Gold amides as anticancer drugs: Synthesis and activity studies

Article abstract of DOI:10.1039/c3ob27460h

Modular gold amide chemotherapeutics: Access to modern chemotherapeutics with robust and flexible synthetic routes that are amenable to extensive customisation is a key requirement in drug synthesis and discovery. A class of chiral gold amide complexes featuring amino acid derived ligands is reported herein. They all exhibit in vitro cytotoxicity against two slow growing breast cancer cell lines with limited toxicity towards normal epithelial cells. The Royal Society of Chemistry 2013.

Full text of DOI:10.1039/c3ob27460h

Organic &
Biomolecular Chemistry
View Article Online
View Journal
PAPER
Gold amides as anticancer drugs: synthesis and activity
studies†
Cite this: DOI: 10.1039/c3ob27460h
Sonya Newcombe,‡b Mariusz Bobin,‡a,b Amruta Shrikhande,‡b Chris Gallop,a
Yannick Pace,a,b Helen Yong,b Rebecca Gates,b Shuvashri Chaudhuri,a,b Mark Roe,a
Eva Hoffmann*b and Eddy M. E. Viseux*a
Modular gold amide chemotherapeutics: Access to modern chemotherapeutics with robust and flexible
synthetic routes that are amenable to extensive customisation is a key requirement in drug synthesis and
discovery. A class of chiral gold amide complexes featuring amino acid derived ligands is reported herein.
They all exhibit in vitro cytotoxicity against two slow growing breast cancer cell lines with limited toxicity
towards normal epithelial cells.
Received 19th December 2012,
Accepted 26th March 2013
DOI: 10.1039/c3ob27460h
www.rsc.org/obc
Introduction
Metal complexes have been used for medical purposes for
almost 5000 years.1,2 In the past 20 years, gold(I) and gold(III)
complexes have been investigated as potential anticancer drug
candidates.3 Two compounds – auranofin and sodium auro-
thiomalate – (Fig. 1) are currently undergoing phase II and
phase I clinical trials for treatment of lymphomas and non-
small cell lung cancer, respectively. Aurothiomalate reportedly
interacts with PKCι whereas auranofin inhibits thioredoxin
reductase (TrxR) via a ligand exchange mechanism with the
selenocysteine residue within the active site of the protein.
TrxR plays a major role in the regulation of the cellular redox
state and is over-expressed in some tumours.4,5
The structures of gold(I) complexes L–Au–X are linear in
nature and feature two ligands, L and X. The L-ligands can be
phosphines, N-heterocyclic carbenes (NHCs) or sulfides,
whereas the X-ligands range from thiolates, featured notably in
the anti-rheumatoid auranofin, bis-triflic amide, chlorides,
alkoxides, sulfonates and NHCs.5–13 Previously the structural
nature of ligands L and X limited their potential for derivatiza-
tion and the corresponding complexes exhibited significant
cytotoxicity. We have developed two complementary
approaches that combine biocompatible ligands with tunable
lipophilic groups to provide a platform for anti-cancer drug
Fig. 1 Current and novel approaches to chemotherapeutic gold(I) complexes.
(a) Sodium aurothiomalate and auranofin are currently in clinical trials for treat-
ment of lymphomas and lung cancer. (b) The development of gold amide drugs
allows easy access to tunable gold complexes featuring biocompatible ligands.
discovery. In the first approach, we take advantage of the
varying physical properties of amino acids to derivatize bio-
compatible ligands bound to gold monomers or dimers
(Fig. 1). In the second approach, we used a polarized gold(I)
complex which features a zwitterionic ligand containing a
delocalized pyridinium cation (Fig. 1). We hypothesized that a
polarized complex would have enhanced efficacy at killing
adenocarcinoma-derived cells, as these are generally hyperpol-
arized as discussed in the Results section.
We have demonstrated that these complexes bearing the
common triflic amide motif, yet featuring a variety of struc-
tural backbones, can act as potent inhibitors of two breast
cancer cell lines in vitro. This validated approach contrasts
with previous use of delocalized lipophilic cations (DLCs),10 as
the X-ligand provides adjustable lability and physical
aSchool of Life Sciences, Department of Chemistry, University of Sussex, Brighton,
BN1 9QJ, UK. Tel: +44 (0) 1273 678621; E-mail: e.m.e.viseux@sussex.ac.uk
bMRC Genome Damage and Stability Centre, University of Sussex, Brighton,
BN1 9RQ, UK. Tel: +44 (0) 1273 678121; E-mail: eh58@sussex.ac.uk
†Electronic supplementary information (ESI) available: Experimental and spec-
troscopic details for all compounds are included, including X-Ray diffraction
spectra for all complexes. CCDC 915632–915634. For ESI and crystallographic
data in CIF or other electronic format see DOI: 10.1039/c3ob27460h
‡These authors contributed equally.
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
properties ideal for cell permeation. The gold amide linkage active compounds containing gold amide adducts are not com-
allows the conjugation of amino acids, which can provide monplace. Amagai et al. have reported nitrogen-containing
increased bioavailability and reduce the cytotoxicity of the compounds featuring pyrimidine dione adducts at the surface
drug in vivo. This divergent approach opens up the possibility of a gold nanoparticle, but the nature of the bond is unclear.
of generating prodrugs based on gold(I) complexes to target Importantly, these ligands derive from 5′-fluorouracil, which is
cancer cells, with reduced side effects.
therapeutically active as a standalone drug.14
Two other alternatives have been reported involving the
modification of NHC ligands with amino acids and the use of
the sulfur-containing cysteine.15 Our approach, however,
allows the use of widely available amino acids and pyridine
analogues that are not limited to specific carbene precursors
or sulfide complexes. Our complexes offer two advantages over
NHC ligands. Firstly, unlike NHCs which are strong σ-donor
ligands, the amide linkage offers the possibility for ligand
exchange. Secondly, the efficiency and ease with which wide-
ranging and varied structures can be generated paves the way
for prodrug development and orthogonal therapeutic strategies
on multiple targets.
Chemistry
Synthesis of the complexes
We herein report on the synthesis of a series of gold(I) com-
plexes 1–6 (Fig. 2 and 3) that contain amino acids with variable
physical properties. The amino acids are ligated to gold(I) by
the use of a triflic amide bond and the complexes feature a
lipophilic di- or tri-phenylphosphine backbone. Therapeutically
Modification of the X ligand
Complexes 1–3 were synthesized from tyrosine and tryptophan,
both aromatic amino acids, and methionine, a non-polar ali-
phatic amino acid. The precursor ligands featured in com-
plexes 1 and 2 were obtained after esterification of the
corresponding amino acids 7 and 8 to their α-amino esters11
followed by triflation of the primary amine with triflic anhy-
dride (Fig. 2).
Complex 3 was synthesised from methionine methyl ester 9
in 3 steps by homologation with succinic anhydride, EDC-
mediated amidification with triflic amide and ligand meta-
thesis with the silver amide derived from silver carbonate
(Fig. 2).
A different aromatic complex 4 was synthesised from
amino-pyridine 10 to contrast the activity of aromatic com-
plexes 1 and 2 with an aromatic zwitterionic complex featuring
a delocalised cationic charge. Of note, the pyridyl ligand did
not complex gold(I) as its triflic amide, but instead as a pyridi-
nium adduct as established by X-ray crystallography. The pyri-
dinium complex 4 was obtained after mono-triflation of
primary amine 10 with triflic anhydride followed by reaction
with silver carbonate and gold(I) chloride (Fig. 2).
Fig. 2 Modification of the L-ligand of the organogold complexes with biocom-
patible ligands. The synthesis of Au(I) complexes 1–6 derived from amino acids
and amino-pyridine 10.
Modification of the L-ligand
Ligand 11 derived from the polar amino acid proline was used
in both complexes 5 and 6 and was synthesized from diphenyl-
phosphine 12 (Fig. 3). Secondary amine 12 was alkylated with
methyl 3-bromopropanoate16 and the ester was subsequently
saponified to the corresponding carboxylic acid. An EDC-
mediated amidification reaction afforded the corresponding
triflic propionamide 11.
Although complexes 5 and 6 feature the same ligand 11 and
are derived from the same dimethylsulfide gold chloride
complex, the two different structures depend on the initial
presence of a silver(I) salt. The unusual dimeric complex 6 was
obtained in one synthetic operation by ligand metathesis
between a silver amide salt generated in situ from ligand 11
Fig. 3 Modification of the L-ligand of the organogold complexes with biocom-
patible ligands. The top scheme describes the synthesis of ligand 11 featured in
both the mononuclear and the dinuclear organogold complexes 5 and 6.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
and displacement of the more labile dimethylsulfide L-ligand
of Me2SAuCl. Complex 5, in contrast, was generated by
L-ligand exchange, isolation of the resulting chloride salt and
ligand metathesis with silver bis-triflic amide (Fig. 3).
Results
Cytotoxicity studies
The cell lines CV-1, MDA-MB-231 and MDA-MB-468 were
obtained from Cancer Research UK. Cells were grown as
described in the ESI.† Cell viability assays, IC50 determination,
and in vitro TrxR and GR assays are described in the ESI.†
In vitro assessment of antiproliferative effects on gold(I)
compounds on two breast cancer cell lines
Previously reported gold(I) complexes were tested for antiproli-
ferative effects on two commonly used and well characterised
breast cancer cell lines (MDA-MB-231 and MDA-MB-468) as
well as CV-1 cells from African green monkey kidney epithelial
cells (‘control’). CV-1 cells can be grown under the same
regime as the two breast cancer cell lines, have been compared
to breast (and other) cancer cell lines in numerous studies,
and the relationship between mitochondrial activity and DLC
accumulation is particularly well-studied in this cell line.17,18
No marked cytotoxicity was observed for Ph3PAuCl. This
was not due to the L-ligand, since Me2SAuCl was also ineffec-
tive in mediating cellular toxicity. The CV-1 cells were
unaffected by either precursors at concentrations up to 50 μM.
Only at concentrations above 30 μM did we observe a slight
effect on the MDA-MB-231; however, at 50 μM, the IC50 had
not yet been reached for either precursor.
In contrast, Ph3PAuNTf2 displayed preferential inhibition of
the two cancer cell lines compared to CV-1 (Fig. 4). This clearly
demonstrates that a gold triflic amide linkage can successfully
be used in gold complexes to induce cytotoxicity in the cancer
cell lines described herein.
Currently, auranofin is undergoing phase II clinical trials
for treatment of chronic lymphocytic leukemia (CCL).
Ph3PAuNTf2 was found to be as cytotoxic as auranofin (Fig. 4).
Although the gold-sulfide has been changed to a gold-sele-
nide,19 further derivatization of auranofin is limited to either
the functionalization of its acetate group or regioselective
transesterification to different esters.
Moreover, side effects of auranofin are common and
prodrug delivery of the gold(I) is therefore desirable. Using
Ph3PAuNTf2 as a model for the amide linkage, we derivatized a
number of amino acids, a common mechanism for prodrug
formulation.§
Fig. 4 Cell viability in response to exposure to gold(I) complexes. Cell viability
was determined by measuring the fluorescence of untreated (‘Medium’),
DMSO-treated (‘0 μM’), and gold(I) treated cells. The ‘No cells’ control was used
as background subtraction. The normalized fluorescence was calculated as the
background-subtracted fluorescence reading, divided by the background-sub-
tracted fluorescence reading of the DMSO-treated control. The error bars corre-
spond to the standard deviation of the mean of four to eight replicates. ND, not
done.
contain non-polar aliphatic groups and therefore were pre-
dicted to result in highly lipophilic complexes. However, the
complexes proved insoluble in an aqueous medium.
Complexes 1 and 2 were derived from aromatic tyrosine and
tryptophan, respectively. These complexes contain a hetero-
atom in the side chain and were soluble. The initial screen of
these compounds revealed marked cytotoxic effects on par
with Ph3PAuNTf2. The MDA-MB-468 cell line was consistently
more sensitive to the gold(I) compounds compared to the
MDA-MB-231. As with Ph3PAuNTf2, complexes 1 and 2 were
comparable to auranofin in the inhibition of the MDA-MB-231
cell line (Fig. 4).
Our initial attempts at derivatization of Ph3P–Au–NTf2
involved the two amino acids valine and leucine, which
§For example next-generation valacyclovir was modified by esterification of acy-
clovir with valine, which improved its bioavailability dramatically.
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem.

View Article Online
Paper
Further analysis revealed IC50 values in the low and even
Organic & Biomolecular Chemistry
sub-micromolar ranges (ESI Fig. S1 online†). Importantly, the
ligands alone did not induce any significant cell toxicity (ESI
Fig. S2 online†). Therefore, generating amino acid derivatives
of Ph3PAuNTf2 does not interfere with its cytotoxic effects.
We also investigated the activity of complex 3 derived from
methionine to model the use of peptidic bonds as linkages for
ligand customisation. This modification also did not impair
the cytotoxicity profile (Fig. 4) and clearly shows that derivati-
zing the NTf2 X-ligand with amino acid R-groups does not
impact on its cytotoxic effects.
The possibility of a delocalised cation in a zwitterionic
complex was investigated with complex 4. This contrasts with
previous work reported by Hickey et al. on delocalised lipophi-
lic cations where two NHC ligands were used in a cationic
complex. Here, the use of one X-ligand based on a pyridinium
salt concomitantly with a liphophilic triphenylphosphine suc-
cessfully mediated cellular toxicity in the two cancer cell lines,
without affecting CV-1 cells. The ligand alone produced no
cytotoxic effect (ESI Fig. S2 online†). These data show that the
gold-amide ligand effects are substantial and that the derived
ligands significantly influence the efficacy of the gold(I) com-
plexes in mediating cellular toxicity.
Fig. 5 Nigericin sensitizes CV-1 cells to gold(I) compounds. CV-1 and
MDA-MB-231 cell lines were treated with increasing concentrations of nigericin
in addition to 4.28 or 5 μM gold(I) complex (black bars) or 0 μM gold(I), DMSO-
treatment (grey bars). Ethanol-only treatment of the cell lines (no gold(I) com-
pounds) was used as ‘Ethanol’ control. ‘No cells’ control was used for back-
ground subtraction. The error bars represent standard deviation of the mean of
three replicates.
To rationalise the effect of the phosphine ligand, pyrroli-
dine 11 was used as an L-ligand by formally replacing a phenyl
ring with a modified proline, as seen in complex 5. Interest-
ingly, this change in the stereoelectronic and physical proper-
ties of the L-ligand did not impact on its cytotoxicity against
either of the cancer cell lines when compared to Ph3P–Au–
NTf2. This broadens the possibilities for further modification
of the gold complexes (Fig. 4).
We have here established that, as long as a triflic amide is
used as an X-ligand, amino acids can also be derivatised to
serve as L-ligands in gold amide complexes without affecting
the cytotoxicity of the complex. It should therefore be possible
to modulate the polarity of L-ligands, which can in turn be
used as prodrugs or modulators for cytotoxicity.
When used in combination with complexes 1 or 2, nigericin
caused a marked further sensitization to the aforementioned
complexes (Fig. 4). In contrast, although MDA-MB-231 cells
also showed a degree of sensitivity to nigericin alone, the
addition of complexes 1 or 2 caused no further decrease in cell
viability (Fig. 5).
Complexes 1 and 2 inhibit TrxR but not GR in vitro
A selenocysteine-containing protein, thioredoxin reductase
(TrxR), has been demonstrated to be inhibited by gold(I) com-
plexes via coordination of the transition metal to the seleno-
cysteine residue.5,22,23 Increasing concentrations of complexes
1 and 2 inhibited the TrxR enzyme in the nM range in vitro
(Fig. 6). Treatment with solely the ligands did not inhibit TrxR,
demonstrating that gold(I) is required for the inhibition of
TrxR. The related protein, glutathionone reductase (GR), which
contains a cysteine, was unaffected, suggesting that the gold(I)
complexes selectively inhibit proteins containing a selenol
function.5,12,24
Finally, we tested the potential of a dinuclear bis-gold(I)
complex 6 in its efficacy to kill cancer cells, but the results
were similar to that of complex 3 and increased activity was
not observed.
Preferential sensitization of CV-1 cells to gold(I) complexes
with treatment of nigericin
Mitochondrial hyperpolarization of adenocarcinomas from the
two cell lines used in this study is hypothesized to be the
cause of the selective inhibition of cell proliferation due to
Conclusions
enhanced accumulation of gold(I) in the mitochondria.5,17 In our search for therapeutically active gold(I) complexes fea-
This predicts that hyperpolarizing the mitochondrial mem- turing an amide linkage stable enough to not dissociate in
branes of CV-1 cells with the K+/H+ ionophore, nigericin, solution in vivo, yet retaining the therapeutic activity of gold(I),
should sensitize them further to the gold complexes.20,21 To we envisaged utilising the complexes described in Fig. 1.
test this hypothesis, CV-1 cells were treated with increasing These contain a gold amide linked to an electron withdrawing
concentrations of nigericin in addition to 0 or 4.28 μM of group and a side group. We have herein reported on their syn-
complex 1 or 0 or 5 μM of complex 2. Nigericin treatment thesis and derivatisation with robust, simple and high-yielding
alone caused a concentration-dependent decrease in cell routes which are amenable to high-throughput synthesis. The
viability.
ligands themselves did not impair the activity of the
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
4; and Dr Peter Karran (CRUK) for cell lines and Dr Judith
Offman for setting up cell lines. We also thank Prof. M. Bagley,
Prof. P. J. Parsons, Prof. S. Ward and Drs O. Navarro and
J. Spencer for their continued interest in this work. We also
wish to thank Chris Dadswell and Aidan Fisher for their work
on the ICP-MS studies and lipophilicity determination
respectively.
Notes and references
1 C. Orvig and M. J. Abrams, Medicinal inorganic chemistry:
introduction, Chem. Rev., 1999, 99, 2201–2204.
2 M. Patra and G. Gasser, Organometallic compounds: an
opportunity for chemical biology?, ChemBioChem, 2012, 13,
1232–1252.
3 I. Ott and R. Gust, Non platinum metal complexes as anti-
cancer drugs, Arch. Pharm., 2007, 340, 117–126.
4 M. Stallings-Mann, L. Jamieson, R. P. Regala, C. Weems,
N. R. Murray and A. P. Fields, A novel small-molecule
inhibitor of protein kinase Ciota blocks transformed
growth of non-small-cell lung cancer cells, Cancer Res.,
2006, 66, 1767–1774.
5 J. L. Hickey, R. A. Ruhayel, P. J. Barnard, M. V. Baker,
S. J. Berners-Price and A. Filipovska, Mitochondria-targeted
chemotherapeutics: the rational design of gold(I) N-hetero-
cyclic carbene complexes that are selectively toxic to cancer
cells and target protein selenols in preference to thiols,
J. Am. Chem. Soc., 2008, 130, 12570–12571.
6 C. Wetzel, P. C. Kunz, M. U. Kassack, A. Hamacher,
P. Bohler, W. Watjen, I. Ott, R. Rubbiani and B. Spingler,
Gold(I) complexes of water-soluble diphos-type ligands: syn-
thesis, anticancer activity, apoptosis and thioredoxin
reductase inhibition, Dalton Trans., 2011, 40, 9212–9220.
7 J. C. Lima and L. Rodriguez, Phosphine–gold(I) compounds
as anticancer agents: general description and mechanisms
of action, Anti-Cancer Agents Med. Chem., 2011, 11, 921–928.
8 C. Gabbiani and L. Messori, Protein targets for anticancer
gold compounds: mechanistic inferences, Anti-Cancer
Agents Med. Chem., 2011, 11, 929–939.
Fig. 6 TrxR activity in vitro is inhibited by gold(I) complexes. Increasing concen-
trations of complexes 1 and 2 or ‘ligand only’ were added to TrxR and GR
enzymes and their activity was monitored by fluorescence. Normalized fluor-
escence values were calculated by taking the fluorescence value of the treated
well and dividing it by the fluorescence value of 0 nM compound (DMSO-only
treatment).
complexes, and were shown to have no cytotoxic effects on
their own. We have also demonstrated the compatibility of a
wide array of functionalities such as indoles, phenol triflates,
esters, methyl sulfide, amides, aliphatic chains, aromatic
groups, diarylalkyl- and triaryl-phosphines. An alternative
chemical environment that was equally active as the gold triflic
amide complexes was also validated. The use of a zwitterionic
ligand bearing a delocalised cation clearly sets a precedent for
the use of therapeutically active analogues of globally electro-
neutral pyridinium gold(I) complexes in treating cancer. The
two approaches described herein open up new possibilities for
an increased diversity in new chemotypes and prodrug deliv-
ery, including peptoid and peptidic precursors as well as
orthogonal anticancer drugs.
The selective accumulation of the gold(I) complexes in
adenocarcinoma cells due to their mitochondrial hyperpolariz-
ation may provide a favourable therapeutic index for the treat-
ment of tumours. This opens up the possibility of targeting
cancers in elderly patients and of slow-growing tumours,
where the conventional cancer therapeutics aimed at rapidly
dividing cells cannot be used.
9 S. P. Fricker, Strategies for the biological evaluation of gold
anticancer agents, Anti-Cancer Agents Med. Chem., 2011, 11,
940–952.
10 X. Zhang, M. Frezza, V. Milacic, L. Ronconi, Y. Fan, C. Bi,
D. Fregona and Q. P. Dou, Inhibition of tumor proteasome
activity by gold-dithiocarbamato complexes via both redox-
dependent and -independent processes, J. Cell. Biochem.,
2010, 109, 162–172.
11 K. Yan, C. N. Lok, K. Bierla and C. M. Che, Gold(I) complex
of N,N′-disubstituted cyclic thiourea with in vitro and
in vivo anticancer properties-potent tight-binding inhi-
bition of thioredoxin reductase, Chem. Commun., 2010, 46,
7691–7693.
Funding sources
E.H. is the recipient of an MRC Senior Fellowship and EMBO
Young Investigator Award. E.M.E.V. is funded by an Alfred
Bader award.
Acknowledgements
We thank Dr Iain Day for the NMR service and Dr Alaa Abdul- 12 C. X. Zhang and S. J. Lippard, New metal complexes as
Sada for the Mass Spectrometry Service at the University of
Sussex; the NCS at Southampton for X-ray data for compound
potential therapeutics, Curr. Opin. Chem. Biol., 2003, 7,
481–489.
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
13 N. Mezailles, L. Ricard and F. Gagosz, Phosphine gold(I)
bis-(trifluoromethanesulfonyl)imidate complexes as new
highly efficient and air-stable catalysts for the cycloisomer-
ization of enynes, Org. Lett., 2005, 7, 4133–4136.
spectroscopic (EXAFS, 197Au Mossbauer, 31P, 1H, 13C, and
77Se NMR, ESI-MS) characterization, biological activity, and
rapid serum albumin-induced triethylphosphine oxide gen-
eration, Inorg. Chem., 2010, 49, 7663–7675.
14 T. Amagai, T. K. Miyamoto, H. Ichida and Y. Sasaki, Prepa- 20 D. T. Poole, T. C. Butler and M. E. Williams, The effects of
ration and crystal-structure of new gold(I) complexes linked
to pyrimidines, Bull. Chem. Soc. Jpn., 1989, 62, 1078–1080.
15 J. Lemke, A. Pinto, P. Niehoff, V. Vasylyeva and N. Metzler-
nigericin, valinomycin, and 2,4-dinitrophenol on intracel-
lular pH, glycolysis, and K + concentration of Ehrlich ascites
tumor cells, Biochim. Biophys. Acta, 1972, 266, 463–470.
Nolte, Synthesis, structural characterisation and anti-prolif- 21 S. Davis, M. J. Weiss, J. R. Wong, T. J. Lampidis and
erative activity of NHC gold amino acid and peptide conju-
gates, Dalton Trans., 2009, 7063–7070.
16 A. Thaqi, J. L. Scott, J. Gilbert, J. A. Sakoff and
A. McCluskey, Synthesis and biological activity of delta-5,6-
L. B. Chen, Mitochondrial and plasma membrane poten-
tials cause unusual accumulation and retention of rhod-
amine 123 by human breast adenocarcinoma-derived
MCF-7 cells, J. Biol. Chem., 1985, 260, 13844–13850.
norcantharimides: importance of the 5,6-bridge, 22 A. G. Cox, K. K. Brown, E. S. Arner and M. B. Hampton, The
Eur. J. Med. Chem., 2010, 45, 1717–1723.
17 L. B. Chen, Mitochondrial membrane potential in living
cells, Annu. Rev. Cell Biol., 1988, 4, 155–181.
thioredoxin reductase inhibitor auranofin triggers apopto-
sis through a Bax/Bak-dependent process that involves per-
oxiredoxin 3 oxidation, Biochem. Pharmacol., 2008, 76,
1097–1109.
18 J. S. Modica-Napolitano, R. Nalbandian, M. E. Kidd,
A. Nalbandian and C. C. Nguyen, The selective in vitro cyto- 23 Y. Omata, M. Folan, M. Shaw, R. L. Messer, P. E. Lockwood,
toxicity of carcinoma cells by AZT is enhanced by concur-
rent treatment with delocalized lipophilic cations, Cancer
Lett., 2003, 198, 59–68.
D. Hobbs, S. Bouillaguet, H. Sano, J. B. Lewis and
J. C. Wataha, Sublethal concentrations of diverse gold com-
pounds inhibit mammalian cytosolic thioredoxin reductase
(TrxR1), Toxicol. in Vitro, 2006, 20, 882–890.
19 D. T. Hill, A. A. Isab, D. E. Griswold, M. J. DiMartino,
E. D. Matz, A. L. Figueroa, J. E. Wawro, C. DeBrosse, 24 M. Bjornstedt, S. Kumar, L. Bjorkhem, G. Spyrou and
W. M. Reiff, R. C. Elder, B. Jones, J. W. Webb and
A. Holmgren, Selenium and the thioredoxin and glutare-
C. F. Shaw, Seleno-auranofin (Et3PAuSe-tagl): synthesis,
doxin systems, Biomed. Environ. Sci., 1997, 10, 271–279.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2013