High levels of thioredoxin reductase 1 modulate drug-specific cytotoxic efficacy

Abstract

The selenoprotein thioredoxin reductase 1 (TrxR1) is currently recognized as a plausible anticancer drug target. Here we analyzed the effects of TrxR1 targeting in the human A549 lung carcinoma cell line, having a very high basal TrxR1 expression. We determined the total cellular TrxR activity to be 271.4 +/- 39.5 nmol min(-1) per milligram of total protein, which by far exceeded the total thioredoxin activity (39.2 +/- 3.5 nmol min(-1) per milligram of total protein). Knocking down TrxR1 by approx 90% using siRNA gave only a slight effect on cell growth, irrespective of concurrent glutathione depletion (> or = 98% decrease), and no increase in cell death or distorted cell cycle phase distributions. This apparent lack of phenotype could probably be explained by Trx functions being maintained by the remaining TrxR1 activity. TrxR1 knockdown nonetheless yielded drug-specific modulation of cytotoxic efficacy in response to various chemotherapeutic agents. No changes in response upon exposure to auranofin or juglone were seen after TrxR1 knockdown, whereas sensitivity to 1-chloro-2,4-dinitrobenzene or menadione became markedly increased. In contrast, a virtually complete resistance to cisplatin using concentrations up to 20 microM appeared upon TrxR1 knockdown. The results suggest that high overexpression of TrxR has an impact not necessarily linked to Trx function that nonetheless modulates drug-specific cytotoxic responses.

Full text

Original Contribution
High levels of thioredoxin reductase 1 modulate drug-specific cytotoxic efficacy
SofiE. Eriksson
a
, Stefanie Prast-Nielsen
a
, Emilie Flaberg
b
, Laszlo Szekely
b
, Elias S.J. Arnér
a,
⁎
a
Division of Biochemistry, Department of Medical Biochemistry and Biophysics, Karolinska Institutet, SE-171 77 Stockholm, Sweden
b
Department of Microbiology, Tumor and Cell Biology, Karolinska Institutet, SE-171 77 Stockholm, Sweden
abstractarticle info
Article history:
Received 17 June 2009
Revised 31 August 2009
Accepted 14 September 2009
Available online 17 September 2009
Keywords:
Thioredoxin reductase
Selenoprotein
Cancer
Cytotoxicity
Chemotherapy
Free radicals
The selenoprotein thioredoxin reductase 1 (TrxR1) is currently recognized as a plausible anticancer drug
target. Here we analyzed the effects of TrxR1 targeting in the human A549 lung carcinoma cell line, having a
very high basal TrxR1 expression. We determined the total cellular TrxR activity to be 271.4 ± 39.5 nmol
min
−1
per milligram of total protein, which by far exceeded the total thioredoxin activity (39.2 ± 3.5 nmol
min
−1
per milligram of total protein). Knocking down TrxR1 by approx 90% using siRNA gave only a slight
effect on cell growth, irrespective of concurrent glutathione depletion (≥98% decrease), and no increase in
cell death or distorted cell cycle phase distributions. This apparent lack of phenotype could probably be
explained by Trx functions being maintained by the remaining TrxR1 activity. TrxR1 knockdown nonetheless
yielded drug-specific modulation of cytotoxic efficacy in response to various chemotherapeutic agents. No
changes in response upon exposure to auranofin or juglone were seen after TrxR1 knockdown, whereas
sensitivity to 1-chloro-2,4-dinitrobenzene or menadione became markedly increased. In contrast, a virtually
complete resistance to cisplatin using concentrations up to 20 μM appeared upon TrxR1 knockdown. The
results suggest that high overexpression of TrxR has an impact not necessarily linked to Trx function that
nonetheless modulates drug-specific cytotoxic responses.
© 2009 Elsevier Inc. All rights reserved.
To protect from oxygen radical-induced damage, cells have
developed multifaceted antioxidant systems [1]. The thioredoxin
(Trx) system is one important enzymatic network for antioxidant
defense and has also a number of additional redox regulatory
functions [1,2]. Trx reduces and thereby supports the activity of
several proteins, e.g., antioxidant peroxiredoxins or methionine
sulfoxide reductase and redox-regulated transcription factors or
other signaling molecules [2,3]. The Trx system also supports
deoxyribonucleotide synthesis, with reduced Trx regenerating a
dithiol in ribonucleotide reductase oxidized upon each cycle of
catalysis [4]. For its enzymatic function, Trx must be reduced by
TrxR (EC 1.8.1.9). In human, three TrxR-encoding genes are found, i.e.,
TXNRD1, encoding TrxR1, which is the major TrxR form in most cells
[5,6];TXNRD2, encoding TrxR2, which is predominantly found in
mitochondria [7]; and TXNRD3, for thioredoxin glutathione reductase
(TGR) mainly expressed in testis [8]. Both TrxR1 and TrxR2 are
essential for embryonic development as shown in knockout mouse
models [9,10], as are their principal substrates Trx1 and Trx2 [11,12].
TrxR proteins are the only enzymes known to reduce Trx and are
thereby believed to be essential for all Trx-dependent reduction
processes. Being selenoproteins, mammalian TrxR's and thereby the
complete Trx systems are fully dependent upon selenium [13,14].
Increasing selenium availability generally results in increased TrxR1
activity, until saturation levels are reached [15–18].
With regard to cancer development and treatment, most studies
have focused on the potential importance of TrxR1 as a drug target,
as reviewed elsewhere [19–22] and as shall briefly be introduced
here. Several observations show that both TrxR1 and Trx1 are often
overexpressed in various cancer forms under basal selenium supply
[21]. Furthermore, TrxR1 has been shown to promote and even be
essential for tumor growth in xenograft cancer models [23,24].
Cancer cells with lowered TrxR1 expression have also been shown to
be more sensitive to UV irradiation or to low doses of cadmium
[25,26]. It is not only for its growth-promoting properties that TrxR1
has been identified as a potential target for anticancer therapy, but
also because of its selenium-dependent enzymological properties.
The Sec residue is an integral part of the TrxR1 active site [27,28].
With a high reactivity of Sec and an easily accessible active site,
TrxR1 has a broad substrate specificity, reducing not only Trx1, but
also other agents, e.g., 5,5′-dithiobis[2-nitrobenzoic acid] (DTNB),
Free Radical Biology & Medicine 47 (2009) 1661–1671
Abbreviations: Trx, thioredoxin; TrxR, thioredoxin reductase; DTNB, 5,5′-dithiobis
[2-nitrobenzoic acid]; Sec, selenocysteine; DNCB, 1-chloro-2,4-dinitrobenzene; cDDP,
cisplatin; Oxa, oxaliplatin; SecTRAP, selenium-compromised thioredoxin reductase-
derived apoptotic protein; A549, human lung carcinoma cell line; PEST, penicillin and
streptomycin; siRNA, small inhibitory RNA; FCS, fetal calf serum; GR, glutathione
reductase; GSH, glutathione; GSSG, glutathione disulfide; PI, propidium iodide; BSO, L-
buthionine sulfoximine; Grx, glutaredoxin; ROS, reactive oxygen species; PBS,
phosphate-buffered saline; NADPH, nicotinamide adenine dinucleotide phosphate.
⁎Corresponding author. Fax: +46 8 31 15 51.
E-mail address: Elias.Arner@ki.se (E.S.J. Arnér).
0891-5849/$ –see front matter © 2009 Elsevier Inc. All rights reserved.
doi:10.1016/j.freeradbiomed.2009.09.016
Contents lists available at ScienceDirect
Free Radical Biology & Medicine
journal homepage: www.elsevier.com/locate/freeradbiomed

lipoic acid, lipoamide, selenite, menadione (vitamin K
3
), ubiquinone,
and vitamin C [1,5,29–33]. Moreover, the N-terminal CVNVGC motif
can reduce certain substrates directly [34]. In addition, the Sec
residue in reduced (but not oxidized) TrxR can also be rapidly
targeted by electrophilic agents. Several gold compounds are potent
inhibitors of TrxR1, e.g., auranofin used in treatment of rheumatoid
arthritis, which inhibits the enzyme even in the nanomolar range
[7,35,36]. Recent results also show that auranofin inhibits overall
selenoprotein synthesis [37]. The first identified inhibitor of TrxR1
was 1-chloro-2,4-dinitrobenzene (DNCB), which also concomitantly
induces an inherent superoxide-producing NADPH oxidase activity in
the derivatized enzyme [38]. Other inhibitors of TrxR1 include
naturally occurring electrophiles such as quinones [39], isothiocya-
nates [40], mercury [41], or various flavonoids [42,43], as well as
several clinically used anticancer drugs including arsenicals [44] and
cisplatin (cDDP) [45–47]. TrxR1 derivatized with cDDP generates an
enzyme inactive for its normal function, but which may gain a new
pro-oxidant function within a cellular context, capable of inducing a
rapid cell death in the form of selenium-compromised thioredoxin
reductase-derived apoptotic proteins, SecTRAPs [48,49]. All of these
properties combined have contributed to the commonly held view
that TrxR1 may be a prime molecular target for anticancer
chemotherapy [19–22].
The purpose of this study was to investigate the impact of an
endogenously high expression of the selenium-dependent TrxR1 in
cancer cells, using knockdown of the enzyme in a human lung
carcinoma cell line (A549) that has among the highest known
overexpression of TrxR1 of all studied cancer cells (www.proteinatlas.
org). By a thoroughanalysis of the phenotype of these cells upon TrxR1
knockdown, we reasoned that insights could be gained regarding the
functional consequences of overexpression of the enzyme in cancer
cells and of its targeting in such cells by chemotherapy. The results
reveal a surprising complexity in the roles of TrxR1 for modulating
cellular responses to different cytotoxic agents.
Experimental procedures
Chemicals and reagents
Recombinant rat TrxR1 was produced as described [50] and
human wt Trx [51] was kindly provided by Arne Holmgren
(Karolinska Institutet, Stockholm, Sweden). Yeast glutathione reduc-
tase (GR) (Cat. No. G3664), reduced glutathione (GSH), and oxidized
glutathione (GSSG) were obtained from Sigma–Aldrich Chemicals
(Steinheim, Germany). Chemicals used in the drug sensitivity assays
were as follows: auranofin from Alexis Biochemicals (San Diego, CA,
USA), cDDP (Platinol) from Bristol–Myers Squibb Pharmaceuticals
(New York, NY, USA), and DNCB, juglone (5-hydroxy-1,4,-naphtho-
quinone), menadione (2-methyl-1,4-naphtoquinone, vitamin K
3
),
and oxaliplatin (Oxa) all from Sigma–Aldrich Chemicals. All other
regular chemicals or reagents were of high purity and purchased from
Sigma–Aldrich Chemicals, unless otherwise specified.
Cell cultures
Human lung carcinoma cells (A549 cells) were obtained from the
American Tissue Culture Collection (CCL-185) and cultivated in
Dulbecco's modified Eagle medium with 4.5 g/l glucose content
(GIBCO/Invitrogen, Carlsbad, CA, USA). The medium was supplemen-
ted with 10% heat-inactivated fetal calf serum (FCS), 2 mM L-
glutamine, 100 μg/ml streptomycin, and 100 units/ml penicillin
(PEST), all from PAA Laboratories (Pasching, Austria). Selenium was
added to the medium in the form of sodium selenite at concentrations
as described in the text. Cells were grown at 37°C in a humidified
atmosphere with 5% CO
2
and kept under logarithmic growth phase for
all experiments unless stated otherwise.
Transient knockdown of TrxR1
Small interfering RNA (siRNA) molecules, specifically targeting the
TrxR1 mRNA, were obtained from Qiagen (Valencia, CA, USA). For
phenotypic confirmation, two different siRNA sequences were used
(herein named Siseq1 and Siseq2), targeting different areas of the
TrxR1 mRNA: Siseq1, sense 5′-(GCAAGACUCUCGAAAUUAU)dTdT-3′,
antisense 5′-(AUAAUUUCGAGAGUCUUGC)dAdG-3′, and Siseq2, sense
5′-(CCUGGCAUUUGGUAGUAUA)dTdT-3′, antisense 5′-(UAUACUAC-
CAAAUGCCAGG)dCdA-3′. Siseq1 targets an mRNA region encoding
the N-terminal redox-active CVNVGC motif in the protein and Siseq2
targets a site in the exon covering the 3′-untranslated region (3′UTR)
of the mRNA downstream of the selenocysteine insertion sequence
element, needed for Sec incorporation [52]. Both siRNA constructs
should thereby knock down all major TrxR1 splice forms (Supple-
mentary Fig. S1). Two different nonsilencing siRNA controls, showing
no apparent homology to any region of the human genome, were
used, i.e., the Alexa 488-labeled AllStar negative control (Qiagen) as
control for transfection efficiency and an unlabeled scramble control
(mock) (Cat. No. 1022076; Qiagen) used for all other experiments.
The sequences for the mock siRNA construct were sense 5′-
(UUCUCCGAACGUGUCACGU)dTdT-3′and antisense 5′-(ACGUGA-
CACGUUCGGAGAA)dTdT-3′. Untransfected control cells as well as
mock-treated cells were included in all siRNA experiments, as
described in the text. For transfection experiments A549 cells were
seeded in six-well plates at a density of about 30,000–32,000 cells per
well 15–18 h before transfection. SiRNA transfection was then
performed according to the manufacturer's protocol by mixing 9 μl
transfection reagent (Hiperfect; Qiagen) and 10 nM siRNA duplexes
in a total volume of 100 μl serum-free medium, per sample. Cells
were incubated for 24 h with the transfection complexes in 2.4 ml
medium without antibiotics, whereupon the medium was replaced
with fresh PEST-containing medium and experiments were con-
ducted as described.
Selenium content in medium
Selenium (m/z 78) content in the fetal calf serum utilized in this
study was analyzed with inductively coupled plasma mass spectrom-
etry as described elsewhere [53] and was determined to be 18.7 ng Se
per gram of FCS, i.e., approx 18.7 μg/L. In the medium supplemented
with 10% FCS, the total selenium concentration would thereby be in
the range of 20–25 nM. In the experiments as described in the text, an
additional 25 nM selenium in the form of sodium selenite was added
to the cells during seeding to saturate the synthesis of TrxR1 in the cell
cultures. Extensive control experiments verified that this was a
required but sufficient selenium supplementation to reach TrxR1
saturation.
75
Se radioisotope labeling of cellular proteins
To verify knockdown of TrxR1, cells were seeded in six-well plates
together with
75
Se-labeled selenite (Research Reactor Center, Uni-
versity of Missouri, Columbia, MO, USA) added to the medium, using
2μCi/ml in a total volume of 1.5 ml. No additional unlabeled selenite
was added. Using the specification from the distributors the total
concentration of selenite added to the cells was calculated to be
within the range of 20 to 100 nM. Fifteen to eighteen hours after
seeding with [
75
Se]selenite, siRNA was added to the cells in the
presence of 2 μCi/ml
75
Se, according to the same procedures as
described above. Cell extracts were subsequently prepared 48, 72, 96,
and 144 h post-siRNA treatment and 25 μg total protein per sample
was analyzed on 4–12% NuPAGE Bis–Tris reducing SDS–PAGE
(equipment, gels, and buffers from Invitrogen). Gels were stained
with Coomassie blue to visualize total protein. The gels were then
dried and exposed on a phosphor screen. The autoradiography was
1662 S.E. Eriksson et al. / Free Radical Biology & Medicine 47 (2009) 1661–1671

visualized with a StormScan PhosphoImager (Molecular Dynamics,
Sunnyvale, CA, USA). Intensities of the
75
Se-labeled bands with an
approx molecular weight of 55 kDa were quantified using Bio-Rad's
Quantity One 1-D 4.6.7 (Hercules, CA, USA).
Preparation of cell extracts
Cells to be used for enzymatic assays, chromatography, or
autoradiography were harvested by trypsinization (GIBCO/Invitrogen),
washed with phosphate-buffered saline, pH 7.4 (PBS; GIBCO/
Invitrogen), and centrifuged (800 g for 5 min). The resulting cell
pellets were resuspended in extraction buffer containing 50 mM Tris–
HCl, pH 7.5, 2 mM EDTA (Merck, Darmstadt, Germany), 0.5 mM
phenyl methyl sulfonyl fluoride, and 0.5% nonionic detergent (Igepal
Ca-630). Cells were lysed by rapid cycles of freezingand thawing using
liquid nitrogen and a 37°C water bath, respectively. The nonsolubilized
fraction was subsequently removed by centrifugation (16,000 g at 4°C
for 5 min) and the supernatant was used as protein source. The
protein concentrations were determined by the Bradford method (Bio-
Rad Laboratories), using bovine serum albumin as a standard.
Anion-exchange chromatography and mass spectrometry
To determine the identity of the remaining TrxR activity in siRNA-
treated cells, the total protein of the corresponding cell extracts was
subjected to chromatography using a strong anion Resource Q column
(1 ml; GE Healthcare, Little Chalfont, UK) on an ÄKTAexplorer 900
purifier HPLC system (GE Healthcare). Buffers for the gradient elution
were as follows: buffer A, 50 mM Tris–Cl and 2 mM EDTA (TE buffer),
pH 7.5, and buffer B, TE buffer, pH 7.5, with 1 M NaCl. The system was
equilibrated with buffer A, whereupon 300 μg of total protein in a total
volume of 0.5 ml was applied to the column, which was then washed
with buffer A and the flowthrough was monitored using absorbance at
280 and 254 nm. Separation was performed using a three-step
gradient as shown, with TrxR activity eluted at 50–200 mM NaCl.
Fractions of a fixed volume (0.5 ml) were continuously collected using
an automated fraction collector (Frac-950; GE Healthcare) and
analyzed for TrxR activity as described. Fractions with the highest
TrxR activity were pooled and concentrated using a 30-kDa cutoff
filter column (Microcon 0.5 ml; Millipore, Bedford, MA, USA).
Concentrated samples were separated by SDS–PAGE and subsequent-
ly subjected to mass spectrometry (MS). Recombinant rat TrxR1
protein was used as migration control in the gel. Briefly, Coomassie-
stained bands in the size range of 55 kDa were cut out from the gel and
subsequently subjected to in-gel trypsin digestion, peptide extraction,
mass mapping, and database searches carried out at the Protein
Analysis Center, Karolinska Institutet. The details of a typical result are
shown in supplementary Fig. S2.
TrxR and Trx activity assays
TrxR activity was determined using the previously described
end-point Trx-dependent insulin reduction method [54], modified
and applied to microtiter plates [39]. Total cellular protein (5 μg)
was incubated with 20 μM recombinant human wt Trx [51] in the
presence of 297 μM insulin, 1.3 mM NADPH (AppliChem, Darmstadt,
Germany), 85 mM Hepes buffer, pH 7.6, and 13 mM EDTA for
40 min at 37°C, in a total volume of 50 μl. The reaction was then
stopped by addition of 200 μl of 7.2 M guanidine–HCl (Acros
Organics, NJ, USA) in 0.2 M Tris–HCl, pH 8.0, containing 1 mM
DTNB. The extent of Trx-dependent formed thiols in the reduced
insulin was then determined by measuring absorbance at 412 nm
(extinction coefficient 13,600 M
−1
cm
−1
)usingaVersaMaxmicro-
plate reader (Molecular Devices, Sunnyvale, CA, USA) with a
background absorbance reference for each sample containing all
components except Trx, incubated and treated in the same manner.
To determine Trx activity in the samples, 5 μg total protein was
incubated with 250 nM rat TrxR1 (22 U/mg) instead of Trx for
60 min, with otherwise the same conditions as described for the TrxR
activity assay. TrxR activity assays on the fractions from the HPLC
purifications were also performed with the Trx-coupled insulin
reduction assay using 10 μl of the fractions and incubation for 90 min.
GR activity assay
To monitor NADPH-dependent GSSG reduction, the spectropho-
tometric method previously described [55] and modified for micro-
titer plates [39] was used. In short, 2 μg of total cellular protein was
diluted in deionized water to a final volume of 80 μl. Then 120 μlof
freshly prepared master mix containing 2 mM GSSG and 200 μM
NADPH in 0.2 M potassium phosphate, pH 7, with 2 mM EDTA (PE
buffer) was added simultaneously to all samples. These assays were
performed in 96-well microtiter plates at 30°C with oxidation of
NADPH determined from the decrease in absorbance at 340 nm for the
first 3 min using a VersaMax microplate reader and an extinction
coefficient of 6200 M
−1
cm
−1
. A background sample containing
everything except GSSG was included in each case and its change in
absorbance was subtracted from the sample values.
Cell growth, cell cycle phase, and viability assessments
To measure cell death and analyze the DNA content profile by flow
cytometry, all floating as well as attached cells were recovered. The
cells were fixed using addition of ice-cold 70% ethanol and then
incubated at room temperature for 30 min. Subsequently DNA
staining was done by incubating for 30 min at 4°C with PBS containing
20 μg/ml propidium iodide (PI), 0.2 mg/ml RNase, and 0.1% Triton X-
100. Samples were subsequently analyzed in a fluorescence-activated
cell sorter (Becton–Dickinson FACScan, Rutherford, NJ, USA). Data
collected from 10,000 cells per sample were analyzed with Becton–
Dickinson Cell Quest Pro version 4.0.2.
GSH depletion
To investigate the importance of GSH-dependent pathways in the
TrxR1-siRNA-treated cells, GSH levels were lowered by preincubation
with 0.25 mM L-buthionine sulfoximine (BSO). After siRNA transfec-
tion BSO was added to the cells and was subsequently present in the
incubation for an additional 48 h before analysis. Total cell numbers
upon treatments were determined using a Bürker hemacytometer
chamber.
Quantification of total GSH and GSSG
Measurements of total intracellular GSH and GSSG concentrations
were performed according to the glutathione reductase−DTNB
recycling assay [56], further improved and modified to a microtiter
plate format [57]. For this, A549 cells were trypsinized, washed with
cold PBS, and centrifuged (800 g for 5 min). The resulting cell pellets
were immediately resuspended in 150 μl ice-cold 10 mM HCl and
lysed by three cycles of freezing and thawing using liquid nitrogen
and a 37°C water bath, respectively. To determine the protein
concentrations in the samples 10-μl aliquots were first transferred
to new separate tubes. Proteins were subsequently precipitated from
the main samples by adding 30 μl 5% (w/v) 5-sulfosalicylic acid (SSA)
to 120 μl total sample volume. The samples were then incubated for
10 min, after which the proteins were removed by centrifugation for
15 min (8000 g, at 4°C). The resulting supernatant was collected and
stored at −80°C until analysis of glutathione content. To microtiter
plate wells 20 μl of sample, background control, or GSH standard
solution was added. To neutralize the pH, 20 μl of 143 mM NaH
2
PO
4
buffer containing 6.3 mM EDTA, pH 7.5 (stock buffer), was added. A
1663S.E. Eriksson et al. / Free Radical Biology & Medicine 47 (2009) 1661–1671

solution of 1.1 mM DTNB and 0.35 mM NADPH (prepared in stock
buffer) was subsequently added to all wells (170 μl/well), and after
5 min incubation at room temperature 40 μl yeast GR solution was
added (final concentration of GR was 1.2 U/ml). The change in
absorbance was followed for 2 min at 412 nm using the Versamax
plate reader and the concentration of GSH in the samples was
calculated using a standard curve generated from GSH solutions of
known concentrations (final concentrations of GSH were in the range
of 0.125–8μM). All GSH standard solutions were diluted in 10 mM HCl
containing 1% SSA, similar to the samples.
Drug sensitivity assays
Cells were cultured and transfected with siRNA as described above.
The cells were trypsinized and subsequently seeded together with the
indicated concentrations of drugs in 384-well plates 48 h after the
siRNA transfection. Depending on the drug used, viability and growth
of cells were examined either 24 or 48 h after incubation at 37°C in 5%
CO
2
. Each well was loaded with a cell suspension containing either
600 cells for the 24-h incubation or 300 cells for the 48-h time point.
Samples were always run in triplicate. After incubation, cells were
stained with a fluorescent dye mixture (Vital dye; Biomarker),
staining living cells green and the nuclei of dead cells red. Each well
of the 384-well plate was imaged using a custom-modified automated
microscope system, including a motorized Nikon Diaphot 200
fluorescence microscope (Nikon, Japan), a motorized XY table
(Märzhauser, Germany), an ORCA ER cold CCD camera with a detector
array 1344×1024 px (Hamamatsu, Japan), and an X-Cite, 120-W
mercury lamp for fluorescence illumination (EXFO, Canada). In this
study a Plan 2.5×/NA 0.08 Pol objective was used. Images were
captured using the program Platefocus_10, developed by two of the
authors (Flaberg and Szekely) in the visual programming language
environment of the Open Lab automator, as a modified version of the
original EFLCM method [58] with the addition of an autofocus
function. For the live and dead discrimination, two images were
captured for each well using either green or red fluorescence,
respectively. The number of red and green cells was then counted
using the Analyze Particle function in ImageJ software. A typical
original image with its resulting analysis is shown in Supplementary
Fig. S3.
Statistics
Values are presented as means ± standard error of the mean
(SEM) and represent at least two independent experiments.
Statistical evaluation of data from the cytotoxicity assays was
performed with the Mann–Whitney test using the GraphPad Prism
computer program, version 5.0 (GraphPad Software, San Diego, CA,
USA). Asterisks denote statistically significant differences: ⁎pb0.05,
⁎⁎pb0.01, and ⁎⁎⁎pb0.001.
Fig. 1. Selenoprotein expression pattern in A549 cells and confirmation of TrxR1 knockdown. (A)
75
Se incorporation experiment. Cells were seeded in the presence of
75
Se-labeled
selenite 15–18 h before siRNA treatment. Cells were harvested 48 h and up to 6 days after the transfection, with the resulting cell extracts separated by SDS–PAGE that was exposed
to autoradiography. The arrows indicate the 55-, 57-, and 130-kDa bands representing three forms of TrxR1. The plot on the right shows the correlation between the optical density
of the 55-kDa band and TrxR activity in the cellular lysate. In (B–D) anion chromatography analyses are shown. Equal protein amounts (300 μg) from (B) mock, (C) TrxR1 Siseq1, or
(D) TrxR1 Siseq2 cell lysates were analyzed on an anion column, with the salt gradient and protein elution profile as shown with all fractions analyzed for TrxR activity (gray bars).
The indicated fractions (asterisks) having the most significant TrxR activity were pooled, concentrated, and separated by SDS–PAGE. Coomassie-stained bands in the size range of
55 kDa from the sample treated with TrxR1 Siseq1 were subsequently subjected to tryptic digestion and MS analysis (for results from MS analysis see Supplementary Fig. S2).
1664 S.E. Eriksson et al. / Free Radical Biology & Medicine 47 (2009) 1661–1671

Results
Confirming knockdown of TrxR1
The aim of this study was to assess the cellular phenotype upon
siRNA-mediated knockdown of the selenium-dependent TrxR1 in the
highly TrxR1-overexpressing A549 lung carcinoma cells. First an Alexa
488-labeled AllStar negative siRNA (Qiagen) was used as a control for
transfection efficiency, and under the utilized transfection conditions
the efficiency was at least 95% (data not shown). Specific siRNA-
mediated knockdown of TrxR1 was first confirmed using
75
Se labeling
of all cellular selenoproteins, which vividly illustrated TrxR as the
Fig. 2. Analysis of TrxR, Trx, and GR activities. (A) Thioredoxin reductase, (B)
thioredoxin, or (C) glutathione reductase activities were measured in cell extracts
from A549 cells 72 h post-siRNA treatment, using the Trx-dependent insulin reduction
assay (A and B) or a direct GSSG reduction assay (C). Means ± SEM are shown and
represent results from at least three independent experiments.
Fig. 3. Cell cycle distribution after TrxR1 knockdown. Cell cycle phase distribution
analysis was performed 72 h after siRNA transfection. Cells were fixed and stained with
PI and the samples (10,000 cells per sample) were then analyzed by flow cytometry.
Results (means ± SEM) are presented as percentage of total cell number in each cell
cycle phase.
Fig. 4. GSH depletion in TrxR1 knockdown cells. (A) Cell growth was analyzed by
counting the total number of A549 cells per sample 72 h post-siRNA transfection with
or without BSO treatment. The initial number of cells at 0 h was inthe rangeof 30,000 to
32,000 in each treatment. Values represent means ± SEM of at least three independent
experiments. (B) Measurement of total glutathione concentrations in cells after BSO
treatment was done using the glutathione reductase–DTNB recycling assay. DTNB is
reduced by GSH to TNB
−
. The resulting GSSG can then be recycled into GSH by GR. The
measured rate of TNB
−
formation is thereby proportional to the sum of total cellular
glutathione present in the reaction and the amount was calculated using a standard
curve generated from GSH solutions of known concentrations. Before the assay was run,
all protein in the samples was precipitated by adding 5-sulfosalicylic acid. Samples and
standards were all measured in duplicate and values represent means ± SEM of at least
two independent experiments.
Table 1
TrxR activity after drug treatment in cells treated with TrxR1-targeting or mock siRNA
Drug
b
TrxR activity (%)
a
Conc. (μM) Mock Siseq1 (TrxR1)
Untreated −100 ± 0.5 7.6 ± 1.5
DNCB 5 63.1 ± 7.4 3.7 ± 0.2
DNCB 20 3.0 ± 1.4 4.5 ± 1.0
Menadione 5 95.8 ± 4.2 14.9 ± 0.5
Menadione 20 92.5 ± 0.2 7.7 ± 0.6
Auranofin 2.5 41.8 ± 5.0 5.3 ± 0.5
Auranofin 10 1.8 ± 0.2 2.3 ± 0.1
Cisplatin 15 55.8 ± 4.7 3.5 ± 0.9
Cisplatin 60 36.5 ± 1.0 2.4 ± 0.1
Oxaliplatin 15 93.2 ± 0.3 7.0 ± 2.8
Oxaliplatin 60 57.5 ± 3.0 3.6 ± 0.3
Juglone 1.25 96.0 ± 1.4 16.3 ± 1.5
Juglone 5 64.2 ± 8.6 5.5 ± 0.8
A549 cells seeded in the presence of 25 nM selenite were treated with various TrxR-
interacting drugs, added 48 h after siRNA transfection, whereupon total cellular TrxR
activity was determined using cell lysates in an insulin-coupled Trx reduction assay.
a
TrxR activity measured in untreated mock cells was set to 100%. Values are
means ± SEM of two independent experiments.
b
Cells treated with DNCB, menadione, and auranofin were incubated for 24 h,
whereas cisplatin, oxaliplatin, and juglone treatment lasted for 48 h. The same
conditions were used for the cell viability analysis shown in Figs. 5A and 5B.
1665S.E. Eriksson et al. / Free Radical Biology & Medicine 47 (2009) 1661–1671

most abundant selenoprotein in A549 cells and also that the method
for knockdown was efficient and long-lasting, up to several days (Fig.
1A). We also noted that, together with the very strong 55-kDa band
efficiently knocked down by the siRNA, corresponding to the major
TrxR1 form TXNRD1_v1 [59], another
75
Se-labeled band of approx
57 kDa and one weak band at approx 130 kDa were also knocked
down (Fig. 1A). The 57-kDa band could correspond to TXNRD1_v2
[59],whereasthe≈130-kDa band could possibly be a yet
unidentified form of TrxR1. The intensity of the 55-kDa band
correlated with the total cellular TrxR activity (Fig. 1A). Generally,
the most efficient knockdown with the lowest cellular TrxR activity
and weakest 55-kDa
75
Se-labeled band intensity was seen between
72 and 96 h after siRNA transfection. Cell extracts from control cells
seeded and grown in selenite-supplemented medium (25 nM) had a
total cellular TrxR activity of 271.4 ± 39.5 nmol min
−1
mg protein
−1
,
whereas TrxR activity in TrxR1 siRNA-treated cells was 31.4 ±
Fig. 5. Drug toxicity in relation to TrxR1 levels. (A and B) Upper four rows show the distribution between live and dead cells (dark gray, live cells, and light gray, dead cells) after the
corresponding treatments as indicated. Means ± SEM from at least two independent experiments done in triplicate are shown. The bottom rows show the percentage of dead cells
of the total cell number per well for mock-or Siseq1-treated cells. Significant differences in sensitivity are indicated: ⁎pb0.05, ⁎⁎pb0.01, or ⁎⁎⁎pb0.001.
1666 S.E. Eriksson et al. / Free Radical Biology & Medicine 47 (2009) 1661–1671

8.5 nmol min
−1
mg protein
−1
72 h posttransfection. Cell extracts from
cells grown without additional selenite in the medium had a cellular
TrxR activity in the range of 90–150 nmol min
−1
mg protein
−1
,
whereas TrxR1 knockdown cells in that case had a total TrxR activity
lowered to around 7.5–15 nmol min
−1
mg protein
−1
(data not
shown). Using both of the siRNA constructs in combination (a mixture
of either 5+ 5 nM or 10+ 10 nM Siseq1 and Siseq2) did not result in
additional down-regulation of TrxR1 compared to using each siRNA
separately (data not shown). We wondered whether the low albeit
remaining cellular TrxR activity in the knockdown cells could derive
from unaffected mitochondrial TrxR2 expression, which would also
give a
75
Se-labeled protein band of about 55 kDa. Thus, to determine
the identity of the remaining TrxR activity upon knockdown, 300 μgof
total cell extracts from either control or TrxR1-knockdown cells was
Fig. 5 (continued).
1667S.E. Eriksson et al. / Free Radical Biology & Medicine 47 (2009) 1661–1671

subjected to anion-exchange chromatography followed by MS analysis
(Figs. 1B, 1C, and 1D). This analysis identified a remaining presence of
low levels of TrxR1 in the knockdown cells (results shown in
Supplementary Fig. S2), whereas no TrxR2 could be detected in the
analyzed fractions under the conditions used. Thus, the low remaining
TrxR activity and
75
Se-labeled 55-kDa band in the siRNA-treated cells
(Fig. 1A) mainly represent a low remaining TrxR1 activity, although
the additional presence of low amounts of TrxR2 cannot be excluded.
Next we found that knocking down the cellular TrxR1 activity to about
10% of controls (Fig. 2A) gave no apparent effects on total cellular Trx
activity (Fig. 2B) or GR activity (Fig. 2C) as determined in crude cell
extracts. The cellular Trx activity, measured under saturating condi-
tions and thus representing the maximal Trx capacity, was in the
range of 39.1 ± 3.3 nmol min
−1
mg total protein
−1
.Notably,the
control A549 cells would thereby have a capacity for about sevenfold
higher TrxR activity than their maximal Trx activity (cf. control cells in
Figs. 2A and 2B). This analysis also showed that in the TrxR1-
knockdown cells, the cellular TrxR1 activity was lowered to the same
range as the total cellular Trx activity found in these cells.
Cell growth, cell cycle phase, and viability assessments
With the TrxR1 activity lowered to about 10% of controls, the
A549 cells showed no significant effects on cell growth compared to
mock, either with or without the addition of 25 nM selenite to the
growth medium (data not shown). However, the transfection
procedures as such caused a general decrease in cell proliferation
of approx 30%. Analysis of cell cycle phases also showed no
significant changes upon TrxR1 knockdown (Fig. 3). Furthermore,
the fraction of cells with a subdiploid DNA content (sub-G1),
identifying dead cells, was less than 4% in both mock and TrxR1-
knockdown cells (Fig. 3). Surprisingly, we found that even TrxR1-
knockdown cells subjected to GSH depletion for 48 h with BSO
showed no alterations in viability or cell growth compared to mock
controls (Fig. 4A), although it gave about 98% reduction in
intracellular glutathione levels (Fig. 4B). The results suggest a
tremendous reserve capacity for ribonucleotide reduction and
basal support of Trx-and GSH-dependent reduction pathways in
A549 cells. The mean concentration of total glutathione measured in
untreated control cells (165 ± 21 nmol per milligram of total
protein) was not significantly different from that in Siseq1-treated
cells (176 ± 9 nmol per milligram of total protein in). After BSO
treatment, the mean values of total glutathione were typically in the
range of 2.5 to 5 nmol per milligram of total protein (Fig. 4B).
Considering the lack of an apparent phenotype upon knocking
down TrxR1 in A549 cells, we next wished to study the potential
effects on the sensitivity of these cells to drugs, using a selection of
compounds known to target this enzyme.
Effects of TrxR1 knockdown on cellular drug sensitivity
For analyses of drug sensitivity, we treated the cells with DNCB,
which may irreversibly derivatize the active site of TrxR and
concomitantly induce an NADPH-oxidase activity in the inhibited
enzyme [38]. Furthermore, we treated cells with menadione, a
nondisulfide substrate of TrxR [32] that may redox cycle with TrxR,
leading to the production of pro-oxidant species, as similarly shown
with juglone [39], which, however, furthermore inhibits the enzyme
[39] and was also included in our analysis. We furthermore wished to
study the cytotoxic effects of auranofin in these cells as well as those
of the two platinum-based anticancer drugs cDDP and Oxa, which
have also been shown to target the enzyme [45–47]. We found that
DNCB and auranofin were obvious potent inhibitors of total cellular
TrxR activity, whereas juglone, cDDP, and Oxa had intermediary
effects and menadione did not inhibit the cellular activity of the
enzyme under the conditions used here and as analyzed in crude cell
extracts (Table 1). Analyzing the potential cell death induced by these
compounds revealed that TrxR1 knockdown significantly sensitized
A549 cells to treatment with DNCB (in the concentration range of 2.5–
10 μM) or menadione (5–10 μM), whereas there was no difference in
sensitivity to auranofin upon TrxR1 knockdown (Fig. 5A). In contrast,
the cells surprisingly became less susceptible to the cytotoxicity of
cDDP (usingb20 μM) upon TrxR1 knockdown, whereas there was no
change in the sensitivity to Oxa or juglone (Fig. 5B).
Discussion
In this study we found that knocking down the endogenous
overexpression of TrxR in a cancer cell line using siRNA leads to
diverse, complex, and unexpected cellular phenotypes. The basal
nonstressed growth of A549 cells was unaffected by an efficient
(approx 90%) TrxR1 knockdown, even in combination with GSH
depletion, illustrating that these cells have a tremendous reserve
capacity in vital redox systems. We also found that the very high
TrxR1 activity in A549 cells exceeded by severalfold the total Trx-
reducing capacity, suggesting that TrxR1 might carry out additional
functions beyond Trx reduction in these cells. Previous studies have
shown that TrxR1 can be highly overexpressed in several forms of
cancer and cancer cell lines [60] (www.proteinatlas.org) and many
studies have indicated that TrxR1 may be a potential target for
anticancer therapy and cancer prevention [20–22,61]. The Trx system
is clearly involved in many important cellular functions, e.g., cell
proliferation, through the disulfide reduction of ribonucleotide
reductase needed for deoxyribonucleotide synthesis [4], and antiox-
idant defense, supporting the activity of, e.g., methionine sulfoxide
reductase and peroxiredoxins [1]. The Trx system also has direct
antiapoptotic functions, e.g., with reduced Trx1 inhibiting apoptosis
signal-regulating kinase 1 [62] and with TrxR1 playing a potential role
in normal p53 maturation [63]. In the latter study it was shown that
modifying TrxR1 with electrophilic lipids resulted in changes in p53
conformation, whereas lowering the TrxR1 levels in the cells before
treatment with the electrophilic agents protected the p53 conforma-
tion [63]. It was also found that certain forms of TrxR1, such as the
enzyme derivatized by cDDP, could gain cell death-inducing proper-
ties in the form of SecTRAPs [48,49]. Based on these prior findings and
the many important endogenous functions of TrxR1 we were at first
surprised that we found only mild effects on the basal cell growth of
A549 cells upon TrxR1 knockdown. Because ribonucleotide reductase
function can be supported by either the Trx or the glutaredoxin (Grx)
system, the latter being dependent upon GSH [64], we initially
reasoned that the Grx system supported the DNA precursor synthesis
upon TrxR1 knockdown. However, depleting the GSH levels by BSO in
the TrxR1-knockdown cells did not cause any significant additional
impact on the cell growth. The answer is likely to be found in the total
cellular capacity of the Trx system in these cells. Knocking down
TrxR1 gave about 90% lower activity compared to the controls but,
importantly, the remaining 31.4 ± 8.5 nmol min
−1
mg protein
−1
of
cellular TrxR1 activity was in the same range as the total Trx capacity
and could thus be enough to sustain the normal ribonucleotide
synthesis and other Trx-dependent enzyme systems. Supposing a
duration of 8 h for S phase in these cells, 12.5 × 10
6
ribonucleotides
would have to be reduced per minute for the synthesis of a complete
genome of 6×10
9
deoxyribonucleotides. Assuming that a typical
A549 cell contains about 1 ng of total protein [60,65], the hereby
determined Trx activity of 39.1 nmol min
−1
mg protein
−1
could
support about 23.5×10
9
reactions per minute and per cell, i.e.,
significantly more than theoretically needed for Trx support of DNA
synthesis. This can explain the lack of effects from TrxR knockdown in
these cells and may explain an earlier study showing that down-
regulation of TrxR in cancer cells from mouse resulted in no changes
in deoxyribonucleotide pools [24]. With this in mind, it should be
emphasized that unless compartmentalization effects or other factors
1668 S.E. Eriksson et al. / Free Radical Biology & Medicine 47 (2009) 1661–1671

exist that increase the intracellular Trx activity beyond that of the
saturating conditions of our enzyme assay, the total TrxR1 capacity in
the A549 cells was about sevenfold higher than their total Trx
capacity. With the remaining 10% TrxR1 activity in the knockdown
cells therefore likely to be enough to sustain most if not all of the Trx1
functions, what other types of functions, if any, would the very high
levels of TrxR1 in A549 cells have? The reactive C-terminal catalytic
selenocysteine residue in TrxR1 clearly makes the enzyme capable of
reducing several other substrates in vitro in addition to Trx and
thereby TrxR could also have additional cellular substrates apart from
Trx [1]. It should also be noted that some splice forms, such as
TXNRD1_v3, lack Trx-reducing activity, which could complicate the
picture even further [59,66,67]. Hatfield and co-workers showed in a
mouse model that xenografts with TrxR1-knockdown cells could not
form tumors in contrast to control cells [23]. Such effects could hardly
be due to impaired replicative capacity, because the cells grew in
culture [23], but could potentially be due to knockdown of TrxR1-
dependent cell–cell interactions not detectable under normal cell
culture conditions. It should be noted that our siRNA constructs
covering either part of the active site of the enzyme (its CVNVGC
motif) or the 3′UTR exon (Supplementary Fig. S1) knocked down at
least three
75
Se-labeled forms of TrxR1 present in A549 cells, even if
the classical 55-kDa form was the most prominent one (Fig. 1). Future
studies with specific siRNA constructs for the various splice forms
would be needed to distinguish between these forms with regard to
any observed phenotypes.
Several previous studies have found a close correlation between
induction of cell death and TrxR1 inhibition by various types of
cytotoxic compounds [20,22,37]. However, our results clearly reveal
that impairment of TrxR1 activity solely by knocking down the
enzyme is not enough to induce cell death in highly overexpressing
A549 cells and, also, that even if TrxR1 obviously affects the extent of
certain drug-induced toxicities, such effects may be due to more
intricate mechanisms than just a lowered TrxR1 activity. The A549
cells with endogenous overexpression of TrxR1 had a markedly
increased sensitivity to the platinum anticancer drug cDDP compared
to TrxR1 knockdown, but they had higher resistance to DNCB and
menadione. The effects of other agents targeting TrxR, including
auranofin, considered to be a potent TrxR1-specific inhibitor, were
unrelated to the expression levels of TrxR1. The cytotoxic effects seen
upon the use of different TrxR inhibitors could, naturally, involve
different off-target effects, but potentially they could also involve
different direct effects on TrxR1. Our group has previously shown that
forms of TrxR1 derivatized with cDDP, or being truncated at the
position of the Sec residue, are completely devoid of Trx reduction
capacity but can show a gain of function in the form of pro-oxidant
cell-killing SecTRAPs [48,49]. Such mechanisms would agree with our
findings herein, showing that A549 cells with their endogenously up-
regulated levels of TrxR1 have a significantly increased sensitivity to
cDDP compared to TrxR1-knockdown cells. It is thus possible that in
A549 cells, where the TrxR1 capacity was found to exceed by far the
activity needed for Trx reduction, the selenoprotein may easily reside
in its reduced selenolate-exposing form [68] (because of less need for
reduction of substrates such as Trx) and thereby could more easily
form lethal SecTRAPs upon exposure to cDDP. If this is indeed the case,
it would suggest that cells having a TrxR1 activity higher than that
needed for Trx reduction would be particularly sensitive to cDDP
because they could easily form SecTRAPs. This concept is supported by
our findings but should be further addressed in future studies. In
contrast to the findings with cDDP, knockdown of TrxR1 in A549 cells
made these cells more sensitive to DNCB and menadione, two
compounds known to act as stressors capable of inducing ROS
production [1,69]. In these cases it seemed like the antioxidant
functions of highly overexpressed TrxR1 could be important for cell
protection against these compounds. DNCB is a very good inhibitor of
TrxR1 [38], whereas menadione is not an inhibitor but rather a pure
substrate [32]. Exposure to menadione has also been shown to result
in up-regulation of TrxR1 in a hepatic cancer cell line [69], probably
owing to an induced oxidative stress with a nuclear factor erythroid-
2-related factor 2-dependent response [70]. It should be noted that
although DNCB is a clear inhibitor of TrxR1 [38], as also shown in our
study (Table 1), it is furthermore a model substrate of glutathione S-
transferases and lowers GSH levels in cells [71]. However, our results
from the combination of TrxR1 knockdown with GSH depletion by
BSO treatment suggested that this combination as such is not
necessarily lethal to A549 cells. Thus, our results clearly demonstrate
that although TrxR1 targeting is important and promising as an
anticancer chemotherapeutic principle, the actual cytotoxic profile
and TrxR1-related mechanisms underlying drug efficacy need to be
highly drug-specific. In the protective effects of TrxR1 against DNCB-
or menadione-induced cell death, it should again be emphasized that
this protection is not necessarily linked to Trx-dependent pathways,
because TrxR1 activity also in the knockdown cells should be
sufficient to sustain the overall Trx capacity in these cells. It is in
this case possible that reduction of low-molecular-weight antiox-
idants or peroxides could explain the protective effects of the very
high TrxR1 levels [1]. To conclude, this study shows that high
overexpression of TrxR1 in cancer cells increases the cytotoxic efficacy
of drugs such as cisplatin, whereas it protects the cells from other
drugs such as DNCB or menadione. It seems clear that TrxR1 is an
interesting target for anticancer chemotherapy, but the molecular
mechanisms underlying such therapy may be more complex than
hitherto fully understood.
Acknowledgments
Determination of selenium content in the fetal calf serum was
kindly performed by Marie Vahter and collaborators, Karolinska
Institutet. Oxaliplatin used in the drug sensitivity assays was a kind
gift from Maria Shoshan, and human thioredoxin was kindly provided
by Arne Holmgren, both at Karolinska Institutet. The kind help of Qing
Cheng and Olle Rengby in the production of TrxR1 is also
acknowledged. This study was supported by funding from the
Swedish Research Council (Medicine), the Swedish Cancer Society,
Hedlunds Stiftelse, Knut and Alice Wallenbergs Stiftelse, and
Karolinska Institutet.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.freeradbiomed.2009.09.016.
References
[1] Nordberg, J.; Arnér, E. S. J. Reactive oxygen species, antioxidants, and the
mammalian thioredoxin system. Free Radic. Biol. Med. 31:1287–1312; 2001.
[2] Arner, E. S.; Holmgren, A. Physiological functions of thioredoxin and thioredoxin
reductase. Eur. J. Biochem. 267:6102–6109; 2000.
[3] Berndt, C.; Lillig, C. H.; Holmgren, A. Thiol-based mechanisms of the thioredoxin
and glutaredoxin systems: implications for diseases in the cardiovascular system.
Am. J. Physiol. Heart Circ. Physiol. 292:H1227–1236; 2007.
[4] Nordlund, P.; Reichard, P. Ribonucleotide reductases. Annu. Rev. Biochem. 75:
681–706; 2006.
[5] Luthman, M.; Holmgren, A. Rat liver thioredoxin and thioredoxin reductase:
purification and characterization. Biochemistry 21:6628–6633; 1982.
[6] Rundlöf, A. K.; Arnér, E. S. J. Regulation of the mammalian selenoprotein
thioredoxin reductase 1 in relation to cellular phenotype, growth, and signaling
events. Antioxid. Redox Signal. 6:41–52; 2004.
[7] Rigobello, M. P.; Callegaro, M. T.; Barzon, E.; Benetti, M.; Bindoli, A. Purification of
mitochondrial thioredoxin reductase and its involvement in the redox regulation
of membrane permeability. Free Radic. Biol. Med. 24:370–376; 1998.
[8] Sun, Q. A.; Su, D.; Novoselov, S. V.; Carlson, B. A.; Hatfield, D. L.; Gladyshev, V. N.
Reaction mechanism and regulation of mammalian thioredoxin/glutathione
reductase. Biochemistry 44:14528–14537; 2005.
[9] Jakupoglu, C.; Przemeck, G. K.; Schneider, M.; Moreno, S. G.; Mayr, N.;
Hatzopoulos, A. K.; de Angelis, M. H.; Wurst, W.; Bornkamm, G. W.; Brielmeier,
M.; Conrad, M. Cytoplasmic thioredoxin reductase is essential for embryogenesis
but dispensable for cardiac development. Mol. Cell. Biol. 25:1980–1988; 2005.
1669S.E. Eriksson et al. / Free Radical Biology & Medicine 47 (2009) 1661–1671

[10] Conrad, M.; Jakupoglu, C.; Moreno, S. G.; Lippl, S.; Banjac, A.; Schneider, M.; Beck,
H.; Hatzopoulos, A. K.; Just, U.; Sinowatz, F.; Schmahl, W.; Chien, K. R.; Wurst, W.;
Bornkamm, G. W.; Brielmeier, M. Essential role for mitochondrial thioredoxin
reductase in hematopoiesis, heart development, and heart function. Mol. Cell. Biol.
24:9414–9423; 2004.
[11] Nonn, L.; Williams, R. R.; Erickson, R. P.; Powis, G. The absence of mitochondrial
thioredoxin 2 causes massive apoptosis, exencephaly, and early embryonic
lethality in homozygous mice. Mol. Cell. Biol. 23:916–922; 2003.
[12] Matsui, M.; Oshima, M.; Oshima, H.; Takaku, K.; Maruyama, T.; Yodoi, J.; Taketo,
M. M. Early embryonic lethality caused by targeted disruption of the mouse
thioredoxin gene. Dev. Biol. 178:179–185; 1996.
[13] Zhong, L.; Holmgren, A. Essential role of selenium in the catalytic activities of
mammalian thioredoxin reductase revealed by characterization of recombinant
enzymes with selenocysteine mutations. J. Biol. Chem. 275:18121–18128; 2000.
[14] Gromer, S.; Eubel, J. K.; Lee, B. L.; Jacob, J. Human selenoproteins at a glance. Cell.
Mol. Life Sci. 62:2414–2437; 2005.
[15] Berggren, M. M.; Mangin, J. F.; Gasdaka, J. R.; Powis, G. Effect of selenium on rat
thioredoxin reductase activity: increase by supranutritional selenium and
decrease by selenium deficiency. Biochem. Pharmacol. 57:187–193; 1999.
[16] Nalvarte, I.; Damdimopoulos, A. E.; Nystom, C.; Nordman, T.; Miranda-Vizuete, A.;
Olsson, J. M.; Eriksson, L.; Bjornstedt, M.; Arner, E. S. J.; Spyrou, G. Overexpression
of enzymatically active human cytosolic and mitochondrial thioredoxin reductase
in HEK-293 cells—effect on cell growth and differentiation. J. Biol. Chem. 279:
54510–54517; 2004.
[17] Hill, K. E.; McCollum, G. W.; Boeglin, M. E.; Burk, R. F. Thioredoxin reductase
activity is decreased by selenium deficiency. Biochem. Biophys. Res. Commun. 234:
293–295; 1997.
[18] Crosley, L. K.; Meplan, C.; Nicol, F.; Rundlof, A. K.; Arner, E. S. J.; Hesketh, J. E.;
Arthur, J. R. Differential regulation of expression of cytosolic and mitochondrial
thioredoxin reductase in rat liver and kidney. Arch. Bio chem. Bioph ys. 459:178–188;
2007.
[19] Papp, L. V.; Lu, J.; Holmgren, A.; Khanna, K. K. From selenium to selenoproteins:
synthesis, identity, and their role in human health. Antioxid. Redox Signal. 9:
775–806; 2007.
[20] Urig, S.; Becker, K. On the potential of thioredoxin reductase inhibitors for cancer
therapy. Semin. Cancer Biol. 16:452–465; 2006.
[21] Gromer, S.; Urig, S.; Becker, K. The thioredoxin system—from science to clinic.
Med. Res. Rev. 24:40–89; 2004.
[22] Arner, E. S. J.; Holmgren, A. The thioredoxin system in cancer. Semin. Cancer Biol.
16:420–426; 2006.
[23] Yoo, M. H.; Xu, X. M.; Carlson, B. A.; Gladyshev, V. N.; Hatfield, D. L. Thioredoxin
reductase 1 deficiency reverses tumor phenotype and tumorigenicity of lung
carcinoma cells. J. Biol. Chem 281:13005–13008; 2006.
[24] Yoo, M. H.; Xu, X. M.; Carlson, B. A.; Patterson, A. D.; Gladyshev, V. N.; Hatfield, D. L.
Targeting thioredoxin reductase 1 reduction in cancer cells inhibits self-sufficient
growth and DNA replication. PLoS ONE e1112:2; 2007.
[25] Yoo, M. H.; Xu, X. M.; Turanov, A. A.; Carlson, B. A.; Gladyshev, V. N.; Hatfield, D. L.
A new strategy for assessing selenoprotein function: siRNA knockdown/knock-in
targeting the 3′-UTR. RNA 13:921–929; 2007.
[26] Nishimoto, M.; Sakaue, M.; Hara, S. Short-interfering RNA-mediated silencing of
thioredoxin reductase 1 alters the sensitivity of HeLa cells toward cadmium.
Biol. Pharm. Bull. 29:543–546; 2006.
[27] Zhong, L.; Arnér, E. S. J.; Holmgren, A. Structure and mechanism of mammalian
thioredoxin reductase: the active site is a redox-active selenolthiol/selenenyl-
sulfide formed from the conserved cysteine–selenocysteine sequence. Proc. Natl.
Acad. Sci. USA 97:5854–5859; 2000.
[28] Cheng, Q.; Sandalova, T.; Lindqvist, Y.; Arner, E. S. Crystal structure and catalysis of
the selenoprotein thioredoxin reductase 1. J. Biol. Chem. 284:3998–4008; 2009.
[29] Cenas, N.; Nivinskas, H.; Anusevicius, Z.; Sarlauskas, J.; Lederer, F.; Arnér, E. S. J.
Interactions of quinones with thioredoxin reductase: a challenge to the
antioxidant role of the mammalian selenoprotein. J. Biol. Chem. 279:2583–2592;
2004.
[30] May, J. M.; Mendiratta, S.; Hill, K. E.; Burk, R. F. Reduction of dehydroascorbate
to ascorbate by the selenoenzyme thioredoxin reductase. J. Biol. Chem. 272:
22607–22610; 1997.
[31] Arnér, E. S. J.; Nordberg, J.; Holmgren, A. Efficient reduction of lipoamide and lipoic
acid by mammalian thioredoxin reductase. Biochem. Biophys. Res. Commun. 225:
268–274; 1996.
[32] Holmgren, A. Thioredoxin. Annu. Rev. Biochem. 54:237–271; 1985.
[33] Björnstedt, M.; Kumar, S.; Björkhem, L.; Spyrou,G.; Holmgren, A. Selenium and the
thioredoxin and glutaredoxin systems. Biomed. Environ. Sci. 10:271–279; 1997.
[34] Lothrop, A. P.; Ruggles, E. L.; Hondal, R. J. No selenium required: reactions
catalyzed by mammalian thioredoxin reductase that are independent of a
selenocysteine residue. Biochemistry 48:6213–6223; 2009.
[35] Omata, Y.; Folan, M.; Shaw, M.; Messer, R. L.; Lockwood, P. E.; Hobbs, D.;
Bouillaguet, S.; Sano, H.; Lewis, J. B.; Wataha, J. C. Sublethal concentrations of
diverse gold compounds inhibit mammalian cytosolic thioredoxin reductase
(TrxR1). Toxicol. In Vitro 20:882–890; 2006.
[36] Gromer, S.; Wissing, J.; Behne, D.; Ashman, K.; Schirmer, R. H.; Flohe, L.; Becker,
K. A. hypothesis on the catalytic mechanism of the selenoenzyme thioredoxin
reductase. Biochem. J 332 (Pt 2):591–592; 1998.
[37] Talbot, S.; Nelson, R.; Self, W. T. Arsenic trioxide and auranofin inhibit
selenoprotein synthesis: implications for chemotherapy for acute promyelocytic
leukaemia. Br. J. Pharmacol. 154:940–948; 2008.
[38] Arnér, E. S. J.; Bjornstedt, M.; Holmgren, A. 1-Chloro-2,4-dinitrobenzene is an
irreversible inhibitor of human thioredoxin reductase: loss of thioredoxin
disulfide reductase activity is accompanied by a large increase in NADPH oxidase
activity. J. Biol. Chem. 270:3479–3482; 1995.
[39] Cenas, N.; Prast, S.; Nivinskas, H.; Sarlauskas, J.; Arner, E. S. Interactions of
nitroaromatic compounds with the mammalian selenoprotein thioredoxin reduc-
tase and the relation to induction of apoptosis in human cancer cells. J. Biol. Chem.
281:5593–5603; 2006.
[40] Brown, K. K.; Eriksson, S. E.; Arner, E. S. J.; Hampton, M. B. Mitochondrial
peroxiredoxin 3 is rapidly oxidized in cells treated with isothiocyanates. Free
Radic. Biol. Med. 45:494–502; 2008.
[41] Carvalho, C. M.; Chew, E. H.; Hashemy, S. I.; Lu, J.; Holmgren, A. Inhibition of the
human thioredoxin system: a molecular mechanism of mercury toxicity. J. Biol.
Chem. 283:11913–11923; 2008.
[42] Wallenborg, K.; Vlachos, P.; Eriksson, S.; Huijbregts, L.; Arner, E. S.; Joseph, B.;
Hermanson, O. Red wine triggers cell death and thioredoxin reductase inhibition:
effects beyond resveratrol and SIRT1. Exp. Cell Res. 315:1360–1371; 2009.
[43] Lu, J.; Papp, L. V.; Fang, J.; Rodriguez-Nieto, S.; Zhivotovsky, B.; Holmgren, A.
Inhibition of mammalian thioredoxin reductase by some flavonoids: implications
for myricetin and quercetin anticancer activity. Cancer Res. 66:4410–4418; 2006.
[44] Lu, J.; Chew, E. H.; Holmgren, A. Targeting thioredoxin reductase is a basis for
cancer therapy by arsenic trioxide. Proc. Natl. Acad. Sci. USA 104:12288–12293;
2007.
[45] Hellberg, V.; Wallin, I.; Eriksson, S.; Hernlund, E.; Jerremalm, E.; Berndtsson, M.;
Eksborg, S.; Arner, E. S.; Shoshan, M.; Ehrsson, H.; Laurell, G. Cisplatin and
oxaliplatin toxicity: importance of cochlear kinetics as a determinant for
ototoxicity. J. Natl. Cancer Inst. 101:37–47; 2009.
[46] Witte, A. B.; Anestal, K.; Jerremalm, E.; Ehrsson, H.; Arner, E. S. J. Inhibition of
thioredoxin reductase but not of glutathione reductase by the major classes of
alkylating and platinum-containing anticancer compounds. Free Radic. Biol. Med.
39:696–703; 2005.
[47] Arnér, E. S. J.; Nakamura, H.; Sasada, T.; Yodoi, J.; Holmgren, A.; Spyrou, G. Analysis
of the inhibition of mammalian thioredoxin, thioredoxin reductase, and
glutaredoxin by cis-diamminedichloroplatinum (II) and its major metabolite,
the glutathione–platinum complex. Free Radic. Biol. Med. 31:1170–1178; 2001.
[48] Anestal, K.; Arner, E. S. Rapid induction of cell death by selenium-compromised
thioredoxin reductase 1 but not by the fully active enzyme containing
selenocysteine. J. Biol. Chem. 278:15966–15972; 2003.
[49] Anestal, K.; Prast-Nielsen, S.; Cenas, N.; Arner, E. S. Cell death by SecTRAPs:
thioredoxin reductase as a prooxidant killer of cells. PLoS ONE e1846:3; 2008.
[50] Rengby, O.; Johansson, L.; Carlson, L. A.; Serini, E.; Vlamis-Gardikas, A.; Karsnas, P.;
Arnér, E. S. J. Assessment of production conditions for efficient use of Escherichia
coli in high-yield heterologous recombinant selenoprotein synthesis. Appl.
Environ. Microbiol. 70:5159–5167; 2004.
[51] Ren, X.; Björnstedt, M.; Shen, B.; Ericson, M. L.; Holmgren, A. Mutagenesis of
structural half-cystine residues in human thioredoxin and effects on the
regulation of activity by selenodiglutathione. Biochemistry 32:9701–9708; 1993.
[52] Zhong, L.; Arnér, E. S. J.; Ljung, J.; Åslund, F.; Holmgren, A. Rat and calf thioredoxin
reductase are homologous to glutathione reductase with a carboxyl-terminal
elongation containing a conserved catalytically active penultimate selenocysteine
residue. J. Biol. Chem. 273:8581–8591; 1998.
[53] Rengby, O.; Cheng, Q.; Vahter, M.; Jornvall, H.; Arner, E. S. Highly active dimeric
and low-activity tetrameric forms of selenium-containing rat thioredoxin
reductase 1. Free Radic. Biol. Med. 46:893–904; 2009.
[54] Arnér, E. S. J.; Zhong, L.; Holmgren, A. Preparation and assay of mammalian
thioredoxin and thioredoxin reductase. Methods Enzymol. 300:226–239; 1999.
[55] Carlberg, I.; Mannervik, B. Glutathione reductase. Methods Enzymol. 113:484–490;
1985.
[56] Tietze, F. Enzymic method for quantitative determination of nanogram amounts
of total and oxidized glutathione: applications to mammalian blood and other
tissues. Anal. Biochem. 27:502–522; 1969.
[57] Vandeputte, C.; Guizon, I.; Genestie-Denis, I.; Vannier, B.; Lorenzon, G. A
microtiter plate assay for total glutathione and glutathione disulfide contents in
cultured/isolated cells: performance study of a new miniaturized protocol. Cell
Biol. Toxicol. 10:415–421; 1994.
[58] Flaberg, E.; Sabelstrom, P.; Strandh, C.; Szekely, L. Extended field laser confocal
microscopy (EFLCM): combining automated gigapixel image capture with in silico
virtual microscopy. BMC Med. Imaging 8:13; 2008.
[59] Rundlöf, A. K.; Janard, M.; Miranda-Vizuete, A.; Arnér, E. S. J. Evidence for
intriguingly complex transcription of human thioredoxin reductase 1. Free Radic.
Biol. Med. 36:641–656; 2004.
[60] Berggren, M.; Gallegos, A.; Gasdaska, J. R.; Gasdaska, P. Y.; Warneke, J.; Powis, G.
Thioredoxin and thioredoxin reductase gene expression in human tumors and cell
lines, and the effects of serum stimulation and hypoxia. Anticancer Res. 16:
3459–3466; 1996.
[61] Yoo, M. H.; Xu, X. M.; Carlson, B. A.; Gladyshev, V. N.; Hatfield, D. L. Thioredoxin
reductase 1 deficiency reverses tumor phenotype and tumorigenicity of lung
carcinoma cells. J. Biol. Chem. 281:13005–13008; 2006.
[62] Saitoh, M.; Nishitoh, H.; Fujii, M.; Takeda, K.; Tobiume, K.; Sawada, Y.; Kawabata,
M.; Miyazono, K.; Ichijo, H. Mammalian thioredoxin is a direct inhibitor of
apoptosis signal-regulating kinase (ASK) 1. EMBO J. 17:2596–2606; 1998.
[63] Cassidy, P. B.; Edes, K.; Nelson, C. C.; Parsawar, K.; Fitzpatrick, F. A.; Moos, P. J.
Thioredoxin reductase is required for the inactivation of tumor suppressor p53
and for apoptosis induced by endogenous electrophiles. Carcinogenesis 27:
2538–2549; 2006.
[64] Avval, F. Z.; Holmgren, A. Molecular mechanisms of thioredoxin and glutaredoxin
as hydrogen donors for mammalian S phase ribonucleotide reductase. J. Biol.
Chem. 284:8233–8240; 2009.
1670 S.E. Eriksson et al. / Free Radical Biology & Medicine 47 (2009) 1661–1671

[65] Fujiwara, N.; Fujii, T.; Fujii, J.; Taniguchi, N. Functional expression of rat
thioredoxin reductase: selenocysteine insertion sequence element is essential
for the active enzyme. Biochem. J. 340 (Pt. 2):439–444; 1999.
[66] Dammeyer, P.; Damdimopoulos, A. E.; Nordman, T.; Jimenez, A.; Miranda-Vizuete,
A.; Arner, E. S. Induction of cell membrane protrusions by the N-terminal
glutaredoxin domain of a rare splice variant of human thioredoxin reductase 1.
J. Biol. Chem. 283:2814–2821; 2008.
[67] Su, D.; Gladyshev, V. N. Alternative splicing involving the thioredoxin reduc-
tase module in mammals: a glutaredoxin-containing thioredoxin reductase 1.
Biochemistry 43:12177–12188; 2004.
[68] Sandalova, T.; Zhong, L.; Lindqvist, Y.; Holmgren, A.; Schneider, G. Three-
dimensional structure of a mammalian thioredoxin reductase: implications for
mechanism and evolution of a selenocysteine-dependent enzyme. Proc. Natl.
Acad. Sci. USA 98:9533–9538; 2001.
[69] Jung, H. I.; Lim, H. W.; Kim, B. C.; Park, E. H.; Lim, C. J. Differential thioredoxin
reductase activity from human normal hepatic and hepatoma cell lines. Yonsei
Med. J. 45:263–272; 2004.
[70] Singh, A.; Boldin-Adamsky, S.; Thimmulappa, R. K.; Rath, S. K.; Ashush, H.; Coulter,
J.; Blackford, A.; Goodman, S. N.; Bunz, F.; Watson, W. H.; Gabrielson, E.; Feinstein,
E.; Biswal, S. RNAi-mediated silencing of nuclear factor erythroid-2-related factor
2 gene expression in non-small cell lung cancer inhibits tumor growth and
increases efficacy of chemotherapy. Cancer Res. 68:7975–7984; 2008.
[71] Armstrong, R. N. Structure, catalytic mechanism, and evolution of the glutathione
transferases. Chem. Res. Toxicol. 10:2–18; 1997.
1671S.E. Eriksson et al. / Free Radical Biology & Medicine 47 (2009) 1661–1671