Investigation of the salicylaldehyde thiosemicarbazone scaffold for inhibition of influenza virus PA endonuclease

Article abstract of DOI:10.1007/s00775-015-1292-0

The influenza virus PA endonuclease is an attractive target for the development of novel anti-influenza virus therapeutics, which are urgently needed because of the emergence of drug-resistant viral strains. Reported PA inhibitors are assumed to chelate t

Full text of DOI:10.1007/s00775-015-1292-0

J Biol Inorg Chem
DOI 10.1007/s00775-015-1292-0
ORIGINAL PAPER
Investigation of the salicylaldehyde thiosemicarbazone scaffold
for inhibition of influenza virus PA endonuclease
1
1
2
1
Dominga Rogolino · Alessia Bacchi · Laura De Luca · Gabriele Rispoli ·
3
4
4
1
Mario Sechi · Annelies Stevaert · Lieve Naesens · Mauro Carcelli
Received: 18 February 2015 / Accepted: 4 August 2015
©
SBIC 2015
Abstract The influenza virus PA endonuclease is an
attractive target for the development of novel anti-influenza
virus therapeutics, which are urgently needed because of
the emergence of drug-resistant viral strains. Reported PA
inhibitors are assumed to chelate the divalent metal ion(s)
involved in the endonuclease activity of the heterotrimeric
influenza polymerase complex. The crystal structure of the
magnesium complex of the o-vanillin thiosemicarbazone
ligand 1 is also described. Moreover, docking studies of
PA endonuclease with compounds 1 and 2 were performed,
to further analyse the possible mechanism of action of this
class of inhibitors.
2
+ 2+
(
Mg or Mn ) in the enzyme’s catalytic site, which is
located in the N-terminal part of PA (PA-Nter). In the
present work, a series of salicylaldehyde thiosemicarba-
zone derivatives have been synthesized and evaluated for
their ability to inhibit the PA-Nter catalytic activity. Com-
pounds 1–6 have been evaluated against influenza virus,
both in enzymatic assays with influenza virus PA-Nter and
in virus yield assays in MDCK cells. In order to establish
a structure–activity relationship, the hydrazone analogue
of the most active thiosemicarbazone has also been evalu-
ated. Since chelation may represent a mode of action of
such class of molecules, we studied the interaction of two
of them, one with and one without biological activity ver-
Keywords Influenza virus · Thiosemicarbazone ·
Endonuclease · Metal chelation · Antiviral
Introduction
Currently available antiviral drugs to treat or prevent influ-
enza infections comprise two classes of agents: those tar-
geting the viral M2 ion-channel (amantadine and riman-
tadine) and those targeting the viral neuraminidase (NA)
(zanamivir and oseltamivir). The M2 inhibitors have lim-
ited clinical utility because of worldwide spread of drug-
resistant mutant viruses among almost all influenza A virus
subtypes, including the 2009 pandemic H1N1 virus [1–3].
Resistance is also a growing concern for oseltamivir [4].
Therefore, there is an urgent need for entirely novel antivi-
ral compounds with a different mode of action and target-
ing a critical step in the viral replication process [5]. The
influenza virus genome consists of eight single-stranded
2
+
sus the PA enzyme, towards Mg , the ion that is probably
Electronic supplementary material The online version of this
article (doi:10.1007/s00775-015-1292-0) contains supplementary
material, which is available to authorized users.
*
Mauro Carcelli
mauro.carcelli@unipr.it
1
2
Dipartimento di Chimica, Università di Parma, Parco Area
delle Scienze 17/A, 43124 Parma, Italy
(
−)RNA segments, which are transcribed and replicated by
the viral RNA-dependent RNA polymerase (RdRp) [6, 7].
This RdRp is widely recognized as a superior target for the
development of new antivirals, since it is highly conserved
among influenza A, B and C viruses, and its functions are
essential for viral genome replication [8]. It is composed of
three subunits: PA, PB1 and PB2. The endonuclease activ-
ity, which resides in the N-terminal part of PA (PA-Nter) [9,
Dipartimento di Scienze del Farmaco e Prodotti per la Salute,
Università di Messina, Polo Universitario SS. Annunziata,
9
8158 Messina, Italy
3
4
Dipartimento di Chimica e Farmacia, Università di Sassari,
Via Vienna 2, 07100 Sassari, Italy
Rega Institute for Medical Research, KU Leuven,
3
000 Louvain, Belgium
1
3

J Biol Inorg Chem
O
OH
OH
Ile120
N
N
O
N
O
O
Glu119
O
N
N
O
O
N
O
O
OH
Asp108
Cl
O
O
M2+
L-742,001
Flutimide
O
N
M2+
H
H
O
N
N
O
N
N
N
N
N
OH
His41
O
N
N
Glu80
2
-(4-(1H-tetrazol-5-yl)phenyl)-
-hydroxypyrimidin-4(3H)-one
5
Fig. 1 Schematic representation of the metal-interacting residues
within the catalytic site of influenza virus PA endonuclease, assuming
the presence of two metal ions (see Ref. [12, 13])
Fig. 2 Chemical structures of some prototype inhibitors of influenza
virus endonuclease [19, 21, 24]
1
0], is required to cleave host cell pre-mRNAs and produce
the 5′-capped primers for transcription of the viral genomic
RNA into mRNA (cap-snatching) [11]. Cap-snatching is
an important event in the life cycle of all members of the
Orthomyxoviridae family of viruses, including influenza
A, B and C viruses. The host cell has no analogous activ-
ity; therefore, inhibitors of cap-snatching may be active
against all influenza virus (sub)types and strains, including
oseltamivir-resistant influenza viruses, without interfering
with normal host cell functions.
2-hydroxyphenyl amide derivatives [22] as well as tetramic
acids [23] and 5-hydroxy pyrimidin-4-one derivatives [24]
(Fig. 2).
Thiosemicarbazones (TSCs) possess a broad range of
biological properties including antitumor, antimalarial and
antimicrobial activity [25]; moreover, they have shown
good antiviral activity against herpes simplex virus [26],
vaccinia and cowpox virus [27], as well as HIV [28, 29].
The biological properties of TSCs are often related to
chelation of metal ions [30] and, therefore, they could be
good candidates as chelating inhibitors of influenza virus
endonuclease. TSCs can coordinate to the metal centre
in an N,S-bidentate mode, but when an additional coordi-
nating group is present, more diversified binding modes
become possible [31]. In particular, the presence of the OH
group in salicylaldehyde thiosemicarbazone derivatives
might provide a more favourable coordination for hard
Several X-ray crystallographic studies have been per-
formed on PA-Nter, revealing the presence of one [10, 12]
or two [9, 13, 14] divalent metal ions within the active site
(
Fig. 1). The two-metal-ion model is also consistent with
biochemical [15] and computational [16] findings. The ions
identified by means of X-ray diffraction are magnesium(II)
or manganese(II), depending on the crystallization condi-
tions. Considering the relative abundance of these two
2
+
2+
metal ions in the cell ([Mg ] is about 1000 times higher
metal ions like Mg that prefers oxygen donor atoms.
2
+
than [Mn ]), magnesium may be more biologically
relevant.
In the present work, we report the synthesis of some
salicylaldehyde thiosemicarbazone derivatives (1–6, Fig. 3)
and their biological evaluation against influenza virus, both
in enzymatic assays with influenza virus PA-Nter and in
virus yield assays in MDCK cells. In order to investigate
the role of the thiosemicarbazone moiety on activity, we
also synthesized and tested compound 7, the hydrazone
analogue of 2, the most active molecule in the thiosemicar-
bazone series.
Metal chelation has emerged as an efficient strat-
egy to develop new inhibitors of metal-dependent viral
enzymes, the most illustrious example being the class of
HIV integrase inhibitors, some of which have already been
approved [17]. In the same way, development of PA-bind-
ing agents with metal-chelating properties may represent
a successful strategy to tackle influenza infections. Sev-
eral chelating molecules have been identified as influenza
endonuclease inhibitors, including 2,4-dioxobutanoic acid
derivatives [18–20], 5-hydroxy-1,6-dihydropyrimidine-
To assess the role of metal chelation in their antivi-
ral mode of action, we have selected 1 and 2 (1 without
activity and 2 with reasonable activity against the PA-Nter
2
+
4
-carboxylic acids [12], flutimide and its derivatives [21],
enzyme) and studied their interactions with Mg , the ion
1
3

J Biol Inorg Chem
Fig. 3 Chemical structures of
ligands 1–7 and the coordina-
tion modes a and b of 1–5,
involving the OH groups
COOH
R
OH
OH
1
R = 3-methoxy (1, H L )
2
2
OH
R = 3-hydroxy (2, H L )
2
R = 4-hydroxy (3)
R = 5-hydroxy (4)
R = 5-methoxy (5)
HC
H
HC
H
N
N
N
N
N
S
S
NH
O
NH2
NH2
NH2
(6)
(7)
O
O
O
O
M
HC
H
HC
M
N
N
N
S
N
S
H
NH2
NH2
a
b
that is probably involved in the endonuclease activity of the
native influenza virus RdRp complex. The crystal structure
Synthesis of the ligands
1
1
of complex Mg(HL ) 2CH OH is also described (1, H L ).
The TSCs 1–6 and the semicarbazone 7 were prepared fol-
lowing reported procedures [32]. Briefly, to a solution of
the aldehyde in absolute ethanol, an equimolar amount of
thiosemicarbazide was added, dissolved in the same sol-
vent. In the synthesis of 7, semicarbazide hydrochloride
was used, which was suspended in ethanol and the pH
adjusted to 7 using KOH 1 M. The mixture was refluxed for
2
3
2
Finally, docking studies on compounds 1 and 2 in complex
with PA-Nter endonuclease were performed, to better elu-
cidate the possible binding mode of these inhibitors.
Materials and methods
Chemistry
6
h, cooled at room temperature and concentrated in vac-
uum. The resulting precipitate was filtered off, washed with
cold ethanol and dried in vacuum.
All reagents of commercial quality were used without fur-
ther purification. Purity of compounds was determined by
elemental analysis and verified to be ≥95 % for all syn-
thesized molecules. NMR spectra were recorded at 25 °C
on a Bruker Avance 400 FT spectrophotometer. The ATR-
IR spectra were recorded by means of a Nicolet-Nexus
N‑(2‑Hydroxy‑)‑3‑methoxybenzylidenethiosemicarbazide
1
1
(1, H L ) Yield = 71 %. H-NMR (DMSO-d , 25 °C), δ:
2
6
3.80 (s, 3H, OCH ), 6.77 (t, 1H; J = 7.9 Hz, ArH), 6.96
3
(d, 1H, J = 7.8 Hz, ArH), 7.52 (d, 1H, J = 7.8 Hz, ArH),
7.87 (s, br, 1H, NH), 8.09 (s, br, 1H, NH), 8.40 (s, 1H;
1
(
Thermo Fisher) spectrophotometer using a diamond crys-
HC=N), 9.20 (s, br, 1H, OH), 11.39 (s, br, 1H; NH). H-
−
1
tal plate in the range of 4000–400 cm . Elemental analy-
ses were performed using a FlashEA 1112 series CHNS/O
analyzer (Thermo Fisher) with gas-chromatographic sepa-
ration. Electrospray mass spectral analyses (ESI–MS) were
performed with an electrospray ionization (ESI) time-of-
flight Micromass 4LCZ spectrometer. MS spectra were
acquired in positive EI mode by means of a DEP-probe
NMR (MeOD-d , 25 °C), δ: 3.90 (s, 3H, OCH ), 6.84 (t, 1H;
4
3
J = 7.9 Hz, ArH), 6.99 (d, 1H, J = 7.8 Hz, ArH), 7.43 (d,
1H, J = 7.8 Hz, ArH), 8.41 (s, 1H; HC=N). MS (EI, 70 eV),
+
−1
m/z (%) = 225.0 ([M] , 100); IR (cm ): ν = 3459,
NH
3339; ν = 3158 (br); ν
= 1596; ν
= 1532, 821;
OH
C=N
C=S
11
ν
= 1258, 1056. Anal. Calcd. for C H N O S·0.5H O:
OCH3
9
3
2
2
C 46.14; H 5.16; N 17.94. Found: C 45.92, H 4.82, N 18.24.
(
Direct Exposure Probe) mounting on the tip a Re-filament
2
with a DSQII Thermo Fisher apparatus, equipped with a
single quadrupole analyzer.
N‑(2,3‑Dihydroxybenzylidene)‑thiosemicarbazide (2, H L )
2
1
Yield = 75 %. H-NMR (DMSO-d , 25 °C), δ: 6.63 (t, 1H;
6
1
3

J Biol Inorg Chem
J = 7.9 Hz, ArH), 6.79 (d, 1H, J = 7.8 Hz, ArH), 7.35 (d,
(s, br, 2H, NH ), 6.64 (t, 1H; J = 7.9 Hz, ArH), 6.76 (d, 1H,
2
1
1
H, J = 7.5 Hz, ArH), 7.86 (s, br, 1H, NH), 8.08 (s, br,
J = 7.8 Hz, ArH), 7.17 (d, 1H, J = 7.5 Hz, ArH), 8.14 (s,
H, NH), 8.36 (s, 1H; HC=N), 9.20 (s, vbr, 2H, OH), 11.37
1H; HC=N), 9.21, 9.39 (s, br, 2H, OH), 10.17 (s, br, 1H;
1
1
(
s, br, 1H; NH). H-NMR (MeOD-d , 25 °C), δ: 6.75 (t,
NH). H-NMR (MeOD-d , 25 °C), δ: 6.75 (t, J = 7.9 Hz,
4
4
J = 7.9 Hz, 1H), 6.86 (dd, J = 7.9, 1.6 Hz, 1H), 7.14 (d,
1H), 6.86 (dd, J = 7.9, 1.6 Hz, 1H), 7.14 (d, J = 7.9 Hz,
1H), 8.30 (s, 1H, HC=N). MS (EI, 70 eV), m/z (%) = 195.0
J = 7.9 Hz, 1H), 8.30 (s, 1H, HC=N). MS (EI, 70 eV), m/z
+
−1
+
−1
(
ν
%) = 210.9 ([M] , 100); IR (cm ): ν = 3364, 3256;
([M] , 100); IR (cm ): ν = 3455, 3350; ν = 3166 (br);
NH
NH
OH
= 3173 (br); ν
= 1620; ν
= 1539, 825. Anal.
ν
= 1694; ν
= 1592. Anal. Calcd. for C H N O : C
OH
C=N
C=S
C=O
C=N
8
9
3
3
Calcd. for C H N O S·0.5H O: C 43.63; H 4.58; N 19.08.
49.23; H 4.65; N 21.53. Found: C 49.16, H 4.48, N 21.40.
8
9
3
2
2
Found: C 43.33, H 4.48, N 18.98.
1
1
Mg(HL ) 2.5H O (8) 0.5 mmol of H L was dissolved in
2
2
2
N‑(2,4‑Dihydroxybenzylidene)‑thiosemicarbazide
20 ml of methanol and 4 eq. of NEt was added. The solu-
3
1
(
6
3) Yield = 78 %. H-NMR (DMSO-d , 25 °C), δ: 6.25–
tion turned yellow and it was stirred at r.t. for 30 min; 0.5 eq.
of magnesium acetate was added and the reaction mixture
was stirred at r.t. for 4 h, concentrated in vacuum and cooled
overnight. The precipitate was filtered off and washed with
water. Crystals suitable for X-ray diffraction analysis were
obtained by slow evaporation of the mother liquors at room
6
.29 (m, 2H; ArH), 7.67 (d, 1H, J = 8.4 Hz, ArH), 7.74 (s,
br, 1H, NH), 7.94 (s, br, 1H, NH), 8.24 (s, 1H; HC=N), 9.78
(
(
s, br, 2H, OH), 11.17 (s, br, 1H; NH). MS (EI, 70 eV), m/z
+
−1
%) = 211.0 ([M] , 100); IR (cm ): ν = 3479, 3347;
NH
ν
= 3262, 3128 (br); ν
= 1630; ν
= 1582, 875.
C=S
OH
C=N
1
Anal. Calcd. for C H N O S 1/3H O: C 44.23; H 4.48; N
temperature. Yield: 74 %. H-NMR (MeOD, 25 °C), δ:
8
9
3
2
2
1
9.34. Found: C 43.94, H 4.05, N 19.18.
3.89 (s, 3H, OCH ); 6.84 (t, br, 1H, ArH), 6.98 (d, br, 1H,
3
ArH), 7.42 (d, br, 1H, ArH), 8.40 (s, 1H, HC=N). MS–ESI,
+
+
N‑(2,5‑Dihydroxybenzylidene)‑thiosemicarbazide
m/z (+, %) = 473 ([M + H] , 40); 495 ([M + Na] , 40).
1
−1
(
(
(
4) Yield = 85 %. H-NMR (DMSO-d , 25 °C), δ: 6.68
IR (cm ): ν = 3334 (br); ν
= 1596; ν
= 1534;
6
NH
C=N
C=S
18 20
m, 2H; ArH), 7.22 (s, 1H, ArH), 7.78 (s, br, 1H, NH), 8.07
ν
= 1238, 1027. Anal. Calcd. for C H N O Mg
OCH3
6
4
s, br, 1H, NH), 8.30 (s, 1H; HC=N), 8.83 (s, br, 2H, OH),
2.5H O: C 41.75, H 4.87, N 16.23, S 12.38. Found: C 41.30,
2
9
.21 (s, br, 2H, OH), 11.33 (s, br, 1H; NH). MS (EI, 70 eV),
H 4.75, N 16.12, S 12.97.
+
−1
m/z (%) = 211.0 ([M] , 100); IR (cm ): ν = 3427, 3328;
NH
2
2
ν
= 3048, 3002 (br); ν
= 1623; ν
= 1559, 942.
Mg(HL ) (9) 0.5 mmol of H L was dissolved in 20 ml
2 2
OH
C=N
C=S
Anal. Calcd. for C H N O S 1/3H O: C 44.23; H 4.48; N
of degassed methanol under nitrogen and 4 eq. of NEt was
8
9
3
2
2
3
1
9.34. Found: C 44.11, H 4.39, N 19.18.
added. The solution turned yellow and it was stirred at r.t.
for 30 min; 0.5 eq. of magnesium acetate was added and the
reaction mixture was stirred at r.t. for 4 h, concentrated in
vacuum and cooled overnight. The precipitate was filtered
off under inert atmosphere, washed with water and kept
N‑(2‑Hydroxy‑5‑methoxybenzylidene)‑thiosemicarbazide
1
(
6
5) Yield = 63 %. H-NMR (DMSO-d , 25 °C), δ:
6
.76–6.84 (m, 2H; ArH), 7.48 (s, 1H, ArH), 8.01 (s, br, 1H,
NH), 8.13 (s, br, 1H, NH), 8.35 (s, 1H; HC=N), 9.44 (s,
under nitrogen. Both the solution and the solid turn dark if
1
br, 2H, OH), 11.36 (s, br, 1H; NH). MS (EI, 70 eV), m/z
exposed to air. H-NMR (MeOD-d , 25 °C), δ: 6.47 (s, br,
4
+
−1
(
ν
%) = 225.0 ([M] , 100); IR (cm ): ν = 3412, 3304;
2H), 6.81 (s, br, 1H), 8.12 (s, br, 1H, HC=N). MS–ESI, m/z
NH
−
2 −
−1
= 3120 (br); ν
= 1626; ν = 1537, 821. Anal.
C=S
(−, %) = 443 ([M] , 10); 210 ([HL ] , 100). IR (cm ):
OH
C=N
Calcd. for C H N O S: C 47.99; H 4.92; N 18.65. Found:
ν
= 3275, 3164 (br); ν
= 1568; ν
= 1534.
9
11
3
2
NH+OH
C=N
C=S
C 48.17, H 4.82, N 18.34.
X‑ray crystallography
N‑(3‑(1,1′‑Biphenyl)‑4‑carboxylic
acid)‑thiosemicar‑
1
1
bazide (6) Yield = 51 %. H-NMR (DMSO-d , 25 °C),
Single crystals of 2 and of Mg(HL ) ·2CH OH were
6
2
3
δ: 7.53 (t, J = 8 Hz, 1H; ArH), 7.78 (m, 2H, ArH), 7.89 (d,
J = 8 Hz, 2H, ArH), 8.03 (d, 2H, ArH), 8.13–8.26 (m, 4H,
selected and mounted on glass fibres to collect data on a
SMART Breeze diffractometer. The crystals were kept at
293 K during data collection. Table 1 reports crystal data
and structure analysis. Using Olex2 [33], the structure
was solved with the SIR2004 [34] structure solution pro-
gram using Direct Methods and refined with the ShelXL
refinement package [35] using least squares minimisation.
Anisotropic displacement parameters were refined for all
non-hydrogen atoms. Hydrogen atoms were partly located
on the Fourier difference map and partly introduced in
ArH + HC=N + NH ), 11.51 (s, br, 1H, NH), 13.00 (s,
2
+
br, 1H, OH). MS (EI, 70 eV), m/z (%) = 298.9 ([M] , 60);
−
1
IR (cm ): ν
= 3432, 3263, 3143 (br); ν
= 1606;
C=N
NH+OH
ν
= 1535. Anal. Calcd. for C H N O S·1.5H O: C
NH2
21 16
2
6
2
5
5.21; H 4.94; N 12.88. Found: C 55.38, H 4.85, N 13.03.
N‑(2,3‑Dihydroxybenzylidene)‑semicarbazide
1
(
7) Yield = 70 %. H-NMR (DMSO-d , 25 °C), δ: 6.36
6
1
3

J Biol Inorg Chem
Table 1 Crystal data and structure refinement for 2 and Mg(HL ) ·2CH OH
1
2
3
1
Mg(HL ) ·2MeOH
2
2
Empirical formula
C H MgN O S
C H N O S
2
0
28
6
6
2
10 15
3
3
Formula weight
536.91
293(2)
257.31
Temperature/K
293
Crystal system
Orthorhombic
P2nn
Monoclinic
Space group
P2 /c
1
a/Å
7.878(1)
10.639(2)
31.921(5)
90
9.3259(6)
b/Å
13.5879(9)
c/Å
10.1652(6)
α/°
90
β/°
90
101.2117(9)
γ/°
90
90
Volume/Å3
2675.4(7)
4
1263.5(1)
Z
4
ρ
g/cm3
1.333
1.353
calc
μ, mm 1
−
0.267
0.257
F(000)
1128.0
544.0
Radiation
MoKα (λ = 0.71073)
2.552–44.106
21,724
MoKα (λ = 0.71073)
4.452–63.216
2
Θ range for data collection/°
Reflections collected
19,770
Independent reflections
Data/restraints/parameters
Goodness-of-fit on F2
3296 [Rint = 0.0818, Rsigma = 0.0481]
3296/7/326
4033 [Rint = 0.0263, Rsigma = 0.0184]
4033/0/174
1.073
1.031
Final R indexes [I ≥ 2σ (I)]
Final R indexes [all data]
R = 0.0702, wR = 0.1916
R = 0.0442, wR = 0.1267
1
2
1
2
R = 0.0855, wR = 0.2057
R = 0.0492, wR = 0.1311
1
2
1
2
−
3
Largest ΔF max/min/e Å
1.09/−0.38
0.43(7)
0.81/−0.71
Flack parameter
calculated positions riding on their carrier atoms. Hydro-
gen bonds were analysed with PARST97 [36] and the Cam-
bridge Structural Database software [37, 38] was used for
the analysis of the crystal packing. Table 1 summarizes
crystal data and structure determination results. Crys-
tallographic data (excluding structure factors) for 2 and
bacterial expression plasmid was transformed into E.
coli BL21-CodonPlus cells (Agilent Technologies, Santa
Clara, CA). These bacteria were grown to an OD of 0.6,
when IPTG was added at a final concentration of 1 mM to
induce expression of recombinant protein for 5 h at 37 °C.
The bacterial cells were ruptured using a French press and
the protein was purified by 6xHis-Ni–NTA chromatogra-
phy (Qiagen, Valencia, CA), followed by buffer exchange
using PD-10 desalting columns (GE Healthcare, Diegem,
Belgium) to keep the protein in storage buffer (50 mM
Tris–HCl pH 8, 100 mM NaCl, 10 mM β-mercaptoethanol,
50 % glycerol). Protein purity was verified by SDS-PAGE
with Coomassie Blue staining, and protein concentration
was determined by Bradford assay. Finally, the purified
protein was divided in aliquots and stored at −80 °C.
1
Mg(HL ) ·2CH OH have been deposited with the Cam-
2
3
bridge Crystallographic Data Centre as supplementary pub-
lication no. CCDC 1049845–1049846. Copies of the data
can be obtained free of charge on application to CCDC, 12
Union Road, Cambridge CB2 1EZ, UK (fax: (+44) 1223-
3
36-033; e-mail: deposit@ccdc.cam.ac.uk).
Biology
Production of recombinant influenza virus PA‑Nter protein
Plasmid‑based endonuclease assay
The coding sequence for PA-Nter (i.e. residues 1–217
from the PA protein of influenza virus strain A/X-31) was
cloned in the pET28a(+) plasmid (Merck KGaA, Darm-
stadt, Germany) with an N-terminal 6xHis-tag, and this
The enzymatic PA-Nter assay was performed according
to a published method [22] with minor modifications [20].
One microgram of recombinant PA-Nter was incubated
1
3

J Biol Inorg Chem
with 1 μg (16.7 nM) of single-stranded circular DNA plas-
mid M13mp18 (Bayou Biolabs, Metairie, Louisiana) in
the presence of the test compounds and at a final volume
of 25 μL. The assay buffer contained 50 mM Tris–HCl pH
concentration (EC ) was defined as the compound con-
5
0
centration causing 50 % reduction in the vRNP-driven
firefly luciferase signal, as compared to cells receiving
medium instead of compound. These EC50 values were
calculated by interpolation assuming a semi-log dose–
response effect. In parallel, the compound cytotoxicity,
expressed as CC50 (50 % cytotoxic concentration) was
determined using the spectrophotometric MTS cell viabil-
ity assay (CellTiter 96 AQueous One Solution Cell Prolif-
eration Assay; Promega). The CC50 values were defined
as the compound concentration reducing cell viability in
untransfected HEK293T cells by 50 %, as compared to the
wells receiving medium instead of compound. Ribavirin
was included as the reference compound.
8
, 100 mM NaCl, 10 mM β-mercaptoethanol and 1 mM
MnCl . The reaction was incubated at 37 °C for 2 h and
2
then stopped by heat inactivation (80 °C, 20 min). The
endonucleolytic digestion of the plasmid was visualized
by gel electrophoresis on a 1 % agarose gel with ethid-
ium bromide staining, and the amount of remaining intact
plasmid was quantified by ImageQuant TL software (GE
Healthcare).
The percentage inhibition of endonuclease activity was
plotted against the compound concentration on a semi-log-
arithmic plot, using GraphPad Prism software (GraphPad
Software, La Jolla, CA). Values were the mean ± S.E.M.
of three independent experiments. The 50 % inhibitory
concentrations (IC50) were obtained by nonlinear least
squares regression analysis. The known PA-Nter inhibitor
Virus yield assay
To determine anti-influenza virus activity in cell culture,
a virus yield assay was performed [40]. One day prior to
infection, MDCK cells were seeded into 96-well plates at
25,000 cells per well. At day 0, the test compounds were
added at serial dilutions, immediately followed by infec-
tion with influenza A/PR/8/34 virus. The multiplicity of
infection (MOI) was 150 CCID50 per well (50 % cell cul-
ture infectious dose; determined by the method of Reed and
Muench [42]). After 24 h of incubation at 35 °C, the super-
natants were harvested and stored at −80 °C. The viral copy
number in these samples was estimated by a one-step quan-
titative real-time reverse transcription (qRT)-PCR assay
(CellsDirect One-Step qRT-PCR kit; Invitrogen, Life Tech-
nologies, Gent, Belgium), with influenza virus M1-specific
primers and probe (see [43] for all details). An M1-plasmid
standard was included to allow absolute quantification.
The EC99 and EC90 values were calculated by interpolation
from data of at least three experiments and defined as the
compound concentration causing, respectively, a 2-log10
and 1-log10 reduction in viral RNA (vRNA) copy number,
as compared to the virus control receiving no compound.
In parallel, uninfected MDCK cells were used to determine
the CC50 values of the compounds after 24 h incubation,
using the MTS cell viability assay (CellTiter 96 AQueous
One Solution Cell Proliferation Assay; Promega). Ribavirin
was included as the reference compound.
2
,4-dioxo-4-phenylbutanoic acid (DPBA) was included as
the reference compound.
Cells and media
Madin–Darby canine kidney (MDCK) cells (a kind gift
from M. Matrosovich, Marburg, Germany) and human
embryonic kidney 293T (HEK293T) cells (purchased from
Thermo Scientific) were grown in Dulbecco’s modified
Eagle medium (DMEM) supplemented with 10 % foetal
calf serum (FCS), 1 mM sodium pyruvate and 0.075 %
sodium bicarbonate. Virus experiments were performed
in MDCK infection medium, consisting of Ultra MDCK
medium (Lonza, Basel, Switzerland) supplemented with
0
2
.0225 % sodium bicarbonate, 2 mM l-glutamine and
μg/mL tosyl phenylalanyl chloromethyl ketone (TPCK)-
treated trypsin (Sigma-Aldrich). The cells were incubated
in a 5 % CO humidified atmosphere.
2
vRNP reconstitution assay
The assay to determine the inhibitory effect of the com-
pounds on reconstituted influenza virus vRNPs is
described in more detail elsewhere [39, 40]. Briefly, the
four relevant expression plasmids derived from influenza
A/PR/8/34 (i.e. pVP-PB1, pVP-PB2, pVP-NP and pVP-
PA; generously donated by M. Kim [41], Korea Research
Institute of Chemical Technology, Daejeon, South Korea)
were combined with a firefly luciferase reporter plasmid
Computational studies
The crystal structure of PA endonuclease in complex with
epigallocatechin-3-gallate was retrieved from the RCSB
Protein Data Bank (entry code 4AWM). The ligand and
water molecules were discarded and the hydrogens were
added to protein by Discovery Studio 2. The epigallocat-
echin-3-gallate, L-742,001 and 2-(4-(1H-tetrazol-5-yl)
phenyl)-5-hydroxypyrimidin-4(3H)-one structures were
(
also a kind gift from M. Kim) and cotransfected into
HEK293T cells using Lipofectamine 2000. After 24 h of
incubation at 37 °C in the presence of the test compounds,
luciferase activity was determined using the ONE-Glo
luciferase assay system (Promega). The 50 % effective
1
3

J Biol Inorg Chem
extracted from their X-ray complexes, and the other struc-
tures of the ligands were constructed using Discovery Stu-
dio 2.5.5 (Accelrys, Discovery Studio) and energy mini-
mized using the Smart Minimizer protocol (1000 steps)
which combines the Steepest Descent and the Conjugate
Gradient methods.
The ligands minimized in this way were docked in their
corresponding proteins by means of Gold Suite 5.0.1 [44].
The region of interest used by Gold program was defined to
contain the residues within 15 Å from the original position
of the ligand in the X-ray structure. The side-chain of the
residue Tyr24 was allowed to rotate according to the inter-
nal rotamer libraries in GOLD Suite 5.0.1. GoldScore was
chosen as fitness function and the standard default setting
was used in all calculations and the ligands were submit-
ted to 100 genetic algorithm runs. The “allow early ter-
mination” command was deactivated. Results differing by
less than 0.75 Å in ligand-all atom RMSD were clustered
together. The best GOLD-calculated conformation was
used both for analysis and representation.
little is known about their interactions with magnesium
ions [52]. Magnesium is a hard Lewis acid and so coordi-
nation of the sulphur atom could be excluded. It is known
that salicylaldimine ligands can be NO-coordinated to mag-
nesium (A, Fig. 3), but in 1 and 2, with two oxygen donors
in ortho, coordination B (Fig. 3) is also plausible. Since
chelation of the divalent metal cofactors within the active
site of the influenza virus PA endonuclease provides the
possible basis for their antiviral activity [17], we decided to
clarify the coordination properties of two of these ligands
1
2
2+
(1 and 2, H L and H L , respectively) towards Mg . The
2
2
ligands were combined with magnesium acetate to react
in the presence of triethylamine as a base to afford the
corresponding metal complexes 8 and 9. It is worth not-
ing that, in the case of 2, the reaction was executed under
inert atmosphere, since the reaction mixture became dark
when exposed to air. In order to verify the influence of the
metal to ligand ratio or the pH, different reaction conditions
were used with 1 (1:1 and 1:2 metal to ligand ratio; up to
4 equivalents of base), but only with a 1:2 metal to ligand
ratio and in the presence of 4 equivalents of base, a metal
complex was isolated.
Results and discussion
Chemistry
The infrared spectrum of both 8 and 9 displays bands
−1
associated with the NH group (3163–3460 cm ), thus indi-
cating that this functionality is not deprotonated. The bands
associated with the C=S stretching are unaffected upon
complexation; therefore, it can be assumed that this group is
not involved in coordination, as confirmed by X-ray diffrac-
tion analysis on 8. In the IR spectrum of 8 the stretching of
the methoxy group is slightly shifted compared to the free
The ligands 1–6 (Fig. 3) were easily prepared in high yields
by condensation of the thiosemicarbazide and the substi-
tuted benzaldehyde; they were characterized by spectro-
scopic tools, mass spectrometry and elemental analysis. 2
was also characterized by single crystal X-ray diffraction.
ligand, suggesting a possible involvement in coordination.
1
In 6 the OH/OCH groups are substituted by a carboxylic
The H-NMR spectrum of 8 registered in d -DMSO shows
3
6
moiety, which retains good affinity for magnesium ions,
following a suggestion emerged in the research for new
inhibitors of HIV RNase H [45], which has two magnesium
ions in its active site. In the IR spectra of 1–6 the C = S
stretching absorptions are at about 1530–1580 and 820–
the presence of two sets of signals: one set corresponds to
the free ligand, while the second one corresponds to the
metal complex. The use of a coordinating solvent, there-
fore, causes the partial decoordination of the ligand. On the
1
contrary, in the H-NMR spectrum in MeOD, where the
−1
1
9
40 cm ; in the H-NMR spectra, the resonances of the
solubility is, however, low, there is a unique set of signals:
the ligand coordinates to the metal in a bidentate fashion
iminic proton (about 8.0 ppm) and of the hydrazonic NH
1
and of NH (11.3 ppm and 7.8–8.0 ppm, approximately)
only when a poor coordinating solvent is present. The H-
2
are evident. Isomerism around the C=N bond is possible,
NMR spectrum of 9 shows very broad signals, which might
indicate that a fluxional behaviour can exist in solution.
It is worth noting that complex 9 is not stable if exposed
to air both in solution and at the solid state. Mass spectra
confirmed the presence in solution of complexes of type
1
but in the H-NMR spectra registered in d -DMSO there is
6
only one set of signals: evidently, all the ligands are in the
E form in solution. Even if these ligands can give rise to
a thione-thiol equilibrium, no evidence was obtained about
the presence of the thiolic form in solution. Analogous con-
siderations can be made for 7, the hydrazonic analogue of
compound 2.
M(HL) and elemental analysis confirmed the proposed sto-
2
1
ichiometry, Mg(HL ) ·2.5H O, for 8.
2
2
Ligands 1–5 are very versatile: they possess both soft
and hard donor atoms and they can act as bidentate or tri-
dentate chelating ligands. Complexes of salicylaldimine
thiosemicarbazones with divalent metals like Pd(II), Zn(II),
Cu(II), Ni(II) have been studied previously [46–51], but
X‑ray crystallography
Ligand 2 was crystallized from ethanol as the solvate form
2·CH CH OH (Table 1). The molecular structure, shown
3
2
in Fig. 4, is perfectly planar, and the thiosemicarbazone
1
3

J Biol Inorg Chem
Fig. 4 Molecular structure and
labelling of 2, with thermal
ellipsoids drawn at the 50 %
probability level. Bottom
supramolecular aggregation of 2
.
CH CH OH in the solid, with
3
2
dotted hydrogen bonds
chain is completely extended, with trans torsion angles
larger than 175° along the chain. The cis configuration
between the amide NH and the imidic CH favours the
aggregation of coplanar molecular pairs by the interac-
tions between these groups and the sulphur (Fig. 4, bot-
tom) (N2–H…S1(i) = 3.401(1)Å, 167.2(1)°, C7–H…
S1(i) = 3.762(1)Å, 152.0(1)°, i = 1 − x, −y, 2 − z). The
dimeric assembly is surrounded by ethanol molecules
linked by hydrogen bonds to the hydroxyl groups of 2
which act as donors: O2–H…O3(ii) = 2.816(2)Å, 141(2)°;
O1-H…O3(iii) = 2.770(2)Å, 168(3)° (ii = 2 − x, y − 1/2,
deprotonated hydroxyl group, forming two planar five-
membered chelation rings; two cis methanol molecules
complete the octahedral coordination (Fig. 5) and the over-
all complex molecule has a pseudo twofold rotation axis
relating the two ligands and the two methanol molecules.
The mer configuration of the two bidentate ligand con-
fers optical activity to the complex; the crystal is, however,
racemic due to the presence of reflection planes. The thio-
semicarbazone arms are protonated on the amidic nitrogen
and are noticeably distorted due to intramolecular inter-
actions. In fact the amide NH and the imidic CH groups
are in trans configuration, and the latter is also turned
3
/2 − z; iii = 2 − x, −y, 1 − z). The ethanol molecule in
turn donates a hydrogen bond to the sulphur atom (O3–H…
away from the hydroxyl group, in contrast to the situa-
2
S1(iv) = 3.218(1)Å, 158(2)°, iv = x + 1, y, z) while the
tion observed for the free ligand 2 (H L ), whose molecu-
2
1
−
NH moiety donates a hydrogen bond to one hydroxyl
group (N3–H…O2(v) = 3.013(2)Å, 178(2)°, v = 1 − x,
lar structure is comparable to 1 (H L ). The torsion angles
2
2
1
−
along the chains of the two ligands (HL ) in the complex
differ significantly from the 0/180° value of the extended
conformation of the free ligand, reported in parentheses:
C1–C2–C8–N1 = 150/155°(−5°), C2–C8–N1–N2 = 7/3°
(180°), C8–N1–N2–C9 = −166/−162°(175°), N1–N2–
C9–N3 = 9/5°(−2°). These conformational distortions
are mainly due to strong intramolecular hydrogen bonds
between the amidic NH and the coordinated deprotonated
1
/2 + y, 3/2 − z), thus generating a three-dimensional
hydrogen bond network.
1
Complex Mg(HL ) ·2CH OH crystallizes from metha-
2
3
nol in the orthorhombic P2nn space group (Table 1). The
neutral complex comprises two monodeprotonated (HL )
1
−
ligands coordinated in a bidentate mode to the magne-
sium cation by means of the methoxy group and of the
1
3

J Biol Inorg Chem
Fig. 5 Molecular structure
and labelling of complex
1
Mg(HL ) ·2CH OH, with
2
3
thermal ellipsoids drawn at
the 50 % probability level
and dashed intramolecular
hydrogen bonds. Bottom
supramolecular aggregation
1
of Mg(HL ) ·2CH OH in the
2
3
solid, with dotted hydrogen
bonds
hydroxyl oxygens (N2–H…O1 = 2.582(1)Å, 147(1)°;
N11–H…O10 = 2.589(1)Å, 153(1)°), and between
the coordinated methanol molecules and the sulphur
atoms (O100–H…S1 = 3.189(1)Å, 155(1)°; O200–H…
HeLa cells; parainfluenza-3 virus and Punta Toro virus,
tested in Vero cells). However, they were found to be
inactive against these diverse viruses, with the exception
of a modest activity of 2 against herpes simplex virus-1
S2 = 3.212(1)Å, 159(1)°). The –NH moiety is involved in
(EC50 = 58 μM) and vaccinia virus (EC = 45 μM).
2
50
an intermolecular hydrogen bond to the sulphur atom gen-
erating an R22(8) ring (Fig. 5, bottom) which expands in
ribbons along the diagonal ab (N3–H…S1(vi) = 3.410(1)
Å, 162(1)°; N12–H…S2(vii) = 3.503(2)Å, 166(1)°,
vi = x + 1, y − 1, z, vii = x − 1, y + 1, z). In both the free
ligand and the complex the sulphur atoms are significantly
involved in intermolecular interactions where they act as
acceptors from –OH and –NH groups.
As can be seen in Table 2, compounds 2 and 4 displayed
moderate activity in the PA-Nter enzymatic assay, with
the catechol derivative 2 presenting the best result (IC :
5
0
37 μM). It is interesting to note that this compound has
an activity profile similar to that of a 2,3-dihydroxyphenyl
amide inhibitor that we recently reported [22], indicating
the importance of the catechol pharmacophore for inhibi-
tion of the PA-Nter enzyme. In line with the enzymatic
data on series 1–6, compounds 2 and 4 were also the only
derivatives showing antiviral activity in both cell culture
assays. Compound 2 displayed an EC99 ≥ 87 μM and
Biological activity
Compounds 1–6 were evaluated for their ability to inhibit
endonuclease activity in an enzymatic assay with recom-
binant PA-Nter, and the results are presented in Table 2.
Besides, their anti-influenza virus activity was determined
in cell culture, in a virus yield assay in MDCK cells and
in a vRNP reconstitution assay in HEK293T cells. We also
evaluated the inhibitory activity of 1–6 against a broad
panel of DNA [i.e. herpes simplex virus type 1, vaccinia
virus and adenovirus, evaluated in infected human embry-
onic lung fibroblast (HEL) cells] and RNA viruses (Cox-
sackie B4 virus and respiratory syncytium virus, tested in
EC90 ≥ 56 μM in the virus yield assay, and an EC value
50
of 63 µM in the vRNP reconstitution assay. For 4, the cor-
responding values were 48, 34 and 22 µM, respectively.
Additionally, compound 5 displayed weak inhibitory activ-
ity in the vRNP reconstitution assay, with an average IC50
value of 150 µM.
Even if the set of data is quite limited, some prelimi-
nary conclusions can be drawn. 1 and 2 can coordinate the
divalent magnesium ion in the B mode, using the methoxy
and OH groups (1) or two OH groups (2); however, 1 is
inactive, while 2 has moderate activity. The same type of
1
3

J Biol Inorg Chem
Table 2 Inhibitory activity of compounds 1–7 in the enzymatic assay with influenza virus PA-Nter endonuclease, or cellular influenza virus
assays based on virus yield or vRNP reconstitution
Compound
Enzyme assay with PA-Ntera
IC50
Virus yield assay in influenza virus-infected MDCK
cells
vRNP reconstitution assay in
HEK293T cells
b
c
Antiviral activity
Cytotoxicity
MCC
Activity
EC50
Cytotoxicity
CC50
EC99
EC90
CC50
(
(
(
(
(
(
(
1)
2)
3)
4)
5)
6)
7)
>500
37
>200
≥87
>200
48
>200
≥56
>200
34
≥200
≥100
>200
≥100
≥200
>200
ND
>200
≥138
>200
146
>200
63
>200
>200
>200
>200
>200
>200
>200
ND
>500
341
>500
>500
24
>200
22
>200
>100
ND
>200
>100
ND
>200
>200
ND
150
>100
107
ND
9.3
DPBAd
5.5
ND
ND
ND
ND
Ribavirin
ND
13
8.5
≥200
>200
>200
ND not determined
a
2+
Recombinant PA-Nter was incubated with the ssDNA plasmid substrate, a Mn -containing buffer and test compounds. Cleavage of the sub-
strate was assessed after 2 h incubation. The IC50 represents the compound concentration (in µM) to obtain 50 % inhibition of cleavage
b
MDCK (Madin–Darby canine kidney) cells were infected with influenza A virus (strain A/PR/8/34) and incubated with the compounds for
2
4 h. The virus yield in the supernatant was assessed by real-time qPCR. The EC99 and EC90 values represent the compound concentrations (in
µM) producing a 2-log10 or 1-log10 reduction in virus titer, respectively. The cytotoxicity, assessed in uninfected MDCK cells, was expressed as
the CC50 value (50 % cytotoxic concentration, determined with the MTS cell viability assay, in µM)
c
HEK293T (human embryonic kidney 293T) cells were cotransfected with the four vRNP-reconstituting plasmids and the luciferase reporter
plasmid in the presence of the test compounds. The EC50 value represents the compound concentration (in µM) producing 50 % reduction in
vRNP-driven firefly reporter signal, estimated at 24 h after transfection. The CC50 (in µM), i.e. the 50 % cytotoxic concentration, was determined
in untransfected HEK293T cells by MTS cell viability assay
d
DPBA, 2,4-dioxo-4-phenylbutanoic acid
substitution, i.e. moving from 4 (5-OH) to 5 (5-methoxy),
has the same effect on biological activity [i.e. 4 is moder-
ately active; 5 has very weak activity (vRNP reconstitution
assay) or is inactive (virus yield assay)]. It is difficult to
correlate this effect to a different coordinating ability of the
ligands, in particular looking at 4 and 5, which can both use
mode A (Fig. 3). Similarly, a different coordination behav-
iour cannot explain the different activity of 3 and 4. On the
other hand, the finding that 4 has lower activity in the enzy-
matic assay than 2 could well be related to their different
coordination mode to the magnesium ion in the active site
of the PA-Nter enzyme.
To further investigate the role of the thiosemicarbazone
moiety on activity, we evaluated the biological activity of
compound 7, the hydrazonic analogue of 2. Compound 7
has three oxygens in the right orientation to chelate simulta-
neously two metal ions. It is worth noting that the replace-
ment of the sulphur atom of 2 with an oxygen in compound
thiosemicarbazone 2, thus suggesting that for the thiosemi-
carbazone compounds inhibition of PA-Nter endonuclease
is not the only mode of action. In influenza virus-infected
cells, in fact, the antiviral activity is higher for compound
4 than for 2, and the other TSCs are inactive. Note that, in
cell culture (i.e. in the virus yield and vRNP reconstitu-
tion assay), 4 produces antiviral activity at concentrations
below 50 μM, while in the enzymatic assay its IC50 value
was 341 μM. This suggests that the suppressive effect of
4 on influenza virus replication may not merely be related
to direct inhibition of the endonuclease enzyme, since the
compound may also affect other steps in the virus replica-
tive cycle. These considerations are in line with our recent
studies [20] showing the complex biological properties of
some molecules proposed as metal chelating inhibitors of
the influenza virus PA endonuclease.
Computational studies
7
results in a slight improvement in the inhibitory activity
against the PA-Nter enzyme, with the IC50 changing from
7 to 24 μM. This can be related to the improved ability
In order to explain the difference of activity between com-
pound 1 and 2, docking simulations by Gold program [44]
were performed using the structural coordinates (pdb code
4AWM) for the PA endonuclease in complex with epigallo-
catechin-3-gallate [14].
3
of the ligand to interact with the metal cofactors, when a
harder donor atom is present. However, the activity of 7
in the cellular vRNP assay is lower than that of the parent
1
3

J Biol Inorg Chem
Fig. 6 Docking studies of PA
endonuclease with compounds
1
(cyan a) and 2 (green b),
respectively. Key residues of
the pocket are presented using
PyMOL [http://www.pymol.
org]. Hydrogen bonds are illus-
trated by dotted lines while the
divalent metal ions are shown
as purple spheres. Schematic
drawing of types of interactions
of the compounds 1 (c) and 2
(
d) generated using Ligplot
[
53]. Dashed lines are hydrogen
bonds and eyelashes show resi-
dues involved in hydrophobic
interactions
Considering that the side-chains of some residues
change depending on which pocket the ligand is occupy-
ing, we superimposed some X-ray structures of complexes
between PA endonuclease and known active ligands.
We observed that the side-chain of the aminoacid Tyr24
showed a greater movement than the other residues and for
this reason we considered it as a flexible residue during the
docking procedure.
agreed well with the experimental binding modes with
rmsd of 0.8, 1.2 and 0.7, respectively.
Later, compounds 1 and 2 were docked and the best
docking poses were selected and analysed by Ligplot [53]
revealing a similar binding mode (Fig. 6a, b).
The thiosemicarbazone portions of 1 and 2 share a
very similar network of interactions with the residues into
the active site (hydrophobic interactions with Gly121 and
Tyr130 and hydrogen bonds with Val122). The salicylalde-
hyde portion is able to chelate one magnesium ion in the
same way (B mode) for both derivatives, but the 3-hydroxy
(2) and 3-methoxy (1) groups play different roles, engaging
different interactions both with the second magnesium ion
First, test docking calculations, using epigallocatechin-
3
5
-gallate, L-742,001 and 2-(4-(1H-tetrazol-5-yl)phenyl)-
-hydroxypyrimidin-4(3H)-one (Fig. 1) were carried out
to compare experimental and predicted binding modes and
to validate our docking protocol. Their best docking poses
1
3

J Biol Inorg Chem
and with some important aminoacidic residues. In particu-
lar, LigPlot analysis displays that the 3-OH of 2 forms an
additional contact with the second metal ion and hydrogen
bonds with the catalytic residues Asp108 and Glu119, sta-
bilizing the molecule in the active site: these interactions,
therefore, could explain the difference of activity between
the two compounds.
different antiviral mechanisms, not only metal chelation.
This underscores that characterization of the activity of a
new compound is possible only if different complementary
assays are used.
Acknowledgments The authors thank the “Centro Interfacoltà
Misure Giuseppe Casnati” of the University of Parma for facilities.
L. Naesens and A. Stevaert acknowledge financial support from the
Geconcerteerde Onderzoeksacties—KU Leuven (GOA/15/019/TBA)
and the technical assistance from Wim van Dam, Leentje Persoons en
Ria Van Berwaer.
Conclusions
We have started a project focused on the synthesis and bio-
logical evaluation of chemical scaffolds that are able to
bind one or two metal ions in the PA-Nter active site and,
in this way, can inhibit influenza virus replication [20, 22,
References
1
.
Deyde VM, Xu X, Bright RA, Shaw M, Smith CB, Zhang Y,
Shu Y, Gubareva LV, Cox NJ, Klimov AI (2007) J Infect Dis
4
0]. From this perspective, salicylaldehyde thiosemicar-
1
96:249–257
bazones seem promising, because of their well-known
coordinating abilities and biological properties. In the first
evaluation reported here, positions 3, 4 and 5 of the phenyl
2. Moscona A (2009) N Engl J Med 360:953–956
3
4
.
.
Memoli MJ, Davis AS, Proudfoot K, Chertow DS, Hrabal RJ,
Bristol T, Taubenberger JK (2011) J Infect Dis 203:348–357
Van der Vries E, Schutten M, Fraaij P, Boucher C, Osterhaus A
(2013) Adv Pharmacol 67:217–246
ring were substituted with OH and OCH donor groups to
3
derive a structure–activity relationship. Compounds 2 and 4
displayed moderate activity in the PA Nter enzymatic assay
and antiviral activity in both cell culture assays used in this
study. In particular, compound 2, with two OH groups in
position 2 and 3 on the phenyl ring, is the most active and
this is in line with the results recently reported by us with
a 2,3-dihydroxyphenyl amide inhibitor [22] and the litera-
ture regarding the use of natural polyphenols for inhibiting
influenza virus infection [54]. Thiosemicarbazones 1–5 are
effectively able to coordinate magnesium(II) ion, as defi-
5. Vanderlinden E, Naesens L (2014) Med Res Rev 34:301–339
6
7
.
.
Honda A, Ishihama A (1997) Biol. Chem 378:483–488
Honda A, Mizumoto K, Ishihama A (2002) Proc Natl Acad Sci
USA 99:13166–13171
8
9
.
.
Das K, Aramini JM, Ma LC, Krug RM, Arnold E (2010) Nat
Struct Mol Biol 17:530–538
Dias A, Bouvier D, Crépin T, McCarthy AA, Hart DJ, Baudin F,
Cusack S, Ruigrok RW (2009) Nature 458:914–918
1
0. Yuan P, Bartlam M, Lou Z, Chen S, Zhou J, He X, Lv Z, Ge R,
Li X, Deng T, Fodor E, Rao Z, LiuY (2009) Nature 458:909–913
11. Plotch SJ, Bouloy M, Ulmanen I, Krug RM (1981) Cell
3:847–858
2
1
12. Zhao C, Lou Z, Guo Y, Ma M, Chen Y, Liang S, Zhang L, Chen
S, Li X, Liu Y, Bartlam M, Rao Z (2009) J Virol 83:9024–9030
nitely proved by the structure of Mg(HL ) ·2CH OH, but
2
3
it seems difficult to correlate their activity exclusively to
this ability. For example, compound 5 is inactive, while 4 is
moderately active, but structural differences between them
1
1
1
3. DuBois RM, Slavish PJ, Baughman BM, Yun MK, Bao J, Webby
RJ, Webb TR, White SW (2012) PLoS Pathog 8:e1002830
4. Kowalinski E, Zubieta C, Wolkerstorfer A, Szolar OH, Ruigrok
RW, Cusack S (2012) PLoS Pathog 8:e1002831
(
5-methoxy group in 5 instead of 5-OH group as in 4) do
5. Doan L, Handa B, Roberts NA, Klumpp K (1999) Biochemistry
not imply differences in chelating properties. The different
activity could depend not only on a different mechanism of
action, but also on the effect that Cohen sharply describes
as “malleable interaction” in metal coordination by metal-
loenzyme inhibitors [55], to describe the large number of
factors (donor atom identity, orientation, electrostatics, van
der Waals interactions,…) that are involved in the inhibi-
tion mechanism. The semicarbazone 7, that has three oxy-
gen atoms and can coordinate simultaneously two metal
ions in the enzyme active site, is more active than the corre-
sponding thiosemicarbazone 2. Following this suggestion,
we are now working on polyhydroxylated acylhydrazones
to exploit their potential as antiviral inhibitors [56].
3
8:5612–5629
16. Xiao S, Klein ML, LeBard DN, Levine BG, Liang H, MacDer-
maid CM, Alfonso-Prieto M (2014) J Phys Chem B 118:873–889
7. Rogolino D, Carcelli M, Sechi M, Neamati N (2012) Coord
Chem Rev 256:3063–3086
8. Tomassini JE, Selnick H, Davies ME, Armstrong ME, Baldwin
J, Bourgeois M, Hastings J, Hazuda D, Lewis J, McClements
W, Ponticello G, Radzilowski E, Smith G, Tebben A, Wolfe A
1
1
(
1994) Antimicrob Agents Chemother 38:2827–2837
1
9. Hastings JC, Selnick H, Wolanski B, Tomassini JE (1996) Anti-
microb Agents Chemother 40:1304–1307
20. Stevaert A, Nurra S, Pala N, Carcelli M, Rogolino D, Shepard C,
Domaoal RA, Kim B, Alfonso-Prieto M, Marras SAE, Sechi M,
Naesens L (2015) Mol Pharmacol 87:323–337
1. Singh SB (1995) Tetrahedron Lett 36:2009–2012
2. Carcelli M, Rogolino D, Bacchi A, Rispoli G, Fisicaro E, Com-
pari C, Sechi M, Stevaert A, Naesens L (2014) Mol Pharm
2
2
Another point that deserves closer examination is the
difference between the data obtained for 2 and 4 using dif-
ferent assays: in the enzyme assay, 2 is more active than
1
1:304–316
2
3. Parkes KEB, Ermert P, Fässler J, Ives J, Martin JA, Merrett
JH, Obrecht D, Williams G, Klumpp K (2003) J Med Chem
4
, but the opposite was seen in virus-infected cells and the
4
6:1153–1164
vRNP reconstitution assay. Compound 4 probably acts by
1
3

J Biol Inorg Chem
2
2
2
4. Sagong HY, Parhi A, Bauman JD, Patel D, Vijayan RS, Das K,
Arnold E, LaVoie EJ (2013) ACS Med Chem Lett 4:547–550
5. Knox JJ, Hotte SJ, Kollmannsberger C, Winquist E, Fisher B,
Eisenhauer EA (2007) Invest New Drugs 25:471–477
6. Kang IJ, Wang LW, Hsu TA, Yueh A, Lee CC, Lee YC, Lee CY,
Chao YS, Shih SR, Chern JH (2011) Bioorg Med Chem Lett
42. Reed LJ, Muench H (1938) Am J Epidemiol 27:493–497
43. Vanderlinden E, Göktas F, Cesur Z, Froeyen M, Reed ML, Rus-
sell CJ, Cesur N, Naesens L (2010) J Virol 84:4277–4288
44. Jones G, Willett P, Glen RC, Leach AR, Taylor R (1997) J Mol
Biol 267:727
45. Gong Q, Menon L, Ilina T, Miller LG, Ahn J, Parniak MA,
Ishima R (2011) Chem Biol Drug Des 77:39–47
2
1:1948–1952
2
2
7. Quenelle DC, Keith KA, Kern ER (2006) Antiv Res 71:24–30
8. Debebe Z, Ammosova T, Breuer D, Lovejoy DB, Kalinowski
DS, Kumar K, Jerebtsova M, Ray P, Kashanchi F, Gordeuk VR,
Richardson DR, Nekhai S (2011) Mol Pharmacol 79:185–196
9. Pelosi G, Bisceglie F, Bignami F, Ronzi P, Schiavone P, Re MC,
Casoli C, Pilotti E (2010) J Med Chem 53:8765–8769
0. Torti SV, Torti FM (2013) Nat Rev Cancer 13:342–355
1. Pelosi G (2010) Open Crystallogr J 3:16–28
2. Sacconi L (1954) Z Anorg Allg Chem 275:249–256
3. Dolomanov OV, Bourhis LJ, Gildea RJ, Howard JAK,
Puschmann H (2009) J Appl Cryst 42:339–341
4. Burla MC, Caliandro R, Camalli M, Carrozzini B, Cascarano
GL, De Caro L, Giacovazzo C, Polidori G, Siliqi D, Spagna R
46. Enyedy ÉA, Zsigó É, Nagy NV, Kowol CR, Roller A, Keppler
BK, Kiss T (2012) Eur J Inorg Chem 2012(25):4036–4047
47. Lobana TS, Kumari P, Hundal G, Butcher RJ (2010) Polyhedron
29:1130–1136
48. Chellan P, Stringer T, Shokar A, Dornbush PJ, Vazquez-Anaya
G, Land KM, Chibale K, Smith GS (2011) J Inorg Biochem
105:1562–1568
49. Ramachandran E, Kalaivani P, Prabhakaran R, Rath NP, Brinda
S, Poornima P, Padma VV, Natarajan K (2012) Metallomics
4:218–227
50. Kalaivani P, Prabhakaran R, Dallemer F, Poornima P, Vaishnavi
E, Ramachandran E, Padma VV, Renganathan R, Natarajan K
(2012) Metallomics 4:101–113
2
3
3
3
3
3
(
2007) J Appl Cryst 40:609–613
51. West DX, Salberg MM, Bain GA, Liberta AE, Valdés-Martínez
J, Hernández Ortega S (1996) Transit Met. Chem. 21:206–212
52. Sahadev RKS, Sindhwani SK (1992) Thermochim Acta
202:291–299
3
3
3
5. Sheldrick GM (2008) Acta Cryst A64:112–122
6. Nardelli M (1995) J Appl Cryst 28:659
7. Allen FH, Kennard O, Taylor R (1983) Acc Chem Res
1
6:146–153
53. Laskowski RA, Swindells MB (2011) J Chem Inf Mod 51:2778
54. Yang ZF, Bai LP, Huang WB, Li XZ, Zhao SS, Zhong NS, Jiang
ZH (2014) Fitoterapia 93:47–53
3
8. Bruno IJ, Cole JC, Edgington PR, Kessler M, Macrae CF,
McCabe P, Pearson J, Taylor R (2002) Acta Crystallogr B
5
8:389–397
55. Martin DP, Blachly PG, McCammon JA, Cohen SM (2014) J
Med Chem 57:7126–7135
56. Rogolino D, Carcelli M, Bacchi A, Compari C, Contardi L, Fisi-
caro E, Gatti A, Sechi M, Stevaert A, Naesens L (2015) J Inorg
Biochem 150:9–17
3
4
4
9. Meneghesso S, Vanderlinden E, Stevaert A, McGuigan C,
Balzarini J, Naesens L (2012) Antiviral Res 94:35–43
0. Stevaert A, Dallocchio R, Dessì R, Pala N, Rogolino D, Sechi M,
Naesens L (2013) J Virol 87:10524–10538
1. Kim M, Kim SY, Lee HW, Shin JS, Kim P, Jung YS, Jeong HS,
Hyun JK, Lee CK (2013) Antiviral Res 100:460–472
1
3