Bifunctional 3-hydroxy-4-pyridinones as effective aluminium chelators: synthesis, solution equilibrium studies and in vivo evaluation

Article abstract of DOI:10.1016/j.jinorgbio.2018.05.017

This paper reports the results on the study of a set of synthesized bifunctional 3-hydroxy-4-pyridinones chelators as potential aluminium sequestering agents. They were N-functionalized with alkyl-amino, -carboxylic and –(amino-carboxylic) groups, envisaging the improvement of the Al³⁺ sequestering capacity, in comparison with the marketed chelating drug deferiprone. The main focus of this work was given to the assessment of their binding ability towards Al³⁺, which was studied by potentiometric and UV–Vis spectrophotometric measurements carried out at T = 298.15 K. The speciation models were characterized by Al_pL_qH_r ^(3p+r?qz) species with different stoichiometry. Depending on ligand side-chain structures and on their thermodynamic properties, different trends of stability was found. Furthermore, the sequestering ability of the ligands towards Al³⁺ was investigated by the calculation of pL_0.5 values at different experimental conditions. These results clearly indicate that the presence of amino-carboxylic groups in the ligands increases the sequestering ability towards Al³⁺. The in silico evaluation of pharmacokinetic descriptors indicated no violation to the Lipinski's rule and drug-likeness properties. Furthermore, the in vivo bioassays on a model of metal-overload mice showed for three investigated ligands a higher metal-sequestering capacity than for the chelating drug deferiprone, thus suggesting their potential interest as Al-chelating drug candidates.

Full text of DOI:10.1016/j.jinorgbio.2018.05.017

Journal of Inorganic Biochemistry 186 (2018) 116–129
Contents lists available at ScienceDirect
Journal of Inorganic Biochemistry
journal homepage: www.elsevier.com/locate/jinorgbio
Bifunctional 3-hydroxy-4-pyridinones as effective aluminium chelators:
synthesis, solution equilibrium studies and in vivo evaluation
a
a
b
a
a
Anna Irto , Paola Cardiano , Karam Chand , Rosalia Maria Cigala , Francesco Crea ,
a
c
a,⁎
b,⁎
Concetta De Stefano , Lurdes Gano , Silvio Sammartano , Maria Amélia Santos
Dipartimento di Scienze Chimiche, Biologiche, Farmaceutiche e Ambientali, Università di Messina, Viale F. Stagno d'Alcontres, 31, 98166 Messina, Italy
Centro de Química Estrutural, Instituto Superior Técnico, Universidade de Lisboa, Av. Rovísco Pais 1, 1049-001 Lisboa, Portugal
a
b
c
Centro de Ciências e Tecnologias Nucleares (C2TN), Instituto Superior Técnico, Universidade de Lisboa, Estrada Nacional 10, 2695-066 Bobadela LRS, Portugal
A R T I C L E I N F O
A B S T R A C T
Keywords:
-Hydroxy-4-pyridinones
Speciation
This paper reports the results on the study of a set of synthesized bifunctional 3-hydroxy-4-pyridinones chelators
3
as potential aluminium sequestering agents. They were N-functionalized with alkyl-amino, -carboxylic and
3+
–
(amino-carboxylic) groups, envisaging the improvement of the Al sequestering capacity, in comparison with
Protonation
the marketed chelating drug deferiprone. The main focus of this work was given to the assessment of their
3+
Al
complexation
3+
binding ability towards Al , which was studied by potentiometric and UV–Vis spectrophotometric measure-
Sequestering ability
In vivo chelation
(
3p+r−qz)
ments carried out at T = 298.15 K. The speciation models were characterized by Al
p
L
H
q r
species with
different stoichiometry. Depending on ligand side-chain structures and on their thermodynamic properties,
different trends of stability was found. Furthermore, the sequestering ability of the ligands towards Al3+ was
investigated by the calculation of pL0.5 values at different experimental conditions. These results clearly indicate
3+
that the presence of amino-carboxylic groups in the ligands increases the sequestering ability towards Al . The
in silico evaluation of pharmacokinetic descriptors indicated no violation to the Lipinski's rule and drug-likeness
properties. Furthermore, the in vivo bioassays on a model of metal-overload mice showed for three investigated
ligands a higher metal-sequestering capacity than for the chelating drug deferiprone, thus suggesting their po-
tential interest as Al-chelating drug candidates.
1. Introduction
developed, i.e. the family of 3-hydroxy-4-pyridinones (3,4-HPs), deri-
vatives of 1,2-dimethyl-3-hydroxy-4-pyridinone, commercially known
The treatment of diseases related to the accumulation of hard metal
as deferiprone (DFP) or Ferriprox®, and approved as oral drug for the
treatment of patients affected by iron overload. These molecules are a
class of bidentate compounds, characterized by an aromatoid N-het-
erocycle with hydroxyl and ketone groups in ortho position. Unlike
DFO, 3,4-HPs can be effective in all biological conditions, do not cause
undesired effects and are also economically affordable [10–13]. These
products can be synthesized from maltol, a natural origin compound
[14].
cations in the human body is based on the chelating therapy. This ap-
proach consists in the administration of chelators to patients, to induce
3
+
3+
the sequestration of the metal ions (e.g. Fe , Al , etc.) and their
systemic excretion [1–4]. Intake of metals in the human body can occur
through the diet, the environment or other external sources; once ab-
sorbed, they reach the blood and human organs, competing with other
essential metals for vital functions. For a long time the Deferoxamine
(
DFO or Desferal®), was used as chelator in the treatment of iron
The 3,4-HP herein studied are extra-functionalized (see Fig. 1), en-
visaging the improvement of their affinity towards biological sites and
drug-likeness properties. Although some of these compounds (L1: 4-(3-
hydroxy-2-methyl-4-oxopyridin-1(4H)-yl)butanoic acid; L5: 1-(3-ami-
nopropyl)-3-hydroxy-2-methylpyridin-4(1H)-one; L4: (S)-2-amino-5-(3-
hydroxy-2-methyl-4-oxopyridin-1(4H)-yl)pentanoic acid) have already
been studied by some of the authors [15–17], the fact that the ligand
with the amino-carboxylic side chain (L4) presented a good in vivo
overload [5]. It is a microbial trishydroxamic acid [6] able to form with
3
+
Fe , as hexadentate ligand, a metal–ligand complex, having high
stability from the thermodynamic point of view (logK[FeL]2+ = 30.6 at
I = 0.10 M in KCl(aq) and T = 298.15 K) [7,8]. On the other hand, it is
featured by various side effects, such as hydrophilic character, oral
inactivity, toxicity, being also very expensive [9]. To overcome these
drawbacks, the employment of a new class of chelating agents has been
⁎
Corresponding authors.
E-mail addresses: ssammartano@unime.it (S. Sammartano), masantos@ist.utl.pt (M.A. Santos).
https://doi.org/10.1016/j.jinorgbio.2018.05.017
Received 14 February 2018; Received in revised form 25 May 2018; Accepted 26 May 2018
Available online 30 May 2018
0162-0134/ © 2018 Elsevier Inc. All rights reserved.

A. Irto et al.
Journal of Inorganic Biochemistry 186 (2018) 116–129
Abbreviations table
UV–Vis spectrophotometry Ultraviolet-Visible spectrophotometry
NMR
Nuclear Magnetic Resonance
DFB
DFP
Deferoxamine or Desferal®
1,2-dimethyl-3-hydroxy-4-pyridinone, deferiprone or Fer-
ESI-MS Electrospray Ionization Mass Spectroscopy
MS-LC Ion Trap Mass-Liquid Chromatography Ion Trap
riprox®
BnCl
MeOH
EtOH
benzyl chloride
methanol
ethanol
L1
L2
L3
L4
4-(3-hydroxy-2-methyl-4-oxopyridin-1(4H)-yl)butanoic
acid
(S)-2-amino-4-((2-(3-hydroxy-2-methyl-4-oxopyridin-
10% Pd/C palladium on activated carbon, 10% (w/w)
1
(4H)-yl)ethyl)amino)-4-oxobutanoic acid
(S)-2-amino-4-((3-(3-hydroxy-2-methyl-4-oxopyridin-
(4H)-yl)propyl)amino)-4-oxobutanoic acid;
TLC
Tin Layer Chromatography
dichloromethane
N-dimethylformamide
dimethyl sulfoxide
acetonitrile
DCM
DMF
DMSO
ACN
1
(S)-2-amino-5-(3-hydroxy-2-methyl-4-oxopyridin-1(4H)-
yl)pentanoic acid
L5
EthylL
1-(3-aminopropyl)-3-hydroxy-2-methylpyridin-4(1H)-one
2-(3-hydroxy-2-ethyl-4-oxopyridin-1(4H)-yl)acetic acid
Et
TMS
2
O
diethyl ether
tetramethylsilane
performance in terms of metal clearance, led us to develop other two
new analogues (L2: (S)-2-amino-4-((2-(3-hydroxy-2-methyl-4-oxopyr-
idin-1(4H)-yl)ethyl)amino)-4-oxobutanoic acid; L3: (S)-2-amino-4-((3-
in NaCl, the main inorganic component of most of natural [18,19] and
biological fluids [20]. The investigations on the acid-base properties of
these 3-hydroxy-4-pyridinones (3,4-HPs) were performed at
T = 298.15 and 310.15 K, with the aim of assessing their behavior also
at physiological conditions. To complete this study, the sequestering
ability of the ligands towards the Al3+ was investigated, by means of a
sigmoidal equation and of pL0.5 parameter, previously proposed by the
research group [21]. In vivo assays on metal-sequestration of a Ga
treated mice model were also carried out for the 3,4-HP-amino-acid
derivatives (L2, L3, L4) in comparison with the market drug (DFP).
(
3-hydroxy-2-methyl-4-oxopyridin-1(4H)-yl)propyl)amino)-4-ox-
obutanoic acid) containing the same terminal group but with an extra
amide bond. So the set of five bifunctional ligands containing as side
groups amino-carboxylic, as well as amino, carboxylic groups (for
comparison purposes) were prepared and evaluated for their Al3+-se-
questering capacity. Their preparation involved standard reactions of
maltol with bifunctional amines, while in some cases further coupling
reactions with a cyclic amino-acid anhydride with the formation of new
amidic bonds [5,11–13]. The synthetic procedures used in the pre-
paration of a set of the bifunctional 3-hydroxy-4-pyridinones are
schematically described in Scheme 1. The acid-base properties were
investigated by Ultraviolet-Visible (UV–Vis) spectrophotometry, spec-
67
2. Materials and methods
2.1. Chemicals for solution studies
3
+
trofluorimetry, while the binding ability towards Al
was studied by
All the reagents were of the highest available purity and the solu-
potentiometry and UV–Vis spectrophotometry. With the purpose of
confirming the speciation models proposed on the basis of the cited
investigations, the protonation behavior of L2 as well as its interaction
−1
tions were prepared with analytical grade water (R = 18 MΩ cm
)
using grade A glassware. Hydrochloric acid and sodium hydroxide so-
lutions were prepared by diluting concentrated ampoules (Riedel-
deHäen) and were standardized against sodium carbonate and po-
tassium hydrogen phthalate, respectively. NaOH solutions were pre-
3
+
1
with Al
was further studied by H Nuclear Magnetic Resonance
(
NMR) spectrometry. The measurements were carried out at I = 0.15 M
served from atmospheric CO
solutions were prepared by weighing AlCl
2
using soda lime traps. The aluminium
hexahydrate, without fur-
3
O
ther purification, and standardized with EDTA standard solutions; their
purity resulted always ≥98% [22]. NaCl aqueous solutions were pre-
pared by weighing the pure salt (Fluka), previously dried in an oven at
T = 383.15 K for at least 2 h.
O
OH
O
a
OH
OH
b
N
c
N
d
N
e
O
2.2. General synthetic information
f
HN
g
OH
OH
DFP
O
NH
2
Analytical grade reagents were purchased from Sigma-Aldrich,
O
Fluka and Acros and were used as supplied. The solvents, if necessary,
were dried according to standard methods [23]. All the reactions were
TLC (Tin Layer Chromatography) controlled and the most common
used mobile phases were dichloromethane (DCM): methanol (MeOH):
L2
L1
O
O
N
O
N
ammonium hydroxide (NH
4
OH) solvents mixtures: S1 (DCM-MeOH
.5:0.5), S2 (DCM-MeOH-NH OH 8.5:1:0.5), S3 (DCM-MeOH-NH OH
:2:0.5) (for more details see Abbreviations table). Moreover, ferric
OH
OH
OH
9
8
4
4
N
O
chloride (to check the presence of phenol groups), ninhydrine (for
amino groups) and Dragendorff's reagent (if quaternary nitrogen groups
were present) assays were carried out. The purity of the synthesized
compounds was checked by 1H and C NMR (proton and 13‑carbon
Nuclear Magnetic Resonance) experiments, performed with Bruker
AVANCE III 300 MHz and 400 MHz spectrometers, in deuterated sol-
O
NH2
OH
OH
NH2
NH2
N
H
13
O
L4
L5
L3
2 4 6 6
vents (D O, Methanol‑d , Dimethyl sulfoxide-d (DMSO‑d )). Melting
Fig. 1. Structures of deferiprone (DFP) and bifunctional 3-hydroxy-4-pyr-
idinones under study. For L2, the letters stand for the 1H NMR titrations peaks
assignment. For each ligand the protonable groups are highlighted.
point of each 3-hydroxy-4-pyridinone was measured using a Leica
Galen III hot stage apparatus. The following abbreviations are used:
s = singlet, d = doublet, t = triplet, m = multiplet. Electrospray
117

A. Irto et al.
Journal of Inorganic Biochemistry 186 (2018) 116–129
a
b
c
Scheme 1. General scheme for the synthesis of the ligands. Reagents and conditions: a) BnCl, NaOH, MeOH/H
T = 348 K, 20 h; c) H , 10% Pd/C, MeOH, p = 4.5 atm, 3–4 h; d) dry DMF, T = 333 K, 3 h.
2 2
O, reflux T = 348 K, 20 h; b) NaOH, EtOH/H O, reflux
2
ionization mass (ESI-MS) spectra were carried out on a 500 MS-LC
Mass-Liquid Chromatography) Ion Trap (Varian Inc., Palo Alto, CA,
USA) mass spectrometer equipped with an ESI ion source, operated in
the positive or negative ion mode. For the target final compounds, the
elemental analyses were performed on a Fisons EA1108 CHNS/O in-
strument at LAIST and were within the limit of ± 0.4%.
(cNaOH = 2 M). This reaction mixture was left under reflux at T = 348 K
for 20 h. After cooling, HCl aqueous solution (cHCl = 2 M) was added
until pH ~1 and ethanol was evaporated. To the remaining oily
solution, water was added (50 mL), followed by extraction with
diethyl ether (4 × 50 mL). The aqueous phase was alkalinized with
NaOH aqueous solution (cNaOH = 10 M) until about pH ~ 12 and
extracted with dichloromethane (3 × 50 mL). The organic phase was
dried over anhydrous sodium sulfate and filtered; then the solvent was
evaporated to dryness. To this dried residue, 3 mL of EtOH were added;
this solution was acidified until pH ~2 with HCl-saturated EtOH to give
a white precipitate that was recrystallized from EtOH-ACN (ACN:
(
2
2
.3. Synthetic procedures
.3.1. Synthesis of 4-(3-hydroxy-2-methyl-4-oxopyridin-1(4H)-yl)butanoic
acid (L1)
This compound was prepared following a synthetic procedure pre-
viously described [24].
acetonitrile), providing
hydrochloride salt (8.80 g, η = 32%). TLCs in S2 mixture, R
m.p. 460–463 K. 1H NMR (400 MHz, D
O), δ (ppm): 8.12 (1H, d, 6-
HPy), 7.39 (5H, s, Ph), 7.10 (1H, d, 5-HPy), 5.07 (2H, s, CH Ph), 4.51
a
pure product as the corresponding
f
= 0.74.
2
2
13
2
1
2
.3.2. Synthesis of (S)-2-amino-4-((2-(3-hydroxy-2-methyl-4-oxopyridin-
(4H)-yl)ethyl)amino)-4-oxobutanoic acid (L2)
.3.2.1. 3-(Benzyloxy)-2-methyl-4-pyrone. To a solution of 3-hydroxy-2-
(
(
1
2H, t, CH
100 MHz, D
29.62, 129.26, 128.84, 114.09, 75.39, 52.37, 38.00, 12.91; m/z (ESI-
MS) = 259 (M + 1).
2
NPy), 3.34 (2H, t, CH
2 2 3
NH ), 2.36 (3H, s, CH ). C NMR
2
O), δ (ppm): 166.39, 150.03, 143.49, 142.28, 135.21,
methyl-4H-pyran-4-one (20.07 g, 159.15 mmol) in 200 mL of MeOH, a
solution of sodium hydroxide (6.98 g, 174.50 mmol) in 22 mL of water
was added to this mixture, followed by dropwise addition of benzyl
chloride (22.0 mL, 191.00 mmol). The reaction mixture was heated to
reflux at T = 348 K for 12 h. The solvent was evaporated under reduced
pressure and the remaining orange oil was dissolved in DCM (80 mL)
and washed with 5% (w/v) of NaOH aqueous solution (3 × 30 mL) and
with water (2 × 30 mL). The organic phase was dried over anhydrous
sodium sulfate and filtered. The solvent was roto-evaporated and dried
2.3.2.3. (S)-4-((2-(3-(Benzyloxy)-2-methyl-4-oxopyridin-1(4H)-yl)ethyl)
amino)-2-(((benzyloxy) carbonyl)amino)-4-oxobutanoic acid (L2a). 1-(2-
aminoethyl)-3-(benzyloxy)-2-methylpyridin-4(1H)-one (0.76 g, 2.59 mmol),
in the form of hydrochloride salt, was neutralized with potassium hydroxide
(0.22 g, 3.92 mmol) in 10 mL of dry MeOH and left stirring under nitrogen
for 1 h. After filtering this solution, to remove the white precipitate of KCl
and evaporating the solvent under vacuum, 10 mL of DMF (N-
dimethylformamide), dried with molecular sieves and distilled, were
added to the flask. Meanwhile, N-Z-L-aspartic anhydride (0.78 g,
in vacuum to give a pale yellow oil product (32.93 g, η = 95%). TLCs
1
were performed in S1 mixture. H NMR (400 MHz, Methanol‑d
4
), δ
(
ppm): 7.82 (1H, d, 6-HPy), 7.33 (5H, s, Ph), 6.35 (1H, d, 5-HPy), 5.03
2H, s, CH Ph), 2.05 (3H, s, CH ). C NMR (100 MHz, Methanol‑d ),
2 3 4
1
3
(
3.13 mmol, Z = benzyloxycarbonyl protecting group) was dissolved in
δ(ppm): 175.81, 160.98, 155.05, 143.49, 136.81, 128.76, 128.19,
5
mL of dried and distilled DMF and added dropwise to the neutralized
128.18, 116.22, 73.37, 13.67; m/z (ESI-MS) = 217 (M + 1).
protected 3-hydroxy-4-pyridinone. The mixture was left on stirring at
T = 333 K under nitrogen. After 3 h, the starting materials were
completely consumed and only one major product was formed. The
solvent was evaporated under vacuum at ca. T = 363 K and the solid
product (0.57 g, η = 44%) was obtained after recrystallization with MeOH-
2.3.2.2. 1-(2-Aminoethyl)-3-(benzyloxy)-2-methylpyridin-4(1H)-one. 3-
Benzyloxy-2-methyl-4-pyrone (20.18 g, 93.35 mmol) was dissolved in a
EtOH (ethanol)/H O mixture (6/8 mL) and added dropwise to a
2
mixture of ethylenediamine (6.90 mL, 102.66 mmol) in 37 mL of
ethanol and 24 mL of water, and 8.5 mL of a NaOH aqueous solution
Et
2 2
O (Et O: diethyl ether). S3 mixture was used as eluent for TLC control of
1
the reaction, R
f
4
= 0.30; m.p. 370–372 K. H NMR (400 MHz, Methanol‑d ),
118

A. Irto et al.
Journal of Inorganic Biochemistry 186 (2018) 116–129
δ (ppm): 7.94 (1H, d, 6-HPy), 7.46 (5H, s, Ph(Z)), 7.37 (5H, s, Ph), 6.91 (1H,
5.12 (2H, s, CH
3.21 (2H, d, CH
1.97 (2H, m, CH
2
Ph), 4.63 (1H, t, CH
2
CHNH), 4.12 (2H, t, CH
NHCO), 2,34 (3H, s, CH
). C NMR (100 MHz, Methanol‑d ), δ (ppm):
2
NPy),
d, 5-HPy), 5.13 (2H, s, CH
2
Ph(Z)), 5.11 (2H, s, CH
2
Ph), 4.56 (1H, t,
NHCO), 2.77 (2H, m,
). C NMR (100 MHz, Methanol‑d ), δ (ppm):
72.47, 172.24, 171.26, 168.48, 156.93, 147.97, 147.31, 144.91, 136.45,
28.71, 128.66, 128.23, 128.10, 127.65, 127.49, 113.88, 74.13, 65.85,
4.09, 51.55, 38.36, 35.96, 12.01; m/z (ESI-MS) = 508 (M + 1).
2
CHNH), 2.81 (2H, t, CH
2
3
),
13
CH
CH
2
CHNH), 4.31 (2H, t, CH
CHNH), 2.42 (3H, s, CH
2
NPy), 3.52 (2H, t, CH
2
2
CH
2
NH
2
4
13
2
3
4
172.97, 172.27, 167.93, 157.02, 147.46, 144.37, 140.90, 136.72,
136.19, 128.81, 128.35, 128.21, 128.07, 127.64, 127.38, 113.93,
74.18, 66.6, 53.20, 51.92, 37.60, 35.38, 29.80, 11.85; m/z (ESI-
MS) = 522 (M + 1).
1
1
5
2
.3.2.4. (S)-2-Amino-4-((2-(3-hydroxy-2-methyl-4-oxopyridin-1(4H)-yl)
2.3.3.3. (S)-2-Amino-4-((3-(3-hydroxy-2-methyl-4-oxopyridin-1(4H)-yl)
propyl)amino)-4-oxobutanoic acid (L3). To a solution of L3a (0.33 g,
0.63 mmol) in dry methanol (50 mL), 10% Pd/C (0.09 g, 0.79 mmol)
ethyl)amino)-4-oxobutanoic acid (L2). To a solution of (S)-4-((2-(3-
(
(
benzyloxy)-2-methyl-4-oxopyridin-1(4H)-yl)ethyl)amino)-2-
((benzyloxy)carbonyl)amino)-4-oxobutanoic acid (L2a) (0.60 g,
2
was added and the mixture was shaked under H (4.5 atm) for 4 h at
1.18 mmol) in of dry methanol (50 mL), 10% Pd/C (palladium on
room temperature. After filtration of the reaction mixture and
activated carbon, 10% (w/w), 0.13 g, 1.20 mmol) was added and the
mixture was shaked under hydrogen (p = 4.5 atm) for 3 h at room
temperature. After filtration of the reaction mixture and evaporation of
the solvent under reduced pressure, the solid product was recrystallized
evaporation of the methanol under reduced pressure, the solid
product was recrystallization from MeOH-Et
2
O, affording pure L3
compound (0.15 g, η = 70%). TLCs in mobile phase S3 mixture,
1
R
f
= 0.35; m.p. 400–403 K. H NMR (400 MHz, methanol‑d
4
),
O), δ(ppm): 7.78 (1H, d, 6-HPy), 6.72
(1H, d, 5-HPy), 4.15 (2H, t, CH NPy), 3.95 (1H, t, CH CHNH ), 3.23
(2H, t, CH NHCO), 2.78 (2H, m, CH CHNH ), 2,45 (3H, s, CH ), 1.97
(2H, m, CH CH
NH). 13C NMR (100 MHz, D
δ
1
with MeOH-Et
2
O, affording pure L2 product (0.20 g, η = 52%). TLCs
= 0.38. m.p.
O), δ (ppm): 7.66 (1H, d, 6-HPy), 6.69
NPy), 3.86 (1H, t, CH CHNH ), 3.55
NHCO), 2.68 (2H, m, CH CHNH ), 2.47 (3H, s, CH
O), δ (ppm): 174.56, 172.65, 171.47, 169.59,
64.25, 143.25, 139.26, 111.54, 54.36, 51.32, 38.15, 35.05, 11.08;
m/z (ESI-MS) = 284 (M + 1). Elemental analysis calcd. for
·0.18 H O: C 50.30, H 6.11, N 14.67%; found: C 50.19, H
.15, N 14.52%.
(ppm): H NMR (400 MHz, D
2
were performed using S3 mixture as mobile phase, R
f
2
2
2
1
3
(
(
90–393 K. H NMR (400 Hz, D
1H, d, 5-HPy), 4.30 (2H, t, CH
2H, t, CH
2
2
2
2
3
2
2
2
2
2
2
O), δ (ppm): 175.02,
13
2
2
2
3
).
C
172.92, 169.38, 165.15, 143.76, 139.79, 138.59, 137.80, 53.00, 48.90,
NMR (100 MHz, D
2
36.73, 35.09, 29.13, 11.78; m/z (ESI-MS) = 298 (M + 1). Elemental
analysis calcd. for C13H N O ·0.1EtOH: C 52.12, H 6.48, N 13.92%;
19 3 5
found: C 52.23, H 6.51, N 13.86%.
1
C
12
H
17
N
3
O
5
2
6
2.3.4. Synthesis of (S)-2-amino-5-(3-hydroxy-2-methyl-4-oxopyridin-
1(4H)-yl)pentanoic acid (L4)
2
1
2
.3.3. Synthesis of (S)-2-amino-4-((3-(3-hydroxy-2-methyl-4-oxopyridin-
(4H)-yl)propyl)amino)-4-oxobutanoic acid (L3)
.3.3.1. 1-(3-Aminopropyl)-3-(benzyloxy)-2-methylpyridin-4(1H)-one. A
Compound L4 was synthesized using a protocol already reported in
the literature [15].
solution of 3-benzyloxy-2-methyl-4-pyrone (10.41 g, 48.16 mol) in a
ethanol/water mixture (6/4 mL) was added dropwise to a solution of
2.3.5. Synthesis of 1-(3-aminopropyl)-3-hydroxy-2-methylpyridin-4(1H)-
one (L5)
1
,3-diaminopropane (4.50 mL, 52.97 mmol) in a 17 mL of EtOH, 13 mL
of H O and NaOH aqueous solution (cNaOH = 2 M, 4.50 mL). The
reaction mixture was left on stirring under reflux at T = 348 K for
0 h. The experimental procedure followed for the workup was the
same reported for the 1-(2-aminoethyl)-3-(benzyloxy)-2-methylpyridin-
(1H)-one. Recrystallization of the white precipitate from EtOH-ACN
gave the product in the form of hydrochloride salt (5.32 g, η = 36%).
TLC control of the reaction was carried out using S2 mixture, R = 0.45.
O), δ (ppm): 8.13 (1H, d, 6-
HPy), 7.38 (5H, s, Ph), 7.11 (1H, d, 5-HPy), 5.08 (2H, s, CH Ph), 4.28
2H, t, CH NPy), 2.99 (2H, t, CH NH ), 2.34 (3H, s, CH ), 2.09 (2H, m,
CH CH NH ). C NMR (100 MHz, D
This product was prepared following literature procedures de-
scribed previously [17].
2
2
2.4. Experimental equipments and procedures
4
2.4.1. UV–Vis spectrophotometric apparatus and procedure
A Varian Cary 50 UV–Vis spectrophotometer equipped with an optic
fiber probe with a fixed 1 cm path length was used for performing the
spectrophotometric measurements. The instrument was connected to a
PC and the acquisition of the signal, the absorbance (A) vs. wavelength
(λ/nm) was carried out by the Varian Cary WinUV software. At the
same time, potentiometric data were collected using a combined glass
electrode (Ross type 8102, from Thermo-Orion), connected to a po-
tentiometric apparatus. The titrant was delivered in the measurement
cell by means of a Metrohm 665 automatic burette; a stirring bar en-
sured the homogeneity of the solutions during the experiment. Before
each measurements, N2(g) was bubbled in the solutions for at least
f
1
m.p. 463–466 K. H NMR (400 MHz, D
2
2
(
2
2
2
3
1
3
2
2
2
2
O), δ (ppm): 164.96, 150.10,
1
3
42.97, 141.92, 135.19, 129.69, 129.27, 128.82, 113.56, 75.43, 53.21,
6.21, 27.28, 12.90; m/z (ESI-MS) = 273 (M + 1).
2.3.3.2. (S)-4-((3-(3-(Benzyloxy)-2-methyl-4-oxopyridin-1(4H)-yl)
propyl)amino)-2-(((benzyloxy)carbonyl)amino)-4-oxobutanoic
L3a). 1-(3-Aminopropyl)-3-(benzyloxy)-2-methylpyridin-4(1H)-one
hydrochloride salt (1.00 g, 3.24 mmol) was firstly neutralized with KOH
0.25 g, 4.45 mmol) in of dry MeOH (20 mL) under N and left stirring
acid
(
5 min, with the aim of excluding the presence of CO2(g) and O2(g)
.
For the investigation of the 3-hydroxy-4-pyridinones protonation
−5
(
2
L
constants, 25 mL of a solution containing the ligands (1.2·10 ≤ c /
−5
for 1 h. The solid precipitate was filtered off, the solvent was
rotoevaporated and the residue was dry under vacuum. N-Z-L-aspartic
anhydride (0.98 g, 3.93 mmol) dissolved in 5 mL of dry and freshly
distilled DMF was added dropwise to the neutralized 1-(3-
aminopropyl)-3-(benzyloxy)-2-methylpyridin-4(1H)-one dissolved in
the same solvent (15 mL). The reaction mixture was left on stirring at
T = 333 K under nitrogen for 3 h. Then, DMF was evaporated under
vacuum at about T = 366 K and the solid residue was dried under
vacuum. Recrystallization from methanol-diethyl ether afforded the
desired product (0.37 g, η = 22%); m.p. 378–381 K. The eluent used for
M ≤ 5.5·10 ) was titrated in the pH range 2.0–10.7. The measure-
ments were performed in the range 200 ≤ λ/nm ≤ 400 at I = 0.15 M in
NaCl(aq) and T = 298.15 K and 310.15 K (physiological conditions). For
the study of the aluminium-ligand interactions, the titrations were
carried out in the wavelengths range as for the study of the acid-base
properties. The following metal and ligand concentrations were used:
3+
−6
−5
−5
Al
L
(5·10 ≤ cAl3+ / M ≤ 1.5·10 ), ligands (1.2·10 ≤ c /
−5
M ≤ 5.5·10 ), at I = 0.15 M in NaCl(aq) and T = 298.15 K, in the pH
range 2.0–11.0.
the TLC control of the reaction was S3 mixture, R
400 MHz, Methanol‑d ), δ (ppm): 8.05 (1H, d, 8 Hz, 6-HPy), 7.37 (5H,
s, Ph(Z)), 7.33 (5H, s, Ph), 7.27 (1H, d, 5-HPy), 5.16 (2H, s, CH Ph(Z)),
f
= 0.45. 1H NMR
2.4.2. Spectrofluorimetric apparatus and procedure
A FluoroMax-4 spectrofluorometer by Horiba Jobin-Yvon, equipped
with F-3006 Autotitration Injector with two Hamilton Syringes (mods.
(
4
2
119

A. Irto et al.
Journal of Inorganic Biochemistry 186 (2018) 116–129
metal ion (3.3·10 3–4·10−3 M), at metal/ligand ratios between 2 and 3,
at pH between approx. 2.0 and 7.0, and I = 0.15 M in NaCl(aq). Prior to
the study of metal containing systems, the protonation behavior of L2
−
Gastight 1725 and 1001 TLLX, 250 μL and 1 mL of capacity, respec-
tively), was used for performing the experiments; the resolutions of
wavelength selectors and titrant additions were 0.3 nm and 0.25 μL,
respectively. The spectrofluorometer was equipped with a Peltier
Sample Cooler (mod. F-3004) controlled by a Peltier Thermoelectric
Temperature Controller model LFI-3751 (5 A - 40 W). The system was
controlled by the FluorEssence 2.1 software by Horiba Jobin-Yvon. The
measurements were performed using a 1 cm light path Hellma type
−2
was also investigated for c
L
= 1·10 M solutions at I = 0.15 M in Na-
Cl(aq), in the pH range 2.0–10.7.
2.5. Calculations
1
01-OS precision cell. In this cell a magnetic stirrer to homogenize the
solution and combined glass microelectrode (model biotrode
.0224.100 purchased from Metrohm) to measure the e.m.f. or the pH
The refinement of all the parameters of the acid–base titration such
0
a
w
as E , pK , liquid junction potential coefficient j
a
, analytical con-
6
centration of reagents was carried out by means of the non-linear least
squares computer program ESAB2M [26]. The determination of the
protonation constants from the UV–Vis spectrophotometric and spec-
trofluorimetric data involved the use of HYPERQUAD 2008 computer
program [27]. For L2 1H NMR solution spectra, since all proton ex-
change reactions were found to be fast on the NMR time-scale, the in-
dividual chemical shifts belonging to each species and the relative
protonation constants were calculated with the HypNMR computer
program [28]. Protonic NMR spectra of Al3+/L2 solutions showed two
sets of resonances, accounting for free and bound ligand, respectively;
the bound average chemical shifts were employed in the HypNMR
calculations (see Results and discussions) thus allowing the refinement
of the stability constants as well as the individual NMR parameters of
each complex species, keeping constant the “free” L2 individual che-
mical shifts previously calculated. The calculation of the formation
constants obtained by potentiometric titrations at different ionic
strengths and temperatures was carried out by means of BSTAC and
HYPERQUAD 2008 computer programs. More details on these latter
computer programs used for the elaboration of the experimental data
are reported [27,29]. HySS program [27] was used to draw the spe-
ciation diagrams and to calculate the species formation percentages.
were inserted. The burette tip and the electrode were both located to
avoid interference with the light beam. The automatic acquisition of
data, in term of emission intensity vs. λ/nm for each volume of titrant
added, was performed using the same FluorEssence 2.1 software. The
best experimental conditions were determined through preliminary
evaluations in which parameters such as equilibration time, scan rate,
scan range and integration time, excitation and emission wavelengths,
were systematically changed to select the values providing the best
signal/noise ratio.
Spectrofluorimetric measurements were performed at I = 0.15 M in
NaCl(aq) and T = 298.15 K and 310.15 K. The experiments were carried
out by titrating, with NaOH standard solutions, 2 mL of a solution
−4
L
containing the ligands in the concentration range 5·10 ≤ c /
−3
M ≤ 1·10
and NaCl to obtain the desired ionic strength. After each
addition of base, the intensity of signal (counts per second, CPS) was
recorded in the wavelengths range 300 ≤ λ/nm ≤ 520 (excitation
−
1
wavelength: λ/nm = 278), with a scan rate of 2 nm s
and an in-
tegration time of 0.5 s, together with the corresponding e.m.f. value.
2.4.3. Potentiometric apparatus and procedure
z−
Potentiometric measurements were carried out at I = 0.15 M in
The protonation constants of all the ligands (L ) are given ac-
H
H
NaCl(aq) and T = 298.15 K by using a Metrohm model 809 Titrando
connected to an automatic burette equipped with a combined glass
electrode (Ross type 8102, from Thermo-Orion). The apparatus was
connected to a PC and automatic titrations were performed by
MetrohmTiAMO 1.2 software with the aim of controlling titrant de-
livery, data acquisition and to check for e.m.f. stability. Estimated ac-
curacy was ± 0.15 mV and ± 0.003 mL for e.m.f. and titrant volume
readings, respectively. The titrations were carried out in thermostatted
cells under magnetic stirring, bubbling through the solution purified
presaturated N2(g) to exclude O2(g) and CO2(g) inside. The measurement
cording to the stepwise (logK
equilibria, respectively:
r
; Eq. (1) and overall (logβ
r
; Eq. (2))
+
−(z−(r−1))
−(z−r)
H + H(r−1)
L
= H
r
L
(1)
and
+
z−
−(z−r)
rH + L = H
r
L
(2)
where r represents the r-th protonation step and z is the charge of the
fully deprotonated ligands.
OH) are given as fol-
r
The overall metal hydrolytic constants (logβ
3
+
−4
solutions were prepared at different Al
M ≤ 7·10 ) and ligand (5·10 ≤ c
(5·10 ≤ cAl3+ /
lows:
−4
−4
−3
L
/ M ≤ 2·10 ) concentration to
pAl3 + rH
+
O = Al
(OH)
(3p−r)
+ rH
+
obtain wide metal/ligand molar ratios.
2
p
r
(3)
For each experiment, independent titrations of strong acid with
standard sodium hydroxide solutions were performed, at the same ex-
perimental conditions of ionic medium, ionic strength (I) and tem-
perature (T) of the systems under study, to obtain the value of electrode
The overall formation constants (logβpqr) for the metal–ligand
complexes investigated refer to the following equilibrium:
3+
z−
+
(3p+r−qz)
pAl + qL + rH = Al
p
L
q
H
r
(4)
0
+
potential (E ), the acidic junction potential (E
j a
= j [H ]), and the ionic
All the protonation and stability constants, concentrations and ionic
product of water (K ). The pH scale used was the free scale, not ac-
w
+
+
strengths are expressed in the molar (c, M) concentration scale.
tivity, so pH = −log [H ], where [H ] is the free proton concentra-
tion. For each titration, 80–100 data points were collected, and the
equilibrium state during titrations was checked, in particular, verifying
the necessary time to reach equilibrium and performing back titrations
2.6. In vivo metal sequestration studies
[
25].
67Ga-citrate injection solution was prepared by dilution of 67Ga ci-
trate from Mallinckrodt Medical B.V. with saline to obtain a final
radioactive concentration of approximately 7.0–8.0 MBq/100 μL.
Biodistribution studies were carried out in groups of 3 female Balb-C
mice (randomly bred, Charles River, from CRIFFA, France) weighing ca.
22–25 g. Mice were intravenously (i.v.) injected with 100 μL
(7.0–8.0 MBq) of Ga citrate via the tail vein immediately followed by
intraperitoneal (i.p.) injection of 0.5 μmol of each ligand in 100 μL
saline. Animals were maintained on normal diet ad libitum and were
sacrificed by cervical dislocation at 1 h and 24 h post-administration.
The administered radioactive dose and the radioactivity in sacrificed
2
.4.4. NMR apparatus and procedure
1
2 2
H NMR spectra at T = 298.15 K in 9:1 H O/D O mixture were
recorded on a Bruker AMX R-300 operating at 300 MHz by employing
presaturation pulse sequence experiments for water signal suppression.
The chemical shifts were measured with respect to tetramethylsilane
67
(
(
TMS) and 1,4-dioxane was used as internal reference
δ
1
3+
dioxane = 3.70 ppm). H NMR titrations on Al /L2 systems were
performed by adding sodium hydroxide solution to mixtures of the li-
−3
−1
−2
L
gand (8·10 ≤ c / mol L ≤ 1·10 ) and various concentrations
120

A. Irto et al.
Journal of Inorganic Biochemistry 186 (2018) 116–129
animals were measured in a dose calibrator (Capintec CRC25R). The
difference between the radioactivity in the injected and sacrificed an-
imal was assumed to be due to whole body excretion. Tissue samples of
main organs were then removed for counting in a gamma counter
3.2. Acid-base properties of the ligands
All the 3-hydroxy-4-pyridinones synthetized were obtained as
0
H
r
(L) neutral species; in general, for this kind of compounds, different
(
Berthold LB2111, Berthold Technologies, Germany). Biodistribution
results were expressed as percent of injected activity per gram of organ
% I.A./g) and presented as mean values ± Std. Deviation.
protonable groups are detected: the eOH on the heterocyclic ring, the
amino and carboxylic groups eventually present on the alkylic chain
and also the pyridinone nitrogen atom (proton supplied by excess of
mineral acid) [16]. The structures of all the investigated compounds are
reported in Fig. 1. Analysis of the UV–Vis spectrophotometric data of
the ligand featuring the simplest lateral chain, L1, allowed the de-
termination of three protonation constants related to the hydroxyl
group, the carboxylic group and the pyridinone nitrogen atom. The
experimental protonation constants at I = 0.15 M in NaCl(aq), at
T = 298.15 K are reported in Table 1. In Fig. S1, the distribution of the
L1 species is shown. The graph evidences that all the species reach high
formation percentages (≥60%), being [H(L1)] the main species
which is present in all the pH range investigated (2.5–10.7).
L2, L3 and L4 ligands, unlike L1, are featured by an amino group on
the alkylic chain, so that four functional groups were considered in the
protonation study. In particular, L2 and L3 are very similar with the
(
3
. Results and discussions
3
.1. Design and synthesis of the ligands
The 3-hydroxy-4-pyridinone derivatives herein reported (Fig. 1)
were designed to have low molecular weight while including extra-
functional groups aimed to facilitate the interaction with biological
sites and to enable the drug absorption from the gastrointestinal (GI)
tract. Thus, besides the hydroxypyridinone chelating cores, the ligands
are featured with different type/number of H-bond donor and acceptor
−
2 2
groups (e.g. eCOOH), (eNH2, eNH CH COOH, eNHCOR), which are
expected to increase their affinity towards biological sites. These
functional groups are usually attached to the pyridinone side of the
molecules by means of n-alkylic spacers, with the aim of improving
their lipophilicity, and of consequently enhancing their ability to cross
biological membranes. On the other hand, the synthesis of excessively
lipophilic ligands should be avoided [15], since they could display high
toxicity, due to a too much easy crossing through the membranes or a
rapid metabolism for gluconization of the eOH group [30]. In silico
assessment of the main pharmacokinetic properties, (see below, Section
difference of an extra –CH
in Fig. 1. The protonation constants of L2 and L3 determined by UV–Vis
spectrophotometric technique at I = 0.15 M in NaCl(aq) and
2
group on L3 alkylic chain, as already shown
,
T = 298.15 K are reported in Table 1. The analysis of the results shows
that the acid-base properties of the ligands are influenced by the lateral
chain length; in particular, the logKH values increase with the length of
the alkylic spacer. As a representative example, the profile of the ti-
tration curves of the UV–Vis spectrophotometric measurements on L2 at
different pH values is reported in Fig. S2. This figure shows that the
absorption spectrum (absorbance (A) vs. wavelengths (λ)) varies sig-
nificantly with pH; this behavior can be explained by considering that
the ligand protonated species absorb at different wavelengths. From the
analysis of the spectra, it is possible to evidence that, upon pH in-
creasing, the absorption band at 278 nm (λmax) is characterized by an
increasing of the intensity up to pH ~ 6.0, then it starts to decrease,
undergoing a bathochromic shift. Moreover, some isosbestic points
occur at different wavelengths.
The last two 3-hydroxy-4-pyridinones, L4 and L5, differ from each
other on a carboxylic group on the end of the lateral chain (L4). The
data obtained at I = 0.15 M in NaCl(aq) and T = 298.15 K, by UV–Vis
spectrophotometric measurements are reported in Table 1, together
with the ones found for the other compounds. The refined constant
values of the hydroxyl, amino and pyridinone nitrogen groups for L4
and L5 are in good agreement.
3.8), indicated adequate molecular descriptors for drug-likeness prop-
erties and bioavailability.
The route for the synthesis of the 3-hydroxy-4-pyridinone deriva-
tives started from the naturally occurring maltol (3-hydroxy-2-methyl-
4H-pyran-4-one), following the reaction general procedure shown in
Scheme 1. The first step of the synthetic route involves the –OH group
protection by a benzyl group, according to Williamson ethers synthesis
starting from alcohols [31,32]. The second step implies a double Mi-
chael-type addition of a primary amine (extrafunctionalized with a
carboxylic (L1), an amine (L5) or an amino-carboxylic group (L4)) with
opening and closure of the heterocyclic ring, according to the procedure
by Harris et al. [33,34]. For compounds L2 and L3 there was a further
extrafunctionalization by reaction of an intermediate containing a N-
alkylamine side chain with N-benzyloxycarbonyl protected aspartic
anhydride to provide an amino-carboxylic acid terminal groups. The
last step in the synthesis of each compound involved always the re-
moval of the protecting groups (O-benzyl for all the compounds and
also N-benzyloxycarbonyl for L2 and L3) via hydrogenolysis catalyzed
by 10% Pd/C.
To confirm the speciation models obtained by UV–Vis spectro-
photometric measurements, spectrofluorimetric titrations were per-
formed as well. Moreover, some checks were also carried out by using
Table 1
a
Protonation constants of the 3-hydroxy-4-pyridinones obtained by different analytical techniques at I = 0.15 M in NaCl(aq) and T = 298.15 K.
H
(logK H
)
r
Analytical technique
logβ
r
L1
L2
L3
L4
L5
H
9.947 ± 0.007b
14.36 ± 0.09 (4.41)
17.74 ± 0.03 (3.38)
UV–Vis spectrophotometry
logβ
logβ
logβ
1
2
3
10.73 ± 0.03
19.52 ± 0.08 (8.79)
24.17 ± 0.08 (4.65)
10.93 ± 0.02
20.70 ± 0.06 (9.77)
25.60 ± 0.01 (4.90)
11.10 ± 0.09
20.44 ± 0.07 (9.34)
24.60 ± 0.05 (4.16)
11.08 ± 0.02
20.468 ± 0.005 (9.388)
23.68 ± 0.03
(logK H
)
2
H
H
(logK
3
)
)
(3.21)
H
H
H
H
H
H
H
H
H
(logK H
logβ
logβ
logβ
logβ
logβ
logβ
logβ
logβ
logβ
4
1
2
3
4
1
2
3
4
4
–
27.43 ± 0.08 (3.26)
10.73 ± 0.03
19.50 ± 0.13 (8.77)
24.18 ± 0.03 (4.68)
27.62 ± 0.04 (3.44)
10.79 ± 0.08
19.37 ± 0.05 (8.58)
24.15 ± 0.07 (4.78)
27.41 ± 0.03 (3.26)
29.02 ± 0.01 (3.42)
11.03 ± 0.01
20.38 ± 0.02 (9.35)
25.99 ± 0.04 (5.61)
27.87 ± 0.04 (3.27)
11.00 ± 0.01
19.77 ± 0.03 (8.77)
24.38 ± 0.03 (4.61)
–
Spectrofluorimetry
9.90 ± 0.04
14.38 ± 0.04 (4.48)
17.73 ± 0.12 (3.35)
–
–
–
–
–
11.20 ± 0.03
20.51 ± 0.03 (9.31)
23.70 ± 0.03 (3.17)
–
–
–
–
–
(logK H
)
)
)
2
H
(logK
(logK
3
H
4
29.69 ± 0.08 (3.70)
27.75 ± 0.03 (3.37)
1H NMR Spectroscopy
–
–
–
–
–
–
–
–
(logK H
)
)
)
2
H
(logK
(logK
3
H
4
a
b
logK H
±
H
r
and logβr refer to Eqs. (1)–(2), respectively.
Std. deviation.
121

A. Irto et al.
Journal of Inorganic Biochemistry 186 (2018) 116–129
the isosbestic points as wavelengths of excitation but no variations were
observed. At each point of the titration (mL of NaOH strong base), the
emission intensity variation vs. wavelength was recorded. All the li-
gands evidenced a lowering of the signal along the pH scale, increasing
the basicity of the solution. In Fig. S3 a tridimensional titration curve of
the ligand L2 is reported, as an example, at I = 0.15 M in NaCl(aq) and
T = 298.15 K.
100
3
3
4
4
1
2
80
60
40
20
0
2
The experimental spectrofluorimetric protonation constants of the
1
3-hydroxy-4-pyridinones studied at I = 0.15 M in NaCl(aq) and
5
T = 298.15 K are reported in Table 1.
The temperature effect on the protonation constants of the ligands is
dependent on the functional group, being different for the hydroxyl
with respect to the one displayed for the other groups (carboxylic,
5
amino and pyridinone nitrogen). Accordingly, analyzing the data re-
H
4
6
8
10
ported in Table 2, it appears that generally the logβ
1
values slightly
pH
increase with temperature whereas for the other protonation constants
this trend is not respected. To better highlight the temperature effect on
the speciation, in Fig. 2 the distribution diagram of L2 at I = 0.15 M in
NaCl(aq) at T = 298.15 K and T = 310.15 K is reported. The graph evi-
dences that the formation of the species at T = 310.15 K is shifted to
lower pH values.
= 3·10−5 M) at I = 0.15 M in NaCl(aq) and
Fig. 2. Distribution diagram of L2 (c
T = 298.15 K (solid line) and T = 310.15 K (dashed line). Species: 1.
L
2+
(L2)]+; 3. [H
0
−
2−
[
H
4
(L2)] ; 2. [H
3
2
(L2)] ; 4. [H(L2)] ; 5. [(L2)]
.
Al3 /3-hydroxy-4-pyridinones complexes obtained by potentiometry
+
The spectrofluorimetric data, reported in Table 2 at I = 0.15 M in
NaCl(aq) and T = 310.15 K, confirmed the trend of the UV–Vis values,
namely a decrease of the protonation constants for the carboxylic,
amino groups and pyridinone nitrogen atom upon temperature in-
creasing. Concerning the hydroxyl group, an opposite trend was ob-
served, except for L1.
st
column), UV–Vis spectrophotometry (2 column), 1H NMR spec-
nd
(
1
troscopy (3 column, see 1H NMR titrations section) and an average
rd
rd
(
4
column) of the cited values. The pH ranges of the experimental
measurements were 2.0–9.2 and 2.0–11.0 from potentiometric and
UV–Vis spectrophotometric data, respectively.
The investigation on Al3+/L1 interactions allowed to define a spe-
+
−
ciation model featured by five species, namely [Al(L1)] , [Al(L1)
2
] ,
.3. Hydrolytic constants of Al3+
0
+
3−
3
[Al(L1) H] , [Al(L1) H ] and [Al(L1)3] , without the formation of
2
(aq)
2 2
3+
possible hydrolytic or sparingly soluble species of Al
.
For the study of the Al3+/L2, Al /L3, Al /L4 and Al3+/L5 sys-
tems, where all the ligands contain an amino group, the same speciation
model was obtained; this model consists of two species, namely
3+
3+
The acid-base properties of the metal cation (Table S1) were already
studied from an accurate analysis of literature data [35–39], at different
conditions (i.e. I, T, etc.) and are reported in a previous paper [40].
[
AlLH](
4−z)
and [AlL]
(3−z)
; the UV–Vis spectrophotometric data con-
.4. Stability of Al3+/Lz− complexes
firmed the speciation schemes obtained by potentiometry, the stability
constant values being also reported in Table 3, found from the two
instrumental techniques in good agreement.
3
The study of the interactions of Al3+ with the ligands of interest
Concerning the [Al(L2)]+ and [Al(L3)]+ species, the stability con-
stant values decrease with alkyl chain length increasing. As an example,
Fig. S5 illustrates a comparison between the Al3+/L2 system titration
(
L1–L5), at I = 0.15 M in NaCl(aq) and T = 298.15 K, allowed to de-
(3p+r-qz)
termine Al
p
L
q
H
r
complexes of different stoichiometry. The most
reliable speciation models were selected taking into account different
variables, namely, (a) the simplicity of the model; (b) the significant
formation percentages of the species considered in the pH range in-
vestigated; (c) the statistical parameters, as standard deviation on sta-
bility constants and on the fit of the systems and in particular (d) the
values of corresponding ratios between single variances in comparison
with those from the accepted model. The high number of measurements
made (and of experimental points collected) allowed us to consider the
differences in variance between the accepted model and other models
to be significant.
curves from potentiometric data at different metal-to-ligand molar
0
ratio. During the experiments the formation of the [Al(OH)
3
]
(s) species
did not occur; however, introducing the solubility product of the
sparingly soluble species in the speciation model of the Al3 /L3 system,
by means of the HySS computer program, its formation can be expected
starting from pH ~6.5. This aspect can be explained in terms of slow
formation kinetic of this species [41].
+
The presence of the carboxylic group in L4 seems to increase the
stability of the species with the metal cation suppressing its hydrolytic
tendency and the possible formation of sparingly soluble species. In
Table 3 reports the experimental overall stability constants of all the
Table 2
a
Protonation constants of the 3-hydroxy-4-pyridinones obtained by UV–Vis spectrophotometry and spectrofluorimetry at I = 0.15 M in NaCl(aq) and T = 310.15 K.
H
(logK H
)
r
Analytical technique
logβ
r
L1
L2
L3
L4
L5
H
10.029 ± 0.007b
13.62 ± 0.01 (3.59)
16.54 ± 0.04 (2.92)
–
UV–Vis spectrophotometry
logβ
logβ
logβ
logβ
1
2
3
4
10.99 ± 0.02
10.93 ± 0.02
11.13 ± 0.03
10.57 ± 0.07
16.53 ± 0.08 (5.96)
19.53 ± 0.05 (3.00)
–
(logK H
)
17.05 ± 0.03 (6.06)
21.02 ± 0.04 (3.97)
24.08 ± 0.05
17.71 ± 0.07 (6.78)
22.50 ± 0.05 (4.79)
25.93 ± 0.05 (3.43)
17.93 ± 0.04 (6.80)
22.00 ± 0.03 (4.07)
25.30 ± 0.05 (3.30)
2
H
H
H
H
(logK
(logK
3
)
)
4
(3.06)
H
H
H
H
Spectrofluorimetry
logβ
logβ
logβ
logβ
1
2
3
4
10.05 ± 0.02
13.88 ± 0.01 (3.83)
16.44 ± 0.01 (2.56)
–
10.397 ± 0.002
10.47 ± 0.01
10.05 ± 0.02
10.62 ± 0.20
16.93 ± 0.04 (6.31)
19.74 ± 0.08 (2.81)
–
(logK H
)
)
)
17.34 ± 0.03 (6.94)
20.69 ± 0.04 (3.35)
22.94 ± 0.07 (2.25)
17.96 ± 0.06 (7.49)
22.00 ± 0.07 (4.04)
25.15 ± 0.09 (3.12)
17.02 ± 0.05 (7.00)
21.17 ± 0.04 (4.15)
24.27 ± 0.01 (3.10)
2
H
H
(logK
(logK
3
4
a
b
logK H
±
H
r
and logβr refer to Eqs. (1)–(2), respectively.
Std. deviation.
122

A. Irto et al.
Journal of Inorganic Biochemistry 186 (2018) 116–129
Table 3
Experimental overalla and averageb stability constants of Al3+/3-hydroxy-4-pyridinone complexes by potentiometric, UV–Vis spectrophotometric and 1H NMR
spectroscopic data at I = 0.15 M in NaCl(aq) and T = 298.15 K.
Species
Potentiometry (ISE – H+
)
UV–Vis Spectrophotometry
1H NMR Spectroscopy
Average stability constants
[
[
[
[
[
[
[
[
[
[
[
[
[
Al(L1)]+
12.30 ± 0.04c
23.44 ± 0.07
27.03 ± 0.12
30.92 ± 0.03
31.66 ± 0.18
23.28 ± 0.02
17.94 ± 0.10
24.27 ± 0.03
17.50 ± 0.10
23.48 ± 0.03
18.32 ± 0.03
20.98 ± 0.02
15.08 ± 0.09
12.85 ± 0.01c
22.90 ± 0.03
28.37 ± 0.07
30.20 ± 0.05
31.80 ± 0.07
23.66 ± 0.01
17.94 ± 0.04
24.50 ± 0.01
17.65 ± 0.10
23.47 ± 0.01
18.88 ± 0.02
20.05 ± 0.10
15.08 ± 0.09
–
–
–
–
12.57 ± 0.23d
23.17 ± 0.23
27.70 ± 0.36
30.56 ± 0.26
31.73 ± 0.11
23.59 ± 0.14
17.90 ± 0.10
24.38 ± 0.15
17.57 ± 0.12
23.47 ± 0.03
18.60 ± 0.23
20.51 ± 0.30
15.08 ± 0.36
−
Al(L1)
Al(L1)
Al(L1)
Al(L1)
Al(L2)H]
Al(L2)]+
Al(L3)H]
Al(L3)]+
2
2
2
3
]
0
H](aq)
H
]
]+
−
2
3
–
2
+
23.67 ± 0.15c
17.75 ± 0.16
2
+
+
+
–
–
–
–
–
–
2
Al(L4)H]
Al(L4)]+
Al(L5)H]
3
Al(L5)]2+
a
logβpqr refers to Eq. (4).
Values obtained by average of potentiometric, UV–Vis spectrophotometric and 1H NMR spectroscopic data.
Std. Deviation.
errors on weighed data.
b
c
±
d
Fig. 3a the distribution of the Al3+/L4 complexes obtained from po-
NaOH showed a single set of average resonances, although several
species involved in protonation equilibria are present in solution, fast
mutual exchange on the NMR time scale was assumed. In addition, the
average chemical shifts of L2 were pH-dependent, thus all the signals,
as a common feature, displayed a shielding effect upon pH increasing.
More in detail, both the two doublets detected in the aromatoid region,
namely a and b (Fig. 1) exhibited a significant decrease in their che-
mical shift in the acidic pH range, indicating that the first deprotona-
tion, as expected, involved the pyridinone nitrogen; then the cited
peaks did not change to a great extent up to pH ~ 9.0. From this pH
value on, the signals started to shift again, being this effect more pro-
nounced for the resonance indicated as “a”, suggesting a deprotonation
occurring at higher pH, involving a nearby functional group, i.e. the
hydroxyl group. A trend quite similar to the ones described was de-
tected at pH < 4.0 also for the methylene d protons, whereas the sig-
nals due to f protons were virtually unchanged up to pH ~7.0, starting
to shift to lower fields at higher pH. More significantly, the triplet due
2
+
tentiometric data is shown. The [Al(L4)H]
lution along the whole pH range of titration, while the [Al(L4)]
species is present in so+-
complex starts to form at pH ~2.7. Fig. 3b shows the distribution of the
3
+
Al /L5 system at the cited experimental conditions. In the case of L5,
the absence of the –COOH group, with respect to L4, should allow the
0
precipitation of [Al(OH)
3
]
(s), as shown in the “shaded zone” represented
3
+
3+
in the figure, where the formation percentages of Al , [Al(L5)H]
species are reported (95%, 96% and 8% at pH ~2.0,
4.1 and ~4.9, respectively). Once again, for the Al /L5 system the
2
+
and [Al(L5)]
3
+
~
0
3
formation of [Al(OH) ](s) was not observed experimentally, but it was
calculated by the HySS program [42], taking into account the pKs0
value of the sparingly soluble species in the speciation model.
A trend of stability of the species can be provided on the basis of the
3
+
investigation carried out on the Al /3-hydroxy-4-pyridinone systems,
i.e. by considering the stability constants of the common [AlL](
species:
3−z)
2
to g eCHe group, which is adjacent to both eCOOH and eNH groups,
L4 > L2 > L3 > L5 > L1
underwent significant shielding effect for pH < 4.5 and even stronger
at pH > 8.0, according to the deprotonation sequence proposed for
similar systems [16,43]. Moreover, since in the pH around neutrality,
1
3
.5. H NMR titrations
L2 ligand should have been twice deprotonated, the spectra recorded at
0
2
pH ~6.5 can be considered with good approximation as the [H (L2)]
With the aim to gain more information about acid-base behavior of
3
+
1
single species spectra (Fig. S4). As already stated, all the spectra from
L2 solutions showed a single set of resonances, arising from the rapidly
exchanging species on the NMR time scale. In this sense, each observed
chemical shift value corresponds to the mol-fraction weighted average
L2 as well as its coordination capability towards Al , several H NMR
spectra on a H O/D O mixture were collected, both on metal free and
2
2
metal containing systems, at different concentrations and metal/ligand
ratios. Since all the spectra coming from the NMR titrations of L2 with
Fig. 3. Distribution diagram of Al3+/L4 and L5
1
00
100
80
60
40
20
0
systems at I = 0.15 M in NaCl(aq) and T = 298.15 K,
2
3
1
2
4
−
4
−3
1
L
cAl3+ = 8·10 M; c = 2·10 M: a) L = L4, species:
8
6
4
2
0
0
0
0
0
2+
+
1.
free Al3+; 2. [Al(L4)H]
L = L5, species: 1. free Al ; 2. [Al(L5)H] ; 3. [Al
L5)] ; 4. [Al(OH)
;
3. [Al(L4)]
;
b)
3
+
3+
2+
0
(
3
]
(s). The pH range from 4.8 to
7
.0 (“shaded zone”) refers to the initial pH value of
0
formation of the sparingly soluble [Al(OH)
3
](s) spe-
cies up to the experimental pH value investigated.
3
2
3
4
5
6
7
2
3
4
5
6
7
pH
pH
3
+
3+
a) (Al /L4)
b) (Al /L5)
123

A. Irto et al.
Journal of Inorganic Biochemistry 186 (2018) 116–129
2
2
2
2
2
2
2
.60
.55
.50
.45
.40
.35
.30
observed
4.1
.0
3.9
observed
calculated
calculated
4
3
3
3
3
.8
.7
.6
.5
3.4
2
4
6
8
10
2
4
6
8
10
pH
pH
1
)
2)
−
2
L
Fig. 4. Observed (□) and calculated (○) values of chemical shifts of c (1) and g (2) nuclei of L2 vs. pH, at c = 1·10 M, I = 0.15 M in NaCl(aq) and T = 298.15 K.
of the values for all the species present at equilibrium, at each in-
vestigated pH. The protonation constants and the individual chemical
shift values, due to each single species, were obtained by employing
HypNMR software. Table 1 shows the calculated protonation constants
for L2, which are totally consistent with the ones obtained from UV–Vis
spectrophotometry and spectrofluorimetry. As can be inferred from
Table S2, besides all the peaks due to the differently protonated L2
species nuclei, the two different resonances due to f AB system were
refined as well. In addition, experimental and calculated chemical shifts
for selected nuclei are compared in Fig. 4, where their excellent
agreement is clearly evidenced.
as also indicated by the almost total overlapping found for the observed
and the calculated average chemical shifts (Fig. S6) found for selected
nuclei of the complexes. Furthermore, although from comparison of the
experimental spectra collected with and without aluminium, the com-
plexation of L2 towards Al3+ is expected to be accomplished on the
pyridinone side of the ligand (i.e. via hydroxo-oxo functionality), from
the calculated individual chemical shifts of the complexes with respect
to the ones refined for the protonated species of L2 (Table S2), an in-
2
teraction of metal with eNH and/or eCOOH could not be discarded.
3.6. Sequestering ability
With the purpose of confirming the speciation model proposed for
3
+ 1
the rationalization of the L2 complexing capability towards Al
,
H
The simple analysis of the stability constants of metal-ligand com-
plexes is not always sufficient to evaluate the sequestering ability of a
ligand towards a metal cation, owing to secondary interactions of the
metal or ligand with other components in the system; similar difficulties
are observed when the sequestering ability must be estimated at dif-
ferent pH, ionic strengths and temperatures. To solve this issue, it was
proposed the introduction of pL0.5, an empirical parameter which re-
presents the total concentration of ligand required to sequester the 50%
3
+
NMR titrations were also carried out for the Al /L2 system, in various
concentrations and metal/ligand ratios, as already pointed out in the
Materials and methods section. For the spectra collected in the pH
range 2.0–5.5, regardless of the metal/ligand ratio employed, two dif-
ferent sets of resonances were observed. More in detail, at lower pH,
2
+
from the speciation data, together with the [H
4
(L2)] , the formation
should be expected. As already observed for several
2
+
of [Al(L2)H]
−12
other aluminium-containing systems [44,45], the presence of the metal
exerted a general broadening of the spectra; nevertheless, by comparing
the spectra recorded at pH ~2.0 for metal free and metal containing
of a metal cation present in trace (10
M) in solution [21,46].
This parameter is described through a sigmoidal type Boltzmann
equation, reported in Eq. (5):
2
+
systems, the peaks due to [H
4
(L2)]
are clearly distinguishable from
1
the new shielded signals, indicated as “bound” along the discussion
xM =
(
pL−pL
)
0.5
1
+ 10
(5)
(
Fig. 5). These new peaks, detected only for a and b aromatic protons, c
2
+
methyl and d methylene, suggest that in the [Al(L2)H]
metal-ligand
where x
M
is the mole fraction of metal complexed by the ligand,
interactions, this part of L2 is involved. It is worth to mention that,
differently from the peaks due to the free ligand, the “bound” signals, as
long as they could be detectable from the spectra, did not shift to a great
extent upon pH increasing. Starting from pH ~5.5, the registered
spectra showed a single set of broad bands, although in these conditions
signals due to [Al(L2)]+ should be observed. In this pH range, the
signals named as a, b, c, d and e appeared downfield shifted with respect
to the metal-free system, whereas f and g peaks were found to be little
affected by the presence of aluminium. This experimental evidence
suggests that all the species forming from pH ~5.5 on are involved in a
fast mutual exchange. Accordingly, the rationalization of the whole
NMR data on metal containing systems has been carried out with
HypNMR software as well. In this sense, the calculation of the com-
plexes stability constants was carried out using, as known values, the
data gained from the protonation data refinement (i.e. protonation
constants and individual species chemical shifts) obtained previously.
The so calculated logβ values (Table 3) were in excellent agreement
with the values derived from potentiometric and spectrophotometric
investigations, thus confirming the reliability of the speciation model,
pL = −log c
L
and pL0.5 = −log c
L
, when x
M
= 0.5. The sequestering
ability can be graphically represented by a dose – response curve,
characterized by asymptotes equal to 1 for pL → −∞ and 0 for pL →
+∞, obtained by plotting the mole fraction of the metal complexed vs.
the pL values. High pL0.5 values account for a greater sequestering
ability of a ligand towards a metal cation. It is a quantitative and ob-
jective parameter which allows to determine the sequestering ability of
a ligand in different experimental conditions (pH, ionic strength, ionic
medium, temperature), and it is independent on the ion analytical
concentration when it is present in trace amount.
Moreover, various factors may influence the systems, such as the
protonation of the ligand, the hydrolysis of the metal cation and the
interactions with other components, which are taken into account in
the speciation model, but are excluded from the calculation of pL0.5
values. The estimation of the sequestering ability can be employed for
several applications such as remediation of polluted sites, detoxification
processes, in chelating therapy and water treatment. All these processes
involve the use of a chelating agent, and the assessment of this para-
meter can be useful for optimizing the working conditions.
124

A. Irto et al.
Journal of Inorganic Biochemistry 186 (2018) 116–129
Fig. 5. 1H NMR spectra of: 1) L2 (c
= 1·10−2 M) and 2) Al3+/L2 system (cAl3+ = 4·10−3 M, c = 8·10−3 M) at pH ~2.0, I = 0.15 M in NaCl(aq) and T = 298.15 K.
L
L
The study of the sequestering ability of the 3,4-HPs towards Al3+
the possible precipitation of [Al(OH)
0
3
]
3+
was performed at different pH values at I = 0.15 M in NaCl(aq) and
questration diagram of the L5 ligand towards Al
3
+
T = 298.15 K. For the Al /3-hydroxy-4-pyridinones complexes, the
lues is reported. The variations of pL0.5 as a function of different
parameters such as pH, temperature and ionic strength often follow
linear or polynomial patterns, as shown in Fig. S7 for the dependence of
pL0.5 for the above system on pH.
3+
pL0.5 values increase with pH, as reported in Table 4. For the Al /L3
3
+
and Al /L5 systems at pH > 6.0 the pL0.5 values decrease, owing to
Table 4
From the evaluation of logβ110 values reported in the Table 3, it
would seem that L4 is featured by the highest capacity to bind the metal
cation with respect to the other ligands, while L1 by the lowest one. On
the other hand, the analysis of the pL0.5 values calculated from po-
tentiometric data at pH = 4.0 (Table 4), suggested that L2, which dis-
pL0.5 values of Al3+/3-hydroxy-4-pyridinones systems from potentio-
metric data at different pH, I = 0.15 M in NaCl(aq) and T = 298.15 K.
System
pH
pL0.5
3
+
Al /L1
4.0
7.0
4.0
7.0
2.0
3.5
4.0
5.0
6.0
7.0
2.5
3.5
4.0
5.0
6.0
7.0
2.5
3.0
4.0
5.0
6.0
7.0
5.8
6.3
7.0
7.7
2.7
5.4
6.5
8.1
8.3
6.4
3.0
5.6
6.6
8.0
8.7
7.4
2.2
3.8
4.4
5.4
5.6
3.9
2
plays two eCH groups on the alkyl chain, unlike all the other com-
3
+
Al /L1
3
+
pounds bearing propyl groups, shows a higher sequestering ability, that
decreases with alkyl chain length increase. Therefore, as already
pointed out, the length of the alkylic spacer plays a key role in the
coordination behavior of these ligands.
Al /L2
3
+
Al /L2
3
+
Al /L3
3
+
Al /L3
3
+
Al /L3
Furthermore, the higher stability constants and pL0.5 values de-
termined for the eCOOH containing ligands can be an indication that
the interaction with Al3+ occurs via chelation. As it can be observed in
3
+
Al /L3
3
+
Al /L3
3
+
Al /L3
3
+
3+
Al /L4
Fig. S8 and from the data in Table 4, the pL0.5 values for the Al /L5
3
+
Al /L4
system are lower than the other ones. This evidence can be confidently
attributed to the absence of carboxylate groups in L5. Taking into ac-
count these considerations and using the stability constants obtained
from potentiometric data, at the experimental conditions already
mentioned, it is possible to observe the following sequestering ability:
3
+
Al /L4
3
+
Al /L4
3
+
Al /L4
3
+
Al /L4
3
+
Al /L5
3
+
Al /L5
3
+
Al /L5
L2 > L4 > L3 > L1 > L5
3
+
Al /L5
3
+
Al /L5
3
+
Al /L5
125

A. Irto et al.
Journal of Inorganic Biochemistry 186 (2018) 116–129
1
0
0
0
0
0
.0
.8
.6
.4
.2
.0
Table 6
(
3−z)
pH = 5.0
Literature stability constants reported for [AlL]
different ionic strengths and ionic media.
species at T = 298.15 K,
System
logβ110
Experimental conditions
Ref.
pH = 4.0
Al3+/deferiprone
11.91
I = 0.15 M in NaCl(aq)
I = 0.10 M in KCl(aq)
[48]
[53]
[37]
[16]
[16]
1
2.20
pH = 6.0
3
+
I = 0.10 M in Na+ supporting electrolyte
I = 0.10 M in KNO3(aq)
Al /aspartic acid
7.87
13.03
13.04
3
+
Al /L1
3
+
pH = 3.5
Al /EthylL
I = 0.10 M in KNO3(aq)
pH = 7.0
pH = 2.5
I = 0.10 M in sodium supporting electrolyte and T = 298.15 K, stated in
H
H
Table 5, are in accordance with the corresponding logK
values determined here for L2 and L3.
2
and logK
3
2
4
6
8
The acid-base properties of the ligands L4 and L5, were studied by
Santos et coworkers [15,17], at T = 298.15 K and I = 0.10 M in
KNO3(aq) and KCl(aq), respectively; they reported values (see Table 5)
that are fairly in agreement with those here listed, at the already
mentioned experimental conditions.
-
logc
L5
Fig. 6. Sequestration diagram of L5 ligand towards Al3+ at different pH values.
Molar fraction of metal cation complexed vs. the total ligand concentration (as
−
logc) at I = 0.15 M in NaCl(aq) and T = 298.15 K.
As regards the study of the interactions of the 3-hydroxy-4-pyr-
idinones with Al3+, the stability of [AlL]
(3−z)
species determined for all
3.7. Literature data comparison
the systems at I = 0.15 M in NaCl(aq) and T = 298.15 K, is higher than
the ones found for Al3+/deferiprone system, as published by Clevette
et al. [48] and Clarke et al. [53] at the same experimental conditions
The protonaton constants of the 3-hydroxy-4-pyridinones (L1–L5)
determined in the present paper at I = 0.15 M in NaCl(aq) and
T = 298.15 K, are in good agreement with the ones obtained for de-
feriprone by Bretti and coworkers [47], Celvette et al. [48] and Ma
et al. [49], at the same experimental conditions, as reported in Table 5.
Crisponi's group [50–52] and Clarke et al. [53] published the results of
DFP protonation studies at I = 0.10 M in KCl(aq) and T = 298.15 K.
Furthermore, Crisponi and coworkers [51] also reported a thermo-
dynamic study on deferiprone acid-base properties at T = 310.15 K and
different ionic strength values (I = 0.1, 0.2, 0.4, 0.5, 1.0 M in KCl(aq)).
Santos et al. [16] investigated at I = 0.10 M in KNO3(aq) and
T = 298.15 K the acid-base behavior of L1 and other 3-hydroxy-4-pyr-
idinones bearing alkylic chains with different length, in particular the
and at I = 0.10 M in KCl(aq) and T = 298.15 K. These data are reported
(3-z)
in Table 6 for [AlL]
species. The same trend was observed also by
3+
comparing the stability constants obtained for the Al /3,4-HPs sys-
3+
tems with the Al /aspartic acid literature data (Table 6) reported by
Martel et al. [37] at the same conditions of temperature and ionic
strength employed by Clarke et al. [53], but in Na+ ionic medium.
Furthermore, the speciation model and the formation constants of
Al3+/L1 complexes obtained in this work are in agreement with the
ones determined for the same system by Santos et al. [16] at I = 0.10 M
(3−z)
in KNO3(aq) and T = 298.15 K, as listed in Table 6 for [AlL]
species.
3+
The author also published stability constant values on the Al /EthylL
two –CH groups on the alkylic chain) interactions, which are lower
(
2
2-(3-hydroxy-2-ethyl-4-oxopyridin-1(4H)-yl)acetic acid, for simplicity
+
than the corresponding [Al(L1)] one, as already seen for the proto-
nation constants, and therefore showing a stability constants increase
upon alkylic chain length increasing.
here called EthylL (2-(3-hydroxy-2-ethyl-4-oxopyridin-1(4H)-yl)acetic
acid), a ligand which is featured by an ethylic spacer between the
pyridinone nitrogen and the carboxylic group. The protonation constant
values of this last compound, reported in Table 5, are lower than the L1
3.8. In vivo studies on metal sequestering capacity
(
propylic spacer) ones here determined at the same temperature and at
I = 0.15 M in NaCl(aq)
Concerning L1, in the same paper Santos et al. [16] characterized
.
The efficacy of the new hydroxypyridinone ligands L2, L3 and L4 as
H
in vivo chelating agent for the treatment of aluminium overload was
the protonable groups, reporting a logK
group, a logK
1
attributed to the hydroxyl
67
H
H
evaluated on the basis of an animal model overload with Ga, namely
Balb-C mice previously injected with 67Ga-citrate. This option was
based on the fact that, as previously reported [13], it has been widely
2
to the carboxylic group and a logK
3
for the pyridyl
nitrogen atom, and values (Table 5) which are in good agreement with
those reported in this paper.
The behavior in aqueous solution of L2 and L3 could be compared
also with that of aspartic acid [37], from which they are derived. The
protonation constants for the amino and the carboxylic groups at
accepted that Ga3+, as an analogue of Fe
3+
and Al , binds to serum
3+
proteins, mostly through transferrin and accesses to cells through the
transferrin receptor, although there is also evidence that in vivo uptake
may not be mediated only by this serum protein. For that purpose the
Table 5
Protonation constant values reported in the literature at different ionic strengths, ionic media and at T = 298.15 K.
logK H
logK H
logK
H
logK
H
Experimental conditions
Ref.
Ligand
1
2
3
4
Deferiprone
9.82
3.69
3.70
3.84
3.66
3.68
3.57
4.13
4.38
3.71
9.01
9.09
–
–
–
–
–
–
3.34
3.53
–
3.83
3.20
–
–
–
–
–
–
–
–
I = 0.15 M in NaCl(aq)
I = 0.15 M in NaCl(aq)
I = 0.15 M in NaCl(aq)
I = 0.10 M in KCl(aq)
[47]
[48]
[49]
[50–52]
[53]
[50–52]
[16]
[16]
[37]
[15]
9.86
9.86
9.82
9.77
9.70
I = 0.10 M in KCl(aq)
I = 0.10 M in KCl(aq), T = 310.15 K
I = 0.10 M in KNO3(aq)
I = 0.10 M in KNO3(aq)
I = 0.10 M in Na+ ionic medium
I = 0.10 M in KNO3(aq)
I = 0.10 M in KCl(aq)
L1
EthylL
aspartic acid
L4
L5
9.83
9.91
9.66
9.85
10.07
–
1.35
–
[17]
126

A. Irto et al.
Journal of Inorganic Biochemistry 186 (2018) 116–129
modifications induced on the usual 67Ga biodistribution profiles de-
pending on the extra-functional groups. The most striking differences
are related with the blood clearance, kidney uptake and retention and
the rate of total excretion, that is significantly faster after co-adminis-
tration of L3 than any of the other ligands (L2, L4). In fact, all these 3-
hydroxy-4-pyridinones present a similarly high chelating capacity in
solution, attributed to cooperation of the terminal amino-carboxylic
group. However, the in vivo metal uptake (sequestration) depends also
on the pharmacokinetic properties and, importantly, on the interaction
of the ligands (and corresponding metal-complexes) with metal-binding
3
3
2
2
1
1
5
0
5
0
5
0
5
0
3
+
plasma proteins, such as transferrin and albumin, which for Al
human plasma can exist in concentrations ~60 and 30%, respectively
49]. Indeed the small structural differences of the extra-functional
in
Blood Liver Intestine Heart
L3 1 h L3 24 h L2 1 h
Lung Kidney Bone Stomach
L2 24 h L4 1 h L4 24 h
[
groups of these three bioassayed compounds resulted in different mo-
lecular descriptors (see Table 7). Therefore, since out of the three
compounds, L4 presents the lowest MW, highest lipophilicity and
strongest interaction with Human Serum Albumin (HSA), it is likely
that L4 and its complex have higher interaction with other proteins,
with concomitantly slower blood clearance and higher kidney retention
along the renal excretion.
Fig. 7. Biodistribution data in the most relevant organs of 67Ga-citrate after
intraperitoneal injection of L3, L2 or L4 expressed as % I. A./g, 1, and 24 h post-
injection in female Balb-C mice (n = 3).
biodistribution and radioactivity excretion of the radionuclide with
intraperitoneal injection of 0.5 μmole of the ligand solution im-
mediately after intravenous administration of the radiotracer was
compared with the well established distribution of the radiotracer, in
the same animal model.
Finally, taking L3 as a representative example of these extra-
functionalized 3,4-HP chelators, the biodistribution and excretion
profiles of 6 Ga after administration of L3 at 1 h and 24 h, in compar-
ison with those of 67Ga-citrate and deferiprone (DFP) is presented in
Fig. 9. Comparison of the metal biodistribution profiles, induced by our
L3 ligand and by the iron-chelating drug DFP at the same experimental
conditions, indicates that the co-administration of the ligand L3 en-
hanced the clearance of the radiometal 67Ga and the overall excretion
rate of radioactivity from whole animal body more efficiently than the
drug. In fact, at 1 h, 66.1% ± 4.9 of the injected activity was already
eliminated, while after administration of DFP only 26.5% of I.A. was
excreted. Altogether these results suggest that, in our animal model, the
radiometal ion is very efficiently chelated by the 3-hydroxy4-pyr-
idinones functionalized with an amino-carboxylic side chain, in parti-
cular by the ligand L3, which appears as a very promising metal se-
questering agent.
7
Biodistribution profiles of 67Ga in the main organs and tissues after
administration of the different chelators, at 1 h and 24 h are presented
in Fig. 7, while the corresponding excretion profiles are reported in
Fig. 8, in comparison with those of DFP. Analysis of data clearly in-
dicates that the co-administration of any of the ligands interferes in the
tissue distribution of the radiometal, inducing a faster clearance from
blood and from the main organs and from whole body. These findings
are in accordance with the effects induced in the distribution and ex-
6
7
cretion of Ga by other similar ligands, including L1, L4 and L5, re-
ported in previous papers [15,16]. However, a direct comparison
cannot be made, since those results were obtained with different mice
strains. Furthermore, analysis of the bioassays results clearly indicates
1
1
20
00
8
6
4
2
0
0
0
0
0
1
h
24 h
1 h
24 h
1 h
24 h
1 h
24 h
L3
L2
L4
DFP
Fig. 8. Excretion of 67Ga-citrate after intraperitoneal injection of L3, L2, L4 and DFP expressed as % I. A. 1, and 24 h post-injection in female Balb-C mice (n = 3).
Table 7
Pharmacokinetic properties as predicted in silico by software QikProp v.2.5 [54].
Comp.
MW
Donors H bonds
Acceptors H bonds
clogP O/W
logK HSA serum binding
log BB
% Human oral absorpt. GI
DFP
L1
L2
L3
L4
139
211
283
297
240
182
1
2
4
4
4
3
3.25
5.25
7.75
7.75
6.25
4.25
0.601
0.723
−2.815
−2.632
−2.082
−0.150
−0.554
−0.708
−1.151
−1.704
−0.800
−0.584
−0.296
−1.587−
−2.417−
−2.507
−1.655
−0.747
78
41
16
21
33
62
L5
127

A. Irto et al.
Journal of Inorganic Biochemistry 186 (2018) 116–129
investigations, which clearly indicated that the coordination of Al
3
+
occurs via the hydroxo-oxo bidentate moiety of the hydroxypyridinone
side of the ligand. However, based on HypNMR calculations for the
complex species, it is suggested that, in some cases, the terminal side-
chain carboxylate groups may also interact with Al3+. The sequestering
ability of the 3-hydroxy-4-pyridinones towards aluminium was in-
vestigated by the calculation of the pL0.5 values, highlighting that the
presence of carboxylic groups in the ligands enhances their affinity
1
00
8
6
4
2
0
0
0
0
0
Excretion
Kidney
Heart
3+
towards Al , since the trend L2 > L4 > L3 > L1 > L5 was ob-
served. Finally, the in vivo metal sequestration indicates for L2, L3 and
L4 a higher capacity for metal clearance than the drug DFP, thus ap-
pearing as promising aluminium chelating drugs.
Blood
1
h
24 h 1 h 24 h 1 h 24 h
Citrate
Liver
DFP
Heart
L3
Kidney
Acknowledgements
Blood
Lung
Bone
Excretion
Fig. 9. Biodistribution of 67Ga-citrate and 67Ga-citrate after intraperitoneal
injection of L3 or DFP expressed as % I. A./g, 1, and 24 h post-injection in
female Balb-C mice (n = 3).
The authors from the University of Messina thank MIUR (Ministero
dell'Istruzione, dell'Università e della Ricerca) for financial support (co-
funded PRIN project with prot. 2015MP34H3).
The authors from (IST) University of Lisbon thank the Portuguese
Fundação para a Ciência e a Tecnologia (FCT) for financial support to
the projects UID/QUI/00100/2013 and PEst-C/SAU/LA0001/2011-
3.9. Pharmacokinetic properties
2
013, and the postdoctoral fellowship (KC). Acknowledgements are
To evaluate the drug-likeness of the 3,4-hydroxypyridinones in
also due to the Portuguese NMR (IST-UL Center) and Mass
Spectrometry Networks (Node IST-CTN) for providing access to their
facilities.
study, some descriptors of their pharmacokinetic profiles were calcu-
lated using QikProp program, v. 2.5 [54]. Thus, parameters such as the
calculated octanol–water partition coefficient (clogP), the ability to
cross the BBB (log BB), the capacity to be absorbed through the gastro-
intestinal gut, and binding with human serum albumin, have been
calculated and are summarized in Table 7. Analysis of the data in this
table show that all the compounds, presented appropriate lipophilicity
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.jinorgbio.2018.05.017.
(
clogP < 5), molecular weights (< 500) and number of H-bond ac-
ceptors (< 10) or H-bond donors (< 5). All the compounds fulfill the
criteria for the Lipinski's rule of five [55], thus indicating a good
bioavailability and oral activity as potential drugs. Furthermore, the
compounds L2, L3, L4 with the best metal sequestering capacity are
more hydrophilic than the others, but they present good capacity for
membrane permeation as indicated by log BB values (> −3), the
highest binding interaction with Human serum albumin (HSA), and
moderate % Human oral absorption by the gastrointestinal (GI) track
References
[
1] S. Chaves, L. Piemontese, A. Hiremathad, M.A. Santos, Curr. Med. Chem. 25 (2018)
7–112.
[2] G. Crisponi, V.M. Nurchi, V. Bertolasi, M. Remelli, G. Faa, Coord. Chem. Rev. 256
2012) 89–104.
9
(
[
3] Z.D. Liu, R.C. Hider, Medicinal Research Reviews, 22 John Wiley & Sons, Inc., 2002,
pp. 26–64.
[4] M.A. Santos, S. Chaves, Future Med. Chem. 7 (2015) 383–410.
[5] S. Gama, P. Dron, S. Chaves, E. Farkas, M.A. Santos, Dalton Trans. (2009)
[
54]. Conversely the other compounds (DFP, L1 and L5) presented
6
141–6150.
higher lipophilicity and % GI absorption but less interaction with HSA
and lower metal sequestering capacity.
[
[
6] J.B. Neilands, Annu. Rev. Biochem. 50 (1981) 715–731.
7] G. Anderegg, F. L'Eplattenier, G. Schwarzenbach, Helv. Chim. Acta, 46 WILEY-VCH
Verlag GmbH, 1963, pp. 1400–1408.
[
8] B.J. Hernlem, L.M. Vane, G.D. Sayles, Inorg. Chim. Acta 244 (1996) 179–184.
4
. Conclusions
[9] J.B. Porter, E.R. Huehns, Baillieres Clin. Haematol. 2 (1989) 459–474.
10] M.A. Santos, Coord. Chem. Rev. 252 (2008) 1213–1224.
11] M.A. Santos, M.A. Esteves, M.C.T. Vaz, M.L.S.S. Gonçalves, J. Chem. Soc. Dalton
Trans. (1993) 927–932.
[
[
A set of extrafunctionalized 3-hydroxy-4-pyridinones (3,4-HPs) was
synthesized; their acid-base properties and binding ability towards the
metal cation Al
[12] M.A. Santos, S. Gama, L. Gano, G. Cantinho, E. Farkas, Dalton Trans. (2004)
3
+
3772–3781.
in NaCl aqueous solutions were investigated in so-
[
[
13] M.A. Santos, S. Gama, L. Gano, E. Farkas, J. Inorg. Biochem. 99 (2005) 1845–1852.
14] G.J. Kontoghiorghes, L.N. Sheppard, T.A. Cresswell, et al. (Ed.), E.P. Office, J. A.
Kemp & Co., 14 South Square Gray's Inn, London WC1R 5EU (GB), EP 0 335 745 A1,
Great Britain, 1989, pp. 1–22.
lution and the in vivo sequestering capacity was studied. The new 3-
hydroxy-4-pyridinones were prepared from a naturally occurring
compound (i.e. maltol) and N-functionalized to introduce differentia-
tion in terms of pharmacokinetic parameters and interaction with
proteins, while keeping the bioavailability parameters. The acid-base
behavior and the binding ability of these 3,4-HPs towards the Al3+
were studied through potentiometric and UV–Vis spectrophotometric
experiments carried out at T = 298.15 K and I = 0.15 M in NaCl(aq)
The speciation models were characterized by Al
with different stoichiometry, such as simple metal-ligand (1:1 stoi-
chiometry) and polynuclear species, depending on the ligands different
charges and structures.
[
[
15] M.A. Santos, M. Gil, L. Gano, S. Chaves, J. Biol. Inorg. Chem. 10 (2005) 564.
16] M.A. Santos, M. Gil, S. Marques, L. Gano, G. Cantinho, S. Chaves, J. Inorg. Biochem.
92 (2002) 43–54.
[17] M.A. Santos, R. Grazina, A.Q. Neto, G. Cantinho, L. Gano, L. Patrıcio, J. Inorg.
Biochem. 304 (2000) 303–311.
[
18] J. Buffle, Complexation Reactions in Aquatic Systems: An Analytical Approach,
1988).
[19] F.J. Millero, Physical Chemistry of Natural Waters, (2001).
.
(
(
3p+r−qz)
p
L
q
H
r
species
[
20] C. Lentner, Geigy Scientific Tables, (1981).
[
21] A. Gianguzza, O. Giuffrè, D. Piazzese, S. Sammartano, Coord. Chem. Rev. 256
(
2012) 222–239.
[22] H.A. Flaschka, EDTA Titration, (1959).
[23] W.L.F. Armarego, D.D. Perrin, Purification of Laboratory Chemicals, (1999).
24] M.A. Santos, M. Gil, S. Marques, L. Gano, G. Cantinho, S. Chaves, J. Inorg. Biochem.
2 (2002) 43–54.
25] A. Braibanti, G. Ostacoli, P. Paoletti, L.D. Pettit, S. Sammartano, Pure Appl. Chem.
9 (1987) 1721.
From data obtained it was concluded that the stability of [AlL](3-z)
[
[
species follows the trend:
9
L4 > L2 > L3 > L5 > L1
5
[
[
26] C. De Stefano, P. Princi, C. Rigano, S. Sammartano, Ann. Chim. 7 (1987) 643–675.
27] P. Gans, A. Sabatini, A. Vacca, Talanta 43 (1996) 1739–1753.
Furthermore, for L2, the models proposed for its protonation as well
3
+
1
as for L2 complexing ability towards Al were confirmed by H NMR
128

A. Irto et al.
Journal of Inorganic Biochemistry 186 (2018) 116–129
[42] L. Alderighi, P. Gans, A. Ienco, D. Peters, A. Sabatini, A. Vacca, Coordin, Chem. Rev.
184 (1999) 311–318.
[28] C. Frassineti, S. Ghelli, P. Gans, A. Sabatini, M.S. Moruzzi, A. Vacca, Anal. Biochem.
31 (1995) 374–382.
2
[
29] C. De Stefano, S. Sammartano, P. Mineo, C. Rigano, A. Gianguzza, E. Pelizzetti,
S. Sammartano (Eds.), Marine Chemistry - An Environmental Analytical Chemistry
Approach, Kluwer Academic Publishers, Amsterdam, 1997, pp. 71–83.
30] C. Agouridas, A. Denis, J.-M. Auger, Y. Benedetti, A. Bonnefoy, F. Bretin, J.-
F. Chantot, A. Dussarat, C. Fromentin, S.G. D'Ambrières, S. Lachaud, P. Laurin, O. Le
Martret, V. Loyau, N. Tessot, J. Med. Chem. 41 (1998) 4080–4100.
31] A.W. Williamson, Justus Liebigs Ann. Chem. 77 (1851) 37–49.
32] A.W. Williamson, Quart. J. Chem. Soc. 4 (1852) 229–239.
[43] S. Chaves, M. Gil, S. Marques, L. Gano, M.A. Santos, J. Inorg. Biochem. 97 (2003)
161–172.
[44] D. Aiello, P. Cardiano, R.M. Cigala, P. Gans, F. Giacobello, O. Giuffrè, A. Napoli,
S. Sammartano, J. Chem. Eng. Data 62 (2017) 3981–3990.
[45] P. Cardiano, F. Giacobello, O. Giuffrè, S. Sammartano, J. Mol. Liq. 232 (2017)
45–54.
[
[
[
[
[
[
[
[46] F. Crea, C. De Stefano, C. Foti, D. Milea, S. Sammartano, Curr. Med. Chem. 21
(2014) 3819–3836.
33] R.L.N. Harris, Aust. J. Chem. 29 (1976) 1335–1339.
[47] C. Bretti, R.M. Cigala, F. Crea, G. Lando, S. Sammartano, J. Chem. Thermodyn. 77
(2014) 98–106.
[48] D.J. Clevette, W.O. Nelson, A. Nordin, C. Orvig, S. Sjoeberg, Inorg. Chem. 28 (1989)
2079–2081.
34] L. Saghaie, M.M. Sadeghi, A. Nikazma, Res. Pharm. Sci. 1 (2006) 40–48.
35] C.F. Baes, R.E. Mesmer, John Wiley & Sons, New York (1976).
36] R.M. Cigala, C. De Stefano, A. Giacalone, A. Gianguzza, Chem. Speciat. Bioavailab.
2
3 (2011) 33–37.
[49] R. Ma, J.J. Reibenspies, A.E. Martell, Inorg. Chim. Acta 223 (1994) 21–29.
[50] G. Crisponi, V.M. Nurchi, M.A. Zoroddu, Thalassemia Reports, 4 (2014), p. 2046.
[51] G. Crisponi, M. Remelli, Coord. Chem. Rev. 252 (2008) 1225–1240.
[52] V.M. Nurchi, G. Crisponi, T. Pivetta, M. Donatoni, M. Remelli, J. Inorg. Biochem.
102 (2008) 684–692.
[
37] A.E. Martell, R.M. Smith, R.J. Motekaitis, NIST Critically Selected Stability
Constants of Metal Complexes Database, 8.0 (2004).
[
[
[
38] F.J. Millero, R. Woosley, Environ. Sci. Technol. 43 (2009) 1818–1823.
39] D. Pettit, K.K. Powell, Stability Constants Database, Academic Software, (1997).
40] P. Cardiano, R.M. Cigala, F. Crea, F. Giacobello, O. Giuffrè, A. Irto, G. Lando,
S. Sammartano, Chemosphere 186 (2017) 535–545.
[53] E.T. Clarke, A.E. Martell, Inorg. Chim. Acta 191 (1992) 57–63.
[54] L.L.C. Schrödinger, New York, NY, (2005).
[
41] L.O. Öhman, Department of Inorganic Chemistry, University of Umeà, Sweden,
[55] C.A. Lipinski, F. Lombardo, B.W. Dominy, P.J. Feeney, Adv. Drug Deliv. Rev. 46
(2001) 3–26.
1
983.
129