 1586, 1564 cmꢀ1; UV
1
252, 279, 338, 354 nm; H NMR (DMSO-d6): d 10.29
(d, 1H, J = 8.8 Hz, H-8), 9.03 (d, 1H, J = 4.2 Hz, H-10),
8.69 (d, 1H, J = 7.2 Hz, H-3), 7.82 (dd, 1H, J = 8.8 and
4.2 Hz, H-9); MS m/z 308/310/312 (M+). Anal.
C13H6Cl2N2O3 (C, H, N).
6.1.6. Preparation of 1-oxo-1,4-dihydro-[4,7]phenantro-
line (14). A suspension of 12 (0.6 g, 1.2 mmol), 10% pal-
ladium–charcoal (0.12 g), and 1 mL of hydrazine
monohydrate (98%) in 15 mL of ethanol was refluxed
under stirring for 2.5 h. The reaction mixture was then
cooled at room temperature, the catalyst was filtered
off, and the solvent and the excess of hydrazine were re-
moved in vacuo. The crude solid was recrystallized by
acetone to yield 0.26 g (56% yield) of 14. Mp 280–
1
(EtOH): kmax 211, 238, 271 nm; H NMR (DMSO-d6):
d 10.07 (d, 1H, J = 8.4 Hz, H-8), 9.13 (d, 1H,
J = 4.2 Hz, H-10), 8.98 (d, 1H, J = 8.4 Hz, H-3), 8.03
(d, 1H, J = 8.4 Hz, H-2), 7.88 (dd, 1H, J = 8.8 and
4.2 Hz, H-9); MS m/z 282/284/286 (M+). Anal.
C12H5Cl2N2 (C, H, N).
281 ꢁC, IR (Nujol): mmax 3472, 3403, 1638, 1605 cmꢀ1
;
6.1.10. Preparation of 1-amine-5,6-dichloro-[4,7]phe-
nantroline (18). A solution of 15 (2.44 g, 7.72 mmol)
in 250 mL of ethanol saturated with dry gaseous
ammonia was heated in a sealed steel vessel at
160 ꢁC under stirring for 72 h. Then the reaction mix-
ture was cooled at room temperature, the solvent was
removed in vacuo, and the solid residue chromato-
graphed on silica gel with a 9:1 mixture of chloro-
form–methanol to give in sequence 18 (1.94 g, 85%)
and 0.14 g (5%) of 16. Mp 261–263 ꢁC (ethanol), IR
(Nujol): m 3450, 3311, 1636, 1583 cmꢀ1; UV (EtOH):
UV (EtOH): kmax 221, 263, 327, 336, 349 nm; 1H
NMR (DMSO-d6): d 11.47 (s, 1H, NH), 10.84 (d, 1H,
J = 8.0 Hz, H-8), 9.16 (d, 1H, J = 4.2 Hz, H-10), 8.51
(d, 1H, 9.2 Hz, H-6), 8.37–8.26 (m, 2H, H-3 + H-5),
8.08 (dd, 1H, J = 8.0 and 4.2 Hz, H-9), 6.72 (d, 1H,
J = 7.0 Hz, H-2); MS m/z 196 (M+). Anal. C12H8N2O
(C, H, N).
6.1.7. Preparation of 5,6-dichloro-1-ethoxy-[4,7]phenantr-
oline (16). A suspension of 12 (1.0 g, 4 mmol) and 1.30 g
(4 mmol) of cesium carbonate (99%) in 20 mL of dehy-
drated dimethylformamide (less than 0.0100% of water)
(DMF) was heated under stirring at 60 ꢁC and slowly
added with a solution of ethyl sulfate (0.74 g, 4.8 mmol)
in 6 mL DMF. After the addition was complete, the reac-
tion mixture was stirred for an additional 4 h and poured
into 200 g of crushed ice. The solid obtained was recrys-
tallized by diethyl ether to yield 0.84 g (72% yield) of 16.
Mp 201–203 ꢁC, IR (Nujol): mmax 1674, 1580 cmꢀ1; UV
(EtOH): kmax 213, 240, 312, 328 nm; 1H NMR (DMSO-
d6): d 9.46 (d, 1 H, J = 8.4 Hz, H-8), 8.89 (d, 1H,
J = 4.4 Hz, H-10), 8.73 (d, 1H, J = 8.0 Hz, H-3), 7.63
(dd, 1H, J = 8.8 and 4.4 Hz, H-9), 7.23 (d, 1H,
J = 8.0 Hz, H-2), 4.30 (q, 2H, J = 7.0 Hz, CH2), 1.33 (t,
3H, J = 7.0 Hz, CH3); MS m/z 292/294/296 (M+). Anal.
C14H10Cl2N2O (C, H, N).
1
kmax 210, 232, 280, 358 nm; H NMR (DMSO-d6): d
9.42 (d, 1H, J = 8.6 Hz, H-8), 8.99 (d, 1H,
J = 4.2 Hz, H-10), 8.49 (d, 1H, J = 8.4 Hz, H-3), 7.77
(dd, 1H, J = 8.6 and 4.2 Hz, H-9), 7.09 (d, 1H,
J = 8 Hz, H-2), 7.04 (s, 2H, NH2); MS m/z 263/265/
267 (M+). Anal. C12H7Cl2N3 (C, H, N).
6.1.11. Preparation of 1-amine-[4,7]phenantroline (19). A
suspension of 18 (0.48 g, 1.82 mmol), 10% palladium–
charcoal (0.10 g), and 2 mL of hydrazine monohydrate
(98%) in 15 mL of ethanol was refluxed under stirring
for 2.5 h. On cooling, the catalyst was filtered off and
the solvent and the excess of hydrazine were removed
in vacuo. The crude solid was recrystallized by ace-
tone to yield 0.34 g (94% yield) of 19. Mp 288–
290 ꢁC, IR (Nujol): mmax 3267, 3096, 1661, 1630,
1608 cmꢀ1; UV (EtOH): kmax 234, 265, 333, 349 nm;
1H NMR (DMSO-d6): d 9.26 (d, 1H, J = 8.6 Hz, H-
8), 9.14 (s, 2H, NH2), 9.06 (d, 1H, J = 4.4 Hz, H-
10), 8.48 (d, 1H, J = 7.0 Hz, H-3), 8.39 (d, 1H,
J = 9.2 Hz, H-6), 8.32 (d, 1H, J = 9.2 Hz, H-5), 7.82
(dd, 1H, J = 8.6 and 4.4 Hz, H-9), 7.28 (d, 1H,
J = 7.0 Hz, H-2); MS m/z 195 (M+). Anal. C12H9N3
(C, H, N).
6.1.8. Preparation of 1-ethoxy-[4,7]phenantroline (17). A
suspension of 16 (0.6 g, 2.4 mmol), 10% palladium–char-
coal (0.12 g), and 2 mL of hydrazine monohydrate
(98%) in 15 mL of ethanol was refluxed under stirring
for 2.5 h. On cooling, the catalyst was filtered off and
the solvent and the excess of hydrazine were removed
in vacuo. The crude solid was recrystallized by acetone
to yield 0.24 g (52% yield) of 17. Mp 172–173 ꢁC, IR
(Nujol): mmax 1662, 1580 cmꢀ1; UV (EtOH): kmax 208,
246, 263, 338 nm; 1H NMR (DMSO-d6): d 9.42 (d,
1H, J = 8.4 Hz, H-8), 8.92 (d, 1H, J = 4.4 Hz, H-10),
8.78 (d, 1H, J = 8.0 Hz, H-3), 8.36 (d, 1 H, J = 9.2 Hz,
H-6), 8.25 (d, 1H, J = 9.2 Hz, H-5), 7.88 (dd, 1H,
J = 8.4 and 4.4 Hz, H-9), 7.59 (d, 1H, J = 8.0 Hz,
6.1.12. Preparation of N-(5,6-dichloro-[4,7]phenantrolin-
1-yl)-acetamide (20). A suspension of 18 (0.24 g,
0.91 mmol) in 10 mL of acetic anhydride was stirred at
rt for 24 h and the crude precipitate was collected by filtra-
tion and recrystallized by ethanol to yield 20 (0.22 g,
78%). Mp 254 ꢁC (dec), IR (Nujol): m 3312, 1662,

1924
A. Carta et al. / Bioorg. Med. Chem. 15 (2007) 1914–1927
1613 cmꢀ1; UV (EtOH): kmax 213, 231, 281, 324, 360 nm;
1H NMR (DMSO-d6): d 12.11 (s, 1H, NH), 9.42 (d, 1H,
J = 8.0 Hz, H-8), 8.99 (d, 1H, J = 4.2 Hz, H-10), 8.48
(d, 1H, J = 8.4 Hz, H-3), 7.76 (dd, 1H, J = 8.0 and
4.2 Hz, H-9), 7.15 (d, 1H, J = 8 Hz, H-2), 1.93 (s, 3H,
CH3); MS m/z 305/307/309 (M+). Anal. C14H9Cl2N3O
(C, H, N).
dilution (in serum-free medium) to give an moi, 0.01
(0.04 in the case of RSV). One hour later, 50 lL of
MEM Earle’s medium (Dulbecco’s modified Eagle’s
medium for Vero/RSV), supplemented with inactivated
fetal calf serum (FCS), 1% final concentration, without
or with serial dilutions of test compounds, was added.
After a 2–3-day incubation at 37 ꢁC, cell viability was
determined by the MTT method. Activity of compounds
against Coxsackie virus, B-2 strain (CVB-2), Polio virus
type-1 (Polio-1), Sabin strain, was determined by plaque
reduction assays in Vero 76 cell monolayers. To this
end, Vero 76 cells were seeded in 24-well plates at a den-
sity of 2 · 105 cells/well and were allowed to form con-
fluent monolayers by incubating overnight in growth
medium at 37 ꢁC in a humidified CO2 (5%) atmosphere.
Then, monolayers were infected with 250 lL of proper
virus dilutions to give 50–100 PFU/well. Following
removal of unabsorbed virus, 500 lL of Dulbecco’s
modified Eagle’s medium supplemented with 1% inacti-
vated FCS and 0.75% methyl cellulose, without or with
serial dilutions of test compounds, was added. Cultures
were incubated at 37 ꢁC for 2 (Sb-1), 3 (CVB-2) days
and then fixed with PBS containing 50% ethanol and
0.8% crystal violet, washed, and air-dried. Plaques were
then counted. 50% effective concentrations (EC50) were
calculated by linear regression technique.
6.2. Cell-based assays
6.2.1. Compounds. Compounds were dissolved in
100 mM DMSO and then diluted in culture medium.
6.2.2. Cells and viruses. Cell lines were purchased from
American Type Culture Collection (ATCC). The ab-
sence of mycoplasma contamination was checked peri-
odically by the Hoechst staining method. Cell lines
supporting the multiplication of RNA viruses were as
follows: Madin Darby Bovine Kidney (MDBK); Baby
Hamster Kidney (BHK-21); CD4+ human T-cells con-
taining an integrated HTLV-1 genome (MT-4); Monkey
kidney (Vero 76) cells.
6.2.3. Cytotoxicity assays. For cytotoxicity tests, run in
parallel with antiviral assays, MDBK, BHK, and Vero
76 cells were resuspended in 96-multiwell plates at an
initial density of 6 · 105, 1 · 106, and 5 · 105 cells/mL,
respectively, in maintenance medium, without or with
serial dilutions of test compounds. Cell viability was
determined after 48–120 h at 37 ꢁC in a humidified
CO2 (5%) atmosphere by the MTT method. The cell
number of Vero 76 monolayers was determined by stain-
ing with the crystal violet dye.
6.3. Molecular modeling
All molecular dynamics simulations were run on a clus-
ter of Silicon Graphics Octane RK12. The entire compu-
tational recipe involved the following program
packages: AutoDock (v. 3.0),17 AMBER 7.0,18,19 Mate-
rials Studio (v.3.2, Accelrys Inc., San Diego, CA, USA),
and in-house developed codes (stand-alone and add-on
to the commercial software).
For cytotoxicity evaluations, exponentially growing cells
derived from human hematological tumors [CD4+
human T-cells containing an integrated HTLV-1 gen-
ome (MT-4)] were seeded at an initial density of
1 · 105 cells/mL in 96-well plates in RPMI-1640 medium
supplemented with 10% fetal calf serum (FCS), 100 U/
mL penicillin G, and 100 lg/mL streptomycin. Cell cul-
tures 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 deter-
mined after 96 h at 37 ꢁC by the 3-(4,5-dimethylthia-
The starting 3-D model of the HCV RNA-dependent
RNA polymerase was based on its X-ray crystallograph-
ic structure (chain B, PBD Code: 1CSJ).20 Missing
hydrogen atoms were added to the protein backbone
and side chains with the parse module of Amber 7.0.
All ionizable residues were considered in the standard
ionization state at neutral pH. The geometry of added
hydrogen was refined for 200 steps (steepest descent)
in vacuum, using the well-validated, all-atom force field
(FF) parameter set (parm94) by Cornell et al.21 Further
protein geometry refinement was carried out using the
Sander module of Amber 7.0 via a combined steepest
descent—conjugate gradient algorithm, using as a con-
vergence criterion for the energy gradient the root-
mean-square of the Cartesian elements of the gradient
zol-2-yl)-2,5-diphenyl-tetrazolium
method.16
bromide
(MTT)
6.2.4. Antiviral assays. Activity of compounds against
yellow fever virus (YFV) was based on inhibition of
virus-induced cytopathogenicity in acutely infected
BHK-21 cells.
˚
equal to 0.01 kcal/(mol A). The generalized Born/surface
Activity against bovine viral diarrhea virus (BVDV), in
infected MDBK cells, was based on inhibition of
virus-induced cytopathogenicity.
area (GB/SA) continuum solvation model.22,23 was used
to mimic a water environment.22,23 As expected, no rel-
evant structural changes were observed between RdRp
relaxed model and the original 3-D structure.
BHK and MDBK were seeded in 96-well plates at a den-
sity of 5 · 104 and 3 · 104, and 2.5 · 104 cells/well,
respectively, and were allowed to form confluent mono-
layers by incubating overnight in growth medium at
37 ꢁC in a humidified CO2 (5%) atmosphere. Cell mon-
olayers were then infected with 50 lL of a proper virus
The model structures of all title compounds were built
and subjected to an initial energy minimization, again
using the Sander module of the Amber 7.0 suite of pro-
grams, with the parm94 version of the Amber force field.
In this case, the convergence criterion was set to

A. Carta et al. / Bioorg. Med. Chem. 15 (2007) 1914–1927
1925
10ꢀ4 kcal/(mol A). A conformational search was carried
˚
the Amber 7.0 suite using the quenched molecular
dynamics (QMD) method.36 In this case, 100 ps MD sim-
ulation at 310 K were employed to sample the conforma-
tional space of the substrate–enzyme complex in the GB/
SA continuum solvation environment.22,23 The integra-
tion step was equal to 1 fs. After each ps, the system
was cooled to 0 K, the structure was extensively mini-
mized, and stored. To prevent global conformational
changes of the enzyme, the backbone of the protein bind-
ing site was constrained by a harmonic force constant of
out using a well-validated, ad hoc developed combined
molecular mechanics/molecular dynamics simulated
annealing (MDSA) protocol.24–28 Accordingly, the re-
laxed structures were subjected to five repeated tempera-
ture cycles (from 310 to 1000 K and back) using constant
volume/constant temperature (NVT) MD conditions. At
the end of each annealing cycle, the structures were again
energy minimized to converge below 10ꢀ4 kcal/(mol A),
˚
and only the structures corresponding to the minimum
energy were used for further modeling. The atomic partial
charges for the geometrically optimized compounds were
obtained using the RESP procedure,29 and the electro-
static potentials were produced by single-point quantum
mechanical calculations at the Hartree–Fock level with
˚
100 kcal/A, whereas the amino acid side chains and the
ligands were allowed to move without any constraint.
The best energy configuration of each complex resulting
˚
from the previous step was allowed to relax in a 55-A
a 6-31G basis set, using the Merz–Singh–Kollman van
*
radius sphere of TIP3P water molecules.37 The resulting
system was minimized with a gradual decrease in the po-
sition restraints of the protein atoms. At the end of the
relaxation process, all water molecules beyond the first
der Waals parameters.30,31 Eventual missing force field
parameters for the phenantroline derivatives were gener-
ated as follows: AM132 geometry optimization of the
˚
structure was followed by RHF/6-31G single point cal-
hydration shell (i.e., at a distance of >3.5 A from any
*
culation to obtain the electrostatic potentials. Next, the
RESP method was used for charge fitting. The missing
bond, torsion angle or van der Waals parameters not
included in the parm94 were transferred33 from the newly
developed parm99 parameter set.34
protein atom) were removed. Finally, to achieve electro-
neutrality, a suitable number of counterions were added,
in the positions of largest electrostatic potential, as
determined by the module Cion within Amber 7.0. To
reduce computational time to reasonable limits, all pro-
˚
tein residues with any atom closer than 20 A from the
center of mass of each bound ligand were chosen to be
flexible in the dynamic simulations. Subsequently, a
The optimized structures of the test compounds were
docked into the HCV polymerase allosteric binding site
by applying a consolidated procedure;25,26,28 according-
ly, it will be described here only briefly. The software
AutoDock 3.017 was employed to estimate the possible
binding orientations of all compounds in the receptor.
In order to encase a reasonable region of the protein sur-
face and interior volume, centered on the crystallo-
˚
spherical TIP3P water cap of radius equal to 20 A was
centered on each inhibitor in the corresponding RdRp
complex, including the hydrating water molecules within
the sphere resulting from the previous step. After energy
minimization of the new water cap for 1500 steps, keep-
ing the protein, the ligand, and the pre-existing waters
rigid, followed by a MD equilibration of the entire water
sphere with fixed solute for 20 ps, further unfavorable
interactions within the structures were relieved by pro-
gressively smaller positional restraints on the solute
˚
graphically identified binding site, the grids were 60 A
˚
on each side. Grid spacing (0.375 A) and 120 grid points
were applied in each Cartesian direction so as to calcu-
late mass-centered grid maps. Amber 12–6 and 12–10
Lennard–Jones parameters were used in modeling van
der Waals interactions and hydrogen bonding (N–H,
O–H, and S–H), respectively. In the generation of the
electrostatic grid maps, the distance-dependent relative
permittivity of Mehler and Solmajer was applied.35
For the docking of each compound to the protein, three
hundred Monte Carlo/Simulated Annealing (MC/SA)
runs were performed, with 100 constant temperature cy-
cles for simulated annealing. For these calculations, the
GB/SA implicit water model22,23 was again used to mim-
ic the solvated environment. The rotation of the angles
/ and u, and the angles of side chains were set free dur-
ing the calculations. All other parameters of the MC/SA
algorithm were kept as default. Following the docking
procedure, the structures of all compounds were subject-
2
˚
(from 25 to 0 kcal/(mol A ) for a total of 4000 steps.
Each system was gradually heated to 310 K in three
intervals, allowing a 5-ps interval per 100 K, and then
equilibrated for 50-ps at 310 K, followed by 400-ps of
data collection runs, necessary for the estimation of
the free energy of binding (vide infra). The MD simula-
tions were performed at constant T = 310 K using the
Berendsen et al. coupling algorithm38 with separate cou-
pling of the solute and solvent to the heat, an integration
time step of 2-fs, and the applications of the Shake
algorithm39 to constrain all bonds to their equilibrium
values, thus removing high frequency vibrations.
Long-range nonbonded van der Waals interactions were
˚
truncated by using a dual cutoff of 9 and 13 A, respec-
tively, where energies and forces due to interactions
˚
˚
ed to cluster analysis with a tolerance of 1 A for an
between 9 and 13 A were updated every 20-time steps.
all-atom root-mean-square (RMS) deviation from a
lower-energy structure representing each cluster family.
In the absence of any relevant crystallographic informa-
tion, the structure of each resulting complex character-
ized by the lowest interaction energy in the prevailing
cluster was selected for further evaluation.
The particle mesh Ewald method40 was used to treat
the long-range electrostatics. For the calculation of the
binding free energy between the RdRp and each phe-
nantroline derivative in water, a total of 400 snapshots
were saved during the MD data collection period
described above, one snapshot per 1 ps of MD simulation.
Each best substrate/RdRp complex resulting from the
automated docking procedure was further refined in
The binding free energy DGbind of each RdRp/drug com-
plex in water was calculated according to the procedure

1926
A. Carta et al. / Bioorg. Med. Chem. 15 (2007) 1914–1927
termed Molecular Mechanic/Poisson–Boltzmann Sur-
face Area (MM/PBSA), and originally proposed by
Srinivasan et al.41 Since the theoretical background of
this methodology is described in details in the original
papers by Peter Kollman and his group,42 it will be only
briefly described below.
tively. Each of those structures was minimized, using a
distance-dependent dielectric constant e = 4r, to account
for solvent screening, and its entropy was calculated
using classical statistical formulas and normal mode-
analysis. To minimize the effects due to different confor-
mations adopted by individual snapshots we averaged
the estimation of entropy over 10 snapshots.
Basically, an MD simulation (typically in explicit sol-
vent) is first carried out which yields a representative
ensemble of structures. The average total free energy
of the system, G, is then evaluated as:
Finally, the IC50 values were calculated from the corre-
sponding binding free energies using the following
relationship:13,14
G ¼ GPB þ GNP þ EMM ꢀ TSsolute
ð1Þ
DGbind ¼ RT lnKdiss ¼ RT lnðIC50 þ 0:50Cenz
ffi RT ln IC50
Þ
in which GPB is the polar solvation energy component,
which is calculated in a continuum solvent, usually a fi-
nite-difference Poisson–Boltzmann (PB) model, and GNP
is the nonpolar contribution to the solvation energy,
which can be obtained from the solvent accessible sur-
face area (SA). EMM denotes the sum of molecular
mechanics (MM) energies of the molecule, and can be
further split into contributions from electrostatic (Eele)
and van der Waals (EvdW) energies:
ð4Þ
The overall quality of the entire procedure described
above, i.e., protein and inhibitors modeling, conforma-
tional analysis, docking, and energetic calculations—
was tested by carrying out the same calculations for
one proprietary inhibitor, Shire A, from Shire BioChem
Inc., characterized by a structure reminiscent of that
characterizing some of the title compounds, and for
which both the crystallographic structures of the relevant
HCV RdRp complexes (PBD entry codes: 1NHU) and
the corresponding IC50 values were available.7
EMM ¼ Eele þ EvdW
The last term in Eq. 1, TSsolute represents the solute
entropy, and is usually estimated by a combination of
classical statistical formulas and normal mode analysis.
ð2Þ
All modeling figures were prepared using Chimera.48
The binding free energy DGbind of a given noncovalent
association can then be obtained as:
References and notes
À
Á
DGbind ¼ Gcomplex ꢀ Gprotein þ Gligand
1. Tan, S. L.; Pause, A.; Shi, Y.; Sonenberg, N. Nat. Rev.
Drug. Discov. 2002, 1, 867.
2. Hayashi, P. H.; Di Bisceglie, A. M. Med. Clin. North Am.
2005, 89, 371.
¼ DEMM þ DGsol ꢀ T DS
ð3Þ
in which DGsol = DGPB + DGNP. The ensemble of struc-
tures for the uncomplexed reactants are generated either
running separate MD simulations for them, or by using
the trajectory of the complex, simply removing the
atoms of the protein or ligand. As reported below, in
this work we followed the successful approach proposed
by Kuhn and Kollman43 and applied the latter variant.
Accordingly, the EMM were calculated with the Anal and
Carnal modules from the AMBER 7.0 suite. The infinite
cutoffs for all interactions and the parm94 force field
parameters were applied.21 The calculations of the polar
solvation term GPB were done with the Delphi package,44
with interior and exterior dielectric constants equal to 1
3. Memon, M. I.; Memon, M. A. J. Viral Hepat. 2002, 9, 84.
4. Echevarria-Mayo, J. M. Enferm. Infecc. Microbiol. Clin.
2006, 24, 45.
5. Bosch, F. X.; Ribes, J.; Cleries, R.; Diaz, M. Clin. Liver
Dis. 2005, 9, 191.
6. Enterovirus surveillance: MMWR Morb. Mortal Wkly
Rep. 2006, 55, 153.
7. Wang, M.; Ng, K. K.-S.; Cherney, M. M.; Chan, L.;
Yannopoulos, C. G.; Bedard, J.; Morin, N.; Nguyen-Ba,
N.; Alaoui-Ismaili, M. H.; Bethell, R. C.; James, M. N. C.
J. Biol. Chem. 2003, 278, 9489.
8. Kaufmann, A.; Radosevic, R. Chem. Ber. 1909, 42, 2613.
9. (a) Albert, A.; Brown, D. J.; Heinz, D. J. Chem. Soc. 1948,
1285; (b) Douglas, B.; Kermack, W. O. J. Chem. Soc.
1949, 1017; (c) Sykes, W. O. J. Chem. Soc. 1953, 3543.
10. Sanna, P.; Carta, A.; Paglietti, G. Heterocycles 2000, 53,
423.
11. Carta, A.; Boatto, G.; Paglietti, G.; Poni, G.; Setzu, M.
G.; Caredda, P. Heterocycles 2003, 60, 833.
12. Pricl, S.; Manfredini, S.; Angusti, A.; Ciliberti, N.; Durini,
E.; Vertuani, S.; Buzzoni, L.; Coslanich, A.; Ferrone, M.;
Fermeglia, M.; Paneni, M. S.; La Colla, P.; Sanna, G.;
Cadeddu, A.; Mura, M.; Loddo, R. In Framing the
Knowledge of Therapeutics for Viral Hepatitis; Schinazi, R.
F., Schiff, E. R., Eds.; IHL Press: Arlington, MA, USA,
2006; p 279.
˚
and 80, respectively. A grid spacing of 2/A, extending
20% beyond the dimensions of the solute, was employed.
The nonpolar component GNP was obtained using the
following relationship:45 GNP = cSA + b, in which
2
˚
c = 0.00542 kcal/(mol A ), b = 0.92 kcal/mol, and the
surface area was estimated by means of the MSMS soft-
ware.46 The last parameter in Eq. 1, i.e., the change in
solute entropy upon association—TSsolute, was calculat-
ed through normal-mode analysis.47 In the first step of
˚
this calculation, an 8-A sphere around the ligand was
cut out from an MD snapshot for each ligand–protein
complex. This value was shown to be large enough to
yield converged mean changes in solute entropy. On
the basis of the size-reduced snapshots, of the complex,
we generated structures of the uncomplexed reactants by
removing the atoms of the protein and ligand, respec-
13. Kroeger Smith, M. B.; Lamb, M. L.; Tirado-Rives, J.;
Jorgensen, W. L.; Michejda, C. J.; Ruby, S. K.; Smith, R.
H., Jr. Protein Eng. 2000, 13, 413.
14. Wang, J.; Morin, P.; Wang, W.; Kollman, P. A. J. Am.
Chem. Soc. 2001, 123, 5221.

A. Carta et al. / Bioorg. Med. Chem. 15 (2007) 1914–1927
1927
15. Miyamoto, S.; Kollman, P. A.. Proc. Natl. Acad. Sci.
U.S.A. 1993, 90, 8402.
29. Bayly, C. I.; Cieplak, P.; Cornell, W. D.; Kollman, P. A.
J. Phys. Chem. 1993, 97, 10269.
16. Pawels, R.; Balzarini, J.; Baba, M.; Snoeck, R.; Schols, D.;
Herdewijn, P.; Desmyster, J.; De Clercq, E. J. Virol.
Methods 1998, 20, 309.
17. Morris, G. M.; Goodsell, D. S.; Halliday, R. S.; Huey, R.;
Hart, W. E.; Belew, R. K.; Olson, A. J. J. Comput. Chem.
1998, 19, 1639.
30. Singh, U. C.; Kollman, P. A. J. Comput. Chem. 1984, 5,
129.
31. Besler, B. H.; Merz, K. M.; Kollman, P. A. J. Comput.
Chem. 1990, 11, 431.
32. Dewar, M. J. S.; Zoebisch, E. G.; Healy, E. F.; Stewart, J.
J. P. J. Am. Chem. Soc. 1985, 107, 3902.
18. Case, D. A.; Pearlman, D. A.; Caldwell, J. W.; Cheatham,
T. E., III; Wang, J.; Ross, W. S.; Simmerling, C. L.;
Darden, T. A.; Merz, K. M.; Stanton, R. V.; Cheng, A. L.;
Vincent, J. J.; Crowley, M.; Tsui, V.; Gohlke, H.; Radmer,
R. J.; Duan, Y.; Pitera, J.; Massova, I.; Seibel, G. L.; Singh,
U. C.; Weiner, P. K.; Kollman, P. A. AMBER 7;
University of California: San Francisco, CA, USA, 2002.
19. Pearlman, D. A.; Case, D. A.; Caldwell, J. W.; Ross, W. S.;
Cheatham, T. E., III; DeBolt, S.; Ferguson, D.; Seibel, G.
L.; Kollman, P. A. Comp. Phys. Commun. 1995, 91, 1.
20. Bressanelli, S.; Tomei, L.; Rey, F. A.; De Francesco, R. J.
Virol. 2002, 76, 3482.
21. Cornell, W. D.; Cieplak, P.; Bayly, C. I.; Gould, I. R.;
Merz, K. M.; Ferguson, D. M.; Spellmeyer, D. C.; Fox,
T.; Caldwell, J. W.; Kollman, P. A. J. Am. Chem. Soc.
1995, 117, 5179.
22. Onufriev, A.; Bashford, D.; Case, D. A. J. Phys. Chem. B
2000, 104, 3712.
23. Feig, M.; Onufriev, A.; Lee, M.; Im, W.; Case, D. A.;
Brooks, C. L., III J. Comput. Chem. 2004, 25, 265.
24. Fermeglia, M.; Ferrone, M.; Pricl, S. Bioorg. Med. Chem.
2002, 10, 2471.
25. Felluga, F.; Pitacco, G.; Valentin, E.; Coslanich, A.;
Fermeglia, M.; Ferrone, M.; Pricl, S. Tetrahedron: Asym-
metry 2003, 14, 3385.
26. Manfredini, S.; Solaroli, N.; Angusti, A.; Nalin, F.;
Durini, E.; Vertuani, S.; Pricl, S.; Ferrone, M.; Spadari,
S.; Focher, F.; Verri, A.; De Clercq, E.; Balzarini, J.
Antiviral Chem. Chemother. 2003, 14, 183.
27. Metullio, L.; Coslanich, A.; Fermeglia, M.; Ferrone, M.;
Fuchs, S.; Paneni, M. S.; Pricl, S. Biomacromolecules 2004,
5, 1371.
33. Huo, S.; Wang, J.; Cieplak, P.; Kollman, P. A.; Kuntz, I.
D. J. Med. Chem. 2002, 45, 1412.
34. Wang, J.; Cieplak, P.; Kollman, P. A. J. Comput. Chem.
2000, 21, 1049.
35. Mehler, E. L.; Solmajer, T. Protein Eng. 1991, 4, 903.
36. Frecer, V.; Kabelac, M.; De Nardi, P.; Pricl, S.; Miertus,
S. J. Mol. Graph. Model. 2004, 22, 209.
37. Jorgensen, W. L.; Chandrasekhar, J.; Madura, J. D.;
Impey, R. W.; Klein, M. L. J. Chem. Phys. 1983, 79,
926.
38. Berendsen, H. J. C.; Postma, J. P. M.; van Gunsteren, W.
F.; DiNola, A.; Haak, J. R. J. Chem. Phys 1984, 81,
3684.
39. Ryckaert, J. P.; Ciccotti, G.; Berendsen, H. J. C.
J. Comput. Phys. 1977, 23, 327.
40. Darden, T.; York, D.; Pedersen, L. J. Chem. Phys. 1993,
98, 10089.
41. Srinivasan, J.; Cheatham, T. E.; Cieplak, P.; Kollman, P.
A.; Case, D. A. J. Am. Chem. Soc. 1998, 120, 9401.
42. Kollman, P. A.; Massova, I.; Reyes, C.; Kuhn, B.; Huo,
S.; Chong, L.; Lee, M.; Lee, T.; Duan, Y.; Wang, W.;
Donini, O.; Cieplak, P.; Srinivasan, J.; Case, D.; Chea-
tham, T. E., III Acc. Chem. Res. 2000, 3, 889.
43. Kuhn, B.; Kollman, P. A. J. Med. Chem. 2000, 43, 3786.
44. Gilson, M. K.; Sharp, K. A.; Honig, B. H. J. Comput.
Chem. 1988, 9, 327.
45. Sitkoff, D.; Sharp, K. A.; Honig, B. H. J. Phys. Chem.
1994, 98, 1978.
46. Sanner, M. F.; Olson, A. J.; Spehner, J. C. Biopolymers
1996, 38, 305.
47. Wilson, E. B.; Decius, J. C.; Cross, P. C. Molecular
Vibrations; McGraw-Hill: New York, NY, 1995.
48. Pettersen, E. F.; Goddard, T. D.; Huang, C. C.; Couch, G.
S.; Greenblatt, D. M.; Meng, E. C.; Ferrin, T. E.
J. Comput. Chem. 2004, 25, 1605.
28. Mamolo, M. G.; Zampieri, D.; Vio, L.; Fermeglia, M.;
Ferrone, M.; Pricl, S.; Scialino, G.; Banfi, E. Bioorg. Med.
Chem. 2005, 13, 3797.