A novel tripodal tris-hydroxypyrimidinone sequestering agent for trivalent hard metal ions: Synthesis, complexation and in vivo studies

Article abstract of DOI:10.1039/c2dt32361c

A new tripodal hexadentate ligand, NTP(PrHPM)₃, having three hydroxypyrimidinone (HPM) chelating units attached to a nitrilotripropionic acid (NTP) has been prepared and studied in terms of thermodynamic stability of the complexes with iron, aluminium and gallium and it has been subsequently in vivo assayed for its capacity to remove hard metal ions from an animal model overloaded with ⁶⁷Ga. The anchoring of the HPM units to the NTP scaffold revealed to be an interesting alternative to the reported hexadentate tris(3-hydroxy-4-pyridinone) analogue, NTP(PrHP)₃, because the new tris-HPM ligand still keeps high chelating capacity for hard metal ions and presents better water-solubility (log P = -1.51). The in vivo studies show that NTP(PrHPM)₃ induces a faster clearance from main organs and an enhancement of overall excretion, as compared with the commercial drug, DFP, or the bidentate HPM compound (HOPY-PrN), albeit slightly lower than the tris-hydroxypyridinone analogue, NTP(PrHP)₃. The solution and in vivo results herein presented encourage further studies envisaging the potential clinical applications of hexadentate HPM derivatives as metal sequestering agents.

Full text of DOI:10.1039/c2dt32361c

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
A novel tripodal tris-hydroxypyrimidinone
sequestering agent for trivalent hard metal ions:
synthesis, complexation and in vivo studies†
Cite this: Dalton Trans., 2013, 42, 6033
Sílvia Chaves,a Anabela Capelo,b Laurinda Areias,b Sérgio M. Marques,a
Lurdes Gano,c M. Alexandra Estevesb and M. Amélia Santos*a
A new tripodal hexadentate ligand, NTP(PrHPM)3, having three hydroxypyrimidinone (HPM) chelating
units attached to a nitrilotripropionic acid (NTP) has been prepared and studied in terms of thermodyn-
amic stability of the complexes with iron, aluminium and gallium and it has been subsequently in vivo
assayed for its capacity to remove hard metal ions from an animal model overloaded with 67Ga. The
anchoring of the HPM units to the NTP scaffold revealed to be an interesting alternative to the reported
hexadentate tris(3-hydroxy-4-pyridinone) analogue, NTP(PrHP)3, because the new tris-HPM ligand still
keeps high chelating capacity for hard metal ions and presents better water-solubility (log P = −1.51).
The in vivo studies show that NTP(PrHPM)3 induces a faster clearance from main organs and an enhance-
ment of overall excretion, as compared with the commercial drug, DFP, or the bidentate HPM compound
(HOPY-PrN), albeit slightly lower than the tris-hydroxypyridinone analogue, NTP(PrHP)3. The solution and
in vivo results herein presented encourage further studies envisaging the potential clinical applications of
hexadentate HPM derivatives as metal sequestering agents.
Received 5th October 2012,
Accepted 15th November 2012
DOI: 10.1039/c2dt32361c
www.rsc.org/dalton
ions.7 However, the physicochemical analogies between Fe3+
Introduction
and other Lewis acid metal ions (e.g. Al3+, Ga3+) have orien-
Despite the biological importance of iron for all living cell, tated the potential use of many iron-chelators as Al-sequester-
when in excess, it may become very toxic to the human body. ing drugs or as radionuclide Ga-chelates for diagnostic drug.8,9
In fact, humans have very little capacity for excretion of excess
The main Al-binding protein in plasma is transferrin (Tf)
iron, which can result from its administration, as in beta- which is only 30% saturated with Fe in normal serum, thus
thalassemia patients, or from excess absorption, as in hemochro- still keeping significant chelating capacity for other trivalent
matosis.1,2 Also, the nonessential ubiquitous aluminum (Al3+), metal ions, such as Al.10 Thus, in principle, Fe/Al-chelators
through different processes of exposure, can accumulate in must be able to compete with transferrin for those metal ions.
specific organs leading to serious problems of toxicity,3,4
The first iron-chelating drug (since ca. 5 decades ago) for
namely bone disorders (vitamin-D-resistant osteomalacia) and beta-thalassemia patients was the hexadentate tris-hydroxa-
neurological diseases, such as dialysis encephalopathy syn- mate Desferrioxamine or Desferal (DFO).11 However, its high
drome, and eventually Alzheimer’s disease (AD).5,6
water-lability and oral inactivity led to the disclosure of other
The therapeutic strategies to minimize the adverse effects orally active chelators, namely the bidentate 3-hydroxy-4-pyridi-
of metal overload and misplaced related illnesses have led to none (3,4-HP) Deferriprone (DFP) (in 1982)12 and the triden-
the development of new chelating agents, which, by far, has tate Deferasirox or Exjade (ICL670), (approved by FDA in
been more focused on iron(III) than any other hard metal 2005),13 although both also with recognized drawbacks. Con-
cerning the aluminum chelators, DFO was also the first chela-
tor introduced in clinical therapy, namely for treatment of
osteomalacia and encephalopathy associated with severe dialy-
sis Al-intoxication,14 but its drawbacks led likewise to the
search for alternatives.8
The need of stronger and orally active chelators for the
decorporation of iron or other specific toxic hard metal ions
such as aluminum or actinides has been worldwide recog-
nized, and so, along the last 2 decades, we have assisted to an
aCentro de Química Estrutural, Instituto Superior Técnico, Universidade Técnica de
Lisboa, Av. Rovisco Pais 1, 1049-001 Lisboa, Portugal. E-mail: masantos@ist.utl.pt;
Fax: +351-21-8464455; Tel: +351-21-8419273
bLNEG, Unidade de Pilhas de Combustível e Hidrogénio, Estr. Paço do Lumiar,
1649-038 Lisboa, Portugal
cIST/ITN, Instituto Superior Técnico, Universidade Técnica de Lisboa, Estrada
Nacional No 10, 2686-953 Sacavém, Portugal
†Electronic supplementary information (ESI) available. See DOI: 10.1039/
c2dt32361c
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 6033–6045 | 6033

View Article Online
Paper
Dalton Transactions
intensive search for new metal sequestering agents, namely
hydroxypyridinone(HP)-based compounds with mono- or poly-
chelating capacity,9,15,16 in view of single or combined drug
administration protocols.17
As a step forward on the discovery of new strong iron chela-
tors, we have decided to explore new ligands with hexadentate
capacity, to guarantee the metal-full coordination, absence of
ligand dilution effects and so iron-scavenging capacity at very
low concentration. Thus, by paralleling a design strategy
recently used in our group for tripodal hexadentate ligands,18
we have substituted the chelating units, 3,4-HP by hydroxypyri-
midinones (HPM), a kind of heterocyclic hydroxamate ana-
logue with high hydrolytic stability, adequate hydrophilicity
and chelating capacity, lying between those of the hydroxamic
acids and the hydroxypyridinones.19
In particular, we describe herein the study of a novel tri-
podal tris-chelator, NTP(PrHPM)3, having three HPM units
appended to nitrilotripropionic acid as the anchoring skeleton
(Scheme 1). After the synthetic methodology, we present the
results of assessing its capacity for complexation with Fe3+
,
Al3+ and Ga3+ in water, as well as the molecular modeling cal-
culations (DFT) on the iron-complex to get some insight into
its structure. The results from the study of the capacity of this
new chelator for metal mobilization in mice pre-administrated
with a radiotracer (67Ga), as an animal model of metal-over-
load, are also analyzed. The discussion of all set of results is
made in comparison with the reported properties for other
synthetic chelators and bioligands.
Scheme 1 Structural formulae for (a) tris-HPM and tris-HP hexadentate com-
pounds and (b) HPM and HP bidentate compounds.
deprotection using a standard hydrogenolysis with 10% Pd/C
as a catalyst.
Solution equilibrium studies
The physico-chemical characterization of the tris-HPM ligand,
NTP(PrHPM)3, was carried out in aqueous solution, namely
the study of the acid–base properties and the lipo-hydrophilic
character, as well as its complexation ability towards hard
metal ions (Fe3+, Ga3+ and Al3+). Discussion of the results is
made on the basis of comparison with previously reported tri-
podal hexadentate tris-HP compounds (NTP(PrHP)3 and NTA-
(BuHP)3)19,22 and also with bidentate compounds, such as the
HPM derivative (HOPY-PrN)20 or the commercial drug DFP
(see Scheme 1).
Results and discussion
Synthesis of the ligand, NTP(PrHPM)3
The hexadentate compound (NTP(PrHPM)3) was obtained
through the attachment of three amino-bearing O-benzyl-
hydroxypyrimidinone units (O-benzyl HOPY-PrN) to the
tris-carboxylic scaffold (nitrilotripropionic acid, NTP) via the
formation of amide linkages and final O-benzyl deprotection
through standard methodologies, as described below.
Acid–base behavior
The general procedure for the synthesis of NTP(PrHPM)3
(3,3′,3′′-nitrilotris(N-(3-(1-hydroxy-2-oxo-1,2-dihydropyrimidin-4- The acid–base properties of NTP(PrHPM)3 were mainly studied
ylamino)propyl)propanamide)) is outlined in Scheme 2. by fitting analysis of the potentiometric titration data, which
In the first step, 1-(benzyloxy)-4-(1′,2′,4′-triazol-1′-yl)-2-(1H)- provided the protonation constants; a 1H NMR spectroscopic
pyrimidinone19 1 was reacted with 3-((tert-butoxycarbonyl)- titration was used to check the protonation sequence; a UV
amino)propylamine20 to give 4-[3-((tert-butoxycarbonyl)amino)- spectrophotometric titration of the ligand was also performed
propylamino]-1-(benzyloxy)-2(1H)-pyrimidinone 2 in 74% yield. to determine the spectral parameters of the various protonated
Removal of the Boc group involved treatment with HCl in 1,4- species to be subsequently introduced in the equilibrium
dioxane21 and provided 4-(3-aminopropylamino)-1-(benzyloxy)- model of the Ga3+/NTP(PrHPM)3 system.
2(1H)-pyrimidinone hydrochloride 3 in quantitative yield. The
The compound was isolated in its neutral form (H3L),
coupling of this amine-bearing HPM side arm to the backbone whereas, when fully protonated, it has seven dissociable
3,3′,3′′-nitrilotripropanoic acid (NTP) involved a pre-activation protons (H7L4+). The stepwise protonation constants, obtained
of this tris-carboxylic acid with TBTU, under anhydrous con- from pH-potentiometric studies for NTP(PrHPM)3, are sum-
ditions, followed by addition of compound 3, after its neutral- marized in Table 1, which also includes the corresponding
ization with N-methylmorpholine (NMM), affording the pre- reported values for some other HPM and HP analogues.
final product 4 (37% yield). The final ligand (NTP(PrHPM)3)
Analysis of calculated stepwise protonation constants
was obtained in almost quantitative yield (91%), by O-benzyl (log Ki) for NTP(PrHPM)3 evidences the existence of a first set
6034 | Dalton Trans., 2013, 42, 6033–6045
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
Scheme 2 Reagents and conditions: (i) H2N(CH2)3NHBoc, dry THF, reflux, 16 h; (ii) 4 M HCl in 1,4-dioxane, 0 °C, 4 h; (iii) TBTU, NMM, dry DMF, r.t., 1 h; (iv) 3, NMM,
dry DMF, r.t., 6 h; (v) H2, Pd/C, MeOH, 2 h, r.t.
Table 1 Stepwise protonation constants (log Ki) and partition coefficient (log P) of NTP(PrHPM)3, as well as global formation constants of the M3+ complexes. Data
for other tripodal compounds, HOPY-PrN and DFP are also presented for comparison (T = 25.0 0.1 °C, I = 0.1 M KCl)
Ligand
log Ki
MmHhLl (m,h,l)
log β (FemHhLl)
log β (GamHhLl)
log β (AlmHhLl)
log P
8.25(2)
7.22(3)
6.78(2)
6.02(3)
3.29(6)
2.61(9)
2.0(1)
(1,5,1)
(1,3,1)
(1,1,1)
(1,0,1)
38.93(4)
36.32(4)
29.36(3)
—
36.61(3)
34.20(6)
28.12(7)
23.50(9)
—
−1.51
31.26(5)
24.57(5)
17.81(8)
pM*
—
23.3
17.7
9.946(9)
9.84(1)
9.091(8)
6.77(1)
3.81(1)
3.14(1)
2.76(2)
(1,5,1)
(1,3,1)
(1,1,1)
(1,0,1)
47.62(5)
45.29(5)
40.56(3)
35.21(1)
46.70(1)b
44.00(4)b
38.79(2)b
33.34(3)b
44.69(6)
40.01(5)
34.72(4)
28.13(3)
−1.24
pM*
29.4
27.5b
22.4
9.98(2)
9.83(3)
8.94(4)
3.88(4)
3.11(5)
2.35(5)
1.4(1)
10.11(2)e
6.84(4)e
2.21(5)e
(1,5,1)
(1,4,1)
(1,2,1)
(1,0,1)
42.30(2)c
40.81(5)c
38.53(3)c
33.50(1)c
—
−1.40
−1.43
—
—
37.45(4)
33.55(1)
27.65(6)
pM*
27.9d
22.0
(111)
(122)
(133)
(123)
pM*
19.88(4)e
38.12(6)e
55.26(5)e
48.04(3)e
16.1
18.74(6)
36.11(4)
52.42(3)
45.63(4)
14.9
16.84(2)e
34.04(4)e
49.67(2)e
13.1
9.77
3.62
pM*
19.3f
19.2g
16.0f
−0.85
−1.03h
*pM values at pH = 7.4 (CM = 10−6 M, CL/CM = 10). a Ref. 18. b Ref. 22. c Values determined in 3% DMSO. d Admitting that there is no precipitation
under the concentration conditions of pM determination. e Ref. 19. f Ref. 23. g Ref. 24. h Ref. 25.
of four values in the range 6.0–8.2, corresponding to three 2.0–3.3, being attributed to the 4-imine groups of the side
hydroxyl groups of the HPM moieties plus the anchoring back- chains. These results are in accordance with the protonation
bone amine group; the second set of values is in the range constants hitherto obtained for the hydroxyl (log K2 = 6.84)
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 6033–6045 | 6035

View Article Online
Paper
Dalton Transactions
Fig. 1 1H NMR titration curves of NTP(PrHPM)3 (CL = 5 × 10−4 M).
and 4-imine (log K3 = 2.21) groups of HOPY-PrN. The protona-
tion constants calculated for the HPM hydroxyl groups are
lower than those obtained for the corresponding HP analogues
(9.1–9.9 for the tris-HP, 9.77 for DFP), which can be due to the
electron withdrawing and resonance effects of the aromatic
ring, as well as to the electron-withdrawing effect of a second
ring-nitrogen, contributing to the stabilization of the negative
charge of the conjugate base. Moreover, the very acidic charac-
ter of the 4-imine groups, requiring the addition of acid excess
to guarantee the ligand full protonation, must be due to the
existence of enamine/imine equilibrium together with the
above referred electron-withdrawing and resonance effects of
the aromatic ring.
Fig. 2 (a) Electronic spectra (pH 0.8–11.5) and (b) species distribution curves
with molar extinction coefficients at maximum absorption wavelengths for NTP
(PrHPM)3 (CL = 5 × 10−5 M).
1
Though the acid–base properties of NTP(PrHPM)3 were
mainly assessed by potentiometry and 1H NMR (see above), we
have also included herein the spectrophotometric titration of
this ligand (see Fig. 2a), a requirement for the subsequent UV
spectrophotometric study of the gallium complex formation
(see below), as well as a combination of spectral data with the
species distribution curves (Fig. 2b).
Analysis of Fig. 2b evidences that in the range 3.2 < pH < 6,
the major protonated species is H4L+; at the physiological pH,
H3L (8%), H2L− (33%), HL2− (51%) and L3− (7%) coexist, while
for pH above 8.3 the ligand is essentially deprotonated (L3−).
This figure also shows that the species H5L2+, H4L+, H3L, H2L−
and HL2− absorb at 251 nm, while H7L4+ absorbs at 290 nm;
the absorptivity values at 318 nm reveal that the main species
absorbing at this wavelength are HL2− and L3−. Therefore,
290 nm and 318 nm seem to correspond to the absorp-
tion maximum wavelengths for the chromophores of the
completely protonated (H7L4+) and deprotonated forms
(HL2− and L3−) of the hydroxypyrimidinone, respectively.
At 251 nm, it is evidenced an equilibrium between species
with different partially protonated hydroxyl/imine moieties
present.
The set of H NMR titration curves (Fig. 1) presents down-
field shifts of the non-labile protons, namely in the ranges pD
6–9 (protons a, b, e, f, g) and pD 2–3.5 (protons a, b, c, d),
which give some support to the attributed protonation
sequence. The protonation of the hydroxyl and the backbone
amine groups is practically simultaneous, as evidenced by a
detailed analysis of these curve profiles, namely the occurrence
of large downfield shifts on protons g and f (due to the nearby
amine group) close to the pH range of hydroxyl group protona-
tion (unfortunately, not clearly evidenced in this figure
because the curve profiles for protons a and b are quite
smooth).
Though the attribution of the four calculated macro-
constants in the higher pH range to individualized protonation
processes renders difficult, analysis of Fig. 1 seems to suggest
that log K3 may be mainly ascribed to the protonation of the
backbone amine, taking into account the isotopic correction.26
The value obtained for NTP(PrHPM)3 (6.78) would be quite
close to the corresponding value for NTP(PrHP)3 (6.77), thus
indicating that for these tripodal ligands, sharing the same
backbone, the protonation of the apical amine is quite inde-
pendent of the appended chelating arms.
6036 | Dalton Trans., 2013, 42, 6033–6045
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
An identical spectrophotometric study was also performed
for the ligand HOPY-PrN, spectra of which presented absorp-
tion maximum at 251, 285 and 318 nm (Fig. S1†).
Lipo-hydrophilic character
The lipo-hydrophilic character of NTP(PrHPM)3 and HOPY-PrN
was assessed via the corresponding partition coefficients
(log P) between 1-octanol and a TRIS-buffered aqueous solution
at the physiological pH. The value obtained for NTP(PrHPM)3
(−1.51) is quite close to others previously reported for analo-
gous compounds, namely hexadentate HP derivatives with
similar anchoring groups.19 This value is slightly higher than
those obtained for NTP(PrHP)3 (−1.24) or NTA(BuHP)3 (−1.40),
which seems according to the higher water solubility of HPM
as compared with HP derivatives. This difference is also sup-
ported by the fact that, at pH 7.4, NTP(PrHPM)3 is mostly nega-
tively charged (ca. 91%, as H2L−, HL2− and L3−) while NTP
(PrHP)3 is predominantly in the neutral form (80% H3L) and
only 20% as a monoprotonated species (H4L+).
The HOPY-PrN compound is also slightly more hydrophilic
(log P = −1.43) than the HP bidentate analogues, such as
the drug DFP (log P = −0.85) or 3,4-HP-PrN (log P = −1.25).
These differences can be partially rationalized on the basis of
distribution curves. At pH 7.4, HOPY-PrN presents as 22%
monoprotonated species (H2L+), while DFP as a 100% neutral
form (HL) and 3,4-HP-PrN as 98% H2L+ species. Of course,
besides the species charge, other factors are also determinant
for the lipo-hydrophilic character, namely the solute–solvent
interaction. In fact, at pH 7.4, DFP is in the neutral form but
presents some hydrophilic character.
Fig. 3 (a) Electronic spectra (pH 0.8–6.5) and (b) species distribution curves
with molar extinction coefficients at maximum absorption wavelengths for the
Fe3+/NTP(PrHPM)3 system (CL/CFe = 1.2, CL = 2.7 × 10−4 M).
Metal chelating capacity
The complexation ability of NTP(PrHPM)3 towards Fe3+, Ga3+
and Al3+ was studied in solution by determination of the
global stability constants of the complexes which allowed the
assessment of species distribution under different pH con- the UV-Vis absorption spectra registered at different pH values
ditions. Those studies involved potentiometric and UV/Vis for the Fe3+/NTP(PrHPM)3 system in Fig. 3a shows one isosbes-
spectrophotometric titrations with subsequent fitting analysis tic point at ca. 500 nm, corresponding to the interconversion
of the experimental data with HYPERQUAD 200827 and of the tetrachelate complex (FeH3L) to the hexachelate complex
PSEQUAD28 programs.
(FeHL). The species distribution curves associated with the
The iron complexation was studied by UV/Vis spectropho- proposed equilibrium model for the Fe3+/NTP(PrHPM)3 system
tometry because at pH = 2 two protonated complex species combined with the absorptivity at two wavelengths (466 and
(FeH5L and FeH3L) were already formed, thus precluding the 510 nm) (Fig. 3b) show that the maximum absorption at
use of potentiometry. This study was performed in two stages: 466 nm (4659 M−1 cm−1) corresponds to the ligand to metal
for pH ≤ 2, a batch titration was used, in which calculated charge-transfer (CT) band of the trischelate (FeHL) complex,
amounts of HCl and KCl were added in order to obtain the while the band at 510 nm (3801 M−1 cm−1) corresponds to the
desired pH value and to keep constant the ionic strength; for CT band of the bischelate (FeH3L) complex. The spectral para-
pH > 2, a standard automatic titration was carried out. The meters obtained for the trischelate are quite close to the
first titration stage allowed the determination of the stability reported values for the CT bands of ferric trischelated com-
constants corresponding to the species FeH5L and FeH3L; plexes with hydroxypyrimidinone analogues such as the hexa-
these values were kept constant and introduced in the equili- dentate 3HOPY5 (465 nm, 4550 M−1 cm−1 29
or the bidentate
)
brium model for pH > 2, which allowed the calculation of HOPY-PrN (465 nm, 5624 M−1 cm−1).19 Thus, these studies evi-
βFeHL. The determination of βFeL became impossible, since denced the high iron chelating capacity of this ligand, since
under our experimental conditions, precipitation occurred the trischelate (FeHL) species, formed through the {O,O} chela-
above pH 6.5, probably due to the formation of the neutral FeL tion of the adjacent N-hydroxo and keto-oxygen groups of the
species or even mixed hydroxo-ligand complexes. Analysis of three arms, is predominant at pH above 3.5.
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 6033–6045 | 6037

View Article Online
Paper
Dalton Transactions
Fig. 5 Species distribution curves for the (1 : 1) Al3+/NTP(PrHPM)3 system
(CL = 4 × 10−4 M).
Table 2 pM* (M = Fe, Ga, Al) values at the physiological pH for NTP(PrHPM)3
and some selected ligands.
Ligand
pFe
pGa
pAl
NTP(PrHPM)3
NTP(PrHP)3
—
23.3
27.5b
—
17.7
22.4a
22.0
18.8
19.0
16.2
15.7
13.2
19.3h
16.0i
14.5m
29.4a
27.9c
25.8
26.3
23.4
24.6
24.3
26.5g
19.3i
20.3k
a
NTA(BuHP)3
d
IDA(HP)2
22.9
24.9
20.2
20.9
18.8
22.4h
19.2j
20.3l
e
EDTA(HP)2
EDTAf
DTPAf
DOTAf
DFO
Fig. 4 (a) Electronic spectra (pH 0.8–8.9) and (b) species distribution curves
with molar extinction coefficients at maximum absorption wavelengths for the
Ga3+/NTP(PrHPM)3 system (CL/CGa = 1.2, CL = 5 × 10−5 M).
DFP
Transferrin
*CM = 10−6 M, CL/CM = 10 (pH = 7.4). a Ref. 18. b Ref. 22. c Admitting
that there is no precipitation for the low concentration values of pM
determination. d Ref. 30. e Ref. 31. f Ref. 32. g Ref. 33. h Ref. 34. i Ref. 23.
j Ref. 24. k Ref. 33. l Ref. 35. m Ref. 10.
Regarding the Ga3+/NTP(PrHPM)3 system, since the
gallium-complex formation also started below pH 2, we have
also performed a two-stage spectrophotometric titration under
1 : 1 metal–ligand stoichiometry, using the same methodology
cited above for the iron complexation (Fig. 4a). The batch titra- millimoles of a base and that of a ligand), which means that,
tion (pH ≤ 2) allowed the determination of the global stability as expected, the backbone amine is not involved in the metal
constants and spectral data of the species GaH5L and GaH3L, coordination. Moreover, analysis of that titration curve also
while the second stage titration (pH > 2) enabled the determi- suggests that the deprotonation of that amine group may be
nation of βGaHL and βGaL. Analysis of the speciation diagram followed by the formation of hydroxo species above pH
for this system (Fig. 4b) discloses the presence of the bische- 7. Nevertheless, once more the high chelating capacity of this
late GaH3L as a major species for pH between 1.7 and 3 and ligand for Al3+ seems evidenced in Fig. 5, since the distri-
above pH 3 the predominance of the trischelate as a mono- bution curves show that the main forms of the aluminium
protonated (GaHL) or a non-protonated (GaL) species. Fig. 4b complex above pH ca. 3.3 are tris-chelated species.
shows also that the fully protonated complex (GaH5L) absorbs
at 293 nm, while the absorbance at 250 nm is mostly due to different acid–base behaviour and denticity is usually made on
the hexacoordinated gallium complexes (GaHL and GaL).
Comparison of the metal chelating capacity of ligands with
the basis of the pM parameter (pM = −log[Mn+], CL/CM = 10,
The 1 : 3 Ga3+/HOPY-PrN system was studied by potentio- CL = 10−5 M at pH = 7.4). Unfortunately, the pFe value could
metry, keeping constant the value of βGaHL previously determined not be calculated herein due to solubility limitations above pH
by spectrophotometric titration under 1 : 1 M–L stoichiometric 6.5, under our experimental conditions. This rendered imposs-
conditions, and the obtained Ga3+ complexation model is ible the determination of βFeL which would result in a calcu-
presented in Table 1.
lated pFe value at pH 7.4 with no real meaning. In order to try
The study of Al3+ complexation was performed by potentio- to make an adequate comparison of the iron chelating ability
metry using
a 1 : 1 metal-to-ligand stoichiometric ratio of this ligand with that of other ligands contained in Table 2,
(Fig. S2†). The Al3+/NTP(PrHPM)3 titration curve shows an the pFe value was determined at pH 6.0. It was found that NTP
inflexion at a = 2 (a is the ratio between the number of (PrHPM)3 (pFe 19.7) is a stronger iron chelator than the drug
6038 | Dalton Trans., 2013, 42, 6033–6045
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
Fig. 6 Metal complexation strength, reported as pGa versus pH, for some
selected ligands.
Fig. 7 DFT-minimized structure of the complex Fe3+–NTP(PrHPM)3. The O–Fe
DFP (pFe 15.4), although it is weaker for iron than the hexa- metal coordination bonds are represented as purple lines and H-bonds as black
lines. C atoms are grey, H atoms white, N atoms blue, O atoms red and Fe atom
dentate tris-HP (pFe 23.7–24.5), the tetradentate bis-HP
(pFe 21.6–22.5) or the commercial chelators EDTA, DTPA and
orange.
DOTA. Concerning the pGa and pAl values determined for NTP
(PrHPM)3, they are considerably high, albeit lower than those
Functional Theory (DFT) to get the optimized structure for the
Fe3+–NTP(PrHPM)3 complex.
for the hexadentate HP analogues (NTP(PrHP)3 and NTA
(BuHP)3). This is according to the reported trend of the biden-
tate HPM derivatives which usually present lower pM values
than HP analogues (e.g. HOPY-PrN and DFP)16 (see Fig. 6).
Moreover, Table 2 shows that NTP(PrHPM)3 presents pM
values in the range of the previously calculated values for bis-
HP compounds (IDA(HP)2 and EDTA(HP)2), but evidencing
higher affinity than other currently used drugs for Al decor-
poration (DFP, EDTA, DTPA, DOTA) or even for diagnostic
probe containing 68Ga (DFP, DFO, EDTA, DTPA, DOTA).
Table 2 also suggests that NTP(PrHPM)3 is able to compete,
from the thermodynamic point of view, with transferrin for the
complexation with those hard metal ions. Of course that this
is a somehow limited view of the problem, since compu-
tational models predicted already significant time-dependent
non-equilibrium binding of Al by ligands in competition with
transferrin,36 which undoubtedly demonstrates the impor-
tance of both kinetic and thermodynamic control for metal
distribution in biological systems.
The DFT modeling studies were carried out using Gaussian
03W software37 in two steps with the B3LYP hybrid functional.
The first step was aimed mainly at optimizing the geometry of
the ligand, while the second one, involving a more complex
basis set (LANL2DZ), was more appropriate to treat the Fe
atom. In order to perform the structural optimization of the
Fe3+–NTP(PrHPM)3 complex, two structures were designed.
The first one, in a so-called “in” conformation, displays the
central N atom with its free orbital pointing inwards the
complex structure, and the second one, with an “out” confor-
mation, has the N atom pointing outwards. The two calcu-
lation processes converged and resulted in two different
structures, which still maintained the original “in” and “out”
conformations of the apical N atom. However, these two struc-
tures presented a difference in their global formation enthalpy
of 0.0080008 Hartree (5.02 kcal mol−1), the “out” conformation
being the most stable (see the final structure in Fig. 7). Inter-
estingly, from previous modeling (DFT calculations) of the cor-
responding Fe3+–NTP(PrHP)3 complex the “out” conformation
also appeared as slightly more stable than the “in” confor-
mation.19 This structure is slightly twisted (see Fig. 7, the right
and left arms are drawn out and flattened over the back arm),
which necessarily results in some energy penalty from this
symmetry break and structure strain. However, such energy
loss must be largely compensated by the stabilization energy
afforded with the formation of three H-bonds between the
three amide NH groups with one N atom and one carbonyl O
atom of the HPM rings (2.48 and 2.18 Å, respectively), and
with a carbonyl O atom of one arm (1.91 Å; see Fig. 7, black
solid lines). Regarding the metal ion coordination sphere, the
optimized structure maintained, as expected, the original
In general, it can be concluded that NTP(PrHPM)3 is a
better chelator for the hard metal ions studied than the drug
DFP, being also more efficient for gallium and aluminium
complexation than several currently used chelators, such as
EDTA, DTPA or DOTA.
Molecular modeling of the Fe3+ complex
It is widely accepted that one of the best techniques to disclose
the lowest energy structure of a molecule is by X-ray diffraction
of single crystal. Unfortunately, until now all our efforts to
obtain good crystals and get the 3D conformation of the ferric
complex with NTP(PrHPM)3 revealed unfruitful. Hence, we
have decided to use molecular simulations based on Density
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 6033–6045 | 6039

View Article Online
Paper
Dalton Transactions
input octahedral geometry with the ferric ion being chelated bioassays and it is herein expected to give some clue about the
through the six HPM-O atoms of the ligand, and Fe–O bond in vivo efficacy of the ligand to mobilize those metals.
lengths ranging between 1.92–1.97 Å. Concerning the final
Biodistribution studies, at specific post-injection times,
“in” geometry structure (not shown), this structure displayed were carried out by i.v. administration of the radiotracer
an even more distorted and asymmetric geometry. It displayed immediately followed by i.p. injection of the ligand solution,
two H-bonds between the NH atoms of two amide groups with as previously reported by us.23 The 67Ga tissue distribution was
one another and with one ring N atom, obviously not stabiliz- compared to its pattern without simultaneous administration
ing enough to make it the lowest energy conformation of the of any chelator (Fig. 8) and the most representative tissue dis-
Fe3+–NTP(PrHPM)3 complex.
tribution data are presented in Table 3.
Analysis of the biodistribution profiles clearly shows that
the co-administration of the ligand NTP(PrHPM)3 and the
radiotracer interferes in the usual tissue distribution of the
radioactive metal, inducing a faster clearance from main
organs and highly enhancing the overall excretion rate of
radioactivity from whole animal body. Furthermore, no signifi-
cant uptake was found in any major organ, except those
related with excretory routes. The high rate of excretion is a rele-
vant sign of the good capacity of this hexadentate ligand for
the in vivo Ga3+ chelation. Hence, this favourable in vivo behav-
iour points to the potential usefulness of this ligand as a de-
corporating agent of hard trivalent metal ions.
In vivo assays
The ability of the new hexadentate chelating agent, NTP
(PrHPM)3, for in vivo mobilisation of 67Ga was assessed in
mice as an animal model of metal-overload pathologies, owing
to the well-known similarities between Ga3+ and other hard
metal ions associated with those diseases (e.g. Fe3+ and Al3+).
In spite of the fact that the suitability of Ga as a substitute for
Al in biological systems seems to be controversial, since some
authors agree38 while others indicate different behaviours,39,40
the 67Ga radionuclide appears to be quite convenient for the
Comparison between the biodistribution data, due to the
administration of the citrate radiotracer followed by injection
of the ligand and due to the administration of the 67Ga–NTP-
(PrHPM)3 complex (previously prepared with high yield and
radiochemical purity) in the same animal model, at 4 h after
injection (data not shown) evidences a very good analogy,
especially on the uptake and clearance from main organs as
well as the excretion rate. Thus, these findings indicate a rapid
kinetics of in vivo complex formation and high affinity of the
ligand to the metal ion.
The effect of this ligand on the metal uptake and clearance
from the main organs of our animal model is illustrated in the
histogram of Fig. 8 (at 1 h and 24 h after administration), in
comparison with the corresponding effect due to the similar
administration of analogues (the hexadentate NTP(PrHP)3 as
well as the bidentates HOPY-PrN and the drug DFP).
Although the administration of both hydroxypyrimidinone-
based compounds alters the metal biodistribution, the
capacity for the 67Ga removal from tissues is obviously higher
for the hexadentate ligand NTP(PrHPM)3 than the bidentate
Fig. 8 Biodistribution data in the most relevant organs, expressed as % I.A./
organ for 67Ga-citrate (iv injection) and 67Ga-citrate with simultaneous intraperi-
toneal injection of the ligands NTP(PrHP)3, NTP(PrHPM)3, HOPY-PrN and DFP,
1 and 24 h after intravenous administration in female mice (n = 3–5).
Table 3 Biodistribution data in the most relevant organs, expressed as % I.A./organ for 67Ga-citrate and 67Ga-citrate with simultaneous intraperitoneal injection of
the ligand NTP(PrHPM)3 at 15 min, 1 h, 4 h and 24 h, after intravenous administration in female mice (n = 3–5).
I.A./organ (%)
67Ga-citrate
15 min
67Ga-citrate + NTP(PrHPM)3
Organs
1 h
3.2 1.9
1.8 0.2
9.8 1.6
0.9 0.1
11.3 3.7
17.1 7.5
17.2 3.6
24 h
2.0 0.8
2.3 1.0
4.3 0.7
1.3 0.3
7.4 0.8
11.8 3.1
35.0 7.2
15 min
1 h
4 h
0.5 0.08
0.73 0.04
1.2 0.1
0.6 0.1
0.8 0.2
4.8 1.0
90.9 0.2
24 h
Blood
Liver
Intestine
Kidney
Muscle
Bone
10.2 2.3
2.7 1.3
4.0 0.4
1.5 0.4
20.2 2.3
13.8 3.5
6.7 2.9
8.5 4.3
2.0 0.8
2.2 0.4
3.9 0.8
12.3 3.3
10.8 2.5
28.7 8.9
1.7 0.2
0.19 0.05
0.5 0.1
0.3 0.1
0.39 0.07
0.15 0.04
2.9 0.7
1.0 0.2
1.3 0.5
1.0 0.1
3.6 0.2
6.2 0.5
78.5 2.2
Excretion
93.8 1.4
6040 | Dalton Trans., 2013, 42, 6033–6045
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
analogue HOPY-PrN. In fact, the administration of HOPY-PrN references TMS (tetramethylsilane), for organic solvents, or
led to slower clearance on blood and soft-tissue, but higher DSS (3-trimethylsilyl-propionic acid-d4 sodium salt) for D2O,
bone retention and no improvement in the overall radioacti- and the coupling constants (J) in Hz. Whenever necessary, the
vity excretion, thus excluding its potential use for chelating peak attribution was aided by performing 2D correlation exper-
therapy.
iments (e.g. COSY, HSQC). Electron Spray Ionization (ESI)
A comparative analysis of the 67Ga mobilization induced by mass spectra (MS) were performed on an ESI-QIT/MS Bruker
the hexadentate ligands shows that NTP(PrHPM)3 led to HCT (electrospray ionisation quadrupole ion trap mass spec-
slightly higher bone radioactivity accumulation than the 3,4- trometer), operated in the positive mode. Microanalyses were
HP analogue. A further comparison between the biodistribu- performed on a Fisons EA1108 CHNF/O instrument.
tion data obtained upon administration of the new ligand and
In the complexation studies, the FeCl3 (0.0177 M), AlCl3
of the chelating drug (DFP) evidences the ability of the new (0.0393 M) and GaCl3 (4.16 × 10−3 M) solutions were standar-
compound for a faster metal clearance from the main organs, dized by atomic absorption (Fe and Al) and inductively
especially from blood and muscle, as well as an enhancement coupled plasma emission (Ga). The metal solutions were pre-
of the overall metal excretion.
pared in acid excess in order to avoid hydrolysis and their
In summary, the high in vivo ability of NTP(PrHPM)3 to exact HCl content was determined by titration with HCl 0.1 M
complex with Ga, promoting its rapid clearance from main (Titrisol) for values of pH ≥ 2. The titrant solution (0.1 M
organs and fast overall excretion rate, anticipates its potential KOH) was prepared from a carbonate-free commercial concen-
usefulness as a drug candidate for metal chelation therapy.
trate (Titrisol) and standardized by titration with potassium
hydrogen phthalate standard. This solution was rejected when-
ever the percentage of carbonate was higher than 0.5% of the
total amount of the base.42 Electronic spectra were recorded
on a Perkin Elmer Lambda 35 spectrophotometer, using 1 cm
path length thermostatic cells (25.0 0.1 °C).
Conclusions
A new tripodal hexadentate hydroxypyrimidinone derivative –
NTP(PrHPM)3 – has been synthesized and studied in solution
and in vivo in order to evaluate its capacity as a metal seques-
tering agent. The developed compound revealed high chelating
capacity towards trivalent hard metal ions (Fe, Al, Ga), with pM
values in the same order of magnitude of the tetradentate bis-
HP compounds and higher values than those of the clinically
used drug DFP. Concerning the capacity of in vivo metal mobi-
lization, NTP(PrHPM)3 presents a favorable biodistribution
profile, a high in vivo chelating efficiency and a faster clearance
from main organs, especially from blood and muscle, as well
as an enhancement of overall excretion, when compared with
the drug DFP. The herein collected data give support to the
potential interest of this chelator in detoxification of hard
metal ions.
Synthesis of NTP(PrHPM)3
The synthesis of the ligand starts with the preparation of the
hydroxypyrimidinone side chains, involving the preliminary
preparation of 1-(benzyloxy)-4-(1′,2′,4′-triazol-1′-yl)-2-(1H)-pyri-
midinone19
1
and 3-((tert-butoxycarbonyl)amino)propyl-
amine,20 according to previously published procedures.
Afterwards, those amine-bearing HPM arms are coupled to the
tris-carboxylic backbone, nitrilotripropionic acid (NTP), via
standard carboxyl activation methodologies. Final catalytic
hydrogenolysis to remove the benzyl protecting groups affords
the final NTP(PrHPM)3 product.
4-[3-((tert-Butoxycarbonyl)amino)propylamino]-1-(benzyloxy)-
2(1H)-pyrimidinone, 2. A solution of 1 (0.50 g, 1.86 mmol)
and 3-((tert-butoxycarbonyl)amino)propylamine (0.39 g,
2.23 mmol) in dry THF (15 mL) was stirred for 16 h at reflux
temperature, under a nitrogen atmosphere. The solvent was
evaporated and H2O added to the residue. The aqueous layer
Experimental
General information and instrumentation
The chemicals were of analytical reagent grade, being used was extracted with CHCl3 (5 × 20 mL) and the combined
without further purification. Whenever necessary, the organic organic extracts were washed successively with 5% citric acid
solvents were dried according to standard methods.41 Chemi- solution, H2O, brine and dried over anhydrous magnesium
cal reactions were followed by thin layer chromatography (TLC) sulfate. Evaporation of the solvent, followed by recrystallization
on silica gel 60 F254 plates with 0.2 mm layer thickness from of the residual solid from ethyl acetate, gave product 2 as a
Macherey-Naguel and the compounds were visualized by illu- white amorphous solid (0.52 g, 74%). M.p. 162–164 °C; IR (KBr
mination under UV light at 254 nm.
pellets): νmax/cm−1 3375, 3259, 3127, 3031, 1711, 1643, 1502,
Melting points were determined in a Reichert Thermovar 1170. 1H-NMR (400 MHz, CDCl3): δ (ppm) 1.43 (9 H, s, OC-
apparatus and are uncorrected. FTIR spectra were recorded on (CH3)3), 1.69 (2 H, q, J = 6.4 Hz, CH2CH2CH2), 3.20 (2 H, t, J =
a Perkin-Elmer Spectrum BX v5.3.1 spectrometer. Fourier 6.6 Hz, NHCH2CH2), 3.51 (2 H, t, J = 7.0 Hz, CH2NHC(O)), 5.20
transform (FT) NMR spectra were run on a BRUKER AVANCE (2 H, s, PhCH2), 5.35 (1 H, d, J = 6.0 Hz, NCHvCHC), 6.89 (1H,
400 MHz Ultra-Shield spectrometer at Faculty of Sciences of d, J = 9.0 Hz, NCHvCHC), 7.38 (5 H, s, Ph-H); 13C-NMR
Lisbon University (compounds 2–3) and at the Instituto (400 MHz, CDCl3): δ (ppm) 28.4 (C(CH3)3), 29.7 (CH2CH2CH2),
Superior Técnico (compounds 4 and NTP(PrHPM)3). The 30.94 (CH2NHC(O)), 37.0 (NHCH2CH2), 78.0 (C6H5CH2),
chemical shifts (δ) are reported in ppm from the internal 79.6 (C(CH3)3), 93.6 (NCHvCHC), 128.7 (Ph-C2 and Ph-C6),
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 6033–6045 | 6041

View Article Online
Paper
Dalton Transactions
129.2 (Ph-C4), 130.1 (Ph-C3 and Ph-C5), 134.3 (Ph-C1), 142.5 (Ph-C2 and Ph-C6), 130.4 (Ph-C4), 131.3 (Ph-C3 and Ph-C5),
(NCHvCHC), 153.2 (NC(O)N), 157.1 (NHC(O)-O), 162.4 135.3 (Ph-C1), 144.0 (NCHvCHC), 155.8 (NC(O)N), 164.4
(NvCNH).
(NvCNH), 175.0 (NCH2CH2CO).
4-(3-Aminopropylamino)-1-(benzyloxy)-2(1H)-pyrimidinone
3,3′,3′′-Nitrilotri(N-(3-(1-hydroxy-2-oxo-1,2-dihydropyrimidin-
hydrochloride, 3. A solution of 4-[3-((tert-butoxycarbonyl)- 4-ylamino)propyl)propanamide), NTP(PrHPM)3. To a solution
amino)propylamino]-1-(benzyloxy)-2(1H)-pyrimidinone 2 (0.5 g, of 4 (0.090 g, 0.090 mmol) in MeOH (10 mL) was added 10%
1.33 mmol) in 4 M HCl 1,4-dioxane (8 mL) was stirred at 0 °C Pd/C (0.020 g), and the suspension was stirred under 2 bar H2
under a nitrogen atmosphere until complete consumption of atmosphere, at r.t. for 1 h. The solid was filtered off and the
2 (4 h) as monitored by TLC. The reaction mixture was evapor- solution was evaporated. The residue was recrystallized from
ated, dry ethanol was added to the residue, and then evapor- MeOH/acetone, affording the pure product as a white solid
ated. This process was repeated 3 times to give product 3 as a (0.060 g, 91% yield). M.p. 120–122 °C; MS (ESI+) m/z 732.5
white solid (0.4 g, 98%). M.p. 169–171 °C; IR (KBr pellets): (100%) for [M + H]+, 754.5 (25%) for [M + Na]+. Elemental
νmax/cm−1 3447, 3033, 1735, 1654, 1275. 1H-NMR (400 MHz, analysis found, C, 49.33; H, 6.65; N, 24.63%;
CD3OD): δ (ppm) 2.06 (2 H, q, J = 8.0 Hz, CH2CH2CH2), 3.08 [C30H45N13O9·0.15 acetone] requires C, 49.39; H, 6.25; N,
(2 H, t, J = 8.0 Hz, CH2CH2NH2), 3.54 (2 H, t, J = 8.0 Hz, 24.59%; 1H-NMR (400 MHz, D2O, pD ca. 9, 1% MeOH):
NHCH2CH2), 5.24 (1.50 H, s, C6H5CH2), 5.27 (0.50 H, s, δ (ppm) 1.76 (6 H, q, J = 6.8 Hz, CH2CH2CH2), 2.41 (6 H, t, J =
PhCH2), 5.98 (0.75 H, d, J = 8.0 H, NCHvCHC), 6.26 (0.25 H, 6.8 Hz, NCH2CH2CO), 2.78 (6 H, t, J = 7.0 Hz, NCH2CH2CO),
d, J = 8.0 Hz, NCHvCHC), 7.44–7.54 (5 H, m, Ph-H), 7.88 3.21–3.36 (12 H, m, CH2CH2CH2), 5.74 (3 H, d, J = 7.2 Hz,
(0.75 H, d, J = 8.0 Hz, NCHvCHC), 8.19 (0.25 H, d, J = 8.0 Hz, NCHvCHC), 7.63 (3 H, d,
J = 7.2 Hz, NCHvCHC);
NCHvCHC); 13C-NMR (400 MHz, CD3OD): δ (ppm) 25.8 13C-NMR (400 MHz, D2O, pD ca. 9, 1% MeOH): δ (ppm)
(CH2CH2CH2), 36.6 (NHCH2CH2), 39.5 (CH2CH2NH2), 79.2 28.5
(C6H5CH2), 92.8 (NCHvCHC), 128.5 (Ph-C2 and Ph-C6), 129.3 (CONHCH2CH2CH2),
(CONHCH2CH2CH2),
39.0
33.5
(NCH2CH2CO),
37.5
49.3.5
(CONHCH2CH2CH2),
(Ph-C4), 129.9 (Ph-C3 and Ph-C5), 133.0 (Ph-C1), 145.6 (NCH2CH2CO), 92.7 (NCHvCHC), 145.3 (NCHvCHC), 158.3
(NCHvCHC), 149.2 (NC(O)N), 158.3 (NvCNH).
(NC(O)N), 161.4 (NvCNH), 175.5 (NCH2CH2CO).
3,3′,3′′-Nitrilotri(N-(3-(1-(benzyloxy)-2-oxo-1,2-dihydropyrimi-
din-4-ylamino)propyl)-propanamide), 4. A mixture of 3,3′,3′′-
nitrilotripropanoic acid (NTP, 0.097 g, 0.414 mmol), TBTU
(0.43 g, 1.37 mmol) and N-methylmorpholine (NMM, 0.30 mL,
2.73 mmol) in dry DMF (15 mL) containing molecular sieves
was stirred at r.t. for 1 h. Meanwhile, a solution of 3 (0.450 g,
1.37 mmol) and NMM (0.30 mL, 2.73 mmol) in dry DMF
(10 mL), also containing molecular sieves, was stirred at r.t. for
1 h. The first mixture was filtrated, the solution was added
dropwise to the second one and the mixture was stirred under
N2 for 6 h. The final mixture was evaporated under vacuum.
The crude material was taken into 0.1 M HCl (75 mL) and it
was washed with CH2Cl2 (4 × 50 mL). Concentrated ammonia
was added until a precipitate appeared (pH ca. 4), and it was
extracted with CH2Cl2 (4 × 75 mL); the pH was raised again
with ammonia (pH ca. 8–9), and the solution was extracted
with more CH2Cl2 (3 × 75 mL). The total organic phase was
dried over anhydrous Na2SO4, and then evaporated. The
residue was washed with water (3 × 5 mL), and then with
acetone. After recrystallization from MeOH/acetone, the final
product was obtained, as a beige solid (0.153 g, 37% yield).
M.p. 92–94 °C; MS (ESI+) m/z 1002.8 (100%) for [M + H]+,
1024.8 (25%) for [M + Na]+; 1H-NMR (400 MHz, CD3OD):
δ (ppm) 1.57 (6 H, q, J = 6.8 Hz, CONHCH2CH2CH2), 2.21 (6 H,
t, J = 6.6 Hz, NCH2CH2CO), 2.61 (6 H, t, J = 6.6 Hz,
Potentiometric studies
Measurements. Titrations of the ligands, alone and in the
presence of aluminum (for NTP(PrHPM)3) or of gallium (for
HOPY-PrN), were performed in aqueous solution at T = 25.0
0.1 °C and ionic strength (I) 0.1 M KCl.43 For all the titrations
involving NTP(PrHPM)3, the total volume was 20 mL, the
ligand concentration (CL) was 4.0–5.0 × 10−4 M and CAl/CL was
0 : 1, 1 : 1.1 or 1 : 1.2. For the Ga3+/HOPY-PrN system, the total
volume was 20 mL, the ligand concentration was 1.0–2.0 ×
10−3 M and CGa/CL was 1 : 3. Each assay was done twice and
the value determined for the ionization constant (pKw) was
13.8.
Calculation of equilibrium constants. The stepwise protona-
tion constants, Ki = [HiL]/[Hi−1L][H] (i = 1–7), and the overall
metal-complex stability constants, βM H L
=
[MmHhLl]/
m
h l
[M]m[H]h[L]l, were determined by fitting analysis of the poten-
tiometric data with the HYPERQUAD 2008 program.27 Al3+ and
Ga3+ hydroxide species40–42 were included in the equilibrium
complexation model and the species distribution curves were
obtained with the HYSS program.27
Spectroscopic studies
Measurements. 1H NMR titration of NTP(PrHPM)3 (5 ×
NCH2CH2CO), 3.03 (6 H, q, J = 6.6 Hz, CONHCH2CH2CH2), 3.19 10−4 M in D2O) was performed as previously described,23 and the
(6 H, t, J = 6.8 Hz, CONHCH2CH2CH2), 4.94 (6 H, s, C6H5CH2), final pD values were determined from the equation pD = pH*
5.41 (3 H, d, J = 7.6 Hz, NCHvCHC), 7.16 (3 H, d, J = 7.6 Hz, + 0.4,26 in which pH* corresponds to the reading of the pH
NCHvCHC), 7.19–7.26 (15 H, m, C6H5); 13C-NMR (400 MHz, meter previously calibrated with aqueous buffers at pH 4 and
CD3OD): δ (ppm) 29.8 (CONHCH2CH2CH2), 34.7 (NCH2CH2CO), 7. The electronic spectra of HOPY-PrN (CL = 1.9 × 10−4 M) were
37.8 (CONHCH2CH2CH2), 39.3 (CONHCH2CH2CH2), 50.7 recorded in the range 200–400 nm and those for the 1 : 1 Ga3+/
(NCH2CH2CO), 79.5 (C6H5CH2), 95.7 (NCHvCHC), 129.8 HOPY-PrN system were obtained in the range 250–380 nm
6042 | Dalton Trans., 2013, 42, 6033–6045
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
(CL = 1.9 × 10−4 M, 1 : 1 stoichiometry). For NTP(PrHPM)3 Alamos ECP and a double-zeta basis set including 3d orbitals
(CL = 5.0 × 10−5 M) the electronic spectra were recorded in the and 3d diffuse functions for the valence shell.
range 200–400 nm, while those of the corresponding Fe3+ and
Ga3+ complexes were collected in the range 300–700 nm (CL =
Biodistribution studies
2.7 × 10−4 M) and 200–400 nm (CL = 5.0 × 10−5 M), respectively,
In vivo biodistribution studies were carried out in groups of
under a 1 : 1.2 stoichiometry. For pH ≥ 2, solutions of com-
plexes were prepared as indicated for the potentiometric
measurements. Both Fe3+–NTP(PrHPM)3 and Ga3+–NTP-
(PrHPM)3 systems were studied for pH below 2 (0.8–2), using a
batch titration (7 points), in which the amount of acid added
(from standard 0.1 or 1 M HCl solutions) was calculated for
the total volume solution used.
Calculation of equilibrium constants. The overall iron and
gallium complex stability constants of NTP(PrHPM)3, as well
as the value of βGaHL for HOPY-PrN, were determined with the
PSEQUAD program,28 by fitting analysis of the spectrophoto-
metric data and including in the equilibrium model the data
for Fe3+ and Ga3+ hydrolytic species.44–46
3–5 female CD1 mice (randomly bred, Charles River, from
CRIFFA, Barcelona, Spain) weighing ca. 25 g, 67Ga-citrate injec-
tion solution was prepared by dilution of 67Ga citrate from
MDS Nordion (Ottawa, Canada) with saline to obtain a final
radioactive concentration of 5–10 MBq per 100 μL. The 67Ga–
NTP(PrHPM)3 complex was synthesized by adding 67Ga-citrate
to a saline solution of NTP(PrHPM)3 and the radiochemical
purity, superior to 95%, was determined by ITLC, as reported
before. Mice were intravenously (i.v.) injected with 100 μL
(5–10 MBq) of 67Ga citrate via the tail vein. In a separate group
of animals the i.v. administration was immediately followed by
intraperitoneal (i.p.) injection of 0.5 μmol of the ligand in
100 μL saline solution. Biodistribution of the 67Ga–NTP-
(PrHPM)3 complex was assessed by i.v. injection of 100 μL of
the complex solution previously prepared. Animals were main-
tained on normal diet ad libitum and were sacrificed by cervi-
Determination of partition coefficients
The octanol–water partition coefficients (log P) of NTP-
(PrHPM)3 and HOPY-PrN were determined by the “shakeflask”
method,47,48 which is based on the concentration ratio of the
compounds between 1-octanol and a TRIS buffered aqueous
phase (pH = 7.4). The species concentrations were assessed by
spectrophotometry, based on the absorbance of the benzenoid
bands (π–π*) of the compounds.
cal dislocation at 15 min,
1 h, 4 h and 24 h post-
administration. The administered radioactive dose and the
radioactivity in sacrificed animals were measured in a dose
calibrator (Aloka, Curiemeter IGC-3, Tokyo, Japan). The differ-
ence between 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, Berthold Technologies,
Germany). Biodistribution results were expressed as percent of
injected activity per total organ (%I.A./organ) and presented
as mean values SD. For blood, bone and muscle, total
activity was calculated assuming, as previously reported, that
these organs constitute 7, 10 and 40% of the total weight,
respectively.
Molecular modeling
The structure of the Fe3+–NTP(PrHPM)3 complex was opti-
mized by quantum mechanical calculations based on Density
Functional Theory (DFT) methods, using the Gaussian-03W
software.37 This energy minimization was carried out using
the functional B3LYP chemical model, and it was performed
in two steps. In the first step the 3-21G basis set was used with
a direct self-consistent field method (SCF) and SCF conver-
gence criterion of 10−5. The results of these calculations were
subjected to a deeper second optimization step using the
LANL2DZ basis with the same SCF settings. The functional
B3LYP has been shown to be an accurate density functional
method,49 and it has proved to give reliable geometries and
energies for complexes with several metal ions, namely tran-
sition metal ions.50 The B3LYP model is a combination of the
Becke three-parameter hybrid functional51 with the Lee–Yang–
Parr correlation functional (which also includes density gradi-
ent terms).52,53 Regarding the basis set, 3-21G is the simplest
of Pople’s split-valence basis sets, and it can be used with rela-
tively good accuracy for molecules containing first and second-
row elements.54 It has been used in our first calculation step,
mostly in order to optimize the conformation of the ligand.
The LANL2DZ basis set specifies the Dunning–Huzinaga
valence double-zeta basis set (D95V) on the first row (all
the atoms of the ligand),55 and Los Alamos ECP plus DZ on
Na–Bi.56 In this way, the Fe atom is described through the Los
Acknowledgements
The authors thank the Portuguese NMR Network (IST-UTL
Center) for providing access to the NMR facilities, the Portu-
guese Fundação para a Ciência e Tecnologia (FCT) (project
PTDC/QUI/65647/2006 and post-Doc grant SFRH/BPD/29874/
2006 – S. M.) as well as PEst-OE/QUI/UI0100/2011 for financial
support.
References
1 R. R. Crichton, in Inorganic Biochemistry of Iron Metabolism:
From Molecular Mechanisms to Clinical Consequences, Wiley,
Chichester, 2001.
2 G. Faa and G. Crisponi, Coord. Chem. Rev., 1999, 184,
291–310.
3 R. A. Yokel and P. J. McNamara, Pharmacol. Toxicol., 2001,
88, 159–167.
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 6033–6045 | 6043

View Article Online
Paper
Dalton Transactions
4 G. Crisponi, V. M. Nurchi, G. Faa and M. Remelli, Monatsh. 32 A. E. Martell, R. M. Smith and R. J. Motekaitis, in Critically
Chem., 2011, 142, 331–340.
Selected Stability Constants of Metal Complexes Database,
College Station TX, version 4, 1997.
33 R. J. Motekaitis and A. E. Martell, Inorg. Chim. Acta, 1991,
183, 71–80.
5 P. Zatta, J. Alzheimers Dis., 2006, 10, 33–37.
6 C. Exley, J. Inorg. Biochem., 1999, 76, 133–140.
7 L. E. Scott and C. Orvig, Chem. Rev., 2009, 109, 4885.
8 G. Crisponi, V. M. Nurchi, V. Bertolasi, M. Remelli and 34 A. Evers, R. D. Hancock, A. E. Martell and R. J. Motekaitis,
G. Faa, Coord. Chem. Rev., 2012, 256, 89–104. Inorg. Chem., 1989, 28, 2189–2195.
9 M. A. Santos, M. A. Esteves and S. Chaves, Curr. Med. 35 W. R. Harris and V. L. Pecoraro, Biochemistry, 1983, 22,
Chem., 2012, 19, 2773–2793. 292–299.
10 W. R. Harris and J. Sheldon, Inorg. Chem., 1990, 29, 36 J. Beardmore and C. Exley, J. Inorg. Biochem., 2009, 103,
119–124. 205–209.
11 R. D. Propper, S. B. Shurin and D. G. Nathan, 37 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
N. Engl. J. Med., 1976, 294, 1421–1423.
12 R. C. Hider, G. Kontoghiorghes, J. Silver and
M. A. Stockham, UK Pat., 2117766, 1982.
13 U. Heinz, K. Hegetschweiber, P. Acklin, B. Faller,
R. Lattmann and H. P. Schnebli, Angew. Chem., Int. Ed.,
1999, 38, 2568–2570.
14 R. A. Yokel, P. Ackrill, E. Burgess, J. P. Day, J. L. Domingo,
T. P. Flaten and J. Savory, J. Toxicol. Environ. Health, 1996,
48, 667–683.
15 Z. D. Liu and R. C. Hider, Med. Res. Rev., 2002, 22, 26–64.
16 M. A. Santos, S. M. Marques and S. Chaves, Coord. Chem.
Rev., 2012, 256, 240–259.
17 M. A. Santos, Coord. Chem. Rev., 2008, 252, 1213–1224.
18 S. Chaves, S. Marques, A. M. F. Matos, A. Nunes, L. Gano,
T. Tuccinardo, A. Martinelli and M. A. Santos, Chem.–Eur.
J., 2010, 16, 10535–10545.
19 M. A. Esteves, A. Cachudo, S. Chaves and M. A. Santos,
Eur. J. Inorg. Chem., 2005, 597–605.
20 D. Boturyn, A. Boudali, J. F. Constant, E. Defrancq and
J. Lhomme, Tetrahedron, 1997, 53, 5485–5492.
M. A. Robb, J. R. Cheeseman, J. A. Montgomery, 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. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko,
P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith,
M. A. Al-Laham, C. Y. Peng, A. Nanayakkara,
M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen,
M. W. Wong, C. Gonzalez and J. A. Pople, GAUSSIAN 03
(Revision C.02), Gaussian, Inc., Wallingford CT, 2004.
21 A. Katoh, Y. Hida, J. Kamitani and J. Ohkanda, J. Chem. 38 C. B. Dobson, J. P. Day, S. J. King and R. F. Itzhaki, Toxicol.
Soc., Dalton Trans., 1998, 3859–3864. Appl. Pharmacol., 1998, 152, 145–152.
22 S. Chaves, A. C. Mendonça, S. M. Marques, M. I. M. Prata, 39 A. Radunovic, H. T. Delves and M. W. B. Bradbury, Biol.
A. C. Santos, A. F. Martins, C. F. G. C. Geraldes and
M. A. Santos, J. Inorg. Biochem., 2011, 105, 31–38.
23 M. A. Santos, M. Gil, S. Marques, L. Gano, G. Cantinho and
S. Chaves, J. Inorg. Biochem., 2002, 92, 43–54.
24 E. T. Clarke and A. E. Martell, Inorg. Chim. Acta, 1992, 196,
185–194.
Trace Elem. Res., 1998, 62, 51–64.
40 R. C. Walton, K. N. White, F. Livens and C. R. McCrohan,
Biometals, 2010, 23, 221–230.
41 W. L. F. Armarego and D. D. Perrin, in Purification of Labo-
ratory Chemicals, Butterworth-Heinemann Press, Oxford,
4th edn, 1999.
25 R. A. Yokel, A. K. Datta and E. G. Jackson, J. Pharmacol. 42 F. J. C. Rossotti and H. Rossotti, J. Chem. Ed., 1965, 42,
Exp.Ther., 1991, 257, 100–106. 375–378.
26 A. K. Covington, M. Paabo, R. A. Robinson and R. G. Bates, 43 S. Chaves, M. Gil, S. Marques, L. Gano and M. A. Santos,
Anal. Chem., 1968, 40, 700–706. J. Inorg. Biochem., 2003, 97, 161–172.
27 A. Gans, A. Sabatini and A. Vacca, Talanta, 1996, 43, 44 C. F. Baes and R. E. Mesmer, in The Hydrolysis of Cations,
1739–1753. Wiley, New York, 1st edn, 1976.
28 L. Zékány, I. Nagypál and G. Peintler, PSEQUAD version 45 L. O. Öhman and W. Forschling, Acta Chim. Scand. Ser. A,
5.01, 2001. 1981, 35, 795–802.
29 J. Okhanda, J. Kamitani, T. Tokumitsu, Y. Hida, 46 R. M. Smith and A. E. Martell, in Critical Stability Constants,
T. Konakahara and A. Katoh, J. Org. Chem., 1997, 62,
3618–3624.
30 M. A. Santos, S. Gama, L. Gano, G. Cantinho and E. Farkas,
Dalton Trans., 2004, 3772–3781.
31 S. Gama, P. Dron, S. Chaves, E. Farkas and M. A. Santos,
Dalton Trans., 2009, 6141–6150.
Plenum Press, New York, 1st edn, 1976, vol. 4, p. 11.
47 M. A. Santos, M. Gil, L. Gano and S. Chaves, J. Biol. Inorg.
Chem., 2005, 10, 564–580.
48 A. Leo, C. Hansch and D. Elkins, Chem. Rev., 1971, 71,
526–616.
49 C. W. Bauschlicher, Chem. Phys. Lett., 1995, 246, 40–44.
6044 | Dalton Trans., 2013, 42, 6033–6045
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
50 A. Ricca and C. W. Bauschlicher, Theor. Chim. Acta, 1995, 54 M. S. Gordon, J. S. Binkley, J. A. Pople, W. J. Pietro and
92, 123–131.
W. J. Hehre, J. Am. Chem. Soc., 1982, 104, 2797–2803.
55 T. H. Dunning Jr. and P. J. Hay, in Modern Theoretical
Chemistry, ed. H. F. Schaefer, III, Plenum, New York, 1976,
vol. 3, pp. 1–28.
51 D. Becke, Phys. Rev. A, 1988, 38, 3098–3100.
52 C. Lee, W. Yang and R. G. Parr, Phys. Rev. B: Condens.
Matter, 1988, 37, 785–789.
53 B. Miehlich, A. Savin, H. Stoll and H. Preuss, Chem. Phys. 56 P. J. Hay and W. R. Wadt, J. Chem. Phys., 1985, 82,
Lett., 1989, 157, 200–206.
299–310.
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 6033–6045 | 6045