Targeting the lysosome: Fluorescent iron(III) chelators to selectively monitor endosomal/ lysosomal labile iron pools

Article abstract of DOI:10.1021/jm8001247

Iron-sensitive fluorescent chemosensors in combination with digital fluorescence spectroscopy have led to the identification of a distinct subcellular compartmentation of intracellular redox-active "labile" iron. To investigate the distribution of labile iron, our research has been focused on the development of fluorescent iron sensors targeting the endosomal/lysosomal system. Following the recent introduction of a series of 3-hydroxypyridin-4-one (HPO) based fluorescent probes we present here two novel HPO sensors capable of accumulating and monitoring iron exclusively in endosomal/lysosomal compartments. Flow cytometric and confocal microscopy studies in murine macrophages revealed endosomal/lysosomal sequestration of the probes and high responsiveness toward alterations of vesicular labile iron concentrations. This allowed assessment of cellular iron status with high sensitivity in response to the clinically applied medications desferrioxamine, deferiprone, and deferasirox. The probes represent a powerful class of sensors for quantitative iron detection and clinical real-time monitoring of subcellular labile iron levels in health and disease.

Full text of DOI:10.1021/jm8001247

J. Med. Chem. 2008, 51, 4539–4552
4539
Targeting the Lysosome: Fluorescent Iron(III) Chelators To Selectively Monitor Endosomal/
Lysosomal Labile Iron Pools
Sarah Fakih,*,† Maria Podinovskaia,§ Xiaole Kong,‡ Helen L. Collins,# Ulrich E. Schaible,§ and Robert C. Hider‡
DiVision of Pharmaceutical Sciences, King’s College London, Franklin-Wilkins-Building, Stamford Street, London SE1 9NH, U.K., Department
of Inorganic Chemistry, Georg-August UniVersity of Go¨ttingen, Tammannstrasse 4, 37077 Go¨ttingen, Germany, Department of Infectious and
Tropical Diseases, London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, U.K., and Department of Infectious
Diseases, King’s College London, Hodgkin-Building, Guy’s Campus, London SE1 1UL, U.K.
ReceiVed February 6, 2008
Iron-sensitive fluorescent chemosensors in combination with digital fluorescence spectroscopy have led to
the identification of a distinct subcellular compartmentation of intracellular redox-active “labile” iron. To
investigate the distribution of labile iron, our research has been focused on the development of fluorescent
iron sensors targeting the endosomal/lysosomal system. Following the recent introduction of a series of
3-hydroxypyridin-4-one (HPO) based fluorescent probes we present here two novel HPO sensors capable
of accumulating and monitoring iron exclusively in endosomal/lysosomal compartments. Flow cytometric
and confocal microscopy studies in murine macrophages revealed endosomal/lysosomal sequestration of
the probes and high responsiveness toward alterations of vesicular labile iron concentrations. This allowed
assessment of cellular iron status with high sensitivity in response to the clinically applied medications
desferrioxamine, deferiprone, and deferasirox. The probes represent a powerful class of sensors for quantitative
iron detection and clinical real-time monitoring of subcellular labile iron levels in health and disease.
Introduction
a highly potent source of intracellular oxidative potential.14
Consequently, exposure to intracellular ROS renders these or-
ganelles particularly susceptible to oxidative damage. Organelle
rupture resulting from intralysosomal Fenton reactions during
oxidative stress situations has been associated with enhanced DNA
damage due to the sudden release of lysosomal labile iron into the
cytoplasm.15,16 Permeabilization of mitochondrial membranes by
released lysosomal enzymes most likely induces a loss of mito-
chondrialmembranepotential,irreversiblyleadingtocelldeath.14,17–19
Administration of the lipophilic chelator salicylaldehyde isonico-
tinoyl hydrazone (SIHa) has been shown to provide protection
against oxidative stress-induced cell death by virtue of its strong
iron chelating properties.20,21 Correspondingly, the predominantly
lysosomal high-affinity iron chelator desferrioxamine (DFO)
sustains lysosomal stability during oxidative stress, preventing
organelle rupture, DNA damage, loss of mitochondrial membrane
potential, and cell death.16,18,19
Most methods to characterize intracellular labile iron distribu-
tion seriously perturb the extent and nature of the labile iron
pool, rendering them inapplicable to living cells. In combination
with noninvasive and high resolution confocal imaging, fluo-
rescent chemosensors have proven to be highly sensitive tools
to investigate the metabolism of labile metal pools in individual
cells with subcellular resolution.13,22–27 To the best of our
knowledge, a fluorescent probe selectively accumulating inside
the endosomes/lysosomes to allow monitoring of labile iron has
not been available to date. To investigate the chemical nature
of intracellular labile iron pools, we recently introduced a series
of coumarin labeled fluorescent 3-hydroxypyridin-4-one (HPO)
based probes for imaging and quantification of cytosolic labile
iron.28–30 Here, we report the synthesis and functional charac-
terization of two novel fluorescein labeled 3-hydroxypyridin-
4-one based fluorescent probes to monitor endosomal/lysosomal
Iron is the most abundant transition metal in cellular systems
holding an outstanding biological importance because of its
presence in the structures of numerous enzymes and proteins.1,2
By virtue of its facile redox chemistry and its high affinity for
oxygen, iron is critically involved in electron transfer reactions
as well as oxygen transport. In contrast to iron bound to proteins,
the intracellular labile iron pool has a considerable toxic
potential. In the presence of reactive oxygen species (ROS)
loosely bound iron is able to redox cycle between its most stable
oxidation states Fe(II)/Fe(III), catalyzing the formation of
oxygen-derived free radicals such as the hydroxyl radical via
the Fenton reaction.3,4 These highly reactive radical species are
capable of interacting with most types of biological material
including sugars, lipids, proteins, and nucleic acids. The resulting
injurious processes covering lipid peroxidation, protein oxida-
tion, DNA/RNA oxidation, and DNA lesions take a critical lead
in the development of pathology in diseases such as hepatitis,
hemochromatosis, liver cirrhosis, cancer, and neurodegen-
eration.3,5–12
Labile iron is generally considered to be cytosolic, its uptake
being driven by receptor-mediated endocytosis of transferrin-
bound iron. However, increasing experimental evidence assigns
only a minor fraction of the total cellular redox-active iron to
the cytosol. Rapid transit of the metal to dominant sites of iron
sequestration and incorporation into proteins can relocate
cytosolic iron to distinct subcellular compartments.13
Considerable accumulation of redox-active iron in the lysosomal
system provided by high turnover degradation of endocytosed as
well as autophagocytosed iron-containing macromolecules creates
* To whom correspondence should be addressed. Phone: +49 (0)551
39 3006. Fax: +49 (0)551 39 3363. E-mail: sarah.fakih@chemie.uni-
goettingen.de.
†
a Abbreviations: HPO, 3-hydroxypyridin-4-one; SIH, salicylaldehyde
isonicotinoyl hydrazone; DFO, desferrioxamine; BMM, bone marrow
macrophages; DFR, deferasirox; DFP, deferiprone; DexTR, dextran Texas
Red; FeDex, iron(III) dextran.
Georg-August University of Go¨ttingen.
London School of Hygiene and Tropical Medicine.
Division of Pharmaceutical Sciences, King’s College London.
Department of Infectious Diseases, King’s College London.
§
‡
#
10.1021/jm8001247 CCC: $40.75
2008 American Chemical Society
Published on Web 07/15/2008

4540 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 15
Scheme 1. Synthesis of the Fluorescent Probe 9
Fakih et al.
Scheme 2. Synthesis of the Fluorescent Probe 20
iron. We analyzed the physicochemical properties of these
probes regarding their sensitivity toward ferric iron. Intracellular
distribution and responsiveness to drug mediated modifications
of iron concentrations have been investigated in murine bone
marrow derived macrophages using flow cytometry and confocal
microscopy.
Intracellular distribution properties of HPO-based compounds
have been found to be regulated mainly by the chemical nature
of the substituent at the nitrogen atom in the 1-position. Probe
9 featuring an unsubstituted nitrogen atom in the 1-position was
designed to function as a nonspecific sensor with no selective
intracellular distribution pattern. Probe 20 on the other hand
was designed as a lysosomotrophic sensor based on the structural
incorporation of a basic 1-dimethylethylamine substituent.
Because of the acidic nature of endosomal/lysosomal compart-
ments, probes featuring basic moieties undergo rapid protonation
Results and Discussion
Chemistry. The structures of the HPO-based fluorescent iron
chemosensors 9 and 20 are illustrated in Schemes 1 and 2.

Endosomal/Lysosomal Fluorescent Iron Chelators
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 15 4541
following their uptake into the vesicles. The resulting charged
molecule species are trapped inside the vesicles because of their
relative membrane impermeability, providing for an efficient
intraorganelle enrichment with the compound.31
of the fluorescent probes 9 and 20 induces quenching of the
fluorescein fluorophor unit depending on the available Fe(III)
concentration. By virtue of the excellent iron selectivity of the
HPO binding moiety, interfering signaling by other biologically
active metals can be mostly precluded. The HPO based
fluorescent sensors specifically do not suffer from interference
from zinc, calcium, and magnesium.29,33 Appreciable interfer-
ence has been observed in the presence of Cu(II). Yet copper
is known to be chaperoned very tightly in mammalian cells by
copper-binding proteins and other ligands. Consequently, the
intracellular labile copper pool is vanishingly small (less than
a single solvated copper ion per cell) and can in practice be
neglected.34 To confirm HPO characteristic Fe(III) selectivity
for the presented fluorescein coupled sensors, representative
interaction of 9 with Ca(II), Mg(II), Cu(II), Ni(II), Zn(II),
Mn(II), and Co(II) was recorded and confirmed fluorescence
insensibility toward the tested metals except for the expected
quenching effect (32.3%) in the case of Cu(II) (data see
Supporting Information).
Illustrated in Figure 1 is a representative fluorescence
emission profile of probe 20 titrated with increasing Fe(III)
concentrations as well as a comparative plot of the averaged
quenching capacities of both probes. Owing to the fluorescein
moiety, characteristic excitation maxima for 9 and 20 are at
494 and 491 nm while emission maxima are at 512 and 514
nm. All measurements were carried out at pH 7.4 in Chelex-
treated MOPS buffer to avoid interfering quenching signals by
iron contamination. Titration of the iron-free buffered probe
solutions with increasing Fe(III) concentrations showed an
overall quenching capacity of 77% for 9 and a notably higher
quenching capacity of 96% for 20. Both quenching profiles
stabilize at an iron concentration of ∼1 µM, indicating
completed iron complex formation. In view of the applied 3
µM sensor concentration, the resulting 3:1 probe/metal molar
ratio at fluorescence stabilization suggests binding of three
bidentate HPO chelators, required to form the anticipated fully
coordinated octahedral iron(III) complex.
Syntheses of compounds 2-5 and 11-14 were based on
procedure described previously.28 Scheme 1 illustrates the
synthetic route employed for the preparation of the 5-fluorescein
labeled probe 9 starting from commercially available 3-hydroxy-
2-methyl-4H-pyran-4-one (maltol, 1). Protection of the 3-hy-
droxyl group was achieved by methylation using dimethyl
sulfate under basic conditions and was followed by exchange
of the ring oxygen with nitrogen via reaction with aqueous
ammonia solution to obtain 3-methoxy-2-methylpyridin-4(1H)-
one 3. Aminomethylation of the activated 5-position was then
accomplished by Mannich reaction conditions using formalde-
hyde and N-benzylmethylamine. Subsequent removal of the
benzyl protection group via hydrogenation yielded the func-
tionalized 3-methoxy-2-methyl-5-[(methylamino)methyl]pyridin-
4(1H)-one 5 furnished with a secondary amine in the 5-position,
which is suitable for fluorophor coupling. The coupling of
fluorescein to the protected pyridinone 5 proceeded via preced-
ing activation of commercially available 5(6)-carboxyfluorescein
6 to yield 7 using N-hydroxysuccinimde and the peptide
coupling reagent dicyclohexylcarbodiimide (DCCI). Following
fluorophor coupling, deprotection of the methyl protected
3-hydroxyl group of 8 was achieved by reaction with BCl3,
furnishing the fully functional iron(III) sensitive fluorescence
chemosenor 9.
The fluorescein labeled probe 20 was prepared following a
multistep synthesis strategy summarized in Scheme 2. Starting
from 5-hydroxy-2-(hydroxymethyl)-4H-pyran-4-one (kojic acid,
10), the 2-hydroxymethyl residue was converted into a methyl
group via preceding chlorination using thionyl chloride (11)
followed by reduction using zinc powder and concentrated
hydrochloric acid to yield 5-hydroxy-2-methyl-4(1H)-pyran-4-
one 12. Reaction with formaldehyde under basic conditions
resulted in a hydroxymethyl substituent in the 6-position (13).
After benzylation of the 3-hydroxyl group (14), the ring oxygen
was exchanged with dimethylethylenediamine to yield 16
following intermediate formation of the 2-tetrahydropyrane
protected 3-(benzyloxy)-6-methyl-2-[(tetrahydro-2H-pyran-2-
yloxy)methyl]-4H-pyran-4-one 15. Conversion of 16 to the
corresponding phthalimido derivative proceeded via the Mit-
sunobu reaction using phthalimide, triphenylphosphine, and
diisopropyl azodicarboxylate. The quaternary phosphonium salt
forming upon reaction of triphenylphosphine with diisopropyl
azodicarboxylate is protonated by the nucleophilic phthalimide
and activates the 2-hydroxyl substituent through formation of
an alkoxyphosphonium salt. Nucleophilic substitution of the
phosphonium moiety by the deprotonated phthalimide furnishes
the desired phthalimide derivative 17.32 Subsequent hydrazi-
nolysis of 17 yielded 2-(aminomethyl)-3-(benzyloxy)-1-[2-
(dimethylamino)ethyl]-6-methylpyridin-4(1H)-one 18 function-
alized in the 2-position with a primary amine for fluorophor
coupling. Coupling of the activated 5(6)-carboxyfluorescein
N-hydroxysuccinimide ester 7 with 18 furnished the protected
probe precursor 19, which was converted to the functional
iron(III) chemosensor 20 by debenzylation of the 3-hydroxyl
group via hydrogenolysis.
Dequenching kinetics of Fe(III) complex solutions of 9 and
20 were investigated using the HPO-based nonfluorescent
chelator 1,2-diethyl-3-hydroxypyridin-4(1H)-one (CP94) and the
high-affinity iron chelator SIH, exhibiting pFe3+ values of 19.7
(CP94)35 and 50 (SIH),36 respectively. Illustrated in Figure 2
is the representative time-dependent dequenching profile of the
Fe(20)3 complex employing each dequencher, expressed as a
percentage of the initial sensor fluorescence in the absence of
iron (set at 100%). By virtue of its higher Fe(III) binding
constant, SIH rapidly restores 95% of the initial fluorescence
of 20 (9, 92%, data not shown) while 87% recovery can be
achieved applying CP94 (9, 87%, data not shown). Optical
properties of both sensors are summarized in Table 1.
pH Sensitivity. For biological application of the sensors and
data analysis on the basis of specific fluorescence emission
intensities, it is necessary to consider the intrinsic pH sensitivity
of fluorescein. In solution fluorescein can in principle exist in
four different prototropic forms, covering the cation, the neutral
molecule, the monoanion, and the dianion.37 Because of complex
prototropic equilibria, spectral properties of fluorescein are
particularly sensitive to pH. In the pH range 6-10, species
distribution is dominated by the monoanion-dianion equilib-
rium; the cation and neutral species virtually do not exist.38,39
As primarily the dianion accounts for the observed fluorescence
intensity at pH 7.4,38,40 the fluorescence emission spectrum
shows a significant decrease of emission intensity at low pH.
A representative pH titration of 20 in Chelex-treated buffer
Optical Characterization. The applicability of a fluorescent
chemosensor to monitor the metabolism and translocation
dynamics of intracellular labile cations is essentially based on
the responsiveness of its fluorophor unit toward the presence
of the targeted ion. Binding of iron to the HPO chelation moiety

4542 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 15
Fakih et al.
Figure 2. Comparative plot of the dequenching kinetics of the fully
quenched Fe(20)3 complex in Chelex-treated MOPS buffer (50 mM)
at pH 7.4. The iron complex was prepared from a 3 µM sensor solution
and a 3 µM NTA-buffered iron(III) solution (iron/NTA, 1:5 molar ratio).
Application of the nonfluorescent iron(III) chelators CP94 (1 mM) and
SIH (100 µM) showed maximal dequenching capacities of 87% and
95%, respectively, indicated by stable plateau formation after ap-
proximately 40 min.
donors: 3-hydroxyl (pKa ≈ 9.8) and 4-carbonyl functions (pKa
≈ 3.7, pKa values relating to 1,2-dimethyl-3-hydroxypyridine-
4-one). The pKa value of the carbonyl function results from
extensive delocalization of the lone pair associated with the ring
nitrogen. pKa values of both sensor molecules as derived from
the spectrophotometric titration are listed in Table 1. For 9 and
20 the oxygen atom of the 3-hydroxyl function shows values
of 9.91 and 9.29, respectively, and the corresponding values
for the oxygen atom of the 4-carbonyl function are 3.85 and
3.81.
Considering the representative speciation plot of 20, both
fluorescent sensors exist as dianions at physiological pH because
of the deprotonated phenol and carboxyl functions of the
fluorescein moieties, showing pKa values of 6.26/1.91 for 9 and
6.44/2.01 for 20. To assess membrane permeation qualities of
the probes, the logarithm of the partition coefficient, log P, and
the distribution coefficient, log D7.4, were determined for both
sensors (Table 1). The partition coefficient, representing the
standard measure of lipophilicity, is determined by distribution
of a compound between two phases (n-octanol/water) applying
pH conditions providing for a neutral molecule. At pH 7.4 the
neutral LH3 sensor species of 9 and 20 represent only 0.0019%
and 7.9% of the overall species composition but exist as the
dominant species within the pH ranges of 2-4 (9) and 3.5-5.5
(20). While 20 exhibits a log P value of -0.35, 9 shows a
surprisingly high log P value of 2.12, suggesting high membrane
permeation qualities of its neutral molecule species. The striking
difference between the log P values of 9 and 20 can be attributed
to the additional functionalization of the HPO chelation unit of
20 in the 1-position. Introduction of the ethyl-bridged secondary
amine strongly influences the overall polarity of the molecule
and renders the probe significantly more hydrophilic than its
unsubstituted analogue. On the basis of the log P measurements,
the corresponding log D7.4 values were calculated, quantifying
the potential membrane permeation qualities of both probes
under physiological conditions. Comparison of the log D7.4
values of 9 (-2.61) and 20 (-3.03) with the clinically applied
3-hydroxypyridin-4-one analogue deferiprone (1,2-dimethyl-3-
hydroxypyridine-4-one, log D7.4 ) -0.77)41 clearly reflects the
magnitude of change in molecular polarity, primarily resulting
Figure 1. (A) Representative fluorescence emission profile of 20 in
the presence of increasing Fe(III) concentrations (arrow). Fe(III) atomic
absorption spectroscopy standard solution was mixed with NTA in a
1:5 molar ratio prior to the measurements. Quenching was carried out
by titration of a probe solution (3 µM) in Chelex-treated MOPS buffer
(50 mM) at pH 7.4 with a series of Fe(III)/NTA concentrations resulting
in probe/metal ratios of 9:1, 6:1, 4.5:1, 3:1, 1.5:1, 1:1, and 1:2.
Fluorescence was excited at 491 nm. (B) Fluorescence intensity maxima
(514 nm) of the quenching spectra plotted against corresponding Fe(III)
concentrations. Measurements were carried out in triplicate. Data shown
represent mean values ( SEM.
(Figure 3) illustrates the decrease of sensor fluorescence upon
acidification in relation to the highest observed fluorescence at
pH 7.4 set at 100%. The pH profile of 20 revealed a 15-67%
decrease in fluorescence intensity between pH 7.4-6 and pH
6-4.5, respectively, corresponding to the region of pH changes
that can be anticipated during transitions of early endosomes
to late endosomes/lysosomes.
Physicochemical Characterization. The specific solution
properties of the fluorescent sensors 9 and 20 were investigated
using an automated spectrophotometric titration system. The
UV/vis profile of the iron free sensor 20 measured over the pH
range 1.65-11.16 is illustrated in Figure 4 together with the
speciation plot derived from the detected spectral changes. As
shown for 20, both probes exist predominantly as the 3-proto-
nated LH species at pH 7.4 (9, 93.0%; 20, 88.9%), thus forming
neutral tris-Fe(III) complexes with intracellular labile iron under
physiological conditions. The affinity of both compounds for
Fe(III) reflects the combined pKa values of the chelating oxygen

Endosomal/Lysosomal Fluorescent Iron Chelators
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 15 4543
Table 1. Physicochemical Properties of the Fluorescent Probes 9 and 20
excitation λ
emission λ
ε a
quenching
[Fe(III)] (%)
dequenching
(SIH/CP94) (%)
pKa
log P b
log D7.4
(nm)
(nm)
×103 m-1 cm-1
9
20
494
491
512
514
49.6
60.1
77
96
92/87
95/87
9.91, 6.26, 3.85, 1.91
9.29, 6.44 5.74, 3.81, 2.12
2.12
-0.35
-2.61
-3.03
a Molar extinction coefficient; b Partition coefficient (citric acid buffer/n-octanol), measured at pH 3 (9) and pH 4.5 (20).
Fluorescence data acquired for the iron-sensitive probe 20
over the entire time-course of the experiments is shown in Figure
6. Curve progression indicates that dequenching/quenching
processes passed through a maximum at 1 h. Converted to
percentage changes of fluorescence emission intensities in
relation to untreated 20-stained cells (set at 0%), a 187.78 (
52.78% fluorescence increase and a 67.33 ( 4.33% decrease
after a 1 h treatment with DFO or iron dextran were recorded.
Comparison of in vitro data for 20 with results for 9 (data not
shown) showed a higher degree of in vitro responsiveness of
20. Applying DFO or iron dextran for 1 h to 9-stained BMML
resulted in significantly lower dequenching (43.65 ( 7.83%)
and quenching processes (41.58 ( 1.82%). This is in agreement
with the superior optical properties of 20 (Figure 1B).
At 2 h after loading, the fluorescent signals of both sensors were
reproducibly reduced when compared to the 1 h time point values.
While 20 shows a considerable difference of 83.35% for dequench-
ing and 11.75% for quenching signals, 9 exhibits less pronounced
changes of 11.20% and 7.87%, respectively (data not shown). To
evaluate possible changes in dequenching and quenching capacities
with regard to intracellular fluorescence signals independent of DFO
or iron dextran, fluorescence intensities of untreated cells stained
with the fluorescent probes or their protected precursors 8 and 19
were compared over the entire 2 h incubation period as illustrated
in Figure 7. Labeling with iron-sensitive or -insensitive probes lead
to a constant intracellular fluorescence of labeled cells after reaching
a steady quenching equilibrium. The data suggest that 20 has
already reached a quenching equilibrium 20 min after loading (t
) 0 min) while 9 has reached an equilibrium 1 h after loading.
Thus, alterations of fluorescence intensities are independent of basal
cellular fluorescence levels and can most likely be accredited to
intracellular iron homeostasis kinetics. The differences between the
fluorescence signals of the iron insensitive precursors compared
to the iron sensors as illustrated in Figure 7 indicate quenching
processes predominantly associated with iron chelation.
Figure 3. pH dependent fluorescence profile of 20. The probe was
measured at pH 3, 4, 5, 6, 7.4, and 8 in Chelex-treated buffer solutions
(pH 3-6, citrate buffer 10 mM; pH 7.4, 8, MOPS buffer 50 mM).
Shown are mean values ( SEM of triple measurements.
from coupling of the fluorescein fluorophor moiety. Thus, the
low distribution coefficient further confirms the results obtained
from spectrophotometric titrations characterizing both probes
as highly hydrophilic and virtually membrane impermeable
under physiological conditions.
Live Cell Measurements. The in vitro sensitivity of the
fluorescein labeled probes was investigated by flow cytometry
(FACS) analysis using bone marrow derived macrophages
(BMML) from C57BL/6 mice. All FACS analyses were carried
out in Chelex-treated PBS to avoid external iron contamination
resulting in artificial quenching signals.
Representative for both probes, the intracellular responsive-
ness of 20 to changes of lysosomal iron concentrations was
measured after cellular iron supplementation or deprivation.
Endosomal/lysosomal iron concentrations were specifically
altered using high molecular weight iron dextran and the
membrane impermeable high affinity chelator desferrioxamine
(DFO), both selectively accumulating in the endosomal/lyso-
somal compartments because of their cellular uptake route via
endocytosis. Studies on cellular uptake kinetics of 9 and 20
showed optimal intracellular probe loading for FACS analysis
after pulsing BMML with 5 µM probe in iron-free PBS for 20
min followed by a 1 or 2 h chase.42
The sensitivity of 20 to changes of intralysosomal iron
concentrations was measured in comparison to its protected and
therefore iron-insensitive fluorescent precursor 19 after 1 and
2 h. Flow cytometric histograms are presented in Figure 5. The
fluorescent probe 19, which is incapable of chelating iron, shows
no response to manipulations of the endosomal/lysosomal iron
content and exhibits constant fluorescence emission intensities
under all conditions. The iron-sensitive probe 20, however,
shows increased fluorescence intensity in cells treated with the
high-affinity iron chelator DFO in comparison to untreated cells.
Correspondingly, decreased fluorescence intensity can be ob-
served following subsequent addition of exogenous iron via iron
dextran.
Constant fluorescence of the iron insensitive precursors 8 and
19 over the 2 h incubation period suggests that changes in
fluorescence intensities due to a shift in endosomal/lysosomal
pH are minor compared to those caused by iron-mediated probe
quenching. This finding is in accordance with the recently
published application of the fluorescein-based probe calcein-
AM to monitor the lysosomal labile iron pool.43
Responsiveness to Clinical Iron Chelators. To evaluate
potential application of the fluorescent probes in the diagnosis
and evaluation of iron-related diseases, sensor responsiveness
was tested against the clinical medications deferasirox (DFR)
and deferiprone (DFP), both orally active iron chelators routinely
used for iron excretion in iron-overloaded patients. The chelation
potency of DFR and DFP was measured by 20 in BMML and
is shown in comparison to the data obtained for DFO (Figure
8). On the basis of the individual pFe3+ values of all three
chelators, the hexadentate iron chelator DFO (pFe3+ ) 26.6)
was expected to exhibit the most efficient iron chelation
correlated with a high fluorescence signal of the sensor.
However, DFO iron chelation is slowed by its membrane
impermeability and endocytic uptake route. In contrast to DFO,
DFR (pFe3+ ) 22.5) benefits from high lipophilicity and low

4544 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 15
Fakih et al.
Figure 4. Spectrophotometric determination of the sensor solution properties: (A) pH dependent UV/vis spectra of 20 (34 µM) over the pH range
0
of 1.65-11.16; (B) speciation plot of 20 (2, LH52+; 4, LH4+; 9, LH3 ; 0, LH2-; b, LH2-; O, L3-).
molecular weight providing for rapid intracellular distribution
and a higher chelation potency within the measured time range.
Correspondingly, the tridentate chelator shows the largest 20
fluorescence increase (62.23%) after the initial 20 min incubation
period followed by DFO (48.22%) in comparison to untreated
20-stained cells set at 0%.
In contrast to DFO and DFR, 20 quenching was observed in
the case of the bidentate chelator DFP (pFe3+ ) 19.5, -7.21%).
This effect is due to the applied staining procedure. As cells
were pretreated with the chelators, subsequent fluorescent probe
pulsing confronted the intracellular chelator-iron complexes
with a huge excess of 20. DFP and 20 both complex iron via
an HPO based chelation unit and show comparable iron
affinities. Confronted with a large 20 excess, intracellular DFP
had no affinity advantage and consequently an iron exchange
between the DFP-iron complex and the probe occurred,
resulting in the observed quenching processes. In similar
experiments applying 9, cells were stained prior to application
of the chelators. With this setup it could be shown that within
the monitored incubation period DFP induced a clear 45%
fluorescence recovery ranging between DFO (40%) and DFR
(77%) in relation to untreated cells set at 0%.42 It can therefore
be concluded that in the case of DFP the observed 20 quenching
is a methodical and not a biochemical effect.

Endosomal/Lysosomal Fluorescent Iron Chelators
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 15 4545
Figure 5. Representative flow cytometric histograms of the benzyl-protected iron-insensitive probe 18 (A) and its iron-sensitive analogue 20 (B) in
response to changes of intralysosomal iron concentrations after 1 h. Data were acquired in the green Fl1 channel. Gates were based on dot-plots of untreated
cell populations. Macrophages were stained with the fluorescent compounds (5 µM) during a 20 min incubation period at 37 °C and 7.5% CO2 in the dark
and subsequently treated for 1 h with iron dextran (500 µg/mL) for iron saturation and DFO (1 mM) for iron depletion. Histograms of unstained cells are
plotted in comparison to histograms of stained cells without treatment as well as after iron dextran and DFO treatment.
Figure 6. Intracellular fluorescence in BMML labeled with 20 during
a 2 h incubation period in response to endosomal/lysosomal iron
supplementation using iron dextran (500 µg/mL) and lysosomal iron
Figure 7. Comparative plots of untreated BMML stained with the
iron-sensitive sensors 9 (A)/20 (B) and their corresponding iron-
insensitive precursors 8/19 measured by FACS after probe loading for
deprivation using DFO (1 mM) according to the procedure described
in Figure 5. Kinetics of intracellular fluorescence (A) as well as
percentage of fluorescence changes between untreated and treated cells
20 min (t ) 0) and at t ) 1 and 2 h. Shown are median values ( SEM
(B) are shown as mean values ( SEM of triple FACS measurements.
of triple measurements.
Intracellular Distribution Studies. In order to characterize
derived macrophages were incubated with 75 µM probe
the intracellular fate of the iron-sensitive probes, bone marrow
solutions in iron-free PBS for 20 min. Because of the negatively

4546 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 15
Fakih et al.
revealed progressive probe trafficking to DexTR positive
compartments (Figure 10F-T) and complete absence of the
probe from DexTR negative compartments after 30 min of chase
(Figure 10T), consistent with the dynamics of endocytic
pathways. Overall probe fluorescence intensity over the 30 min
incubation period was not affected (Figure 10A,F,K,P).
Studies applying DexTR-labeled cells treated with DFO or
iron dextran and subsequently labeled with 20 allowed combi-
natorial monitoring of sensor responsiveness and lysosomal
colocalization. Illustrated in Figure 11A-C are representative
magnifications of macrophages measured in the green and red
channels after a 60 min incubation period. At this later time
point 20 fluorescence colocalized predominantly with DexTR
labeled late endosomes/lysosomes. DFO-treated cells exhibited
considerably brighter 20 fluorescence, suggesting significant
intracellular probe dequenching by this iron chelator. Iron
dextran treated cells illustrated in Figure 11I-K show the same
colocalization pattern in combination with a reduced 20
fluorescence signal following vesicular probe quenching.
Further characterization of cellular sensor location was carried
out in macrophages stained for endosomes, endoplasmic reticu-
lum, and mitochondria, using DexTR, ER-Tracker Red, and
MitoTracker Deep Red, respectively. Mixed solutions of tracer
and 20 were applied to cells for 20 min. At this early time point,
DexTR and 20 are predicted to locate to endosomes, whereas
lysosomes are expected to be tracer- and sensor-free. As
illustrated in Figure 12A-E, there is partial colocalization
between the sensor and DexTR; however, DexTR-negative 20-
containing vesicles and DexTR-positive 20-negative vesicles are
also present. These findings suggest partially divergent uptake
pathways for DexTR and 20. ER-labeling (Figure 12F-J) and
mitochondrial labeling (Figure 12K-O) reveal characteristic
staining of these organelles and no colocalization with the iron
sensor, confirming the sensor confinement to the endocytic
pathways.
Figure 8. Responsiveness of intracellular 20 fluorescence to clinically
approved iron chelators. BMMØ was treated for 40 min with DFO
(100 µM), DFR (100 µM), and DFP (100 µM) and subsequently stained
with 20 for 20 min at 37 °C and 7.5% CO2. Fluorescence was measured
by FACS as described in Figure 5. Shown are median values ( SEM
of triplicate measurements.
charged nature of both probes at physiological pH, their
intracellular localization, i.e., trafficking, accumulation, and
distribution, is expected to be confined to the endosomal/
lysosomal system. Preliminary confocal microscopy analysis of
9-labeled cells revealed a distinct vesicular localization of the
probe 20 min after loading, most likely representing endosomal/
lysosomal compartments.35
A similar localization within intracellular vesicles was
observed for 20. Illustrated in Figure 9 are confocal microscopy
images of 20-labeled cells acquired after a 20 min pulse with
20 and a subsequent chase over a 60 min period, initially in the
presence of DFO and its subsequent replacement with iron
dextran. As shown in Figure 9A, a 20 min incubation period
resulted in efficient probe saturation for most of the macroph-
ages. Extracellular probe sticking to outer cell membranes
indicates that cells are still in the process of probe uptake, which
possibly represents a preliminary stage to endocytic probe
internalization. On the other hand, images acquired later in the
60 min chase period illustrate completed cellular probe uptake
and reveal a defined vesicular probe localization (Figure 9B-D)
as observed for 9.
To verify functionality of vesicular 20, fluorescence intensity
was measured in response to DFO treatment for 20 and 40 min
during the chase period. Considerable fluorescence recovery in
probe loaded vesicles could be observed upon incubation with
DFO, showing equilibrated dequenching conditions after a 20
min incubation period with no notable increase after 40 min
(Figure 9B,C). Cross-check of sensor responsiveness by sub-
sequent application of iron dextran for an additional 20 min
incubation period to the DFO treated macrophages resulted in
significant intracellular fluorescence decrease as excessive
exogenous iron quenched the fluorescent sensor (Figure 9D).
To characterize the intracellular compartments accumulating
the probes, colocalization studies using Texas Red labeled
dextran (DexTR) as a lysosomal tracer were carried out in 20-
labeled macrophages. The representative colocalization pattern
acquired after the initial 20 min probe incubation period is
illustrated in Figure 10A-C and shows a direct comparison of
cellular fluorescence recorded in the green and red channels.
At this early time point after loading, 20 was only present within
DexTR negative compartments most likely representing early
endosomes as shown in magnification in Figure 10E. Subsequent
monitoring of 20 localization over a 30 min chase period
Probe Uptake. To further specify endocytic probe uptake
mechanisms, bone marrow derived macrophages were treated
with cytochalasin D, a commonly used inhibitor of actin-
dependent macropinocytosis. The cells were incubated with
cytochalasin D for 30 min and subsequently treated with 20 or
DexTR (10 kDa) for 20 min. Confocal microscopy revealed a
reduction in DexTR uptake, whereas uptake of 20 was not
affected (data not shown), suggesting that in contrast to DexTR,
20 is engulfed by actin-independent pinocytosis.
Conclusions
Lysosomes have been suggested to play a major role in the
intracellular regulation and homeostasis of iron metabolism.
Because of the breakdown of iron-containing endocytosed as
well as autophagocytosed material, significant concentrations
of labile iron accumulate in the lysosomes, potentially catalyzing
organelle rupture and subsequent cell death. To study lysosomal
iron and its influence on cellular iron homeostasis, we have
synthesized the novel fluorescein-labeled iron(III) chemosensors
9 and 20 featuring substituted 3-hydroxypyridin-4-one (HPO)
based chelation moieties specifically targeting the endosomal/
lysosomal system for iron monitoring. Detection of cellular
quenching and dequenching processes showed a clear difference
in sensor responsiveness to manipulations of intracellular iron
concentrations, with 20 being of magnitudes more responsive.
Because pKa values give no indication of differences in iron
binding properties of both probes, substitution in the 1-position
of 20 is most likely responsible for the observed differences.
Owing to the deprotonated fluorescein moiety, both probes
are highly hydrophilic and virtually membrane impermeable at

Endosomal/Lysosomal Fluorescent Iron Chelators
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 15 4547
Figure 9. Representative confocal microscopy images of the sensor distribution in BMMØ. Cells were cultured on µ-channel slides and treated
with a 75 µM solution of 20 for 20 min. Fluorescence was excited at a 488 nm excitation wavelength using a 505 nm long-pass filter for emission.
Fluorescence data are shown in comparison to the corresponding bright field images. Untreated 20-labeled cells were monitored after the initial 20
min incubation period (A). Following exchange of the incubation buffer with a 1 mM DFO solution, cells were monitored after an additional 20
min (B) and 40 min (C) incubation period. Subsequent application of a 500 µg/mL solution of iron dextran (FeDex) was measured after a further
20 min (D).
were cultured overnight, and nonadherent cells were washed off
prior to manipulations. No antibiotics were used in the culture
medium.
physiological pH. Correspondingly, cellular imaging experi-
ments using confocal microscopy in macrophages revealed
compartmentation of both probes in endosomal/lysosomal
compartments due to a pinocytic uptake route. Furthermore,
application of iron(III) insensitive precursors could demonstrate
that endosomal maturation accompanied by acidification did not
influence iron related measurements. Correspondingly, the
probes proved to be highly sensitive and reproducible in
assessing the chelation potency of the clinically employed
chelators, deferiprone, desferrioxamine, and deferasirox. De-
velopment of fluorescein coupled HPO-based iron(III) sensors
therefore represents a considerable step forward toward devel-
oping techniques to study subcellular iron distribution. Com-
bination of exclusive endosomal/lysosomal accumulation with
the high HPO iron specificity/affinity and reproducibly high
responsiveness toward intracellular changes of free iron resulted
in a promising class of new sensors to study iron metabolism
under physiological and pathophysiological conditions. Ap-
plication of these probes in real-time monitoring of the
endosomal/lysosomal labile iron pool will allow assessment of
iron status in patients under iron supplementation or chelation
therapy. These novel diagnostic tools will also provide valuable
insights into the role of intracellular trafficking and homeostasis
of redox-active labile iron for diseases as diverse as infections,
autoimmunity, and cancer. They will significantly contribute
to a facilitated development of more efficient iron chelating
agents.
Manipulation of the intracellular iron content was carried out
by iron supplementation or deprivation treatments. Cells were
incubated for up to 2 h prior to staining in 500 µg/mL iron dextran
(Pharmacosmos A/S, Holbaek, Denmark) for cellular iron overload-
ing and in 1 mM/100 µM deferoxamine mesylate (DFO, Sigma
Aldrich, Gillingham, U.K.), 100 µM deferiprone (DFP, Sigma
Aldrich, Gillingham, U.K.), or deferasirox (DFR) for iron deprivation.
Inhibition of macropinocytosis by cytochalasin D (Calbiochem
Merck Chemicals, Nottingham, U.K.) was performed as follows.
Cytochalasin D was provided as a 1 mg/mL stock solution in DMSO
and diluted with iron-free PBS to a 10 µg/mL solution. BMML
was incubated with the compound for 30 min, washed with Chelex-
treated PBS, and subsequently treated with a 75 µM solution of
the fluorescent iron sensor or a 100 µg/mL solution of DexTR
(Molecular Probes Invitrogen, Paisley, U.K.).
Flow Cytometry. For flow cytometric measurements nonadher-
ent cells were washed off from cultivated cell monolayers, which
were subsequently incubated at 4 °C for 20 min in cold PBS (pH
7.4) treated with Chelex-100 (sodium form, Sigma Aldrich) to
remove iron contaminations. The cells were harvested, washed three
times with cold iron-free PBS, and split into FACS tubes at 105
cells per tube. All subsequent washes and procedures were carried
out on ice unless otherwise stated. The cells were incubated with
the fluorescent probes, provided as 10 mM stock solutions in DMSO
and diluted with cold iron-free PBS at a final concentration of 5
µM, for 20 min at 37 °C and 7.5% CO2. Cells were washed and
resuspended in cold Chelex-treated PBS. Measurements were
carried out on a FACSCalibur flow cytometer (BD Biosciences),
and analysis proceeded via CellQuest and FlowJo softwares. Gates
were based on dot-plots of untreated cell populations. Medians of
at least 10 000 events were recorded and corrected for cell
autofluorescence. Mean values were calculated from three inde-
pendent experiments.
Experimental Section
Cell Culture and Reagents. Bone marrow derived macrophages
(BMML)were isolated from the femurs and tibias of 6-8 month
old C57BL/6 mice (Harlan Bioproducts, Bicester, U.K.) according
to standardized procedures. Cells were plated in Dulbecco’s
modified Eagles medium (DMEM) supplemented with 10% heat-
inactivated fetal calf serum (FCS), 5% horse serum, 1% L-glutamine,
and 20% L929 cell culture supernatant as a source of macrophage
colony-stimulating factor (M-CSF; Biochrome, Berlin, Germany).
Cells were cultivated at 37 °C and 7.5% CO2. Mature adherent
macrophages were harvested after 7 days of culture and stored at
-80 °C in 10% DMSO to be replated when required. Thawed cells
Confocal Microscopy. Confocal microscopic imaging of
intracellular probe distribution was carried out using BMML
plated onto uncoated µ-channel slides (Ibidi, Munich, Germany)
at 105 cells per slide. For lysosomal colocalization experiments,
cells were cultured with 100 µg/mL DextranTR (10 kDa) for
4 h and chased overnight. The cells on the µ-channel slide were
washed three times with 600 µL portions of Chelex-treated PBS

4548 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 15
Fakih et al.
Figure 10. Representative confocal microscopy images of intracellular colocalization studies of 20 in BMMØ with dextran Texas Red labeled
lysosomes. Cells were cultured on µ-channel slides, treated with a 100 µg/mL dextran Texas Red solution for 4 h, and chased overnight. Dextran
Texas Red labeled cells were stained for 20 min with a 75 µM solution of 20, followed by subsequent exchange of the incubation solution with
iron-free PBS. Intracellular fluorescence of stained cells was excited at 543 nm (dextran texas red) or at 488 nm (20) using a 560 and a 505 nm
long-pass filter for emission. Fluorescence of stained cells was measured in the green (A, F, K, P) as well as in the red (B, G, L, Q) channel, and
data were merged for evaluation of colocalization in yellow (C, H, M, R). Fluorescence data are shown in comparison to the corresponding bright
field images (D, I, N, S). Magnifications of merged fluorescence data show detailed changes in localization of the probe with respect to the tracer
at various stages of cellular probe uptake (E, J, O, T). Parts A-E were taken immediately after probe loading. Parts F-J were taken after a 10 min
chase. Parts K-O were taken after a 20 min chase. Parts P-T were taken after a 30 min chase. Parts U-X are BMMØ labeled with dextran Texas
Red only as a control for the broad emission spectrum of the red channel.
to remove cell culture medium and incubated for 20 min with a
75 µM probe solution in iron-free PBS at 37 °C and 7.5% CO2.
For endosomal, mitochondrial, and endoplasmic reticulum colo-
calization experiments, 100 µg/mL DextranTR, 1 µM ER-Tracker
Red (Invitrogen, Paisley, U.K.), or 250 µM MitoTracker Deep
Red (Invitrogen, Paisley, U.K.) was added to a 100 µM probe
solution in iron-free PBS and applied to cells for 20 min at 37
°C and 7.5% CO2. The cells were repeatedly washed with iron-
free PBS and kept on ice for subsequent measurements or at 37
°C for probe chasing. Fluorescence was measured using an
inverted Zeiss confocal fluorescent microscope using a 63× oil
immersion objective. Images were acquired and analyzed by
Zeiss LSM image browser software.
and a reference electrode (6.0733.100) was used. A 0.1 M KCl
electrolyte solution was used to maintain the ionic strength. For
the measurements the temperature of the probe solution was
maintained in a thermostatic jacketed titration vessel at 25 ( 0.1
°C using a Techne TE-8J temperature controller. The solution was
stirred vigorously during the experiment. A Gilson Mini-plus#3
pump with speed capability (20 mL/min) was used to circulate the
probe solution through a Hellem quartz flow cuvette. A cuvette
path length of 10 mm was used and the flow cuvette mounted on
a HP 8453 UV-visible spectrophotometer. All instruments were
interfaced to a computer and controlled by a Visual Basic program.
The pH of the probe solution was increased by 0.1 pH unit by the
addition of KOH from the autoburet. When pH readings varied by
<0.001 pH unit over a 3 s period, an incubation period of 1 min
was adopted. At the end of the equilibrium period, the spectrum of
the solution was recorded. The cycle was repeated automatically
until the defined end point pH value was achieved. Titration data
were analyzed by pHab.44
Physical Measurements. Optical characterization of the fluo-
rescent probes 9 and 20 was performed on a Perkin-Elmer
spectrophotometer type UV/vis Lamda2S and a Perkin-Elmer
spectrofluorometer type LS 50B.
1H NMR spectra were recorded using a Bruker 360 MHz NMR
spectrometer. Chemical shifts δ are reported in ppm.
pKa Determination. The pH electrodes were calibrated by
titrating a volumetric standard strong acid HCl (0.15 mL, 0.2092
M) in KCl (15 mL, 0.1 M) with KOH (0.1 M) under an argon
atmosphere at 25 °C. The E0, slope of the electrode, and pKw of
the solution were both optimized by GLEE.45 Following electrode
calibration, a 34 µM solution of 20 in 11.162 mL of 0.1 M KCl at
an initial pH value of 1.646 were alkalimetrically titrated to pH
Mass spectra (ESI) analyses were carried out by the Mass
Spectrometry Facility, School of Health and Life Sciences, 150
Stamford Street London SE1 9NH, U.K.
For pKa determinations, an automatic titration system comprising
an autoburet (Metrohm Dosimat 765, l mL syringe) and a Mettler
Toledo MP230 pH meter with a Metrohm pH electrode (6.0133.100)

Endosomal/Lysosomal Fluorescent Iron Chelators
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 15 4549
Figure 11. Representative confocal microscopy images of intracellular colocalization studies of 20 in BMMØ using dextran Texas Red labeled
lysosomes. Colocalization studies according to the experimental setup described in Figure 10 were carried out in DFO (E, F) and iron dextran (I,
J) pretreated cells as well as untreated cells (A, B), with 1 h of incubation with 75 µM 20. Findings are illustrated together with merged fluorescence
data showing a high degree of lysosomal colocalization in yellow at a later stage of cellular probe processing (C, G, K). Fluorescence data are
shown in comparison to the corresponding bright field images (D, H, L).
Figure 12. Representative confocal microscopy images of intracellular colocalization studies of 20 in BMMØ with dextran Texas Red labeled
endosomes (A-E), ER-Tracker labeled endoplasmic reticulum (F-J) and MitoTracker labeled mitochondria (K-O). Cells were treated simultaneously
with 100 µM 20 and 100 µg/mL dextran Texas Red, 1 µM ER-Tracker Red for 20 min, or 250 µM MitoTracker Deep Red, followed by subsequent
exchange of the incubation solution with iron-free PBS. Fluorescence data were collected as described in Figure 10. Green (A, F, K) and red (B,
G, L) fluorescence data were merged together (C, H, M) to represent colocalization in yellow, also in detail in magnifications (E, J, O). Fluorescence
data are shown in comparison to the corresponding bright field images (D, I, N).
11.162. The spectra of 20 and the titration experimental data at
different pH values were analyzed by refining the extinction
coefficients of the protonated and deprotonated species.
interval was repeated three times, and the layers were subsequently
separated by centrifugation at 25 °C. The aqueous phases were again
measured spectrophotometrically showing a decrease of probe
absorbance due to partition into the organic phase. Lipophilicity,
directly correlating with the detected decrease in probe absorbance,
was derived from the recorded UV spectra according to the
following equation:
log P Determination. The log P values of 9 and 20 were
determined by the shake-flask method using a n-octanol-aqueous
buffer system at 25 °C. Because the log P is strictly defined for
neutral molecules, the partition of both probes was measured in
citric acid buffer at pH 3 and pH 4.5, respectively, providing
experimental conditions where according to probe speciation the
neutral LH3 species predominates. The absorbance of a buffered
probe solution of known concentration was recorded spectropho-
tometrically, and different ratios of buffered probe solution and
n-octanol were then shaken by vigorous stirring for 1 h. The stirring
(A0 - A1)Vaq
P )
A1Vo
where A0 is the initial absorbance of the aqueous phase, A1 is the
absorbance of the aqueous phase after separation of the aqueous

4550 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 15
Fakih et al.
phase by centrifugation, Vaq is the total volume of the aqueous
phase, and Vo is the total volume of the organic phase.
was stirred at room temperature overnight in the dark to complete
the intermediate formation of the activated fluorophor 7. Subse-
quently, the formed white N,N-dicyclohexylurea (DCU) precipitate
was filtered off and an amount of 266 mg (1,46 mmol) of 5 was
added to the red filtrate. The mixture was stirred at room
temperature for 24 h in the dark, followed by solvent removal under
reduced pressure. The orange residue was taken up in 25 mL of
1.25% sodium hydroxide solution and the aqueous layer washed
successively with chloroform (3 × 10 mL). The pH of the aqueous
reaction solution was adjusted with 4 N hydrochloric acid to pH 3,
resulting in the formation of a thick orange precipitate. After
completion of the precipitation the orange solid was filtered off
and dried under high vacuum. Purification of the crude product
was conducted by preparative HPLC methods using a reversed
phase C18 column and application of an isocratic water/0.1% TFA
(buffer A) and acetonitrile (buffer B) gradient (432 mg, 0.80 mmol,
60%). m/z: 563 (M + Na+).
4(and 5)-[(3-Hydroxy-2-methyl-4-oxopyridin-5-yl)(methyl-
amino)methylcarbamoyl]-2-(3-hydroxy-6-oxo-6H-xanthen-
9-yl)benzoic acid (9). A suspension of 100 mg (0.19 mmol) 8 in
5 mL of anhydrous dichloromethane was flushed with nitrogen and
cooled to 0 °C in an ice/water bath. An amount of 2 mL of BCl3
solution (1 M in dichloromethane) was added dropwise to the cooled
solution, and the mixture was stirred at room temperature in the
dark until complete conversion of the starting material. An amount
of 10 mL of methanol was added, and the mixture was stirred for
30 min to eliminate excess of BCl3. The solvent was removed under
reduced pressure, and after additional drying of the residue in high
vacuum the crude product was purified by recrystallization.
Dissolution of the solid in a minimal amount of methanol and
subsequent precipitation with excess acetone yielded the pure yellow
hydrochloride salt of the fluorescent product as an amorphous
powder. Purity of the compound was verified by HPLC using a
reversed phase C18 column and applying an isocratic water/0.1%
TFA (buffer A) and acetonitrile (buffer B) method (for data, see
Supporting Information) (25 mg, 0.04 mmol, 23%). m/z: 527 (M
+ 1)+.
Syntheses. All starting materials were obtained from commercial
sources and used without further purification. 3-Hydroxy-2-methyl-
4H-pyran-4-one (maltol, 1) was purchased from Cultor Food
Science, and 5-hydroxy-2-(hydroxymethyl)-4H-pyran-4-one (kojic
acid, 10) and 5(6)-carboxy fluorescein 6 were purchased from Fluka.
Syntheses of 2-5 and 11-14 were based on procedures described
previously.28 Because of the applied mix of isomers of 5(6)-carboxy
fluorescein, no NMR data were recorded for 9 and 20 and
compound identity was established on the basis of mass spectrom-
etry measurements.
3-Methoxy-2-methyl-4H-pyran-4-one (2). A solution of 63.06
g (500 mmol) maltol in 300 mL of 10% potassium hydroxide
solution was cooled to 0 °C in an ice/water bath. An amount of
52.04 mL (69.37 g, 550 mmol) of dimethyl sulfate was added
dropwise to the solution, which was kept at 0 °C throughout the
reaction. Stirring continued for 4 h and was followed by extraction
of the reaction solution with dichloromethane (3 × 100 mL). The
combined organic extracts were washed with 5% KOH solution
and water (3 × 100 mL), dried over sodium sulfate, and filtered.
The solvent was removed in vacuum to yield the product as an
1
orange oil (49.85 g, 356 mmol, 71%). H NMR (CDCl3) δ: 2.34
(s, 3H), 3.86 (s, 3H), 6.36 (d, 1H), 7.66 (d, 1H). m/z: 141 (M +
1)+.
3-Methoxy-2-methylpyridin-4(1H)-one (3). To a solution of
10.51 g (75 mmol) 2 in 80 mL of water were added 10 mL portions
of aqueous ammonia solution under reflux at 60 °C until all starting
material had completely reacted. Subsequently, the excess ammonia
was evaporated under reduced pressure and the aqueous reaction
solution dried in high vacuum. The solid brown residue was
thoroughly resuspended and washed in acetone, filtered, and dried
to yield the product as a light-brown solid (8.15 g, 59 mmol, 78%).
1H NMR (CDCl3) δ: 2.43 (s, 3H), 3.80 (s, 3H), 6.47 (d, 1H), 7.56
(d, 1H). m/z: 140 (M + 1)+.
5-{[Benzyl(methyl)amino]methyl}-3-methoxy-2-methyl-pyridin-
4(1H)-one (4). In 120 mL of ethanol, an amount of 46.46 mL (43.62
g, 360 mmol) of N-benzylmethylamine was mixed with 14.62 mL
(5.41 g, 180 mmol) of a 37% aqueous formaldehyde solution and
stirred at room temperature for 30 min. An amount of 8.35 g (60
mmol) of 3 was added, and the brown solution was boiled under
reflux until the reaction was complete. The solvent was removed
under reduced pressure, and the resulting brown oil was additionally
dried in high vacuum. The viscous residue was covered with acetone
and left at 4 °C overnight to allow crystallization of the product.
The resulting precipitate was extensively washed with acetone,
filtered, and dried in high vacuum to yield 4 as a white solid (14.22
2-(Chloromethyl)-5-hydroxy-4(1H)-pyran-4-one (11). A solu-
tion of 4.97 g (35 mmol) of 2-hydroxymethyl-5-hydroxypyrone (10,
kojic acid) in 20 mL of distilled thionyl chloride was allowed to
stir for 1 h at room temperature. Precipitation of the crude product
in the form of a brownish powder occurred, which was collected
by filtration and washed with petrol ether to furnish 11 as a white
1
amorphous solid (4.50, 28 mmol, 80%). H NMR (DMSO-d6) δ:
4.66 (s, 2H), 6.57 (s, 1H), 8.13 (s, 1H). m/z: 161 (M + 1)+.
5-Hydroxy-2-methyl-4(1H)-pyran-4-one (12). In a three-neck
round-bottom flask fitted with reflux condenser and thermometer,
an amount of 10.44 g (65 mmol, 1 equiv) of 11 was suspended in
50 mL of water and heated to 50 °C. Under vigorous stirring, an
amount of 8.52 g (130.34 mmol, 2 equiv) of zinc powder was slowly
added to the suspension while maintaining a constant reaction
temperature of 50 °C. The reaction solution was subsequently heated
to 70 °C and a total of 19.5 mL of concentrated hydrochloric acid
was added dropwise keeping the reaction temperature between 70
and 80 °C. Stirring continued for another 3 h at 70 °C, and excess
of zinc powder was immediately removed by filtration of the hot
reaction solution. The cooled filtrate was extracted with dichlo-
romethane (3 × 50 mL), and the combined organic extracts were
dried over sodium sulfate and filtered. The solvent was removed
under reduced pressure to yield 12 as a white powder (6.29 g, 50
1
g, 52 mmol, 87%). H NMR (DMSO-d6) δ: 2.09 (s, 3H), 2.19 (s,
3H), 3.34 (s, 2H), 3.50 (s, 2H), 7.27 (m, 5H), 7.49 (s, 1H). m/z:
273 (M + 1)+.
3-Methoxy-2-methyl-5-[(methylamino)methyl]pyridin-4(1H)-
one (5). An amount of 5.45 g (20 mmol) of 4 was dissolved in 100
mL of ethanol in a hydrogenolysis flask and debenzylated overnight
via hydrogenolysis in the presence of a 10% Pd/C catalyst. The
catalyst was subsequently filtered off and the solvent removed under
reduced pressure. The residual viscous brown oil was covered with
dichloromethane and kept at 4 °C overnight to allow crystallization
of the product. The resulting solid was filtered off and the filtrate
stored at 4 °C to be used in several corresponding cycles of
precipitation until the formed product was completely collected.
The combined precipitates were extensively washed with dichlo-
romethane, filtered, and dried in high vacuum to yield 5 as a white
1
mmol, 77%). H NMR (DMSO-d6) δ: 2.23 (s, 3H), 6.23 (s, 1H),
7.96 (s, 1H), 8.98 (s, 1H). m/z: 127 (M + 1)+.
1
solid (3.46 g, 19 mmol, 95%). H NMR (D2O) δ: 2.24 (s, 3H),
3-Hydroxy-2-(hydroxymethyl)-6-methyl-4H-pyran-4-one
(13). An amount of 15.13 g (120 mmol) of 12 was dissolved in
120 mL of an aqueous 1.1 M sodium hydroxide solution and stirred
at room temperature for 5 min. After dropwise addition of 10.22
mL (3.78 g, 126 mmol) of formaldehyde solution (37% in water),
the mixture was stirred for 24 h and subsequently the pH of the
solution was adjusted with concentrated hydrochloric acid to pH
1. The formation of a thick white precipitate occurred, which was
kept at 4 °C overnight to ensure complete product precipitation.
2.46 (s, 3H), 3.63 (s, 3H), 3.79 (s, 2H), 7.64 (s, 1H). m/z: 183 (M
+ 1)+.
4(and 5)-[(3-Methoxy-2-methyl-4-oxopyridin-5-yl)(methyl-
amino)methylcarbamoyl]-2-(3-hydroxy-6-oxo-6H-xanthen-9-
yl)benzoic acid (8). To a solution of 500 mg (1.33 mmol) of 5(6)-
carboxyfluorescein (6) in 15 mL of DMF were added 329 mg (1.59
mmol) of dicyclohexylcarbodiimide (DCCI) and 183 mg (1.59
mmol, 1.2 equiv) of N-hydroxysuccinimde. The reaction solution

Endosomal/Lysosomal Fluorescent Iron Chelators
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 15 4551
The solid was filtered off, washed with cold ether, and dried in
high vacuum to yield 13 as a white crystalline powder (14.76 g,
95 mmol, 79%). 1H NMR (DMSO-d6) δ: 2.26 (s, 3H), 4.39 (s,
2H), 6.22 (s, 1H), 8.88 (br s, 1H). m/z: 157 (M + 1)+.
2-(Aminomethyl)-3-(benzyloxy)-1-[2-(dimethylamino)ethyl]-
6-methylpyridin-4(1H)-one (18). A solution of 5 mL of methanol
containing 500 mg (1.12 mmol) of 17 and 69.91 µL (71.94 mg,
2.24 mmol) of hydrazine hydrate was refluxed overnight. The
solution was allowed to cool to room temperature, inducing
precipitation of side products. Cooling to 0 °C in an ice/water bath
and adjusting the solution to pH 1 using concentrated hydrochloric
acid led to a complete precipitation of the side products, which
were removed from the solution by filtration. The filtrate was
evaporated to dryness and dissolved in 5 mL of water. By use of
a 10 M sodium hydroxide solution, the aqueous mixture was
adjusted to pH 8, avoiding product precipitation, and added directly
to the dimethyl sulfoxide solution of the activated fluorescein
derivative 7 for fluorophor coupling.
4(and 5)-[(3-Benzyloxy-1-[2-(dimethylamino)ethyl]-6-methyl-
4-oxopyridin-2-yl)methylcarbamoyl]-2-(3-hydroxy-6-oxo-
6H-xanthen-9-yl)benzoic Acid (19). An analogous reaction as
described for 8 using a basic aqueous solution of 18 gave 19 (367
mg, 0.54 mmol, 41%). m/z: 674 (M + 1)+.
4(and 5)-[(1-[2-(Dimethylamino)ethyl]-3-hydroxy-6-methyl-
4-oxopyridin-2-yl)methylcarbamoyl]-2-(3-hydroxy-6-oxo-
6H-xanthen-9-yl)benzoic acid (20). An analogous procedure
starting with 50 mg (0.74 mmol) of 19 as described for 9 gave 20.
Purity of the compound was verified by two different HPLC
methods (for data, see Supporting Information). Method A had the
following parameters: reversed phase C18 column, isocratic method,
water/0.1% TFA (buffer A) and acetonitrile (buffer B). Method B
had the following parameters: reversed phase C18 column, gradient,
aqueous 1-heptanesulfonic acid (5 mM), pH 2 (HCl) (buffer A),
and acetonitrile (buffer B) (21 mg, 0.03 mmol, 46%). m/z: 584 (M
+ 1)+.
3-(Benzyloxy)-2-(hydroxymethyl)-6-methyl-4H-pyran-4-one
(14). To a solution of 15.61 g (100 mmol) of 13 in methanol
was added a total of 10 mL of a 11 M aqueous sodium hydroxide
solution, and the mixture was heated to reflux. An amount of
12.66 mL (13.92 g, 110 mmol, 1.1 equiv) of benzyl chloride
was added dropwise to the solution, and the mixture was refluxed
overnight. The NaCl precipitate was removed by filtration, and
the solvent evaporated under reduced pressure. The brown
viscose residue was taken up in dichloromethane and washed
with 5% aqueous sodium hydroxide solution and water. The
mixture was dried over soldium sulfate and filtered, and the
solvent was evaporated under reduced pressure. Recrystallization
of the crude product followed dissolution of the residue in a
minimal amount of dichloromethane and the addition of an
excess of petrolether. After filtration and drying of the precipitate
the product was obtained as a white powder (15.36 g, 62 mmol,
1
62%). H NMR (CDCl3) δ: 2.24 (s, 3H), 4.31 (s, 2H), 5.16 (s,
2H), 6.22 (s, 1H), 7.34 (m, 5H). m/z: 247 (M + 1)+.
3-(Benzyloxy)-1-[2-(dimethylamino)ethyl]-2-(hydroxy
methyl)-6-methylpyridin-4(1H)-one (16). An amount of 7.39 g
(30 mmol) of 14 was dissolved in 120 mL of dichloromethane. To
the solution were added 5.47 mL (5.05 g, 60 mmol) of 3,4-dihydro-
2H-pyran and a catalytic amount of p-toluenesulfonic acid (90 mg).
The mixture was stirred at room temperature for 6 h and
subsequently was washed with 5% sodium hydroxide solution and
water (3 × 40 mL). After solvent evaporated under reduced
pressure, 15 was yielded as a colorless oil and reacted in the next
step without further purification. The tetrahydropyrane protected
intermediate was taken up in 130 mL of water and 20 mL of ethanol.
An amount of 9.88 mL (7.93 g, 90 mmol) of N,N-dimethylethyl-
enediamine was added dropwise to the solution, and the mixture
was heated to reflux until 15 had completely reacted. The solvent
was removed under reduced pressure, and the residual brown oil
dissolved in 60 mL of ethanol. The solution was adjusted to pH 1
using concentrated hydrochloric acid and further refluxed for 5 h.
After evaporation of the solvent the dry residue was dissolved in
160 mL of water and extensively washed with diethyl ether (3 ×
70 mL). The pH of the aqueous solution was adjusted to pH 12
using 10 M sodium hydroxide solution and extracted with dichlo-
romethane (5 × 50 mL). The combined organic extracts were dried
over sodium sulfate, filtered, and dried in vacuum to yield the crude
product as a light-brown powder. Purification of 16 was carried
out by column chromatography on silica gel (eluent, chloroform/
methanol, 15:1 v/v) and furnished a white amorphous powder (3.99
g, 13 mmol, 42%). 1H NMR (CDCl3) δ: 2.22 (s, 6H), 2.27 (s, 3H),
2.51 (t, 2H), 4.08 (t, 2H), 5.15 (s, 2H), 6.32 (s, 1H), 7.34 (m, 5H).
m/z: 317 (M + 1)+.
Acknowledgment. This work was supported by a fellowship
for S.F. within the postdoctoral program of the German
Academic Exchange Service (DAAD). M.P. is a MRC PhD
Studentship holder, and U.E.S. is supported by the Wolfson
Research Fellowship by the Royal Society. The authors thank
Prof. Albert Haas for kindly providing tracers for colocalization
studies.
Supporting Information Available: HPLC results of 9 and 20
and metal selectivity data of 9. This material is available free of
charge via the Internet at http://pubs.acs.org.
References
(1) Lippard, S. J.; Berg, J. M. Principles of Bioinorganic Chemistry;
University Science Books: Mill Valley, CA, 1994.
(2) Kaim, W.; Schwederski, B. Bioinorganic Chemistry, 2nd ed.; B. G.
Teubner: Stuttgart, Germany, 1995.
(3) Halliwell, B.; Gutteridge, J. M. C. Role of free radicals and catalytic
metal ions in human disease: an overview. Methods Enzymol. 1990,
186, 1–85.
3-(Benzyloxy)-1-[2-(dimethylamino)ethyl]-6-methyl-2-
phthalimidomethyl-pyridin-4(1H)-one (17). An amount of 3.15 g
(12 mmol) of triphenylphosphine was dissolved in 50 mL of
tetrahydrofurane followed by 1.77 mg (12 mmol) of phthalimide.
Subsequently, an amount of 3.16 g (10 mmol) of 16 was added
and the resulting suspension was cooled to 0 °C in an ice/water
bath. An amount of 2.36 mL (2.43 g, 12 mmol) of diisopropyl
azodicarboxylate was added slowly to the suspension, and the
mixture was stirred overnight at room temperature. The thin white
precipitate formed after complete reaction of 16 was spinned down
in a centrifuge, and the brown supernatant containing the product
was additionally filtered through a syringe filter. The solvent was
evaporated under reduced pressure to yield the crude product as a
brown solid. Purification of 17 was carried out by column
chromatography on silica gel (eluent, chloroform/methanol, 30:1)
and furnished a white amorphous powder (1.02 g, 2.28 mmol, 23%).
1H NMR (CDCl3) δ: 2.16 (s, 6H), 2.38 (s, 3H), 3.47 (s, 2H), 4.23
(t, 2H), 4.72 (br s, sH), 5.38 (s, 2H), 6.36 (s, 1H), 7.25 (m, 5H),
7.71 (m, 2H), 7.77 (m, 2H). m/z: 446 (M + 1)+.
(4) Halliwell, B.; Gutteridge, J. M. C. Biologically relevant metal ion-
dependent hydroxyl radical generation. An update. FEBS Lett. 1992,
307, 108–112.
(5) Gaeta, A.; Hider, R. C. The crucial role of metal ions in neurodegen-
eration: the basis for a promising therapeutic strategy. Br. J. Phar-
macol. 2005, 146, 1041–1059.
(6) Molina-Holgado, M.; Hider, R. C.; Gaeta, A.; Williams, R.; Francis,
P. Metals ions and neurodegeneration. BioMetals, in press.
(7) Bacon, B. R.; Britton, R. S. Hepatic injury in chronic iron overload.
Role of lipid peroxidation. Chem.-Biol. Interact. 1989, 70, 183–226.
(8) Bacon, B. R.; Britton, R. S. The pathology of hepatic iron overload:
a free radical-mediated process. Hepatology 1990, 11, 127–137.
(9) Kehrer, J. P. The Haber-Weiss reaction and mechanisms of toxicity.
Toxicology 2000, 149, 43–50.
(10) Kowdley, K. V. Iron, hemochromatosis, and hepatocellular carcinoma.
Gastroenterology 2004, 127, S79–S86.
(11) Valko, M.; Rhodes, C. J.; Moncol, J.; Izakovic, M.; Mazur, M. Free
radicals, metals and antioxidants in oxidative stress-induced cancer.
Chem.-Biol. Interact. 2006, 1, 1–40.
(12) Valko, M.; Leibfritz, D.; Moncol, J.; Cronin, M. T. D.; Mazur, M.; et
al. Free radicals and antioxidants in normal physiological functions
and human disease. Int. J. Biochem. Cell Biol. 2007, 39, 44–84.

4552 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 15
Fakih et al.
(13) Petrat, F.; De Groot, H.; Rauen, U. Subcellular distribution of
chelatable iron: a laser scanning microscopic study in isolated
hepatocytes and liver endothelial cells. Biochem. J. 2001, 356, 61–
69.
(14) Yu, Z.; Persson, H. L.; Eaton, J. W.; Brunk, U. T. Intralysosomal
iron: a major determinant of oxidant-induced cell death. Free Radical
Biol. Med. 2003, 34, 1243–1252.
(15) Kurz, T.; Leake, A.; von Zglinicki, T.; Brunk, U. T. Lysosomal redox-
active iron is important for oxidative stress-induced DNA damage.
Ann. N.Y. Acad. Sci. 2004, 1019, 285–288.
(16) Kurz, T.; Leake, A.; von Zglinicki, T.; Brunk, U. T. Relocalized redox-
active lysosomal iron is an important mediator of oxidative-stress-
induced DNA damage. Biochem. J. 2004, 378, 1039–1045.
(17) Terman, A.; Gustafsson, B.; Brunk, U. T. The lysosomal-mitochondrial
axis theory of postmitotic aging and cell death. Chem.-Biol. Interact.
2006, 163, 29–37.
chelators with fluorescent sensors. J. Med. Chem. 2004, 47, 6349–
6362.
(29) Luo, W.; Ma, Y.; Quinn, P. J.; Hider, R. C.; Liu, Z. Design, synthesis
and properties of novel iron(III)-specific fluorescent probes. J. Pharm.
Pharmacol. 2004, 56, 529–536.
(30) Ma, Y.; De Groot, H.; Liu, Z.; Hider, R. C.; Petrat, F. Chelation and
determination of labile iron in primary hepatocytes by pyridinone
fluorescent probes. Biochem. J. 2006, 395, 49–55.
(31) Kaufmann, A. M.; Krise, J. P. Lysosomal sequestration of amine-
containing drugs: analysis and therapeutic implications. J. Pharm. Sci.
2007, 96, 729.
(32) Mitsunobu, O. The use of diethyl azodicarboxylate and triphenylphos-
phine in synthesis and transformation of natural products. Synthesis
1981, 1, 28.
(33) Zhou, T.; Neubert, H.; Liu, D. Y.; Liu, Z. D.; Ma, Y.; et al. Iron
binding dendrimers: a novel approach for the treatment of haemo-
chromatosis. J. Med. Chem. 2006, 49, 4171–4182.
(34) Rae, T. D. Undetectable intracellular free copper: the requirement of
a copper chaperone for superoxide dismutase. Science 1999, 284, 805–
808.
(18) Kurz, T.; Gustafsson, B.; Brunk, U. T. Intralysosomal iron chelation
protects against oxidative stress-induced cellular damage. FEBS J.
2006, 273, 3106–3117.
(19) Persson, H. L.; Yu, Z.; Tirosh, O.; Eaton, J. W.; Brunk, U. T.
Prevention of oxidant-induced cell death by lysosomotropic iron
chelators. Free Radical Biol. Med. 2003, 34, 1295–1305.
(20) Horackova, M.; Ponka, P.; Byczko, Z. The antioxidant effects of a
novel iron chelator salicylaldehyde isonicotinoyl hydrazone in the
prevention of H2O2 injury in adult cardiomyocytes. CardioVasc. Res.
2000, 47, 529–536.
(21) Simunek, T.; Boer, C.; Bouwman, R. A.; Vlasblom, R.; Versteilen,
A. M. G.; et al. SIH, a novel lipophilic iron chelator, protects H9c2
cardiomyoblasts from oxidative stress-induced mitochondrial injury
and cell death. J. Mol. Cell. Cardiol. 2005, 39, 345–354.
(22) Petrat, F.; Rauen, U.; De Groot, H. Determination of the chelatable
iron pool of isolated rat hepatocytes by digital fluorescence microscopy
using the fluorescent probe phen green SK. Hepatology 1999, 29,
1171–1179.
(23) Petrat, F.; De Groot, H.; Rauen, U. Determination of the chelatable
iron pool of single intact cells by laser scanning microscopy. Arch.
Biochem. Biophys. 2000, 376, 74–81.
(24) Glickstein, H.; El, R. B.; Link, G.; Breuer, W.; Konijn, A. M.; et al.
Action of chelators in iron-loaded cardiac cells: accessibility to
intracellular labile iron and functional consequences. Blood 2006, 108,
3195–3203.
(25) Glickstein, H.; El, R. B.; Shvartsman, M.; Cabantchik, Z. I. Intracellular
labile iron pools as direct targets of iron chelators: a fluorescence study
of chelator action in living cells. Blood 2005, 106, 3242–3250.
(26) Petrat, F.; Weisheit, D.; Lensen, M.; De Groot, H.; Sustmann, R.; et
al. Selective determination of mitochondrial chelatable iron in viable
cells with a new fluorescent sensor. Biochem. J. 2002, 362, 137–147.
(27) Rauen, U; Springer, A; Weisheit, D; Petrat, F; H.G, K. Assessment
of chelatable mitochondrial iron by using mitochondrion-selective
fluorescent iron indicators with different iron-binding affinities.
ChemBioChem 2007, 8, 341–352.
(35) Liu, Z. D.; Khodr, H. H.; Liu, D. Y.; Lu, S. L.; Hider, R. C. Synthesis,
physicochemical characterization, and biological evaluation of 2-(1′-
hydroxyalkyl)-3-hydroxypyridin-4-ones: novel iron chelators with
enhanced pFe3+ values. J. Med. Chem. 1999, 42, 4814–4823.
(36) Wis Vitolo, L. M.; Hefter, G. T.; Clare, B. W.; Webb, J. Iron chelators
of the pyridoxal isonicotinoyl hydrazone class part II. Formation
constants with iron(III) and iron(II). Inorg. Chim. Acta 1990, 170,
171–176.
(37) Zanker, V.; Peter, W. Die prototropen formen des fluoresceins. Chem.
Ber. 1958, 91, 572–580.
(38) Sjo¨back, R.; Nygren, J.; Kubista, M. Absorption and fluorescence of
fluorescein. Spectrochim. Acta A 1995, 51, L7–L21.
(39) Alvarez-Pez, J. M.; Ballesteros, L.; Talavera, E.; Yguerabide, J.
Fluorescein excited-state proton exchange reactions: nanosecond
emission kinetics and correlation with steady-state flurescence intensity.
J. Phys. Chem. A 2001, 105, 6320–6332.
(40) Klonis, N.; Sawyer, W. H. Spectral properties of the prototropic forms
of fluorescein in aqueous solution. J. Fluoresc. 1996, 6, 147–157.
(41) Liu, Z. D.; Liu, D. Y.; Hider, R. C. Iron Chelator Chemistry. Iron
Chelation Therapy; Kluwer Academic/Plenum Publishers: New York,
2002; pp 141-166.
(42) Fakih, S.; Podinovskaia, M.; Kong, X. L.; Schaible, U. E.; Collins,
H. L.; Monitoring intracellular labile iron pools: a novel fluorescent
iron(III) sensor as a potentially advanced and non-invasive diagnosis
tool. J. Pharm. Sci., in press.
(43) Tenopoulou, M.; Kurz, T.; Doulias, P. T.; Galaris, D.; Brunk, U. T.
Does the calcein-AM method assay the total cellular “labile iron pool”
or only a fraction of it? Biochem. J. 2007, 403, 261.
(44) Gans, P.; Sabatini, A.; Vacca, A. Determination of equilibrium
constants from spectrophotometric data obtained from solutions of
known pH: the program pHab. Ann. Chim. 1999, 89, 45–49.
(45) Gans, P.; O’Sullivan, B. GLEE, a new computer program for glass
electrode calibration. Talanta 2000, 51, 33–37.
(28) Ma, Y.; Luo, W.; Quinn, P. J.; Liu, Z. D.; Hider, R. C. Design,
synthesis, physicochemical properties, and evaluation of novel iron
JM8001247