Pyridoxal isonicotinoyl hydrazone analogs induce apoptosis in hematopoietic cells due to their iron-chelating properties

Article abstract of DOI:10.1016/S0006-2952(02)01512-5

Analogs of pyridoxal isonicotinoyl hydrazone (PIH) are of interest as iron chelators for the treatment of secondary iron overload and cancer. PIH, salicylaldehyde isonicotinoyl hydrazone (SIH), and 2-hydroxy-1-naphthylaldehyde isonicotinoyl hydrazone (NIH), the toxicity of which vary over two orders of magnitude, were selected for a study of their mechanisms of toxicity. PIH analogs and their iron complexes caused concentration- and time-dependent apoptosis in Jurkat T lymphocytes and K562 cells. Bcl-2 overexpression was partially anti-apoptotic, suggesting mitochondrial mediation of apoptosis. Since the pan-caspase inhibitor zVAD-fmk did not reduce lysosomal and mitochondrial destabilization, these events occur upstream of caspase activation. In contrast, phosphatidylserine externalization and the development of apoptotic morphology were inhibited significantly, indicating the role of caspases in mediating these later events. Since overexpression of CrmA had no effect on apoptosis, caspase-8 is not likely involved. Fe³⁺ complexes of SIH and NIH, which accumulated in ⁵⁹Fe-labeled mouse reticulocytes during incubation with the chelators, also caused apoptosis. BSA, which promotes release of the complexes from cells, reduced the toxicity of SIH and NIH, suggesting that the induction of apoptosis by PIH analogs involves toxic effects mediated by their Fe³⁺ complexes. Moreover, analogs of these agents lacking the iron-chelating moiety were non-toxic.

Full text of DOI:10.1016/S0006-2952(02)01512-5

Biochemical Pharmacology 65 (2003) 161±172
Pyridoxal isonicotinoyl hydrazone analogs induce apoptosis in
hematopoietic cells due to their iron-chelating properties
Joan L. Bussa, Jiri Neuzilb,c,1, Nina Gellertc, Christian Weberc,d, Prem Ponkaa,*
aDepartments of Physiology and Medicine, McGill University, and Lady Davis Institute for Medical Research,
Sir Mortimer B. Davis Jewish General Hospital, 3755 chemin de la Cote-Ste-Catherine, Montreal, Que., Canada H3T 1E2
b
È
Division of Pathology II, Faculty of Health Sciences, University Hospital, Linkoping, Sweden
cInstitute for Prevention of Cardiovascular Diseases, Ludwig Maximilians University, Munich, Germany
dDepartment of Cardiovascular Molecular Biology, University Hospital, Aachen, Germany
Received 5 December 2001; accepted 13 June 2002
Abstract
Analogs of pyridoxal isonicotinoyl hydrazone (PIH) are of interest as iron chelators for the treatment of secondary iron overload and
cancer. PIH, salicylaldehyde isonicotinoyl hydrazone (SIH), and 2-hydroxy-1-naphthylaldehyde isonicotinoyl hydrazone (NIH), the
toxicity of which vary over two orders of magnitude, were selected for a study of their mechanisms of toxicity. PIH analogs and their iron
complexes caused concentration- and time-dependent apoptosis in Jurkat T lymphocytes and K562 cells. Bcl-2 overexpression was
partially anti-apoptotic, suggesting mitochondrial mediation of apoptosis. Since the pan-caspase inhibitor zVAD-fmk did not reduce
lysosomal and mitochondrial destabilization, these events occur upstream of caspase activation. In contrast, phosphatidylserine
externalization and the development of apoptotic morphology were inhibited signi®cantly, indicating the role of caspases in mediating
these later events. Since overexpression of CrmA had no effect on apoptosis, caspase-8 is not likely involved. Fe3‡ complexes of SIH and
NIH, which accumulated in 59Fe-labeled mouse reticulocytes during incubation with the chelators, also caused apoptosis. BSA, which
promotes release of the complexes from cells, reduced the toxicity of SIH and NIH, suggesting that the induction of apoptosis by PIH
analogs involves toxic effects mediated by their Fe3‡ complexes. Moreover, analogs of these agents lacking the iron-chelating moiety
were non-toxic.
# 2002 Published by Elsevier Science Inc.
Keywords: Apoptosis; Pyridoxal isonicotinoyl hydrazone; Jurkat; K562; Iron chelators
1. Introduction
the production of reactive oxygen species that damage cell
membranes, nucleic acids, and proteins. Excretion of iron is
limited to cell desquamation, as there is no speci®c pathway
by which it is eliminated. Iron homeostasis in the organism
depends upon ef®cient recycling of its body stores and
tightly controlled intestinal absorption. When iron intake
exceeds the requirements of the organism, as is the case for
hereditary hemochromatosis and secondary iron-overload
disorders, it gradually accumulates in the parenchyma.
When the capacity to safely store iron in ferritin is
exceeded, the target organs, namely the liver, kidney, and
heart, begin to exhibit pathophysiology characteristic of
iron overload. The observed toxicity of iron is thought to be
due to iron-based oxidative stress. There is increasing
evidence that iron may play a role in the pathophysiology
of numerous conditions in individuals who are not iron-
overloaded, including ischemia/reperfusion injuries such as
Iron is essential for life due to its roles in oxygen
transport and one-electron redox chemistry. Complex
mechanisms have evolved to allow delivery of iron to
tissues [1], which prevent it from acting as a catalyst for
* Corresponding author. Tel.: ‡1-514-340-8222, Ext. 5289;
fax: ‡1-514-340-7502.
E-mail address: prem.ponka@mcgill.ca (P. Ponka).
1 Present address: Heart Foundation Research Center, School of Health
Sciences, Grif®th University, Southport, Queensland, Australia.
Abbreviations: AO, acridine orange; BBH, benzaldehyde benzoyl
hydrazone; BIH, benzaldehyde isonicotinoyl hydrazone; DFO, desferriox-
amine; FBS, fetal bovine serum; IRP, iron regulatory protein; MSDH,
O-methylserine dodecylamide hydrochloride; MTS, 3-(4,5-dimethylthiazol-
2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium; NIH,
2-hydroxy-1-naphthylaldehyde isonicotinoyl hydrazone; PIH, pyridoxal
isonicotinoyl hydrazone; SIH, salicylaldehyde isonicotinoyl hydrazone.
0006-2952/02/$ ± see front matter # 2002 Published by Elsevier Science Inc.
PII: S 0 0 0 6 - 2 9 5 2 ( 0 2 ) 0 1 5 1 2 - 5

162
J.L. Buss et al. / Biochemical Pharmacology 65 (2003) 161±172
stroke and myocardial infarction, in¯ammatory diseases,
and cancer. Although iron chelators are used to treat
secondary iron overload, they have not been studied exten-
sively as therapeutic agents to treat free radical-mediated
tissue damage, despite their potential value.
purities of both compounds were established by 1H NMR.
Stock solutions of 2 mM were made in RPMI-1640 med-
ium containing 10% FBS and stored at À208 until used.
2.2. Preparation of chelators and their Fe3‡ complexes
Secondary iron-overload disorders, such as b-thalasse-
mia, are caused by the hyperabsorption of iron and chronic
blood transfusions, which are required as a source of
functional erythrocytes. Concomitant chelation therapy
is required for these patients to prevent liver and cardiac
dysfunction, which occur by the age of 25 in the absence of
iron-chelation therapy. At present, the only clinically
accepted drug for this purpose is DFO which, due to its
dosage form, is costly and inconvenient. There is, there-
fore, an ongoing search for an orally available iron-che-
lator, which would result in its accessibility to a greater
number of patients.
In biological systems, iron chelators have multiple
effects, which may or may not contribute to their effec-
tiveness and/or toxicity. An improved understanding of the
properties required of effective chelators will improve the
screens for identi®cation of therapeutically valuable com-
pounds. Several iron chelators have been shown to induce
apoptosis [2±6]. Iron chelators have many biological
effects, which may result in apoptotic cell death, including
inhibition of ribonucleotide reductase, redistribution of
cellular iron, and formation of iron complexes that cause
oxidative stress.
PIH, SIH, and NIH were prepared as previously
described [14]. Stock solutions of 5 or 10 mM were pre-
pared in 50 or 100 mM NaOH, and stored at À208 until
used. Storage under these conditions did not result in
signi®cant hydrolysis over 6 months. Fe3‡ complexes of
the chelators were prepared at a minimum concentration of
25 mM by the addition of 5 mM FeCl3 (Fisher) in 100 mM
sodium citrate (Bioshop) and chelator in a 1:2 molar ratio,
and incubated at room temperature for 60 min. Chelator
stability under the storage conditions described and com-
plex formation with Fe3‡ were assessed at 25 and 378 using
a Cary 1 spectrophotometer.
2.3. Cell culture
Jurkat T lymphoma cells, their mutants overexpressing
Bcl-2 [15] or CrmA [16], and K562 cells were maintained
in RPMI-1640 medium supplemented with 10% FBS.
Anti-Fas IgM (Immunotech) was used at 20 ng/mL to
induce apoptosis in control and Bcl-2- and CrmA-over-
expressing Jurkat cells.
PIH is among the most effective orally available iron
chelators known. It effectively causes iron excretion in
humans [7] and rats [8], demonstrating its capacity to
access biological iron rapidly. These effects were observed
at doses that caused no signi®cant adverse effects, and were
2.4. Cell toxicity experiments
K562 and Jurkat T cells were cultured in phenol red-free
RPMI-1640 medium (Cellgro) supplemented with 10%
FBS (Gibco) and 2 mM glutamine (Gibco). K562 or Jurkat
cells from log phase cultures were plated in 96-well plates
at 14,000 or 20,000 cells/well, respectively, in a ®nal
volume of 200 mL/well. Cell numbers were assessed using
either trypan blue staining or the MTS assay (Promega). In
the latter assay of mitochondrial dehydrogenase activity,
phenomethazine sulfonate (Aldrich) mediates the reduc-
tion of MTS by cells to a compound that can be detected
spectrophotometrically. Absorbance measurements were
made at 490 nm, and corrected for background absorbance
at 630 nm, on an ELx800 microplate reader (Bio-tek
Instruments) and, when necessary, were corrected for
the absorbance of the iron-chelator complexes. Calibra-
tion curves of both cell types were linear over the range
used in the toxicity experiments described, and treatment
of cells with the chelators did not affect MTS reduction
(data not shown). This was also the case for the treatment
of Jurkat cells with iron complexes. Treatment of K562
cells with iron-chelator complexes appears to enhance
their capacity to reduce MTS, since absorbance measure-
ments were much higher than expected based on cell
number [17]. For this reason, the toxicity of iron-chelator
complexes toward K562 cells was not assessed in this
study.
well below LD values established in rats [9]. In vitro
50
[10,11] and in vivo [12] studies have identi®ed analogs that
are far more effective than PIH itself. Previous studies have
also demonstrated that the factors governing the toxicity of
these chelators, which varies over two orders of magnitude
[10], are complex and do not depend simply on the amount
of iron mobilized [13]. Thus, PIH and its analogs warrant
further study as a class of iron chelators that may be of
value in treating diseases in which iron plays a signi®cant
role.
2. Materials and methods
2.1. Synthesis of BIH and BBH
BIH and BBH were synthesized using the general
method previously described [14]. Brie¯y, 0.01 mol ben-
zaldehyde and 0.01 mol isonicotinoyl or benzoyl hydra-
zide (all from Lancaster) were dissolved in 175 mL of 66%
aqueous ethanol, re¯uxed for 30 min, cooled, ®ltered, and
washed with water to yield BIH and BBH, respectively.
BIH was recrystallized from ethanol. The identities and

J.L. Buss et al. / Biochemical Pharmacology 65 (2003) 161±172
163
2.5. Mobilization of 59Fe from mouse reticulocytes
JC-1 in RPMI medium for 15 min at 378, washed with PBS,
and analyzed by ¯ow cytometry. The lysosomal stability
assay was performed using the weak base AO as described
[19]. Cells were washed and resuspended in 2.5 mL RPMI
medium with 10 mM HEPES and 5 mg/mL of AO, incu-
bated at 378 for 15 min in the dark, washed, and resus-
pended in PBS; the loss of red ¯uorescence was estimated
using ¯ow cytometry. Flow cytometric data measuring
mitochondrial destabilization (DCm) using JC-1 and lyso-
somal destabilization using AO were analyzed similarly to
the annexin-V data. For both DCm and lysosomal desta-
bilization, chelator treatment resulted in a loss of ¯uores-
cence as compared with controls. The gate was set on the
control cells, and cells with ¯uorescence below this range
were de®ned as having ``low DCm'' or ``pale'' lysosomes
for these assays, respectively.
Caspase-3 activity was measured essentially as pub-
lished elsewhere [21]. In brief, 106 cells were lysed in
100 mL lysis buffer (10 mM Tris±HCl, pH 7.4, 130 mM
NaCl, 0.1% Triton X-100, 10 mM NaH2PO4), mixed with
900 mL reaction buffer (20 mM HEPES, pH 7.4, 10%
glycerol, 2 mM dithiothreitol), and incubated for 2 hr at
378 with the substrate for caspase-3, Ac-DEVD-AMC
(Calbiochem), at a concentration of 50 mM. The ¯uores-
cence of the liberated 7-amino-4-methylcoumarin (AMC)
was then measured at lex ˆ 380 nm and lem ˆ 435 nm.
Apoptotic cellular morphology was assessed by Giemsa
staining, performed according to a standard protocol.
Brie¯y, cells were centrifuged, incubated with Giemsa
for 15 min, and washed. Cells were mounted in Mowiol
and examined using light microscopy. Photographs were
taken with an RGB camera using Spot32 software and
processed in PhotoShop. The intact nuclei of normal cells,
which are stained blue, were easily distinguished from the
condensed, fragmented, less intensely stained nuclei of
apoptotic cells. Approximately 100 cells were evaluated
per sample, and were scored as ``apoptotic'' or ``normal''
based on the presence or absence of condensed nuclei. Cell
treatments were assessed in triplicate samples.
Reticulocyte-enriched mouse blood labeled with 59Fe
was prepared. Brie¯y, reticulocytosis was induced in CD1
mice by intraperitoneal injection of 50 mg/kg of phenyl-
hydrazine on three consecutive days, and blood was col-
lected 3 days after the last injection. A 30% cell suspension
was incubated in 1 mM succinylacetone, which inhibits
heme biosynthesis, at 378 for 30 min, at which time 10 mM
human 59Fe2-transferrin (Sigma) [18] was added, and the
incubation continued for 60 min. Cells were washed, and
the total 59Fe taken up by the cells was measured by g-
counting, corresponding to 100% 59Fe. Packed cells
(35 mL) were incubated with the chelators in a ®nal volume
of 500 mL MEM for 2 hr. Cells were centrifuged, and the
total radioactivity in the supernatants was calculated by g-
counting small samples to determine the amounts of 59Fe
released from the cells. Cells were washed and lysed
osmotically by the addition of 200 mL of distilled water.
Intracellular 59Fe in ethanol-soluble forms was separated
from protein-associated 59Fe by the precipitation of pro-
teins using 1 mL of ice-cold ethanol, and the total radio-
activity in both fractions was g-counted. Thus, binding of
59Fe by the chelators (the sum of 59Fe in the supernatant
and the ethanol-soluble fraction), and release of the result-
ing 59Fe-chelator complexes from the cells (59Fe in the
supernatant) were measured. Under these conditions, cell
viability in the presence of PIH analogs at concentrations
of 50 mM was over 90% for all treatments, as assessed by
the absence of hemoglobin release into the extracellular
medium.
2.6. Assessment of apoptotic markers
Cells were seeded at 0:5 Â 106/mL in RPMI medium
containing 10% FBS and treated as indicated. Phosphati-
dylserine externalization, loss of the inner mitochondrial
membrane potential (DCm), and lysosomal destabilization
were assessed by ¯ow cytometry as described below.
Phosphatidylserine externalization was measured by the
binding of annexin-V-FITC (PharMingen) [19]. Brie¯y,
cells were washed in PBS, resuspended in binding buffer
(10 mM HEPES/NaOH, pH 7.4, 140 mM NaCl, 2.5 mM
CaCl2), incubated for 20 min at room temperature with
2 mL annexin-V-FITC, and analyzed by ¯ow cytometry
using a FACScan instrument (Becton Dickinson), which
measures the ¯uorescence intensity of individual cells.
Since a small percentage of the control cells were highly
¯uorescent, the gating was set to encompass the approxi-
mately 95% of cells in the low-¯uorescence peak. Cells
with ¯uorescence above this range were considered apop-
totic.
3. Results
3.1. Antiproliferative effects of the chelators
PIH, and two of its analogs, SIH and NIH, whose
structures are shown in Fig. 1, were chosen for this study
as they are among the best characterized of the many
analogs synthesized in this laboratory [14]. Despite their
structural similarity, their toxicities, as measured in a
neuroblastoma cell line [10], range from the least toxic,
PIH, to one of the most toxic, NIH.
The mitochondrial inner membrane potential (DCm)
was assessed using 5,50,6,60-tetrachloro-1,10,3,30-tetra-
ethylbenzimidazolyl-carbocyanine iodide (JC-1, Molecu-
lar Probes) [20]. Cells were washed, incubated with 10 mM
PIH and two of its analogs evaluated in this study (Fig. 1)
had antiproliferative effects on K562 cells (Fig. 2A), which
were assessed after a 72-hr incubation. Concentration±
response data, as measured by the MTS assay, were ®tted

164
J.L. Buss et al. / Biochemical Pharmacology 65 (2003) 161±172
Fig. 1. Structures of the PIH analogs examined in this study.
to a three-parameter logistic equation to obtain IC
values, at which cell growth is half that of untreated cells
(Cellsmax):
50
Cellsmax
b
Relative cell number ˆ
(1)
ꢀc=ic50† ‡ 1
in which c is the chelator concentration, and b is the Hill-
type coef®cient. PIH, which has been demonstrated to be
well tolerated in vivo [8,9,12], had an IC of 90 mM, and
50
was the least toxic, whereas the IC50 values of SIH and NIH
were 12 and 1 mM, respectively.
The toxicity of PIH analogs toward Jurkat cells (Fig. 2B)
was similar to the results obtained with K562 cells,
although the IC values were approximately half those
50
obtained using K562 cells, indicating the relative sensitiv-
ity of Jurkat cells to the toxic effects of the chelators. As
was the case for neuroblastoma cells [10], toxicity of the
chelators increased in the order PIH < SIH < NIH. This
trend has been observed previously in K562 cells [17] and
in a neuroblastoma cell line [10], and is consistent with the
hypothesis that the toxicity of PIH analogs is correlated
with their lipophilicity [17]. The antiproliferative effects of
PIH, SIH, and NIH on Jurkat cells were time-dependent
(Fig. 3).
Fig. 2. Antiproliferative effects of PIH, SIH, NIH, and their Fe3‡
complexes. Cells were incubated for 72 hr with PIH (*), SIH (&), NIH
(~), or their corresponding Fe3‡ complexes. Cell viability was assessed
spectrophotometrically, using the MTS assay. (A) K562 cells treated with
chelators. (B) Jurkat cells treated with chelators. (C) Jurkat cells treated
with iron-chelator complexes. Data points are the averages of six replicates
that were normalized with respect to untreated cells, and the error bars
represent standard deviations. Solid lines represent non-linear weighted
least squares fits of the data to Eq. (1), except in the case of FePIH2
treatment of Jurkat cells, which caused no cell death.
3.2. Characterization of PIH analogs that do not
mobilize iron
PIH analogs have been shown to have a number of
effects on cells, including inhibition of transferrin-mediated

J.L. Buss et al. / Biochemical Pharmacology 65 (2003) 161±172
165
Fig. 3. Time- and concentration-dependence of PIH (A), SIH (B), and NIH (C) toxicity toward Jurkat cells. Cell death was assessed by trypan blue staining at
24 hr (*), 48 hr (*), and 72 hr (!). Results shown are from a single experiment, which was performed twice.
iron uptake [10], inhibition of DNA synthesis [13], and
mobilization of cellular iron [10,18], all of which have
been explained by their direct interaction with iron. It is
unknown, however, whether this class of iron chelators
may cause toxicity via other biological effects. Two ana-
logs, BIH and BBH, the structures of which are shown in
Fig. 1, were synthesized. These compounds are similar to
SIH, but lack the phenolic group that is essential for strong
iron af®nity, as demonstrated in X-ray crystallography
[22], and by the shifts in infrared spectral bands that occur
when PIH analogs bind iron [23]. That BIH and BBH do
not bind iron was con®rmed spectrophotometrically by the
absence of ligand-to-metal charge transfer absorption,
which is characteristic of PIH analogs [23,24].
The antiproliferative effects of BIH and BBH toward
K562 cells were assessed over 72 hr, under conditions
identical to those of Fig. 2A (data not shown). BIH, which
was tested up to 1000 mM, was not toxic. The solubility
limit of BBH was approximately 200 mM, below which no
toxicity was observed. The toxicity of iron-citrate (pre-
pared as described in Section 2) toward K562 cells was
unaffected by the presence of 100 mM BIH or BBH (data
not shown). Thus, it appears that the complete iron-binding
motif, consisting of the phenolic oxygen, imine nitrogen,
and carbonyl oxygen atoms, is necessary for the antipro-
liferative activity of PIH analogs.
59Fe in reticulocytes in which heme synthesis was inhibited
with succinylacetone (Fig. 4); in this experiment the total
59Fe taken up by the cells during the 1-hr incubation with
59Fe2-transferrin corresponds to 100%. The amounts of
59Fe bound by the chelators are represented by the total
height of the bars, which are divided into 59Fe released into
the extracellular medium (dark shading) and intracellular,
ethanol-soluble 59Fe (light shading), which is presumably
in the form of 59Fe-chelator complexes.
As previously observed [24], PIH bound less 59Fe than
did its analogs. Although the total amounts of 59Fe bound
by SIH and NIH were high (Fig. 4), as is the case for most
PIH analogs [24], the fraction remaining in the cell (light
shading) was much higher than for the most effective
analogs (approximately 1% under the same experimental
conditions [24]). The accumulation of 59Fe-chelator com-
plexes in the cytosol, which appears to be due to the
lipophilicity of the complexes [17], is correlated with
toxicity, suggesting that intracellular iron-chelator com-
plexes may have some cytotoxic effects.
3.3. 59Fe mobilization from reticulocytes by PIH analogs
The capacity of iron chelators to mobilize intracellular
iron was assessed using murine reticulocytes labeled with
59Fe [18]. In this model, heme synthesis is inhibited by the
incubation of reticulocytes with succinylacetone, which
inhibits d-aminolevulinic acid synthase, and 59Fe2-trans-
ferrin, the biological iron donor. The ef®cacy of chelators
at chelating intracellular iron and causing its release from
cells may be evaluated in this system. PIH analogs have
been studied using similar models [11], and the results
agree well with the capacity of the chelators to cause the
excretion of 59Fe in rats [12]. PIH, SIH, and NIH bound
Fig. 4. Mobilization of 59Fe from reticulocytes. Mouse reticulocytes were
labeled with non-heme 59Fe as described in Section 2 and incubated with
50 mM iron chelators for 2 hr. Dark bars represent 59Fe bound and released
into the extracellular medium, while light bars represent 59Fe bound inside
the cells. Data are expressed as a percentage of the total 59Fe taken up by
the cells during the labeling step. Data are the averages of triplicate
samples, and standard deviations were not more than 2%.

166
J.L. Buss et al. / Biochemical Pharmacology 65 (2003) 161±172
As expected, neither BIH nor BBH bound signi®cant
amounts of 59Fe (Fig. 4). It is unlikely that this is due to an
inability of these compounds to penetrate the cell mem-
brane, since their structures are similar to that of SIH,
which readily crosses membranes to access 59Fe (Fig. 4).
Since these analogs do not bind iron strongly, they are
unable to mobilize 59Fe, and consequently have no anti-
proliferative effects. This strongly corroborates the hypoth-
esis that iron binding is central to the biological activities
of PIH analogs.
3.4. Iron chelators and apoptosis
The preceding experiments did not distinguish among
the three mechanisms by which PIH analogs may affect
cell growth: reduced growth rate, necrosis, and apoptosis.
From the data in Fig. 3, concentrations of the chelators
were chosen that caused signi®cant antiproliferative
effects: 100, 20, and 2 mM for PIH, SIH, and NIH, respec-
tively. The death pathway taken by Jurkat cells was
evaluated using two apoptotic markers: annexin-V binding,
which measures phosphatidylserine externalization, and
morphological evaluation following Giemsa staining. At
all time points, the percentage of cells that were positive for
markers of apoptosis was greater than or equal to that of
dead cells (Fig. 5). Thus, at these concentrations of the
chelators, apoptosis is the major, if not the only, mechan-
ism of cell death triggered by PIH analogs.
Several other apoptotic and pro-apoptotic markers were
evaluated after treatment of Jurkat cells with 100, 20, and
2 mM PIH, SIH, and NIH, respectively (Fig. 6). Mitochon-
drial destabilization, which occurs as a result of the open-
ing of the mitochondrial permeability transition (MPT)
pore, was measured using JC-1 [19], which shifts from red
¯uorescence to green when the mitochondrial potential
drops below 100 mV, due to disaggregation of the dye. The
percentage of cells containing green, destabilized mito-
chondria increased in a time-dependent manner (Fig. 6A).
Fig. 6. Apoptosis induction in Jurkat cells by PIH analogs. Apoptosis was
assessed in control cells (*), and at 100 mM PIH (*), 20 mM SIH (!),
and 2 mM NIH (5) by measuring the collapse of the mitochondrial
membrane potential (DCm) with JC-1 (A), lysosomal destabilization with
acridine orange (B), caspase-3 activity with Ac-DEVD-AMC (C),
phosphatidylserine externalization with annexin-V-FITC (D), and mor-
phological changes by Giemsa staining (E). Data points are the averages of
three replicates, and error bars represent standard deviations. For some
data points, error bars are smaller than the symbols.
Fig. 5. Correlation of Jurkat cell death induced by PIH analogs with apoptosis markers. Cell death, assessed by trypan blue staining (*), phosphatidylserine
externalization, assessed by annexin-V-FITC binding (*), and morphological changes, assessed by Giemsa staining (!), were measured over time at
100 mM PIH (A), 20 mM SIH (B), and 2 mM NIH (C). Data shown are from a single experiment, which was performed three times.

J.L. Buss et al. / Biochemical Pharmacology 65 (2003) 161±172
167
The ¯uorophore AO is a weak base that accumulates in
the acidic environment of the lysosome. A decrease in AO
¯uorescence, detectable by ¯ow cytometry, indicates the
loss of the lysosomal proton gradient [25], and is thus a
marker of lysosomal membrane integrity. Incubation of
Jurkat cells with PIH analogs resulted in a signi®cant
increase in the number of ``pale cells,'' indicating that
these chelators cause lysosomal destabilization (Fig. 6B).
Other cytotoxic stimuli also have this effect on lysosomes,
including the redox-cycling quinone, naphthazirin [26], the
lysosome-speci®c detergent MSDH [27], lysosomally tar-
geted photo-oxidative damage [28], and Fas/APO-1 [29].
Similar effects of chelators on K562 cells were observed
for all apoptotic markers (Fig. 7), although the extent of
apoptosis was generally lower, due to the greater resistance
of these cells to the antiproliferative effects of these
chelators (Fig. 2). Because they had similar effects on
two different cell lines, the observed induction of apoptosis
by the chelators is likely to be a general effect on different
cell types.
The sequence of events in the apoptotic cascade was
probed by assessing various markers of apoptosis in the
presence and absence of the pan-caspase inhibitor zVAD-
fmk (Fig. 8). At equitoxic concentrations, all analogs had
similar effects on both cell lines after a 48-hr incubation.
As was shown in Figs. 6 and 7 for Jurkat and K562 cells,
respectively, incubation of cells with PIH analogs caused
mitochondrial and lysosomal destabilization, caspase-3
activation, phosphatidylserine externalization, and mor-
phological changes associated with apoptosis (Fig. 8). In
the presence of the caspase inhibitor zVAD-fmk, lysosomal
and mitochondrial effects were unchanged (Student's t-
test, P > 0:05 in all cases), indicating that these events
occur upstream of caspase activation. Phosphatidylserine
externalization and morphological alterations, however,
were inhibited signi®cantly by zVAD-fmk: P < 0:05 in
all cases except for the assessment of K562 cells with SIH,
in which the P-value for phosphatidylserine externalization
was 0.051. One explanation for these observations is that
these events are caspase-dependent, and therefore occur
downstream of caspase activation in both Jurkat and K562
cells. However, 25 mM zVAD-fmk was unable to protect
Jurkat cells from the toxicity of PIH analogs over a 72-hr
incubation (data not shown). Therefore, it is possible that
apoptosis may proceed through multiple mechanisms,
including a caspase-independent pathway such as apopto-
sis-inducing factor [30], or that in the absence of caspase
activity damaged cells die by necrosis.
To further characterize the pathway by which apoptosis
is induced in response to PIH analogs, Jurkat cells trans-
fected with CrmA [15], an inhibitor of caspase-8 [31], were
examined. As a positive control, the inhibitory effect of
CrmA overexpression on apoptosis induced by anti-Fas
IgM was con®rmed (Fig. 9). Wild-type Jurkat cells were
sensitive to apoptosis induced by anti-Fas IgM; 68 Æ 7% of
the cells were annexin-positive, whereas apoptosis in
CrmA-overexpressing cells was almost completely inhib-
ited. In contrast, the extent of apoptosis induced by PIH,
SIH, or NIH in CrmA-overexpressing cells was not sig-
ni®cantly different from that in control cells (Fig. 9). Thus,
as is the case for many chemical inducers of apoptosis, the
main pro-apoptotic pathway appears to be independent of
caspase-8.
Jurkat cells overexpressing Bcl-2 [15], an anti-apoptotic
mitochondrial protein [32], were partially protected
against anti-Fas, IgM-induced apoptosis. This cell line
was also relatively resistant to apoptosis induced by
PIH, SIH, and NIH (Fig. 9); approximately 45% of
wild-type cells were annexin-positive, as compared to
approximately 17% of Bcl-2 overexpressing cells. This
effect of Bcl-2 demonstrates the importance of mitochon-
dria in the pathway of apoptosis induced by PIH analogs.
Fig. 7. Apoptosis induction in K562 cells by PIH analogs. Apoptosis was
assessed in control cells (*), and at 100 mM PIH (*), 20 mM SIH (!),
and 2 mM NIH (5) by measuring the collapse of the mitochondrial
membrane potential (DCm) with JC-1 (A), lysosomal destabilization with
acridine orange (B), caspase-3 activity with Ac-DEVD-AMC (C),
phosphatidylserine externalization with annexin-V-FITC (D), and mor-
phological changes by Giemsa staining (E). Data points are the averages of
three replicates, and error bars represent standard deviations. For some
data points, error bars are smaller than the symbols.

168
J.L. Buss et al. / Biochemical Pharmacology 65 (2003) 161±172
Fig. 8. Effect of zVAD-fmk on the inhibition of apoptotic events induced by PIH analogs. Jurkat (A±C) and K562 (D±F) cells were incubated for 48 hr with
100 mM PIH (A and D), 20 mM SIH (B and E), and 2 mM NIH (C and F) in the absence (dark bars) and presence (light bars) of 25 mM zVAD-fmk. The
following apoptotic markers were assessed: loss of the mitochondrial membrane potential (DCm), lysosomal destabilization (AO), caspase-3 activity (C-3),
morphological changes (M), and phosphatidylserine externalization (PS). The data represent the averages of three replicates, and the error bars represent
standard deviations. In no case was DCm or AO significantly different in the presence and absence of zVAD-fmk. C-3, M, and PS were significantly different
(P < 0:05) in each case, except in E, in which P ˆ 0:051 (Student's t-test) for PS.
3.5. Apoptosis induction by Fe3‡-chelator complexes
Table 1
Induction of apoptosis by the Fe3‡ complexes of PIH analogs
Phosphatidylserine externalizationa (%)
Toxicity of the iron-chelator complexes was assessed in
Jurkat cells (Fig. 2C), and analyzed as described for the
24 hr
48 hr
66 hr
free chelators. Toxicity of FePIH2 was not observed below
its solubility limit (approximately 150 mM). FeSIH2 and
Controlb
8.2 Æ 5.1
13.2 Æ 2.8
12.1 Æ 3.5
17.1 Æ 2.1
16.2 Æ 4.1
7.2 Æ 1.2
10.8 Æ 2.1
11.2 Æ 2.5
11.8 Æ 2.3
15.1 Æ 4.2
12.3 Æ 4.1
13.1 Æ 3.2
12.9 Æ 3.7
15.2 Æ 4.1
16.5 Æ 2.9
53.9 Æ 1.8d
28.9 Æ 4.2
62.2 Æ 7.2
92.0 Æ 0.2d
c
37.5 mM FePIH2
75 mM FePIH2
c
FeNIH2 were toxic, with IC values of 77 Æ 5 and
50
85 Æ 7 mM, respectively. That incubation of Jurkat cells
with FeSIH2 and FeNIH2 induced apoptosis in a time- and
concentration-dependent manner, as assessed by annexin-
V binding, is shown in Table 1. Caspase activation paral-
leled annexin-V binding (data not shown). As expected
from the lack of toxicity of FePIH2 toward Jurkat cells
(Fig. 2C), FePIH2 did not induce apoptosis at any con-
centration.
37.5 mM FeSIH2
75 mM FeSIH2
150 mM FeSIH2
37.5 mM FeNIH2
75 mM FeNIH2
150 mM FeNIH2
9.9 Æ 4.2
15.2 Æ 3.8
47.2 Æ 4.9
21.3 Æ 5.8
a Jurkat cells were incubated with Fe3‡ complexes of the ligands at the
indicated concentrations. Annexin-V-FITC binding and caspase activation
were assessed by ¯ow cytometry as described in Section 2. Values are
averages of triplicate measurements, and errors are standard deviations.
b Annexin binding for control cells at time zero was 7:5 Æ 3:5%.
c Concentrations are indicated in terms of Fe.
Evidence for a contribution to the toxicity of PIH
analogs by their iron complexes is demonstrated in
Fig. 10, which shows the protective effect of BSA against
the toxicity of PIH, SIH, and NIH. PIH analogs bind
d These data are from a separate experiment, in which caspase
activation was not assessed.

J.L. Buss et al. / Biochemical Pharmacology 65 (2003) 161±172
169
4. Discussion
4.1. Relationship of the toxicity of PIH analogs to
interaction with intracellular iron
The hypothesis that the biological effects of PIH and its
analogs are related to their effects on cellular iron was
tested by synthesizing and evaluating the effects of BIH
and BBH, PIH analogs that do not bind iron. These ligands
were designed to be as similar as possible to SIH, which
enters reticulocytes readily and binds intracellular 59Fe
(Fig. 4), and activates protein-1 IRP-1 in the macrophage
cell line RAW264.7 [33]. Its Fe3‡ complex [34], as well as
that of PIH [35], delivers iron to cells, supporting cell
growth in the absence of transferrin, the biological iron
donor. No interaction of BIH or BBH with iron was
observed, either spectrophotometrically (data not shown)
or in the reticulocyte 59Fe mobilization assay (Fig. 4). The
structural changes that abrogated iron af®nity also abol-
ished toxicity toward K562 cells. These data demonstrate
that, as hypothesized, iron binding is central to the activity
of PIH analogs.
Fig. 9. Roles of caspase-8 and Bcl-2 in apoptosis induced by PIH analogs.
Jurkat cells (J) overexpressing Bcl-2 (B) or the caspase-8 inhibitor CrmA (C)
were exposed to anti-Fas IgM, 100 mM PIH, 20 mM SIH, and 2 mM NIH for
48 hr. Apoptosis was assessed by phosphatidylserine externalization. Data
represent averages of three replicates, and error bars represent standard
deviations. There was no significant difference among the cell lines in the
absence of the apoptogenic agent. CrmA and Bcl-2 overexpression
significantly protected Jurkat cells against anti-Fas IgM-induced apoptosis
(P < 0:05, one-tailed t-test). Bcl-2 overexpression significantly protected
cells against PIH analog-induced apoptosis (P < 0:05, one-tailed t-test),
whereas CrmA overexpression had no effect.
intracellular iron, resulting in the transient accumulation of
iron-chelator complexes inside cells [24], as shown in
Fig. 4. The rate of diffusion of the complexes through
the cell membrane into the extracellular medium deter-
mines the intracellular concentration of iron-chelator [17].
Because of its high af®nity for iron-chelator complexes
[24], the addition of BSA to the extracellular medium
enhances the release of 59Fe-chelator complexes from the
cell, and therefore reduces the accumulation of intracel-
lular 59Fe-chelator [24]. This effect caused a signi®cant
decrease in the toxicity of PIH analogs, as shown in Fig. 10.
Thus, the toxicity of PIH analogs is due, in part, to the
toxicity of the iron-chelator complexes that are formed
intracellularly.
4.2. PIH analog-induced apoptosis via mitochondrial
damage
The toxicity results of PIH, SIH, and NIH in Jurkat T
lymphocytes and K562 cells found in this study (Fig. 2) are
consistent with the results of a study describing the toxicity
of PIH analogs as assessed in a neuroblastoma cell line
[10], in that SIH and NIH had much stronger antiproli-
ferative effects than did PIH. CrmA overexpression did not
affect the toxicity of PIH analogs (Fig. 9), suggesting that
caspase-8, involved in receptor-mediated apoptosis [36],
does not play a signi®cant role in the induction of apoptosis
by these chelators. Bcl-2 overexpression, on the other hand,
inhibited the progression of chelator-induced apoptosis
Fig. 10. Effect of BSA on the toxicity of PIH analogs toward K562 cells. Cells were exposed to PIH (A), SIH (B), and NIH (C) in the absence (*) and
presence (*) of 5% BSA for 72 hr. Cell viability was assessed using the MTS assay. Data points represent the averages of six replicates, and the error bars
represent standard deviations. Solid lines represent non-linear weighted least squares fits of the data to Eq. (1), except for the treatment with NIH and BSA; in
this case cell viability in the presence of 3 mM NIH was greater than 0.5.

170
J.L. Buss et al. / Biochemical Pharmacology 65 (2003) 161±172
signi®cantly (Fig. 9). Bcl-2 appears to protect cells against
mitochondrially-mediated apoptosis [32] through a bal-
ance of pro- and anti-apoptotic factors. Thus, protection by
Bcl-2 against apoptosis induced by PIH analogs supports
the hypothesis that they act by damaging mitochondria,
which then causes caspase activation.
concentrations below 1 mM, delivered iron to proliferating
mouse lymphocytes in an apparently non-speci®c manner
[41], resulting in the accumulation of iron in pools that
had escaped or overloaded the normal intracellular iron
traf®cking pathways. Non-speci®c iron delivery may
inappropriately regulate biochemical processes that are
sensitive to intracellular iron levels and/or oxidative
stress, such as the IRP system [33] or cyclin-dependent
kinases [42].
Treatment of Jurkat and K562 cells with PIH analogs
caused apoptosis in which lysosomal and mitochondrial
destabilization occurred upstream of, or independently of,
caspase activation, whereas phosphatidylserine externali-
zation and development of apoptotic morphology were
inhibited by the caspase inhibitor zVAD-fmk (Fig. 8). This
sequence of events is consistent with the ``mitochondrial''
pathway of apoptosis [36], in which mitochondrial damage
causes cytochrome c release and caspase activation, which
result in cell death. Together with the protection by Bcl-2
against PIH-induced apoptosis (Fig. 9), these data support
a mechanism of action in which mitochondrial damage and
subsequent cytochrome c release cause caspase activation.
However, since caspase inhibition by zVAD-fmk was
unable to protect Jurkat cells against the toxicity of the
PIH analogs FeSIH2 or FeNIH2 (data not shown), the
possibility that cells die through multiple mechanisms
must be considered.
An early event in apoptosis induced by PIH analogs was
lysosomal destabilization (Fig. 8), which has also been
observed in apoptosis mediated by a number of diverse
stimuli: Fas/APO-1 [29], MSDH [27], and oxidants includ-
ing photo-oxidation [28] and naphthazirin [26,27]. Time±
course studies have demonstrated that lysosomal destabi-
lization is one of the earliest detectable events, involving
the loss of the proton gradient followed by leakage of
lysosomal contents [28], and occurring prior to loss of the
mitochondrial membrane potential [37].
4.4. Mechanisms of toxicity of PIH analogs
The capacity of PIH analogs to mobilize iron in vitro
[10,11] and in vivo [8,12] is well characterized, and iron
chelation may contribute to the toxicity of PIH analogs by
depleting cells of the iron they require for proliferation.
However, iron depletion cannot be the only mechanism by
which PIH analogs are toxic. The capacity of PIH
analogs, aside from PIH itself, which was signi®cantly
less effective than all the analogs tested, to bind 59Fe in
59Fe-labeled reticulocytes and K562 cells was similar
[24] (although there were great differences in the capacity
of the analogs to ef¯ux the 59Fe), whereas the toxicity
of these analogs varied over two orders of magnitude
[17]. Furthermore, 59Fe mobilization from K562 cells
reaches a maximum level at 10 mM for all the chelators
tested [17], whereas the toxicity of some PIH analogs is
not observed until much higher concentrations are
reached. Thus, the effect of diverting iron from its normal
cellular ligands, while it may contribute to the overall
toxicity observed, is insuf®cient to account for either the
observed antiproliferative effects, or the structure±activ-
ity relationships governing the toxicity of this class of
chelators.
The main factor that correlated with the toxicity of PIH
analogs is their lipophilicity, which has been the focus of a
number of studies [17,24,43,44]. PIH analogs appear to
enter cells with ease, as demonstrated by their similar
capacity to bind cellular 59Fe [24]. Thus, the role of
lipophilicity in determining toxicity is unlikely to be due
to the membrane permeability of the chelators. Toxicity is
correlated with high lipophilicity of the Fe3‡ complexes of
PIH analogs, which inhibits their release from the cell [17].
Thus, the simplest explanation for the observed structure±
activity relationships is that the intracellular concentration
of the iron-chelator, at least in part, determines the toxicity.
In support of this hypothesis is the protective effect of BSA
against the toxicity of PIH analogs (Fig. 10). BSA binds
Fe3‡ complexes of PIH analogs tightly, and improves iron
mobilization by PIH analogs in reticulocytes and K562 cells
[24]. BSA presumably acts by providing an extracellular
sink for the lipophilic iron complexes. BSA, at its plasma
concentration, also diminishes the toxicity of all PIH ana-
logs tested in K562 cells (Fig. 10, [17]). This is apparently
the result of the reduction of intracellular levels of the iron
complexes, which occurs because BSA expedites their
4.3. Apoptosis induction by iron-chelator complexes
Treatment of Jurkat T cells with the Fe3‡ complexes of
PIH, SIH, and NIH caused cell death (Fig. 2C), as pre-
viously reported in other cell lines for FePIH2 [38,39] and
the Fe3‡ complexes of other PIH analogs, including SIH
and NIH [17]. Although under the experimental conditions
described, FePIH2 was not toxic below its solubility limit,
the IC50 values of FeSIH2 and FeNIH2 were approximately
77 and 85 mM for Jurkat T cells (Fig. 2C), respectively,
values far higher than those of the free ligands (Fig. 2). As
was the case for the free chelators, the Fe3‡ complexes
caused apoptosis (Table 1). In contrast to the present
®ndings, a previous study found that the Fe3‡ complex
of NIH, formed in a 1:1 ratio, had no antiproliferative
effects toward SK-N-MC cells below 50 mM [40], which
may be the result of the relative resistance of this cell line,
or a consequence of the difference in the stoichiometry
of the complex. The mechanism by which iron complexes
of PIH analogs cause toxicity is unknown, but may be
related to their capacity to deliver iron to cells; FePIH, at

J.L. Buss et al. / Biochemical Pharmacology 65 (2003) 161±172
171
[15] Hockenbery DM, Oltvai ZN, Yin XM, Milliman CL, Korsmeyer SJ.
Bcl-2 functions in an anti-oxidant pathway to prevent apoptosis. Cell
1993;75:241±51.
release from the cell [17]. This observation is a strong argu-
ment for the hypothesis that accumulation of iron com-
plexes in the cell is, at least in part, the cause of the toxicity
induced by PIH analogs.
[16] Gamen S, Anel A, Perez-Galan P, Lasierra P, Johnson D, Pineiro A,
Naval J. Doxorubicin treatment activates a Z-VAD-sensitive caspase,
which causes deltapsim loss, caspase-9 activity, and apoptosis in
Jurkat cells. Exp Cell Res 2000;258:223±35.
[17] Buss JL, Arduini E, Shephard KC, Ponka P. Lipophilicity of analogs of
pyridoxal isonicotinoyl hydrazone (PIH) determines efflux of Fe
complexes and toxicity in K562 cells. Biochem Pharmacol, in press.
[18] Huang AR, Ponka P. A study of the mechanism of action of pyridoxal
isonicotinoyl hydrazone at the cellular level using reticulocytes loaded
with non-heme 59Fe. Biochim Biophys Acta 1983;757:306±15.
[19] Neuzil J, Svensson I, Weber T, Weber C, Brunk UT. a-Tocopheryl
succinate-induced apoptosis in Jurkat T cells involves caspase-3
activation, and both lysosomal and mitochondrial destabilisation.
FEBS Lett 1999;445:295±300.
Acknowledgments
The ®nancial assistance of the CIHR, in the form of
operating funds, and of the National Cooley's Anemia
Foundation, in the form of a postdoctoral fellowship
(J.L.B.), is gratefully acknowledged. J.N. was partially
È
supported by a University of Linkoping Grant (83081030).
È
[20] Neuzil J, Weber T, Schroder A, Lu M, Ostermann G, Gellert N, Mayne
References
Á
Â
GC, Olejnicka B, Negre-Salvayre A, Stõcha M, Coffey RJ, Weber C.
Induction of cancer cell apoptosis by a-tocopheryl succinate: mole-
cular pathways and structural requirements. FASEB J 2001;15:403±15.
[21] Neuzil J, Schroder A, von Hundelshausen P, Zernecke A, Weber T,
Gellert N, Weber C. Inhibition of inflammatory endothelial responses
by a pathway involving caspase activation and p65 cleavage. Bio-
chemistry 2001;40:4686±92.
[1] Ponka P, Beaumont C, Richardson DR. Function and regulation of
transferrin and ferritin. Semin Hematol 1998;35:35±54.
[2] Hileti D, Panayiotidis P, Hoffbrand AV. Iron chelators induce apop-
tosis in proliferating cells. Br J Haematol 1995;89:181±7.
[3] Tanaka T, Satoh T, Onozawa Y, Kohroki J, Itoh N, Ishidate Jr M, Muto
N, Tanaka K. Apoptosis during iron-chelator-induced differentiation
in F9 embryonal carcinoma cells. Cell Biol Int 1999;23:541±50.
[4] Juckett MB, Shadley JD, Zheng Y, Klein JP. Desferrioxamine en-
hances the effects of gamma radiation on clonogenic survival and the
formation of chromosomal aberrations in endothelial cells. Radiat Res
1998;149:330±7.
[22] Murphy TB, Johnson DK, Rose NJ, Aruffo A, Schomaker V. Structural
studies of iron(III) complexes of the new iron-binding drug, pyridoxal
isonicotinoyl hydrazone. Inorg Chim Acta 1982;66:L67±8.
[23] Ponka P, Grady RW, Wilczynska A, Schulman HM. The effect of
various chelating agents on the mobilization of iron from reticulocytes
in the presence and absence of pyridoxal isonicotinoyl hydrazone.
Biochim Biophys Acta 1984;802:477±89.
[5] Rakba N, Loyer P, Gilot D, Delcros JG, Glaise D, Baret P, Pierre JL,
Brissot P, Lescoat G. Antiproliferative and apoptotic effects of O-
Trensox, a new synthetic iron-chelator, on differentiated human
hepatoma cell lines. Carcinogenesis 2000;21:943±51.
[24] Buss JL, Arduini E, Ponka P. Mobilization of intracellular iron by
analogs of pyridoxal isonicotinoyl hydrazone (PIH) is determined by
the membrane permeability of the iron-chelator complexes. Biochem
Pharmacol 2002;64:1689±701.
[6] Simonart T, Degraef C, Andrei G, Mosselmans R, Hermans P, Van
Vooren JP, Noel JC, Boelaert JR, Snoeck R, Heenen M. Iron chelators
inhibit the growth and induce the apoptosis of Kaposi's sarcoma cells
and of their putative endothelial precursors. J Invest Dermatol
2000;115:893±900.
[25] Brunk UT, Svensson I. Oxidative stress, growth factor starvation and
Fas activation may all cause apoptosis through lysosomal leak. Redox
Rep 1999;4:3±11.
[26] Roberg K, Ollinger K. Oxidative stress causes relocation of the
lysosomal enzyme cathepsin D with ensuing apoptosis in neonatal
rat cardiomyocytes. Am J Pathol 1998;152:1151±6.
[7] Brittenham GM. Pyridoxal isonicotinoyl hydrazone. Effective iron-
chelation after oral administration. Ann NY Acad Sci 1990;612:
315±26.
[27] Li W, Yuan X, Nordgren G, Dalen H, Dubowchik GM, Firestone RA,
Brunk UT. Induction of cell death by the lysosomotropic detergent
MSDH. FEBS Lett 2000;470:35±9.
[8] Cikrt M, Ponka P, Necas E, Neuwirt J. Biliary iron excretion in rats
following pyridoxal isonicotinoyl hydrazone. Br J Haematol 1980;
45:275±83.
[28] Brunk UT, Dalen H, Roberg K, Hellquist HB. Photo-oxidative dis-
ruption of lysosomal membranes causes apoptosis of cultured human
fibroblasts. Free Radic Biol Med 1997;23:616±26.
[9] Sookvanichsilp N, Nakornchai S, Weerapradist W. Toxicological
study of pyridoxal isonicotinoyl hydrazone: acute and subchronic
toxicity. Drug Chem Toxicol 1991;14:395±403.
[29] Deiss LP, Galinka H, Berissi H, Cohen O, Kimchi A. Cathepsin D
protease mediates programmed cell death induced by interferon-g,
Fas/APO-1 and TNF-a. EMBO J 1996;15:3861±70.
[30] Susin SA, Lorenzo HK, Zamzami N, Marzo I, Snow BE, Brothers GM,
Mangion J, Jacotot E, Costantini P, Loeffler M, Larochette N, Goodlett
DR, Aebersold R, Siderovski DP, Penninger JM, Kroemer G. Mole-
cular characterization of mitochondrial apoptosis-inducing factor.
Nature 1999;397:441±6.
[10] Richardson DR, Tran EH, Ponka P. The potential of iron chelators of
the pyridoxal isonicotinoyl hydrazone class as effective antiprolifera-
tive agents. Blood 1995;86:4295±306.
[11] Ponka P, Baker E, Edward JT. Effect of pyridoxal isonicotinoyl
hydrazone and other hydrazones on iron release from macrophages,
reticulocytes and hepatocytes. Biochim Biophys Acta 1988;967:
122±9.
[12] Blaha K, Cikrt M, Nerudova J, Fornuskova H, Ponka P. Biliary iron
excretion in rats following treatment with analogs of pyridoxal iso-
nicotinoyl hydrazone. Blood 1998;91:4368±72.
[31] Zhou Q, Snipas S, Orth K, Muzio M, Dixit VM, Salvesen GS. Target
protease specificity of the viral serpin CrmA. Analysis of five cas-
pases. J Biol Chem 1997;272:7797±800.
[13] Richardson DR, Ponka P. The iron metabolism of the human neuro-
blastoma cell: lack of relationship between the efficacy of iron-chela-
tion and the inhibition of DNA synthesis. J Lab Clin Med 1994;124:
660±71.
[32] Hockenbery D, Nunez G, Milliman C, Schreiber RD, Korsmeyer SJ.
Bcl-2 is an inner mitochondrial membrane protein that blocks pro-
grammed cell death. Nature 1990;348:334±6.
[33] Kim S, Ponka P. Effects of interferon-g and lipopolysaccharide on
macrophage iron metabolism are mediated by nitric oxide-induced
degradation of iron regulatory protein 2. J Biol Chem 2000;275:
6220±6.
[14] Edward JT, Gauthier M, Chubb FL, Ponka P. Synthesis of new
acylhydrazones as iron-chelating compounds. J Chem Eng Data
1988;33:538±40.

172
J.L. Buss et al. / Biochemical Pharmacology 65 (2003) 161±172
[34] Laskey J, Webb I, Schulman HM, Ponka P. Evidence that transferrin
supports cell proliferation by supplying iron for DNA synthesis. Exp
Cell Res 1988;176:87±95.
iron(III) complex: an iron-chelator with anti-tumour activity. J Biol
Inorg Chem 1999;4:266±73.
[41] Djeha A, Brock JH. Uptake and intracellular handling of iron from
transferrin and iron chelates by mitogen stimulated mouse lympho-
cytes. Biochim Biophys Acta 1992;1133:147±52.
[35] Ponka P, Schulman HM, Wilczynska A. Ferric pyridoxal isonicotinoyl
hydrazone can provide iron for heme synthesis in reticulocytes.
Biochim Biophys Acta 1982;718:151±6.
[42] Darnell G, Richardson DR. The potential of iron chelators of the
pyridoxal isonicotinoyl hydrazone class as effective antiproliferative
agents III: the effect of the ligands on molecular targets involved in
proliferation. Blood 1999;94:781±92.
[36] Hengartner MO. The biochemistry of apoptosis. Nature 2000;407:
770±6.
È
[37] Roberg K, Johansson U, Ollinger K. Lysosomal release of cathepsin D
precedes relocation of cytochrome c and loss of mitochondrial trans-
membrane potential during apoptosis induced by oxidative stress. Free
Radic Biol Med 1999;27:1228±37.
[43] Edward JT, Chubb FL, Sangster J. Iron chelators of the pyridoxal
isonicotinoyl hydrazone class. Relationship of the lipophilicity of the
apochelator to its ability to mobilize iron from reticulocytes in vitro:
reappraisal of reported partition coefficients. Can J Physiol Pharmacol
1997;75:1362±8.
[38] Tsao MS, Sanders GH, Grisham JW. Regulation of growth of cultured
hepatic epithelial cells by transferrin. Exp Cell Res 1987;171:52±62.
[39] Brock JH, Stevenson J. Replacement of transferrin in serum-free
cultures of mitogen-stimulated mouse lymphocytes by a lipophilic
iron-chelator. Immunol Lett 1986;15:23±5.
[44] Ponka P, Richardson DR, Edward JT, Chubb FL. Iron chelators of
the pyridoxal isonicotinoyl hydrazone class. Relationship of the
lipophilicity of the apochelator to its ability to mobilise iron
from reticulocytes in vitro. Can J Physiol Pharmacol 1994;72:
659±66.
[40] Richardson DR, Bernhardt PV. Crystal and molecular structure of 2-
hydroxy-1-naphthylaldehyde isonicotinoyl hydrazone (NIH) and its