Investigation of the selenium metabolism in cancer cell lines

Abstract

The aim of this work was to compare different selenium species for their ability to induce cell death in different cancer cell lines, while investigating the underlying chemistry by speciation analysis. A prostate cancer cell line (PC-3), a colon cancer cell line (HT-29) and a leukaemia cell line (Jurkat E6-1) were incubated with five selenium compounds representing inorganic as well as organic Se compounds in different oxidation states. Selenomethionine (SeMet), Se-methylselenocysteine (MeSeCys), methylseleninic acid (MeSeA), selenite and selenate in the concentration range 5-100 μM were incubated with cells for 24 h and the induction of cell death was measured using flow cytometry. The amounts of total selenium in cell medium, cell lysate and the insoluble fractions was determined by ICP-MS. Speciation analysis of cellular fractions was performed by reversed phase, anion exchange and size exclusion chromatography and ICP-MS detection. The selenium compounds exhibited large differences in their ability to induce cell death in the three cell lines and the susceptibilities of the cell lines were different. Full recovery of selenium in the cellular fractions was observed for all Se compounds except MeSeA. Speciation analysis showed that MeSeA was completely transformed during the incubations, while metabolic conversion of the other Se compounds was limited. Production of volatile dimethyl diselenide was observed for MeSeA and MeSeCys. MeSeA, MeSeCys and selenite showed noticeable protein binding. Correlations between cell death induction and the Se compounds transformations could not be demonstrated.

Full text

162 Metallomics, 2011, 3, 162–168 This journal is cThe Royal Society of Chemistry 2011
Cite this:
Metallomics
, 2011, 3, 162–168
Investigation of the selenium metabolism in cancer cell lines
Kristoffer Lunøe,*
a
Charlotte Gabel-Jensen,
a
Stefan Stu
¨rup,
a
Lars Andresen,
b
Søren Skov
b
and Bente Gammelgaard*
a
Received 24th August 2010, Accepted 29th November 2010
DOI: 10.1039/c0mt00091d
The aim of this work was to compare different selenium species for their ability to induce cell
death in different cancer cell lines, while investigating the underlying chemistry by speciation
analysis. A prostate cancer cell line (PC-3), a colon cancer cell line (HT-29) and a leukaemia
cell line (Jurkat E6-1) were incubated with five selenium compounds representing inorganic
as well as organic Se compounds in different oxidation states. Selenomethionine (SeMet),
Se-methylselenocysteine (MeSeCys), methylseleninic acid (MeSeA), selenite and selenate in the
concentration range 5–100 mM were incubated with cells for 24 h and the induction of cell death
was measured using flow cytometry. The amounts of total selenium in cell medium, cell lysate and
the insoluble fractions was determined by ICP-MS. Speciation analysis of cellular fractions was
performed by reversed phase, anion exchange and size exclusion chromatography and ICP-MS
detection. The selenium compounds exhibited large differences in their ability to induce cell death
in the three cell lines and the susceptibilities of the cell lines were different. Full recovery of
selenium in the cellular fractions was observed for all Se compounds except MeSeA. Speciation
analysis showed that MeSeA was completely transformed during the incubations, while metabolic
conversion of the other Se compounds was limited. Production of volatile dimethyl diselenide was
observed for MeSeA and MeSeCys. MeSeA, MeSeCys and selenite showed noticeable protein
binding. Correlations between cell death induction and the Se compounds transformations could
not be demonstrated.
Introduction
Selenium is essential to the proper functioning of the organism,
but toxic in high doses.
1,2
The tolerable upper intake level for
selenium is set at 400 mg/d which is less than ten times the
recommended dose.
2
This is a narrow therapeutic window and
potentially represents a problem when using dietary selenium
supplements. The upper limits for selenium intake are
based on total selenium alone, but the toxicity of selenium
compounds has been shown to vary greatly with the structure.
3
The toxicity may be exerted by the native selenium compound
or by metabolic products. Therefore elucidation of the meta-
bolism of nutritionally available selenium compounds is of
great interest. The state of the art in this field is combination of
efficient separation methods with the high sensitivity offered
by element specific ICP-MS detectors. Identification of new
species is commonly performed by the complementary use
of molecular MS analysis giving the structural information
inherently lost in LC-ICP-MS analysis. Reviews on selenium
speciation with focus on LC-ICP-MS,
4
selenium speciation of
biological samples,
5
the complementary use of element specific
and molecular mass spectrometry
6,7
and the significance of
finding new matabolites in selenometabolomics
8
have recently
been given.
The interest in metabolism of selenium increased signifi-
cantly when Clark et al.
9
published the results of the Nutritional
Prevention of Cancer trial (NPC), showing that selenium in
supra nutritional doses administered to male human test
subjects in form of selenized yeast exhibited cancer chemo-
preventive effects. The following SELECT-trial,
10
in which
SeMet was the selenium source, however, did not show the
same beneficial effect. This disparity may be explained by
the different selenium species applied
11
and their metabolic
products
12
and the need for speciation in the design of clinical
trials is increasingly being recognized.
13
The presently accepted common theory of the cancer
preventive effect assigns the metabolite methylselenol, MeSeH
a key role and selenium species like MeSeA and MeSeCys that
would be metabolised into this species would be the more
protective species.
14
An alternative theory is that the cancer
preventive effect is exerted in a prooxidative fashion as opposed
to the antioxidant status normally associated with selenium.
The mechanism of action would then be inhibition of relevant
a
Department of Pharmaceutics and Analytical Chemistry,
Faculty of Pharmaceutical Sciences, University of Copenhagen,
Universitetsparken 2, DK-2100 Copenhagen, Denmark
b
Department of Disease Biology, Faculty of Life Sciences,
University of Copenhagen, Denmark
Metallomics Dynamic Article Links
www.rsc.org/metallomics PAPER
Published on 16 December 2010. Downloaded by New Copenhagen University on 21/07/2016 12:49:37.
View Article Online
/ Journal Homepage
/ Table of Contents for this issue

This journal is cThe Royal Society of Chemistry 2011 Metallomics, 2011, 3, 162–168 163
proteins by oxidation and selenite and MeSeA would be the
more reactive species according to this theory.
15
The role of selenium in cell proliferation, tumour cell
invasion and apoptosis was recently reviewed.
16,17
This
research is based on molecular biology approaches. Insight
into the underlying selenium metabolism and chemistry in
such studies might synergistically enhance the knowledge of
the anti cancer effect of selenium.
Only few speciation studies investigating the metabolism of
well defined selenium compounds in cancer cells have been
reported. One study showed the transformation of MeSeA to
MeSeCys and the volatile species dimethylselenide, DMeSe
and dimethyldiselenide, DMeDSe in the human B-cell
lymphocyte cell line DHL-4 by elemental and molecular
MS.
18
Another study showed that MeSeCys protected liver
cancer cell line HepG2 from oxidative stress and LC-ICP-MS
analysis showed no transformation of the applied MeSeCys in
medium and cell homogenates.
19
The aim of this study was to apply speciation analysis
for investigation of the metabolism of different selenium
compounds in cancer cell lines and examine if the metabolic
patterns observed could be correlated to the cytotoxic
effects of the species. For this purpose, selenium compounds
representing inorganic as well as organic compounds and
different oxidation states were incubated with three cancer
cell lines representing different cancer forms.
Experimental
Apparatus
HPLC. All systems were combinations of modules from the
Agilent 1100 and 1200 series (Agilent, Waldbronn, Germany).
The general system setup comprised a degasser, a pump and
an autosampler. All fluid connections were made of PEEK
(Upchurch scientific, Oak Harbor, WA, USA) with an inner
diameter of 0.127 mm or less. Columns, mobile phases and
flow rates used in the specific setups are listed in Table 1.
ICP-MS. The ICP-MS instruments were an Elan DRC-e
(Perkin Elmer SCIEX, Norwalk, Connecticut, USA) fitted
with a jacketed cyclonic spray chamber equipped with a
MicroMist concentric nebulizer (AR30-1-FM02E) (all from Glass
Expansion, West Melbourne, Vic, Australia) and an Elan 6000
equipped with a PC
3
peltier cooled cyclonic spray chamber
(Elemental Scientific Inc., Omaha, NE, USA) fitted with a
micro concentric nebulizer (Glass Expansion). The DRC
parameters were: cell gas flow: 0.55 L min
1
and RPq: 0.30.
Methane was used as reaction gas. Nebulizer gas flow, lens
voltage and RF power were optimized daily on a 100 mgSeL
1
standard in eluent. The instruments were run using Elan
software version 3.3 (Perkin Elmer) and the chromatographic
data integrated and translated into Excel readable data using
TotalChrome 6.1 or WSearch32 (www.wsearch.com.au).
77
Se,
78
Se and
82
Se were monitored on the Elan 6000, while
78
Se,
80
Se and
82
Se were monitored on the DRCe. The dwell time
was 250 ms per isotope with one sweep per reading. The
number of readings was adjusted to meet the run time for
the HPLC analysis.
Reagents
All reagents were analytical grade. Methylseleninic acid
(MeSeA), sodium selenite, sodium selenate, ammonium
formate and formic acid were from Sigma-Aldrich (Steinheim,
Germany). Selenomethionine (SeMet) was from Scandilink
(Virum, Denmark). Methanol was from VWR (VWR-Bie &
Berntsen, Herlev, Denmark). NaOH and nitric acid were from
Merck (Darmstadt, Germany); the nitric acid was purified by
subboiling twice before use.
Cell lines
Adherent cells. Cells of human prostate carcinoma cell
line PC-3 and human colon adenocarcinoma cell line HT-29
(both from DSMZ, Braunschweig) were grown in DMEM
medium and, unless stated otherwise, 10% Foetal Bovine
Serum (FBS). The cells were incubated with selenium standards
at 5 mMor50mM for either 4 or 24 h. The medium was removed
and the cells were released from the surface by addition of
0.25% trypsin-0.38 g L
1
EDTA solution (Invitrogen, Taastrup,
Denmark), washed twice in PBS, lysed by addition of 10%
MeOH and centrifuged. The resulting pellet and supernatant
were separated. Samples of media to be analyzed by RP-HPLC
3 were diluted 1:1 with methanol. Samples were stored at
41Cuntilanalysis.
Cells in suspension. Acute T-cell leukemia cell line Jurkat
E6-1 (from American Type Culture Collection, LGC
Standards AB, Boras, Sweden) were grown at a cell density
of 1 10
6
cells mL
1
in RPMI-1640 with, unless stated
otherwise, 10% FBS was added. The cells were incubated with
the selenium compounds at 10 mMor50mM for periods of
4 or 24 h. The medium was removed and the cells washed in
PBS twice and following lysed by addition of 10% or 50%
MeOH and the suspension centrifuged. The resulting pellet
and supernatant were separated. The methanol from the lysing
of the cells was evaporated by nitrogen flux and the lysates
reconstituted in mobile phase. Samples were kept at 4 1C until
analysis. Samples of medium to be analyzed by RP-HPLC 3
were diluted 1 : 1 with methanol.
Table 1 Columns, eluents and flow rates of the LC systems
Separation type RP-LC 1 RP-LC 2 RP-LC 3 AE-LC SEC-LC
Column Gemini C18
250 2mm
Gemini C18
250 2mm
Gemini C18
50 1mm
AS-11-HC
250 2mm
BioSep-2000
300 4.6 mm
Eluent 0.1% formic acid
in 2%MeOH
20 mM ammonium
formiate in 2% MeOH
0.1% formic acid in
40% MeOH
25mM NaOH in
2% MeOH
20 mM ammonium
formiate in 2% MeOH
Flow rate 250 mL min
1
250 mL min
1
100 mL min
1
250 mL min
1
350 mL min
1
Inj. vol. 1–10 mL
Published on 16 December 2010. Downloaded by New Copenhagen University on 21/07/2016 12:49:37.
View Article Online

164 Metallomics, 2011, 3, 162–168 This journal is cThe Royal Society of Chemistry 2011
Toxicity measurement. Jurkat E6-1 cells, HT29 and PC3
cells were incubated with selenium compounds as indicated.
24 h later the cells were washed once in PBS (Sigma) and
stained with propidium iodide (PI) (Sigma) and Annexin
V-FITC (BD Biosciences) and the number of dead (PI-positive)
and apoptotic (Annexin V-FITC positive) cells was measured
using flow cytometry. Adherent cells were washed in PBS and
released from the culture plate using Trypsin. The released
cells together with cells from the first wash were collected in a
tube by centrifugation, resuspended in PBS and stained with
PI and Annexin V-FITC before flow cytometry analysis. Data
are presented as percent dead cells (PI-positive cells divided
with the total number of analysed cells).
Determination of total Se
FIA-ICPMS. The tubes containing media, lysates and
pellets were stored at 4 1C until analysis. To each tube con-
taining a pellet was added 50 mL 65% nitric acid, and the tube
was vortexed and left to stand a day at 4 1C. The acid digested
pellets were then diluted 100 times before analysis. The media
and lysates were sampled without further pretreatment.
Separate matrix matched standard curves were constructed
for the three sample types to correct for difference in the
matrices. Carrier fluid was mobile phase from either RP-LC 1
or 2 and the injection volume was 5 mL.
Infusion-ICP-MS. Pellet samples were added 100 mL con-
centrated nitric acid, vortex mixed and left to stand for a day.
The pellet samples were then diluted 100 times in water.
Samples of media and lysates were diluted to appropriate
concentrations and added nitric acid to obtain a final concen-
tration of 0.67% similar to the concentration of the pellet
samples. One point standard additions were performed together
with external calibration by a calibration curve of standards in
a 0.67% nitric acid matrix. The infusion rate was 700 mLmin
1
and 5 replicates of 25 readings were collected.
Results and discussion
Induction of cell death
Incubation of the three different cell lines resulted in remarkably
different mortalities of the cells. From the data in Table 2
it appears that the T-cell leukocyte Jurkat, the colon cancer
HT-29 and the prostate cancer PC-3 cell lines underwent
apoptosis and cell death to significantly different extent.
The HT-29 cell line seemed to be the more robust cell line of
the three. Selenite and MeSeA induced cell death in all cell
lines; in Jurkat their effects were similar, while selenite was
vastly more toxic in PC-3 and HT-29. SeMet induced death in
PC-3 cells at a level comparable with that observed for MeSeA
at all concentrations levels, but was devoid of toxicity in
HT-29 and only exhibited toxicity in Jurkat at the highest
concentration level. MeSeCys gave rise to a very slight increase
in cell death in PC-3 and HT-29, but only at the highest
concentration. Selenate did not produce any elevation in cell
death in any of the investigated cell lines.
Overall, Table 2 shows that most selenium compounds had
an impact on the cell lines. The two consistently toxic species
MeSeA and selenite are in the oxidation state +2 and +4,
respectively, while MeSeCys (oxidation state-2) and selenate
(+6) exhibited a general lack of toxicity. SeMet, which is in
the same oxidation state as MeSeCys (2), however, exhibited
cytotoxicity in the PC-3 cell line. This may indicate that
selenium in oxidation state +2 and +4 are in general more
toxic, maybe due to production of reactive oxygen species
(ROS) as describe by Hail et al.
20
The suggested cancer
protective effect of methylated Se compounds like MeSeA
and MeSeCys was not confirmed in these cell models as
MeSeCys appeared to be the least toxic species of the organic
compounds.
Total Se in cell fractions
The ICP-MS sample introduction technique depended on the
scale of the cell experiment. Large sample volumes were
analysed by infusion, while flow-injection was used for small
sample volumes. The results are presented in Table 3. As can
be seen from the Table 3, nearly total recovery was obtained
for four of the species after the incubations, while the recovery
for incubations with MeSeA stood out as being below 25%in
all cell lines. Selenite exhibited a lowered recovery only in
HT-29 but not in PC-3 or Jurkat. Total recovery was seen for
Table 2 Cell death (%) of three cell lines after incubation with
selenocompounds at three concentration levels. The numbers are
corrected for cell death measured in controls
mM Selenate Selenite MeSeA MeSeCys SeMet
50 1 0 0 0
HT-29 10 0 21 2 0 0
100 0 39 14 3 0
5 0 52 12 0 11
PC-3 10 0 78 16 0 17
100 1 80 26 6 30
5 0 22 29 0 0
Jurkat 10 0 46 50 0 0
100 0 72 76 1 15
Table 3 Selenium recovery in cell fractions from 24 h incubations
with selenium compounds
Recovery of added Se/%
Selenate Selenite MeSeA MeSeCys SeMet
PC-3 (prostate cancer) 5 mM24h
Medium 107 99 24 114 72
Lysate o1o1o1o119
Pellet o1o1o1o1o1
Total 107 99 24 114 91
HT-29 (colon cancer) 5 mM24h
Medium 68 6 92 97
Lysate 4 o1311
Pellet 3 o1o13
Total 75 695 111
Jurkat (leukemia) 10 mM24h
Medium 13 88
Lysate o113
Pellet o12
Total 13 103
Jurkat (leukemia) 50 mM4h
Medium 107 5 125 111
Lysate o1o1o1o1
Pellet o1o1o1o1
Total 107 5125 111
Published on 16 December 2010. Downloaded by New Copenhagen University on 21/07/2016 12:49:37.
View Article Online

This journal is cThe Royal Society of Chemistry 2011 Metallomics, 2011, 3, 162–168 165
MeSeCys and selenate. The low recovery of MeSeA is due to
escaped volatiles. The formation of DMeSe and DMeDSe
from MeSeA has been demonstrated in several previous
reports.
18,21,22
The presence of volatile selenium species
implies that recoveries could be overestimated due to induced
sensitivity from vapour enhancement effect exhibited by the
volatiles as demonstrated by Juresa et al.
23
However, the
extensive sample handling leaves only trace amounts of
volatiles and these were not considered to influence the results.
The Se content of the individual cell fractions revealed that
only Se from SeMet was found in the cell cytosol in all three
cell lines. It is not surprising that SeMet is taken up in the cells
as SeMet is indistinguishable from methionine and shares the
transport mechanism of methionine. More surprising is the
lack of-or very limited-uptake when incubating the cells with
the other species. The HT-29 cell line showed the largest
amounts of selenium in pellets and lysates. Significant
amounts were observed in the cell lysates of cells incubated
with selenite, MeSeCys and SeMet. The amounts found in the
pellets were only noticeable for selenite and SeMet.
Comparison of the cell death induction and the total
selenium content did not reveal any obvious correlations.
Although selenite was the most toxic selenium compound in
the tested cell lines, it did not enter the cells in noticeable
amounts while SeMet, which was only mildly toxic, did enter
the cell cytosol.
Speciation analysis
Medium and lysate samples from the incubations with organic
compounds were analysed by reversed phase chromatography
LC-ICP-MS. Examples of chromatograms are shown in Fig. 1–3.
Some samples were analysed for volatiles (Fig. 4) and some
samples were analysed by size exclusion chromatography
(SEC) to demonstrate protein interaction (Fig. 5). Fig. 5 also
shows analysis of aqueous standards on the SEC column,
illustrating the overlapping peaks of the low molecular
weight selenium compounds. Samples from incubation with
the inorganic compounds were mainly analysed by anion
exchange LC-ICP-MS (Fig. 6).
MeSeA (+2). Speciation analysis of the medium and lysate
samples (Fig. 1 and 2) by the reversed phase chromatography
revealed only a tiny MeSeA signal. When foetal bovine serum
(FBS) was present in the medium during incubation, no signal
was observed. When MeSeA was incubated without cells for
control, MeSeA remained in the sample as can be seen in
Fig. 3. Very small unidentified peaks designated Uwere
observed in samples as well as controls, and must therefore
stem from reactions with constituents of the cell medium.
Column recovery for the medium sample analyzed by
RP-HPLC 1 it was only about 4%, indicating that the
selenium still present in the samples was not eluting from the
column. When the medium samples following were analyzed
by RP-HPLC 3 using an eluent containing 40% MeOH, a
more lipophilic compound (Fig. 4) was eluted. This was
identified as DMeDSe by coelution with a spiked standard.
Production of DMeDSe has been suggested as an indication of
production of methylselenol (MeSeH).
22
The production of
volatile species also explains the low recovery for this species.
Analysis by SEC-ICP-MS showed protein bound selenium
(Fig. 5). The column recovery for this method was around
50%. The presence of a lipophilic volatile selenium species
together with protein-bound selenium explains the very low
column recovery for the RP-HPLC 1 method. Incubations in
medium containing FBS without presence of cells showed that
Se was bound to the serum proteins leading to a diminished
peak for free MeSeA. Another interesting result obtained from
the speciation analysis of the MeSeA samples was the presence
of a peak coeluting with MeSeCys. The presence of MeSeCys
has previously been seen when incubating hepatocytes or
lymphoma cell line DHL-4 with MeSeA.
18,24
No Se signal
was obtained in lysates, which was expected from the lack of
Se in the determination of total Se in this fraction.
In summary, MeSeA exhibited a high degree of reactivity,
protein binding and significant toxicity. These observed
properties support that MeSeA is readily transformed to the
cancer protective MeSeH. The elution of a selenium compound
in the protein fraction in SEC from incubations without
presence of cells indicates that MeSeA interacted directly with
serum proteins, probably via the protein thiol groups.
SeMet (2). SeMet exhibited toxicity only in PC-3 cells but
the compound entered the cell cytosol in all cell lines. Analysis
by RP-HPLC 1 revealed that the main peaks were SeMet itself
and an early eluting species ascribed to the oxidation of SeMet
to selenomethionine-oxide (SeOMet) (Fig. 1 and 2). This
assignment was based on previous reports,
25,26
the presence
of the peak in control samples and its growth when submitting
samples to freeze-thaw cycles (data not shown). The peak
ascribed to SeOMet was markedly larger in control samples
where SeMet had been incubated with medium alone (data not
shown). This may indicate that the presence of cells prevented
the oxidation that would otherwise take place. The minor peak
eluting just before the peak ascribed to SeOMet in Fig. 1 and 2
was not identified but was also seen in control samples without
cells present. The column recovery was around 70% for the
Fig. 1 Chromatograms of media samples from incubation of Jurkat
cells with selenite, MeSeA, MeSeCys and SeMet. The standard
contained MeSeA, MeSeCys and SeMet at a concentration of 1 mM
in mobile phase. System: RP-LC 2 (Table 1). Injection volume: 5 mL.
Published on 16 December 2010. Downloaded by New Copenhagen University on 21/07/2016 12:49:37.
View Article Online

166 Metallomics, 2011, 3, 162–168 This journal is cThe Royal Society of Chemistry 2011
medium samples and SEC analysis of the samples showed no
appreciable protein binding (Fig. 5). Speciation results from
incubations with Jurkat and PC-3 were quite similar and did
not reveal an explanation for the different toxicity of SeMet in
these cell lines. Analysis of the lysates produced the same
overall results as the medium analyses; SeMet was the primary
species found, while two minor peaks eluted in or just after
the void volume (Fig. 2).
In summary, SeMet deviated from the other compounds by
entering the cell cytosol in large amounts. The species did
not react with proteins and was only changed into the oxidized
form, which could have a function or may only be an
analytical artefact.
MeSeCys (2). The test for induction of apoptosis revealed
that MeSeCys was almost non-toxic in all three cell lines
and speciation by RP-HPLC 1 and 2 revealed no apparent
transformation of MeSeCys (Fig. 1). The column recovery for
analysis of media samples from PC-3 incubations by HPLC 1
Fig. 2 Chromatograms of lysate samples from incubation of Jurkat
cells with selenite, MeSeA, MeSeCys and SeMet. The standard
contained MeSeA, MeSeCys and SeMet at a concentration of 1 mM
in mobile phase. System: RP-LC 2 (Table 1). Injection volume: 5 mL.
Fig. 3 Chromatogram of a medium sample (a) from incubation
of Jurkat cells with 50 mM MeSeA for 4h and a control sample
(b) without presence of cells. System: RP-LC 1. Injection volume: 5 mL.
Fig. 4 Chromatograms of media samples from incubation of PC-3
cells with selenite, MeSeA and SeMet in concentration of 50 mM. The
bottom chromatogram represents a standard containing DMeDSe and
DMeSe in a concentration of 2 mM. System used was RP-LC 3.
Injection volume: 1 mL for samples and 5 mL for the standard.
Fig. 5 Chromatograms from size exclusion analysis (SEC-LC):
(a) Stacked chromatograms of aqueous selenium standards in con-
centrations of 5 mM. (b) Stacked chromatograms from analysis of
media from incubation of PC-3 cells with MeSeA, MeSeCys, SeMet,
selenate and selenite. The interval marked with the dashed vertical
lines is where the protein bound selenium elutes. System: SEC-LC.
Injection volume: 1 mL.
Published on 16 December 2010. Downloaded by New Copenhagen University on 21/07/2016 12:49:37.
View Article Online

This journal is cThe Royal Society of Chemistry 2011 Metallomics, 2011, 3, 162–168 167
was around 66%. SEC analysis revealed protein binding
(Fig. 5), comparable to the level of MeSeA protein binding,
indicating that protein binding is not directly linked to toxic
response, as MeSeA induced cell death while MeSeCys was
almost devoid of toxicity.
Presence of DMeDSe was demonstrated in medium samples
from MeSeCys incubation in Jurkat cells without FBS added.
The presence of volatile selenium species should lead to loss
of selenium from the samples leading to lowered recovery,
but repeated determinations of total selenium showed total
recovery. This means that either the conversion to volatiles
was very limited or the volatile selenium still present in the
medium sample was able to make up for the selenium lost due
to its higher sensitivity, as mentioned earlier. The presence of
DMeDSe indicate production of MeSeH, which would be
expected to produce a toxic response. This was not the case,
but may be due to the limited production being lower than the
level necessary for a toxic response.
Selenite (+4). Samples from cells incubated with selenite
were analyzed by the anion exchange method. Examples of
chromatograms are shown in Fig. 6. It appears that selenite
did not undergo any significant metabolic conversion. The
main peak observed was selenite and the tiny peak for selenate
originated from impurities in the selenite standard. As also
indicated in Fig. 6, the column recovery for media samples
containing FBS from cell incubations was significantly lower
than for an aqueous standard. This low column recovery was
explained by the SEC results displayed in Fig. 5. It appears
that selenite was bound to proteins to a significant extent.
Volatile selenium species were not detected in samples from
incubation of selenite with PC-3 cells. Analysis of the lysates
revealed only selenite itself; thus selenite was entering the cells
in trace amounts.
In summary, the results of the speciation analysis could
not explain the cytotoxic effects of selenite. The toxicity of
selenite may be caused by binding to other essential cellular
components as indicated by the protein binding.
Selenate (+6). Selenate showed total recovery and no
toxicity, indicating that this species does not interact with
biological systems. Speciation by AE-LC supported this
assumption as only selenate was observed (Fig. 6). The column
recovery was quite high and did not change if FBS was
removed from the incubation medium. Thus, there was no
indication of protein binding of selenate, which was confirmed
by SEC analysis of the medium (Fig. 5).
According to this work, selenate is the least interesting
species as a possible cancer preventive agent, as the species
exhibited a total lack of conversion in the three cell lines.
In summary, the five selenium compounds showed differ-
ences in release of volatiles, uptake in cell cytosol and protein
binding. However, no general relations between these
observations and the differences in cell death induction
were deduced. Production of volatile selenium species and
protein binding was for toxic and nontoxic compounds alike,
and no metabolite was identified as important for the toxic
response. Furthermore protein binding was not necessary
for cell death induction as SeMet showed toxicity in PC-3
cells but did not bind to serum proteins as MeSeA and
selenite did.
It was demonstrated that the induction of cell death was
different in the three different cell lines for selenite, MeSeA and
SeMet and that some cell lines were more susceptible than
others to the individual compounds. The very low recovery of
selenium after incubations with MeSeA was experienced
across the three cell lines and the production of volatile species
illustrates the reactive nature of MeSeA. As MeSeA was less
toxic to the cells than selenite, the prevalent metabolic
conversion of MeSeA does not explain the toxicity of MeSeA
or possibly indicate that MeSeA is toxic through another
pathway than selenite. The toxic response of four cell lines
was recently studied by Suzuki et al.
27
Cells were treated
with selenite, MeSeCys and SeMet in concentrations between
1 and 1000 mM and it was shown that apoptosis was
effectuated by two different pathways, one path for selenite
and another for MeSeCys and SeMet. Interestingly, toxicity of
MeSeCys was observed in three out of four cell lines at
concentration levels comparable to what was used in the
present study. This is not in accordance with our results as
none of the cell lines tested in this work showed susceptibility
to this compound. This discrepancy underlines how the choice
of cell line is crucial to the result. The toxicity of a given
selenium compound in a particular cell line may rest on the
presence of mandatory factors for the metabolism pathway
through which the selenium compound cause toxicity. The
lack of toxicity of MeSeCys in this study may thus be
explained by lacking of the b-lyase enzyme necessary for
converting MeSeCys to methylselenol, an enzyme normally
found in the organism.
27
Another problem arising in inter-
pretation of cell line experiments is that the Se compounds in
an in vivo system would be presented to the cancer cells
in a different form. When selenium is administered orally,
the passage through the gastro intestinal tract, absorption and
first pass metabolism, all have the potential to alter the
administered species.
24,28,29
In conclusion, no relation between cell death induction and
the result of the speciation analysis could be deduced.
Fig. 6 Chromatograms from analysis of media from incubation of
PC-3 cells with 5 mM selenate and selenite along with an aqueous
standard mixture of the two species at 5 mM.w: The column recovery
was determined for separate aqueous 5 mM standards of selenite and
selenate. System: AE-LC. Injection volume: 1 mL.
Published on 16 December 2010. Downloaded by New Copenhagen University on 21/07/2016 12:49:37.
View Article Online

168 Metallomics, 2011, 3, 162–168 This journal is cThe Royal Society of Chemistry 2011
Acknowledgements
This project was supported by a grant (271-07-0302) from the
Danish Research Council.
References
1 P. D. Whanger, Br. J. Nutr., 2004, 91, 11.
2 M. P. Rayman, Br. J. Nutr., 2008, 100, 254.
3 C. Ip, J. Nutr., 1998, 128, 1845.
4 C. B’Hymer and J. A. Caruso, J. Chromatogr., A, 2006, 1114,1.
5 A. Polatajko, N. Jakubowski and J. Szpunar, J. Anal. At.
Spectrom., 2006, 21, 639.
6 E. Dumont, F. Vanhaecke and R. Cornelis, Anal. Bioanal. Chem.,
2006, 385, 1304.
7 B. Gammelgaard, C. Gabel-Jensen, S. Stu
¨rup and H. R. Hansen,
Anal. Bioanal. Chem., 2008, 390, 1691.
8 Y. Ogra and Y. Anan, J. Anal. At. Spectrom., 2009, 24, 1477.
9 L. C. Clark, G. F. Combs, B. W. Turnbull, E. H. Slate, D. K. Chalker,
J.Chow,L.S.Davis,R.A.Glover,G.F.Graham,E.G.Gross,
A. Krongrad, J. L. Lesher, H. K. Park, B. B. Sanders, C. L. Smith and
J. R. Taylor, JAMA,J.Am.Med.Assoc., 1996, 276, 1957.
10 S. M. Lippman, E. A. Klein, P. J. Goodman, M. S. Lucia,
I. M. Thompson, L. G. Ford, H. L. Parnes, L. M. Minasian,
J. M. Gaziano, J. A. Hartline, J. K. Parsons, J. D. Bearden,
E. D. Crawford, G. E. Goodman, J. Claudio, E. Winquist,
E. D. Cook, D. D. Karp, P. Walther, M. M. Lieber,
A. R. Kristal, A. K. Darke, K. B. Arnold, P. A. Ganz,
R. M. Santella, D. Albanes, P. R. Taylor, J. L. Probstfield,
T. J. Jagpal, J. J. Crowley, F. L. Meyskens, L. H. Baker,
C. A. Coltman and P. C. Walsh, J. Urol., 2009, 181, 1686.
11 K. El-Bayoumy, Nutrition and Cancer, 2009, 61, 285.
12 M. I. Jackson and G. F. Combs, Curr. Opin. Clin. Nutr. Metab.
Care, 2008, 11, 718.
13 R. Hurst, C. N. Armah, J. R. Dainty, D. J. Hart, B. Teucher,
A. J. Goldson, M. R. Broadley, A. K. Motley and
S. J. Fairweather-Tait, Am. J. Clin. Nutr., 2010, 91, 923.
14 C. Ip, H. J. Thompson, Z. Zhu and H. E. Ganther, Cancer
Research, 2000, 60, 2882.
15 E. N. Drake, Med. Hypotheses, 2006, 67, 318.
16 H. Zeng and G. F. Combs, J. Nutr. Biochem., 2008, 19,1.
17 H. Zeng, Molecules, 2009, 14, 1263.
18 H. Infante, S. P. Joel, E. Warburton, C. Hopley, R. Hearn and
S. Ju
¨liger, J. Anal. At. Spectrom., 2007, 22, 888.
19 S. Cuello, S. Ramos, R. Mateos, M. A. Martin, Y. Madrid,
C. Camara, L. Bravo and L. Goya, Anal. Bioanal. Chem., 2007,
389, 2167.
20 N. Hail, M. Cortes, E. N. Drake and J. E. Spallholz, Free Radical
Biol. Med., 2008, 45, 97.
21 C. Gabel-Jensen, S. Bak, F. Lauritsen, H. Hansen, L. Badolo and
B. Gammelgaard, J. Anal. At. Spectrom., 2009, 24,
949.
22 C. Gabel-Jensen, K. Lunoe and B. Gammelgaard, Metallomics,
2010, 2, 167.
23 D. Juresa, D. Kuehnelt and K. A. Francesconi, Anal. Chem., 2006,
78, 8569.
24 C. Gabel-Jensen, J. Odgaard, C. Skonberg, L. Badolo and
B. Gammelgaard, J. Anal. At. Spectrom., 2009, 24, 69.
25 P. O. Amoako, P. C. Uden and J. F. Tyson, Anal. Chim. Acta,
2009, 652, 315.
26 B. Gammelgaard, C. Cornett, J. Olsen, L. Bendahl and
S. H. Hansen, Talanta, 2003, 59, 1165.
27 Y. Suzuki, Y. Hashiura, K. Matsumura, T. Matsukawa,
A. Shinohara and N. Furuta, Metallomics, 2010, 2, 126.
28 C. Gabel-Jensen, K. Lunoe, K. G. Madsen, J. Bendix, C. Cornett,
S. Sturup, H. R. Hansen and B. Gammelgaard, J. Anal. At.
Spectrom., 2008, 23, 727.
29 C. Gabel-Jensen, B. Gammelgaard, L. Bendahl, S. Sturup and
O. Jons, Anal. Bioanal. Chem., 2006, 384, 697.
Published on 16 December 2010. Downloaded by New Copenhagen University on 21/07/2016 12:49:37.
View Article Online