Novel second-generation Di-2-pyridylketone thiosemicarbazones show synergism with standard chemotherapeutics and demonstrate potent activity against lung cancer xenografts after oral and intravenous administration in vivo

Article abstract of DOI:10.1021/jm300768u

We developed a series of second-generation di-2-pyridyl ketone thiosemicarbazone (DpT) and 2-benzoylpyr-idine thiosemicarbazone (BpT) ligands to improve the efficacy and safety profile of these potential antitumor agents. Two novel DpT analogues, Dp4e4mT and DpC, exhibited pronounced and selective activity against human lung cancer xenografts in vivo via the intravenous and oral routes. Importantly, these analogues did not induce the cardiotoxicity observed at high nonoptimal doses of the first-generation DpT analogue, Dp44mT. The Cu(II) complexes of these ligands exhibited potent antiproliferative activity having redox potentials in a range accessible to biological reductants. The activity of the copper complexes of Dp4e4mT and DpC against lung cancer cells was synergistic in combination with gemcitabine or cisplatin. It was demonstrated by EPR spectroscopy that dimeric copper compounds of the type [CuLCl]₂, identified crystallographically, dissociate in solution to give monomeric 1:1 Cu:ligand complexes. These monomers represent the biologically active form of the complex.

Full text of DOI:10.1021/jm300768u

Article
pubs.acs.org/jmc
Novel Second-Generation Di-2-Pyridylketone Thiosemicarbazones
Show Synergism with Standard Chemotherapeutics and
Demonstrate Potent Activity against Lung Cancer Xenografts after
Oral and Intravenous Administration in Vivo
David B. Lovejoy,† Danae M. Sharp,† Nicole Seebacher,† Peyman Obeidy,† Thomas Prichard,†
Christian Stefani,† Maram T. Basha,‡ Philip C. Sharpe,‡ Patric J. Jansson,† Danuta S. Kalinowski,†
Paul V. Bernhardt,‡ and Des R. Richardson*,†
†Iron Metabolism and Chelation Program, Department of Pathology and Bosch Institute, University of Sydney, Sydney, New South
Wales 2006, Australia
‡School of Chemistry and Molecular Biosciences, University of Queensland, Brisbane, Queensland 4072, Australia
S
* Supporting Information
ABSTRACT: We developed a series of second-generation di-
2-pyridyl ketone thiosemicarbazone (DpT) and 2-benzoylpyr-
idine thiosemicarbazone (BpT) ligands to improve the efficacy
and safety profile of these potential antitumor agents. Two
novel DpT analogues, Dp4e4mT and DpC, exhibited
pronounced and selective activity against human lung cancer
xenografts in vivo via the intravenous and oral routes.
Importantly, these analogues did not induce the cardiotoxicity
observed at high nonoptimal doses of the first-generation DpT
analogue, Dp44mT. The Cu(II) complexes of these ligands
exhibited potent antiproliferative activity having redox
potentials in a range accessible to biological reductants. The
activity of the copper complexes of Dp4e4mT and DpC
against lung cancer cells was synergistic in combination with gemcitabine or cisplatin. It was demonstrated by EPR spectroscopy
that dimeric copper compounds of the type [CuLCl]2, identified crystallographically, dissociate in solution to give monomeric
1:1 Cu:ligand complexes. These monomers represent the biologically active form of the complex.
shown marked antitumor activity are the di-2-pyridyl ketone
thiosemicarbazones (DpT; Figure 1) and their 2-benzoylpyr-
idine thiosemicarbazone (BpT; Figure 1) analogues.9−12 Some
of the DpT ligands overcome the disadvantages of 3-AP, in that
they do not induce methemoglobinemia13 and are markedly
more active and selective.10,14−16
The most extensively evaluated DpT series analogue is di-2-
pyridyl ketone 4,4,-dimethyl-3-thiosemicarbazone (Dp44mT;
Figure 1), which demonstrated marked and selective activity
against tumor xenografts in nude mice.10,12 Significantly,
Dp44mT was shown to mediate marked antitumor activity by
Fe and copper (Cu) chelation and redox-cycling of the so-
formed Fe and Cu complexes to generate reactive oxygen
species (ROS).12,17−19 Other modes of anticancer activity
reported for Dp44mT and other chelators include up-
regulation of the metastasis suppressor protein, N-myc
downstream regulated gene-1 (NDRG1),14,20−22 and modu-
lation of the expression of the cyclin family of proteins (e.g.,
INTRODUCTION
■
Targeting the essential nutrient iron (Fe) is a promising
approach for the treatment of cancer, reflecting the fact that
neoplastic cells have higher Fe utilization due to their rapid
replication relative to normal cells.1,2 In fact, this increased use
of iron is required for the activity of key Fe-containing proteins
that catalyze critical reactions involved in energy metabolism,
respiration, and DNA synthesis.3−5 Consequently, due to this
increased Fe requirement, neoplastic cells compared to their
normal counterparts express higher levels of the transferrin
receptor 1 that delivers iron from the Fe transport protein,
transferrin (Tf).1,2
Considering the key role of Fe in tumor cell growth, the
development of novel Fe-binding drugs (chelators) is a
promising anticancer strategy,1,5,6 as demonstrated by the
entrance into clinical trials of the anticancer agent, 3-
aminopyridine-2-carboxaldehyde thiosemicarbazone (3-AP).
However, this chelator results in relatively low antitumor
activity and its dose-limiting side effects such as methemoglo-
binemia and hypoxia have seriously limited its clinical
potential.7,8 Other classes of potential ligands that have
Received: June 3, 2012
Published: August 3, 2012
© 2012 American Chemical Society
7230
dx.doi.org/10.1021/jm300768u | J. Med. Chem. 2012, 55, 7230−7244

Journal of Medicinal Chemistry
Article
yet less toxic DpT chelators, we designed a second-generation
of DpT ligands. These new agents were designed based upon a
key structure−activity relationship derived from our previous
studies25 in which the replacement of the terminal H at N4
with an alkyl group increases antiproliferative activity.
The current investigation describes seven second-generation
DpT and BpT analogues, and of these, two ligands, namely di-
2-pyridyl ketone 4-ethyl-4-methyl-3-thiosemicarbazone
(Dp4e4mT) and di-2-pyridyl ketone 4-cyclohexyl-4-methyl-3-
thiosemicarbazone (DpC), were identified as having the best
activity and safety profile in vivo. Major findings include that
both agents exhibited marked activity in vivo against a human
lung cancer xenograft after both intravenous and oral
administration and that they did not induce cardiotoxicity,
unlike the first-generation analogue, Dp44mT.10 Indeed, the
ability of these agents to be administered orally provides a
significant pharmacological advantage for treatment relative to
the more cumbersome intravenous route. Additionally, for the
first time, we demonstrate that these agents and their copper
complexes act synergistically with standard chemotherapeutics
for lung cancer, namely gemcitabine and cisplatin.
RESULTS AND DISCUSSION
■
Chemistry Studies. Ligand and Complex Synthesis and
Characterization. The second-generation thiosemicarbazone
analogues were synthesized by standard methods involving a
Schiff base condensation between the N4-disubstituted
thiosemicarbazide with di-2-pyridyl ketone or 2-benzoyl-
pyridine, as reported for analogous compounds.19,26 The
ligands were not appreciably soluble in water at high
concentrations but were highly soluble in polar aprotic solvents
Figure 1. Line drawings of the structures of the chelators,
desferrioxamine (DFO), and 2-hydroxy-1-naphthaldehyde isonicoti-
noyl hydrazone (NIH; also known as 311) and the general structures
of the DpT and BpT series of chelators. Substituents defining the first-
and second-generation ligands are indicated.
1
cyclin D1), GADD family of proteins (e.g., GADD45), and
cyclin-dependent kinase 2.23,24 Significantly, the potent and
selective antitumor activity of Dp44mT has been independently
verified by others.15,16,22 In particular, Rao and colleagues
reported topoisomerase IIα inhibition and specific antitumor
activity,15 while Liu et al. demonstrated that Dp44mT markedly
inhibits tumor cell metastasis in vivo through induction of
NDRG1.22 However, our laboratory reported that, at high
nonoptimal doses, Dp44mT was found to induce cardiotoxicity
in nude mice.10 Hence, in an effort to develop highly potent,
such as DMF, MeCN, and DMSO. The H NMR spectral
properties of the second-generation DpT and BpT ligands were
very similar to those of the first-generation DpT and BpT
chelators.9,19
The CuII and FeIII complexes of thiosemicarbazones are
known to possess potent antitumor activity and were
synthesized according to previously published procedures.19,27
Spectroscopic measurements (UV−vis and IR) confirmed the
formation of each complex, and the properties mirrored those
previously reported.19,27 In DMF, the CuII complexes exhibited
Figure 2. ORTEP views of: (A) the dimeric [Cu(Dp4e4mT)Cl]2 complex and (B) the monomeric [Cu(Dp4e4mT)Cl] asymmetric unit. The 30%
probability ellipsoids are shown. Selected bond lengths (Å) and angles (°): Cu1−N1 2.030(3), Cu−N3 1.987(3), Cu−S1 2.255(1), Cu−Cl1
2.285(1), Cu−N2′ 2.373(3), N1−Cu−S1 163.13(9), N3−Cu−Cl1 148.04(9), N1−Cu−N3 80.12(1), N3−Cu−S1 83.06(9), N2′−Cu-Cl1 94.88(7).
7231
dx.doi.org/10.1021/jm300768u | J. Med. Chem. 2012, 55, 7230−7244

Journal of Medicinal Chemistry
Article
a weak visible absorption maximum (as a shoulder) in the
region of 600 nm, which is most likely of d−d origin on the
basis of their extinction coefficients and their energies which are
as expected for a CuIIN2SO or CuIIN2SCl chromophore. The
dominant peaks are due to ligand-centered transitions in the
near UV. Overall, the electronic spectra of the CuII complexes
mirror those of the related thiosemicarbazones such as
[Cu(Dp44mT)(OAc)].27 Similarly, the FeIII complexes are
spectroscopically indistinguishable from previously published
pyridyl-thiosemicarbazone FeIII complexes.9,19,28
types will form and mediate biological activity, as shown for the
closely related DpT analogue, Dp44mT.18
Solution Structure. The dimeric solid state structure of
[Cu(Dp4e4mT)Cl]2 (Figure 2A) certainly does not persist in
solution. In fact, EPR spectra of [Cu(Dp4e4mT)(OAc)],
[Cu(Dp4e4mT)Cl], [Cu(DpC)(OAc)], and [Cu(DpC)Cl]
were measured in frozen DMF solution (Figure 3), with each
Structural Characterization. The crystal structure of the
CuII complex of Dp4e4mT (with chloride as the coligand) was
determined considering the potent antitumor activity of this
complex (see Biological Studies) and its likelihood to be
involved in the biological activity of this thiosemicarbazone.18
Although the expected tridentate N2S coordination mode of the
ligand was found, the structure is of a centrosymmetric, dimeric
complex [Cu(Dp4e4mT)Cl]2 where the distal pyridyl ring
forms a bridge to an adjacent CuII ion (Figure 2A). The
monomeric asymmetric unit comprising the equatorially
coordinated Dp4e4mT− and Cl− ligands is also shown (Figure
2B). The Cu(Dp4e4mT) moiety is essentially planar, with all
three donor atoms and the metal lying within 0.03 Å of the
least-squares plane defined by the CuN2S array. The chlorido
ligand is displaced by 1.16 Å from the CuN2S plane to
minimize repulsion of the cis-coordinated axial pyridyl ligand
(C11−H11···Cl1′ 2.80 Å). The coordination sphere is
completed by a weakly bound axial pyridyl ligand where the
axial elongation is due to a Jahn−Teller distortion which
markedly lowers the energy of the dz2 orbital relative to the
dx2−y2 orbital. The overall form of the five-coordinate geometry
can be quantified by the τ parameter29 which defines the degree
of distortion between the extremes of square pyramidal (τ = 0)
and trigonal bipyramidal (τ = 1) geometry on the basis of the
coordinate angles [τ = (α − β)60, where α is the largest
coordinate angle and β is the second largest coordinate angle].
For the structure of [Cu(Dp4e4mT)Cl]2, a value of τ = 0.25
was calculated, reflecting a distorted square pyramidal
geometry.
Dimeric structures of this type have been seen in other Schiff
base complexes comprising the di-2-pyridyl ketone moiety.30,31
However, we have not structurally characterized any 1:1 CuII
thiosemicarbazone complexes derived from chelators of the
DpT class, and it appears likely that the self-assembly of Cu
complex dimers, such as that seen in Figure 2A, is a
characteristic feature of this group of ligands.
Because of the ability of these compounds to act as potent Fe
chelators in cells in culture (see Biological Studies), the
structure of the FeIII complex [Fe(DpC)2]ClO4 was also
examined. There were no unusual features of this complex
which is similar to other low spin FeIII thiosemicarbazone
complexes reported by our laboratory.28 The anionic ligands
coordinate to Fe in a tridentate manner, typical for chelators
from the BpT and DpT series, in their imine-thiolate resonance
forms as NNS donors.9,19 The three donor atoms are in a syn
disposition as found previously for other members of this
family.9,19 These data above demonstrating complexation of Fe
and Cu are significant, as the biological activity of this class of
ligands is mediated through the binding of both these
intracellular metals with the subsequent formation of cytotoxic
redox-active complexes.12,18,19 Indeed, we show below that Cu
complexation leads to potent antitumor efficacy and it is likely
that because cells contain both Fe and Cu, complexes of both
Figure 3. X-band EPR spectra of [Cu(Dp4e4mT)(OAc)], [Cu-
(Dp4e4mT)Cl], [Cu(DpC)(OAc)], and[Cu(DpC)Cl] (1 mM) in
frozen DMF solution (140 K).
spectrum being consistent with a monomeric CuII complex in a
tetragonally elongated square pyramidal coordination environ-
ment (gz ≫ gx,y, Az ≫ Ax,y).32 No evidence for dimer formation
through dipole−dipole coupling in the EPR spectrum was
apparent, thus complete dissociation into monomeric 1:1
Cu:ligand complexes occurs in a solution of DMF and water
(see below). The EPR spectra were simulated (see Supporting
Information Figures S1−S4) in order to obtain accurate spin
Hamiltonian parameters. Partial chlorido ligand dissociation of
[Cu(Dp4e4mT)Cl] and [Cu(DpC)Cl] in DMF solution was
suppressed by the addition of an excess of Et4NCl. In contrast,
the acetato complexes were more stable and no evidence of
dissociation was found. In addition to hyperfine coupling with
the Cu nucleus (63Cu, 65Cu: I = 3/2), which generates three
sets of four lines centered at gz, gy, and gx, ligand-centered
superhyperfine coupling was superposed in the spectra of the
OAc− complexes. The two cis coordinated N-donors in the xy
plane bear a nuclear spin of I = 1 and split each peak into a
quintet. Long range superhyperfine coupling, probably from the
noncoordinated N4 (Cu···N4 2.96 Å), may be responsible for
the added fine structure, but this was not modeled in the
simulation. This superhyperfine coupling was completely
absent in the spectra of the two chlorido complexes,
[Cu(Dp4e4mT)Cl] and [Cu(DpC)Cl], due presumably to
broadening from unresolved superhypefine coupling with the
chlorido ligands (35Cl, 37Cl: I = 3/2).
In each case, the EPR spectra in Figure 3 mimic those of the
previously reported monomeric [Cu(Dp44mT)(OAc)] com-
7232
dx.doi.org/10.1021/jm300768u | J. Med. Chem. 2012, 55, 7230−7244

Journal of Medicinal Chemistry
Article
plex27 and unequivocally show that the solid state structure is
not representative of its solution structure. In previous studies,
we isolated six-coordinate 1:2 Cu:ligand complexes of related
thiosemicarbazones27 but found that partial dissociation was
spontaneous, leading to the corresponding 1:1 Cu:ligand
complex in solution.
generation DpT and BpT chelators were designed by specific
replacement of the H atom at the terminal N4 atom (Figure 1)
with alkyl and aryl groups in order to increase their lipophilicity
and potentially enhance antitumor activity.25 In these experi-
ments, the first-generation chelators still bearing the terminal H
atom were directly compared to the novel second-generation
analogues in terms of their antiproliferative activity in vitro. For
these studies, we initially utilized the SK-N-MC cell line as its
response to chelators has been extensively examined in our
laboratory.25,34,35 Additionally, the positive control chelators,
namely desferrioxamine (DFO), 2-hydroxy-1-naphthaldehyde
isonicotinoyl hydrazone (NIH; also known as 311), and
Dp44mT, were also assessed as their antiproliferative activity
has been well characterized in this cell line.25,34,35
Electrochemistry. Cyclic voltammetry of the Cu complexes
in mixed DMF:H2O (2:1) was carried out (Figure 4) to
In good agreement with previous studies,12,19 DFO and NIH
exhibited moderate antiproliferative activity against SK-N-MC
cells, with NIH being more active than DFO (Table 1). The
Table 1. Antiproliferative Activity (IC50) of the First-
Generation DpT and BpT Chelators Compared to the
Second-Generation Analogues and the Positive Control
Chelators, DFO, NIH and Dp44mT, as Determined Using
a
SK-N-MC Cells after a 72 h Incubation
IC50 (μM)
SK-N-MC
chelator
DFO
10
1
NIH
1.1 0.3
Figure 4. Cyclic voltammograms of [Cu(Dp4e4mT)Cl], [Cu-
(Dp4e4mT)(OAc)], [Cu(DpC)Cl], and [Cu(DpC)(OAc)] (1 mM
in DMF:H2O 2:1) at a sweep rate of 100 mV s−1.
Dp44mT
0.004 0.002
first-generation analogues
Dp4mT
0.3 0.1
0.06 0.01
0.02 0.01
0.018 0.004
0.012 0.002
Dp4eT
determine the CuII/I redox potentials and to assess whether the
complexes might be able to mediate Cu-catalyzed Fenton
chemistry. This solvent combination was used to obtain
sufficiently high concentrations of the copper complexes in
solution. Under these conditions, the four complexes gave
almost identical voltammograms with quasireversible CuII/I
couples at −90 mV versus the non-hydrogen electrode
(NHE; Figure 4). This suggests that the Cl− and OAc− ligands
are replaced by water in mixed DMF:water. This was confirmed
by equivalent voltammetry experiments conducted in 100%
DMF, where significantly different voltammograms were
obtained for the acetato and chlorido complexes (see
Supporting Information Figure S5). Partial dissociation of the
chlorido complexes (in this case with no additional chloride
ions present) was evident to give the DMF-coordinated
analogues from the presence of a dominant high potential
wave from [Cu(DpC)(DMF)]+/0 and [Cu(Dp4e4mT)-
(DMF)]+/0 (Supporting Information Figure S5) and a minor
low potential wave from the corresponding chlorido complexes.
Very similar behavior has been reported for related
dithiocarbazate Schiff base complexes of Cu(II).33 In either
solvent (DMF or water/DMF), the voltammograms were
insensitive to the change from ethyl (Dp4e4mT) to cyclohexyl
(DpC). The reversible electrochemical behavior of these Cu
complexes observed at a relatively high potential suggests that
the Cu(II) complexes can easily be reduced in vivo and
participate in redox cycling and ROS generation. This may be a
factor in their potent antiproliferative activity (see Biological
Studies).
Dp4pT
Bp4mT
Bp4eT
second-generation analogues
Dp4e4mT
Dp44eT
0.004 0.002
0.005 0.002
0.006 0.001
0.010 0.002
0.013 0.002
0.006 0.001
0.011 0.003
Dp4p4mT
Dp4p4eT
DpC
Bp4e4mT
Bp44eT
a
Results are mean SD (three experiments).
concentrations of DFO and NIH required to inhibit cell growth
to 50% of the untreated control (IC50) were 10 1 and 1.1
0.3 μM, respectively (Table 1). However, Dp44mT potently
inhibited proliferation of SK-N-MC cells (IC50: 0.004 0.002
μM), being far more active than DFO or NIH, as shown
previously.12,19 The new second-generation compounds also all
exhibited significantly (p < 0.001) greater antiproliferative
activity than DFO or NIH against SK-N-MC cells (Table 1).
The analogues Dp4e4mT, Dp44eT, and Dp4p4mT were the
most effective second-generation DpT series chelators, having
IC50 values (0.004−0.006 μM) that were similar to Dp44mT
(Table 1).
A significant (p < 0.001) increase in antiproliferative activity
was apparent when examining the least active first-generation
ligand, Dp4mT (IC50: 0.3 0.1 μM; Table 1), with the least
active second-generation analogue, DpC (IC50: 0.013 0.002
μM; Table 1), which showed 23-fold greater activity. This
demonstrated the success of our ligand design strategy (Table 1
Biological Studies. Novel Second-Generation DpT and
BpT Analogues Exhibit Potent Antiproliferative Activity
against SK-N-MC Neuroepithelioma Cells. The second-
7233
dx.doi.org/10.1021/jm300768u | J. Med. Chem. 2012, 55, 7230−7244

Journal of Medicinal Chemistry
Article
and Figure 5A) and could be due to their increased lipophilicity
and membrane permeability. The first-generation BpT
chelators were found to lie within this optimal lipophilic range
(Figure 5B). The optimal lipophilicity of the second-generation
DpT and BpT chelators may be a critical factor in their
increased antiproliferative activity in comparison to their parent
first-generation analogues, which were generally less lipophilic
(Figure 5B). We have recently observed a similar relationship
with a related class of thiosemicarbazones,36 and this
demonstrates the important correlation between optimal
lipophilicity and potent antiproliferative activity of thiosemi-
carbazone chelators.
Novel Second-Generation DpT and BpT Analogues Inhibit
Cellular 59Fe Uptake from 59Fe Transferrin and Mediate
Efflux of 59Fe from Cells. Our previous studies have related
chelator antiproliferative activity to the prevention of cellular Fe
uptake and to the ability to increase Fe mobilization from
cells.35,36 Hence, the efficacy of the second-generation DpT and
BpT chelators to inhibit 59Fe uptake from the serum Fe-binding
protein, Tf, in SK-N-MC neuroepithelioma cells was assessed
and compared to first-generation chelators. Again, these cells
were used as their response to chelators in Fe uptake and efflux
studies has been well characterized in our laboratory.25,34,35 The
positive control ligands, NIH and Dp44mT, were very effective
in limiting cellular 59Fe uptake to 5.3 1.2% and 3.5 1.4% of
the control, respectively (Figure 6A), and were far more active
than the hydrophilic chelator, DFO, that only limited 59Fe
uptake to 84 5% of the control (Figure 6A). These results
were in good agreement with our previous studies.9,25
Generally, the majority of the second-generation DpT and
BpT chelators were slightly less effective than their first-
generation counterparts (Figure 6A). For instance, while
Dp4e4mT showed 75-fold greater antiproliferative efficacy
than its first-generation parent, Dp4mT, in terms of preventing
59Fe uptake, Dp4e4mT was almost half as effective as Dp4mT
(12.3% cf. 6.9% of the control; Figure 6A). The only exception
was the second-generation chelator, Dp44eT, which was
approximately twice as effective at inhibiting 59Fe uptake as
the relative first-generation ligand, Dp4eT. However, all
second-generation compounds showed markedly and signifi-
cantly (p < 0.001) greater activity than DFO (Figure 6A).
Assessing the ability of these ligands to mobilize 59Fe from
SK-N-MC cells, and in agreement with previous investiga-
tions,9,12 the positive control chelators, DFO, NIH, and
Figure 5. (A) A marked increase in antiproliferative activity is apparent
for the second-generation analogues (Dp4e4mT and DpC) relative to
their first-generation counterpart (Dp4mT) due to the replacement of
the N4-terminal H with alkyl or aryl substituents. These studies
demonstrate the success of the design strategy implemented in the
current investigation. Results are from MTT proliferation assays
examining the effects of the ligands on the growth of SK-N-MC cells
over a 72 h incubation. Results are means of three experiments. (B)
Relationship between the antiproliferative activity (IC50) and lip-
ophilicity (log Pcalc) of the chelators using SK-N-MC neuroepithelioma
cells. The IC50 (measured after a 72 h incubation) and log Pcalc values
for the second-generation DpT and BpT analogues and their parent
compounds (Dp4mT, Dp4eT, Dp4pT, Bp4mT, and Bp4eT) were
plotted. Lines were fitted in (A) and (B) using Microsoft Excel 2007
(Microsoft, Redmond, WA).
Dp44mT, were able to release 15.2
and 45.0
0.9%, 44.9
1.5%,
3.0% cellular 59Fe, respectively, while control
chelators, Bp4mT and Bp4eT, showed antiproliferative activity
against SK-N-MC cells that was also markedly more potent
than DFO or NIH in accordance with previous studies.9 The
second-generation Bp4e4mT and Bp44eT chelators demon-
strated similar or greater antiproliferative efficacy than the
corresponding parent BpT first-generation chelators (Table 1).
To assess the optimal octanol:water partition coefficient
required for potent antiproliferative activity in SK-N-MC cells,
the log Pcalc values of the first- and second-generation DpT and
BpT analogues were plotted against their IC50 values (Figure
5B). This demonstrated that an optimal lipophilic range existed,
with ligands having log Pcalc values between ∼2.5 and 4.0,
demonstrating the greatest antiproliferative efficacy. Signifi-
cantly, optimal activity was observed at a log Pcalc value of ∼3.2
(Figure 5B). Importantly, more hydrophilic compounds of this
class with log Pcalc values lower than 2.5 may have difficulty
crossing the cell membrane, while more lipophilic thiosemi-
carbazones with log Pcalc values greater than 4.0 may become
trapped in the lipid membrane. Significantly, the log Pcalc values
of the majority of the second-generation DpT and BpT
medium alone only led to the release of 6.4 0.7% of 59Fe
(Figure 6B). While all the second-generation chelators were
more active than DFO at mobilizing cellular 59Fe, they were
generally less active than the corresponding first-generation
analogues. The most effective second-generation chelators at
mobilizing cellular 59Fe were Dp4e4mT, Dp44eT, and
Dp4p4mT, which had comparable efficacy as their first-
generation counterparts (Figure 6B). However, notably lower
activity at inducing 59Fe efflux was identified for the second-
generation chelators, Dp4p4eT, DpC, Bp4e4mT, and Bp44eT,
relative to the first-generation analogues (Figure 6B).
An optimal log Pcalc value was demonstrated by examining
the relationship between the log Pcalc values of the first- and
second-generation analogues and their ability to inhibit 59Fe
uptake (Figure 6C) or mobilize cellular 59Fe (Figure 6D). In
fact, analogues with a log Pcalc value of 2−2.5 were able to
effectively inhibit 59Fe uptake and mobilize cellular 59Fe, while
the second-generation chelators that generally had higher log
Pcalc values were less efficient (Figures 6C,D). This may reflect
7234
dx.doi.org/10.1021/jm300768u | J. Med. Chem. 2012, 55, 7230−7244

Journal of Medicinal Chemistry
Article
Figure 6. The effect of the first-generation analogues relative to their second-generation counterparts on: (A) inhibition of internalized 59Fe uptake
from 59Fe-transferrin (Tf) by SK-N-MC cells and (B) promotion of cellular 59Fe release from prelabeled SK-N-MC cells. The relationships between
(C) the inhibition of internalized 59Fe uptake from 59Fe-Tf and (D) the cellular 59Fe release against lipophilicity (log Pcalc) of the second-generation
DpT and BpT analogues and their parent compounds (Dp4mT, Dp4eT, Dp4pT, Bp4mT, and Bp4eT) using SK-N-MC neuroepithelioma cells were
plotted. (A) Cells were labeled with 59Fe-Tf (0.75 μM) for 3 h/37 °C in the presence or absence of chelators (25 μM) and then washed four times
on ice with ice-cold PBS. Internalized 59Fe uptake was determined using Pronase (1 mg/mL). (B) Cells were prelabeled with 59Fe-Tf (0.75 μM) for
3 h/37 °C and then washed fpir times on ice with ice-cold PBS. The cells were then reincubated in the presence or absence of chelators (25 μM) for
3 h/37 °C and 59Fe release examined. The results in (C) and (D) are plotted against log Pcalc, which was calculated using ChemDraw v4.5
(CambridgeSoft, Cambridge, MA) and lines fitted using Microsoft Excel 2007. Results are mean SD (three experiments).
the optimal log P value necessary for the resultant hydrophobic
iron complex to pass through cellular membranes and
efficiently mobilize iron from cells. In fact, the most lipophilic
second-generation analogues may form highly hydrophobic
iron complexes that become trapped within cellular mem-
branes, preventing iron mobilization.
Because the majority of the second-generation ligands
showed greater antiproliferative efficacy than first-generation
analogues, despite showing less Fe chelation activity, suggests
their antitumor effect does not solely depend on cellular Fe
complexation. Indeed, we showed that chelators of this class
form highly redox active Fe and Cu complexes that generate
cytotoxic ROS.12,18,19 Additionally, DpT and BpT class
7235
dx.doi.org/10.1021/jm300768u | J. Med. Chem. 2012, 55, 7230−7244

Journal of Medicinal Chemistry
Article
chelators were also demonstrated to have multiple molecular
targets including ribonucleotide reductase,3 thiol-containing
systems,37 molecules involved in cell cycle control e.g., cyclin
D1,23,38 and the metastasis suppressor, NDRG-1.14,20,22
To improve the solubility of these second-generation
chelators for oral in vivo tumor studies (see below),
hydrochloride salts of the agents were prepared. Comparing
their antiproliferative activity, the hydrochloride salts of
Dp4e4mT and DpC showed similar antiproliferative activity
to their base forms (Table 2). Another important pharmaco-
logical aspect of chemotherapy is the ability to target tumor
cells while leaving normal cells unaffected. Hence, we next
examined the effect of the chelators on the growth of mortal
MRC-5 fibroblasts (Table 2). Significantly, these cells were at
least a 1000-fold more resistant to the effects of the second-
generation ligands compared to lung cancer cells, suggesting
the existence of an exploitable “chemotherapeutic window” of
selectivity.
Current chemotherapy for cancer often uses combinations of
two or more drugs.40−42 This approach offers enhanced
synergistic therapeutic efficacy, improved selectivity, lower
toxicity, and helps overcome drug resistance.43 The chemo-
therapy of small cell lung carcinoma often consists of a
platinum drug in combination with etoposide, whereas a
combination of cisplatin and gemcitabine is frequently used for
treatment of nonsmall cell lung carcinoma.40−42 Considering
this, we then examined if DpC·HCl or Dp4e4mT·HCl work
synergistically with standard chemotherapy regimens for lung
cancer (Figure 7A). To potentially provide mechanistic
insights, we included in these experiments DFO and
Dp44mT as relative controls and the Cu(II) complexes of
Activity of the Second-Generation Analogues Against
Lung Cancer In Vitro and In Vivo: Identification of Dp4e4mT
and DpC as Potent Anti-Tumor Agents. Considering the
marked antiproliferative activity of the second-generation
analogues (Table 1) and also the impressive efficacy of the
first-generation analogues against lung cancer cells and a lung
tumor xenograft,10,11 we then investigated the activity of the
second-generation series against this tumor type. Preliminary
investigations in vivo of all second-generation analogues
demonstrated that Dp4e4mT and DpC were the most effective
against this tumor (data not shown). Moreover, these
analogues were novel compounds which enabled patent
protection39 and potential further development as pharmaceut-
icals. Hence, the studies reported below in vitro and in vivo
focus on these two highly active agents.
Dp4e4mT and DpC Exhibit Marked Antiproliferative
Activity Against Lung Cancer Cells in Vitro and Act
Synergistically with Currently Used Chemotherapeutics.
The antiproliferative activity of Dp4e4mT and DpC was
examined using DMS-53 small cell lung cancer cells and A549
non-small cell lung carcinoma cells in vitro over a 72 h
incubation (Table 2). Both Dp4e4mT and DpC potently
Table 2. Antiproliferative Activity (IC50) of Dp4e4mT and
DpC and their HCl Salts Compared to the the Positive
Control Chelators DFO, NIH and Dp44mT and the
Chemotherapeutics, Gemcitabine, Cisplatin, and Etoposide
as Determined Using DMS-53 and A549 Lung Carcinoma
a
Cells after a 72 h incubation
IC50 (μM)
agents
DFO
DMS-53
10
A549
MRC-5
1
7.5 0.6
>10
>10
>10
>10
>10
>10
>10
NIH
1.1 0.3
0.010 0.002
0.008 0.002
0.006 0.001
0.008 0.004
0.007 0.004
0.009 0.003
2.2 0.8
2.2 0.2
Dp44mT
Dp4e4mT
Dp4e4mT·HCI
DpC
0.020 0.008
0.010 0.004
0.001 0.004
0.004 0.001
0.004 0.001
0.02 0.01
DpC·HCI
gemcitabine
cisplatin
8
3
etoposide
1.5 0.6
0.8 0.3
a
The antiproliferative activity of these agents is far less in mortal
MRC-5 fibroblasts relative to the lung cancer cells. Results are mean
SD (three experiments).
inhibited DMS-53 and A549 cell proliferation, generally
showing slightly more activity than the first-generation chelator,
Dp44mT (Table 2). For example, against A549 cells, DpC was
5-fold more active than Dp44mT. Significantly, Dp4e4mT and
DpC were also at least 137-fold more active than DFO or NIH
(Table 2). Importantly, in comparison to the standard
chemotherapeutics, cisplatin and etoposide, that are used for
lung cancer treatment,40−42 Dp4e4mT and DpC exhibited
pronounced and significantly (p < 0.001) greater activity
against DMS-53 and A549 cells (Table 2). Additionally, these
second-generation chelators exhibited comparable or better
activity than the chemotherapeutic, gemcitabine,40−42 again
confirming their promise as anticancer agents.
Figure 7. Demonstration of synergy of DpC, Dp4e4mT, or their
copper complexes when combined with either gemcitabine or cisplatin.
(A) Combination index (CI) plot of DpC, Dp4e4mT, or their copper
complexes when combined with either gemcitabine or cisplatin after a
72 h incubation with either DMS-53 or A549 lung cancer cells. These
results are compared with the controls, DFO, Dp44mT, CuCl2,
gemcitabine, or etoposide when combined with gemcitabine or
cisplatin. (B) Fractional affect−combination index plot for the
combination of gemcitabine and the DpC copper complex ([Cu-
(DpC)Cl]) or cisplatin and [Cu(DpC)Cl] using DMS-53 cells and a
72 h incubation with these agents. Plots were calculated according to
the Chou−Talalay method44−46 using CalcuSyn software (Biosoft,
Cambridge, UK). Results are mean SD (three experiments).
7236
dx.doi.org/10.1021/jm300768u | J. Med. Chem. 2012, 55, 7230−7244

Journal of Medicinal Chemistry
Article
DpC or Dp4e4mT that can probably form intracellularly. The
latter complexes were also assessed by considering that the
Cu(II) complexes of related DpT ligands show marked
antitumor activity due to their ability to redox cycle and
generate cytotoxic ROS.18 In these studies, we used the method
of Chou−Talalay,44,45 which provides a rigorous assessment of
drug−drug interactions based on drug−response curves to
generate a combination index (CI) that quantifies synergistic
activity.44,46 In fact, CI values of <1, 1, and >1 indicate
synergistic, additive, and antagonistic effects, respectively.44,46
Interestingly, DFO exhibited moderate synergism with
gemcitabine or cisplatin against DMS-53 cells, as defined by
studies in vivo against DMS-53 tumor xenografts in nude mice
were important to perform to assess efficacy. This tumor was
used to provide a relevant comparison to our investigations
using Dp44mT and other ligands.10,11 Initial studies examined
chelator efficacy when given by iv administration, as our
previous studies demonstrated that this route of administration
showed good tolerability with potent antitumor activity.10
When DMS-53 xenografts had reached 100 mm3 in volume,
treatment began with either the vehicle control, the positive
control chelator, Dp44mT, which was given as the base form in
accordance with our previous studies (0.75 mg/kg iv; 5 days/
week),10 Dp4e4mT·HCl (3 or 6 mg/kg iv; 5 days/week), or
DpC·HCl (4 or 8 mg/kg iv; 5 days/week). The 0.75 mg/kg
dose of Dp44mT was used considering our in vivo studies
where it led to marked inhibition of tumor growth.10 The doses
selected for DpC·HCl and Dp4e4mT·HCl represented the
maximum tolerated (MTD) and half-MTD doses (data not
shown).
the CI values of 0.80
0.04 and 0.75
0.08, respectively
(Figure 7A). However, against A549 cells, DFO interactions
with gemcitabine or cisplatin and were nearly additive (Figure
7A). The first-generation chelator Dp44mT showed similar
results to DFO against DMS-53 cells. In addition, this chelator
showed a slightly greater synergistic effect in combination with
cisplatin than with gemcitabine in DMS-53 cells (Figure 7A).
Moderate synergism was also evident for Dp4e4mT·HCl or
DpC·HCl in combination with gemcitabine or cisplatin against
DMS-53 cells. However, only Dp4e4mT·HCl retained
synergism against A549 cells in the presence of gemcitabine
(Figure 7A).
Surprisingly, the copper complexes of both DpT analogues
(i.e., [Cu(Dp4e4mT)Cl] and [Cu(DpC)Cl]) were markedly
and significantly (p < 0.001) more synergistic than the free
ligands in combination with gemcitabine or cisplatin against
DMS-53 and A549 cells. Specifically, [Cu(DpC)Cl] in
combination with gemcitabine or cisplatin against DMS-53
After 21 days of treatment, tumors in nude mice receiving the
vehicle control grew rapidly, reaching an average net volume of
922
94 mm3 (Figure 8A). Importantly, Dp44mT,
Dp4e4mT·HCl, and DpC·HCl all significantly (p < 0.01−
0.001) reduced the net growth of DMS-53 tumor xenografts
relative to the control after the 21 day treatment schedule
(Figure 8A). However, the most effective agent was DpC·HCl
(8 mg/kg), which markedly and significantly (p < 0.001)
decreased tumor volume versus the control (Figure 8A). In fact,
tumor volume began to decrease in size relative to day 0
(untreated mice) after 16 days of treatment with DpC·HCl (8
mg/kg; Figure 8A). At a lower dose of 4 mg/kg, DpC·HCl also
significantly (p < 0.001) limited tumor volume to 79 14 mm3
relative to the vehicle control (922 94 mm3) after 21 days of
treatment. Notably, DpC·HCl (4 or 8 mg/kg) was significantly
(p < 0.001) more effective and less toxic (see Toxicology) than
the first-generation chelator, Dp44mT (0.75 mg/kg), that
significantly (p < 0.001) reduced net tumor volume to 136
28 mm3 after 21 days of treatment relative to the vehicle
control (Figure 8A). Notably, Dp4e4mT·HCl (3 and 6 mg/kg)
also significantly (p < 0.001−0.05) reduced net tumor volume
to 512 101 mm3 and 171 67 mm3, respectively, after 21
days treatment when compared to the vehicle-treated control
mice (Figure 8A). At 6 mg/kg, Dp4e4mT·HCl did not show
any significant difference in activity to Dp44mT (0.75 mg/kg;
Figure 8A). However, importantly, Dp4e4mT·HCl showed less
toxicity (see Toxicology). The length of treatment used in this
experiment was limited by tumor size in the vehicle control
group. As tumor size approached the maximum limit (1000
mm3) prescribed by the local animal ethics committee by day
21, it was not possible to continue the experiment. After 21
days of treatment, the mice were sacrificed and tumors were
excised and weighed. Control tumors weighed 0.83 0.11 g,
whereas mice receiving 4 or 8 mg/kg of DpC·HCl led to
significantly (p < 0.001) smaller tumors that weighed 0.15
cells resulted in CI values of 0.42
0.09 and 0.44
0.09,
respectively (Figure 7A). To ensure that these results were due
to the formation of copper thiosemicarbazone complexes, we
included CuCl2 alone as a relative control. In clear contrast,
these results showed that copper ions were nearly additive or
produced mildly antagonistic effects in combination with
gemcitabine or cisplatin against both cell types (Figure 7A).
As it is possible that the nature of drug−drug interactions can
change as a function of concentration or activity, we generated
a fractional affect−combination index plot to assess the
synergism between [Cu(DpC)Cl] and gemcitabine or
cisplatin.44,46 These data show that the interaction of
[Cu(DpC)Cl] with both these chemotherapeutics is synergistic
over a range of activity levels (Figure 7B).
Considering these results in Figure 7, it is conceivable that
the chelation of cellular copper may be important in terms of
the mechanism of synergism observed with the ligands and the
chemotherapeutics. Importantly, the synergism exhibited by the
second-generation chelators and their copper complexes was
generally comparable to standard drug therapy regimens. In
particular, DFO, Dp44mT, Dp4e4mT·HCl, and DpC·HCl were
all more synergistic against DMS-53 cells than the combination
of etoposide and cisplatin that is the foundation of chemo-
therapy regimens for small cell lung carcinoma.41,42 Against
A549 cells, [Cu(Dp4e4mT)Cl] and [Cu(DpC)Cl] exhibited
comparable synergism to the standard chemotherapy regimen
of gemcitabine and cisplatin. Collectively, Dp4e4mT·HCl and
DpC·HCl have synergistic potential in combination with
standard chemotherapy regimens for lung cancer.
0.02 g and 0.06
0.02 g, respectively (Figure 8B). Tumors
from mice given Dp44mT (0.75 mg/kg) weighed 0.27 0.04
g, a result that was comparable to tumors excised from
Dp4e4mT·HCl (6 mg/kg)-treated mice which weighed 0.40
0.10 g (Figure 8B). Hence, DpC·HCl was significantly (p <
0.05) more effective than Dp44mT at reducing tumor weight.
Importantly, all tumor weights were consistent with measured
tumor volumes.
Dp4e4mT and DpC Exhibit Potent and Selective In Vivo
Antitumor Activity Against Lung Cancer Xenografts in Mice
by both the Intravenous and Oral Routes. The marked
activity of Dp4e4mT and DpC observed in vitro indicated that
Toxicology Following iv Administration. Body and Organ
Weights. Measurement of these indices provided an initial
7237
dx.doi.org/10.1021/jm300768u | J. Med. Chem. 2012, 55, 7230−7244

Journal of Medicinal Chemistry
Article
hematological indices. As Fe chelating agents have the potential
to induce iron deficiency, it was vital to carefully assess these
parameters. Importantly, no significant difference was found in
the red blood cell (RBC) count between the vehicle-treated
control and most treatment groups (Supporting Information
Table S2). This was probably due to the low doses of ligands
utilized. However, in mice treated with DpC·HCl (8 mg/kg), a
significant (p < 0.05−0.01) decrease in RBC count and
hemoglobin level in addition to a significant (p < 0.01) increase
in reticulocyte count and % reticulocytes was found, indicating
a mild anemia. The higher dose of Dp4e4mT·HCl (6 mg/kg)
also resulted in a significant (p < 0.01) increase in reticulocyte
counts and % reticulocytes, while RBC count, hematocrit, and
hemoglobin level were not significantly altered (Supporting
Information Table S2). A significant (p < 0.05−0.01) elevation
of white blood cell (WBC) count in nearly all chelator
treatment groups was observed, except for the lower DpC·HCl
(4 mg/kg) dose. A significant (p < 0.05−0.01) increase in
platelet count was also evident in the 3 and 4 mg/kg doses of
Dp4e4mT·HCl and DpC·HCl, respectively.
Histology. Histological examination of the spleen, kidney,
liver, heart, lungs, and brain was performed by an independent
veterinary pathologist (Rothwell Consulting, Sydney, NSW)
after staining tissues with hematoxylin and eosin (H & E; to
assess general ultrastructural pathology), Perls’ (to detect the
presence of Fe-loading), and Gomori Trichrome (for detection
of fibrosis) by standard methods.10,11 Previous tumor xenograft
studies with the chelators, 3-AP and Dp44mT, showed that 3-
AP increased the number of hematopoietic cells in the splenic
red pulp, whereas Dp44mT induced cardiac fibrosis (as
demonstrated by Gomori Trichrome staining of cardiac
sections) at high nonoptimal doses.10 In the current study,
no increase in hematopoietic cells in the splenic red pulp was
found as a result of 21 days of treatment under all conditions
(data not shown). Assessment of Perl’s stained spleen sections
showed no alteration in hemosiderin derived from splenic
macrophages that are generally present in the splenic red pulp
of normal mice.11,47 The lack of these changes was consistent
with the observation that RBC numbers did not markedly
decrease in any treatment group with the exception of only a
modest decrease in RBC count in mice treated with the higher
DpC·HCl dose (8 mg/kg; Supporting Information Table S2).
In agreement with several studies from our laboratory,10,11
Dp44mT (0.75 mg/kg) induced fibrotic legions that were
particularly evident in the wall of the right ventricle and also in
the myocardium beneath the endocardium of the left ventricle
(Figure 8D). In contrast, treatment of mice with iv
Dp4e4mT·HCl (3 or 6 mg/kg) or DpC·HCl (4 or 8 mg/kg)
did not result in any evidence of cardiac fibrosis despite the
much higher doses used (Figure 8D).
Figure 8. DpC and Dp4e4mT demonstrate marked antitumor activity
against human DMS-53 lung tumor xenografts in nude mice after
intravenous (iv) administration and do not induce cardiotoxicity. (A)
Net tumor volume as a function of time of DMS-53 xenografts after iv
administration (tail vein) of vehicle control (30% propylene glycol/
saline), Dp44mT (0.75 mg/kg), Dp4e4mT·HCl (3 or 6 mg/kg), or
DpC·HCl (4 or 8 mg/kg). (B) Tumor weights (wet weight) from the
study in (A) after sacrifice of the animals on day 21. (C) Average body
weights of nude mice for the study in (A) during treatment with the
agents for 21 days. (D) Histological sections of the hearts taken from
the mice in (A) after 21 days of treatment and stained with Gomori
Trichrome to detect fibrosis (see arrow; magnification 100×). Results
in (A−C) are mean
SD (n = 6−7 mice/condition), while the
photographs in (D) are typical of the histology observed.
assessment of toxicity following euthanasia. In all treatment
groups, body weight did not significantly change from the
pretreatment measurement (Figure 8C), indicating that the
chelators did not affect appetite or otherwise induce cachexia.
Preliminary studies defined the MTD over a 42 day period, and
some moderate weight loss was expected. However, due to the
early termination after 21 days due to rapid tumor growth in
vehicle-treated mice, no weight loss was found. Additionally, no
significant differences were found in any organ weights between
the different treatment groups, apart from significant (p <
0.05−0.01) increases in liver weight in Dp4e4mT·HCl (3 and 6
mg/kg) and DpC·HCl (8 mg/kg) treated mice relative to the
vehicle control (Supporting Information Table S1).
A previous investigation from our laboratory showed that the
structurally related ligand, Bp44mT, caused partially reversible
cytoplasmic vacuolation of the liver, particularly when given
orally.11 In the current study, mild vacuolation was evident in
the livers of mice treated with Dp44mT and at the higher doses
of Dp4e4mT·HCl (6 mg/kg) and DpC·HCl (8 mg/kg).
However, there was no evidence of hepatocellular necrosis or
fibrosis. The mild cytoplasmic vacuolation of the liver in these
groups correlated to a small, but significant (p < 0.05−0.01)
increase in liver weights in these groups (Supporting
Information Table S1). All other H & E stained organs
appeared normal (data not shown), which agreed with the lack
Hematological Indices. To further explore whether chelator
treatment induced other toxic effects, we examined a panel of
7238
dx.doi.org/10.1021/jm300768u | J. Med. Chem. 2012, 55, 7230−7244

Journal of Medicinal Chemistry
Article
Figure 9. DpC and Dp4e4mT demonstrate marked antitumor activity against primary human DMS-53 lung tumor xenografts in nude mice after oral
administration and do not induce cardiotoxicity. (A) Net tumor volume as a function of time of DMS-53 xenografts after oral administration
(gavage) of vehicle control (30% propylene glycol/saline) or Dp4e4mT·HCl (7.5 or 10 mg/kg). (B) Tumor weights (wet weight) from the study in
(A) after sacrifice of the animals on day 20. (C) Net tumor volume as a function of time of DMS-53 xenografts after oral administration (gavage) of
vehicle control (30% propylene glycol/saline) or DpC·HCl (10 or 20 mg/kg). (D) Tumor weights (wet weight) from the study in (C) after sacrifice
of the animals on day 20. (E) Average body weights of nude mice for the study in (A) during treatment with the agents for 20 days. (F) Average
body weights of nude mice for the study in (C) during treatment with the agents for 20 days. (G) Histological sections of the hearts taken from the
mice in (A,C) after 20 days of treatment and stained with Gomori Trichrome to detect fibrosis (magnification 100×). It should be noted that
Dp44mT could not be given by the oral route as a relative comparison due to acute toxicity.11 Results in (A−F) are mean SD (n = 6−11 mice/
condition), while the photographs in (G) show typical heart histology from the treatment groups.
7239
dx.doi.org/10.1021/jm300768u | J. Med. Chem. 2012, 55, 7230−7244

Journal of Medicinal Chemistry
Article
of change in the weights of these organs in the control and
chelator-treated groups (Supporting Information Table S1).
Novel Second-Generation Chelators are also Orally Active
against DMS-53 Lung Carcinoma Xenografts. To further
assess the potential clinical utility of the second-generation
analogues, we also examined whether they exhibited activity
after oral administration. Assessment of “Lipinski’s Rule of
Five” indicated that the DpT class of thiosemicarbazones
should be orally available.48 Initial dose-range finding MTD
experiments conducted in DMS-53 tumor-bearing mice
suggested good oral activity and showed that the MTDs of
Dp4e4mT·HCl and DpC·HCl were 10 and 20 mg/kg,
respectively, over a 28-day treatment period (data not
shown). Hence, these doses were chosen for examination in
more comprehensive tumor growth studies. Because of the
rapid toxicity of Dp44mT after oral administration,11 it was not
possible to include it as a positive control. In these studies, mice
were gavaged with the vehicle control, Dp4e4mT·HCl, or
DpC·HCl 3 times/week using an established protocol in our
laboratory for thiosemicarbazones.11
mg/kg Dp4e4mT·HCl-treated group, RBC count (p < 0.05)
and hemoglobin level (p < 0.01) were significantly decreased
(Supporting Information Table S5). For both the 7.5 and 10
mg/kg Dp4e4mT·HCl-treated groups, the hematocrit was also
significantly (p < 0.05) reduced. However, in contrast, no
significant differences were found in any hematological
parameters in the DpC·HCl-treated mice (Supporting In-
formation Table S6), demonstrating the high tolerability of this
ligand.
Histology. There were no detectable changes in the spleen,
kidney, liver, heart, lungs, or brain after treatment with the two
analogues relative to the vehicle control. Importantly, in
agreement with the iv studies, oral treatment of mice with
Dp4e4mT·HCl or DpC·HCl at any of the doses tested did not
induce cardiac fibrosis (Figure 9G).
CONCLUSIONS
■
New drug therapies that are able to overcome existing problems
of drug resistance and poor efficacy are urgently required for
lung and other solid tumors. We have been actively exploring
the activity of a range of Fe chelators as potential cancer
treatments. Over the past decade, our research has provided
chelator structure−activity relationships that have culminated in
the design of the second-generation DpT and BpT analogues
presented herein. These new analogues are the most potent
antitumor agents our laboratories have designed and overcome
the limitations of earlier first-generation chelators, in particular
the cardiotoxicity evident with Dp44mT at high, nonoptimal
doses.10
After 20 days of treatment, mice bearing DMS-53 tumors
receiving the vehicle control grew quickly, reaching an average
net volume of 997
155 mm3, which was close to the
maximum ethical limit (1000 mm3) prescribed by the local
ethics committee, and thus, the experiment was terminated at
this time (Figure 9A). Doses of Dp4e4mT·HCl at 7.5 and 10
mg/kg markedly and significantly (p < 0.01−0.001) reduced
the net growth of DMS-53 tumor xenografts to 231 66 mm3
and 119 20 mm3, respectively (Figure 9A). Importantly, as
observed in the iv studies, measurement of tumor volumes in
mice that were treated orally with Dp4e4mT·HCl agreed with
excised tumor weights (Figure 9B). Additionally, DpC·HCl
when given orally at 10 or 20 mg/kg over 20 days, also
significantly (p < 0.05−0.01) inhibited tumor growth relative to
mice receiving the vehicle control (Figure 9C), and again, these
tumor volumes reflected the excised tumor weights (Figure
9D). Notably, because of the significant inhibition of tumor
growth observed after oral administration of Dp4e4mT·HCl
and DpC·HCl, our results suggest that these chelators may be
more effective than the orally administered chemotherapeutic
and ribonucleotide reductase inhibitor, hydroxyurea, that shows
poor efficacy.48
The specific structural design insight incorporated into the
second-generation analogues was “saturation” at the terminal
N4 atom with alkyl or aryl groups. This increased the
lipophilicity of the second-generation analogues and resulted
in superior antiproliferative activity relative to first-generation
chelators bearing one alkyl group and a hydrogen atom at the
N4 atom (Figure 5A). The best performing chelators,
DpC·HCl and Dp4e4mT·HCl, showed pronounced activity
against lung tumor xenografts without marked toxicological
effects by both the intravenous and oral routes. In particular,
the high efficacy and low toxicity of orally administered DpC is
of significant note. Clearly, oral administration demonstrates a
marked advantage for treating patients relative to the more
cumbersome intravenous route. It was also demonstrated that
these chelators acted synergistically with standard chemo-
therapeutics for lung cancer. These results provide a strong
basis for rapid preclinical development of Dp4e4mT·HCl and
especially DpC·HCl as potential antitumor agents.
Toxicology Following Oral Administration. Body and
Organ Weights. In agreement with the iv administration
study, body weights of chelator-treated animals did not
significantly change relative to the vehicle control (Figure
9E,F). Furthermore, in most cases, organ weights did not
significantly alter between the treated groups and the vehicle
control (Supporting Information Table S3 and S4), with the
exception that lung and heart weights were slightly, but
significantly (p < 0.05) higher in the 10 mg/kg Dp4e4mT·HCl-
treated group, and the spleen weight was significantly (p <
0.05) lower in the 7.5 mg/kg Dp4e4mT·HCl-treated group
(Supporting Information Table S3). However, these alterations
in weight could not be correlated to histology which was
normal (data not shown). These data further indicate that
treatment with these two analogues was well tolerated after oral
administration.
EXPERIMENTAL PROCEDURES
■
All commercial reagents were used without further purification. DFO
was obtained from Novartis (Basel, Switzerland). The chemo-
therapeutic agents, cisplatin, etoposide, and gemcitabine, were
obtained from Sigma-Aldrich (St. Louis, MO). Dp44mT and the
first-generation chelators, Dp4mT, Dp4eT, Dp4pT, Bp4mT, and
Bp4eT, were synthesized as described.9,19 Desferrioxamine (DFO) was
from Novartis (Basel, Switzerland).
Second-Generation DpT and BpT Chelators: General Syn-
thesis. The second-generation DpT and BpT chelators were
synthesized using a combination of established methods.19,26 Briefly,
carbon disulfide (0.2 mol) was added dropwise to the appropriate
amine (0.2 mol; e.g., N-methylcyclohexylamine for DpC) in NaOH
solution (250 mL, 0.8 M). This mixture was allowed to react until the
organic layer disappeared, and this occurred after approximately 4 h.
Next, sodium chloroacetate (0.2 mol) was added to the aqueous
Hematological Indices. Hematological indices were also
examined after oral administration of the two ligands
(Supporting Information Tables S5 and S6). In the 7.5 mg/
kg Dp4e4mT·HCl-treated group, the hemoglobin level was
slightly, but significantly (p < 0.05) reduced, while in the 10
7240
dx.doi.org/10.1021/jm300768u | J. Med. Chem. 2012, 55, 7230−7244

Journal of Medicinal Chemistry
Article
extract and allowed to react overnight at room temperature. The
addition of concentrated HCl (25 mL) gave the solid carboxymethyl
thiocarbamate intermediate. Approximately 0.08 mol of carboxymethyl
thiocarbamate was dissolved in 20 mL of hydrazine hydrate plus 10
mL of water. This was followed by five cycles of gentle heating (until
fuming) and cooling. The solution was then allowed to stand until fine
white crystals of thiosemicarbazide intermediate formed. A solution of
this intermediate (10 mmol) in water (15 mL) was added to di-2-
pyridyl ketone (10 mmol) dissolved in EtOH (15 mL). Then, 5 drops
of glacial acetic acid were added and the mixture was refluxed for 2 h
and cooled to 5 °C to give the thiosemicarbazone precipitate.
Preparation of the HCl salt for assessment of in vivo antitumor activity
was performed by dissolving the ligand (10 mM) in a minimum
volume of warm toluene (∼60 °C). When the solution had cooled to
approximately 40 °C, equimolar HCl was then added to give the HCl
salt.
turned dark brown. A fine olive-green−brown powder formed on
standing, which was filtered off, washed with EtOH (5 mL) and then
diethyl ether (5 mL), and dried in a vacuum desiccator overnight
(yield 41%). Anal. Calcd for C17H19N5CuSO2·H2O: C, 46.51%; H,
4.82%; N, 15.95%. Found: C, 46.35%; H 4.73%; N 16.20%. Electronic
spectra: (DMF) 302 nm (ε 17700 M−1 cm−1), 361 (9100), 434
(18200), (DMF:H2O 2:1) 302 nm (ε 18700 M−1 cm−1), 427 (21000).
IR (main peaks) 1488s, 1501s, 1433 m, 1370 m, 1311s, 1238 m, 1084
m, 970 m, 880 m, 824 m, 788s, 744 m. MS m/z (%) 361.3 (M − OAc,
100).
Cu(Dp4e4mT)Cl. Yield: 58%. Anal. Calcd for C15H16N5CuSCl: C,
45.34%; H, 4.06%; N, 17.62%. Found: C, 45.25%; H 4.01%; N 17.83%.
Electronic spectra: (DMF) 308 nm (ε 15200 M−1 cm−1), 364 (8400),
437 (15700), (DMF:H2O 2:1) 303 nm (ε 18000 M−1 cm−1), 427
(20000). IR (main peaks) 2973 m, 1593s, 1502s, 1464s, 1398s, 1371
m, 1289 m, 1237 m, 1079w, 1060 m, 984s, 822 m, 782vs, 741 m, 661
m, 634 m. MS m/z (%) 361.3 (M − Cl, 59).
Dp4e4mT and Dp4e4mT·HCl·5H2O. Yield: 44% (from CS2). Anal.
Calcd for C15H17N5S·H2O: C, 56.76%; H, 6.03%; N, 22.06%. Found:
C, 56.61%; H, 6.12%; N, 21.84%. Dp4e4mT·HCl·5H2O: Yield: 91%
(from Dp4e4mT). Anal. Calcd for C15H17N5S·HCl·5H2O: C, 41.42;
Cu(DpC)(OAc). Yield: 51%. Anal. Calcd for C21H24N5CuSO2: C,
53.21%; H, 5.01%; N, 14.77%. Found: C, 52.87%; H, 5.24%; N,
14.89%. Electronic spectra: (DMF) 306 nm (ε 17700 M−1 cm−1), 370
(9700), 437 (19500), (DMF:H2O 2:1) 304 nm (ε 16600 M−1 cm−1),
430 (19500). IR (main peaks) 2923 m, 1615 m, 1593 m, 1448 m,
1375vs, 1304vs, 1247s, 1182 m, 1152 m, 1085 m, 1003s, 926 m, 883
m,827w, 782s, 739s, 649 m. MS m/z (%) 415.4 (M − OAc, 100).
Cu(DpC)Cl. Yield: 43%. Anal. Calcd for C19H22N5CuSCl: C,
50.55%; H, 4.91%; N, 15.51%. Found: C, 50.38%; H 4.85%; N
15.75%. Electronic spectra: (DMF) 310 nm (ε 17200 M−1 cm−1), 371
(9900), 440 (18800), (DMF:H2O 2:1) 305 nm (ε 16200 M−1 cm−1),
430 (19900). IR (main peaks) 2924 m, 1592s, 1448vs, 1398s, 1370s,
1247s, 1301vs, 1244vs, 1179 m, 1154s, 1083 m, 1002vs, 883 m, 785s,
739 m, 655 m. MS m/z (%) 415.4 (M − Cl, 100).
1
H, 6.72; N, 16.10%. Found: C, 41.31; H, 6.45; N, 15.95%. H NMR
(DMSO-d6): δ 8.84 (d, 1H), 8.66 (d, 1H), 8.13−8.09 (t, 1H), 8.04−
7.97 (dt, 2H), 7.67−7.59 (dt, 3H), 3.32 (s, 3H), 1.20 (s, 3H). MS m/z
(%) 322.0 (M + Na, 46), 620.87 (M; dimer, + Na, 100).
Dp44eT. Yield: 38% (from CS2). Anal. Calcd for C16H21N5OS·H2O:
C, 57.98%; H, 6.56%; N, 21.13%. Found: C, 57.87%; H, 6.52%; N,
1
20.89%. H NMR (DMSO-d6): 8.60 (d, 1H), 8.80 (d, 1H), 7.92 (d,
1H), 7.62 (m, 1H), 7.59 (dd, 1H), 7.43 (t, 1H), 1.25 (t, 6H), 1.14 (q,
4H). MS m/z (%) 314.0 (M + H, 100), 336.1 (M + Na, 24), 648.9
(M; dimer, + Na, 14).
DpC and DpC·HCl·5H2O. Yield: 64% (from CS2). Anal. Calcd for
C19H23N5S: C, 64.56%; H, 6.56%; N, 19.81%. Found: C, 64.51%; H,
6.47%; N, 20.04%. DpC·HCl·5H2O: Yield 87% (from Dp4cycH4mT).
Anal. Calcd for C19H23N5S·HCl·5H2O: C, 47.54%; H, 7.14%; N,
1
Physical Methods. H NMR (400 MHz) spectra were acquired
using a Bruker Avance 400 NMR spectrometer with DMSO-d6 as the
solvent and internal reference (Me2SO: 1H NMR δ 2.49 ppm and 13
C
1
NMR δ 39.5 ppm vs TMS). Infrared spectra were measured with a
Varian Scimitar 800 FT-IR spectrophotometer with compounds being
dispersed as KBr discs or a Perkin-Elmer 1600 series spectrometer
using an ATR sample holder. Electronic spectra were measured with a
Perkin-Elmer Lambda 35 spectrophotometer. Cyclic voltammetry was
determined with a BAS 100B/W potentiostat utilizing a glassy carbon
working electrode and a Pt counter electrode. For experiments done in
DMF:water (7:3), a Ag/AgCl reference electrode was used. Potentials
are given relative to the NHE. For experiments in 100% DMF, a
nonaqueous Ag/Ag+ (DMF) electrode was constructed and the
potentials were referenced externally against the ferrocene/ferroce-
nium couple. All solutions contained 0.1 M Et4NClO4 as supporting
electrolyte and were purged with Ar before measurement. Electron
paramagnetic resonance (EPR) spectra were measured on a Bruker
ER200 instrument at X-band frequency (∼9.3 GHz) in 1 mM DMSO
frozen solutions at 77 K. Spectra were simulated with the program
EPR50F.49 Log Pcalc values were the average log P values calculated in
ChemDraw v4.5 using Crippen’s fragmentation,50 Viswanadhan’s
fragmentation,51 and Broto’s methods.52
X-ray Crystallography. Crystallographic data were acquired at
293 K on an Oxford Diffraction Gemini CCD diffractometer
employing graphite-monochromated Cu Kα radiation (1.5418 Å)
and operating within the range 2 < 2θ < 125°. Data reduction and
empirical absorption corrections (multiscan) were performed with the
Oxford Diffraction CrysAlisPro software. The structure was solved by
direct methods with SHELXS and refined by full-matrix least-squares
analysis with SHELXL-97.53 All non-H atoms were refined with
anisotropic thermal parameters. Molecular structure diagrams were
produced with ORTEP3,54 and all calculations were carried out within
the WinGX package.55 The data in CIF format has been deposited at
the Cambridge Crystallographic Data Centre with deposition number
CCDC 883394.
14.59%. Found: C, 47.06%; H, 6.65%; N, 14.95%. H NMR (DMSO-
d6): 8.82 (d, 1H), 8.61 (d, 1H), 8.04−7.90 (m, 3H), 7.62−7.56 (t,
2H), 7.50−7.46 (t, 1H), 3.18 (s, 3H), 1.83−1.49 (m, 7H), 1.40−1.10
(m, 3H). MS m/z (%) 376.1 (M + Na, 22), 729.0 (M; dimer + Na,
100).
Dp4p4mT. Yield: 51% (from CS2). Anal. Calcd for C19H17N5: C,
65.68%; H, 4.93%; N, 20.15%. Found: C, 65.59%; H, 5.16%; N,
1
20.17%. H NMR (DMSO-d6): 8.55 (s, 1H), 7.95 (t, 1H), 7.84 (q,
1H), 7.75 (d, 1H), 7.65−7.58 (m, 3H), 7.46−7.43 (m, 5H), 7.36 (t,
1H), 2.09 (s, 3H). MS m/z (%) 370.0 (M + Na, 16), 716.8 (M; dimer
+ Na, 100).
Dp4e4pT. Yield: 57% (from CS2). Anal. Calcd for C20H19N5: C,
66.45%; H, 5.29%; N, 19.37%. Found: C, 66.06%; H, 5.31%; N,
19.46%. 1H NMR (DMSO-d6): 8.85 (dt, 1H), 7.94 (t, 1H), 7.82−7.87
(m, 2H), 7.73 (d, 1H), 7.34 (dd, 1H), 7.39 (d, 2H), 7.44 (m, 2H),
7.67−7.59 (m, 3H), 4.25 (q, 2H), 1.17 (t, 3H). MS m/z (%) 384.0 (M
+ Na, 38), 744.8 (M; dimer + Na, 100).
Bp4m4eT. Yield: 77% (from 4-methyl-4ethyl-3-thiosemicarbazide).
Anal. Calcd for C16H18N4S: C, 64.40%; H, 6.07%; N, 18.78%. Found:
C, 64.47%; H, 6.18%; N, 18.65%. 1H NMR (DMSO-d6): 8.87 (d, 1H),
8.02 (dt, 1H), 7.63−7.57 (dd, 1H), 7.56−7.47 (m, 5H), 7.34 (d, 1H),
3.78 (q, 2H), 1.25 (t, 3H). MS m/z (%) 320.8 (M + Na, 27), 618.8
(M; dimer + Na, 100).
Bp44eT. Yield: 68% (from 4,4-diethyl-3-thiosemicarbazide). Anal.
Calcd for C17H20N4S: C, 65.35%; H, 6.45%; N, 17.93%. Found: C,
1
65.32%; H, 6.68%; N, 17.94%. H NMR (DMSO-d6): 8.84 (d, 1H),
8.02 (dt, 1H), 7.62−7.58 (dd, 1H), 7.58−7.47 (m, 5H), 7.34 (d, 1H),
3.86 (q, 4H), 1.28 (t, 6H). MS m/z (%) 313.1 (M + H, 92), 335.1 (M
+ Na, 100), 646.9 (M; dimer + Na, 19).
Copper Complexes of Selected Second-Generation DpT
Chelators. Copper complexes of DpC and Dp4e4mT were
synthesized in accordance with established procedures,27 as described
briefly here for [Cu(Dp4e4mT)(OAc)]. Dp4e4mT·H2O (0.317 g, 1
mmol) was dissolved in DMF (7 mL) with gentle heating and stirring.
A solution of Cu(OAc)2·3H2O (0.20 g, 1 mmol) in water (7 mL) was
added dropwise with stirring, and the ligand solution immediately
Cell Culture. Human SK-N-MC neuroepithelioma cells, human
MRC5 fibroblasts, human DMS-53 small cell lung carcinoma cells, and
human A549 nonsmall cell lung carcinoma cells were obtained from
the American Type Culture Collection (Manassas, VA) and grown as
7241
dx.doi.org/10.1021/jm300768u | J. Med. Chem. 2012, 55, 7230−7244

Journal of Medicinal Chemistry
Article
described.10,11 Cells were used within 2 months of purchase after
resuscitation of frozen aliquots. Cell lines were authenticated based on
viability, recovery, growth, morphology, and also cytogenetic analysis,
antigen expression, DNA profile, and iso-enzymology by the provider.
Cellular Proliferation Assay and Synergism Calculations. The
effect of chelators on cellular proliferation after a 72 h/37 °C
incubation was assessed using MTT assays as previously described.35
Direct cell counts were used to validate MTT assays.35 The effect of
multiple drug combinations on the proliferation of DMS-53 and A549
cells was also examined using the MTT assay. Combinations of
gemcitabine and cisplatin with Dp4e4mT·HCl and DpC·HCl, as well
as the control chelators, DFO and Dp44mT, were assessed. The
dose−response curves were analyzed using the methodology of Chou
and Talalay.45 Briefly, the median-effect plot and multiple-drug
equation derived by Chou and Talalay45 was employed to determine
the CI using CalcuSyn (Biosoft, Cambridge, MA).
mice were either iv injected or gavaged with the vehicle alone (30%
propylene glycol/0.9% saline).
Hematology. Blood was collected from the hearts of anesthetized
mice by cardiac puncture at the end of the experiment. Serum clinical
and hematological parameters were determined using a Konelab 20i
analyzer (Thermo-Electron Corporation, Vantaa, Finland) and Sysmex
K-4500 analyzer (TOA Medical Electronics Co., Kobe, Japan),
respectively.
Histology. Organs were dissected, fixed in 10% formalin, sectioned,
and stained with hematoxylin and eosin, Perl’s, or Gomori Trichrome
stains for microscopic examination, as described.11
Statistics. Results are expressed as mean
standard deviation
(SD). Statistical analysis comparing two groups was performed using
Student’s t-test. Data were considered statistically significant when p <
0.05.
Preparation of 56Fe- and 59Fe-Tf. Human Tf (Sigma-Aldrich)
was labeled with 56Fe or 59Fe (Dupont NEN, MA) to produce 59Fe-Tf
and 56Fe-Tf, respectively, as previously described.35,56,57 Briefly,
unbound 56Fe or 59Fe was removed by passage through a Sephadex
G25 column and exhaustive vacuum dialysis against a large excess of
0.15 M NaCl buffered to pH 7.4 with 1.4% NaHCO3 by standard
methods.56,57
ASSOCIATED CONTENT
* Supporting Information
Body and organ weights, hematological indices from in vivo
tumor xenograft experiments, experimental and simulated EPR
spectra, and cyclic voltammetry of Cu complexes. This material
is available free of charge via the Internet at http://pubs.acs.org.
■
S
Effect of Chelators on 59Fe Efflux from Cells. Iron efflux
experiments examining the ability of various chelators to mobilize 59Fe
from SK-N-MC cells were performed using established techniques.35
Briefly, following prelabeling of cells with 59Fe-Tf (0.75 μM) for 3 h/
37 °C, the cell cultures were washed on ice four times with ice-cold
PBS and then subsequently incubated with each chelator (25 μM) for
3 h/37 °C. The overlying media containing released 59Fe was then
separated from the cells using a Pasteur pipet. The radioactivity was
measured in both the cell pellet and the supernatant using a γ-
scintillation counter (Wallac Wizard 3, Turku, Finland). In these
studies, the novel ligands were compared to the previously
characterized chelators, DFO, NIH, and Dp44mT.12,35
AUTHOR INFORMATION
Corresponding Author
*Phone: +61-2-9036-6548. Fax: +61-2-9036-6549. E-mail: d.
richardson@med.usyd.edu.au.
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by: a project grant from the National
Health and Medical Research Council (NHMRC) Australia to
D.R.R. (grant 632778), a NHMRC Senior Principal Research
Fellowship to D.R.R. (grant 571123), a Cancer Institute New
South Wales Early Career Development Fellowship to D.B.L.
(grant 07/ECF/1-19), a Cancer Institute New South Wales
Early Career Development Fellowship to P.J.J. (grant 10/ECF/
2-15) and D.S.K. (grant 08/ECF/1-30), a Cancer Institute
Research Innovation Grant to D.R.R. and D.S.K. (grant 10/
RFG/2-50), a Australian Research Council Discovery Grant to
P.V.B. (grant DP1096029), and a King Abdul Aziz University
Scholarship to M.B. (grant Jeddah, Saudi Arabia).
Effect of Chelators at Preventing 59Fe Uptake from 59Fe-Tf
by Cells. The ability of the chelators to prevent cellular 59Fe uptake
from the serum Fe transport protein, 59Fe-Tf, was examined using
established techniques.35 Briefly, cells were incubated with 59Fe-Tf
(0.75 μM) for 3 h/37 °C in the presence of each of the chelators (25
μM). The cells were then washed four times with ice-cold PBS, and
internalized 59Fe was determined by standard techniques by incubating
the cell monolayer for 30 min/4 °C with the general protease, Pronase
(1 mg/mL; Sigma-Aldrich). The cells were removed from the
monolayer using a plastic spatula and centrifuged at 14000 rpm/1
min. The supernatant represents membrane-bound, Pronase-sensitive
59Fe that was released by the protease, while the Pronase-insensitive
fraction represents internalized 59Fe.35 The novel ligands were
compared to the previously characterized chelators, DFO, NIH, and
Dp44mT.12,35
ABBREVIATIONS USED
■
Tumor Xenograft Model in Nude Mice. All procedures for the
animal experiments were approved by the University of Sydney Animal
Ethics Committee. Female BALB/c nu/nu mice were used at 8−10
weeks of age (Laboratory Animal Services, University of Sydney).
Mice were housed under a 12 h light−dark cycle, routinely fed basal
rodent chow, and watered ad libitum. DMS-53 human tumor cells
were harvested and resuspended in a solution containing 1:1 ratio of
RPMI and Matrigel (BD Biosciences, San Jose, CA). Viable cells (5 ×
106 cells), as determined using trypan blue dye exclusion, were
subcutaneously injected into the right flanks of mice. Tumor size was
measured using digital Vernier calipers after engraftment and volume
was calculated, as described.58,59 The chelator treatment began after
the tumors reached a volume of 100 mm3.
In Vivo Chelator Administration. Chelators were dissolved in
30% propylene glycol in 0.9% saline. During the intravenous (iv)
studies, chelators were injected (100 μL) into the tail vein of nude
mice using a 19 gauge needle, 5 days/week (Monday−Friday). In the
oral administration experiments, chelators were given via oral gavage
(300 μL) on alternate days, 3 times/week, for the indicated time
periods (see Results) by standard methods.11 As a relevant control,
3-AP, 3-aminopyridine-2-carboxaldehyde thiosemicarbazone;
DFO, desferrioxamine; BpT, 2-benzoylpyridine thiosemicarba-
zone; Bp4mT, 2-benzoylpyridine 4-methyl-3-thiosemicarba-
zone; Bp4eT, 2-benzoylpyridine 4-ethyl-3-thiosemicarbazone;
Bp44mT, 2-benzoylpyridine 4,4-dimethyl-3-thiosemicarbazone;
Bp4e4mT, 2-benzoylpyridine 4-ethyl-4-methyl-3-thiosemicar-
bazone; Bp44eT, 2-benzoylpyridine 4,4-diethyl-3-thiosemicar-
bazone; CI, combination index; DpT, di-2-pyridyl ketone
thiosemicarbazone; Dp44mT, di-2-pyridyl ketone 4,4-dimethyl-
3-thiosemicarbazone; DpC, di-2-pyridyl ketone 4-cyclohexyl-4-
methyl-3-thiosemicarbazone; Dp4e4mT, di-2-pyridyl ketone 4-
ethyl-4-methyl-3-thiosemicarbazone; Dp44eT, di-2-pyridyl ke-
tone 4,4-diethyl-3-thiosemicarbazone; Dp4p4mT, di-2-pyridyl
ketone 4-phenyl-4-methyl-3-thiosemicarbazone; Dp4p4eT, di-
2-pyridyl ketone 4-phenyl-4-ethyl-3-thiosemicarbazone; MTD,
maximum tolerated dose; NIH, 2-hydroxy-1-naphthaldehyde
isonicotinoyl hydrazone; ROS, reactive oxygen species; Tf,
transferrin
7242
dx.doi.org/10.1021/jm300768u | J. Med. Chem. 2012, 55, 7230−7244

Journal of Medicinal Chemistry
Article
(17) Jansson, P. J.; Hawkins, C. L.; Lovejoy, D. B.; Richardson, D. R.
The Iron Complex of Dp44mT is Redox-Active and Induces Hydroxyl
Radical Formation: An EPR Study. J. Inorg. Biochem. 2010, 104, 1224−
1228.
(18) Lovejoy, D. B.; Jansson, P. J.; Brunk, U. T.; Wong, J.; Ponka, P.;
Richardson, D. R. Antitumor Activity of Metal-Chelating Compound
Dp44mT is Mediated by Formation of a Redox-Active Copper
Complex that Accumulates in Lysosomes. Cancer Res. 2011, 71, 5871−
5880.
(19) Richardson, D. R.; Sharpe, P. C.; Lovejoy, D. B.; Senaratne, D.;
Kalinowski, D. S.; Islam, M.; Bernhardt, P. V. Dipyridyl Thiosemi-
carbazone Chelators with Potent and Selective Antitumor Activity
Form Iron Complexes with Redox Activity. J. Med. Chem. 2006, 49,
6510−6521.
(20) Kovacevic, Z.; Chikhani, S.; Lui, G. Y.; Sivagurunathan, S.;
Richardson, D. R. The Iron-Regulated Metastasis Suppressor NDRG1
Targets NEDD4L, PTEN, and SMAD4 and Inhibits the PI3K and Ras
Signaling Pathways. Antioxid. Redox Signal. 2012, [online early access]
PMID 22462691, published online May 8, 2012 , DOI 10.1089/
ars.2011.4273 .
(21) Le, N. T.; Richardson, D. R. Iron Chelators with High
Antiproliferative Activity Up-Regulate the Expression of a Growth
Inhibitory and Metastasis Suppressor Gene: A Link Between Iron
Metabolism and Proliferation. Blood 2004, 104, 2967−2975.
(22) Liu, W.; Xing, F.; Iiizumi-Gairani, M.; Okuda, H.; Watabe, M.;
Pai, S. K.; Pandey, P. R.; Hirota, S.; Kobayashi, A.; Mo, Y. Y.; Fukuda,
K.; Li, Y.; Watabe, K. N-myc Downstream Regulated Gene 1
Modulates Wnt-Beta-Catenin Signalling and Pleiotropically Suppresses
Metastasis. EMBO Mol. Med. 2012, 4, 93−108.
(23) Gao, J.; Richardson, D. R. The Potential of Iron Chelators of the
Pyridoxal Isonicotinoyl Hydrazone Class as Effective Antiproliferative
Agents, IV: The Mechanisms Involved in Inhibiting Cell-Cycle
Progression. Blood 2001, 98, 842−850.
(24) Saletta, F.; Rahmanto, Y. S.; Siafakas, A. R.; Richardson, D. R.
Cellular Iron Depletion and the Mechanisms Involved in the Iron-
Dependent-Regulation of the Growth Arrest and DNA Damage
Family of Genes. J. Biol. Chem. 2011, 286, 35396−35406.
(25) Lovejoy, D. B.; Richardson, D. R. Novel “Hybrid” Iron
Chelators Derived from Aroylhydrazones and Thiosemicarbazones
Demonstrate Selective Antiproliferative Activity Against Tumor Cells.
Blood 2002, 100, 666−676.
(26) Scovill, J. P. A Facile Synthesis of Thiosemicarbazides and
Thiosemicarbazones by the Transamination of 4-Methyl-4-phenyl-3-
thiosemicarbazide. Phosphorus, Sulfur Silicon Relat. Elem. 1991, 60, 15−
19.
REFERENCES
■
(1) Kovacevic, Z.; Kalinowski, D. S.; Lovejoy, D. B.; Quach, P.;
Wong, J.; Richardson, D. R. Iron Chelators: Development of Novel
Compounds with High and Selective Anti-Tumour Activity. Curr.
Drug Delivery 2010, 7, 194−207.
(2) Yu, Y.; Wong, J.; Lovejoy, D. B.; Kalinowski, D. S.; Richardson,
D. R. Chelators at the Cancer Coalface: Desferrioxamine to Triapine
and Beyond. Clin. Cancer Res. 2006, 12, 6876−6883.
(3) Nyholm, S.; Mann, G. J.; Johansson, A. G.; Bergeron, R. J.;
Graslund, A.; Thelander, L. Role of Ribonucleotide Reductase in
Inhibition of Mammalian Cell Growth by Potent Iron Chelators. J.
Biol. Chem. 1993, 268, 26200−26205.
(4) Saletta, F.; Kovacevic, Z.; Richardson, D. R. Iron chelation:
Deciphering Novel Molecular Targets for Cancer Therapy. The Tip of
the Iceberg of a Web of Iron-Regulated Molecules. Future Med. Chem.
2011, 3, 1983−1986.
(5) Yu, Y.; Gutierrez, E.; Kovacevic, Z.; Saletta, F.; Obeidy, P.;
Rahmanto, Y. S.; Richardson, D. R. Iron Chelators for the Treatment
of Cancer. Curr. Med. Chem. 2012, 19, 2689−2702.
(6) Pahl, P. M.; Horwitz, L. D. Cell Permeable Iron Chelators as
Potential Cancer Chemotherapeutic Agents. Cancer Invest. 2005, 23,
683−691.
(7) Knox, J. J.; Hotte, S. J.; Kollmannsberger, C.; Winquist, E.; Fisher,
B.; Eisenhauer, E. A. Phase II Study of Triapine in Patients with
Metastatic Renal Cell Carcinoma: A Trial of the National Cancer
Institute of Canada Clinical Trials Group (NCIC IND.161). Invest.
New Drugs 2007, 25, 471−477.
(8) Ma, B.; Goh, B. C.; Tan, E. H.; Lam, K. C.; Soo, R.; Leong, S. S.;
Wang, L. Z.; Mo, F.; Chan, A. T.; Zee, B.; Mok, T. A Multicenter
Phase II Trial of 3-Aminopyridine-2-carboxaldehyde Thiosemicarba-
zone (3-AP, Triapine) and Gemcitabine in Advanced Non-Small-Cell
Lung Cancer with Pharmacokinetic Evaluation using Peripheral Blood
Mononuclear Cells. Invest. New Drugs 2008, 26, 169−173.
(9) Kalinowski, D. S.; Yu, Y.; Sharpe, P. C.; Islam, M.; Liao, Y. T.;
Lovejoy, D. B.; Kumar, N.; Bernhardt, P. V.; Richardson, D. R. Design,
Synthesis, and Characterization of Novel Iron Chelators: Structure−
Activity Relationships of the 2-Benzoylpyridine Thiosemicarbazone
Series and their 3-Nitrobenzoyl Analogues as Potent Antitumor
Agents. J. Med. Chem. 2007, 50, 3716−3729.
(10) Whitnall, M.; Howard, J.; Ponka, P.; Richardson, D. R. A Class
of Iron Chelators with a Wide Spectrum of Potent Antitumor Activity
that Overcomes Resistance to Chemotherapeutics. Proc. Natl. Acad.
Sci. U.S.A. 2006, 103, 14901−14906.
(11) Yu, Y.; Rahmanto, Y. S.; Richardson, D. R. Bp44mT: An Orally
Active Iron Chelator of the Thiosemicarbazone Class with Potent
Anti-Tumour Efficacy. Br. J. Pharmacol. 2012, 165, 148−166.
(12) Yuan, J.; Lovejoy, D. B.; Richardson, D. R. Novel Di-2-pyridyl-
Derived Iron Chelators with Marked and Selective Antitumor Activity:
In Vitro and In Vivo Assessment. Blood 2004, 104, 1450−1458.
(13) Quach, P.; Gutierrez, E.; Basha, M. T.; Kalinowski, D. S.;
Sharpe, P. C.; Lovejoy, D. B.; Bernhardt, P. V.; Jansson, P. J.;
Richardson, D. R. Methemoglobin Formation by 3-AP, Dp44mT and
Other Anti-Cancer Thiosemicarbazones: Identification of Novel
Thiosemicarbazones and Therapeutics that Prevent this Effect. Mol.
Pharmacol. 2012, 82, 105−114.
(14) Kovacevic, Z.; Chikhani, S.; Lovejoy, D. B.; Richardson, D. R.
Novel Thiosemicarbazone Iron Chelators Induce Up-Regulation and
Phosphorylation of the Metastasis Suppressor N-Myc Down-Stream
Regulated Gene 1: A New Strategy for the Treatment of Pancreatic
Cancer. Mol. Pharmacol. 2011, 80, 598−609.
(27) Jansson, P. J.; Sharpe, P. C.; Bernhardt, P. V.; Richardson, D. R.
Novel Thiosemicarbazones of the ApT and DpT Series and their
Copper Complexes: Identification of Pronounced Redox Activity and
Characterization of their Antitumor Activity. J. Med. Chem. 2010, 53,
5759−5769.
(28) Richardson, D. R.; Kalinowski, D. S.; Richardson, V.; Sharpe, P.
C.; Lovejoy, D. B.; Islam, M.; Bernhardt, P. V. 2-Acetylpyridine
Thiosemicarbazones are Potent Iron Chelators and Antiproliferative
Agents: Redox Activity, Iron Complexation and Characterization of
their Antitumor Activity. J. Med. Chem. 2009, 52, 1459−1470.
(29) Addison, A. W.; Rao, T. N.; Reedijk, J.; Vanrijn, J.; Verschoor,
G. C. Synthesis, Structure, and Spectroscopic Properties of Copper(II)
Compounds Containing Nitrogen Sulfur Donor LigandsThe
Crystal and Molecular Structure of Aqua[1,7-bis(N-methylbenzimida-
zol-2′-yl)-2,6-dithiaheptane]copper(II) perchlorate. Dalton Trans.
1984, 1349−1356.
(15) Rao, V. A.; Klein, S. R.; Agama, K. K.; Toyoda, E.; Adachi, N.;
Pommier, Y.; Shacter, E. B. The Iron Chelator Dp44mT causes DNA
Damage and Selective Inhibition of Topoisomerase IIalpha in Breast
Cancer Cells. Cancer Res. 2009, 69, 948−957.
(30) Ali, M. A.; Mirza, A. H.; Butcher, R. J.; Bernhardt, P. V.; Karim,
M. R. Self-Assembling Dicopper(II) Complexes of Di-2-pyridyl
Ketone Schiff Base Ligands Derived from S-Alkyldithiocarbazates.
Polyhedron 2011, 30, 1478−1486.
(16) Tian, J.; Peehl, D. M.; Zheng, W.; Knox, S. J. Antitumor and
Radiosensitization Activities of the Iron Chelator HDp44mT are
Mediated by Effects on Intracellular Redox Status. Cancer Lett. 2010,
298, 231−237.
(31) Ali, M. A.; Mirza, A. H.; Nazimuddin, M.; Ahmed, R.; Gahan, L.
R.; Bernhardt, P. V. Synthesis and Characterization of Mono- and Bis-
Ligand Zinc(II) and Cadmium(II) Complexes of the Di-2-
pyridylketone Schiff Base of S-Benzyl Dithiocarbazate (Hdpksbz)
7243
dx.doi.org/10.1021/jm300768u | J. Med. Chem. 2012, 55, 7230−7244

Journal of Medicinal Chemistry
Article
and the X-ray Crystal Structures of the [Zn(dpksbz)(2)] and
[Cd(dpksbz)NCS](2) Complexes. Polyhedron 2003, 22, 1471−1479.
(32) Hathaway, B. J. A New Look at the Stereochemistry and
Electronic-Properties of Complexes of the Copper(II) Ion. Struct.
Bonding (Berlin, Ger.) 1984, 57, 55−118.
(33) Hamid, M. H. S. A.; Ali, M. A.; Mirza, A. H.; Bernhardt, P. V.;
Moubaraki, B.; Murray, K. S. Magnetic, Spectroscopic and X-ray
Crystallographic Structural Studies on Copper(II) Complexes of
Tridentate NNS Schiff Base Ligands formed from 2-Acetylpyrazine
and S-Methyl- and S-Benzyldithiocarbazates. Inorg. Chim. Acta 2009,
362, 3648−3656.
(34) Kalinowski, D. S.; Sharpe, P. C.; Bernhardt, P. V.; Richardson,
D. R. Design, Synthesis, and Characterization of New Iron Chelators
with Anti-Proliferative Activity: Structure−Activity Relationships of
Novel Thiohydrazone Analogues. J. Med. Chem. 2007, 50, 6212−6225.
(35) Richardson, D. R.; Tran, E. H.; Ponka, P. The Potential of Iron
Chelators of the Pyridoxal Isonicotinoyl Hydrazone Class as Effective
Antiproliferative Agents. Blood 1995, 86, 4295−4306.
(36) Stefani, C.; Punnia-Moorthy, G.; Lovejoy, D. B.; Jansson, P. J.;
Kalinowski, D. S.; Sharpe, P. C.; Bernhardt, P. V.; Richardson, D. R.
Halogenated 2′-Benzoylpyridine Thiosemicarbazone (XBpT) Chela-
tors with Potent and Selective Antineoplastic Activity: Relationship to
Intracellular Redox Activity. J. Med. Chem. 2011, 54, 6936−6948.
(37) Yu, Y.; Suryo Rahmanto, Y.; Hawkins, C. L.; Richardson, D. R.
The Potent and Novel Thiosemicarbazone Chelators Di-2-pyridylke-
tone-4,4-dimethyl-3-thiosemicarbazone and 2-Benzoylpyridine-4,4-di-
methyl-3-thiosemicarbazone Affect Crucial Thiol Systems Required for
Ribonucleotide Reductase Activity. Mol. Pharmacol. 2011, 79, 921−
931.
(38) Nurtjahja-Tjendraputra, E.; Fu, D.; Phang, J. M.; Richardson, D.
R. Iron Chelation Regulates Cyclin D1 Expression via the Proteasome:
A Link to Iron Deficiency-Mediated Growth Suppression. Blood 2007,
109, 4045−4054.
(39) Richardson, D. R.; Lovejoy, D. B. Dipyridyl thiosemicarbazone
compounds and use in therapy. Provisional Patent Application number
2010905539, 2010.
(40) D’Addario, G.; Fruh, M.; Reck, M.; Baumann, P.; Klepetko, W.;
Felip, E. Metastatic Non-Small-Cell Lung Cancer: ESMO Clinical
Practice Guidelines for Diagnosis, Treatment and Follow-Up. Ann.
Oncol. 2010, 21 (Suppl 5), v116−v119.
(41) Simon, G. R.; Turrisi, A. Management of Small Cell Lung
Cancer: ACCP Evidence-Based Clinical Practice Guidelines (2nd
edition). Chest 2007, 132, 324S−339S.
(50) Ghose, A. K.; Crippen, G. M. Atomic Physicochemical
Parameters for Three-Dimensional-Structure-Directed Quantitative
Structure−Activity Relationships. 2. Modeling Dispersive and Hydro-
phobic Interactions. J. Chem. Inf. Comput. Sci. 1987, 27, 21−35.
(51) Viswanadhan, V. N.; Ghose, A. K.; Revankar, G. R.; Robins, R.
K. Atomic Physiochemical Parameters for 3-Dimensional-Structure
Directed Quantitative Structure−Activity Relationships 4. Additional
Parameters for Hydrophobic and Dispersive Interactions and their
Application for an Automated Superposition of Certain Naturally-
Occuring Nucleoside Antibiotics. J. Chem. Inf. Comput. Sci. 1987, 29,
163−172.
(52) Broto, P.; Moreau, G.; Vandycke, C. Molecular Structures:
Perception, Autocorrelation Descriptor and SAR Studies. System of
Atomic Contributions for the Calculation of the n-Octanol/Water
Partition Coefficients. Eur. J. Med. Chem. Chim. Theor. 1984, 19, 71−
78.
(53) Sheldrick, G. A Short History of SHELX. Acta Crystallogr., Sect.
A: Found. Crystallogr. 2008, 64, 112−122.
(54) Farrugia, L. J. ORTEP-3 for WindowsA Version of ORTEP-
III with a Graphical User Interface (GUI). J. Appl. Crystallogr. 1997,
30, 565.
(55) Farrugia, L. J. WINGXAn Integrated System of Windows
Programs for the Solution, Refinement and Analysis of Single Crystal
X-Ray Diffraction Data. J. Appl. Crystallogr. 1999, 32, 837−838.
(56) Richardson, D. R.; Baker, E. The Uptake of Iron and Transferrin
by the Human Malignant Melanoma Cell. Biochim. Biophys. Acta 1990,
1053, 1−12.
(57) Richardson, D. R.; Baker, E. Two Mechanisms of Iron Uptake
from Transferrin by Melanoma Cells. The Effect of Desferrioxamine
and Ferric Ammonium Citrate. J. Biol. Chem. 1992, 267, 13972−
13979.
(58) Balsari, A.; Tortoreto, M.; Besusso, D.; Petrangolini, G.;
Sfondrini, L.; Maggi, R.; Menard, S.; Pratesi, G. Combination of a
CpG-Oligodeoxynucleotide and a Topoisomerase I Inhibitor in the
Therapy of Human Tumour Xenografts. Eur. J. Cancer 2004, 40,
1275−1281.
(59) Sanceau, J.; Poupon, M. F.; Delattre, O.; Sastre-Garau, X.;
Wietzerbin, J. Strong Inhibition of Ewing Tumor Xenograft Growth by
Combination of Human Interferon-alpha or Interferon-beta with
Ifosfamide. Oncogene 2002, 21, 7700−7709.
(42) Sorensen, M.; Pijls-Johannesma, M.; Felip, E. Small-Cell Lung
Cancer: ESMO Clinical Practice Guidelines for Diagnosis, Treatment
and Follow-Up. Ann. Oncol. 2010, 21 (Suppl 5), v120−v125.
(43) Chou, T. C. Preclinical versus Clinical Drug Combination
Studies. Leuk. Lymphoma 2008, 49, 2059−2080.
(44) Chou, T. C. Drug Combination Studies and their Synergy
Quantification using the Chou−Talalay Method. Cancer Res. 2010, 70,
440−446.
(45) Chou, T. C.; Talalay, P. Quantitative Analysis of Dose−Effect
Relationships: The Combined Effects of Multiple Drugs or Enzyme
Inhibitors. Adv. Enzyme Regul. 1984, 22, 27−55.
(46) Chou, T. C. Theoretical Basis, Experimental Design, and
Computerized Simulation of Synergism and Antagonism in Drug
Combination Studies. Pharmacol. Rev. 2006, 58, 621−681.
(47) Suttie, A. W. Histopathology of the Spleen. Toxicol. Pathol.
2006, 34, 466−503.
(48) Yu, Y.; Kalinowski, D. S.; Kovacevic, Z.; Siafakas, A. R.; Jansson,
P. J.; Stefani, C.; Lovejoy, D. B.; Sharpe, P. C.; Bernhardt, P. V.;
Richardson, D. R. Thiosemicarbazones from the Old to New: Iron
Chelators that are More than just Ribonucleotide Reductase
Inhibitors. J. Med. Chem. 2009, 52, 5271−5294.
(49) Martinelli, R. A.; Hanson, G. R.; Thompson, J. S.; Holmquist,
B.; Pilbrow, J. R.; Auld, D. S.; Vallee, B. L. Characterization of the
Inhibitor Complexes of Cobalt Carboxypeptidase A by Electron
Paramagnetic Resonance Spectroscopy. Biochemistry 1989, 28, 2251−
2258.
7244
dx.doi.org/10.1021/jm300768u | J. Med. Chem. 2012, 55, 7230−7244