Heterocyclic dithiocarbazate iron chelators: Fe coordination chemistry and biological activity

Article abstract of DOI:10.1039/c2dt12387h

The iron coordination and biological chemistry of a series of heterocyclic dithiocarbazate Schiff base ligands is reported with regard to their activity as Fe chelators for the treatment of Fe overload and also cancer. The ligands are analogous to tridentate heterocyclic hydrazone and thiosemicarbazone chelators we have studied previously which bear NNO and NNS donor sets. The dithiocarbazate Schiff base ligands in this work also are NNS chelators and form stable low spin ferric and ferrous complexes and both have been isolated. In addition an unusual hydroxylated ligand derivative has been identified via an Fe-induced oxidation reaction. X-ray crystallographic and spectroscopic characterisation of these complexes has been carried out and also the electrochemical properties have been investigated. All Fe complexes exhibit totally reversible Fe^III/II couples in mixed aqueous solvents at potentials higher than found in analogous thiosemicarbazone Fe complexes. The ability of the dithiocarbazate Schiff base ligands to mobilise Fe from cells and also to prevent Fe uptake from transferrin was examined and all ligands were effective in chelating intracellular Fe relative to known controls such as the clinically important Fe chelator desferrioxamine. The Schiff base ligands derived from 2-pyridinecarbaldehyde were non-toxic to SK-N-MC neuroepithelioma (cancer) cells but those derived from the ketones 2-acetylpyridine and di-2-pyridyl ketone exhibited significant antiproliferative activity.

Full text of DOI:10.1039/c2dt12387h

View Online / Journal Homepage / Table of Contents for this issue
This paper is published as part of a Dalton Transactions themed issue entitled:
Application of inorganic chemistry for
non-cancer therapeutics
Guest Editor: Katherine J. Franz
Published in issue 21, 2012 of Dalton Transactions
Image reproduced with permission of Christophe Biot
Articles published in this issue include:
The therapeutic potential of metal-based antimalarial agents: Implications for the
mechanism of action
Christophe Biot, William Castro, Cyrille Y. Botté and Maribel Navarro
Dalton Trans., 2012, DOI: 10.1039/C2DT12247B
Chelation therapy in Wilson's disease: from D-penicillamine to the design of selective
bioinspired intracellular Cu(I) chelators
Pascale Delangle and Elisabeth Mintz
Dalton Trans., 2012, DOI: 10.1039/C2DT12188C
Therapeutic potential of selenium and tellurium compounds: Opportunities yet
unrealised
Edward R. T. Tiekink
Dalton Trans., 2012, DOI: 10.1039/C2DT12225A
Visit the Dalton Transactions website for more cutting-edge bioinorganic chemistry
research
www.rsc.org/dalton

Dynamic Article Links
Dalton
Transactions
Cite this: Dalton Trans., 2012, 41, 6536
www.rsc.org/dalton
PAPER
Heterocyclic dithiocarbazate iron chelators: Fe coordination chemistry and
biological activity†
Maram T. Basha,a Jy D. Chartres,a Namfon Pantarat,b Mohammad Akbar Ali,c Aminul Huq Mirza,c
Danuta S. Kalinowski,b Des R. Richardson*b and Paul V. Bernhardt*a
Received 11th December 2011, Accepted 13th January 2012
DOI: 10.1039/c2dt12387h
The iron coordination and biological chemistry of a series of heterocyclic dithiocarbazate Schiff base
ligands is reported with regard to their activity as Fe chelators for the treatment of Fe overload and also
cancer. The ligands are analogous to tridentate heterocyclic hydrazone and thiosemicarbazone chelators
we have studied previously which bear NNO and NNS donor sets. The dithiocarbazate Schiff base ligands
in this work also are NNS chelators and form stable low spin ferric and ferrous complexes and both have
been isolated. In addition an unusual hydroxylated ligand derivative has been identified via an Fe-induced
oxidation reaction. X-ray crystallographic and spectroscopic characterisation of these complexes has been
carried out and also the electrochemical properties have been investigated. All Fe complexes exhibit
totally reversible FeIII/II couples in mixed aqueous solvents at potentials higher than found in analogous
thiosemicarbazone Fe complexes. The ability of the dithiocarbazate Schiff base ligands to mobilise Fe
from cells and also to prevent Fe uptake from transferrin was examined and all ligands were effective in
chelating intracellular Fe relative to known controls such as the clinically important Fe chelator
desferrioxamine. The Schiff base ligands derived from 2-pyridinecarbaldehyde were non-toxic to SK-N-
MC neuroepithelioma (cancer) cells but those derived from the ketones 2-acetylpyridine and di-2-pyridyl
ketone exhibited significant antiproliferative activity.
administered by long and frequent periods of subcutaneous infu-
Introduction
sion (12–24 h per day, 5–6 times a week). In 2005, the triazole
deferasirox (Exjade®, Chart 1),5–8 was approved by the US
In the treatment of acute iron (Fe) overload disorders, the admin-
istration of Fe chelating ligands is essential.1,2 In humans, intesti-
Food and Drug Administration (FDA) for oral administration in
the treatment of Fe overload and was the first drug of its type to
nal Fe absorption is tightly regulated3 and when Fe levels are
adequate, further Fe uptake is inhibited. In cases where severe Fe
overload occurs, commonly through necessary blood transfu-
sions in the treatment of anaemias such as β-thalassemia, Fe
overload becomes acute, as humans have no mechanism for
actively excreting excess Fe. Failure to remove excess Fe leads to
its accumulation in vital organs such as the heart and liver, and if
left untreated, leads to irreversible organ damage and death.
Few Fe chelators have been successful in treating Fe overload
in the clinic. Desferrioxamine (DFO; Chart 1) was for many
years the only drug approved worldwide for the treatment of Fe
overload. Its widespread use continues today, despite its major
drawback of not being orally active.4 In fact, DFO must be
be approved worldwide. However, possible side effects such as
renal and liver damage and gastrointestinal haemorrhage in some
patients were added as warnings to the packaging of this drug in
2010.‡
The hydroxypyridinones (such as deferiprone also known as
L1 and more recently Ferriprox®) emerged in the 1980s9 and
also offered the hope of oral activity, but controversy surround-
ing their efficacy and toxicity10,11 hampered drug development
over the years. In 2011, the FDA finally granted approval for the
‘second line’ use of deferiprone as an orally administered drug
in the treatment of Fe overload,§ although there are still unre-
solved questions over its efficacy. Notably, the most promising
clinical results with deferiprone have been obtained when it is
co-administered with DFO where it appears to be effective
against Fe loading in heart tissue.12
aSchool of Chemistry and Molecular Biosciences, University of
Queensland, Brisbane 4072, Australia. E-mail: p.bernhardt@uq.edu.au
bDiscipline of Pathology, Sydney Medical School, University of Sydney,
Sydney 2006, Australia. E-mail: d.richardson@med.usyd.edu.au
cFaculty of Science, Universiti Brunei Darussalam, Gadong BE1410,
Brunei Darussalam
†Electronic supplementary information (ESI) available. CCDC reference
numbers 858720–858723. For ESI and crystallographic data in CIF or
other electronic format see DOI: 10.1039/c2dt12387h
The tridentate 2-pyridinecarbaldehyde isonicotinoyl hydra-
zone (HPCIH) family of Fe chelators (Chart 1) have shown
‡http://www.fda.gov/safety/medwatch/safetyinformation/safetyalertsfor-
humanmedicalproducts/ucm200850.htm
§http://www.fda.gov/downloads/Drugs/InformationOnDrugs/
UCM086233.pdf
6536 | Dalton Trans., 2012, 41, 6536–6548
This journal is © The Royal Society of Chemistry 2012

Chart 1 Fe chelators in clinical use for the treatment of Fe overload disease (i.e. DFO, deferiprone and deferasirox) as well as hydrazone, thiohydra-
zone, thiosemicarbazone and hydrazine Fe chelators with known biological activity. Donor atoms are in bold type.
promise as alternatives to existing compounds for the treatment
of Fe overload.13–16 Investigations by our laboratories have
shown that they form stable FeII complexes, in contrast to the
FeIII complexes formed by DFO, deferiprone and deferasirox.1,2,4
Subsequent studies have shown that substitutions to the C-atom
adjacent to the pyridyl ring lead to a marked change in biological
activity, for example the HPCIH analogues based on 2-acetylpyr-
idine (HAPIH series)17 or di-2-pyridyl ketone (HPKIH ana-
logues)18 exhibit significant cytotoxicity in contrast to the
benign HPCIH chelators.16 Although deleterious for the treat-
ment of chronic Fe overload (where chelators must be adminis-
tered lifelong), cytotoxic Fe chelators may prove to be a novel
and effective method in cancer therapy.19–26 Of interest, a signifi-
cant increase in cytotoxicity and activity against neoplastic cells
was found by substitution of the O-donor of the HPKIH and
HAPIH hydrazones with an S-atom leading to the thiohydrazone
(HPKTBH)27 and thiosemicarbazone (HDp44mT)28–30 families
of chelators (Chart 1).
In the current investigation, we have assessed a series of
dithiocarbazate Schiff base analogues (Chart 2) in their reactions
with Fe. The ligands share the same NNS, tridentate donor set as
the HDp44mT and HPKTBH analogues, but the terminal substi-
tuent is a mercaptomethyl or mercaptobenzyl group rather than
an amine or aromatic ring. Dithiocarbazate Schiff bases are well
described and a number of transition metal complexes are
known31–41 and these compounds have shown a wide spectrum
of biological activity. Little is known about the Fe coordination
chemistry of these dithiocarbazate Schiff base ligands42,43 and
given the resemblance to the highly active thiosemicarbazone Fe
chelators we have studied previously,21,28–30 this was the focus
of the current work. Of particular interest are the relationships
between chelator structure, redox properties of the Fe complex
and the biological activity (including anti-proliferative efficacy).
Modification of the substituents adjacent to the donor group is
known to have a marked effect on redox potentials in com-
pounds from the heterocyclic thiosemicarbazone28–30 and
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 6536–6548 | 6537

Chart 2 Dithiocarbazate Schiff base ligands investigated in this work and related ligands. Donor atoms are in bold type.
hydrazone16–18 series due to extended conjugation along the
Free ligand syntheses
ligand backbone. Replacement of the terminal amino group
with a thioether brings about both differences in inductive and
resonance effects, which then are communicated to the coordi-
nated metal ion. In this contribution, we investigate the Fe
coordination chemistry of these dithiocarbazate ligands, their
effectiveness at mobilising intracellular Fe and their anti-proli-
ferative activity.
2-Acetylpyridine S-methyldithiocarbazate (HAPSMC). A sus-
pension of S-methyl dithiocarbazate (6.11 g, 0.05 mol) in 75 mL
of isopropanol was heated until it dissolved. This mixture was
cooled to room temperature then filtered and the residue was
washed with isopropyl alcohol then discarded. The filtrate was
treated with 2-acetylpyridine (6.05 g, 0.05 mol) added drop-
wise. The mixture was stirred for 24 h at room temperature. The
resulting yellow precipitate was filtered off, washed with isopro-
Experimental
panol and recrystallised from ethanol to give a yellow crystalline
product. Yield (7.3 g, 65%). Microanalysis found C, 48.1; H,
5.0; N, 18.8%; Calculated for C9H11N3S2: C, 48.0; H, 4.9; N,
Safety note
18.7%. IR: ν (cm−1); 3144m, 2913w, 2114w, 1665w, 1615w,
ˉ
Perchlorate salts are potentially explosive. Although we experi-
enced no problems with the compounds described below they
should never be heated in the solid state or scraped from sintered
glass frits.
1568m, 1562m, 1484m, 1460s, 1428s, 1366m, 1336m, 1254vs,
1144w, 115w, 1060vs, 1045m, 990w, 949vs, 886w, 774vs,
1
734vs, 678m, 642s, 620w, and 562m. H NMR (DMSO-d6): δ
12.57 (s, 1H, NH), 8.62 (m, 1H, py), 8.07 (m, 1H, py), 7.88 (m,
1H, py), 7.45 (m, 1H, py), 2.52 (s, 3H, S–CH3), 2.45 (s, 3H, C–
CH3). 13C NMR (DMSO-d6): 12.9, 17.0, 120.3, 124.6, 136.8,
148.8, 151.7, 154.3 and 200.6 ppm.
Reagents
S-Methyldithiocarbazate,44 and S-benzyldithiocarbazate45 were
prepared according to literature syntheses. All other reagents
were obtained commercially and used without further
purification.
2-Acetylpyridine S-benzyldithiocarbazate (HAPSBC). S-
Benzyl dithiocarbazate (1.98 g, 0.01 mol) was dissolved in hot
absolute ethanol (50 mL). To this solution was added drop-wise
6538 | Dalton Trans., 2012, 41, 6536–6548
This journal is © The Royal Society of Chemistry 2012

an equimolar amount of 2-acetylpyridine (1.21 g, 0.01 mol). The
mixture was heated with stirring for 30 min and allowed to stand
at room temperature until the product precipitated. The product
was filtered off and washed with ethanol to afford dark yellow
crystals. Yield (2.29 g, 76%). Microanalysis found C, 59.5; H,
5.1; N, 14.0%; Calculated for C15H15N3S2: C, 59.8; H, 5.0; N,
ethanol to give pale yellow crystals. Yield (7.08 g, 67%). Micro-
analysis found C, 45.6; H, 4.3; N, 20.1%; Calculated for
C H N S : C, 45.5; H, 4.3; N, 19.9%. IR: ν (cm−1); 3090w,
ˉ
8
9 3 2
2773m, 2081w, 1582m, 1529s, 1465s, 1434m, 1311m, 1278vs,
1428s, 1145m, 1106m, 1034vs, 997s, 927s, 875m, 774vs, 725m,
1
678s, and 596m. H NMR (DMSO-d6): δ 13.44 (s, 1H, NH),
13.9%. IR: ν (cm−1); 3171m, 2969w, 2120w, 1668w, 1600w,
8.62 (m, 1H, py), 8.26 (s, 1H, CH), 7.91 (m, 1H, py), 7.87 (m,
1H, py), 7.44 (m, 1H, py), 2.53 (s, 3H, SCH3). 13C NMR
(DMSO-d6): 16.8, 120, 124.9, 137, 146.4, 149.8, 152.3, and
199.3 ppm.
ˉ
1580m, 1563m, 1489s, 1461vs, 1416s, 1367m, 1338m, 1268s,
1230m, 1146w, 1116m, 1062vs, 990w, 961m, 919w, 889w,
1
776vs, 720s, 681m, 622vs, 589w, and 566m. H NMR (DMSO-
d6): δ 12.61 (s, 1H, NH), 8.59 (m, 1H, py), 8.01 (m, 1H, py),
7.83 (m, 1H, py), 7.25 (m, 1H, py), 7.43–7.27 (m, 5H, Ph), 4.48
(s, 2H, CH2), 2.45 (s, 3H, CH3). 13C NMR (DMSO-d6): 12.9,
37.9, 120.4, 124.6, 127.2, 128.5, 129.3, 136.7, 136.8, 148.8,
152.2, 154.2 and 198.8 ppm.
2-Pyridinecarbaldehyde S-benzyldithiocarbazate (HPCSBC).
S-Benzyldithiocarbazate (1.98 g, 0.01 mol) was dissolved in
35 mL of absolute ethanol and mixed with 2-pyridinecarbalde-
hyde (1.08 g, 0.01 mol). The mixture was refluxed for 15 min
and allowed to stand for 1 h. The yellow powder which formed
was filtered off, washed with ethanol and dried in a vacuum
desiccator. It was recrystallised twice from absolute ethanol to
obtain light yellow crystals. Yield (2.29 g, 79.9%). Microanaly-
sis found C, 58.3; H, 4.6; N, 14.6%; Calculated for C14H13N3S2:
Di-2-pyridyl ketone S-methyldithiocarbazate (HPKSMC). S-
Methyl dithiocarbazate (1.22 g, 0.01 mol) was dissolved in
ethanol (25 mL) and added to a solution of di-2-pyridyl ketone
in ethanol (15 mL). The mixture was refluxed for 1 h and then
left to stand overnight at room temperature. The product was
filtered off and recrystallised from ethanol to give bright yellow
crystals. Yield (2.19 g, 76%). Microanalysis found C, 54.2; H,
4.2; N, 19.8%; Calculated for C13H12N4S2: C, 54.1; H, 4.2; N,
C, 58.5; H, 4.6; N, 14.6%. IR: ν (cm−1); 2923w, 2794m, 1582w,
ˉ
1566w, 1533s, 1464s, 1438w, 1366w, 1317s, 1278vs, 1108m,
1044s, 1000m, 928s, 830w, 773vs, 714m, 680s, 633, and 600.
1H NMR (DMSO-d6): δ 13.49 (s, 1H, NH), 8.60 (m, 1H, py),
8.26 (s, 1H, CH), 7.86 (m, 2H, py), 7.45–7.25 (m, 5H, Ph, 1H,
py), 4.49 (s, 2H, CH2). 13C NMR (DMSO-d6): 37.7, 120.2,
124.9, 127.3, 128.5, 129.3, 163.5, 137, 146.7, 149.8 and
152.2 ppm.
19.4%. IR: ν (cm−1); 3042w, 2981w, 2910w, 1689w, 1564w,
ˉ
1584m, 1481m, 1452s, 1416vs, 1327s, 1281m, 1242s, 1130s,
1060vs, 998m, 962vs, 890m, 802s, 733vs, 692s, 693m, 614m,
1
and 589s. H NMR (CDCl3): δ 15.28 (s, 1H, NH), 8.78 (m, 1H,
py), 8.61 (m, 1H, py), 8.03 (m, 1H, py), 7.82 (m, 2H, py), 7.70
(m, 1H, py), 7.37 (m, 2H, py), 2.65 (s, 3H, SCH3). 13C NMR
(CDCl3): 17.4, 123.9, 124.3, 124.5, 127.2, 137.1, 137.3, 142.1,
148.1, 151.1, 155.4 and 202.3 ppm.
FeIII complexes
[FeIII(APSMC)2](ClO4). HAPSMC (1 g, 4.4 mmol) was dis-
solved in 15 mL of hot ethanol and then triethylamine
(0.617 mL, 4.4 mmol) was added. Solid Fe(ClO4)3·6H2O
(1.025 g, 2.2 mmol) was added directly to the basic ligand sol-
ution with stirring and the mixture was gently refluxed for
30 min. After cooling to room temperature, the resulting solid
was filtered off and washed with ethanol (10 mL), followed by
diethyl ether (10 mL). It was then dried in a vacuum desiccator
and recrystallised from ethanol to give dark brown powder. Yield
(1.09 g, 82%). Microanalysis found C, 36.1; H, 3.3; N, 14.2%;
Calculated for C18H20ClFeO4 N6S4: C, 35.8; H, 3.3; N, 13.9%.
Di-2-pyridyl ketone S-benzyldithiocarbazate (HPKSBC). Di-
2-pyridyl ketone (1.84 g, 0.01 mol) was dissolved in EtOH
(15 mL) and mixed with a hot solution of S-benzyl dithiocarba-
zate (1.98 g, 0.01 mol) in EtOH (30 mL). The mixture was
refluxed for 1 h and then left to stand overnight at room tempera-
ture. The product was filtered off and recrystallised from ethanol
to obtain yellow crystals. Yield (2.94 g, 80.71%). Microanalysis
found: C, 62.6; H, 4.4; N, 15.7%; Calculated for C19H16N4S2:
C, 62.6; H, 4.4; N, 15.3%. IR: ν (cm−1); 3054w, 2083w, 1584m,
ˉ
1480m, 1451s, 1419s, 1327m, 1281m, 1246m, 1121s, 1053vs,
995m, 961m, 845.w, 802s, 737m, 690s, 639m, 615w, and 591s.
1H NMR (CDCl3): δ 15.33 (s, 1H, NH), 8.81(m, 1H, py), 8.61
(m, 1H, py), 8.03 (m, 1H, py), 7.82 (m, 2H, py), 7.72 (m, 2H,
py), 7.45–7.27 (m, 5H, Ph and 1H, py), 4.57 (s, 2H, CH2). 13C
NMR (CDCl3): 39.2, 123.9, 124.4, 124.5, 127.3, 127.4, 128.6,
129.5, 136.2, 137.2, 137.3, 142.2, 148.1, 151.1, 155.2 and
202.5 ppm.
IR: ν (cm−1); 3078w, 2926w, 1700w, 1597m, 1481w, 1423s,
ˉ
1310m, 1148w, 1084vs, 1032vs, 945s, 818m, 774s, 742s, 621s,
and 557w. Electronic spectrum (MeCN): 425 nm (11 300 M−1
cm−1), 343 (23 400).
[FeIII(APSBC)2](ClO4). HAPSBC (0.7 g, 2.3 mmol) was dis-
solved in 30 mL of hot ethanol and Fe(ClO4)3·6H2O (0.54 g,
1.2 mmol) was added directly to the solution. Et3N (0.32 mL,
2.3 mmol) was added and the dark brown mixture was refluxed
for 30 min. The reaction mixture was filtered while hot to obtain
an initial crop of the desired product. The filtrate was allowed to
cool at room temperature, and was then put in the refrigerator
overnight. The brown precipitate was collected by vacuum fil-
tration and washed with ethanol and diethyl ether to give a
second crop. Total yield (0.62 g, 70%). Microanalysis C, 47.6;
H, 3.7; N, 11.1%; Calculated for C30H28ClFe N6O4S4: C, 47.7;
2-Pyridinecarbaldehyde S-methyldithiocarbazate (HPCSMC).
S-Methyldithiocarbazate (6.11 g, 0.05 mol) was suspended in
isopropanol (75 mL) and refluxed until it dissolved. The solution
was allowed to cool to room temperature and the residue was
filtered off, washed with isopropyl alcohol and discarded. The
filtrate was treated with 2-pyridinecarbaldehyde (5.35 g,
0.05 mol) added drop-wise and the mixture was stirred for 24 h
at room temperature. Then the resultant pale yellow precipitate
was filtered, washed with isopropanol then recrystallised from
H, 3.7; N, 11.1%. IR: ν (cm−1); 3093w, 1598m, 1492w, 1427vs,
ˉ
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 6536–6548 | 6539

1330m, 1242w, 1170m, 1077vs, 1027vs, 942s, 812w, 785s,
694s, and 620s. Electronic spectrum (MeCN): 428 nm (8530
M−1 cm−1), 350 (25 600).
Solid Fe(ClO4)3·6H2O (1.08 g, 2.41 mmol) was added to the
basic ligand solution with stirring and the mixture was gently
refluxed for 1 h. On cooling to room temperature the resulting
crude dark brown solid was filtered off and washed with 10 mL
of ethanol followed by 10 mL of diethyl ether. Column chrom-
atography on silica (see previous synthesis) was carried out
using a solvent gradient of CH3OH (in increasing quantities) in
DCM. After elution of a green band of [FeII(PCSMC)2], the
desired mixed-ligand complex [FeIII(PCSMC)(PSMC)] eluted.
This was evaporated to dryness. Yield 50%. The complex was
recrystallised with acetone–diethyl ether. Microanalysis found C,
38.6; H, 4.1; N, 12.1%; calculated for C22H35FeN6O7S4: C,
[FeIII(PCSBC)2](ClO4)·1.5H2O. HPCSBC (0.7 g, 2.4 mmol)
was dissolved in 300 mL of hot ethanol and then triethylamine
(0.34 mL, 2.4 mmol) was added. Solid Fe(ClO4)3·6H2O (0.54 g,
1.2 mmol) was added to the basic ligand solution with stirring
and the mixture was gently refluxed for 1 h. The reaction
mixture was cooled to room temperature and the resulting crude
dark brown solid containing a mixture of compounds was
filtered off, washed with 10 mL of ethanol followed by 10 mL of
diethyl ether. A column of dry silica (100 × 32 mm diameter,
Scharlau or Merck Silica gel 60, 230–400 mesh) was prepared.
The crude product (0.3 g) was dissolved in CH3CN (∼50 mL)
and adsorbed onto Celite (∼1.3 g) by evaporation to dryness on
a rotary evaporator at room temperature (see Safety Note). The
resulting solids were packed dry onto the column which was
tapped gently to remove air pockets. The column was eluted
with a step-wise gradient of CH3CN–DCM starting with 0.1%
CH3CN–DCM with the application of air pressure. A green band
of [FeII(PCSBC)2] began to elute with 5% CH3CN–DCM. The
polarity of the solvent mixture was gradually increased until all
of this compound had eluted. This compound was characterised
separately (see FeII complex section below). The brown second
component ([FeIII(PCSBC)2](ClO4)) began to elute with 25%
CH3CN–DCM. Evaporation to dryness at room temperature gave
the product as a brown solid. Yield (0.4 g, 43%). Microanalysis
found C, 44.76; H, 3.62; N, 10.77%; Calculated for
38.9; H, 5.2; N, 12.4%. IR: ν (cm−1); 3401m, 2921m, 1583vs,
ˉ
1371s, 1283m, 1192w, 1073vs, 995m, 916m, 761, 686w, and
620m. Crystals suitable for crystallography were obtained by
slow diethyl ether vapour diffusion into a concentrated acetone
solution. Electronic spectrum (MeCN): 618 nm (1440 M−1
cm−1), 441 (3850), 300 (11 300).
[FeIII(PKSMC)2](ClO4). HPKSMC (0.8 g, 2.8 mmol) was
dissolved in 30 mL of hot ethanol and (0.64 1.4 mmol) of solid
Fe(ClO4)3·6H2O was added directly to the solution. Et3N
(0.39 mL, 2.8 mmol) was added to the dark brown mixture
which was gently refluxed for 30 min. The reaction mixture was
filtered while still hot to obtain an initial crop. The filtrate was
allowed to cool at room temperature and refrigerated overnight.
The dark brown product was collected by vacuum filtration and
washed with ethanol and diethyl ether to give a second batch of
brown solid. Yield (0.83 g, 83%). Microanalysis found C, 43.0;
H, 3.2; N, 15.4%; Calculated for C26H22ClFe·N8O4S4: C, 42.8;
C28H27ClFeO5.5 N S : C, 44.5; H, 3.6; N, 11.1%. IR: ν (cm−1);
ˉ
6
4
H, 3.0; N, 15.4%. IR: ν (cm−1); 3090w, 1620m, 1584m, 1529w,
3029w, 2922w, 1602m, 1579m, 1477w, 1404s, 1342m, 1219w,
1066vs, 974s, 910m, 883w, 680m, 619s, and 563w. Electronic
spectrum (MeCN): 434 nm (8130 M−1 cm−1), 352 (23 800).
ˉ
1413s, 1340s, 1313m, 1159m, 1073vs, 1000s, 821w, 792m,
743s, and 615vs. Crystals of [FeIII(PKSMC)2]·MPPT·(ClO4)2
(MPPT being a triazinium product of HPKSMC desulfuration,
see Scheme 1) suitable for X-ray work were obtained from
diethyl ether vapour diffusion into a concentrated acetone sol-
ution of the crude first crop. Electronic spectrum (MeCN):
445 nm (12 100 M−1 cm−1), 362 (27 000).
[FeIII(PCSMC)2](ClO4). HPCSMC (1 g, 4.7 mmol) was dis-
solved in 75 mL of hot ethanol and solid Fe(ClO4)3·6H2O
(0.54 g, 2.4 mmol) was added directly to the solution. Triethyl-
amine (0.66 mL, 4.7 mmol) was added to the dark brown
mixture which was gently refluxed for 30 min. The reaction
mixture was filtered while still hot to obtain an initial crop of
dark brown compound. The filtrate was allowed to cool at room
temperature and then refrigerated overnight. A second crop was
obtained by vacuum filtration and was washed with ethanol and
diethyl ether. Attempts to recrystallise the complex were unsuc-
cessful. Column chromatography (as described in the previous
synthesis) enabled separation of [FeII(PCSMC)2] (green, first
band) from the desired product [FeIII(PCSMC)2](ClO4) which
was evaporated to dryness at room temperature (see Safety Note)
to afford a brown solid. Yield (0.68 g, 50%). Microanalysis
found C, 33.75; H, 2.90; N, 14.32%; Calculated for
[FeIII(PKSBC)2](ClO4)·2.5H2O. HPKSBC (0.8 g, 2.2 mmol)
was dissolved in 65 mL of hot ethanol and solid Fe
(ClO4)3·6H2O (0.51 g, 1.1 mmol) was added directly to the sol-
ution. Triethylamine (0.31 mL, 2.2 mmol) was added to the dark
brown mixture which was gently refluxed for 30 min. The reac-
tion mixture was filtered while still hot and the filtrate was
allowed to cool at room temperature and refrigerated overnight.
The dark brown product was collected by vacuum filtration and
washed with diethyl ether. Yield (0.46 g, 48%). Microanalysis
found C, 49.3; H, 3.6; N, 12.3%; Calculated for C38H35ClFe
N8O6.5S : C, 49.2; H, 3.8; N, 12.1%. IR:ν (cm−1); 3059w,
ˉ
4
C H ClFeO N S : C, 33.37; H, 2.80; N, 14.59%. IR: ν
ˉ
16 16
4 6 4
1585m, 1494w, 1408vs, 1335s, 1243w, 1074vs, 989s, 820w,
745m, 700m, and 617vs. Electronic spectrum (MeCN): 431 nm
(9370 M−1 cm−1), 356 (20 100).
(cm−1); 3098w, 1603m, 1477w, 1405vs, 1344m, 1306m, 1216m,
1166w, 1062s, 978s, 914w, 766m, and 619s. Vapour diffusion of
diethyl ether into an acetone solution of the complex afforded
crystals suitable for X-ray work. Electronic spectrum (MeCN):
431 nm (9370 M−1 cm−1), 356 (20 100).
FeII complexes
[FeIII(PCSMC)(PSMC)]·2Me2CO·4H2O. HPCSMC (1 g,
4.7 mmol) was dissolved in 100 mL of hot ethanol and then tri-
ethylamine (0.66 mL, 4.7 mmol) was added to the solution.
The FeII complexes were synthesised by the following general
method. 2.4 mmol of the appropriate dithiocarbazate ligand was
dissolved in ethanol (30–150 mL) and the mixture was purged
6540 | Dalton Trans., 2012, 41, 6536–6548
This journal is © The Royal Society of Chemistry 2012

Scheme 1
with nitrogen for 10 min. Et3N (0.34 mL, 2.4 mmol) was added
3063w, 2119w, 1581m, 1481m, 1379vs, 1322s, 1239w, 1112w,
1061vs, 1015w, 980vs, 858w, 785s, 766m, 740s, 699m, 601s,
and 556w. Electronic spectrum (MeCN): 684 nm (7120 M−1
cm−1), 441 (sh, 8000), 331 (17 800).
followed by solid Fe(ClO4)2·6H2O (0.44 g, 1.2 mmol) with stir-
ring. The mixture was gently refluxed for 30 min under nitrogen.
Upon cooling, the green product was filtered off and washed
with 10 mL of ethanol followed by 10 mL diethyl ether.
[FeII(APSMC)2]·H2O. Yield (0.93 g, 80%). Microanalysis
Physical methods
found C, 41.0; H, 3.9; N, 15.9%; Calculated for C18H22FeN6S4:
C, 41.4; H, 4.2; N, 16.1%. IR: ν (cm−1); 3414b, 2927w, 1629w,
UV-vis spectra were recorded on a Perkin Elmer Lambda 40
spectrophotometer. Solutions were analysed immediately before
any redox reactions occurred. This was particularly important for
the FeII complexes which slowly oxidised in aerated solutions.
IR spectra were recorded as undiluted solids on a Perkin Elmer
1600 FTIR spectrometer in attenuated total reflectance mode.
NMR spectra were acquired on a Bruker Avance 300 MHz spec-
trometer and all resonances were given relative to tetramethylsi-
lane (δ = 0 ppm). Cyclic voltammetry was recorded with a
BAS100B/W potentiostat employing a glassy carbon working
electrode, Pt wire counter and Ag–AgCl reference electrode. The
solvent was MeCN–H2O (7 : 3) and the supporting electrolyte
was 0.1 M Et4NClO4. All solutions were purged with N2 before
measurement. Approximately 1 mM solutions of analyte were
used.
ˉ
1589m, 1467w, 1395s, 1322vs, 1140m, 1093m, 1030vs, 933m,
816w, 738s, 647w, 627m, and 556w. Electronic spectrum
(MeCN): 676 nm (7360 M−1 cm−1), 429 (sh, 15 100), 331
(28 300).
[FeII(APSBC)2]·2H2O. Yield (0.58 g, 72%). Microanalysis
found C, 51.4; H, 4.2; N, 12.4%; Calculated for C30H32Fe
N O S : C, 52.0; H, 4.7; N, 12.1%. IR: ν (cm−1); 3395b,
ˉ
6
2 4
1627w, 1590s, 1400vs, 1323vs, 1286m, 1234w, 1142m, 1092m,
1031vs, 942s, 765s, 694s, 630m, and 567. Electronic spectrum
(MeCN): 679 nm (5920 M−1 cm−1), 429 (sh, 12 500), 333
(26 300).
[FeII(PCSMC)2]·2H2O. Yield (0.95 g, 81%). Microanalysis
found C, 37.9; H, 3.5; N, 16.0%; Calculated for C16H20Fe
N O S : C, 37.5; H, 3.9; N, 16.4%. IR: ν (cm−1); 3374b,
ˉ
6
2 4
2970w, 2111m, 1594s, 1495m, 1463w, 1429vs, 1372vs, 1322m,
1204m, 1105m, 1049vs, 953s, 882m, 844m, 754s, 666w, and
566m. Electronic spectrum (MeCN): 684 nm (7210 M−1 cm−1),
441 (sh, 7210), 346 (17 800).
Biological assays
Cell culture. The chelators were dissolved to make 10 mM
stock solutions in DMSO and diluted in medium containing
10% foetal calf serum (Commonwealth Serum Laboratories,
Melbourne, Australia). A final [DMSO] < 0.5% (v/v) was used
to ensure that DMSO had no effect on proliferation, 59Fe uptake,
or 59Fe mobilisation from cells.46 The human SK-N-MC neuro-
epithelioma cell line (American Type Culture Collection, Mana-
ssas, VA) was grown using standard procedures27 in a humidified
atmosphere of 5% CO2–95% air at 37 °C in an incubator (Forma
Scientific, Marietta, OH).
[FeII(PCSBC)2]·0.5H2O. Yield (0.4 g, 52%). Microanalysis
found C, 52.7; H, 4.0; N, 12.9%; Calculated for C28H25Fe
N O S : C, 52.7; H, 4.0; N, 13.2%. IR: ν (cm−1); 3374b,
ˉ
6
5 4
2970w, 2111m, 1594s, 1495m, 1463w, 1429vs, 1372vs, 1322m,
1204m, 1105m, 1049vs, 953s, 882m, 844m, 754s, 666w, and
566m. Electronic spectrum (MeCN): 664 nm (6320 M−1 cm−1),
429 (sh, 10 500), 331 (28 400).
[FeII(PKSMC)2]·1.5H2O. Yield (0.8 g, 70%). Microanalysis
found C, 47.4; H, 3.8; N, 16.7%; Calculated for
C26H25Fe·N8O1.5S : C, 47.5; H, 3.8; N, 17.0%. IR: ν (cm−1);
3059w, 2083w, 1595s, 1496s, 1394vs, 1333m, 1287m,
1207.61m, 1153w, 1108s, 1040s, 968w, 852m, 759s, 699s, and
561w. Electronic spectrum (MeCN): 682 nm (6320 M−1 cm−1),
445 (sh, 10 500), 344 (25 300).
Effect of the chelators on cellular proliferation
ˉ
4
The effect of the chelators on cellular proliferation was deter-
mined using the [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetra-
zolium] (MTT) assay via standard methods.27,46 The SK-N-MC
cell line was seeded at 1.5 × 104 cells per well in 96-well micro-
titer plates in medium containing the Fe uptake protein, human
59Fe2-transferrin (Fe2-Tf; 1.25 μM) and chelators at a range of
concentrations (0–25 μM). The control samples contained
medium with Fe2-Tf (1.25 μM) in the absence of the ligands.
FeII(PKSBC)2]·0.5H2O. Yield (0.5 g, 64%). Microanalysis
found C, 57.5; H, 3.9; N, 14.4%; Calculated for
C38H31FeN8O0.5S : C, 57.6; H, 4.0; N, 14.2%. IR: ν (cm−1);
ˉ
4
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 6536–6548 | 6541

Table 1 Crystal data
HAPSMC
[Fe(PCSMC)2](ClO4)
[Fe(PKSMC)2](MPPT)(ClO4)2
[Fe(PSMC)(PCSMC)]·Me2CO
Formula
Mol. Wt
Crystal system
a (Å)
b (Å)
c (Å)
C9H11N3S2
225.33
C16H16ClFeN6O4S4
575.89
Monoclinic
33.437(2)
8.8580(4)
14.865(1)
C39H33Cl2FeN12O8S5
1084.82
Triclinic
C19H21FeN6O2S4
549.51
Trigonal
Triclinic
5.8633(5)
8.148(1)
11.757(2)
81.91(1)
79.334(9)
84.497(8)
545.1(1)
0.71073
293
7.9883(3)
14.5465(7)
20.3055(9)
89.484(4)
82.777(4)
84.060(4)
2328.2(2)
1.54184
36.376(5)
9.3420(7)
α (°)
β (°)
90.851(5)
γ (°)
V (Å3)
λ (Å)
4402.4(4)
0.71073
293
10 705(2)
0.71073
293
T (K)
293
Z
2
P1
0.452
8
C2/c
1.224
2
P1
6.302
18
R3
1.014
ˉ
ˉ
ˉ
Space group
μ (mm−1
)
Total refs
3488
1905 (0.0292)
0.0432
14 700
3881 (0.0425)
0.0406
0.0908
21 135
7365 (0.0398)
0.0424
8404
4169 (0.0729)
0.0933
Indep. refs (Rint
R1 (obs. data)
wR2 (all data)
)
0.0826
0.1120
0.2546
The cells were then incubated in a humidified atmosphere con-
taining 5% CO2 and 95% air at 37 °C for 72 h. Subsequently,
10 μL of MTT (5 mg mL−1) was added to each well and incu-
bated at 37 °C for 2 h. The cells were solubilised with 100 μL of
10% SDS–50% isobutanol in 10 mM HCl and the plates were
then read at 570 nm using a scanning multi-well spectropho-
tometer. The inhibitory concentration (IC50) was defined as the
chelator concentration needed to decrease the absorbance to 50%
of the untreated control. Absorbance was shown to be directly
proportional to cell counts, as shown previously.46
established methods.46 Cells were incubated with 59Fe2-Tf
(0.75 μM) for 3 h at 37 °C in the presence of the chelators
(25 μM). The cells were washed on ice four times with ice-cold
PBS and the internalised 59Fe was determined by standard pro-
cedures by incubating the monolayer for 30 min at 4 °C with the
protease, Pronase (1 mg mL−1; Sigma).46,47 The cells were
removed using a plastic spatula and centrifuged at 14 000 rpm
for 1 min. The pronase-insensitive fraction represents interna-
lised 59Fe, while the supernatant represents the membrane-
bound, pronase-sensitive 59Fe that was released by the pro-
tease.46,47 The ligands were compared to the previously charac-
terised chelators, DFO, 311 and HDp44mT.19,46
Preparation of 56Fe- and 59Fe-Tf
Human Tf (Sigma) was labelled with either 56Fe or 59Fe
(Dupont NEN, MA) to produce 56Fe2-Tf and 59Fe2-Tf, respect-
ively, as described previously.47,48 Any unbound 59Fe was
removed by passage through a Sephadex G25 column and
exhaustive vacuum dialysis against an excess of 0.15 mM NaCl
buffered to pH 7.4 with 1.4% NaHCO3 via standard methods.47
Crystallography
Crystallographic data were acquired at 23 °C on an Oxford Dif-
fraction Gemini CCD diffractometer employing either Mo-Kα or
Cu-Kα radiation (see Table 1) and operating in the ω-scan mode.
Data reduction and empirical absorption corrections (multi-scan)
or analytical absorption corrections were performed with Oxford
Diffraction CrysAlisPro software. Structures were solved by
direct methods with SHELXS and refined by full-matrix least-
squares analysis with SHELXL-97.49 All non-H atoms were
refined with anisotropic thermal parameters. Molecular structure
diagrams were produced with ORTEP3.50 All calculations were
carried out within the WinGX graphical user interface.51 In the
structure of [Fe(PCSME)(PSME)]·Me2CO the acetone molecule
was modelled as a diffuse contribution to the overall scattering
without specific atom positions using SQUEEZE/PLATON.52
Crystal and refinement data are summarised in Table 1.
Effect of chelators on 59Fe efflux from SK-N-MC cells
The ability of the chelators to mobilise 59Fe from SK-N-MC
cells was performed by established techniques.46 After prelabel-
ling cells with 59Fe2-Tf (0.75 μM) for 3 h at 37 °C, the cell
monolayer was washed four times on ice with ice-cold PBS and
subsequently incubated with the chelators (25 μM) for 3 h at
37 °C. The overlying media containing released 59Fe was separ-
ated from the cells using a Pasteur pipette. Radioactivity in the
cell pellet and the supernatant was measured using a γ-scintil-
lation counter (Wallac Wizard 3, Turku, Finland). In these
studies, the novel ligands were compared to the well-character-
ised chelators, DFO, H2NIH (311) and HDp44mT.19,46,47
Results and discussion
Free ligand characterisation
Effect of chelators at preventing 59Fe uptake from 59Fe2-Tf by
SK-N-MC cells
The dithiocarbazate Schiff base ligands in Chart 2 were prepared
by variations on published procedures31,33 involving conden-
sation between S-methyl (or S-benzyl) thiocarbazate and the cor-
responding aldehyde or ketone. These Schiff base ligands
The ability of the novel ligands to prevent cellular 59Fe uptake
from the Fe transport protein, 59Fe2-Tf, was examined via
6542 | Dalton Trans., 2012, 41, 6536–6548
This journal is © The Royal Society of Chemistry 2012

Fig. 1 ORTEP view of the free ligand HAPSMC (30% probability
ellipsoids).
present a homologous series bearing a common NNS donor set.
As a representative example, the single crystal structure of
HAPSMC was determined. A view of the ligand appears in
Fig. 1. The compound adopts an approximately planar confor-
mation. Of note is that the donor atoms (N1, N2 and S1) are not
in the required conformation for tridentate coordination and 180°
rotation of the C8–N3 and C5–C6 bonds are required. The C8–
S1 bond has double bond character in contrast to the S-methyl
single bond (S2–C9). The C8–S2 bond is slightly shorter than
S2–C9 indicating some electronic delocalisation. Selected bond
lengths appear in Table 2.
Fig. 2 UV-vis spectra of HAPSBC and its FeII and FeIII complexes
measured in MeCN.
trivalent state during synthesis. Each ferrous complex exhibited a
characteristically green colour due to a broad and intense MLCT
(FeII → pyridine) transition around 670 nm (Fig. 2). Similar
maxima have been identified in the FeII complexes of the
HPCIH16–18 and HDp44mT28–30 analogues. In the solid state,
the FeII complexes are stable and do not undergo oxidation. The
FeIII complexes were prepared by direct reaction with ferric per-
chlorate in EtOH. The ferric complexes are dark brown and
exhibit a series of broad, overlapping and intense electronic tran-
sitions that span most of the visible region (Fig. 2). The free
ligands exhibit no electronic maxima in the visible region and
are pale yellow in colour.
Purification of the Fe complexes was generally possible by
recrystallisation, but column chromatography (silica) was some-
times necessary to separate mixtures of FeII and FeIII complexes
that co-precipitated. This is a new method for purification of
these complexes and facilitated isolation of compounds for
single crystal X-ray work.
Fe complexation
Deprotonation of the thioamide N atom accompanies tridentate
coordination of each dithiocarbazate Schiff base ligand. The Fe
complexes of these ligands were the focus of our investigation
here. Given the parallels with our previous work on the HPCIH
(hydrazone)16 and HDp44mT (thiosemicarbazone) series,28 the
present set of dithiocarbazate compounds exhibit chemical and
biological properties reminiscent of both. The FeII complexes
were prepared by reaction of the ligands with ferrous perchlorate
in EtOH under an inert atmosphere to avoid oxidation to the
Table 2 Selected bond lengths and angles (ligands ‘a’ and ‘b’ defined by labels in Fig. 3 and 4). For the heteroligand complex [Fe(PSMC)
(PCSMC)] bonds to the oxidised PSMC2− ligand are in italics
[Fe(PCSMC)2]+,
Ligand a, b
[Fe(PKSMC)2]+,
Ligand a, b
[Fe(PSMC)(PCSMC)],
PSMC, PCSMC
HAPSMC
C8–S1
C8–S2
C8–N3
C9–S2
N2–N3
N2–C6
Fe–N1a/b
Fe–N2a/b
Fe–S1a/b
N1a/b–Fe–N2a/b
N2a/b–Fe–S1a/b
N2a–Fe–N2b
1.645(3)
1.762(3)
1.333(3)
1.791(3)
1.374(3)
1.282(3)
—
—
—
—
—
1.720(3), 1.713(3)
1.710(3),1.713(3)
1.282(4), 1.288(4)
1.764(3), 1.767(4)
1.368(3), 1.371(3)
1.277(3), 1.277(3)
1.968(2), 1.956(3)
1.880(2), 1.879(2)
2.1891(9), 2.201(1)
80.9(1), 80.5(1)
84.97(7), 85.03(8)
176.5(1)
1.735(3), 1.735(3)
1.738(3), 1.733(3)
1.291(4), 1.301(4)
1.794(4), 1.796(4)
1.384(3), 1.385(3)
1.308(4), 1.309(4)
1.984(3), 1.984(2)
1.910(2), 1.912(2)
2.2255(9), 2.2273(9)
80.4(1), 80.4(1)
1.73(1), 1.751(9)
1.765(9), 1.740(9)
1.29(1), 1.26(1)
1.75(1), 1.78(1)
1.365(9), 1.369(9)
1.38(1), 1.27(1)
1.992(6), 1.990(7)
1.911(7), 1.942(7)
2.223(2), 2.265(3)
81.6(3), 79.7(3)
84.8(2), 82.8(2)
176.5(3)
85.37(8), 85.56(8)
177.8(1)
—
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 6536–6548 | 6543

complex) are in the region of the coordinated S-donors. The C8–
S1 and C8–S2 bond lengths in [Fe(PCSMC)2]+ are identical in
the complex due to electron delocalisation across the N3–C8
(S1)–S2 moiety in contrast to what is seen in the structure of the
analogue HAPSMC where the C8vS1 bond has definite double
bond character. The C3–N3 bonds in each ligand shorten
appreciably upon complexation, defining an effective imine–
thiolate (–NvC–S−) resonance form of the coordinated ligand
anion. No significant variations in the N2–N3 or N2–C6 bonds
are apparent on deprotonation and complexation.
The crystal structure of the homologous complex [Fe
(PKSMC)2]+ was also determined (Fig. 4). The geometry of the
complex cation mirrors that of [Fe(PCSMC)2]+. Again the co-
ordinate bonds and angles are consistent with low spin FeIII com-
plexes of the dithiocarbazate Schiff base ligands (Table 2) but
the Fe–N and Fe–S bonds are slightly longer than seen in [Fe
(PCSMC)2]+. This may be attributed to a combination of interli-
gand repulsion brought about by introduction of the non-coordi-
nating pyridyl rings and also electron-withdrawing effects of
these groups that may slightly weaken the donor strength of the
PKSMC− ligand.
Fig. 3 ORTEP view of the [Fe(PCSMC)2]+ cation (30% probability
ellipsoids).
Curiously, the complex [Fe(PKSMC)2]+ co-crystallised with a
cationic pyridotriazinium derivative (MPPT+, Fig. 4); a by-
product of ligand desulfuration. It is not known whether Fe cata-
lyses this reaction, although the loss of sulfide (or hydrogen
sulfide) is probably assisted by the presence of free metal ions. A
tentative mechanism is proposed in Scheme 1. It is known that
elimination of methanethiol followed by ring closure is an
alternative possibility in dithiocarbazates of this type, leading to
an exocyclic thioamide (rather than thioether) group.53
Depending on the order of addition of reagents (ligand, ferric
perchlorate and base), different products were observed in the
synthesis of the FeIII complex of HPCSMC. When the ferric salt
was added last, an unusual mixed-ligand FeIII complex was
obtained following column chromatography, namely [Fe
(PCSMC)(PSMC)]. The ligand dianion PSMC2− (Chart 2) is
Fig. 4 ORTEP view of the [Fe(PKSMC)2]+ (left) and MPPT+ (right)
cations (30% probability ellipsoids).
The crystal structure of [Fe(PCSMC)2](ClO4) was determined.
A view of the complex cation appears in Fig. 3. The meridional
configuration of each ligand is apparent to produce a complex
with approximate C2 symmetry. The Fe–N and Fe–S bond
lengths (Table 2) are characteristic of low spin FeIII complexes
seen in related thiosemicarbazone and thiohydrazone ferric com-
plexes from the HDp44mT30 and HPKTBH27 series. There is no
conformational flexibility in the complex cation except for the S-
methyl groups which each adopt an anti-conformation with
respect to the adjacent coordinated S-donor. The most significant
variations in intraligand bond lengths (going from free ligand to
Fig. 5 ORTEP view of [Fe(PCSMC)(PSMC)] (30% probability
ellipsoids).
6544 | Dalton Trans., 2012, 41, 6536–6548
This journal is © The Royal Society of Chemistry 2012

effectively a hydroxylated derivative of PCSMC−. The FeIII
complex, [Fe(PCSMC)(PSMC)] exhibited some characteristic
spectroscopic features, including an intense CvO vibrational
band in the IR spectrum (1583 cm−1), which was absent in all
other IR spectra reported here.
The structure of the complex was confirmed by X-ray crystal-
lography. A view of the neutral complex [Fe(PCSMC)(PSMC)]
is shown in Fig. 5. The presence of O1a on the PSMC2− ligand
is apparent. This has some important structural influences on the
complex and particularly on the PCSMC− co-ligand. The Fe–
N2b and Fe–S1b bonds lengthen considerably in comparison
with the symmetrical [Fe(PCSMC)2]+ and [Fe(PKSMC)2]+ com-
plexes. This may be partially attributed to a trans influence of
the dianionic PSMC2− ligand. The Fe–N2b bond (from
PCSMC−) is trans to the Fe–N2a bond (from PSMC2−) where
N2a formally carries a negative charge due to amide deprotona-
tion. The extension of the Fe–S1b bond may be a consequence
of Fe–N1b elongation as both donor atoms are part of the same
rigid chelate ring. Another structural feature is the change in
bond order of C6–N2 from a formal (imine) double bond in the
dithiocarbazates (Table 2), to an intermediate bond order closer
to one (C6a–N2a 1.38(1) Å) characteristic of a deprotonated
amide in PSMC2−
.
Fig. 6 Cyclic voltammetry of the Fe dithiocarbazate complexes in this
work: sweep rate 100 mV s−1, MeCN–H2O 7 : 3 solvent and 0.1 M
Et4NClO4 supporting electrolyte.
Ligand hydroxylation of this kind is unusual but not without
precedent. Our previous work has uncovered other Schiff base
ligands derived from 2-pyridinecarbaldehyde which undergo
FeIII catalysed oxidation reactions34,54 leading to hydrazine (as
opposed to hydrazone) analogues (the H2IPH analogues,
Chart 1). The mechanism of this reaction is very complicated,39
but nucleophilic attack by water on the C6vN2 double bond fol-
lowed by ligand dehydrogenation are involved. The oxidant is
FeIII and not dioxygen. The properties of Fe complexes from the
H2IPH series are very different, and in fact, only FeIII complexes
have been isolated in contrast to the FeII complexes from the cor-
responding HPCIH series.
complex must be able to be reduced by intracellular reductants
(eqn (2)) so the rate of both eqn (1) and (2) will be important in
determining the overall yield of ˙OH. The higher potential FeII
complexes will be less reactive toward H2O2 but more easily
reduced from their reactive FeIII form.
Cyclic voltammetry of the Fe complexes of the dithiocarba-
zate ligands was undertaken and the results are summarised in
Fig. 6 and Table 3. In all cases the FeII and FeIII dithiocarbazate
complexes gave identical and totally reversible cyclic voltammo-
grams confirming their common structural features (identical first
coordination spheres). Only the voltammograms of the FeIII
complexes are shown in Fig. 6. Any differences in redox poten-
tials are attributed to changes in inductive effects of the substitu-
ents. It is immediately apparent that altering the terminal
S-substitutent (from methyl to benzyl) has little effect (ca.
Redox properties
The FeIII/II redox potentials are an important determinant of bio-
logical activity and affinity for Fe in each of its oxidation states.
Complexes with exceptionally high redox potentials will be most
stable in the FeII form under biological conditions, while those
with low (large negative) redox potentials will be spontaneously
oxidised to FeIII. Between these extremes, both FeIII and FeII
may co-exist and interconvert. This is germane in a cellular
environment due to the ability of redox-active Fe complexes to
catalyse the production of hydroxyl radicals (Fenton chemistry,
eqn (1)),55 which leads to oxidative stress and damage to impor-
tant biomolecules.
Table 3 FeIII/II redox potentials and biological data for the chelators in
this work
Eo(FeIII/II
(mV vs.
NHE)
)
59Fe efflux
IC50
(μM)
(% total cell 59Fe uptake
iron released) (% control)
Control
DFO
H2NIH
(311)
—
—
4
10
40
0.5
2
3
100
86
7
4
½FeIIL2ꢀ þ H2O2 ! ½FeIIIL2ꢀþ þ ˙OH þ OHꢁ
½FeIIIL2ꢀþ þ red ! ½FeIIL2ꢀ þ ox
Previous investigations on Fe complexes of thiosemicarbazone
complexes28–30 have shown that compounds possessing FeIII/II
redox potentials in the range 0 to +200 mV vs. NHE are also
able to catalyse the production of hydroxyl radicals in the pres-
ence of hydrogen peroxide. For eqn (1) to be catalytic, the FeIII
ð1Þ
ð2Þ
−475 (ref. 56) 11.4 0.8
0.75 0.26
7
1
—
HDp44mT +166 (ref. 28) 0.009 0.004 40
2
1
4.9 0.8
HAPSMC +229
HAPSBC +239
HPKSMC +383
HPKSBC +401
HPCSMC +357
HPCSBC +367
0.05 0.02
0.05 0.02
0.37 0.11
0.09 0.06
>10
32
18
46
21
43 11
17
42
1
7
1
22.4 0.7
36
28
38.4 0.9
27
2
1
2
7
>10
3
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 6536–6548 | 6545

as that of the thiocarbazate complex analogues reported here.
Replacing the thiohydrazone S-donor (e.g. [Fe(PKTBH)2]+) with
an O-donor ([Fe(PKBH)2]+, a hydrazone) shifts the FeIII/II
couple to a very high potential (ca. +500 mV) and in fact the
highly oxidising FeIII complexes of these and related hydrazone
ligands are not stable in aqueous solution.16–18
The redox potential of the unusual mixed-ligand complex [Fe
(PSMC)(PCSMC)] was much lower than the symmetric bis-
dithiocarbazate complexes. The voltammetry at three different
sweep rates is shown in Fig. 7. Again a totally reversible FeIII/II
couple is seen, but this time at a potential of −100 mV vs. NHE;
approximately 450 mV lower than the related [Fe(PCSMC)2]+
complex. This large stabilisation of the trivalent oxidation state
is a consequence of the PSMC2− ligand which bears two for-
mally anionic donor atoms; the deprotonated amide N2a (Fig. 5)
and the thiolate S-donor. This cathodic shift mirrors the behav-
iour seen in the related hydrazone (HPCIH)16 to hydrazine
(H2IPH) series,57 where the FeIII/II redox potential of
[FeIII(IPH)2]− is ca. 750 mV lower than [FeII(PCIH)2]. In the
case of [FeIII(PSMC)(PCSMC)], only one of the dithiocarbazate
Schiff base ligands has been altered, so the effect on the redox
potential is not as pronounced.
Fig. 7 Cyclic voltammetry of [Fe(PSMC)(PCSMC)] at 20, 50 and
100 mV s−1 sweep rate: MeCN–H2O (7 : 3) with 0.1 M Et4NClO4 sup-
porting electrolyte.
10 mV) on the FeIII/II redox potential. This is to be expected
given the fact that substitution of a phenyl ring with a H-atom
occurs at a position remote from the metal. The substituent on
atom C6 (as defined in Fig. 1) has a stronger influence on the
redox potential, as found in our investigations of the HPCIH17
and HDp44mT30 families and their Fe complexes. The electron-
withdrawing non-coordinating pyridyl ring (in PKSMC− and
PKSBC−) yields Fe complexes with the highest redox potentials
of the series. The electron-donating methyl group gave lower
potential Fe complexes of APSMC− and APSBC−. These poten-
tials were born out in the properties of the FeIII and FeII com-
plexes; [FeII(APSMC)2] and [Fe(APSBC)2] being air-sensitive in
solution upon standing.
Fe mobilisation studies and cytotoxicity
The ability of the dithiocarbazates in Chart 2 to bind intracellular
59Fe was examined in two separate, but complementary assays
using 59Fe2-Tf which donates 59Fe to cells.46,47 The ability of
the chelators to mobilise cellular 59Fe (Fig. 8A) from human
SK-N-MC neuroepithelioma cells pre-labelled with 59Fe2-Tf was
expressed as the percentage of 59Fe released in the presence of
each ligand. During this assay, the chelators act as “shuttles” to
mobilise 59Fe out of the cell. As expected, the control (medium
alone with no chelator) resulted in only a small amount (∼4%)
of 59Fe being released from the cell over the time course of the
experiment. Positive controls were the known Fe chelators, DFO,
HDp44mT and H2NIH (311; Chart 1).19,46,47 All dithiocarbazate
ligands were more effective than DFO at releasing intracellular
Fe and HPCSMC was the most active of all compounds being
almost as effective as HDp44mT (Fig. 8A). Furthermore, the
three S-methyl dithiocarbazates (HAPSMC, HPCSMC and
The thiosemicarbazone complex [Fe(Dp44mT)2]+ (see Chart 1
for ligand structure) exhibits a redox potential (+166 mV)28
more than 200 mV below the analogous thiocarbazate complexes
[Fe(PKSMC)2]+ and [Fe(PKSBC)2]+; all three compounds being
of Schiff bases derived from di-2-pyridyl ketone. The thiohydra-
zone analogue [Fe(PKTBH)2]+ (see Chart 1) has a FeIII/II redox
potential of +383 mV vs. NHE27 which is essentially the same
Fig. 8 The effect of the dithiocarbazate chelators on (A) cellular 59Fe release from pre-labelled SK-N-MC neuroepithelioma cells and (B) 59Fe
uptake from 59Fe2-Tf by SK-N-MC neuroepithelioma cells. The control is treated with medium only. Results are mean SD (3 experiments). See
Charts 1 and 2 for ligand abbreviations.
6546 | Dalton Trans., 2012, 41, 6536–6548
This journal is © The Royal Society of Chemistry 2012

HPKSMC) were more effective at mobilising intracellular 59Fe
than their corresponding S-benzyl analogues (HAPSBC,
HPCSBC and HPKSBC, respectively).
Health and Medical Research Council of Australia for a Senior
Principal Research Fellowship and Project Grants. D.S.K.
acknowledges the Cancer Institute New South Wales for Early
Career Development Fellowship and a Research Innovation
Grant. M.T.B. is grateful for scholarship support from the King
Abdul Aziz University (Jeddah, Saudi Arabia).
During the second assay (Fig. 8B), the ability of the chelators
to inhibit 59Fe uptake from 59Fe2-Tf was assessed. All results are
relative to the untreated control (i.e. 59Fe in the presence of incu-
bation medium alone). The results mirror those obtained with the
59Fe efflux assay. The three S-methyl dithiocarbazates are con-
sistently more effective at preventing 59Fe uptake from 59Fe2-Tf
than the three S-benzyl chelators. Again, the novel chelators pre-
pared in this investigation were all more active than DFO, but
less effective than H2NIH (311) or HDp44mT (Fig. 8B).
References
1 P. V. Bernhardt, Dalton Trans., 2007, 3214–3220.
2 P. C. Sharpe, D. R. Richardson, D. S. Kalinowski and P. V. Bernhardt,
Curr. Top. Med. Chem., 2011, 11, 591–607.
Potential chelators for the treatment of Fe overload disorders1,2
must have low toxicity as the condition will in many cases
require life-long administration of the drug.1,2 Thus, the toxicity
of the chelators was examined against the same SK-N-MC neu-
roepithelioma cell line using DFO as a negative control and the
highly potent ligand, HDp44mT, as a positive control.19,29,46
The results are summarised in Table 3, where it can be seen that
the two dithiocarbazates derived from 2-pyridinecarbaldehyde
exhibit no apparent cytotoxicity (>10 μM). This property, in
combination with their high activity in sequestering intracellular
Fe, present ideal properties for a chelator in the treatment of Fe
overload. The four dithiocarbazates derived from 2-acetylpyri-
dine and di-2-pyridyl ketone show moderate to potent anti-prolif-
erative activity, and as such, may be problematic as useful drugs
for the treatment of Fe overload. However, anti-proliferative
activity can be advantageous in the treatment of cancer if a large
therapeutic index (anti-proliferative activity against cancer cells
relative to normal cells) can be defined, and the thiosemicarba-
zone relatives are an example of this.21,22 This aspect is the
subject of a separate investigation.
3 D. M. Frazer, S. J. Wilkins, E. M. Becker, C. D. Vulpe, A. T. McKie,
D. Trinder and G. J. Anderson, Gastroenterology, 2002, 123, 835–844.
4 D. S. Kalinowski and D. R. Richardson, Pharmacol. Rev., 2005, 57,
547–583.
5 S. Steinhauser, U. Heinz, M. Bartholomae, T. Weyhermueller, H. Nick
and K. Hegetschweiler, Eur. J. Inorg. Chem., 2004, 4177–4192.
6 R. Galanello, Ann. N. Y. Acad. Sci., 2005, 1054, 183–195.
7 A. Piga, R. Galanello, G. L. Forni, M. D. Cappellini, R. Origa, A. Zappu,
G. Donato, E. Bordone, A. Lavagetto, L. Zanaboni, R. Sechaud,
N. Hewson, J. M. Ford, H. Opitz and D. Alberti, Haematologica, 2006,
91, 873–880.
8 J. B. Porter, Drugs Today, 2006, 42, 623–637.
9 G. J. Kontoghiorghes, M. A. Aldouri, L. N. Sheppard and A.
V. Hoffbrand, Lancet, 1987, 329, 1294–1295.
10 N. F. Olivieri, G. M. Brittenham, C. E. McLaren, D. M. Templeton, R.
G. Cameron, R. A. McClelland, A. D. Burt and K. A. Fleming, N.
Engl. J. Med., 1998, 339, 417–423.
11 J. Savulescu, Br. Med. J., 2004, 328, 358–359.
12 F. H. Mourad, A. V. Hoffbrand, M. Sheikh-Taha, S. Koussa, A.
I. Khoriaty and A. Taher, Br. J. Haematol., 2003, 121, 187–189.
13 E. Becker and D. R. Richardson, J. Lab. Clin. Med., 1999, 134, 510–521.
14 C. S. M. Wong, J. C. Kwok and D. R. Richardson, Biochim. Biophys.
Acta, Mol. Basis Dis., 2004, 1739, 70–80.
15 D. B. Lovejoy, D. S. Kalinowski, P. V. Bernhardt and D. R. Richardson,
Hemoglobin, 2006, 30, 93–104.
16 P. V. Bernhardt, P. Chin, P. C. Sharpe and D. R. Richardson, Dalton
Trans., 2007, 3232–3244.
17 P. V. Bernhardt, G. J. Wilson, P. C. Sharpe, D. S. Kalinowski and D.
R. Richardson, J. Biol. Inorg. Chem., 2008, 13, 107–119.
18 P. V. Bernhardt, L. M. Caldwell, T. B. Chaston, P. Chin and D.
R. Richardson, J. Biol. Inorg. Chem., 2003, 8, 866–880.
19 J. Yuan, D. B. Lovejoy and D. R. Richardson, Blood, 2004, 104, 1450–
1458.
20 M. Whitnall, J. Howard, P. Ponka and D. R. Richardson, Proc. Natl.
Acad. Sci. U. S. A., 2006, 103, 14901–14906.
21 Y. Yu, D. S. Kalinowski, Z. Kovacevic, A. R. Siafakas, P. J. Jansson,
C. Stefani, D. B. Lovejoy, P. C. Sharpe, P. V. Bernhardt and D.
R. Richardson, J. Med. Chem., 2009, 52, 5271–5294.
22 Z. Kovacevic, D. S. Kalinowski, D. B. Lovejoy, Y. Yu, Y. S. Rahmanto,
P. C. Sharpe, P. V. Bernhardt and D. R. Richardson, Curr. Top. Med.
Chem., 2011, 11, 483–499.
23 J. G. Cory, A. H. Cory, G. Rappa, A. Lorico, M.-C. Liu, T.-S. Lin and A.
C. Sartorelli, Biochem. Pharmacol., 1994, 48, 335–344.
24 M.-C. Liu, T.-S. Lin, P. Penketh and A. C. Sartorelli, J. Med. Chem.,
1995, 38, 4234–4243.
25 M.-C. Liu, T.-S. Lin and A. C. Sartorelli, Prog. Med. Chem., 1995, 32,
1–35.
26 R. A. Finch, M. Liu, S. P. Grill, W. C. Rose, R. Loomis, K. M. Vasquez,
Y. Cheng and A. C. Sartorelli, Biochem. Pharmacol., 2000, 59, 983–991.
27 D. S. Kalinowski, P. C. Sharpe, P. V. Bernhardt and D. R. Richardson, J.
Med. Chem., 2007, 50, 6212–6225.
Conclusions
This study has focused on the Fe coordination chemistry of a
series of dithiocarbazate Schiff base ligands. Both FeIII and FeII
complexes were isolated and characterised. Cyclic voltammetry
showed that all undergo totally reversible redox reactions at
potentials dependent on substituent effects inherent to their
parent carbonyl compound (2-acetylpyridine, di-2-pyridyl ketone
or 2-pyridinecarbaldehyde). All ligands were more effective in
chelating intracellular Fe relative to the clinically proven chelator
DFO, and the compounds derived from 2-pyridinecarbaldehyde
exhibited no marked anti-proliferative activity. The results
reported here mirror the trends observed with the HPCIH,
HPKIH and HAPIH hydrazones.17 In those cases, Schiff bases
derived from aldehydes (bearing a H-atom at C6) exhibited no
pronounced cytotoxicity in contrast to ketone-derived Schiff
bases bearing a methyl, phenyl or pyridyl group at C6 which all
exhibited significant anti-proliferative efficacy against cancer
cells. Exactly the same trend is seen here, suggesting that the 2-
pyridinecarbaldehyde group is an ideal pharmacophore for an Fe
chelator with low toxicity as seen in the HPCIH analogues.16
28 D. R. Richardson, P. C. Sharpe, D. B. Lovejoy, D. Senaratne, D.
S. Kalinowski, M. Islam and P. V. Bernhardt, J. Med. Chem., 2006, 49,
6510–6521.
29 D. S. Kalinowski, Y. Yu, P. C. Sharpe, M. Islam, Y.-T. Liao, D.
B. Lovejoy, N. Kumar, P. V. Bernhardt and D. R. Richardson, J. Med.
Chem., 2007, 50, 3716–3729.
Acknowledgements
30 D. R. Richardson, D. S. Kalinowski, V. Richardson, P. C. Sharpe, D.
B. Lovejoy, M. Islam and P. V. Bernhardt, J. Med. Chem., 2009, 52,
1459–1470.
P.V.B. acknowledges financial support from the Australian
Research Council (DP1096029). D.R.R. thanks the National
This journal is © The Royal Society of Chemistry 2012
Dalton Trans., 2012, 41, 6536–6548 | 6547

31 M. Akbar Ali, A. H. Mirza, M. Nazimuddin, H. Rahman and R.
J. Butcher, Polyhedron, 2001, 20, 2431–2437.
32 M. R. Maurya, S. Khurana, W. Zhang and D. Rehder, Eur. J. Inorg.
Chem., 2002, 1749–1760.
33 M. Akbar Ali, A. H. Mirza, M. Nazimuddin, R. Ahmed, L. R. Gahan and
P. V. Bernhardt, Polyhedron, 2003, 22, 1471–1479.
34 A. M. Akbar, A. H. Mirza, T. B. S. A. Ravoof and P. V. Bernhardt, Poly-
hedron, 2004, 23, 2031–2036.
35 A. M. Akbar, A. H. Mirza, F. H. Bujang, M. H. S. A. Hamid and P.
V. Bernhardt, Polyhedron, 2006, 25, 3245–3252.
36 A. B. Beshir, S. K. Guchhait, J. A. Gascon and G. Fenteany, Bioorg.
Med. Chem. Lett., 2008, 18, 498–504.
37 T. B. S. A. Ravoof, K. A. Crouse, M. I. M. Tahir, F. N. F. How, R. Rosli
and D. J. Watkins, Transition Met. Chem., 2010, 35, 871–876.
38 S. Shrivastava, N. Fahmi, D. Singh and R. V. Singh, J. Coord. Chem.,
2010, 63, 1807–1819.
39 A. M. Akbar, A. H. Mirza, R. J. Butcher, P. V. Bernhardt and M.
R. Karim, Polyhedron, 2010, 30, 1478–1486.
43 R. Raina and T. S. Srivastava, Inorg. Chim. Acta, 1984, 91, 137–140.
44 W.-X. Hu, W. Zhou, C.-N. Xia and X. Wen, Bioorg. Med. Chem. Lett.,
2006, 16, 2213–2218.
45 P. Bera, C.-H. Kim and S. I. Seok, Polyhedron, 2008, 27, 3433–
3438.
46 D. R. Richardson, E. H. Tran and P. Ponka, Blood, 1995, 86, 4295–
4306.
47 D. R. Richardson and E. Baker, Biochim. Biophys. Acta, Mol. Cell Res.,
1990, 1053, 1–12.
48 D. R. Richardson and K. Milnes, Blood, 1997, 89, 3025–3038.
49 G. M. Sheldrick, Acta Crystallogr., Sect. A, 2008, 64, 112–122.
50 L. J. Farrugia, J. Appl. Crystallogr., 1997, 30, 565.
51 L. J. Farrugia, J. Appl. Crystallogr., 1999, 32, 837–838.
52 A. L. Spek, Acta Crystallogr., Sect. D, 1990, 65, 148–155.
53 J. P. Scovill and J. V. Silverton, J. Org. Chem., 1980, 45, 4372–4376.
54 D. Richardson, P. V. Bernhardt and E. M. Becker, 2-Pyridinecarboxalde-
hyde Isonicotinoylhydrazone Iron Chelators and Uses thereof, University
of Queensland, Heart Research Institute Ltd., Australia, WO, 2001.
55 H. B. Dunford, Coord. Chem. Rev., 2002, 233–234, 311–318.
56 I. Spasojevic, S. K. Armstrong, T. J. Brickman and A. L. Crumbliss,
Inorg. Chem., 1999, 38, 449–454.
40 B. K. Koo, Bull. Korean Chem. Soc., 2011, 32, 1729–1732.
41 A. Molter, J. Rust, C. W. Lehmann, G. Deepa, P. Chiba and F. Mohr,
Dalton Trans., 2011, 40, 9810–9820.
42 M. Akbar Ali, S. E. Livingstone and D. J. Phillips, Inorg. Chim. Acta,
1971, 5, 493–498.
57 P. V. Bernhardt, P. Chin, P. C. Sharpe, J.-Y. C. Wang and D.
R. Richardson, J. Biol. Inorg. Chem., 2005, 10, 761–777.
6548 | Dalton Trans., 2012, 41, 6536–6548
This journal is © The Royal Society of Chemistry 2012