Design, synthesis, and physicochemical and biological characterization of a new iron chelator of the family of hydroxychromenes

Article abstract of DOI:10.1021/jm021022u

Increasing evidence suggests that iron plays an important role in tissue damage both during chronic iron overload diseases (i.e., hemochromatosis) and when, in the absence of actual tissue iron overload, iron is delocalized from specific carriers or intracellular sites (inflammation, neurodegenerative diseases, postischaemic reperfusion, xenobiotic intoxications, etc.). In the present work, we appropriately modified an iron chelator of the hydroxychromene family in order to obtain a tridentate chelator that would inactivate the iron redox cycle after its complexation, with a view to using this molecule in human therapy and/or in disease prevention. We synthesized such a chelator for the first time and show, by different physicochemical analysis, its tridentate nature and, importantly, its capacity to chelate iron with enough strength to inhibit both iron-dependent H₂O₂ generation and lipid peroxidation in in vitro biological systems.

Full text of DOI:10.1021/jm021022u

5
776
J . Med. Chem. 2002, 45, 5776-5785
Design , Syn th esis, a n d P h ysicoch em ica l a n d Biologica l Ch a r a cter iza tion of a
New Ir on Ch ela tor of th e F a m ily of Hyd r oxych r om en es
,
†
‡
‡
‡
‡
‡
Marco Ferrali,* Sabrina Bambagioni, Antonio Ceccanti, Donato Donati, Gianluca Giorgi, Marco Fontani,
‡
‡
§
|
Franco Laschi, Piero Zanello, Mario Casolaro, and Antonello Pietrangelo
Departments of Physiopathology and Experimental Medicine, Chemistry and Chemical and Biosystem Sciences and
Technology, University of Siena, via Aldo Moro, I-53100 Siena, Italy, and Department of Medicine and Medical Specialities,
University of Modena and Reggio Emilia, Largo del Pozzo, 71, I-41100 Modena, Italy
Received August 6, 2002
Increasing evidence suggests that iron plays an important role in tissue damage both during
chronic iron overload diseases (i.e., hemochromatosis) and when, in the absence of actual tissue
iron overload, iron is delocalized from specific carriers or intracellular sites (inflammation,
neurodegenerative diseases, postischaemic reperfusion, xenobiotic intoxications, etc.). In the
present work, we appropriately modified an iron chelator of the hydroxychromene family in
order to obtain a tridentate chelator that would inactivate the iron redox cycle after its
complexation, with a view to using this molecule in human therapy and/or in disease prevention.
We synthesized such a chelator for the first time and show, by different physicochemical
analysis, its tridentate nature and, importantly, its capacity to chelate iron with enough strength
to inhibit both iron-dependent H
systems.
2
O
2
generation and lipid peroxidation in in vitro biological
In tr od u ction
Ch a r t 1
It is well-known that iron plays a primary role in
mammalian life and in all living cells. It is usually
transported and stored by specific proteins which allow
the physicochemical control of the metal during and
after its redox activity. If, for different reasons, iron is
released from these proteins, or is excessively absorbed
by liver from food, it becomes extremely dangerous for
cells and the tissues by entering a redox cycle which
generates toxic free radicals.1,2 These chemically active
species, which originate from the reaction of free iron
with molecular oxygen or hydrogen peroxide by the
Fenton and/or Haber-Weiss reactions, cause the cells
Undesirable effects of prolonged treatments with such
drugs are obviously related to direct toxicity of the
molecules or to iron deprivation of the bone marrow.
17,18
The ideal chelator should remove only the excess of iron
from its natural carriers, but producing such a chelator
is obviously extremely difficult. The picture is compli-
cated by the fact that the chelator should not be toxic,
should be suitably lipophilic to enter the cell, should
chelate iron intracellularly, and should be rapidly
excreted by urine and/or bile. Due to the wide variety
of diseases in which iron chelators need to be utilized,
it is important to continue to search for novel chela-
3
4
to undergo lipid peroxidation, protein oxidation, and
5
DNA damage.
There is increasing evidence of direct or indirect
involvement of iron in human diseases such as haemo-
chromatosis, thalassemias,7 neurodegenerative dis-
6
8
9
eases, cancer, and in all pathological states of free
19
tors. This can be achieved either by improving known
radical-mediated tissue damage.1
0-12
Thus the goal of
drugs or by proposing suitable new molecules in order
to obtain compounds showing different affinities for iron.
In a previous paper we suggested that the 3-hydroxy-
chromen-4-ones could be assessed as iron-chelating
improving the quality of life of people affected by severe
iron overload (in particular, by haemochromatosis and
thalassemia) has been the stimulus to develop drugs
that would increase iron excretion in man.
20
agents in vivo. We showed that 3-hydroxy-2-methyl-
To this end several naturally occurring molecules and
synthetic products have been tested in the past decades.
Despite their disadvantages, desferrioxamine, hydroxy-
pyridones, and pyridoxal hydrazones are the most
chromen-4-one (IC1, Chart 1) chelates iron(III) in vitro
and possesses a good affinity for the metal ion. This
could guarantee the lack of effect of the chelator toward
iron transported by biological carriers, but it remains
to be seen whether iron should be still redox active after
chelation. As a bidentate ligand, IC1 leaves two coordi-
nation sites of the Fe ion unoccupied, making it
likely to not be sufficiently inert for redox reactions. In
fact, IC1 is not able to inhibit iron-induced lipid peroxi-
dation in hepatic microsomal preparations in vitro
1
3-16
effective chelators currently used in human therapy.
III
20
*
To whom correspondence should be addressed. Phone: +39-0577-
2
34007. Fax:+39-0577-234009. E-mail: ferrali@unisi.it.
†
Department of Physiopathology and Experimental Medicine,
University of Siena.
‡
Department of Chemistry, University of Siena.
Department of Chemical and Biosystem Sciences and Technology,
§
(unpublished personal observations).
University of Siena.
|
With a view to developing a new synthetic iron
Department of Medicine and Medical Specialities, University of
Modena and Reggio Emilia.
chelator with a higher affinity for iron and keeping the
1
0.1021/jm021022u CCC: $22.00 © 2002 American Chemical Society
Published on Web 11/15/2002

New Hydroxychromene Iron Chelator
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 26 5777
Sch em e 1
IC1 structure as the basic molecule, we synthesized a
new molecule: 3-hydroxy-2-(5-hydroxypentyl)chromen-
4
-one (IC2, Chart 1). The aim was to obtain a tridentate
chelator capable of enhancing the stability constant of
the iron complex in order to obtain a complete inhibition
of the redox cycle of iron in biological systems. We chose
to use the basic IC1 structure, rather than considering
alternative molecular structures due to the fact that the
chromanic ring is common to tocopherols and this
results in good lipophilicity and low toxicity for most
synthetic derivatives. The characterization of the mol-
ecule and preliminary in vitro and in vivo studies on
its possible use in the biological environment are
presented in this paper.
F igu r e 1. X-ray crystal structure of 3-hydroxy-2-(5-hydroxy-
pentyl)chromen-4-one (IC ).
2
...
squares plane. The distance O(2) O(3), which consti-
tutes the two most probable sites for iron chelation, is
equal to 2.767(4) Å, which is very close to the value of
Resu lts a n d Discu ssion
2
0
The synthesis of the 3-hydroxy-2-(5-hydroxypentyl)-
chromen-4-one (IC2) was accomplished as described in
Scheme 1.
The synthesis of IC2 starts from the monolateral
protection of an alcoholic group of 1,6-hexanediol by
reaction with 2,4-dihydropyran. This procedure allows
the next monolateral oxidation of the 6-(tetrahydropy-
ran-2-yloxy)hexan-1-ol to 6-(tetrahydropyran-2-yloxy)-
hexanal (Scheme 1, A) with pyridinium chlorochromate
2.734(3) Å found in IC1. In the crystal lattice, O(2)-H
participates in intra- and intermolecular hydrogen
bonding with O(3) and stacking interactions occur
between the heterocyclic systems.
Spectrophotometric analysis of the dynamics of the
complex formation with iron(III) shows that 2 mol of
the chelator are engaged in the complexation of 1 mol
of iron (Figure 2). The isosbestic point of the spectra at
335.5 nm is evident until the concentration of the iron
rises from zero to half the concentration of the chelator
(Figure 2A-D). When the iron concentration exceeds
this value (Figure 2E), spectral changes are evident,
reasonably due to the presence of an excess of uncom-
plexed FeCl . In fact a solution of 100 µM FeCl in the
(see Experimental Section for further details). The
carbonyl group is the means by which the condensation
of the last product with 2′-hydroxyacetophenone takes
place in the presence of pyrrolidine. The subsequent
cyclization in the pyranic ring, after water is lost, is
immediate and spontaneous giving 2-[5-(tetrahydropy-
ran-2-yloxy)pentyl]chroman-4-one (Scheme 1, B). The
hydroxylation of the latter compound is obtained by the
reaction with isoamyl nitrite after removal of the pyrane
protection to avoid possible interference of the pyrane
displaced during the reaction, yielding IC2. The com-
3
3
same buffer shows an absorbance of 0.180 and 0.250 au
at 335.5 and 300 nm, respectively (not shown).
Ma ss Sp ectr om etr y of th e Com p lex. Different
mass spectrometry ionization techniques, i.e., LSIMS
and ESI, were used for the structural characterization
of the complex obtained by mixing FeCl and IC2 in
3
1
bined analyses of the purified product obtained, by H
methanol. Both LSIMS and ESI show the presence of
1
3
NMR, C NMR in CDCl3, mass spectrometry, and X-ray
diffraction (Figure 1) assured the identity of the final
product (see Experimental Section). The IC2 is a color-
less solid with a melting point of 112-114 °C; it is a
fluorescent product having an exmax at 335 nm and an
emmax at 485 nm in CH2Cl2 solution. Its mass spectrum
shows the molecular ion at m/z 248 (rel inten 15%) and
two very abundant ions at m/z 189 (75%) and 176 (100%)
both corresponding to the fragmentation of the aliphatic
side chain.
intense ions at m/z 550 corresponding to an Fe(III):IC2
ratio of 1:2, while species with different molar ratios
were not detectable.
2
4
The product ion spectra, ranging from MS to MS ,
obtained by ESI of the iron complex, are reported in
Figure 3. The main decompositions obtained by selecting
the ions at m/z 550 as the main beam (Figure 3A) consist
in the loss of one and two water molecules yielding ions
3
at m/z 532 and 514, respectively. The MS spectrum of
ions at m/z 532 (Figure 3B) shows the loss of 54 mass
units corresponding to C4H6 from the side chain produc-
ing the species at m/z 478. Also ions at m/z 514 eliminate
54 mass units yielding ions at m/z 460 (Figure 3C).
The X-ray structure of IC2 is illustrated in Figure 1.
The heterocyclic system is planar with the largest
deviation (0.047(4) Å) shown by C(4) from its least-

5
778 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 26
Ferrali et al.
F igu r e 2. Spectrophotometric detection of Fe(III) complexation with IC
2
: (A) spectrum of 200 µM IC
2
in Tris-Maleate buffer pH
7
.4; (B-E) spectra of IC after the addition of 50, 70, 100, and 200 µM FeCl , respectively. See Experimental Section for details.
2
3
F igu r e 3. Electrospray MS experiments of the [FeIII-((IC
n
)
2
- 3H)] complex in methanol. (a) MS/MS of the species at m/z 550;
2
3
4
(
b) MS of the cations at m/z 532; (c) MS of the species at m/z 514.

New Hydroxychromene Iron Chelator
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 26 5779
After having unambiguously determined that the
species formed by mixing FeCl3 and IC2 is detected at
m/z 550, we have to establish whether this species is a
charged ion already present in solution or whether it is
produced by a protonation reaction due to ionization.
The study of negative ions was able to clarify this point.
When we used LSIMS in negative mode we obtained
very intense ions at m/z 549, while under ESI (-)
conditions a species is detected at m/z 548, with very
scarce abundance. This apparent disagreement can be
rationalized by considering that under LSIMS (-)
conditions a radical anion can be formed. On the other
hand, this process is highly unfavored under the ESI
(-) regime where radical ions are only seldom observed.
The prevailing process under ESI (-) conditions is the
abstraction of a proton that for this complex is very
unfavored, as shown by the scarce abundance of ions
at m/z 548. It follows that the species detected at m/z
F igu r e 4. Computed species distribution plot of the ligand
IC and Fe(III) system. (IC /Fe molar ratio of 2:1; [IC ] ) 0.4
mM, [Fe ] ) 0.2 mM).
5
50 both under LSIMS and ESI is formed by protonation
2
2
2
III
during the ionization, and it is reasonably attributed
+
to [Fe(IC2 - H)(IC2 - 2H) + H] . On the other hand,
close to 2.5. Only in this case the stoichiometry of the
the negative ions at m/z 549 (LSIMS) and m/z 548 (ESI)
III
-
•
complex species of Fe (IC2)2H-3 fits the potentiometric
are due to [Fe(IC2 - H)(IC2 - 2H)] and [Fe(IC2 - H)-
-
curve and an estimated stability constant log â is
obtained. The log â value of -1.76 for the reaction: Fe-
(
IC2 - 2H) - H] , respectively.
This suggests that by mixing FeCl3 and IC2 a neutral
III
+
(
III) + 2IC2 ) Fe (IC2)2H-3 + 3H suggests deproto-
complex with stoichiometry [Fe(IC2 - H)(IC2 - 2H)] is
obtained. It implies a deprotonation of one ligand on the
oxygen at position 3 of the pyran ring, while the other
ligand undergoes a double deprotonation of both the
hydroxy groups. This suggests that the oxygen of the
long side chain participates in the chelation of the iron-
nation of the coordinating ligand. In particular, three
protons for each Fe(III) ion are titrated before the sharp
equivalent point. The splitting-off of three protons in
the acidic region is compatible with the very high
′
stability constant value (log â )26) of the direct
III
reaction: Fe(III) + IC2H-2 + IC2H -1 ) Fe (IC2)2H -3.
(III) ion.
It is noteworthy that the iron(III) complex forms even
in the low pH region.
Mass spectrometry experiments aimed at evaluating
the chelating strength of IC2 toward Fe(III) were also
Figure 4 shows a plot of the species distribution
curves obtained by taking into account only the complex
formation equilibrium at the ligand/metal molar ratio
of 2:1. It is evident that complexation reaches its
maximum at pH 5, when the concentration of the free
Fe(III) ion is close to zero. Extrapolation to the latter
concentration at pH 7.4 gives a pFe(III) ) 16.3. This
value is obtained when the calculation is made with
III
carried out. By mixing a solution of the complex [Fe
(
-
+
IC1 - H)2] with IC2, in a 1:1 molar ratio, the coordina-
tion strength of IC2 causes the removal of the iron from
the complex with IC1and ions at m/z 550 are produced
in very high abundance. This suggests that IC2 has
much stronger coordinating properties than IC1 toward
Fe(III) ions.
P oten tiom etr y. Potentiometric measurements in
aqueous solution have been performed to further study
the formation of the iron(III) complex with IC2. The free
ligand IC2 is scarcely soluble in water at low pH, but
its solubility increases sharply above pH 7. The original
colorless solution becomes yellow in the alkaline region
because of the ionization of the OH group(s) that may
trigger electron delocalization along the carbon rings.
Protonation of the ionized groups was easily recognized
from the buffered region appearing in the potentiometric
curve. The computed basicity constant value (log K )
-5
total concentrations of ligand and Fe(III) ion of 10 and
-6
15
1
0
M, respectively. The calculated pFe(III) value can
be considered on the borderline where the hydrolysis
reaction of the metal may occur. In our case this reaction
occurs only in alkaline conditions and above the physi-
ological pH.
From the potentiometric data we can reasonably
hypothesize that the Fe(III) ion coordinates with the
tridentate ligand according to an octahedral geometry
giving rise to a neutral complex.
X-Ba n d EP R An a lysis. Figure 5 compares the
8
.88 ( 0.04) is close to that of the homologous ligand
III
X-band EPR spectra of the Fe (IC2)2 complex recorded
IC1 (8.90 ( 0.03), thus suggesting that ionization of the
ligand really involves the OH group directly linked to
the ring.
III
in CH2Cl2 solution (Figure 5A,B) with that of Fe Cl3
Figure 5C). The line shape analysis confirms the metal-
(
centered nature of the paramagnetic signal (gi * gelectron
) 2.0023) for the two iron complexes, and the pertinent
parameters are consistent with a S ) 5/2, Fe(III) high
spin state (Table 1). Such features can be accounted for
the strong magnetic effect of the Zero Field Splitting
Unlike protonation, the complex formation equilibri-
um occurs only in the low pH region. The dark-violet
solution of the iron(III) complex becomes clearer as the
ligand improves its solubility in the presence of Fe(III)
ions. As the pH reaches the alkaline region, the iron
complex is destroyed and the solution again assumes
the yellow color of the free ligand. The best potentio-
metric curve is obtained with a ligand/metal molar ratio
2
1,22
interaction, typical of S > 1/2 system.
Because of the strong interplay between the ligand
framework and the S ) 5/2 electrons of the Fe(III) ion,
such an interaction affords significant variations in the

5
780 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 26
Ferrali et al.
Ch a r t 2
increase in ∆Haniso values of FeIII(IC1)2 with respect to
III
Fe (IC2)2 likely reflects differences in apical coordina-
tion (z axis). The bidentate IC1 ligand leaves the axial
position open to different coordinating species, whereas
the tridentate IC2 ligand occupies all the coordination
sites, thus causing significant geometrical constraints
(
see the geometry optimization section).
In fluid solution, the intense adsorption of FeIII(IC2)2
becomes quite symmetrical (giso ) 2.017(6)) (first and
second derivative spectra) and unresolved, due to the
overall isotropic line width, ∆Hiso ) 135(8) G, which
largely overlaps the underlying S ) 5/2 fine structure.
III
Geom etr y Op tim iza tion of F e (IC2)2 Com p lex.
On the light of satisfactory performances in geometry
2
5
F igu r e 5. X-Band EPR spectra of the [FeIII-((IC
optimization of transition metal complexes, we adopted
PM3 semiempirical method to optimize the molecular
2
)
2
- 3H)]
at room temperature (298 K) (A), 100 K
at 100 K (C).
complex in CH
B), and FeCl
2 2
Cl
III
(
3
structure of Fe (IC2)2 in the gas phase. The complex
assumes an highly distorted octahedral coordination.
The mer-isomer results only 17 kJ /mol more stable than
the fac-isomer (see Chart 2).
Ta ble 1. X-band EPR Parameters of the FeIII-L and FeIII-Cl3
Complexes in CH2Cl2 Solutiona
complex
giso
ganiso
∆Hiso
∆Haniso
Electr och em istr y. Since electrochemical investiga-
tions are commonly employed to determine the coordi-
FeIII-IC2
2.017(6)
2.012(8)
135(8)
70(8)
100(8)
120(6)
130(6)
220(30)
340(30)
4
.257(8)
2.016(6)
.259(6)
2.014(8)
.738(8)
2
0,26-29
nating ability of siderophores,
a cyclic voltam-
FeIII-IC1
FeIII-Cl3
2.014(5)
2.027(8)
130(6)
metric study was performed in order to compare the
affinity of the IC2 ligand for Fe(III) ion with those of
the IC1 precursor and of the well-known desferrioxam-
ine B (DEF) (Figure 6A).
4
290(30)
3
a
giso, ∆Hiso: T ) 298 K; ganiso, ∆Haniso: T ) 100 K; ∆H values
in Gauss.
Given that, in agreement with the above-mentioned
potentiometric and mass spectrometric measurements,
preliminary voltammetric experiments on each ligand
confirm that both IC and IC form stable complexes
line shape and important dependence of the paramag-
netic parameters upon different experimental condi-
tions. Accordingly, the signals are broad and unresolved
1
2
with Fe(III) ion in a ligand-to-metal ratio of 2:1, whereas
DEF forms the 1:1 complex, it is immediately evident
2
3,24
(Figure 5).
No metal hyperfine or ligand superhy-
2
0
perfine interactions (if any) are present either in glassy
that, as expected, the original Fe(III)/Fe(II) reduction
of FeCl3 (Epc ) +0.29V, vs SCE) upon complexation
undergoes cathodic shifts which are a function of the
nature of the different ligands. In addition, the quasi-
reversible character of the one-electron reduction of
5
7
or fluid solutions ( Fe: natural abundance ) 2.15%,
5
7
I- Fe ) 1/2).
Table 1 points out the significant effect played by IC1
III
and IC2 ligands on the S ) 5/2 Electron Spin Fe
system in comparison with FeIIICl3: a noticeable dif-
FeCl3 (∆Ep ) 410 mV at 0.05 Vs ) becomes irreversible
-1
ference in both ∆Hi and gi values occurs. Moreover, the
upon complexation (see below).

New Hydroxychromene Iron Chelator
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 26 5781
Ta ble 2. Stability Constant Values of Fe(III) Complexes
Derived from Electrochemical Data.
ligand
log âFe(III)/âFe(II)
log âFe(III)
log âFe(II)
IC1
IC2
DEF
8.0
8.8
15.8
-
-
19.2
15.2
g26a
a,b
31
a
b
From potentiometric data in acid solution; From ref 32.
Since log âFe(III) values have been obtained in acid
solutions, it is evident that the log âFe(II) values calcu-
lated here are purely speculative. For example, a log
âFe(II) value of 10.0 has been calculated from the
electrochemical investigation of Fe(III)(DEF) in strong
2
8
alkaline solution (pH 10.5).
In support of the deprotonation reaction observed both
by potentiometric or mass spectrometric methods upon
mixing FeCl3 and IC2, Figure 6B shows the cyclic
voltammogram obtained at a platinum electrode in a
methanol solution of FeCl3 in the presence of IC2 (in
1
:3 ratio). As seen, the previously discussed irreversible
III
Fe (IC2)2 reduction (Ep ) -0.01 V) is followed by a
major (about three time more intense) reversible reduc-
tion (E°′ ) -0.27 V). Such a reduction process was
coincident with the proton reduction experimentally
observed on a methanol solution containing 3 mM HCl.
Under the same experimental conditions the use of a
glassy carbon electrode makes the proton reduction
irreversible and shifts toward much more negative
potential values (Ep ) -1.0 V), thus explaining the clear
picture illustrated in Figure 6A.
Finally, we point out that the irreversibility of the Fe-
F igu r e 6. (A) Cyclic voltammetric profiles recorded at a glassy
(
III)/Fe(II) reduction upon complexation suggests that
in all cases the instantaneously electrogenerated Fe-
II) complex monoanion is not stable under the original
carbon electrode on methanol solutions containing [NBu
4
]ClO
(1.6 mM) + IC (4.8
(1.7 mM) +
4
(0.1 M) and (a) FeCl
3
(1.6 mM), (b) FeCl
3
1
(
mM), (c) FeCl (1.8 mM) + IC
3
2
(5.4 mM); (d) FeCl
3
ligand composition which stabilizes the Fe(III) ion.
Ir on -Dep en d en t H 2O2 a n d TBAR S F or m a t ion .
The central focus of our work is to build a chelator able
to inhibit iron-dependent biological oxidative stress, that
is, to make iron electrochemically inert in the biological
environment after its complexation. Therefore we per-
formed a series of experiments to evaluate this impor-
tant aspect. We previously realized that the bidentate
chelator IC1 was not able to inactivate the iron-redox
cycle and thus prevent oxidative stress in in vitro
biological systems (Ferrali, unpublished results). The
higher affinity of the new chelator for Fe(III), as deduced
from electrochemistry, potentiometry, and mass spec-
trometric measurements, suggests that IC2 could pre-
vent the redox activity of the Fe(III) ion. To verify this
hypothesis, rat liver microsomes were incubated in vitro
in an iron-dependent pro-oxidant system and, sepa-
rately, hydrogen peroxide production and lipid peroxi-
dation evolution were evaluated. It is well-known that,
in the presence of dioxygen, iron ions easily generate
hydrogen peroxide which, in turn, catalyzes the hy-
droxyl radical formation widely assumed to be respon-
sible for oxidative alteration of membrane phospholip-
DEF (3.4 mM). (B) Cyclic voltammogram recorded at a
platinum electrode on a methanol solution containing [NBu ]-
ClO (0.1 M) and FeCl (1.8 mM) + IC (5.4 mM). Scan rate
.05 V s
4
4
3
2
-
1
0
.
From a qualitative viewpoint, Figure 6A immediately
proves that the coordinating ability of IC2 ligand toward
Fe(III) ion is somewhat higher than that of IC1 (Epc )
-0.01V (IC2) and + 0.05V (IC1) vs SCE), but notably
lower than that of DEF (Epc ) -0.42V, vs SCE). This
was further supported by competition experiments. As
a matter of fact, both the addition of IC2 ligand to
solution of Fe(IC1)2, and addition of DEF ligand to Fe-
(IC2)2 ultimately leads to the voltammetric profiles of
Fe(IC2)2 and Fe(DEF), respectively.
The shift of the redox potential of free metal ions upon
complexation is strictly related to the stability constant
values of the involved redox couples according to the
3
0
relationship:
âox.
(E°′)complex ) (E°′)free - 0.059 log
âred.
2
which holds in those cases in which the complex formed
by each member of the redox couple involves the same
number of ligand molecules. In these cases the electrode
potential of the redox processes does not vary with the
ligand concentration. Assuming that the electrode po-
tential of free Fe(III)/Fe(II) ion in aqueous solution (E°′
ids. The biological membranes we prepared allowed us
to obtain microsomes with sufficient lysosome contami-
nation, to have a moderate initiation and propagation
of the iron-dependent lipid peroxidation without addi-
3
3
tion of NADPH or other reducing agents.
Figure 7 shows the results obtained for H2O2 forma-
tion. As can be observed, the production of H2O2 strongly
increases when iron-ADP complex is added as an
)
+0.51 V, vs SCE) 31 also holds in methanol solution,
the data summarized in Table 2 are obtained.

5
782 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 26
Ferrali et al.
F igu r e 8. Iron-dependent TBARS evolution in rat liver
microsomal preparations (MC). A: Control MC; B: MC plus
1
00 µM FeCl3, 2.5 mM ADP; C: as in B plus 100 µM IC
2
; D:
as in B plus 200 µM IC
was 1.5 mg/mL. Values represent the mean of four separate
2
. The microsomal protein concentration
experiments ( SE.
means of the ADP-Fe complex causes the appearance
of TBARS in the incubation medium at a concentration
of about 10 µM (Figure 8B) whereas the presence of 200
µM IC2 completely prevents TBARS formation (Figure
2 2
F igu r e 7. Iron-dependent H O formation in rat liver mi-
crosomal preparations (MC). A: Control MC with the only
addition of 2.5 mM ADP; B: MC as in A plus DCFA; C:
8
D). Again, as with H2O2 formation, 100 µM of the
complete system (plus 100 µM FeCl
IC ; E: as in C plus 200 µM IC ; F: control MC plus 2.5 mM
ADP plus15 µM IC plus DCFA. (for further details see
3
); D: as in C plus 100 µM
chelator practically does not affect the extent of lipid
peroxidation (Figure 8C). Similarly to H2O2 formation,
TBARS evolution has been evaluated also in the pres-
ence of 0.2 mM NADPH as reductant. In this case ADP-
iron complex causes a TBARS concentration increase
at 40 µM after 60 min of incubation being about 30 µM
just after 10 min. Nevertheless, as seen with H2O2
formation, the presence of 200 µM IC2 completely
prevents TBARS development (about 1.5 µM after 60
min; not shown). Therefore on the basis of these biologi-
cal experiments, we can conclude that the new chelator,
IC2, maintains the iron in a sufficiently stable complex
to prevent its redox cycle in biological systems.
2
2
2
Experimental Section). The final microsomal protein concen-
tration was about 0.08 mg/mL. Values represent the mean of
three separate experiments ( SE.
oxidant trigger (Figure 7C) to control microsomes
(Figure 7B). The concomitant presence of the chelator
at a concentration double that of Fe(III) ion completely
prevents its oxidative effect (Figure 7E). On the other
hand, when the concentration of the chelator is the same
as that of the metal, leaving 50 µM iron free, no
significant inhibition is observed (Figure 7D). This
suggests that the new chelator, IC2, is able to inactivate
the iron redox cycle only when all the metal is seques-
tered, excluding an aspecific antioxidant activity of the
molecule. The H2O2 production in this case is almost
the same of that produced by 100 µM iron because, at
these concentrations (that is 30-100 µM), the relation-
ships between dose-effect is not linear. Also worthy of
attention is the finding that the addition of 15 µM IC2
to the control microsomes (Figure 7F) completely abol-
ishes the spontaneous production of H2O2, the residual
fluorescence being due to ADP per se (Figure 7A). This
suggests that redox active iron contamination in liver
microsomal preparations is no more than 6-7 µM.
Measurements of H2O2 production have also been
performed in the presence of 0.2 mM NADPH as
biological reducing agent. In the presence of ADP-iron,
DCFA oxidation, in this case, is about double when
compared to that obtained without reductant. Neverthe-
less, the presence of 200 µM IC2 completely inhibits
again H2O2 formation (data not shown).
Con clu sion s
We previously suggested that, when suitably modified
to chelate iron, the chromanic ring, which characterizes
the polar head of vitamin E, could be useful as iron
sequestering agent both in vitro and in vivo. The first
molecule that we proposed (IC1) binds iron but does not
inhibit the iron redox cycle in vitro so that it is difficult
to think that it could be suitable for iron chelation
therapy. Therefore we undertook to modify IC1 so as to
increase its affinity for iron sufficiently to prevent its
redox cycle after its complexation. Our novelty is to
substitute the methyl group in IC1 with an aliphatic
alcoholic chain long enough to carry the terminal
alcoholic oxygen in the core of the iron coordination site
of the expected complex. The designed molecule is
unknown, and its synthesis has been reported for the
first time and described in detail in the present work.
The synthetic procedure appears to be suitable to
provide a few grams of the final product at a time. A
series of chemicophysical investigations, such as spec-
trophotometry, mass spectrometry, potentiometry, EPR,
and electrochemistry demonstrated that IC2 binds Fe-
(III) in a 2:1 molar ratio. IC2 acts as a tridentate and,
importantly, the affinity for iron (log â g 26) is high
enough to inhibit the iron redox cycle in the biological
environment despite the fact that it is lower than that
In another series of experiments using higher con-
centration of microsomes, we measured thiobarbituric
acid-reactive substance (TBARS) formation as an index
of lipid peroxidation, which should approximately be
proportional to H2O2 formation.
As can be seen in Figure 8, while the control rat liver
microsomes do not produce any significant TBARS
formation (Figure 8A), the addition of 100 µM iron by

New Hydroxychromene Iron Chelator
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 26 5783
of desferrioxamine and hydroxypyridones (log â ) 31
and log â ) 36, respectively). The electrochemical study
of the complexation between FeCl3 and IC2 shows that
the reduction potential of its complex is higher than that
of IC1 but lower than that of ferrioxamine, thus con-
firming that desferrioxamine has higher affinity than
IC2 for iron. In our opinion, this finding is very impor-
tant in view of the possible use of the new chelator as
an in vivo sequestering agent; it is possible, in fact, that
the undesirable side effects of the chelators already in
use are due to their excessive affinity for body iron.
It is really important that once iron is chelated it must
be not able to enter the redox cycle. The experiments
carried out on iron-dependent hydrogen peroxide forma-
tion and TBARS evolution clearly demonstrate that
when iron is chelated by IC2 it is not redox active in
biological systems such as hepatic microsomal suspen-
sions not even in the presence of biological reductans.
This means that, at least in the liver, an environment
suitable to reach the reduction potential of the IC2-
complexed iron does not exist.
mixture is stirred at room temperature. After 90 min, the
solvent is evaporated and the residual slurry is washed several
times with a petroleum ether/ethyl ether (1:1) solution. The
yield of this oxidation step was about 60%, as estimated by
GC-MS.
2
-(5-Hyd r oxyp en tyl)ch r om a n -4-on e. According to the
35
work of Kabbe and Widdig, about 15 g of crude products from
the former reaction are placed in a two-neck round-bottom
flask and 2.1 g (30 mmol) of pyrrolidine are added, drop by
drop, under magnetic stirring, keeping the solution at room
temperature by a water bath. A 4.1 g (30 mmol) amount of
2
′-hydroxyacetophenone were then added and the formation
of the condensation product was monitored by TLC (silica gel,
with petroleum ether/ethyl ether 3:1 as eluent). After about
9
0 min, the pyrrolidine is removed by vacuum evaporation,
and water is added to the mixture and extracted by ethyl ether.
After evaporation, an oily residue containing the 2-[5-(tet-
rahydropyran-2-yloxy)pentyl]chroman-4-one is obtained.
To remove the protecting pyranyl group, the residue is
dissolved in about 20 mL of ethanol with the addition of
concentrated HCl (25 mL). The formation of a compound with
lower R was monitored by TLC using ethyl acetate/petroleum
f
ether (30:70) as the eluent. After about 60 min at room
temperature, the ethanol is evaporated. The aqueous mixture
IC2 is a lipophilic compound and, like IC1,20 crosses
erythrocyte membranes, is absorbed in rodents after
both intraperitoneal and intragastric administration,
and is excreted within 24-48 h via the urine (unpub-
lished personal observations).
These observations strongly suggest that our expecta-
tion of improving the iron affinity of the chromanic ring
in IC1 by adding the aliphatic alcoholic chain in the 2
position of the IC1 has been completely vindicated. The
new molecule acts as a tridentate chelator and, impor-
tantly, inhibits the iron redox cycle after its chelation
in the in vitro biological systems. Therefore it is reason-
able to continue to study this compound especially in
view to demonstrate its ability to remove, in a redox
inactive form, the excess of body iron in the cases of iron
overload.
2 3
is neutralized with a K CO solution and then extracted with
ethyl ether. After evaporation of the organic phase, the residue
was chromatographed using a silica column with ethyl acetate/
petroleum ether (70:30) as the eluent, obtaining the final
1
product 2-(5-hydroxypentyl)chroman-4-one. Yield: 50%.
NMR (CDCl
H
3
): δ 6.95-7.85 (m, 4H, aromatic), 4.45 (tdd,1H,
J ) 7.9, J ) 8.20, J ) 6.25 Hz, H(2)), 3.67 (t, 2H, J ) 7.4 Hz,
H(5′)), 2.69 (dd, 2H, J ) 6.25, J ) 8.20 Hz, H(3)), 1.4, 1.9 (m,
8
H, H(1′), H(2′), H(3′), H(4′)).
-H yd r oxy-2-(5-h yd r oxyp en t yl)ch r om en -4-on e (IC
3
2
).
Following the procedure of Geissman and Armen, 10 mmol
2.3 g) of 2-(5-hydroxypentyl)chroman-4-one is dissolved in 70
3
6
(
mL of 90% ethanol at 45-50 °C, and 30 mmol of isoamyl nitrite
and 100 mmol of concentrated HCl are added, drop by drop
contemporaneously. The reaction is carried out until the
starting compound disappears in TLC (about 90 min). Then
2 3
the reaction mixture is neutralized with K CO -saturated
water solution and extracted with ethyl acetate. The oily
residue is fractionated using silica gel column chromatography
and eluting with ethyl acetate/petroleum ether (70:30). After
evaporation of the solvent a yellow solid is recovered. Several
crystallizations from acetone at first and then from ethyl ether
allowed us to obtain a gas chromatographically pure, colorless,
Exp er im en ta l Section
1
,6-Hexanediol was from Fluka, 2′-hydroxyacetophenone,
3
,4-dihydro-2H-pyran, pyridinium chlorochromate, pyrrolidine,
and isoamyl nitrite were supplied by Aldrich, and ADP and
dichlorofluorescein were from Sigma. Others salts and solvents
were of analytical grade. The NMR analyses were performed
using a Bruker AC 200 MHz instrument; melting points were
detected by a Mettler TA instrument DSC 12 E. Elementary
analysis was evaluated by a 240 Perkin-Elmer analyzer. Gas
chromatography-mass spectrometry (GC-MS) experiments
were carried out by a Varian Saturn 2000 ion trap instrument
coupled with a gas chromatograph Varian Star 3400 CX
equipped with a column J &W DB5-MS (30m × 0.25 mm, film
thickness 0.25 µm) by using ultrapure helium as a carrier gas.
Fluorimetric analysis were performed by a 650-10S Perkin-
Elmer fluorimeter.
solid of IC
found 67.78; H, calcd 6.45, found 6.74. Mp 112-114 °C; H
NMR (CDCl ): δ 7-3-8.3 (m, 4H, aromatic), 3.68 (t, 2H, J )
6.3 Hz, H(5′)), 2.88 (t, 2H, J ) 7.4 Hz, H(1′)), 1.5, 1.9 (m, 6H,
H(2′), H(3′), H(4′)); 13C NMR (CDCl
2 16 4
. Yield about 50%. Anal. (C14H O ) C, calcd 67.74,
1
3
3
): δ 25.36, 26.43, 28.83,
35.35, 62.66, 118.07, 121.50, 124,28, 125.46, 132.95, 138.25,
+
152.34, 155.57, 172.250; m/z (EI) 248 (M , 15), 189 (75), 176
(100); HRMS Calcd for C14
248.1049.
16 4
H O m/z 248.1049, found: m/z
X-r a y Cr ysta llogr a p h y. Single crystals of IC2 were ob-
tained by dissolving a few milligrams of IC2 in acetone and
allowing the solution to concentrate at room temperature. A
colorless single crystal of approximate dimensions 0.15 × 0.35
× 0.15 mm was submitted to X-ray analysis using a Siemens
P4 four-circle diffractometer with graphite monochromated
Mo-KR radiation (λ ) 0.71069 Å). Lattice parameters were
determined by least-squares refinement on 46 randomly
selected and automatically centered reflections. The ω/2Θ scan
technique was used and the data were collected in the 2e Θe
6
-(Tetr a h yd r op yr a n -2-yloxy)h exa n -1-ol. A 100 g (0.5
mol) amount of 1,6-hexanediol are dissolved in 38 mL (0.5 mol)
of 2,4-dihydropyran and maintained at room temperature in
a water bath After adding about 50 mg of p-toluensulfonic acid,
the mixture is allowed to react for about 2 h under magnetic
2 3
stirring. An aqueous solution of K CO is then added, and the
reaction products are extracted with diethyl ether. Three
different compounds are detected by GC-MS, and these are
the unreacted 1,6-hexanediol, 1,6-dioxopyranylhexane, and
about 70% of 6-(tetrahydropyran-2-yloxy)hexan-1-ol.
26° scan range. Crystal system: monoclinic; space group: P2 /c
1
(no. 14); a ) 12.335(2), b ) 5.4090(10), c ) 18.721(2) Å, â )
3
3
95.97(1)°, V ) 1242.3(3) Å , Z ) 4, D ) 1.327 g/cm . 3400
c
reflections were collected at 22 °C of which 2445 are unique
(Rint) 0.06). Absorption correction by using the Ψ method was
applied. The structure was solved by direct methods imple-
mented in the SHELXS-97 program.38 The refinement was
6
-(Tetr a h yd r op yr a n -2-yloxy)h exa n a l. According to the
3
4
work of Kasmai et al., a solution of 15 g of pyridinium
chlorochromate in 400 mL of dichloromethane is added to 20
g of the crude mixture from the former reaction, and the
2
carried out by full-matrix anisotropic least-squares on F for

5
784 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 26
Ferrali et al.
all reflections for non-H atoms by using the SHELXL-97
homogenized in 40 mL of 50 mM Tris-maleate buffer pH 7.4
(TM) containing 150 mM KCl and 100 µM EDTA to prevent
spontaneous oxidations. The homogenate was centrifuged at
10000 rpm for 20 min. The supernatant was collected and
centrifuged at 80000 × g for 40 min.
3
9
program. The final refinement converged to R
1
) 0.068, wR
2
)
0.115 for I > 2σI, goodness-of-fit ) 0.982. Min. max height
-3
in last ∆F map of -0.184 and 0.195 eÅ . Full crystallographic
details will be given elsewhere.
-F e 3 Com p lex. Mass
+
Ma ss Sp ectr om etr y of th e IC
2
The microsomal sediment was resuspended in 36 mL of 50
mM TM. This suspension was further diluted in TM buffer at
the concentration of microsomes equivalent to 100 mg of liver
(about 0.25 mg of protein) per 10 mL. Solutions of 5 mM ADP
and 5 mM ADP plus 200 µM ferric chloride, respectively, were
prepared in the same buffer. Two separate solutions of 20 mM
spectrometry (MS) and MS/MS analyses were performed with
a VG 70-250S instrument (Micromass, Manchester) using
electron ionization (EI) and liquid secondary ion mass spec-
trometry (LSIMS) techniques; electrospray (ESI) measure-
ments were carried out on a LCQ DECA mass spectrometer
IC
2 2
ethanol were prepared. H O
2
and 40 mM dichloro fluorescein diacetate43 (DCFA) in
formation was detected for 5 min
(ThermoFinnigan). One drop of a solution, obtained by dis-
solving a few micrograms of IC and ferric chloride in metha-
2
nol, was put on the stainless steel tip of the LSIMS probe,
mixed with glycerol as a matrix, and exposed to the Cs+ beam
for the desorption. In ESI experiments, methanol solutions in
the micromolar range were injected by direct infusion at 5 µL
in a fluorimeter cuvette containing 0.5 mL of microsomal
suspension in buffer, 1 mL of Fe-ADP solution, and 5 µL of
2
DCFA, to which different additions of IC were made. Control
measurements adding 1 mL of 5 mM ADP alone instead of
ADP-Fe were also performed.
In h ibition of Ir on -Dep en d en t Micr osom a l Lip id P er -
oxid a tion . MC were prepared as described above and resus-
pended in TM buffer at the concentration of MC equivalent to
-
1
min . The spray voltage and the temperature were +4.5 kV
and 180 °C, respectively.
Sp ectr op h otom etr ic Mea su r em en ts. Fresh stock solu-
tions of 10 mM IC
chloride in water were made. The spectrum of 200 µM IC
Tris-maleate buffer pH 7.4 was detected between 300 and 400
nm. Then increasing amounts of FeCl (50, 70, 100, and 200
2
in ethanol and 2 mM anhydrous ferric
4
g of liver in 36 mL. The suspension was then adjusted, after
2
in
protein measurements, at 3 mg of protein per mL. Solutions
of iron-ADP and IC
2
were prepared as above. Amounts of 0.6
were pipetted into separate flasks. The
3
and 1.2 µmol of IC
2
µM, respectively) were added and each new spectrum detected
by a DU 640 Beckman spectrophotometer.
P oten tiom etr ic Mea su r em en ts. Titrations were per-
formed at a constant temperature (25 °C) with a digital PHM-
solvent was dried and the residue dissolved in 3 mL of the
MC suspension with vigorous shaking, after which 3 mL of
the iron-ADP solution was added. Control microsomes were
incubated adding 3 mL of buffer instead of ADP-Fe solution.
All the flasks were incubated for 60 min at 37 °C. At the
indicated times, aliquots were withdrawn to measure TBARS
formed.44
8
4 Radiometer potentiometer, equipped with a pHG211 High
pH glass electrode and a ref 201 reference electrode (Radiom-
eter). A Metrohm Multidosimat piston buret connected to a
computer (Olivetti M20) controlled and automatically recorded
potentiometric data (emf reading, mV, in relation to volume,
Su p p or tin g In for m a tion Ava ila ble: Crystal data, atomic
coordinates, anisotropic dispacement parameters, bond lengths
and angles (Tables 1S-4S) of IC . This material is available
2
free of charge via the Internet at http://pubs.acs.org.
3
9
2
mL, of added titrant). A 0.1169 mmol amount of IC was
dissolved in 95 mL of 0.1 M NaCl, and 3 mL of 0.1 M NaOH
was added to help its dissolution under magnetic stirring.
Titration with 0.1 M HCl was performed under a presaturated
nitrogen stream flowing over the surface of the solution to
avoid contamination with CO from the outside atmosphere.
2
At the end of the titration, a measured quantity (0.0460 mmol)
Ack n ow led gm en t. We thank MURST, Rome, Italy
and Consorzio Interuniversitario di Ricerca in Chimica
dei Metalli nei Sistemi Biologici for financial support.
of FeCl
ratio IC
3
was added to the solution in order to have a molar
/Fe(III) of 2.5. The obtained dark-violet solution was
2
Refer en ces
titrated with a standardized NaOH solution (0.1078 M).
Potentiometric data were processed with the Superquad
(
1) Halliwell, B.; Gutteridge, J . M. C. Oxygen Free Radicals and
Iron in Relation to Biology and Medicine: Some Problems and
Concepts. Arch. Biochem. Biophys. 1986, 246, 501-514.
2) Ryan, T. P.; Aust, S. D. The Role of Iron in Oxygen-Mediated
Toxicities. Crit. Rev. Toxicol. 1992, 22, 119-141.
4
0
program to calculate the basicity and stability constants.
EP R Sp ectr a . X-band EPR spectra were recorded on the
BRUKER 200D-SRC spectrometer operating at 9.44 GHz. The
microwave frequency was tested with a XL Microwave Fre-
quency Counter 3120 and the external magnetic field H
calibrated by using a DPPH powder sample (gDPPH ) 2.0036).
The spectrometer was equipped with a low temperature device
(
(
3) Minotti, G.; Aust, S. D. Redox Cycling of Iron and Lipid
Peroxidation. Lipids 1992, 27, 219-226.
(4) Dean, R. T.; Fu, S.; Stocker, R.; Davies, M. J . Biochemistry and
Pathology of Radical-mediated Protein Oxidation. Biochem. J .
1
997, 324, 1-18.
(Bruker ER 401 VT). The solid-state and solution samples were
(
5) Linn, S. DNA Damage by Iron and Hydrogen Peroxide in vitro
and in vivo. Drug Metab. Rev. 1998, 30, 313-326.
placed into calibrated quartz capillaries permanently posi-
tioned in the EPR resonance cavity. The X-band EPR averaged
parameters were evaluated on the background line of the first
derivative line shape of the different experimental spectra
recorded at different temperatures (298 and 100 K) for both
the solid-state and solution samples.
(
6) Pietrangelo, A. Metals, Oxidative Stress, and Hepatic Fibrogen-
esis. Semin. Liver Dis. 1996, 16, 13-30.
(7) Hershko, C.; Graham, G.; Bates, G. W.; Rachmilewitz, E. A. Non-
Specific Serum Iron in Thalassemia: An Abnormal Serum
Fraction of Potential Toxicity. Brit. J . Haematol. 1978, 40, 255-
2
63.
Geom etr y Op tim iza tion of F eIII(IC
semiempirical method, as implemented in Hyperchem, has
been used to optimize the structure of the Fe (IC ) complex.
2 2
) Com p lex. PM3
2 2
(8) Dexter, D. T.; J enner, P. Alterations in Iron in Neurodegenera-
tive Disorders: Implications and Possible Therapeutic Agents.
In Mineral and Metal Neurotoxicology; Yasui, M. et al., Eds.;
CRC Press: Boca Raton, 1997; pp 365-378.
4
1
42
III
The staring geometry of the ligand was that obtained by the
X-ray study. The structure was fully optimized without any
symmetry constraint. The complex has an high spin state,
therefore the unrestricted formalism has been adopted for a
system with S ) 5/2. Calculations have been performed for
both the fac- and mer-isomers.
(
9) Weinberg, E. D. Development of Clinical Methods of Iron
Deprivation for Suppression of Neoplastic and Infectious Dis-
eases. Cancer Invest. 1999, 17, 507-513.
(10) Galey, J . B. Potential Use of Iron Chelators Against Oxidative
Damage. Adv. Pharmacol. 1997, 38, 1167-203.
(
11) Ferrali, M.; Signorini, C.; Sugherini, L.; Pompella, A.; Lodovici,
M.; Caciotti, B.; Ciccoli, L.; Comporti, M. Release of Free, Redox-
Active Iron in the Liver and DNA Oxidative Damage Following
Phenylhydrazine Intoxication. Biochem. Pharmacol. 1997, 53,
1743-1751.
Volta m m etr y. Materials and apparatus for cyclic voltam-
2
0
metry have been described elsewhere. Potential values were
referenced to the saturated calomel electrode (SCE). Under
the present experimental conditions the one-electron oxidation
(
12) de Valk, B.; Marx, J . J . M. Iron, Atherosclerosis, and Ischemic
Heart Disease. Arch. Int. Med. 1999, 26, 1542-1548.
13) Giardina, P. J .; Grady, R. W. Chelation Therapy in â-Thalas-
semia: The Benefits and Limitations of Desferrioxamine. Semin.
Hematol. 1995, 32, 304-312.
0
′
of ferrocene occurs at E ) +0.34V, vs SCE.
In h ibition of Ir on -Dep en d en t H F or m a tion in Mi-
cr osom a l Su sp en sion . A 4 g amount of rat liver was
(
2
O
2

New Hydroxychromene Iron Chelator
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 26 5785
(
14) Olivieri, N. F.; Brittenham, G. M. Iron-Chelating Therapy and
the Treatment of Thalassemia. Blood 1997, 89, 739-761.
15) Tilbrook, G. S.; Hider, R. C. Iron Chelators for Clinical Use. In
Metal Ions in Biological Systems; Sigel, A.;. Sigel, H., Eds.;
Marcel Dekker: New York, 1998; Vol. 35, pp 691-730.
16) Richardson, D. R.; Ponka, P. Pyridoxal Isonicotinoyl Hydrazone
and its Analogues: Potential Orally Effective Iron-Chelating
Agents for the Treatment of Iron Overload Disease. J . Lab. Clin.
Med. 1998, 131, 306-315.
(30) Heyrovsky, I.; Kuta, J . Reversible Processes Controlled by
Diffusion of Complex Ions. Principles of Polarography; Academic
Press: New York, 1966; pp 147-160.
(
(
31) Kolthoff, I. M.; Lingane J . J . Iron, Cobalt, Nichel and Platinum
Metals. Polarography; Interscience Publishers: New York, 1952;
pp 475-492.
(
(
32) Evers, A.; Hancock, R. D.; Martell, A. E.; Motekaitis, R. J . Metal
Ion Recognition in Ligands with Negatively Charged Oxygen
Donor Groups. Complexation of Fe(III), Ga(III), In(III), Al(III),
and Other Highly Charged Metal Ions. Inorg. Chem. 1989, 28,
2189-2195.
(
17) Diav-Citrin, O.; Koren, G. Oral Iron Chelation with Deferiprone.
Pediatr. Clin. North Am. 1997, 44, 235-247.
(
18) Cregg, L.; Hebbel, R. P.; Miller, W.; Solovey, A.; Selby, S.;
Enright, H. The Iron Chelator L1 Potentiates Oxidative DNA
Damage in Iron-Loaded Liver Cells. Blood 1998, 32, 632-638.
19) Hoffbrand, A. V. Iron Chelation Therapy: The Need for Orally
Active Drugs. J . Lab. Clin. Med. 1998, 131, 290-291.
20) Ferrali, M.; Donati, D.; Bambagioni, S.; Fontani, M.; Giorgi, G.;
Pietrangelo, A. 3-Hydroxy-(4H)-benzopyran-4-ones as Potential
Iron Chelating Agents In Vivo. Bioorg. Med. Chem. 2001, 9,
(
33) Fong, K. L.; McCay, P. B.; Poyer, J . L. Evidence that Peroxidation
of Lysosomal Membranes is Initiated by Hydroxyl Free Radicals
Produced during Flavin Enzyme Activity. J . Biol. Chem. 1973,
(
(
2
48, 7792-7797.
(
34) Kasmai, H. S.; Mischke, S. G.; Blake, T. J . 18-Crown-6 Com-
plexes of n-Butylammonium and Pyridinium Chlorochromates.
Mild and Selective Oxidizing Agents for Alcohols. J . Org. Chem.
1995, 60, 2267-2270.
3
041-3047.
(
21) Abragam, A.; Bleaney, D. Electron Paramagnetic Resonance of
Transition Ions; Dover Publ.: New York, 1970.
(
35) Kabbe, H. J .; Widdig, A. Synthesis and Reactions of 4-Chro-
manones. Angew. Chem., Int. Ed. Engl. 1982, 21, 247-256.
36) Geissman, T. A.; Armen, A. The Rearrangement of 2-Acethyl-
and 2-Benzoylcoumarone Oxime p-Toluenesulfonates. J . Am.
Chem. Soc. 1955, 77, 1623-1627.
(
22) Mabbs, F. E.; Collison, D. Electron Paramagnetic Resonance of
d Transition Metal Compounds, Elsevier Publ.: New York, 1992;
Vol. 16.
23) Drago, S. Electron Paramagnetic Resonance Spectra of Transi-
tion Metal ion Complexes. Physical Methods for Chemists;
Saunders College Publishing: New York, 1992; pp 559-603.
24) Field, D.; George, A. V.; Laschi, F.; Malouf, Y.; Zanello P.
Electrochemistry of Acetylide Complexes of Iron. J . Org. Chem.
(
(
(37) Sheldrick, G. M. SHELXS-97, Rel. 97-2, University of G o¨ ttin-
(
gen, 1997.
(38) Sheldrick, G. M. SHELXL-97, Rel. 97-2, University of G o¨ ttin-
1
992, 435, 347-56.
gen, 1997.
(
25) Cundari, T. R.; Deng, J . PM3 Analysis of Transition-Metal
Complexes. J . Chem. Inf. Comput. Sci. 1999, 39, 376-381.
26) Harris, W. R.; Carrano, C. J .; Cooper, S. R.; Sofen, S. R.; Avdeef,
A. E.; McArdle, J . V.; Raymond, K. N. Coordination Chemistry
of Microbial Iron Transport Compounds. 19. Stability Constants
and Electrochemical Behaviour of Enterobactin and Model
Complexes. J . Am. Chem. Soc. 1979, 101, 6097-6104.
(
(
(
39) Casolaro, M.; Bignotti, F.; Ferruti, P. Amphoteric Poly(amido-
amine) Polymers Containing the Ethylendiamine-N,N′-Diacetic
Acid Moiety. Stability of Copper(II) and Calcium(II) Chelates.
J . Chem. Soc., Faraday Trans. 1997, 3663-3668.
40) Gans, P.; Sabatini, A.; Vacca, A. Superquad: An Improved
General Program for Computation of Formation Constants from
Potentiometric Data. J . Chem. Soc., Dalton Trans. 1985, 1195-
1200.
41) Stewart, J . J . P. Optimisation of parameters for semiempirical
methods. I. Method. J . Comput. Chem. 1989, 10, 209-220.
42) Hyperchem 7.0, Hypercube Inc., Gainesville, FL.
43) Keston, A. S.; Brandt, R. The Fluorimetric Analysis of Ultramicro
Quantities of Hydrogen Peroxide. Anal. Biochem. 1965, 11, 1-5.
44) Comporti, M.; Hartman, A.; Di Luzio, N. R. Effects of in vivo
and in vitro Ethanol Administration on Liver Lipid Peroxidation.
Lab. Invest. 1967, 16, 616-624.
(
(
27) Nesset, M. J . M.; Shokhirev, N. V.; Enemark, P. D.; J acobson,
S. E.; Walker, F. A. Models of the Cytochromes. Redox Properties
and Thermodynamic Stabilities of Complexes of Hindered Iron-
(III) and Iron(II) Tetraphenylporphyrinates with Substituted
Pyridines and Imidazoles. Inorg. Chem. 1996, 35, 5188-5200.
28) Spasojeviæ, I.; Armstrong, S. K.; Brickman, T. J .; Crumbliss,
A. L. Electrochemical Behaviour of the Fe(III) Complexes of
Cyclic Hydroxamate Siderophores Alcaligin and Desferrioxamine
E. Inorg. Chem. 1999, 38, 449-454.
(
(
(
(
(
29) Serratrice, G.; Galey, J .-B.; Aman, E. S.; Dumats, J . Iron(III)
Complexation by New Aminocarboxylate Chelators. Thermody-
namic and Kinetic Studies. Eur. J . Inorg. Chem. 2001, 471-
4
79.
J M021022U