New tris(hydroxypyridinones) as iron and aluminium sequestering agents: Synthesis, complexation and in vivo studies

Article abstract of DOI:10.1002/chem.201001335

Two new tripodal tris(3-hydroxy-4-pyridinone) hexadentate chelators - NTA(BuHP)₃ and NTP(PrHP)₃ (NTA = nitrilotriacetic acid, NTP = nitrilotripropionic acid, HP = hydroxypyridone)- have been developed and studied in solution for their iron and aluminium binding affinity, and also assayed in vivo for their capacity to remove metal from an animal model that is overloaded. These chelators are positional isomers, possessing identical general structures based on aminotricarboxylic acid skeletons attached to three bidentate 3-hydroxy-4-pyridinones (3,4-HPs), but differing in the position of the amide linkage along the chelating arm . In spite of expected differences in the tripodal ligands, such as acidity and hydrogen-bonding networks, they share important properties, namely, a mild hydrophilic character (log P ca. -1.2 to -1.4) and a strong chelating affinity for Fe and Al (pFe = 27.9 and pAl = 22.0 for NTA(BuHP)₃; pFe = 29.4 and pAl = 22.4 for NTP(PrHP)₃). They also evidenced identical effects on the biodistribution and on the excretion of a radiotracer (⁶⁷Ga) previously administered to mice, as models of iron overload animals. Comparison of the new compounds with reported analogues shows good improvement in terms of solution and in vivo sequestering properties, thus giving support to expectations about their potential clinical application as metal removal agents.

Full text of DOI:10.1002/chem.201001335

FULL PAPER
DOI: 10.1002/chem.201001335
New Tris(hydroxypyridinones) as Iron and Aluminium Sequestering Agents:
Synthesis, Complexation and In Vivo Studies
Sꢀlvia Chaves,[a] Sꢁrgio M. Marques,[a] Andrꢁ M. F. Matos,[a] Andreia Nunes,[a]
Lurdes Gano,[b] Tiziano Tuccinardi,[c] Adriano Martinelli,[c] and M. Amꢁlia Santos*[a]
Abstract: Two new tripodal tris(3-hy-
droxy-4-pyridinone) hexadentate chela-
nones (3,4-HPs), but differing in the
position of the amide linkage along the
chelating “arm”. In spite of expected
differences in the tripodal ligands, such
as acidity and hydrogen-bonding net-
works, they share important properties,
namely, a mild hydrophilic character
(logP ca. À1.2 to À1.4) and a strong
chelating affinity for Fe and Al (pFe=
pFe=29.4 and pAl=22.4 for NTP-
ACHTUNGTNER(NUNG PrHP)3). They also evidenced identi-
tors—NTA
G
N
cal effects on the biodistribution and
on the excretion of a radiotracer (67Ga)
previously administered to mice, as
models of iron overload animals. Com-
parison of the new compounds with re-
ported analogues shows good improve-
ment in terms of solution and in vivo
sequestering properties, thus giving
support to expectations about their po-
tential clinical application as metal re-
moval agents.
(NTA=nitrilotriacetic acid, NTP=ni-
trilotripropionic acid, HP=hydroxypyr-
idone)—have been developed and
studied in solution for their iron and
aluminium binding affinity, and also as-
sayed in vivo for their capacity to
remove metal from an animal model
that is overloaded. These chelators are
positional isomers, possessing identical
general structures based on aminotri-
carboxylic acid skeletons attached to
three bidentate 3-hydroxy-4-pyridi-
27.9 and pAl=22.0 for NTA
ACHTUNGTRENNUNG
Keywords: biodistribution
·
lates · hydroxypyridinones · iron ·
sequestering agents
Introduction
(e.g., aluminium) in certain organs, resulting in serious dis-
eases in the brain (dialysis encephalopathy) or in the bone
(osteomalacia).[2,3] To prevent or relieve these problems, the
design of new orally active specific chelators has been a
major objective and a number of synthetic chelators have
been reported.[4,5]
Iron is an essential bioelement, but when in excess it is
toxic. Iron overload diseases, such as hemochromatosis or
beta-thalassemia-associated transfusion hemosiderosis, can
be fatal unless patients follow iron-chelating therapy proto-
cols.[1] Environmental exposure or therapeutic contamina-
tion can also lead to accumulation of other hard metal ions
Desferrioxamine B (DFO) was the first and the only hex-
adentate ligand approved by the Food and Drug Adminis-
tration (FDA) for chelating therapy of patients with transfu-
sion iron overload. However, the drawbacks associated with
this therapeutic, such as the low patient compliance due to
the need for daily subcutaneous (sc) or intravenous (iv) ad-
ministration, led to the investigation of new chelators.[4]
Major advances in terms of new iron chelators appeared
in form of the 3-hydroxy-4-pyridinone (3,4-HP) family of
compounds, that is, after the disclosure of Deferiprone
(DFP, 1,2-dimethyl-3-hydroxypyridin-4-one) as an orally
active drug for the treatment of iron overload.[6] This type of
compound has a high affinity for hard metal ions and fa-
vourable toxicity profiles; on the other hand, they can be
easily designed for extra-functionalisation to guarantee high
bioavailability as well as targeting properties. Therefore, this
family of bidentate chelators has long been considered good
[a] Prof. Dr. S. Chaves, Dr. S. M. Marques, Dipl.-Chem. A. M. F. Matos,
Dipl.-Chem. A. Nunes, Prof. Dr. M. A. Santos
Centro de Quꢀmica Estrutural, Instituto Superior Tꢁcnico
Av. Rovisco Pais 1, 1049-001 Lisboa (Portugal)
Fax : (+351)21-8464455
E-mail: masantos@ist.utl.pt
[b] Dr. L. Gano
Instituto Tecnolꢂgico e Nuclear, Estrada Nacional N8 10
2686-953 Sacavꢁm (Portugal)
[c] Dr. T. Tuccinardi, Prof. Dr. A. Martinelli
Universitꢃ degli Studi di Pisa
Dipartimento di Scienze Farmaceutiche
Via Bonanno 6, 56126 Pisa (Italy)
Chem. Eur. J. 2010, 16, 10535 – 10545
ꢄ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
10535

pharmaceutical targets for chelating therapeutics[7–9] or even
metallodrugs.[10] Nevertheless, some drawbacks (toxicity) as-
sociated with the low denticity of DFP and concomitant
high dosage demands for competition with bioligands led to
the search for ligands with increased denticity. Aside from
the recent emergence of new tridentate iron chelators ap-
proved by the FDA (desferasirox or ICL670),[11] several re-
search groups have been implementing drug design strat-
egies to increase the ligand denticity. Specifically, some poly-
dentate 3,4-HP-based chelators have been developed, that
is, tetradentate ligands, to be used alone or in combined-
ligand therapeutic protocols,[12,13] and hexadentate ligands,
to fulfil the metal coordination by the formation of 1:1 M3+
–L (metal–ligand) complexes with improved thermodynamic
and kinetic stability.[14,15]
pounds, this paper is focused on the study of the physico-
chemical properties of ligands and complexes in solution, es-
pecially those related to the lipo-hydrophilic character and
complexation ability towards ironACHTUNTRGENN(GU III) and aluminiumACHTUNGTRENNUNG(III),
as well as some modelling studies on the iron complex struc-
tures. Their in vivo behaviour as potential sequestering
agents is also assessed, that is, their ability for the 67Ga mo-
bilisation from mice previously injected with this radiotrac-
er, as an animal model of Fe or Al overload.
Results and Discussion
Design and chemistry of the compounds: Notwithstanding
the high interest in 3,4-HPs as iron chelating agents and the
importance of polydenticity as a determining factor for the
thermodynamic stability of the metal complexes, up to
around five years ago only bidentate 3,4-HP compounds had
been explored, as opposed to the other HP classes.[15] There-
fore, we have implemented a strategy for developing poly-
dentate 3,4-HP-based compounds by attachment of two or
three N-alkylamine-3,4-HP (“arms”) to suitable polycarbox-
ylate-containing skeletons to form, respectively, tetra- and
hexadentate ligands pre-organised to wrap and coordinate
Herein, we report results on the development and study
of two new hydroxypyridinone-based tripodal hexadentate
chelators with three 3,4-HP units appended to amino-tris-
carboxylic anchoring skeletons, namely, nitrilotriacetic acid
(NTA) and nitrilotripropionic acid (NTP) (Scheme 1a). Be-
sides the design and the synthetic strategy of the new com-
the metal ions.[16] For the hexadentate ligands, tris
ACTHUNTGRENUNG(3,4-HP),
we first appended three of the chelating arms to a Kemp
acid, a cyclohexane-based scaffold containing three carbox-
ylic groups in axial positions (Scheme 1b).[15] Despite the
high stability of the corresponding M3+ complexes, the fully
coordinated species were quite insoluble in water. There-
fore, to bring some hydrophilicity to the chelator/complex
system, we designed a new set of compounds based on ami-
notricarboxylic acid skeletons with a central amine group as
the anchoring point for the three carboxylic acids. Based on
a preliminary modelling study, it seemed that, between the
apical tertiary amine and the 3,4-HP chelating units, a
spacer containing five methylenic groups plus an amide
bond (for the linkage between the anchoring carboxylic and
the amine groups of the HP arms) was well suited for wrap-
ping the iron. Accordingly, two trisACTHNUTRGNEUNG(3,4-HP) chelators were
À
À
À
developed with arms of the type
in which p+m=5, but with different peptide bond posi-
tions: p=1, m=4 for NTA(BuHP)3; p=2, m=3 for NTP-
(PrHP)3.
The synthetic route for the preparation of the compounds
ACHTUNGTREN(NUG CH2)p C(O)NH ACHTUNGTRENNUNG(CH2)m,
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
involved the coupling of the aminotriacid skeletons (NTA
and NTP) with the free amine group of an N-alkylamine, O-
benzylated monohydroxypyridinone arm (alkyl=butyl or
propyl), through activation of the anchoring carboxylic
groups with O-(benzotriazol-1-yl)-N,N,N’,N’-tetramethyluro-
nium tetrafluoroborate (TBTU), in dry DMF at low temper-
ature and under nitrogen. This methodology afforded the
corresponding benzyl-protected tris-3,4-HP derivatives,
which were isolated by recrystallisation or silica-gel chroma-
tography. The final products were obtained by removal of
the protecting groups through standard methods of hydroge-
nolysis.
Scheme 1. Structural formulae of the studied (a) and reported (b) hexa-
dentate 3,4-HP ligands.
10536
www.chemeurj.org
ꢄ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 10535 – 10545

Iron and Aluminium Sequestering Agents
FULL PAPER
Physico-chemical characterisation of the compounds:
three other protonation constants with values around 2.3–
3.9 attributed to the protonation of the pyridinyl nitrogen
atoms of the same moieties (see Table 1). The calculated
protonation constants for these two processes are analogous
to the corresponding values found in the literature (9.2–
10.2/2.9–4[15] and 9.77/3.62[17]) for the Kemp derivatives and
DFP, respectively. Moreover, the titration curve of NTP-
The new tripodal tris(hydroxypyridinones) (NTA
ACHTUNGERTN(NUNG BuHP)3
and NTP(PrHP)3) were studied in terms of their acid–base
ACHTUNGTRENNUNG
properties and lipo-hydrophilic character. Both compounds
were isolated in the neutral form (H3L), but when fully pro-
tonated (H7L4+) they have seven dissociable protons. The
protonation processes were mainly studied by potentiometry
and fitting analysis of the corresponding titration data, but,
ACHTUNGTERN(NGNU PrHP)3 shows a further intermediate buffer zone at around
1
in the case of NTA
N
pH 7, which is attributed to the protonation of the backbone
amine group. Its logK value (6.77) is much lower than that
for NTP (logK1 =9.65),[20] which can be ascribed to intramo-
lecular hydrogen bonds of the type CONH···N, in a probable
trifurcated geometry with the apical tertiary amine of the
tion was also performed to identify the respective protona-
tion sequence and to determine the very low logK7 value
(<2). Table 1 summarises the stepwise protonation con-
stants (logKi), calculated for the compounds under study, as
well as reported values for the tris(hydroxypyridinone)–
NTP skeleton. In the case of NTAACHTNUTRGNEUNG(BuHP)3, the logK value
Kemp acid derivatives (Kemp
(BuHP)3)[15] and for the commercial drug DFP.[17]
The potentiometric studies revealed that NTA
and NTP(PrHP)3 have three protonation constants with
G
corresponding to the protonation of the apical amine ap-
peared to be quite low when calculated by potentiometry
and so a 1H NMR spectroscopic titration was carried out
ACHTUNGTRENNUNG
1
N
(see Figure 1). Analysis of H NMR titration curve profiles
values in the range of 8.9–10, attributable to the hydroxyl
groups of the three hydroxypyridinone moieties, as well as
suggests that the NTAACTHNGUTERN(NUG BuHP)3 protonation sequence fol-
lows the order: hydroxyl groups (at pDꢀ11, peaks 1 and 2
are downfield shifted), pyridin-
yl nitrogen atoms (downfield
shifts of peaks 1–3 and 6–8 at
Table 1. Stepwise protonation constants (logKi) and partition coefficients (logP) for a set of tripodal com-
pounds and DFP, as well as the global formation constants of their M3+ complexes (T=(25.0Æ0.1)8C, I=
0.1m KCl).
pDꢀ4.5);
apical
N-amine
group (downfield shift of
peak 4 for pD<3). This lowest
pD buffer region indicates an
enormous decrease in the pro-
tonation constant of the apical
nitrogen compared with the
corresponding parent com-
pound, NTA (logK1 =9.2).[20]
This unexpectedly low value
obtained for the protonation of
this apical amine group can be
explained by the formation of
intramolecular hydrogen-bond-
ing (CONH···N) networks that
Ligand
logKi
MpHqLr (p,q,r) logb (FepHqLr) logb (AlpHqLr) logP
9.98(2)
9.83(3)
8.94(4)
3.88(4)
3.11(5)
2.35(5)
1.4(1)[b]
A
42.30(2)[a]
40.81(5)[a]
38.53(3)[a]
33.50(1)[a]
–
R
37.45(4)
33.55(1)
27.65(6)
AHCTUNGTRENNUNG
G
À1.40
pM[c]
(1,5,1)
(1,3,1)
(1,1,1)
(1,0,1)
27.9[d]
22.0
9.946(9)
9.84(1)
9.091(8)
6.77(1)
3.81(1)
3.14(1)
2.76(2)
N
47.62(5)
45.29(5)
40.56(3)
35.21(1)
44.69(6)
40.01(5)
34.72(4)
28.13(3)
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
E
À1.24
À1.04
À0.62
are stronger for NTA
ACHTUNGTRENNUNG
pM[c]
(1,4,1)
(1,2,1)
(1,0,1)
29.4
22.4
(p=1) than for NTPACHTUNGTRENNUNG
10.07(4)
9.89(4)
9.18(5)
3.98(3)
3.25(3)
2.91(4)
G
41.50(2)
38.91(2)
33.98(3)[e]
39.33(2)
33.62(3)
27.21(1)
(p=2), due to the formation of
five- and six-membered ring in-
termediates, respectively (see
Scheme 1).
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
[15]
Kemp
Kemp
ACHTUNGTRENNUNG(PrHP)3
pM[c]
(1,4,1)
(1,2,1)
(1,0,1)
28.0
21.2
Despite the acidity differen-
ces in the backbone ammoni-
um proton, the remaining pro-
tonation constants for both li-
gands are very similar because
the corresponding basic centres
are quite far away from the an-
choring skeleton and only very
negligible differences in the
electron-donation effect could
result from the homologous al-
kylic spacers.
10.15(3)
9.91(3)
9.21(4)
4.05(4)
3.38(3)
3.10(3)
G
41.98(2)
39.08(3)
32.97(6)[e]
39.84(3)
33.87(1)
27.95(1)
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
[15]
ACHTUNGTRENNUNG(BuHP)3
pM[c]
(1,0,1)
(1,0,2)
(1,0,3)
pM[c]
26.8
21.8
9.77
3.62
G
15.14
26.68
35.92
19.3[19]
12.20
23.25
32.62
16.0[19]
E
À0.85
DFP[17]
A
À1.03[18]
[a] Values determined in 3% DMSO. [b] Value determined by 1H NMR titration. [c] pM values at pH 7.4 (CM
= 10À6 m, CL/CM = 10). [d] No precipitation in the concentration conditions of pM determination. [e] Deter-
mined spectrophotometrically by competition with ethylenediaminetetraacetic acid (EDTA).
Chem. Eur. J. 2010, 16, 10535 – 10545
ꢄ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10537

M. A. Santos et al.
constants calculated herein for the M3+ complexes are pre-
sented in Table 1, together with values previously deter-
mined for the Kemp derivatives[15] and DFP.[17–19]
Data analysis of the electronic spectra obtained for the
Fe3+–NTA
ACHTUNGTRENNUNG
Figure 1. 1H NMR spectroscopy titration curves of NTA
ACTHNUTRGEN(UGN BuHP)3 (CL =
7.3ꢅ10À4 m).
At physiological pH, the predominant form for both com-
pounds is the neutral species H3L (ca. 82% for NTP(PrHP)3
and 98% for NTA(BuHP)3). Thus, some lipophilic nature
AHCTUNGTRENNUNG
ACHTUNGTRENNUNG
can be expected for both ligands, although the lipo-hydro-
philic character is not only determined by the molecular
charge, but also by the capacity to establish solute–solvent
interactions. The lipophilicity of these compounds was as-
sessed through the calculation of their partition coefficients
(logP) between 1-octanol and a Tris-buffered aqueous phase
at pH 7.4. Analysis of the results given in Table 1 shows that
NTAACHTUNGTRENNUNG(BuHP)3 and NTPACHTUTGNREN(NGUN PrHP)3 present similar logP values,
but due to the hydrophilic character of the anchoring groups
these compounds are more hydrophilic than the Kemp de-
rivatives and DFP. Although for both the compounds the
molecular weight exceeded the value suggested by the Lip-
inski rule for oral bioavailability (500 Da);[21] tissue and cell
permeability can also be favoured by other descriptors.
Figure 2. Top) Electronic spectra for the 1:1 Fe3+–NTA
ACTHNUGTRNE(UNG BuHP)3 system,
pH 0.6–5.0. Bottom) Species distribution curves for the same system with
molar extinction coefficients at the maximum absorption wavelengths
(CL/CFe =1, CL =1.8ꢅ10À4 m).
Solution metal-complexation studies: The chelating ability
of the compounds towards ironACHTNUTRGENN(UG III) and aluminiumAHCTUNGTRENN(UGN III) was
estimated on the basis of the global formation constants of
their complexes, which were determined by UV/Vis spectro-
photometric (PSEQUAD program)[22] and potentiometric
(HYPERQUAD program)[23] techniques. Iron complexation
was studied by spectrophotometry due to the high iron-che-
lating affinity of both compounds and concomitant complete
formation of the complexes at the beginning of the potentio-
metric titration. To overcome the low water solubility of the
FeH2L or FeL, which have one, two or three hydroxypyridi-
none chelating moieties, respectively, coordinated to FeIII.
This model is supported by analysis of the data shown in
Figure 2, bottom, containing the species distribution curves
for the proposed equilibrium model merged with some ex-
perimental spectral data, that is, the absorptivity at two
characteristic wavelengths (510 and 460 nm). The maximum
absorption at 510 nm (e=3353mÀ1 cmÀ1) corresponds to the
charge-transfer band of the tetra-coordinated (bischelate)
complex (FeH2L2+), whereas that at 460 nm (e=
4153mÀ1 cmÀ1) corresponds to the hexa-coordinated (tris-
chelate) complex (FeL). These spectral parameters are ac-
cording to literature data for Fe3+–(3-hydroxy-4-pyridinone)
complexes.[24] The refined model also includes two bis-coor-
dinated complexes, FeH4L4+ and FeH5L5+, appearing at
pH<2, with a maximum absorption at around 560 nm, the
neutral FeL complex with NTAACHTNUGTRENUNG(BuHP)3, the predominant
species above around pH 2.5, the iron complexation was
studied in a 3% (v/v) DMSO/water medium. Even so, it was
impossible to pursue the titration over pHꢀ5.5 since precip-
itation occurred, probably due to the formation of the neu-
tral complex and/or mixed hydroxo-ligand complexes under
these experimental conditions. For pH<2, both systems
were also studied spectrophotometrically by using a batch ti-
tration in which the amount of acid to be added was calcu-
lated for the total volume of solution used. The stability
10538
www.chemeurj.org
ꢄ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 10535 – 10545

Iron and Aluminium Sequestering Agents
FULL PAPER
last complex species with the apical backbone nitrogen pro-
tonated.
the major hexa-coordinated complex (tris-chelated species).
Also, some solubility problems, due to the presence of the
neutral FeL complex, are less critical for the Fe3+–NTP-
Regarding the Fe3+–NTP
ACTHNUTRGEN(UNG PrHP)3 system, the spectropho-
tometric titration curves (Figure 3, top) also show two ab-
sorption maxima at 510 (e=3490mÀ1 cmÀ1) and 460 nm (e=
4750mÀ1 cmÀ1), which, according to our model, should corre-
spond to the tetra- (FeH3L) and hexa-coordinated complex
species (FeHL or FeL), respectively (see Figure 3, bottom).
Below pH 2, both the bis-coordinated (FeH5L) and the
tetra-coordinated (FeH3L) species are present.
ACHTUNGTERN(NGNU PrHP)3 system because in this case that species is only pre-
dominant at pH values higher than about 5.5.
The study of Al3+ complexation in solution was per-
formed by potentiometry under 1:1 stoichiometric condi-
tions. Figure 4 shows the species distribution curves for both
Figure 4. Species distribution curves for the 1:1 Al3+–L systems with
NTA
CAHUTTGNRENN(GU BuHP)3 (top) and NTPACHTUNGTREN(NGUN PrHP)3 (bottom) (CAl/CL =1, CL =1.0ꢅ
10À3 m).
Al3+–ligand systems, evidencing the predominance of the
hexa-coordinated tris-chelated species even in acidic condi-
tions (pH>3). ESI-MS spectra of 1:1 solutions for the Al3+
Figure 3. Top) Electronic spectra for the 1:1 Fe3+–NTP
ACTHNUGTRNE(UNG PrHP)3 system,
pH 0.8–8.0. Bottom) Species distribution curves for the same system with
molar extinction coefficients at the maximum absorption wavelengths
(CL/CFe =1, CFe =1.6ꢅ10À4 m).
–NTA
systems confirmed the presence of the complex species [Al-
(BuHP)3 (pH 3.89) and Al3+–NTP
ACHUTGTNRNENUG CAHTUTGNREN(NUGN PrHP)3 (pH 4.00)
2+
+
À
À
ACHUTNGREN(NUG L H)] (375.9) and [AlACHTUNGTREN(NGUN L 2H)] (750.6).
Analysis of the data reported in Table 1, namely, the pM
values at physiological pH, indicates that the chelating abili-
ties of both compounds towards aluminium are quite simi-
ESI-MS measurements, performed for solutions of the
Fe3+–NTA(BuHP)3 and Fe3+–NTP
ACHTUNGTRENNUNG HCATUGNTERN(NUGN PrHP)3 systems under
(1:1) stoichiometric conditions at pH 3.21 and 3.66, respec-
tively, confirmed the existence of the iron complexes as the
lar, whereas NTPACTHNGUTERNNU(G PrHP)3 seems to be a slightly better iron
chelator. This small difference in the complexation behav-
iour of these compounds towards the two metal ions may be
ascribed to some extra difficulty in disrupting the hydrogen-
2+
+
À
À
species [Fe
both systems.
ACHTUNGTRENNUNG(L H)] (390.3) and [FeACHTUNGTREN(NNGU L 2H)] (779.6) for
In summary, both ligands present a similar iron complexa-
tion behaviour, although considerable differences are associ-
ated with the protonation of the apical nitrogen atom in the
ligand backbone structure, namely, the unexpectedly low
bonding network of NTAACTHNGUTER(NNUG BuHP)3 for wrapping the metal
ion with larger dimensions, such as Fe3+ compared with
Al3+ (ionic radius 65 and 52 pm, respectively). Moreover,
NTA
G
ACHTUNGTRENNUNG
logK7 value found for NTA
A
chelators than Kemp
A
ACHTUNGTRENNUNG
A
E
tively. That may be due to the higher backbone flexibility of
the nitrilo- than the Kemp-based derivatives, with enhance-
ment of the metal wrapping ability and the complex stabili-
ty. Furthermore, in opposition to the Kemp derivatives,
monochelated species FeH5L appears only at very low pH
values, whereas the last deprotonation of the Fe3+–NTP-
ACHTUNGTRENNUNG(PrHP)3 system occurs at higher pH values and so FeHL is
Chem. Eur. J. 2010, 16, 10535 – 10545
ꢄ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10539

M. A. Santos et al.
identical arm flexibility could be expected for these com-
pounds under study, due to the same number and type of
atoms between the anchoring tertiary amine and the chelat-
ing moieties, but the different positioning of the attachment
point (amide linkage) of both sub-arms contribute to some
differences in complex formation.
Comparison between the metal (M) chelating affinities of
ligands with different proton dependency and denticity is
usually made on the basis of pM values (pM=Àlog[M] with
CL/CM =10 and CM =10À6 m) at a specific pH, usually physio-
logical pH (7.4). Therefore, pM values were calculated for
the Fe and Al complexes with the ligands under study and
are depicted on Table 2, together with the corresponding
hexadentate ligands towards Ga3+ (pGa=27.29 for NTP-
AHCTUNGTRENNUNG
(PrHP)3[33]), a gallium radiotracer (67Ga) was used as in vivo
model for the iron mobilisation studies (see below).
Molecular modelling of the iron complexes: Since efforts to
obtain good crystalline samples of the ferric complexes for
X-ray diffraction have been unsuccessful up to now (al-
though trials are still on going), it was decided to carry out
molecular modelling studies to provide further insight into
both ferric complex structures. Therefore, molecular simula-
tions were carried out with full geometry optimisation of the
iron complexes by quantum mechanical calculations based
on DFT methods included in the Gaussian 03 program soft-
ware,[34] with the B3LYP hybrid functional.[35] No symmetry
constrains were enforced during geometry optimisation.
Both energy-minimised complex structures are quite similar
in terms of the metal hexacoordination and octahedral ge-
ometry (Figure 5), in agreement with our expectations and
solution spectroscopic parameters.
Table 2. The pM[a] values at physiological pH for the ligands under study
and other reported chelators.
Ligand
NTP(PrHP)3
NTA(BuHP)3
pFe
pAl
G
29.4
27.9[b]
28.0
26.8
26.7
25.5
27.2
25.8
26.3
30.9
23.4
24.6
26.5
19.3
20.3
22.4
22.0
21.2
21.8
–
ACHTUNGTRENNUNG
[15]
Kemp
ACHTUNGTRENNUNG(PrHP)3
[15]
KempACHTUNGTRENNUNG(BuHP)3
Tren(Me-3,2-HOPO)[24]
3,4-LI(Me-3,2-HOPO)[25]
–
–
CP254[26]
[27]
IDA
N
18.8
19.0
–
16.2
15.7
19.3
16.0
14.5
[12]
EDTAACHTUNGTRENNUNG(3,4-HP)2
HBED[28]
EDTA[29]
DTPA[29]
DFO[30,31]
DFP[19]
transferrin[30,32]
[a] CM =10À6 m, CL/CM =10. [b] No precipitation for the low concentration
values of pM determination.
values for a set of synthetic and biological ligands, namely,
polyhydroxypyridinone-based ligands, such as hexadentate
3,2-HP (tripodal Tren(Me-3,2-HOPO) (Tren=tris(2-amino-
ethyl)amine, Me-3,2-HOPO = 1-methyl-3,2-hydroxypyridi-
none),[24] linear 3,4-LI(Me-3,2-HOPO) (3,4-LI = N1-(3-ami-
nopropyl)butane-1,4-diamine),[25] a hexadentate tripodal 3,4-
HP (CP254),[26] tetradentate 3,4-HP (IDA
ACHTUNGTRNE(NUNG 3,4-HP)2 (IDA =
iminodiacetic acid),[27] EDTA(3,4-HP)2[12]), and some well
AHCTUNGTRENNUNG
known non-hydroxypyridinone strong synthetic chelators
such as N,N’-bis(2-hydroxybenzyl)ethylenediamine-N,N’-di-
acetic acid (HBED),[28] EDTA,[29] DTPA= diethylene tri-
amine pentaacetic acid,[29] desferrioxamine (DFO),[30,31]
DFP,[19] and the endogenous ligand, transferrin.[30,32] Analysis
reveals that the herein developed hexadentate chelators are
more potent than the others, except HBED, reported in
Table 2 towards iron. Thus, the high metal chelating affinity
revealed by these novel hexadentate hydroxypyridinones
undoubtedly constitutes a strong motivating factor for pur-
suing biodistribution studies with both ligands to evaluate
their removal capacity towards hard metal ions.
Figure 5. DFT-minimised structures of complexes Fe3+–L with NTA-
À
(BuHP)3 (top) and NTP
U
Based on the known similar high chelating affinity of 3,4-
HP ligands for hard metal ions (M3+: M=Fe, Al, Ga)[15] and
on a preliminary evaluation of the capacity of one of these
tahedral coordination bonds (grey); hydrogen bond (black). For clarity,
only heteroatoms are labelled, while carbon atoms are coloured grey and
hydrogen atoms are white.
10540
www.chemeurj.org
ꢄ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 10535 – 10545

Iron and Aluminium Sequestering Agents
FULL PAPER
According to chemical intuition, both minimum-energy
structures have an “out” conformation, which means that
the tertiary amine N atom and the free electron pair is
pointing away from the metal ion. However, interestingly,
the “in” conformation was found in the X-ray structure of
the ferric complex with a tripodal hexadentate 3-hydroxy-2-
pyridinone (FeCP130).[36] Direct comparison between the
studied and the reported model structures cannot be made
due to differences in the anchoring skeletons (nitrilotricar-
boxylic acids versus trisaminoethylamine) and in the chelat-
ing units (3-hydroxy-4-pyridinone and 3-hydroxy-2-pyridi-
none). However, in the present case, it is believed that the
“in” conformation may be favoured for the free ligand, due
to an increased number of stabilising intramolecular hydro-
gen bonds, whereas these interactions have to be disrupted
to pre-organise the ligand geometry for the metal wrapping
and coordination by the chelating arms. All of the amide
protons point outwards from the molecule, but in the Fe–
Biodistribution studies: To evaluate the efficacy of the new
hexadentate ligands (NTAACHTUNRGTENNUG(BuHP)3 and NTPACHTUGNTERN(NUGN PrHP)3) as
chelating agents for mobilisation of hard metal ions, the
effect of these compounds on the usual biodistribution pro-
file of the 67Ga–citrate was studied through tissue distribu-
tion biokinetics in female mice, after intravenous adminis-
tration of the radiotracer. The biological distribution of this
radiotracer was compared with that involving the simultane-
ous intraperitoneal administration of 0.5 mmol of each
ligand in saline solution. The most representative tissue dis-
tribution data are shown in Figure 6. Both hexadentate li-
NTAACHTUNGTRENNUNG(BuHP)3 complex one of them is pointing inwards and
forms an intramolecular hydrogen bond with a chelating car-
bonyl O-atom from the neighbouring arm (hydrogen-bond
length NH···O 2.23 ꢆ; hydrogen-bond angle N-H-O 1708).
This type of intramolecular hydrogen bond is quite often
found in this family of complexes[37] and in the NTA deriva-
tive it can be due to the higher number of hydrogen-bond
donors and acceptors in a closer distance as well as to the
higher flexibility of the HP arm, relative to the NTP deriva-
tive. The oxygen atom participation in that hydrogen bond
should result in a change in its electronegativity, thus justify-
À
ing the lengthening of the corresponding O Fe bond
(1.999 ꢆ), compared with the average value (1.950 ꢆ) found
in these complexes, although still in the range of values re-
ported in the literature for the X-ray crystal structure of the
ferric complex with other 3-hydroxy-4-pyridinones (DFP).[37]
Since both complexes under study have the same number
and type of atoms, their stability can be compared on the
basis of differences between their total energies. The calcu-
lated minimum energies are very close (but (Fe–NTP-
ACHTUNGTRENNUNG(PrHP)3 is slightly more stable than Fe–NTAAHCUTNGTREN(NUGN BuHP)3; DE=
0.004718 hartree=2.960 kcalmolÀ1), thus suggesting analo-
gous stabilities for both complexes. Based on the above
cited calculated hydrogen-bond parameters, this hydrogen
bond should have a weak to moderate strength (2–7 kcal
molÀ1).[38,39] However, it has been determined that, in the
case of weak interactions, DFT procedures may be deficient
due to their inability to take into account the dispersion
energy.[39] Therefore, this intramolecular hydrogen bond
may not have an important contribution to the molecular
stability of the NTA complex derivative, or its stabilising
effect is overcome by other structural (metal ion wrapping)
and electrostatic contributions that determine the overall
stability of these ferric complexes. These simulations are
somehow corroborated by the solution complexation studies,
Figure 6. Biodistribution data in the most relevant organs, expressed as
%I. D. per organ for 67Ga–citrate with simultaneous intraperitoneal in-
jection of the ligand NTAACTHNURGTENN(UG BuHP)3 (top) or NTPACHTUNGTNER(NUGN PrHP)3 (bottom) at
15 min, 1 h, 4 h and 24 h, after intravenous administration in female mice
(n=3–5).
gands induce similar modifications on the radiotracer biodis-
tribution, but NTA
ACHTUNGTRENNUNG
and bone clearance than NTPAHCTNUGTRENNNUG
tantly slower hepatobiliar and whole animal body radioac-
tivity excretion. However, these differences are actually
very small, as expected from the similarity of the ligand
properties in terms of lipophilicity and metal chelating affin-
ity.
showing that the Fe–NTP
ACHTUNGTRENNUNG
more stable than Fe–NTA
AHCTUNGTRENNUNG
the amide linkage in the chelating arm does not seem to be
very dependent on the stability of these complexes.
Chem. Eur. J. 2010, 16, 10535 – 10545
ꢄ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10541

M. A. Santos et al.
These results clearly show that the co-administration of
any of these ligands induces a rapid clearance of the radio-
tracer from most tissues, especially blood, muscle and bone,
and greatly increases the overall excretion rate of radioac-
tivity from the whole animal body. Moreover, no significant
uptake in any major organ, except those related with excre-
tory routes was found. Such improvement in the elimination
of the radiotracer illustrates the clear ability of these ligands
to coordinate in vivo with gallium, thus acting as removal
agents for this metal. Therefore, this biodistribution pattern
suggests the potential application of these ligands to com-
plex and to eliminate in vivo overdoses of hard trivalent
metal ions.
major organs and especially on the excretion rate. All of the
reported hexa- and tetradentate ligands[15,27] are highly effi-
cient to eliminate radioactivity from the animal body. Never-
theless, the administration of NTA
led to a even faster clearance from main organs than Kemp-
(PrHP)3 or IDA(PrHP)2, which can be attributed to im-
ACHUTNGTNERNUG(BuHP)3 or NTPACHTUNGTRENNUNG(PrHP)3
A
ACHTUNGTRENNUNG
proved properties of the new compounds in terms of lipo-
hydrophilic balance and metal affinity. Concerning the com-
mercially available iron-chelating drug, DFP, the most signif-
icant differences (data in the histogram of Figure 7) are the
fastest clearance from main organs induced by the co-ad-
ministration of the hexadentate ligands (NTA
ACHTUNGTERNN(UNG BuHP)3 and
NTP(PrHP)3) and the overall excretion enhancement. In
AHCTUNGTRENNUNG
To confirm the high in vivo chelating efficiency of the
ligand for the radiometal, the 67Ga complexes of these li-
gands were prepared with high labelling efficiency and their
biodistribution evaluated in the same animal model 4 h after
injection (data not shown). Comparison between the biodis-
tribution profiles of the complexes and those of the radio-
tracer (67Ga–citrate) followed by injection of the related li-
gands do not present significant differences in the uptake
and clearance from the main organs as well as on the excre-
tion rate. These observations indicate that the biodistribu-
tion of the tracer with co-administration of the ligand
should correspond to the in vivo formation of the 67Ga com-
plex.
fact, the overall excretion induced by the new ligands is
quite rapid, as indicated by the 67Ga I.D values at 15 min,
1 h, 4 h and 24 h of 36.5, 77.2, 92.2 and 97.0%, respectively,
in the presence of NTP
ACHTUNGTRENNUNG
96.0%, respectively, in the presence of NTAACHTUNGTRENNUNG
In summary, the favourable biodistribution and excretion
profiles associated with the administration of these new che-
lators, especially the biological profile of the radiotracer
with co-administration of the hexadentade ligands, indicates
the in vivo formation of a complex with fast tissue clearance
and rapid excretion. These are essential requirements to get
a high ratio between target/not target organs and highlight
the interest to synthesise new radioactive complexes with li-
gands containing a similar backbone for specific targeting.
Comparison between 67Ga-mobilisation ability of the new
ligands (NTA
based compounds (e.g., hexadentate (Kemp
dentate (IDA(PrHP)2) and bidentate (DFP) ligands) can be
ACHTUNGTRENNUNG(BuHP)3, NTPACHTGNUTRENNUNG
AHCTUNGTRENNUNG
N
Conclusion
made by analysis of the data shown in Figure 7, which illus-
trates the effect on biodistribution of 67Ga in mice, when the
radiotracer is co-administrated with the chelators.
Two novel tripodal hexadentate 3,4-HP-based chelators
have been designed, prepared and then studied in solution
and in vivo for their properties as potential iron or alumini-
um sequestering agents for metal chelating therapy.
Although each of the five ligands can modify the 67Ga
biological distribution pattern, enhancing its total excretion,
significant differences were observed on the clearance from
Both compounds have aminotricarboxylic anchoring back-
bones attached to aminoalkyl bidentate 3,4-HP chelating
moieties, but differing in the positioning of the peptidic
bonds connecting the tripodal carboxylate (NTA or NTP)
and HP-chelating segments. Although both isomeric ligands
have the same arm size, considerable differences result from
the hydrogen-bonding networks associated with the apical
amine, namely, the five- and six-membered ring hydrogen-
bonded intermediates associated with the NTA and NTP de-
rivatives, respectively. These differences were mostly reflect-
ed in the acidity of the corresponding ammonium protons.
Both compounds revealed higher chelating capacity than
other reported hydroxypyridinone-based ligands, although
NTP
ACHTUNGTERNUN(NG PrHP)3 is a slightly stronger chelator than NTA-
AHCTUNGTRENNUNG
the calculated minimum-energy structures of the iron com-
plexes and respective total energies. Concerning the in vivo
metal mobilisation, both evidenced excellent capacities, es-
pecially in comparison with other reported polydentate 3,4-
HP chelators. Furthermore, the favourable biodistribution
profiles and high excretion rates, resulting from the adminis-
Figure 7. Biodistribution data in the most relevant organs, expressed as
%I. D. per organ for 67Ga–citrate and 67Ga–citrate with simultaneous in-
traperitoneal injection of the ligands NTP
(PrHP)3, IDA(PrHP)2 or DFP, 1 and 24 h after intravenous administra-
tion in female mice (n=3–5).
ACHUTGTNERN(GUN PrHP)3, NTACAHTUNGTRNE(NGUN BuHP)3, Kemp-
A
ACHTUNGTRENNUNG
tration of NTAACHTUNTRGENNUG(BuHP)3 or NTPACHTUGNTREN(NUNG PrHP)3, support expecta-
10542
www.chemeurj.org
ꢄ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 10535 – 10545

Iron and Aluminium Sequestering Agents
FULL PAPER
J=7.2 Hz, 3H; 6-PyH), 7.36 (m, 15H; PhH), 6.75 (d, J=7.2 Hz, 3H; 5-
PyH), 5.02 (s, 6H; CH2Ph), 4.01 (t, J=7.2 Hz, 6H; CH2Py), 3.34 (s, 6H;
tions on the potential clinical application of these novel li-
gands as useful metal removal agents.
NCH2CONH), 3.14 (t, J=6.6 Hz, 6H; CONHCH
CH3), 1.58 (q, J=7.5 Hz, 6H; CONH(CH2)2CH2CH2Py), 1.40 ppm (q, J=
7.4 Hz, 6H; CONHCH2CH2A
(CH2)2Py); m/z (FAB): 997 [M+H]+.
2,2’,2’’-Nitrilotris(N-{4-[3-hydroxy-2-methyl-4-oxopyridin-1(4H)-yl]butyl}-
acetamide) (NTA(BuHP)3): 10% Pd/C (0.090 g) was added to a solution
2ACHTUNGTREN(UNGN CH2)3Py), 2.10 (s, 9H;
AHCTUNGTRENNUNG
CTHUNGTRENNUNG
Experimental Section
AHCTUNGTRENNUNG
of 2 (0.18 g, 0.18 mmol) in methanol (10 mL) and the mixture was stirred
under H2 (1.5 bar) for 3 h. After filtration and removal of the solvent, re-
crystallisation from dry methanol/ethyl ether afforded the pure com-
pound as a pale yellow solid (0.12 g, 89%). M.p. 159–1608C; 1H NMR
(D2O): d=7.94 (d, J=7.2 Hz, 3H; 6-PyH), 7.00 (d, J=6.9 Hz, 3H; 5-
PyH), 4.23 (t, J=7.6 Hz, 6H; CH2Py), 4.04 (s, 6H; NCH2CONH), 3.17 (t,
General remarks: All the chemicals were of analytical reagent grade and
used as supplied without further purification. Whenever necessary, the or-
ganic solvents were dried according to standard methods.[40] In the com-
plexation studies, the solutions of FeCl3 (0.0177m) and AlCl3 (0.0393m)
from Merck were standardised by atomic absorption. These solutions
were prepared in excess acid to prevent hydrolysis and their exact con-
centration in HCl was determined by titration with HCl 0.1m (Titrisol)
for values of pHꢁ2.
J=6.9 Hz, 6H; CONHCH
7.2 Hz, 6H; CONH(CH2)2CH2CH2Py), 1.48 ppm (q, J=7.3 Hz, 6H;
CONHCH2CH2A
(CH2)2Py); m/z (FAB): 726 [M+H]+; elemental analysis
2ACHTGNURTEN(NUNG CH2)3Py), 2.45 (s, 9H; CH3), 1.72 (q, J=
AHCTUNGTRENNUNG
1H NMR spectra were recorded on a Varian Unity 300 MHz or a Bruker
Advance II 300 spectrometer. Chemical shifts are reported in ppm (d)
from internal references TMS (tetramethylsilane) for organic solvents
and DSS ([D4]3-trimethylsilylpropionic acid sodium salt) for D2O. The ti-
trant (0.1m KOH) was prepared from a carbonate-free commercial con-
centrate (Titrisol) and standardised by titration with a solution of potassi-
um hydrogen phthalate. It was discarded whenever the percentage of car-
bonate (Granꢇs method)[41] was higher than 0.5% of the total amount of
base.
CTHUNGTRENNUNG
calcd (%) for C36H51N7O9·0.8 MeOH: C 58.82, H 7.27, N 13.05; found: C
58.90, H 7.01, N 12.81.
3,3’,3’’-Nitrilotris(N-{3-[3-benzyloxy-2-methyl-4-oxopyridin-1(4H)-yl]pro-
pyl}propanamide) (3): A procedure similar to that used for the prepara-
tion of 2 was followed, starting from NTP and 1-(3’-aminopropyl)-3-ben-
zyloxy-2-methylpyridin-4(1H)-one dihydrochloride.[27] After performing
the extractions and evaporating the organic solvent, flash chromatogra-
phy was performed on silica gel using 2:1:0.02 acetonitrile/water/concd
ammonia as the eluent (Rf =0.27). After evaporation of the solvent, the
inorganic matter was removed by the addition of (2:1) dry acetonitrile/
methanol and filtration of the mixture. Evaporation of the solvent afford-
ed the pure product as a pale hygroscopic solid (0.39 g, 39%). 1H NMR
(D2O, pD ca. 3): d=8.14 (d, J=7.1 Hz, 3H; 6-PyH), 7.37 (s, 15H; PhH),
7.15 (d, J=7.1 Hz, 3H; 5-PyH), 5.07 (s, 6H; CH2Ph), 4.22 (t, J=7.5 Hz,
6H; CH2Py), 3.45 (t, J=5.8 Hz, 6H; NCH2CH2CONH), 3.12 (t, J=
6.7 Hz, 6H; CONHCH2CH2CH2Py), 2.83 (t, J=5.7 Hz, 6H;
NCH2CH2CONH), 2.30 (s, 9H; CH3), 1.88 ppm (q, J=7.4 Hz, 6H; CON-
HCH2CH2CH2Py); m/z (FAB): 997 [M+H]+.
Elemental analyses were performed on a Fisons EA1108 CHNF/O instru-
ment. Electronic spectra were recorded with a Perkin–Elmer Lambda 9
spectrophotometer, using 1 cm path length cells that were thermostated
at (25.0Æ0.1)8C by a Grant W6 equipment.
1-(4-Aminobutyl)-3-benzyloxy-2-methylpyridin-4(1H)-one (1): A solution
of 2m NaOH (2 mL) was added to a solution of 1,4-diaminobutane
(3 mL, 30 mmol) in 1:1 ethanol/water (50 mL) and heated to 508C; then
a solution of 3-benzyloxy-2-methyl-4-pyrone, prepared as previously de-
scribed[27] (6.0 g, 28 mmol) in ethanol (25 mL), was added dropwise, and
the mixture was stirred overnight at 608C. The solvent was removed and
the residue was dried under vacuum. The crude product was dissolved in
water (30 mL), the solution was acidified to pH 1 with concentrated HCl,
and then it was washed with dichloromethane (3ꢅ30 mL); afterwards the
aqueous solution was neutralised with 5% NaOH and then further ex-
tracted with dichloromethane (3ꢅ30 mL). The aqueous phase was basi-
fied to pH 12 (with 5% NaOH) and then was extracted with dichlorome-
thane (5ꢅ30 mL). The organic phases were dried over anhydrous Na2SO4
and the solvent was removed under vacuum. An HCl-saturated metha-
nolic solution was added to acidify the residue until obtaining pH 1 and
the solvent was then removed. The pure dihydrochloric salt of 1 was ob-
tained after recrystallisation from dry methanol/acetonitrile, as a yellow-
ish solid (4.1 g, 48%). M.p. 115–1178C; 1H NMR (D2O): d=7.75 (d, J=
7.2 Hz, 1H; 6-PyH), 7.41 (s, 5H; PhH), 6.61 (d, J=7.2 Hz, 1H; 5-PyH),
5.05 (s, 2H; CH2Ph), 4.03 (t, J=7.4 Hz, 2H; CH2Py), 2.98 (t, J=7.5 Hz,
3,3’,3’’-Nitrilotris(N-{3-[3-hydroxy-2-methyl-4-oxopyridin-1(4H)-yl]pro-
pyl}propanamide) (NTP
ACHTUNGTRENNUNG
formed as described for NTAAHCTNUGTRENNNUG
was filtered and the solvent was removed under vacuum, then the residue
was washed with a dichloromethane/acetonitrile mixture. Final recrystal-
lisation from dry methanol/ethyl ether afforded the pure product as a
pale hygroscopic solid (0.14 g, 88%). 1H NMR (D2O, pD ca. 3): d=8.02
(d, J=7.2 Hz, 3H; 6-PyH), 7.08 (d, J=6.9 Hz, 3H; 5-PyH), 4.34 (t, J=
7.5 Hz, 6H; CH2Py), 3.47 (t, J=5.8 Hz, 6H; NCH2CH2CONH), 3.23 (t,
J=6.6 Hz, 6H; CONHCH2CH2CH2Py), 2.86 (t, J=5.7 Hz, 6H;
NCH2CH2CONH), 2.54 (s, 9H; CH3), 2.01 ppm (q, J=7.3 Hz, 6H; CON-
HCH2CH2CH2Py); m/z (FAB): 726 [M+H]+; elemental analysis calcd
(%.) for C36H51N7O9: C 59.57, H 7.08, N 13.51; found: C 59.65, H 7.42, N
13.48.
2H; NH2CH2ACHTUNGTRENNUNG(CH2)3Py), 2.10 (s, 3H; CH3), 1.71 (q, J=7.2 Hz, 2H;
(CH2)2CH2CH2Py), 1.62 ppm (q, J=6.8 Hz, 2H; CH2CH
2ACHTUNGTRNEN(UNG CH2)2Py); m/z
Potentiometric titrations
(FAB): 287 [M+H]+.
Measurements: Potentiometric titrations of the ligands and their alumini-
um complexes were performed in water at ionic strength (I) 0.1m KCl,
T=(25.0Æ0.1)8C and the equipment used was previously described.[42]
2,2’,2’’-Nitrilotris(N-{4-[3-benzyloxy-2-methyl-4-oxopyridin-1(4H)-yl]bu-
tyl}acetamide) (2): TBTU (1.1 mL, 3.3 mmol) and dry NMM (0.70 mL,
6.3 mmol) were added to a solution of NTA (0.19 g, 1.0 mmol) in dry
DMF (10 mL) at 08C and the reaction mixture was stirred for 45 min
For the system Fe3+–NTA
ACTHUNTRGNE(UGN BuHP)3 study, the calibration procedure was
analogous to that in aqueous medium, but in a water/DMSO (97:3) sol-
vent mixture. For all of the samples prepared, the total volume was
20 mL, the ligand concentration (CL) was 1.0–1.7ꢅ10À3 m and the metal
ion to ligand molar ratio was 0:1 or 1:1. Each titration was performed
twice and the value determined for the ionisation constant (pKw) was
13.85 and 13.80 for the water and 3% DMSO media, respectively.
under N2. Meanwhile, compound
1 (1.11 g, 3.1 mmol) in dry DMF
(15 mL) was neutralised with NMM (0.68 mL, 6.2 mmol) and mixture
was stirred at RT for 30 min; the first solution was added dropwise to the
free amine and it was left stirring overnight. After evaporation of the re-
action-mixture solvent, the residue was dissolved in dichloromethane
(50 mL) and extractions were performed with 5% aqueous NaOH (2ꢅ
20 mL) and brine (2ꢅ20 mL); the final organic phase was dried over an-
hydrous Na2SO4 and evaporated. Flash chromatography was performed
on silica gel, using 2:0.6:0.01 acetonitrile/water/concd ammonia as the
eluent (Rf =0.32). After evaporation of the solvent, dry acetonitrile was
added to the residue and the insoluble inorganic matter was filtered off.
Evaporation of the solvent afforded the pure product as a pale yellow
Calculation of equilibrium constants: The stepwise protonation constants,
Ki =[HiL]/
ACHTUNGTREN[NUG HiÀ1L][H], (excluding the logKD value for NTAACHTUNGTRENNNUG
the overall aluminium complex stability constants, b
Mm Hh Ll
[Al]m[H]h[L]l, were calculated by the fitting analysis of the respective po-
tentiometric data with the HYPERQUAD 2003 program.[23] The Al3+ hy-
drolytic species[43] were included in the equilibrium model and the species
distribution curves were plotted with the HYSS program.[23]
1
solid (0.61 g, 61%). M.p. 90–918C; H NMR (D2O, pD, ca. 3): d=7.82 (d,
Chem. Eur. J. 2010, 16, 10535 – 10545
ꢄ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10543

M. A. Santos et al.
Spectroscopic titrations
(Aloka, Curiemeter IGC-3, Aloka, Tokyo, Japan). The difference be-
tween the radioactivity in the injected and sacrificed animal was assumed
to be due to whole body excretion. Tissue samples of main organs were
then removed for counting in a gamma counter (Berthold LB2111, Bert-
hold Technologies, Germany). Biodistribution results were expressed as
percent of injected dose per total organ (%ID/organ) and presented as
mean values Æstandard deviation. For blood, bone and muscle, total ac-
tivity was calculated by assuming, as previously reported, that these
organs constitute 7, 10 and 40% of the total weight, respectively.
Measurements: 1H NMR titration of a solution of NTA
ACTHUNTRGNE(UGN BuHP)3 (7.3ꢅ
10À4 m) in D2O was performed as previously described.[42,44]
The electronic spectra of the FeIII complexes with both ligands were re-
corded in the range 300–750 nm and 1:1 stoichiometry (CL =1.8ꢅ10À4
m
for NTAACHTUNGTRENNUNG ACHTUGNTREN(NUGN PrHP)3). Solutions of com-
(BuHP)3 and 1.6ꢅ10À4 m for NTP
plexes with pHꢁ2 were prepared as indicated for the potentiometric
measurements. Both Fe3+–L systems were studied for pH below 2 (0.6–2,
for L=NTAACHTUNGTRENNUNG(BuHP)3); 0.8–2 for L=NTPACHTUTGNREN(NGUN PrHP)3) using a batch titration
Animal experiments were carried out by a researcher holding certifica-
tion by the National Authority in accordance with the EU recommenda-
tions on the use of living animal in scientific investigation and followed
the principles of laboratory animal care.
(7–13 points), in which the amount of acid to be added (from standard
0.1 or 1m HCl solutions) was calculated for the total volume solution
used. To improve the low water solubility of the FeL complex of NTA-
ACHTUNGTRENNUNG(BuHP)3, the iron complexation of that ligand was studied in a 3% (v/v)
DMSO/water medium.
Calculation of equilibrium constants: The logKD value was calculated by
fitting analysis of the experimental data, using the PSEQUAD pro-
gram[22] and then that value (determined in D2O) was converted into the
corresponding value in H2O by the equation pKD =0.32+1.044pKH.[45]
The overall iron complex stability constants were determined with the
PSEQUAD program,[22] by fitting of the spectrophotometric data and in-
cluding the Fe3+ hydrolytic species[43] in the equilibrium model.
Acknowledgements
The authors thank Dr. Joaquim MarÅalo from Instituto Tecnolꢂgico e
Nuclear (ITN), Sacavꢁm, for kindly performing the ESI-MS measure-
ments. The QIT/MS at ITN is part of RNEM-Rede Nacionalde Espectro-
metria de Massa, supported by the Fundażo para a CiÞncia e Tecnolo-
gia. We also thank Portuguese NMR Network (IST-UTL Center) for pro-
viding access to the NMR facilities. The authors thank the Portuguese
Fundażo para a CiÞncia e Tecnologia (FCT) (project PCDT/QUI/56985/
04 and post-doctoral grant SFRH/BPD/29874/2006 (S.M.)) for financial
support.
Partition coefficients: The octanol–water partition coefficients (logP)
were determined by the “shakeflask” method,[8,46] based on the concen-
tration ratio of the compound between 1-octanol and a Tris-buffered
aqueous phase (pH 7.4). The species concentrations were evaluated by
spectrophotometry, based on the absorbance of the benzenoid bands (p–
p*) of the compounds.
ESI-MS measurements: The electrospray mass spectra were obtained on
ESI-QIT/MS Bruker HCT (electrospray ionisation quadrupole ion trap
mass spectrometer), operated in the positive mode. The temperature of
the heated capillary was set at 2508C. The flow rate of the electrospray
solution was 2.5 mLminÀ1. Other parameters, including capillary, capillary
exit, skimmer voltage (4 kV, 150 V, 40 V), nebuliser/N2 pressure 8.0 psi.
Solutions of metal and ligand (CM/CL =1, CL =1ꢅ10À4 m) in water were
[1] R. R. Crichton, Inorganic Biochemistry of Iron Metabolism: From
Molecular Mechanisms to Clinical Consequences, Wiley, New York,
2001.
[2] T. Kiss, K. Gajda-Schrantz, P. Zatta, Neurodegenerative Diseases and
Metal Ions, Wiley, New York, 2006.
[3] M. K. Ward, T. G. Feest, H. E. Ellis, I. S. Parkinson, D. N. S. Kerr, J.
Herrington, G. L. Goode, Lancet 1978, 311, 841–845.
[4] G. Crisponi, M. Remelli, Coord. Chem. Rev. 2008, 252, 1225–1240.
[5] M. A. Santos, Coord. Chem. Rev. 2002, 228, 187–203.
[6] G. J. Kontoghiorghes, Lancet 1985, 325, 817–817.
[7] a) M. Y. Moridani, G. S. Tilbrook, H. H. Khodr, R. C. Hider, J.
Pharm. Pharmacol. 2002, 54, 349–364; b) R. C. Hider, Z. D. Liu,
Curr. Med. Chem. 2003, 10, 1051–1064.
[8] M. A. Santos, M. Gil, L. Gano, S. Chaves, J. Biol. Inorg. Chem.
2005, 10, 564–580.
[9] D. E. Green, C. L. Ferreira, R. V. Stick, B. O. Patrick, M. J. Adam,
C. Orvig, Bioconjugate Chem. 2005, 16, 1597–1609.
[10] K. H. Thompson, C. A. Barta, C. Orvig, Chem. Soc. Rev. 2006, 35,
545–556.
[11] E. Nisbet-Brown, N. F. Olivieri, P. J. Giardina, R. W. Grady, E. J.
Neufeld, R. Sechaud, A. J. Krebs-Brown, J. R. Anderson, D. Alberti,
K. C. Sizer, D. G. Nathan, Lancet 2003, 361, 1597–1602.
[12] S. Gama, E. Farkas, P. I. Dron, S. Chaves, M. A. Santos, Dalton
Trans. 2009, 6141–6150.
[13] a) M. A. Santos, Coord. Chem. Rev. 2008, 252, 1213–1224; b) S.
Chaves, P. I. Dron, F. A. Danalache, D. Sacoto, L. Gano, M. A.
Santos, J. Inorg. Biochem. 2009, 103, 1521–1529.
analysed at appropriate pH values: 2.96 (Fe3+–NTA
–NTA
(BuHP)3), 3.50 (Fe3+–NTP(PrHP)3) and 3.96 (Al3+–NTP
Molecular modelling: The structure of Fe3+–L complexes with NTA-
(BuHP)3 and NTP(PrHP)3 were fully optimised by quantum mechanical
ACHTUNGTRENNUNG
A
R
ACHTUNGTRENNUNG
A
ACHTUNGTRENNUNG
calculations based on DFT methods included in the Gaussian 03W pro-
gram software.[34] A first optimisation step was carried out using the
B3LYP chemical model with a LANL2MB basis set, a direct SCF calcula-
tion and an SCF convergence criterion to 10À5. The B3LYP chemical
model has been shown to be an accurate density functional method,[47]
and it gives as good or better geometries and energies as MP2 ab initio
methods for first-row transition-metal complexes.[48] The B3LYP model is
a combination of the Becke three-parameter hybrid functional[35] with
the Lee–Yang–Parr correlation functional (which also includes density
gradient terms).[49,50] With regards to the basis set, LANL2MB specifies
the STO-3G on first row[51] and, Los Alamos ECP plus MBS on Na-Bi.[52]
The results of these calculations were then subjected to a deeper optimi-
sation using the B3LYP model with the LANL2DZ basis set that de-
scribes the Fe atom through the Los Alamos ECP and an essentially
double-zeta basis set including 3d orbitals and 3d diffuse functions for
the valence shell. In this basis set, the rest of the atoms are described
through the Dunning–Huzinaga full double-z basis set.[53]
Biodistribution studies: A 67Ga–citrate injection solution was prepared
by dilution of 67Ga–citrate purchased from MDS Nordion (Ottawa,
Canada) with saline to obtain a final radioactive concentration of 5–
10 MBq/100 mL. Animal studies were carried out in groups of 3–5 female
CD1 mice (randomly bred, Charles River, from CRIFFA, Barcelona,
Spain) weighing around 25 g. Mice were intravenously (i.v.) injected with
67Ga–citrate (100 mL, 5–10 MBq) through the tail vein. In a separate
group of animals, the i.v. administration was immediately followed by in-
traperitoneal (i.p.) injection of the ligand (0.5 mmol) in saline solution
(100 mL). Animals were maintained on a normal diet ad libitum and were
sacrificed by cervical dislocation at 5 min, 30 min, 1 h and 24 h post-ad-
ministration. The administered radioactive dose and the radioactivity in
sacrificed animals were determined by counting in a dose calibrator
[14] S. Piyamongkol, T. Zhou, Z. D. Liu, H. Khodr, R. C. Hider, Tetrahe-
dron Lett. 2005, 46, 1333–1336.
[15] R. Grazina, L. Gano, J. Sebestik, M. A. Santos, J. Inorg. Biochem.
2009, 103, 262–273.
[16] M. A. Santos, R. Grazina, L. Gano, S. Gama, Port. Pat. PT102660,
2003.
[17] E. T. Clarke, A. E. Martell, Inorg. Chim. Acta 1992, 196, 185–194.
[18] R. A. Yokel, A. K. Datta, E. G. Jackson, J. Pharmacol. Exp. Ther.
1991, 257, 100–106.
[19] M. A. Santos, M. Gil, S. Marques, L. Gano, G. Cantinho, S. Chaves,
J. Inorg. Biochem. 2002, 92, 43–54.
[20] R. M. Smith, A. E. Martell, Critical Stability Constants, Vol 5,
Plenum, New York, 1982, p. 62.
10544
www.chemeurj.org
ꢄ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 10535 – 10545

Iron and Aluminium Sequestering Agents
FULL PAPER
[21] C. A. Lipinski, F. Lombardo, B. W. Dominy, P. J. Feeney, Adv. Drug
Delivery Rev. 1997, 23, 3–25.
[22] PSEQUAD version 5.01, L. Zꢁkꢈny, I. Nagypꢈl, G. Peintler, 2001.
[23] P. Gans, A. Sabatini, A. Vacca, Talanta 1996, 43, 1739–1753.
[24] I. Turcot, A. Stintzi, J. D. Xu, K. N. Raymond, J. Biol. Inorg. Chem.
2000, 5, 634–641.
Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen,
M. W. Wong, C. Gonzalez, J. A. Pople, Gaussian, Inc., Wallingford
CT, 2004.
[35] A. Becke, Phys. Rev. A 1988, 38, 3098–3100.
[36] G. Xiao, D. Helm, R. C. Hider, P. S. Dobbin, Inorg. Chem. 1995, 34,
1268–1270.
[25] R. J. Abergel, K. N. Raymond, Inorg. Chem. 2006, 45, 3622–3631.
[26] T. Zhou, X. L. Kong, Z. D. Liu, D. Y. Liu, R. C. Hider, Biomacro-
molecules 2008, 9, 1372–1375.
[27] M. A. Santos, S. Gama, L. Gano, G. Cantinho, E. Farkas, Dalton
Trans. 2004, 3772–3781.
[28] A. E. Martell, R. M. Smith, Critical Stability Constants, Vol. 1,
Plenum, New York, 1974, p. 112; A. E. Martell, R. M. Smith, Critical
Stability Constants, Vol. 6, Plenum, New York, 1989, p. 55.
[29] A. E. Martell, R. M. Smith, R. J. Motekaitis, Critically Selected Sta-
bility Constants of Metal Complexes Database, Plenum, New York,
1997.
[30] R. J. Motekaitis, A. E. Martell, Inorg. Chim. Acta 1991, 183, 71–80.
[31] Z. D. Liu, R. C. Hider, Coord. Chem. Rev. 2002, 232, 151–171.
[32] W. R. Harris, J. Sheldon, Inorg. Chem. 1990, 29, 119–124.
[33] A. C. MendonÅa, S. Chaves, S. Marques, M. I. M. Prata, A. C.
Santos, C. F. G. C. Geraldes, M. A. Santos, J. Inorg. Biochem. in
press.
[34] Gaussian 03, Revision C.02, M. J. Frisch, G. W. Trucks, H. B. Schle-
gel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgom-
ery, Jr., T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S.
Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani,
N. Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K.
Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda,
O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian,
J. B. Cross, V. Bakken, C. Adamo, J Jaramillo, R. Gomperts, R. E.
Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W.
Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J.
Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C.
Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari,
J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cio-
slowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaro-
mi, R. L Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A.
[37] J. Charalambous, A. Dodd, M. McPartlin, S. O. C. Matondo, N. D.
Pathirana, H. R. Powell, Polyhedron 1988, 7, 2235–2237.
[38] J. Hermans, H. J. C. Berendsen, W. F. van Gunsteren, J. P. M.
Postma, Biopolymers 1984, 23, 1513–1518.
[39] H. Szatyłowicz, J. Phys. Org. Chem. 2008, 21, 897–914.
[40] Purification of Laboratory Chemicals, 4th ed. (Eds:. W. L. F. Armar-
ego, D. D. Perrin), Butterworth-Heinemann, Oxford, 1999.
[41] F. J. C. Rossotti, H. Rossotti, J. Chem. Educ. 1965, 42, 375–378.
[42] S. Chaves, M. Gil, S. Marques, L. Gano, M. A. Santos, J. Inorg. Bio-
chem. 2003, 97, 161–172.
[43] a) The Hydrolysis of Cations (Eds.: C. F. Baes, R. E. Mesmer),
Wiley, New York, 1976; b) L. O. ꢉhman, W. Forschling, Acta Chem.
Scand. Ser. A 1981, 35, 795–802.
[44] A. K. Covington, M. Paabo, R. A. Robinson, R. G. Bates, Anal.
Chem. 1968, 40, 700–706.
[45] R. Delgado, J. J. R. Frafflsto da Silva, M. T. S. Amorim, M. F. Cabral,
S. Chaves, J. Costa, Anal. Chim. Acta 1991, 245, 271–282.
[46] A. Leo, C. Hansch, D. Elkins, Chem. Rev. 1971, 71, 525–616.
[47] C. W. Bauschlicher, Chem. Phys. Lett. 1995, 246, 40–44.
[48] A. Ricca, C. W. Bauschlicher, Theor. Chim. Acta 1995, 92, 123–131.
[49] C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785–789.
[50] B. Miehlich, A. Savin, H. Stoll, H. Preuss, Chem. Phys. Lett. 1989,
157, 200–206.
[51] J. B. Collins, P. v. R. Schleyer, J. S. Binkley, J. A. Pople, J. Chem.
Phys. 1976, 64, 5142–5151.
[52] P. J. Hay, W. R. Wadt, J. Chem. Phys. 1985, 82, 299–310.
[53] T. H. Dunning, Jr., P. J. Hay in Modern Theoretical Chemistry, Vol. 3
(Ed.: H. F. Schaefer III), Plenum, New York 1976, pp. 1–28.
Received: May 17, 2010
Published online: July 21, 2010
Chem. Eur. J. 2010, 16, 10535 – 10545
ꢄ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
10545