Metal-chelating 2-hydroxyphenyl amide pharmacophore for inhibition of influenza virus endonuclease

Article abstract of DOI:10.1021/mp400482a

The influenza virus PA endonuclease is an attractive target for development of novel anti-influenza virus therapeutics. Reported PA inhibitors chelate the divalent metal ion(s) in the enzyme's catalytic site, which is located in the N-terminal part of PA (PA-Nter). In this work, a series of 2-hydroxybenzamide- based compounds have been synthesized and biologically evaluated in order to identify the essential pharmacophoric motif, which could be involved in functional sequestration of the metal ions (probably Mg²⁺) in the catalytic site of PA. By using HL¹, H₂L², and HL³ as model ligands with Mg²⁺ ions, we isolated and fully characterized a series of complexes and tested them for inhibitory activity toward PA-Nter endonuclease. H₂L² and the corresponding Mg²⁺ complex showed an interesting inhibition of the endonuclease activity. The crystal structures of the uncomplexed HL¹ and H ₂L² and of the isolated magnesium complex [Mg(L ³)₂(MeOH)₂]·2MeOH were solved by X-ray diffraction analysis. Furthermore, the speciation models for HL¹, H₂L², and HL³ with Mg²⁺ were obtained, and the formation constants of the complexes were measured. Preliminary docking calculations were conducted to investigate the interactions of the title compounds with essential amino acids in the PA-Nter active site. These findings supported the two-metal coordination of divalent ions by a donor triad atoms chemotype as a powerful strategy to develop more potent PA endonuclease inhibitors.

Full text of DOI:10.1021/mp400482a

Article
pubs.acs.org/molecularpharmaceutics
Metal-Chelating 2‑Hydroxyphenyl Amide Pharmacophore for
Inhibition of Influenza Virus Endonuclease
Mauro Carcelli,*,† Dominga Rogolino,† Alessia Bacchi,† Gabriele Rispoli,† Emilia Fisicaro,‡
Carlotta Compari,‡ Mario Sechi,§ Annelies Stevaert,∥ and Lieve Naesens∥
†Dipartimento di Chimica,‡Dipartimento di Farmacia, Universita
̀
di Parma, Parco Area delle Scienze 17/A, 43124 Parma, Italy
§Dipartimento di Chimica e Farmacia, Universita
̀
di Sassari, Via Vienna 2, 07100 Sassari, Italy
∥Rega Institute for Medical Research, KU Leuven, B-3000 Leuven, Belgium
ABSTRACT: The influenza virus PA endonuclease is an
attractive target for development of novel anti-influenza virus
therapeutics. Reported PA inhibitors chelate the divalent metal
ion(s) in the enzyme’s catalytic site, which is located in the N-
terminal part of PA (PA-Nter). In this work, a series of 2-
hydroxybenzamide-based compounds have been synthesized and biologically evaluated in order to identify the essential
pharmacophoric motif, which could be involved in functional sequestration of the metal ions (probably Mg2+) in the catalytic site
of PA. By using HL1, H2L2, and HL3 as model ligands with Mg2+ ions, we isolated and fully characterized a series of complexes
and tested them for inhibitory activity toward PA-Nter endonuclease. H2L2 and the corresponding Mg2+ complex showed an
interesting inhibition of the endonuclease activity. The crystal structures of the uncomplexed HL1 and H2L2 and of the isolated
magnesium complex [Mg(L3)2(MeOH)2]·2MeOH were solved by X-ray diffraction analysis. Furthermore, the speciation models
for HL1, H2L2, and HL3 with Mg2+ were obtained, and the formation constants of the complexes were measured. Preliminary
docking calculations were conducted to investigate the interactions of the title compounds with essential amino acids in the PA-
Nter active site. These findings supported the “two-metal” coordination of divalent ions by a donor triad atoms chemotype as a
powerful strategy to develop more potent PA endonuclease inhibitors.
KEYWORDS: metalloenzymes, influenza virus endonuclease, metal chelation, antiviral, magnesium complexes
among influenza A, B, and C viruses, and its functions are
essential for viral genome replication. The RdRp is composed
of three subunits: PA, PB1, and PB2. The endonuclease activity,
which resides in the N-terminal part of PA (PA-Nter), is
required to cleave host cell pre-mRNAs to produce the 5′-
capped primers for transcription of the viral genomic RNA into
mRNA.12 The protein structure of PA-Nter was described in
recent crystallographic studies, which located two13 or one14
divalent metal ions in the active site. Recently, Crepin et al.15
demonstrated that Mn2+ binds tightly to one site and both
Mg2+ and Mn2+ can bind with lower affinity at a second site.
The two-metal-ion model16 is consistent with biochemical
studies indicating cooperative activation by the two-metal
ions.17 Similar to other metal-dependent viral enzymes, that is,
HIV integrase (IN) or RNase H,17 metal chelation represents a
relevant strategy to develop inhibitors of the PA endonuclease
because compounds that efficiently coordinate the metal ions
within the active site can block the enzymatic activity and
prevent processing of the biological substrates (Figure 1).19,20 A
major breakthrough in this field was the introduction of
Raltegravir, the first HIV IN inhibitor to be approved by the
INTRODUCTION
■
Seasonal human influenza A or B virus infections are an
important cause of morbidity and mortality, particularly in
children and elderly, chronically ill, or immunocompromised
individuals. In addition, influenza A viruses are responsible for
sporadic pandemics such as the 1918 “Spanish flu”1 or the
swine-origin A/H1N1 virus that emerged in 2009.2 There is
concern that highly virulent avian viruses such as A/H5N1 or
the recently emerged A/H7N9 virus may cause severe
pandemics when acquiring human-to-human transmissibil-
ity.3−5 Influenza vaccination is the most widely used
prophylactic measure, but since the development of a new
pandemic vaccine requires several months, effective antiviral
drugs should be available at all times. Anti-influenza virus drugs
that target the viral M2 ion channel (amantadine and
rimantadine) or neuraminidase (NA) (zanamivir and oseltami-
vir) have proven to be efficient drugs.6,7 However, due to the
emergence of influenza viruses showing resistance to M2 or NA
inhibitors, there is an urgent need for entirely novel antiviral
compounds with a different mode of action and targeting a
critical step in the viral replication process.8,9
The influenza virus genome consists of eight single-stranded
(-)RNA segments, which are transcribed and replicated by the
viral RNA-dependent RNA polymerase (RdRp).10,11 The
influenza virus RdRp represents an attractive target for the
development of new antiviral agents, since it is highly conserved
Received: August 14, 2013
Revised: October 24, 2013
Accepted: November 8, 2013
Published: November 8, 2013
© 2013 American Chemical Society
304
dx.doi.org/10.1021/mp400482a | Mol. Pharmaceutics 2014, 11, 304−316

Molecular Pharmaceutics
Article
Figure 1. (A) (Top) Crystal structure of the active site of PAN with the inhibitor (c) (adapted from ref 31); (bottom) some representative influenza
endonuclease inhibitors with their common chelating chemotype in bold. (B) (Top) Crystal structure of the HIV-1 IN inhibitor Dolutegravir within
the active site of PFV intasome [Mn(II) ions are depicted as orange spheres (adapted from Hare, S.; et al. Mol. Pharmacol., 80 (2011), 565)];
(bottom) the most important metal-chelating inhibitors of HIV-1 IN with the common chelating chemotype in bold.
U.S. FDA. Similarly, diverse metal-chelating molecules have
been reported to inhibit the influenza virus endonuclease
activity, starting from the β-diketoacid family (exemplified by
2,4-dioxo-4-phenylbutanoic acid or DPBA and L-742 001,
Figure 1), which are structurally related to some well-known
HIV-1 IN inhibitors.21,22 Metal-chelating inhibitors of the
influenza virus endonuclease belonging to other chemical
classes include (Figure 1): flutimide,23,24 N-hydroxamic acids
(a),25 and N-hydroxyimides [(b) and (c)].25,26 These
prototype compounds share a common pharmacophoric
motif that is likely responsible for functional sequestration of
the metal ions in the catalytic site of PA-Nter. Other
compounds such as marchantins, green tea catechins27 and
dihydroxy phenethylphenylphthalimides,28,29 bear a dihydroxy-
phenethyl group, and their action may be attributed to the
dihydroxy functionality which may chelate the divalent cations
within PA-Nter.28 Looking at these chemotypes, it is clear that
as in the case of HIV IN,30 a triad of donor atoms (i.e., O, “X”,
and “Y”, Figure 2) is not sufficient for potent inhibition. Rather,
their optimal arrangement appears essential, along with the
matching of their Lewis acid/base character. This was indeed
observed in two recent studies in which the cocrystal structures
of PA-Nter in complex with a variety of metal-chelating
inhibitors were obtained.31,32
Recently, the salicylic acid and catechol pharmacophores
were merged in a novel scaffold to create effective HIV IN
305
dx.doi.org/10.1021/mp400482a | Mol. Pharmaceutics 2014, 11, 304−316

Molecular Pharmaceutics
Article
N-(4-Fluorobenzyl)-2-methoxybenzamide (1). Carbon-
yl diimidazole (2.56 g, 15.77 mmol) was added under nitrogen
to a solution of 2-methoxy benzoic acid (2 g, 13.14 mmol) in
dry methylene chloride (45 mL). The solution was stirred at
room temperature for 1 h. Then, 4-fluoro benzyl amine (2.25
mL, 19.71 mmol) was added, and the white suspension was
stirred at room temperature for 24 h. The reaction was
monitored by TLC (SiO2, eluent dichloromethane/methanol
99:1). After the complete conversion of the starting material,
the reaction was quenched with 1 M HCl. The aqueous phase
was extracted with methylene chloride (3 × 50 mL). The
organic phases were washed with brine until neutral pH, dried
on sodium sulfate, filtered, and the solvent was eliminated
under reduced pressure to give the product as a light yellow oil.
Yield = 97%. The product was used without further purification.
1H NMR (CDCl3, 25 °C): δ = 3.93 (s, 3H, OCH3) 4.65 (d, J =
5.7 Hz, 2H, NHCH2Ar), 6.97−7.12 (m, 4H, ArH), 7.32−7.36
(m, 2H, ArH), 7.44−7.49 (m, 1H, ArH), 8.23−8.26 (m, 2H,
CONHCH2Ar, ArH). 13C NMR (CDCl3, 25 °C): δ = 165.4,
163.3, 160.8, 157.5, 134.7, 133.0, 132.4, 129.2, 121.4, 115.5,
115.3, 111.4, 56.0, 43.0. MS (EI, 70 eV) m/z (%) = 259.1 (49),
135 (100), 124.0 (31).
Figure 2. One metal (HL1) versus two metals binding (H2L2 and
HL3) 2-hydroxyphenyl amides.
inhibitors, capable of chelating two metal ions through the
adjacent carboxylic and hydroxyl groups.33 Due to the
functional resemblance between the active sites of HIV IN
and PA-Nter,18 it can be hypothesized that similar molecules
may be designed as effective PA endonuclease inhibitors.
Accordingly, we focused our attention on three derivatives,
previously tested as HIV IN inhibitors,30 containing the salicylic
group. In HL1 (Figure 2), the hydroxyl and amide groups
provide two essentially coplanar oxygen atoms to potentially
bind one metal ion in the active site of PA-Nter. On the other
hand, H2L2 (Figure 2) contains a “two-metal” binding motif
(i.e., a “O, O, O” donor atoms set) that can be involved in
coordination of both ions; a similar coordinative effect can be
predicted also for HL3 (Figure 2). We herein report the
synthesis and characterization of these three 2-hydroxybenza-
mides (HL1, H2L2, HL3) and the related metal complexes. We
confirmed the structure of HL1, H2L2, and [Mg-
(L3)2(MeOH)2]·2MeOH by X-ray crystallographic analysis.
The metal-bonding ability of HL1, H2L2, and HL3 toward
magnesium was studied in solution and solid state. Finally,
enzymatic assays were undertaken to investigate the activity of
the simple ligands and of related complexes in inhibiting the
endonuclease reaction performed by PA-Nter. Preliminary
computational docking studies of the model ligands in the PA-
Nter active site were also performed to elucidate the key
features that contribute to effective metal chelation by these
compounds.
N-(4-Fluorobenzyl)-2,3-dimethoxybenzamide (2). The
same as for (1), by using 2,3-dimethoxy benzoic acid (2 g,
10.98 mmol). The light beige product was used without further
1
purification. Yield = 86%. H NMR (CDCl3, 25 °C): δ = 3.82
(s, 3H, OCH3), 3.90 (s, 3H, OCH3), 4.64 (d, J = 5.7 Hz, 2H,
NHCH2Ar), 7.01−7.08 (m, 3H, ArH), 7.17 (t, J = 7.8 Hz, 1H,
ArH), 7.32−7.37 (m, 2H, ArH), 7.73 (dd, J = 6.3 Hz, J = 1.8
Hz, 1H, ArH), 8.32 (bt, 1H, CONHCH2Ar). 13C NMR
(CDCl3, 25 °C): δ = 165.2, 163.3, 160.9, 152.6, 147.5, 129.4,
129.3, 126.5, 124.5, 122.8, 115.6, 115.5, 115.4, 61.3, 56.1, 43.1.
MS (EI, 70 eV) m/z (%) = 289.1 (96), 258 (30), 165.0 (100),
124.0 (53), 109.0 (40).
N1,N3-Bis(4-fluorobenzyl)-2-methoxyisophthalamide
(3). The same as (1) by using 2-methoxy isophthalic acid (2 g,
10.2 mmol). A light pink solid was obtained after the usual
1
workup. Yield = 90%. H NMR (CDCl3, 25 °C): δ = 3.65 (s,
3H, OCH3), 4.61 (d, J = 6.0 Hz, 4H, NHCH2Ar), 7.02−7.06
(m, 4H, ArH), 7.31−7.35 (m, 4H, ArH,) 7.64 (bt, 2H,
CONH), 8.12 (d, J = 7.6 Hz, 2H, ArH). 13C NMR (CDCl3, 25
°C): δ = 164.8, 163.4, 161.0, 155.8, 134.6, 133.9, 129.7, 127.7,
125.3, 115.8, 115.6, 63.5, 43.3. MS (EI, 70 eV) m/z (%) =
410.1 (89), 379 (7), 124 (43), 109 (100).
N-(4-Fluorobenzyl)-2-hydroxybenzamide (HL1) (4). To
a solution of (1) (240.8 mg, 0.93 mmol) in methylene chloride
(10 mL), a 1 M solution of boron tribromide in methylene
chloride (5.6 mL, 5.6 mmol) at 0 °C was added, under nitrogen
atmosphere. The light yellow suspension was stirred for 24 h.
The reaction was monitored with TLC (SiO2; eluent
dichloromethane/methanol 98:2). After the complete con-
version of the starting material, the solution was diluted at 0 °C
with methanol/water 2:8 (20 mL). The aqueous phase was
extracted with dichloromethane (3 × 20 mL), the organic
phases were washed with brine until neutral pH, dried on
sodium sulfate, filtered, and the solvent was eliminated under
reduced pressure to give a brown liquid. The crude product was
purified by recrystallization with methanol−water to obtain the
EXPERIMENTAL SECTION
■
Materials and Methods. Chemistry. All reagents of
commercial quality were used without further purification.
Purity of compounds was determined by elemental analysis and
verified to be ≥95% for all synthesized 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 (Thermo Fisher) spectrophotom-
eter by using a diamond crystal plate in the range of 4000−400
cm−1. Elemental analyses were performed by using a FlashEA
1112 series CHNS/O analyzer (Thermo Fisher) with gas-
chromatographic separation. Electrospray mass spectral anal-
yses (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
(direct exposure probe) mounting on the tip of a Refilament
with a DSQII Thermo Fisher apparatus, equipped with a single
quadrupole analyzer. HL1, H2L2, and HL3 were synthesized
according to slightly modified literature procedures.30
1
product as white solid. Yield = 77%. H NMR (DMSO-d6, 25
°C): δ = 4.50 (d, J = 5.7 Hz, 2H, CONHCH2Ar), 6.88−6.93
(m, 2H, ArH), 7.14−7.19 (m, 2H, ArH), 7.36−7.44 (m, 3H,
ArH), 7.90 (dd, J = 7.5 Hz, J = 1.2 Hz, 1H, ArH), 9.36 (btr, J =
5.7 Hz, 1H, CONH), 12.49 (s, 1H, OH). 13C NMR (DMSO-
306
dx.doi.org/10.1021/mp400482a | Mol. Pharmaceutics 2014, 11, 304−316

Molecular Pharmaceutics
Article
Table 1. Crystal Data and Structure Refinement for HL1, H2L2, and [Mg(L3)2(MeOH)2]·2MeOH
HL1
H2L2
[Mg(L3)2(MeOH)2]·2MeOH
empirical formula
formula weight
temperature/K
crystal system
space group
C14H12NO2F
245.25
C14H12NO3F
261.25
C48H50N4O10F4Mg
471.61
293(2)
293(2)
293(2)
orthorhombic
P212121
orthorhombic
Pc21n
monoclinic
P21/n
a/Å
6.494(1)
11.669(2)
15.492(3)
90.00
8.8140(5)
10.3890(5)
13.0280(7)
90.00
15.072(2)
8.754(1)
18.684(2)
90.00
b/Å
c/Å
α/°
β/°
γ/°
volume/Å3
90.00
90.00
107.648(2)
90.00
90.00
90.00
1174.0(4)
4
1193.0(1)
4
2349.2(5)
2
Z
ρcalcmg/mm3
μ/mm−1
1.388
1.455
1.333
0.104
0.113
0.116
F(000)
512.0
544.0
988.0
2Θ range /°
reflections collected
independent refls [R(int)]
data/restraints/parameters
goodness-of-fit on F2
final R1, wR2 [I ≥ 2σ (I)]
final R1, wR2 [all data]
largest ΔF max/min/e Å−3
Flack parameter
4.38−59.42
17 752
5.58−63.8
18 770
3.06−63.88
37 797
3321[0.0480]
3321/0/165
0.894
3911[0.0214]
3911/1/181
0.910
7717[0.0363]
7717/1/313
1.055
0.0484, 0.1276
0.0723, 0.1470
0.14/−0.17
−0.1(11)
0.0334, 0.0943
0.0349, 0.0962
0.32/−0.19
−0.1(5)
0.0473, 0.1322
0.0633, 0.1445
0.53/−0.29
d6, 25 °C): δ = 169.4, 162.9, 160.6, 135.7, 135.6, 134.3, 129.8,
129.7, 128.3, 119.1, 117.9, 115.6, 115.4, 42.2. MS (EI, 70 eV)
m/z (%) = 245.0 (26), 124 (10), 121 (18), 109 (100). IR
(cm−1): νNH = 3044; νCO = 1632. mp 118.9−120.9 °C. Anal.
Calcd for C14H12FNO2: C, 68.56; H, 4.93; N, 5.71. Found: C,
68.81; H, 5.15; N, 5.59. X-ray diffraction quality crystals were
obtained by recrystallization from methanol/water.
= 5.37 Hz, 4H, CH2Ar), 7.03 (t, J = 7.7 Hz, 1H, Ar), 7.14−7.18
(m, 4H, Ar), 7.38−7.41 (m, 4H, Ar), 8.07 (d, J = 7.7 Hz, 2H,
Ar), 9.30 (bs, 2H, CONHCH2Ar), 14.72 (s, 1H, OH). 13C
NMR (DMSO-d6, 25 °C): δ = 167.6, 162.9, 160.1, 135.6, 133.2,
129.8, 118.8, 115.6, 115.4, 42.4. MS (EI, 70 eV) m/z (%) = 396
(5), 124 (100), 109 (65). IR (cm−1): νNH = 3326, νCO
=
1645. mp 124.9−126.3 °C. Anal. Calcd for C22H18F2N2O3: C,
66.66; H, 4.58; N, 7.07. Found: C, 66.34; H, 4.70; N, 7.03.
Mg(L1)2·3H2O (7). An aqueous solution of sodium
hydroxide (1.02 mmol, 0.255 mL) was added to a solution of
HL1 (168 mg, 0.68 mmol) in methanol (8 mL). The mixture
was stirred at room temperature for 1 h. Then, a suspension of
MgCl2·6H2O (69.12 mg, 0.34 mmol) in water (2 mL) was
added. The white suspension was stirred for 24 h. The solvent
was eliminated under reduced pressure to give a white solid,
which was washed with water until neutral pH and then dried
under vacuum. Yield = 45%. 1H NMR (DMSO-d6, 25 °C): δ =
4.45 (bs, 2H, NHCH2Ar), 6.46 (bs, 1H, Ar), 6.66 (bs, 1H, Ar),
7.10−7.14 (m, 3H, Ar), 7.33−7.36 (m, 2H, Ar), 7.72 (bs, 1H,
Ar), 10.62 (bs, 1H, NH). ESI-MS (+, m/z): 513 [Mg(L1)2 +
N-(4-Fluorobenzyl)-2,3-dihydroxybenzamide (H2L2)
(5). A 1 M solution of boron tribromide in methylene chloride
(70.20 mL, 70.20 mmol) at 0 °C and in nitrogen atmosphere
was added to a solution of (2) (3.3 g, 11.7 mmol) in methylene
chloride (130 mL). A yellow solid was obtained after the
workup already described for HL1. The crude product was
recrystallized in methanol−water. Yield = 84%. 1H NMR
(DMSO-d6, 25 °C): δ = 4.48 (d, J = 5.7 Hz, 2H,
CONHCH2Ar), 6.70 (t, J = 8.1 Hz, 1H, Ar), 6.93 (d, J = 7.2
Hz, 1H, Ar), 7.14−7.20 (m, 2H, Ar), 7.32−7.40 (m, 3H, Ar),
9.19 (s, 1H, OH), 9.34 (brt, J = 5.4 Hz, 1H CONH), 12.61 (s,
1H, OH). 13C NMR (DMSO-d6, 25 °C): δ = 170.2, 162.9,
150.1, 146.7, 135.6, 129.8, 129.7, 119.4, 118.5, 117.7, 115.7,
115.5, 115.4, 42.1. IR (cm−1): νNH = 3351, νCO = 1644. mp
203.3−204.9 °C. Anal. Calcd for C14H12FNO3: C, 64.37; H,
4.62; N, 5.35. Found: C, 63.90; H, 4.66; N, 5.34. X-ray
diffraction quality crystals were obtained by recrystallization
from methanol/water.
N1,N3-Bis(4-fluorobenzyl)-2-hydroxyisophthalamide
(HL3) (6). A 1 M solution of boron tribromide in methylene
chloride was added (62 mL, 62 mmol) to a solution of (3) (4.2
g, 10.3 mmol) in methylene chloride (110 mL), at 0 °C and in
nitrogen atmosphere. The light yellow suspension was stirred
under nitrogen for 24 h. A gray solid was obtained after the
workup already described for HL1 and H2L2. The crude
product was recrystallized in methanol−water to obtain a white
solid. Yield = 70%. 1H NMR (DMSO-d6, 25 °C): δ = 4.53 (d, J
H+], 535 [Mg(L1)2 + Na+]. IR (cm−1): νNH = 3380, νCO
=
1614. Anal. Calcd for C28H22F2MgN2O4·3H2O: C, 59.33; H,
4.98; N, 4.94. Found: C, 59.76; H, 4.84; N, 4.68.
The complexes (8) and (9) were synthesized by a similar
procedure.
Mg2(L2)2·2.5H2O (8). H2L2 (150 mg, 0.57 mmol) in
methanol (10 mL), sodium hydroxide (1.14 mmol, 0.570
mL), and an aqueous suspension of MgCl2·6H2O (59 mg, 0.29
mmol) were used. The product was isolated as a white solid.
1
Yield = 43%. IR (cm−1): νNH = 3434, νCO = 1613. H NMR
(DMSO-d6, 25 °C): δ = 3.38 (vbs, 2H, NHCH2Ar), 6.48−7.68
(vbs, 7H, Ar), 8.59 (vbs, 1H, NH). ESI-MS (+, m/z): 284
[MgL2 + H]+, 589 [Mg2(L2)2 + Na]+. Anal. Calcd for
307
dx.doi.org/10.1021/mp400482a | Mol. Pharmaceutics 2014, 11, 304−316

Molecular Pharmaceutics
Article
C28H20F2Mg2N2O6·2.5H2O: C, 54.94; H, 4.12; N, 4.58. Found:
C, 55.02; H, 3.67; N, 4.53.
Potentiometric titrations were carried out by a fully automated
apparatus equipped with a CRISON GLP 21−22 digital
voltmeter (resolution, 0.1 mV) and a 5 mL Metrohm Dosimat
655 autoburet, both controlled by a homemade software in
BASIC, working on an IBM computer. The electrodic chain
(Crison 5250 glass electrode and 0.1 M KCl in methanol/water
= 9:1 v/v calomel electrode, Radiometer 401) was calibrated in
terms of [H+] by means of a strong acid−strong base titration
by Gran’s method,41 allowing the determination of the standard
potential, E° (371.5 0.4 mV), and of the ionic product of
water, Kw (pKw = 14.40 0.05) in the experimental conditions
used. The software HYPERQUAD42 was used to evaluate the
protonation and complexation constants from emf data.
Biology. Production of Recombinant Influenza Virus
PA-Nter Protein. 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),
and the protein was expressed in E. coli BL21-CodonPlus cells
(Agilent Technologies) and purified by 6xHis-Ni-NTA
chromatography,27 followed by buffer exchange.
Enzymatic Assay with Influenza PA-Nter Endonu-
clease. The enzymatic assay was performed according to a
published procedure,13 with some modifications. The reaction
mixture (25 μL volume) contained 1 μg of recombinant PA-
Nter, 1 μg (16.7 nM) of single-stranded circular DNA plasmid
M13mp18 (Affymetrix), buffer (50 mM Tris-HCl pH 8, 100
mM NaCl, 1 mM MnCl2, and 10 mM β-mercaptoethanol), and
serial dilutions of the test compounds. 2,4-Dioxo-4-phenyl-
butanoic acid (DPBA, purchased from Interchim), a diketo acid
compound known to inhibit the influenza virus endonuclease,21
was included as reference compound. After a 1 h incubation at
37 °C, the reaction was 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
ethidium bromide staining. The amount of remaining intact
plasmid was quantified by ImageQuant TL software (GE
Healthcare).
Mg(L3)2·3H2O (9). HL3 (100 mg, 0.25 mmol) in methanol
(8 mL), sodium hydroxide (0.38 mmol, 0.095 mL), and an
aqueous suspension of MgCl2·6H2O (26.43 mg, 0.13 mmol,)
1
were used. Yield = 50%. IR (cm−1): νCO = 1633. H NMR
(DMSO-d6, 25 °C): δ = 4.21 (bs, 4H, NHCH2Ar), 6.22−8.04
(bs, 11H, Ar), 9.25 (bs, 1H, NH), 11.47 (bs, 1H, NH), 12.16
(bs, 2H, NH). ESI-MS (+, m/z): 419 [HL3 + Na]+, 837
[Mg(L3)2 + Na]+. Anal. Calcd for C44H34F4MgN4O6·3H2O: C,
60.81; H, 4.64; N, 6.45. Found: C, 60.70; H, 4.30; N, 6.38. X-
ray diffraction quality crystals were obtained by recrystallization
from methanol as [Mg(L3)2(MeOH)2]·2MeOH.
X-ray Crystallography. Crystals suitable for X-ray
diffraction were obtained by slow evaporation of methanol/
water (HL1 and H2L2) and methanol ([Mg(L3)2(MeOH)2]·
2MeOH) solutions, respectively. X-ray diffraction data were
collected at room temperature on a Bruker APEX II
diffractometer equipped with CCD detector, using graphite
monochromated Mo Kα (λ = 0.71069 Å) radiation. Resulting
crystal decay was negligible. Data were processed by Lorentz
and polarization corrections with the SAINT package,34
corrected for absorption effects by the SADABS35 procedure.
The phase problem was solved by direct methods36 and refined
by full matrix least-squares on all F2 .37 Anisotropic displace-
ment parameters were refined for all non hydrogen atoms,
while hydrogen atoms were introduced in idealized positions;
only the hydrogen atoms of the OH groups of the methanol
molecules in [Mg(L3)2(MeOH)2]·2MeOH were located on
the difference map and refined independently. The final maps
were featureless. Details of data collection and structure
refinement are presented in Table 1. For the discussion, use
was made of the Cambridge Structural Database software38 and
of Olex2.39 CCDC-942580-942582 contains the supplementary
crystallographic data for this paper. These data can be obtained
free of charge via www.ccdc.cam.ac.uk/conts/retrieving.html
(or from the Cambridge Crystallographic Data Centre, 12
Union Road, Cambridge CB2 1EZ, UK; fax: (44) 1223-336-
033; or e-mail: deposit@ccdc.cam.ac.uk).
For each active compound, the percentage inhibition of PA
endonuclease activity was plotted against the compound
concentration on a semilogarithmic plot, using GraphPad
Potentiometry. The Mg(II) stock solutions were prepared
from MgCl2·6H2O (Aldrich). Their concentrations were
determined using EDTA as a titrant and the sodium salt of
Eriochrome black T in the presence of triethanolamine and
hydroxylamine chloride as an indicator. Equilibrium constants
for protonation and complexation reactions were determined
by means of potentiometric measurements in methanol/water
= 9:1 v/v solution at ionic strength 0.1 M KCl, carried out at 25
0.1 °C under nitrogen in the pH range 2.5−11. Temperature
was controlled to 0.1 °C by using a thermostatted circulating
water bath (ISCO GTR 2000 IIx). Appropriate aliquots of
ligand solution, prepared by weight, were titrated with standard
KOH (methanol/water = 9:1 v/v, I = 0.1 M KCl) with and
without metal ions, applying constant speed magnetic stirring.
Freshly boiled methanol and double-distilled water, kept under
nitrogen, were used throughout. The experimental procedure to
reach high accuracy in the determination of the equilibrium
constants in this mixed solvent has been described in detail
elsewhere.40 The protonation constants of HL1, H2L2, and HL3
were obtained by titrating 20−50 mL of samples of each ligand
(3 × 10−3 to 5 × 10−3 M). For the complex formation
constants, the titrations were performed with different ligand/
metal ratios (1 up to 5). At least two measurements (about 60
experimental points each) were performed for each system.
Prism software. Values were the mean
SEM of three
independent experiments. The 50% inhibitory concentrations
(IC50) were obtained by nonlinear least-squares regression
analysis.
Computational Methods. Ligand docking studies were
performed using the Molecular Operating Environment
(MOE) software suite.43 For docking experiments, the crystal
structure of avian influenza virus PAN bound to L-742 001 was
retrieved from the PDB Data Bank (http://www.rcsb.org/ -
PDB code: 4E5H).31 To reduce significant differences between
the A/X-31 and the H5N1 sequence of the X-ray crystal model,
the residue Ala20, located quite near the active site metals, has
been replaced with a threonine (A20T), which was submitted
to a minimization procedure using AMBER99 force field with a
generalized Born solvation model. Moreover, minimization of
Arg84 has been performed to balance the shift of such residue
due to the presence of L-742 001 in the cocrystal structure.
Water molecules have been removed, and the partial charges
were automatically calculated. The structures of compounds
HL1, H2L2, and HL3 were built in MOE and minimized before
the docking, using the MMFF94x force field, with the
systematic algorithms until a RMSD gradient of 0.001 kcal
mol−1 Å−1 was reached. Rigid receptor−flexible ligand (bearing
308
dx.doi.org/10.1021/mp400482a | Mol. Pharmaceutics 2014, 11, 304−316

Molecular Pharmaceutics
Article
constrained chelating groups) docking calculations were
performed using the docking simulation feature MOE-dock
by setting grid sizes that included the entire macromolecule.
The triangle matcher was used as a placement method to
generate docking poses.44 Only the best-scored poses generated
in the docking experiments were retained and examined with
MOE.
Discussion of the X-ray Structures of HL1, H2L2, and
[Mg(L3)2(MeOH)2]·2MeOH. The molecular structures of HL1
and H2L2 are remarkably similar, with a root-mean-square
deviation less than 0.16 Å for the superimposition of the atoms
common to the two molecules (Figure 3). In fact, the main
conformational degrees of freedom for both ligands are the
torsion around the bonds C1−C7, N1−C8, C8−C9, with
values of C2−C1−C7−N1 = 175.5(1), 165.8(2)°; C7−N1−
C8−C9 = −94.0(1), −96.0(2)°; N1−C8−C9−C14 = 73.4(1),
79.8(2)°, respectively, for H2L2 and HL1. In both cases, the
torsion angle around the bond C1−C7 is locked to a planar
conformation by the intramolecular hydrogen bond between
the amidic carbonyl and the o−OH group on the adjacent
aromatic ring (O1...O2 = 2.491(1) and 2.523(2) Å, O−H...O =
158(2) and 147.9(1)°, respectively, for H2L2 and HL1).
Combined with the constrained planarity of the amide group,
this results in the overall planarity of the ligand from C4 to C8.
By contrast, the p-F-phenyl group is in both cases perpendicular
to the rest of the molecule, resulting in a remarkably bent
conformation (Figure 4). The crystal packing of HL1 (Figure 4)
is simply based on the complete balance of hydrogen bond
donors and acceptors present in the molecule, since besides the
intramolecular hydrogen bond, the remaining NH donor links
to the OH acceptor in chains where molecules are translated
along the a axis (N1−H...O1(i) = 2.992(3) Å, 143.4(1)°, i = x
+ 1, y, z). In analogy to HL1, in H2L2 the NH group acts as
donor toward the OH already involved in the intramolecular
hydrogen bond (N1−H...O1(ii) = 2.885(1) Å, 140(2)°, ii = 3/
2 − x, y, z − 1/2), thus forming screw chains where consecutive
molecules are rotated by 180°; the same screw motif is found in
the structure of the N-benzyl-2-hydroxybenzamide.45 The
second OH group links to the amide carbonyl (O3−H...O2(iii)
= 2.675(1) Å, 169.5(1)°, iii = x + 1/2, y + 1/2, 3/2 − z),
originating a three-dimensional network (Figure 4) with all
fluorines segregated in square channels.
RESULTS AND DISCUSSION
■
Synthesis and Characterization of HL1, H2L2, HL3 and
of the Corresponding Magnesium Complexes (7), (8),
and (9). The amides (1)−(3) were obtained with good yields
(Scheme 1) and characterized by the usual spectroscopic tools.
a
Scheme 1. Synthesis of Amide Intermediates
a
Reagents and conditions: (i) carbonyldiimidazole, dichloromethane
dry, r.t. 24 h.
Deprotection of the hydroxyl groups (Scheme 2) was carried
out with a 1 M solution of boron tribromide to obtain HL1,
H2L2, and HL3. The complexes Mg(L1)2·3H2O (7), Mg2(L2)2·
2.5H2O (8), and Mg(L3)2·3H2O (9) were synthesized by
reacting in methanol the deprotonated ligands with magnesium
chloride in a 2:1 ligand-to-metal ratio (Scheme 2). The use of a
base is necessary to ensure deprotonation of the hydroxyl
moiety and subsequent coordination to the metal. In the IR
spectra of all the complexes, the OH absorption disappeared
(2800−3000 cm−1 in the free ligands), and the CO bands
shifted from 1632 to 1644 cm−1 in HL1, H2L2, and HL3 to
1613−1614 cm−1 in the complexes because of the involvement
of the amide group in coordination. In the 1H NMR spectra of
the magnesium complexes, a broadening and a downfield shift
of the signals of about 0.2 ppm was observed versus the free
ligands, and the resonance of the OH proton, detectable in the
free ligand, disappeared. In particular, in complex (8), both the
hydroxyl groups are deprotonated, and all data are in agreement
The molecular structure of [Mg(L3)2(MeOH)2]·2MeOH,
displayed in Figure 5, is a neutral centrosymmetric octahedral
complex, with two deprotonated ligands on the equatorial plane
and two coordinated methanol molecules at the apexes (Table
2). Two methanol molecules cocrystallize in the complex,
enriching the packing network. Coordination occurs by
chelation of the deprotonated OH and one of the two amidic
carbonyl oxygens on each ligand, while the second amidic C
O is free and points away from the metal in order to allow the
formation of an intramolecular hydrogen bond between the
NH and the central coordinated oxygen (N2−H...O2 =
2.591(1) Å, 136.9(1)°). The two ligand arms show therefore
opposite configuration with respect to the C9−C8 and C13−
C15 equivalent bonds (C10−C9−C8−N1 = −171.6(1)° for
the coordinated amide oxygen, and C12−C13−C15−O3 =
1
with a 2:2 metal to ligand ratio. In the H NMR spectrum of
(9), it is also possible to distinguish the signals of the two
different arms of the ligand, the coordinated and the
uncoordinated one. Elemental analysis and mass spectra
confirm the proposed stoichiometry.
a
Scheme 2. Synthesis of the Ligands and the Corresponding Magnesium Complexes
a
Reagents and conditions: (i) BBr3, dichloromethane dry, T = 0 °C to r.t., 24 h; (ii) NaOH 2 M r.t. 30 min, methanol; (iii) MgCl2 r.t. 5 h.
309
dx.doi.org/10.1021/mp400482a | Mol. Pharmaceutics 2014, 11, 304−316

Molecular Pharmaceutics
Article
Figure 3. Molecular structures and labeling of HL1 (top) and H2L2 (bottom), with thermal ellipsoids drawn at the 50% probability level.
with different orientations relatively to the NH vectors: H1−
N1−C7−C4 = −12.9(2)° for the chelating ligand arm, and
H2−N2−C16−C17 = −92.4(1)° for the free arm.
Potentiometric Measurements and Calculations. The
metal-binding behavior of some hydroxybenzamides has been
previously investigated by computational docking studies to
elucidate their interactions with the active site of the prototype
foamy virus IN,30 but nothing is known about the speciation of
this class of ligand with magnesium. This information is
essential in order to rationalize their inhibitory ability and
mechanism of action. In this perspective, we carried out
potentiometric studies about the solution behavior of HL1,
H2L2, and HL3 with Mg2+. To be consistent with our previous
studies,46 the titrations have been performed in mixed solvent
methanol/water = 9/1 and ionic strength 0.1 M KCl.
Figure 4. Hydrogen bond network in HL1 (left) and H2L2 (right).
The round inset highlights the screw chain motif in H2L2.
The acidity (pKa) of the salicylic OH of the ligands resulted
9.51(0.01) for HL1, 9.10(0.01) for H2L2, and 7.28(0.01) for
HL3, in which the presence of two identical amido moieties in
ortho-position to the acidic OH allows for a greater
delocalization of the charge of the anion. The second acidity
constant (pKa2) for H2L2 cannot be correctly evaluated by
potentiometry because it is greater than 12 and so out of the
8.2(2)° for the free amide). The six-membered chelation rings
are practically planar, with a slight out of plane deformation at
O1. The intramolecular NH...O hydrogen bond, together with
the chelation and the intrinsic planarity of the amide moieties,
renders the central core of the complex quite planar (maximum
deviation for the skeleton from C7 to C16 = 0.27 Å for O1).
The four p-F-phenyl rings point away from this mean plane,
310
dx.doi.org/10.1021/mp400482a | Mol. Pharmaceutics 2014, 11, 304−316

Molecular Pharmaceutics
Article
Figure 5. Molecular structure of the neutral complex [Mg(L3)2(MeOH)2]·2MeOH, with thermal ellipsoids drawn at the 50% probability level. The
two cocrystallized methanol molecules are omitted.
Table 2. Coordination Bond Lengths (Å) and Angles (°) of
Magnesium in [Mg(L3)2(MeOH)2]·2MeOH with SD in
Parentheses
Table 3. Logarithms of Formation Constants (βpqr =
[MpLqHr]/[M]p[L]q[H]r) in Methanol/Water = 9:1 v/v, I =
0.1 M KCl at 25 °C for the ligands under study with Mg2+.
SDs are given in parentheses. Charges are omitted for
simplicity
MgO1
1.9997(8)
1.9997(8)
1.9813(8)
1.9813(8)
2.153(1)
2.153(1)
180.00
MgO11
MgO2
HL1
MgO21
p
q
r
Mg2+
MgO41
1
1
1
2
−1
−2
−7.98 (0.06)
−16.75 (0.48)
MgO4
O11 MgO1
O1MgO4
O1MgO41
O11MgO41
O11MgO4
O21MgO11
O2MgO11
O21MgO1
O2MgO1
O2MgO21
O21MgO41
O2MgO4
O21MgO4
O2MgO41
O4MgO41
L + H = LH
9.51 (0.01)
H2L2
89.91(4)
90.09(4)
89.91(4)
90.09(4)
86.57(3)
93.43(3)
93.43(3)
86.57(3)
180.00
p
q
r
Mg2+
1
1
1
1
0
2.71 (0.08)
1
−5.53 (0.06)
LH + H = LH2
9.10 (0.01)
HL3
p
q
r
Mg2+
1
0
1
2
−1
0
−9.84 (0.07)
2.66 (0.16)
5.85 (0.09)
90.22(4)
90.22(4)
89.78(4)
89.78(4)
180.00
1
1
0
L + H = LH
7.28 (0.01)
1
1
sphere, rather than bulkier ligands. This also explains why many
magnesium salts are extensively hydrated in the crystalline
form, with the anion located in the outer coordination sphere.
Crystallographic data showed that outer sphere coordination by
hexahydrate magnesium is the normal binding mode of Mg2+ to
oligonucleotides.47,48 Moreover, being a “hard” ion, it prefers
binding to “hard” oxygen-containing ligands, such as carbonyls,
carboxylates, phosphates, hydroxyls, and water.
Symmetry operator: 1 − x, 1 − y, −z
limits of reliability of the glass electrode. The best fit of the
experimental titration curves, carried out in presence of the
metal ion, was obtained by the sets of species shown in Table 3.
The proposed models are corroborated by very good statistical
parameters and by nice overlapping between experimental and
computed titration curves. In order to understand the
speciation in solution, it must be recalled that the wide
prevalence of magnesium as the metal cofactor for enzymes is
due not only to its high natural abundance but also to its
peculiar chemical−physical properties.47−49 The Mg2+ ion is
characterized by a small ionic radius, high charge density, and
tendency to bind water molecules in the inner coordination
Despite the similar pKa values of the salicylic OH, HL1 and
H2L2 behave very differently in solution with Mg2+ ions (see
Table 3). For HL1, the species [MgLH−1] and [MgL2H−2]2−
were found. The species distribution plot for HL1 in Figure 6a
clearly shows that the metal is not bound at physiological pH;
rather, these complexes are formed in basic solution above pH
9 at which one or two water molecules of the inner
311
dx.doi.org/10.1021/mp400482a | Mol. Pharmaceutics 2014, 11, 304−316

Molecular Pharmaceutics
Article
Figure 6b shows that the [MgHL]+1 complex changes into
[MgL], due to the acid dissociation of the second OH of the
ligand, so forming an inner sphere chelate in a very cooperative
manner. The presence on the ligand of three oxygens, over
which the charges derived by the dissociation of the two OH
groups are delocalized, is the condition favoring the formation
of an inner sphere complex.
The situation is different for HL3, which bears only one
acidic OH between two amidic groups. It is important to
underline that its coordination ability toward magnesium ions
seems to be lower than that of H2L2. The best fit is obtained by
a model considering the species [MgL]−1, [MgL2], and
[MgH−1]−1. The last one is a hydroxide species, necessary in
order to achieve convergence of the fit and obtain good
statistical parameters. The distribution diagram of Figure 6c
shows that the complex formation starts at a pH about 5,
indicating that the interaction with Mg2+ lowers the pKa value
of the ligand, but, notwithstanding the quite high ligand/metal
ratio, the metal is never completely complexed by HL3 and its
concentration decreases when the hydroxide species starts
forming. This behavior does not allow the discrimination
between an inner or outer sphere complex, but, considering
that HL3 is a bulky ligand, the last hypothesis appears more
likely.
Inhibition of PA-Nter Endonuclease Activity. The
hydroxybenzamide-based ligands and their complexes were
tested for their ability to inhibit the endonuclease activity of
PA-Nter in an enzymatic assay with recombinant PA-Nter. Two
compounds, H2L2 and the corresponding magnesium complex
(8), showed a dose-dependent inhibition of the endonuclease
activity (Figure 7), thus demonstrating the effectiveness of this
novel 2,3-dihydroxybenzamide scaffold. Interestingly, complex
(8) proved to be the most active compound, with an IC50 value
of 18 μM, which is 6-fold higher than that of the reference
compound DPBA. The uncomplexed ligand H2L2 (IC50 = 33
μM) had an 11-fold lower potency than DPBA. The IC50 value
that we obtained for DPBA (3 μM) is comparable to that
published by others,50 supporting the relevance of our enzyme
inhibition data. The other four compounds (HL1, HL3, and the
complexes (7) and (9)) did not inhibit the endonuclease
reaction at 500 μM, the highest concentration tested.
It is worth noting that the complex (8) displayed a 2-fold
higher potency than its free ligand H2L2. This suggests that the
effective inhibitor of the activity of the enzyme could be a metal
complex and not a free ligand. In fact, the activity of H2L2 could
be due to the formation of the corresponding complex, when it
is added to the enzymatic assay in the presence of divalent
metal ions.
Molecular Modeling. To understand the binding mode of
HL1, H2L2, and HL3 to PA-Nter protein, and to propose a
plausible mechanistic hypothesis, preliminary ligand−receptor
docking studies were conducted. The coordinates for PA-Nter
protein (PDB code: 4E5H) were used for computational
docking. On the basis of the structure and the pKa values of the
model ligands (see potentiometry section), we built com-
pounds HL1, H2L2, and HL3 in monodeprotonated form.
Graphical representations of top-ranking binding modes
obtained for these ligands are depicted in Figure 8. Analysis
of the docking results showed different binding modes with
variable affinities within the amino acid pocket located near the
catalytic site that includes the following residues: Thr20, Glu23,
Tyr24, Glu26, Lys34, Ile38, His41, Glu80, Gly81, Arg82,
Asp83, Arg84, Thp88, Lys104, Phe105, Leu106, Asp108,
Figure 6. Distribution diagrams for the systems under investigation at
L:Mg = 4:1 (the concentration of Mg2+ is 25 mM): (a) L = HL1, (b) L
= H2L2, (c) L = HL3.
coordination sphere loose a proton, justifying the negative
index of H in the above formulas. Moreover, the complexes are
formed when the ligand is in the deprotonated form, which
means that the acidity of the ligand is not affected by the
presence of Mg2+ ion, indicating an outer sphere coordination.
For H2L2, the species [MgHL]+1 and [MgL] give rise to the
best fit of the experimental data. Figure 6b outlines the
difference in respect to HL1. In fact, the formation of the
complexes starts at pH below 7, showing that metal ion
coordination lowers the pKa value of the salicylic OH. This
strong effect suggests an inner sphere coordination. Moreover,
312
dx.doi.org/10.1021/mp400482a | Mol. Pharmaceutics 2014, 11, 304−316

Molecular Pharmaceutics
Article
Figure 7. Dose−response curves for inhibition of influenza PA-Nter
endonuclease activity by the two active compounds, H2L2 and its
magnesium complex 8. DPBA was included as the reference
compound21. Values represent the percentage inhibition (versus the
no compound control) and are the mean SEM of three independent
tests.
Phe117, Glu119, Ile120, Gly121, Val122, Tyr130, Lys134,
Lys137, and Ile138 (Figure 8, A−C). The residues His41,
Asp108, Glu119, and Lys134 are involved in metal binding and
play a role in the catalytic process. In particular, the most
favorable conformation places HL1 in proximity to His41,
located near the metal ions (Figure 8A). The inactivity of HL1
in the enzymatic assay may be explained by the observation that
this compound does not tightly interact with the metal ion and
that the flexibility and the orientation of the p-fluorobenzyl
moiety away from the pocket could reduce the formation of
favorable interactions (such as π-stacking) with the hydro-
phobic residues.
Figure 8. Representative view of in silico docking of HL1, H2L2, and
HL3 within their putative binding site in the active site of PA-Nter.
(A−C): Hypothetical disposition of the best energy docking poses.
The full PA-Nter protein is represented by cartoon and top-ranked
energy poses as cyan (HL1), yellow (H2L2), and magenta (HL3) sticks.
The two metal ions are orange.
On the other hand, in all docking clusters, H2L2 was found to
bind to one metal ion cofactor through the catechol group, and
it also establishes a coordination bond with the second cation
by a combination of the α-hydroxyl and the amide-linked
carbonyl groups (Figure 8B). Moreover, for the binding pose
with the most favorable binding energies, the fluorobenzyl
moiety of H2L2 directs toward Arg84, similarly to L-742 001,
where its benzyl ring is positioned in a narrow hydrophobic
cavity comprising Arg84, Trp88, Phe105, and Leu106 (Figure
9). Interestingly, the chelating motif of H2L2 strictly overlaps
313
dx.doi.org/10.1021/mp400482a | Mol. Pharmaceutics 2014, 11, 304−316

Molecular Pharmaceutics
Article
8 displayed a 2-fold higher potency than its free ligand H2L2. As
far as we know, it is the first time that an inhibition of the PA-
Nter endonuclease activity by a metal complex has been
reported. The results of the potentiometric studies and the
molecular modeling analysis suggest that H2L2 could interact
effectively with the two metal ions in the active site of the
enzyme, by using the catechol and the amide oxygen, and, in
this way, it could exercise its biological activity. Regarding the
activity of 8, it is not clear whether this compound acts in the
form of the starting complex, since it may rearrange under the
conditions of the enzymatic assay. However, since the enzyme
assay is performed with 1 mM MnCl2, we find it unlikely that
the ligand H2L2 would be uncomplexed.
On the basis of the present data, we propose that the 2,3-
dihydroxyamide chemotype represents a versatile platform for
designing influenza PA endonuclease inhibitors, and studies are
underway to explore this scaffold further using extensive
chemical modifications.
Figure 9. Binding model of H2L2 (yellow) compared with the
disposition of L-742 001 (gray) into the L-742 001-PA-Nter complex
(PDB entry 4E5H).30 The two metal ions are orange.
AUTHOR INFORMATION
Corresponding Author
*Tel.: +39 0521 905427.
■
Author Contributions
with the three coplanar oxygen atoms of L-742 001 (i.e., a
carboxylate oxygen, a lone oxygen pair of the α-hydroxyl group,
and the oxygen atom at the γ-position), which are involved in
metal chelation of the divalent ions.
The manuscript was written through contributions of all
authors. All authors have given approval to the final version of
the manuscript.
Notes
Finally, HL3 behaves similarly to HL1, since its bulky planar
conformation does not result in any significant contacts with
amino acid residues in the active site pocket (Figure 8C). This
confirms that the PA-Nter active site is quite large and relatively
shallow and therefore able to accommodate large variations in
ligand size and shape with different binding affinities. Moreover,
the two rotatable p-fluorobenzyl amides substituents at the
ortho-position on the hydroxyl phenol group do not readily
engage in a correct bridging mode for metal ion coordination
and this may have a negative effect on the activity of this
compound.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
“Centro Interdipartimentale Misure Giuseppe Casnati” and
“Laboratorio di Strutturistica Mario Nardelli” of the University
of Parma are thanked for facilities. The authors thank Dr.
Andrea Brancale and Dr. Nicolino Pala for their assistance on
the elaboration of the molecular modeling calculation. M.S. is
grateful to “Fondazione Banco di Sardegna” for its partial
financial support. A.S. is holder of a PhD grant from the Agency
for Innovation by Science and Technology (IWT). This study
was supported by a grant from the Geconcerteerde
Onderzoeksacties (GOA/10/014) from the KU Leuven.
On the basis of these results, the inhibitory effect of H2L2 can
be explained by its arrangement in the active site of PA-Nter. In
particular, the “two-metal” coordination of the cations by a
unique “O,O,O” chelating triad, which is formed by an optimal
combination of the catecholic and the carbonyl oxygens,
appears as the important determinant for the inhibitory activity
toward PA-Nter. This chelating motif therefore represents a
relevant scaffold for designing novel inhibitors of the influenza
virus PA endonuclease.
REFERENCES
■
(1) Reid, A. H.; Taubenberger, J. K.; Fanning, T. G. The 1918
Spanish influenza: integrating history and biology. Microbes Infect.
2001, 3, 81−87.
(2) Smith, G. J. D.; Vijaykrishna, D.; Bahl, J.; Lycett, S. J.; Worobey,
M.; Pybus, O. G.; Ma, S. K.; Cheung, C. L.; Raghwani, J.; Bhatt, S.;
Peiris, J. S. M.; Guan, Y.; Rambaut, A. Origins and evolutionary
genomics of the 2009 swine-origin H1N1 influenza A epidemic. Nature
2009, 459, 1122−1125.
(3) Chowell, G.; Bertozzi, S. M.; Colchero, M. A.; Lopez-Gatell, H.;
Alpuche-Aranda, C.; Hernandez, M.; Miller, M. A severe respiratory
disease concurrent with the circulation of H1N1 influenza. N. Engl. J.
Med. 2009, 361, 674−679.
(4) Garten, R. J.; Davis, C. T.; Russell, C. A; Shu, B.; Lindstrom, S.;
Balish, A.; Sessions, W. M.; Xu, X.; Skepner, E.; Deyde, V.; Okomo-
Adhiambo, M.; Gubareva, L.; Barnes, J.; Smith, C. B.; Emery, S. L.;
Hillman, M. J.; Rivailler, P.; Smagala, J.; De Graaf, M.; Burke, D. F.;
CONCLUSIONS
■
This study was focused on the design, synthesis, and biological
evaluation of a chemical scaffold containing a chelating motif
able to bind one or two metal ions in the PA-Nter active site.
Analysis of the potentiometric and X-ray data demonstrated the
metal-complexing ability of 2-hydroxyamide-based scaffolds. All
three ligands (HL1, H2L2, and HL3) are found to chelate Mg2+
ions, forming complexes with different stoichiometric ratios.
However, their coordination ability was found to be different.
In particular, at physiological pH, potentiometric studies
proved that H2L2 is the most efficient one, and this also
resulted in inhibition of the PA-Nter endonuclease activity.
Two compounds, H2L2 and the corresponding magnesium
complex 8, proved to be effective inhibitors of the PA-Nter
endonuclease activity, and it is interesting to note that complex
Fouchier, R. A. M.; Pappas, C.; Alpuche-Aranda, C. M.; Lop
́
ez-Gatell,
H.; Olivera, H.; Lopez, I.; Myers, C. A.; Faix, D.; Blair, P. J.; Yu, C.;
́
Keene, K. M.; Dotson, P. D.; Boxrud, D.; Sambol, A. R.; Abid, S. H.; St
George, K.; Bannerman, T.; Moore, A. L.; Stringer, D. J.; Blevins, P.;
Demmler-Harrison, G. J.; Ginsberg, M.; Kriner, P.; Waterman, S.;
Smole, S.; Guevara, H. F.; Belongia, E. A.; Clark, P. A.; Beatrice, S. T.;
314
dx.doi.org/10.1021/mp400482a | Mol. Pharmaceutics 2014, 11, 304−316

Molecular Pharmaceutics
Article
Donis, R.; Katz, J.; Finelli, L.; Bridges, C. B.; Shaw, M.; Jernigan, D. B.;
Uyeki, T. M.; Smith, D. J.; Klimov, A. I.; Cox, N. J. Antigenic and
genetic characteristics of swine-origin 2009 A(H1N1) influenza viruses
circulating in humans. Science 2009, 325, 197−201.
(5) Gao, R.; Cao, B.; Hu, Y.; Feng, Z.; Wang, D.; Hu, W.; Chen, J.;
Jie, Z.; Qiu, H.; Xu, K.; Xu, X.; Lu, H.; Zhu, W.; Gao, Z.; Xiang, N.;
Shen, Y.; He, Z.; Gu, Y.; Zhang, Z.; Yang, Y.; Zhao, X.; Zhou, L.; Li, X.;
Zou, S.; Zhang, Y.; Li, X.; Yang, L.; Guo, J.; Dong, J.; Li, Q.; Dong, L.;
Zhu, Y.; Bai, T.; Wang, S.; Hao, P.; Yang, W.; Zhang, Y.; Han, J.; Yu,
H.; Li, D.; Gao, G. F.; Wu, G.; Wang, Y.; Yuan, Z.; Shu, Y. Human
infection with a novel avian-origin influenza A (H7N9) virus. N. Engl.
J. Med. 2013, 368, 1888−1897.
(6) Krug, R. M.; Aramini, J. M. Emerging antiviral targets for
influenza A virus. Trends Phramacol. Sci. 2009, 30, 269−277.
(7) Du, J.; Cross, T. A.; Zhou, H.-X. Recent progress in structure-
based anti-influenza drug design. Drug Discovery Today 2012, 17,
1111−1120.
(8) Reece, P. A. Neuraminidase inhibitor resistance in influenza
viruses. J. Med. Virol. 2007, 1586, 1577−1586.
(9) Collins, P. J.; Haire, L. F.; Lin, Y. P.; Liu, J.; Russell, R. J.; Walker,
P. A.; Skehel, J. J.; Martin, S. R.; Hay, A. J.; Gamblin, S. J. Crystal
structures of oseltamivir-resistant influenza virus neuraminidase
mutants. Nature 2008, 453, 1258−1261.
(10) Honda, A.; Ishiama, A. Biol. Chem. 1997, 378, 483−488.
(11) Honda, A.; Mizumoto, K.; Ishihama, A. Minimum molecular
architectures for transcription. Proc. Natl. Acad. Sci. U.S.A. 2002, 99,
1−6.
(12) Plotch, S. J.; Bouloy, M.; Ulmanen, I.; Krug, R. M. Cell 1981, 23,
847−858.
́
(13) Dias, A.; Bouvier, D.; Crepin, T.; McCarthy, A.; Hart, D. J.;
Flutimide, a novel endonuclease inhibitor of influenza virus.
Tetrahedron Lett. 1995, 36, 2005−2008.
(25) Cianci, C.; Chung, T. D. Y.; Meanwell, N.; Putz, H.; Hagen, M.;
Colonno, R. J.; Krystal, M. Identification of N-hydroxamic acid and N-
hydroxyimide compounds that inhibit the influenza virus polymerase.
Antiviral Chem. Chemother. 1996, 7, 353−360.
(26) Parkes, K. E.; Ermert, P.; Fassler, J.; Ives, J.; Martin, J. A.;
̈
Merrett, J. H.; Obrecht, D.; Williams, G.; Klumpp, K. Use of a
pharmacophore model to discover a new class of influenza
endonuclease inhibitors. J. Med. Chem. 2003, 1153−1164.
(27) Iwai, Y.; Murakami, K.; Gomi, Y.; Hashimoto, T.; Asakawa, Y.
Anti-influenza activity of marchantins, macrocyclic bisbibenzyls
contained in liverworts. PLoS One 2011, 6, e19825.
(28) Kuzuhara, T.; Iwai, Y.; Takahashi, H.; Hatakeyama, D.; Echigo,
N.; Bunri, T. Green tea catechins inhibit the endonuclease activity of
influenza A virus RNA polymerase. PLoS One Currents 2012, 1−8.
(29) Song, J.-M.; Lee, K.-H.; Seong, B.-L. Antiviral effect of catechins
in green tea on influenza virus. Antiviral Res. 2005, 68, 66−74.
(30) Agrawal, A.; DeSoto, J.; Fullagar, J. L.; Maddali, K.; Rostami, S.;
Richman, D. D.; Pommier, Y.; Cohen, S. M. Probing chelation motifs
in HIV integrase inhibitors. Proc. Natl. Acad. Sci. U.S.A. 2012, 109,
2251−2256.
(31) Dubois, R. M.; Slavish, P. J.; Baughman, B. M.; Yun, M.; Bao, J.;
Webby, R. J.; Webb, T. R.; White, S. W. Structural and biochemical
basis for development of influenza virus inhibitors targeting the PA
endonuclease. PLoS Pathogens 2012, 8 (8), e1002830.
(32) Kowalinski, E.; Zubieta, C.; Wolkerstorfer, A.; Szolar, O. H. J.;
Ruigrok, R. W. H.; Cusack, S. Structural analysis of specific metal
chelating inhibitor binding to the endonuclease domain of influenza
pH1N1 (2009) Polymerase. PLoS Pathogens 2012, 8 (8), e1002831.
(33) Fan, X.; Zhang, F.-H.; Al-Safi, R. I.; Zeng, L.-F.; Shabaik, Y.;
Debnath, B.; Sanchez, T. W.; Odde, S.; Neamati, N.; Long, Y.-Q.
Design of HIV-1 integrase inhibitors targeting the catalytic domain as
well as its interaction with LEDGF/p75: a scaffold hopping approach
using salicylate and catechol groups. Bioorg. Med. Chem. 2011, 19,
4935−4952.
Baudin, F.; Cusack, S.; Ruigrok, R. W. H. The cap-snatching
endonuclease of influenza virus polymerase resides in the PA subunit.
Nature 2009, 458, 914−918.
(14) Yuan, P.; Bartlam, M.; Lou, Z.; Chen, S.; Zhou, J.; He, X.; Lv, Z.;
Ge, R.; Li, X.; Deng, T.; Fodor, E.; Rao, Z.; Liu, Y. Crystal structure of
an avian influenza polymerase PA N reveals an endonuclease active
site. Nature 2009, 458, 909−913.
(34) SAINT: SAX, Area Detector Integration. Siemens Analytical
instruments, Inc.: Madison, WI.
(35) SADABS: Area-Detector Absorption Correction. Siemens
́
(15) Crepin, T.; Dias, A.; Palencia, A.; Swale, C.; Cusack, S.; Ruigrok,
R. W. H. Mutational and metal binding analysis of the endonuclease
domain of the influenza virus polymerase PA subunit. J. Virol. 2010,
84, 9096−9104.
Industrial Automation, Inc.: Madison, WI, 1996.
(36) Burla, M. C.; Caliandro, R.; Camalli, M.; Carrozzini, B.;
Cascarano, G. L.; De Caro, L.; Giacovazzo, C.; Polidori, G.; Spagna, R.
SIR2004: an improved tool for crystal structure determination and
refinement. J. Appl. Crystallogr. 2005, 38, 381−388.
(37) SHELXL97; Sheldrick, G. M. Acta Crystallogr. 2008, A64, 112.
(38) Allen, F. H. The Cambridge Structural Database: a quarter of a
million crystal structures and rising. Acta Crystallogr. 2002, B58, 380−
388.
(39) Dolomanov, O. V.; Bourhis, L. J.; Gildea, R. J.; Howard, J. A. K.;
Puschmann, H. OLEX2: a complete structure solution, refinement and
analysis program. J. Appl. Crystallogr. 2009, 42, 339−341.
(40) Fisicaro, E.; Braibanti, A. Talanta 1988, 10, 769.
(16) Steitz, T.; Steitz, J. A general two-metal-ion mechanism for
catalytic RNA. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 6498−6502.
(17) Doan, L.; Handa, B.; Roberts, N.; Klumpp, K. Metal ion
catalysis of RNA cleavage by the influenza virus endonuclease.
Biochemistry 1999, 38, 5612−9.
(18) Rogolino, D.; Carcelli, M.; Sechi, M.; Neamati, N. Viral enzymes
containing magnesium: metal binding as a successful strategy in drug
design. Coord. Chem. Rev. 2012, 256, 3063−3086.
(19) Kirschberg, T.; Parrish, J. Metal chelators as antiviral agents.
Curr. Opin. Drug Discovery Dev. 2007, 10, 460−472.
(20) Liao, C.; Marchand, C.; Burke, T. R., Jr; Pommier, Y.; Nicklaus,
M. C. NIH Public Access. Future Med. Chem. 2010, 2, 1107−1122.
(21) Tomassini, J.; Selnick, H.; Davies, M. E.; Armstrong, M. E.;
Baldwin, J.; Bourgeois, M.; Hastings, J.; Hazuda, D.; Lewis, J.;
Mcclements, W.; Ponticello, G.; Radzilowski, E.; Smith, G.; Tebben,
A.; Wolfe, A. Inhibition of cap (m7GpppXm)-dependent endonuclease
of influenza virus by 4-substituted 2,4-dioxobutanoic acid compounds.
Antimicrob. Agents Chemother. 1994, 38, 2827−2837.
(22) Hastings, J. C.; Selnick, H.; Wolanski, B.; Tomassini, J. E. Anti-
influenza virus activities of 4-substituted 2,4-dioxobutanoic acid
inhibitors. Antimicrob. Agents Chemother. 1996, 40, 1304−1307.
(23) Tomassini, J. E.; Davies, M. E.; Hastings, J. C.; Lingham, R.;
Mojena, M.; Raghoobar, S. L.; Singh, S. B.; Tkacz, J. S.; Goetz, M. A
novel antiviral agent which inhibits the endonuclease of influenza
viruses. Antimicrob. Agents Chemother. 1996, 40, 1189−1193.
(24) Hensens, O. D.; Goetz, M. A.; Liesch, J. M.; Zink, D. L.;
Raghoobar, S. L.; Helms, G. L.; Singh, S. B. Isolation and structure of
(41) Gran, G. Analyst 1952, 77, 661.
(42) Gans, P.; Sabatini, A.; Vacca, A. Talanta 1996, 43, 1739.
(43) Molecular Operating Environment (MOE 2010.11). Chemical
Computing Group, Inc.: Montreal, Quebec, Canada.
(44) Feher, M.; Williams, C. I. J. Chem. Infect. Model 2009, 49, 1704.
(45) Zhang, Q. X.; Zhang, B. Acta Crystallogr. 2008, E 64, o884.
(46) Sechi, M.; Bacchi, A.; Carcelli, M.; Compari, C.; Duce, E.;
Fisicaro, E.; Rogolino, D.; Gates, P.; Derudas, M.; Al-Mawsawi, L. Q.;
Neamati, N. From ligand to complexes: Inhibition of human
immunodeficiency virus type 1 integrase by β-diketo acid metal
complexes. J. Med. Chem. 2006, 49, 4248−4260.
(47) Cowan, J. Metal activation of enzymes in nucleic acid
biochemistry. Chem. Rev. 1998, 98, 1067−1088.
(48) Dudev, T.; Cowan, J.; Lim, C. Competitive binding in
magnesium coordination chemistry: water versus ligands of biological
interest. J. Am. Chem. Soc. 1999, 121, 7665−7673.
315
dx.doi.org/10.1021/mp400482a | Mol. Pharmaceutics 2014, 11, 304−316

Molecular Pharmaceutics
Article
(49) Cowan, J. Structural and catalytic chemistry of magnesium-
dependent enzymes. Biometals: an international journal on the role of
metal ions in biology, biochemistry, and medicine 2002, 15, 225−235.
(50) Noble, E.; Cox, A.; Deval, J.; Kim, B. Endonuclease substrate
selectivity characterized with full-length PA of influenza A virus
polymerase. Virology 2012, 433, 27−34.
316
dx.doi.org/10.1021/mp400482a | Mol. Pharmaceutics 2014, 11, 304−316