Integrated biometal science
Hollmann et al.
Detection of methylbismuth species
by GC/EI-MS/ICP-MS as evidence for
bismuth methylation in hepatic cells
Bertini and Cavallaro
Bioinformatics in bioinorganic
Volume 2 | Number 1 | 2010 Metallomics
www.rsc.org/metallomics Volume 2 | Number 1 | January 2010 | Pages 1–92
Parallel on-line detection of a methylbismuth species by hyphenated
GC/EI-MS/ICP-MS technique as evidence for bismuth methylation
by human hepatic cellsw
Alfred Vitalis Hirner
Received 22nd June 2009, Accepted 16th September 2009
First published as an Advance Article on the web 6th October 2009
Methylation of metal(loid)s by bacteria or even mammals is a well known process that can lead
to increased toxicity for humans. Nevertheless, reliable analytical techniques and tools are
indispensable in speciation analysis of trace elements, especially since environmental or biological
samples are usually characterised by complex matrices. Here the methylating capability of hepatic
cells was observed in vitro. HepG2 cells were incubated with colloidal bismuth subcitrate, bismuth
cysteine and bismuth glutathione, respectively for a period of 24 h. For identiﬁcation the cell
lysate was ethylated by sodium tetraethyl borate under neutral conditions. After cryo focussing by
purge and trap, the bismuth speciation was carried out via GC/EI-MS/ICP-MS. Colloidal
bismuth subcitrate and bismuth cysteine were methylated by HepG2 cells, while no methylated
bismuth species was detected after incubation with bismuth glutathione.
Inorganic bismuth compounds are known to show low
so bismuth is commonly used as a lead substitute,
e.g. in paintings or alloys.
Furthermore, it has been used in
medicine for many decades as an anti-gastritic and anti-ulcer
In contrast, organic bismuth compounds like the
permethylated trimethylbismuth (TMBi) may cause encephalo-
and are proven to be highly toxic as shown
in experiments with cats and dogs.
After application of
inorganic bismuth compounds many cases of bismuth poisoning
are well documented, e.g. epidemic-like encephalopathic
diseases in France and Australia.
It is commonly assumed
that the observed symptoms were caused by conversion of
inorganic bismuth to TMBi.
This can be caused by intestinal
microbes as shown by Michalke et al.
Boertz et al.
determinated TMBi in the human body after ingestion of
colloidal bismuth subcitrate.
Von Recklinghausen et al.
observed the uptake of bismuth glutathione and bismuth citrate
by human erythrocytes, lymphocytes and hepatocytes.
authors assumed that biomethylation of inorganic bismuth
compounds occurs in the human intestinum, however, the
methylation of bismuth in the liver is also imaginable. As
Styblo et al. have shown for arsenic, the lighter homologue of
bismuth, hepatomic cells (HepG2) can methylate inorganic
It is already known that heavy metals in living organisms
form complexes with sulfur containing molecules like cysteine
Especially the formation of methylmercury
cysteine in ﬁsh
as a biologically active substance indicates the
importance of small ‘‘biomolecules’’ for transport processes in
animals. The formation of bismuth cysteine and bismuth
glutathione complexes is also possible as shown by Burford
Since alkylated bismuth compounds tend to decompose in
water, the importance of bismuth cysteine complexes in the
ﬁeld of bismuth methylation is emphasized by a study proving
the existence of methylbismuth cysteine in aqueous solution.
Speciation analysis of bismuth compounds is usually carried
out by liquid chromatography coupled to mass spectrometry
as well as LT-GC/ICP-MS.
For quantiﬁcation of
bismuth in alloys, stable volatile ethyl derivatives are used
for quantitative determination of bismuth by ICP-AES.
very same technique has been used for volatilization and
determination of methylated bismuth species in human blood
samples after ingestion of bismuth(
In the present qualitative study, the transformation of
inorganic bismuth to methylated bismuth species by human
hepatic cells (HepG2) is investigated by derivatization of these
species with sodium tetraethyl borate. Then the volatile
bismuth species are detected simultaneously by EI-MS and
ICP-MS after gas chromatographic separation, revealing all
bismuth containing compounds directly through correlation
of both detector signals. This unique analytical system has
Institute of Environmental Analytical Chemistry, University of
Duisburg-Essen, Universitaetsstrasse 3-5, 45141 Essen, Germany.
E-mail: email@example.com; Fax: +49 (0)201 1833951;
Tel: +49 (0)201 1833238
Institute of Hygiene and Occupational Medicine, University of
Duisburg-Essen, Hufelandstrasse 55, 45122 Essen, Germany.
E-mail: firstname.lastname@example.org; Fax: +49 (0)201 7234546;
Tel: + 49( 0)201 7234578
w Electronic supplementary information (ESI) available: Fig. S1–S5,
EI-MS spectra. See DOI: 10.1039/b911945k
z New correspondence address: European Commission, Joint
Research Centre, Institute for Reference Materials and Measurements
(IRMM), Reference Materials Unit, Retieseweg 111, B-2440 Geel,
Belgium. Fax: +32 (0)14571548; Tel: +49 (0)14573005; E-mail:
52 | Metallomics, 2010, 2, 52–56 This journal is
The Royal Society of Chemistry 2010
PAPER www.rsc.org/metallomics | Metallomics
recently proven its usefulness in the determination of volatile
arsenic species generated by the (intestinal) microﬂora in
2.1 Reagents and standards
All reagents used were of analytical grade or better and are
either purchased from Sigma Aldrich (Buchs, Switzerland)
or from Alfa Aesar (Karlsruhe, Germany). Milli-Q water
(Millipore, Milford, MA, USA) was used for the preparation
of bismuth glutathione and bismuth cysteine. Colloidal bismuth
subcitrate (CBS) was prepared according to Asato et al.
bismuth complexes of glutathione and cysteine were prepared
according to Burford et al.
In brief: bismuth cysteine
and bismuth glutathione were synthesized by adding solid
III) chloride (BiCl
) to a saturated solution of
L-cysteine or glutathione, in ultra-pure water in a molar bismuth
to ligand ratio of 2 : 1. The mixture was stirred at room
temperature (20 1C) under an ultra pure argon atmosphere.
In a further step we isolated both complexes by adding small
amounts of methanol until precipitation of a (slightly) yellow
solid occurred. For isolation the crystalline product was
subsequently ﬁltered through a ﬁbreglass ﬁlter. Finally the
solid was dried in a vacuum desiccator containing silica gel for
Colloidal bismuth subcitrate was prepared as follows: to a
solution of bismuth citrate in aqueous ammonia (25%) and
Dulbecco’s phosphate buﬀered saline (D-PBS, Gibco
Invitrogent, Karlsruhe, Germany) hydrochloric acid (37%, p.a.,
n, Seelze, Germany) was added dropwise until a
pH value of 8.5 was reached. The resulting CBS-suspension
was used without further treatment and characterization.
The solution of each bismuth compound was prepared
separately by dissolving CBS, bismuth cysteine or bismuth
glutathione in D-PBS yielding a solution of 2000 mg bismuth
compound per kg.
The 6890 N gas chromatographic system (Agilent Technologies,
Waldbronn, Germany) was equipped with a UNIS 2000 inlet
system (Joint Analytical Systems, Moers, Germany) for
programmed temperature vaporisation (PTV) after purge
and trap sampling.
Two detection systems were used simultaneously: a 5973 N
EI-MS (Agilent Technologies, Waldbronn, Germany), which
served as a molecule selective detector for species conformation,
and a 7500a ICP-MS (Agilent Technologies) for sensitive
and element selective detection of the analytes. Working
parameters for all instruments are listed in Table 1.
More details of this unique system are described elsewhere.
2.3 Chromatographic and analytical conditions
To analyze alkylated bismuth species, constant ﬂow separation
was employed on a DB-5 MS capillary column (30 m 250 mm
25 mm, J&W Scientiﬁc, Agilent Technologies, Waldbronn,
Germany). The mobile phase was helium (5.0, Air Liquide,
Duesseldorf, Germany). Internal standardization of ICP-MS
was done by continuously adding a solution of 10 ng/ml
2.4 Cell culture
Tumorigenic human hepatocellular carcinoma cells (HepG2,
HB 8065, ATCC, USA) were cultured in minimal essential
medium (MEM, CC-Pro, Neustadt, Germany) with Earle’
BSS and sodium bicarbonate (CC-Pro) supplemented with
10% heat-inactivated foetal calf serum (FCS, Gibco
Invitrogent, Karlsruhe, Germany), nonessential amino acids
(MEM-NEAM, 0.1 mM), sodium pyruvate (1 mM), and
gentamycin (10 mg/mL, all reagents from CC-Pro) at 37 1C
under 5% CO
(99.7%, Air Liquide, Duesseldorf, Germany)
and 95% air in a water-jacket incubator (Thermo Forma,
Ohio, USA). HepG2 is an adherent cell line that grows as a
2.5 Incubation of cells
In a 25 cm
cell culture ﬂask 10
HepG2 cells were cultivated in
5 ml Minimum Essential Medium (MEM) for 24 h. Following
substitution of the medium with 10 ml HEPES and evaluation
of the cell viability via light microscopy the cells were exposed
to 1 ml of each solution containing either CBS, bismuth
cysteine or bismuth gluthathione for 24 h at 37 1C. The ﬁnal
Bi concentrations for incubation were 105 mg/kg (CBS),
93 mg/kg (BiCys
) and 122 mg/kg (BiGSH
Cell culture ﬂasks were closed with gas tight caps and Teﬂon
tape. A butyl-rubber septum was glued on the gas tight cap to
avoid the loss of volatile compounds during sample transfer.
2.6 Sample transfer & derivatization
Transfer and derivatization procedures were slightly modiﬁed
as described by Boertz et al.
After incubation the cell media
ﬂasks were stored at 80 1C for three hours and subsequently
Table 1 Operating conditions for the GC/EI-MS/ICP-MS instrument
Column DB-5 MS, 30 m 250 mm25mm
Initial head pressure 254.8 kPa
Inlet Conditions PTV
Split 1 : 50
Initial temperature 100 1C for 10 min
Heating rate 800 1C/min
Final temperature 250 1C for 11 min
Initial temperature 60 1C
Cooling rate 1 100 1C/min
Final temperature 1 35 1C for 10 min
Heating rate 2 30 1C/min
Final temperature 2 230 1C for 10 min
Mass window 200–400 amu
Transfer line temperature 280 1C
MS Quad temperature 150 1C
Ionisation energy 70 eV
Argon ﬂow 15 l/min
Carrier gas 0.79 l/min
Makeup gas 0.23 l/min
RF-Power 1540 W
Sampling depth 5.0 mm
Isotopes monitored (dwell time)
Bi (0.1 s)
This journal is
The Royal Society of Chemistry 2010 Metallomics, 2010, 2, 52–56 | 53
warmed to room temperature for lysis of the cells. Each liquid
phase was transferred by a 10 ml syringe, penetrating both
septum and cap of the cell media ﬂask, into an empty 25 ml
glass vial closed with a butyl-rubber septum. Subsequently the
vial was placed in a helium purge gas ﬂow and 1 ml of sodium
tetraethylborate solution (1% w/w) (Galab, Geesthacht,
Germany, stabilized by 2% ( w/w) KOH-solution) and
100 ml Antifoam 289
were added for ethylation. Finally the
headspace was purged for 10 minutes to a packed liner that
was cooled to 100 1C.
3. Results and discussion
3.1 Standard identiﬁcation
For identiﬁcation of bismuth cysteine and bismuth glutathione
ESI-HR-TOF-MS (positive mode) was performed. The spectrum
of bismuth cysteine shows dominant ions at m/z = 449.0033
[M + H]
, which are consistent with the expected molecular
ion (calculated 449.0037 amu) indicating the formula
was identiﬁed at m/z =
822.1396 [M + H]
in good agreement with the calculated
value of the chemical formula BiC
822.1402 amu). No signiﬁcant impurities could be detected in
3.2 Bismuth detection after incubation
Correlation of the ICP-MS signal and TIC of EI-MS revealed
all bismuth containing compounds. As shown in Fig. 1a, two
bismuth containing compounds (peaks b and d) were detected
by ICP-MS as well as EI-MS after incubation of HepG2 with
both bismuth cysteine and CBS.
A magniﬁed view of the TIC (EI-MS) gave six representative
peaks which were named as shown in Fig. 1. Taking the
ICP-MS signal into account only peaks b and d represented
bismuth compounds. All other peaks had their origin in
siloxanes which could be derived from either the GC column
or the Antifoam agent, which was added during sample
Peaks b and d were identiﬁed by interpreting the fragmentation
of these compounds as shown in Fig. 2a and b.
Consequently, these compounds could be identiﬁed as
methylbismuth and an inorganic bismuth species forming
diethyl methylbismuth and triethyl bismuth after derivatization
by sodium tetraethyl borate.
Both chromatograms and mass spectra of bismuth cysteine
and CBS, were identical. Peaks a, c, e and f could be identiﬁed
as siloxanes D3, D4, D5 and D6 by matching the spectra with
the NIST database. Considering the blank values (ESI,w
Fig. S2 and Fig. S4) their origin is probably the added
In contrast, incubation with bismuth glutathione did not
lead to the methylated bismuth compound as shown in
The methylbismuth species giving peak b did not occur in
the chromatogram. For ethylation it is known that artifact
formation can happen.
So we observed carefully the formation
of monomethyl species during ethylation of inorganic bismuth
As Fig. 3 demonstrates, no methylated species occurred
during ethylation of the cell culture medium without hepatic
cells containing bismuth compounds, nor did methylated
bismuth species occur after adding sodium tetraethyl borate
to a non-incubated sample of HepG2 in cell culture medium.
For ethylation of bismuth it turned out to be important to
buﬀer the solution to a pH value of about 7–8.
Many independent studies have shown that TMBi is a
volatile bismuth species detected in biological and environmental
although the heat of formation DH
= 194 kJ/mol
is rather high. This means permethylation of bismuth is an
energetically inappropriate process. Nevertheless TMBi was
detected, which can be explained by catalytically mediated
biomethylation. Likewise, an exchange of methyl groups
by monomethylated bismuth species as well as a stepwise
methylation in the human liver is possible. An earlier work
of Hirner and co-workers underlines the time shifted transfer
of methyl groups to bismuth by monitoring the increase
and decrease of methylated bismuth species in the blood of
bismuth exposed probands.
Nevertheless, further investigations have to be carried out to
close the gap between mono- and trimethylated bismuth
species both in vitro and in vivo.
One possible explanation of why bismuth cysteine enters
the cells is that it uses amino acid carriers. Since bismuth
glutathione is a possible excretion of HepG2, an intake into
Fig. 1 (a) Ethylation of hepatic cells incubated with BiCys
two bismuth containing compounds b and d. (b) Using BiGSH
incubation instead, peak b does not occur in the chromatogram.
54 | Metallomics, 2010, 2, 52–56 This journal is
The Royal Society of Chemistry 2010
these cells is very unlikely. This mechanism is well studied for
mercury cysteine by Clarkson et al.
The intake of CBS can be
explained by the possible formation of an amino acid complex
of bismuth by amino acids that are part of the MEM solution.
Due to the lack of suitable bismuth standards and reference
materials a quantitative reﬂection of the results is hardly
reliable. Nevertheless, assuming that both methylated and
non-methylated bismuth species have similar derivatization
eﬃciencies, comparison of their resulting peak areas in ICP-MS
chromatograms shows that approximately 2–3% of the inorganic
bismuth species were methylated.
The values listed in Table 2 are given to make a rough
estimate of bismuth conversion rates by hepatic cells. Since
there are no reliable standards available for analytical
balancing, we had to assume that there are no diﬀerent eﬀects
on methylated and non-methylated bismuth compounds
during derivatization and ionization, respectively.
Furthermore, to prevent the potential loss of volatile
bismuth species, no gas was exchanged during incubation,
although HepG2 cells require proper gas exchange for
viability. So the methylation yields could also be limited by
cell lifetime without oxygen.
Since this study was planned as a qualitative experiment
only, we did not determine LODs. Typical LODs for volatile
bismuth species on our analytical system are 0.1 to 0.3 ng per
of gaseous sample.
In summary, it was shown that CBS as well as bismuth
cysteine is methylated by HepG2 cells, in contrast to bismuth
glutathione which is not methylated. This indicates a species
dependent intake of bismuth into human hepatic cells and
supports the ﬁndings of von Recklinghausen et al.
signiﬁcant uptake of bismuth gluthathione was observed.
Further investigations for clariﬁcation of which is the
dominant species in the cells will be carried out. Likewise it
has to be speciﬁed whether the monomethyl bismuth species
were excreted by the cells or if they were still inside the cells and
could only be detected because of the lysis. With respect to the
analytics used, the sensitivity and ability of elemental detection
of ICP-MS hyphenated to GC/EI-MS providing structural
information is a powerful tool in bismuth speciation analysis.
The ICP-MS signal directly indicates where to look for
bismuth-containing compounds. Otherwise an identiﬁcation
of bismuth compounds by EI-MS spectra alone would be more
complicated, since bismuth is a monoisotopic element which
subsequently has no characteristic isotope pattern. Moreover
selected ion monitoring at m /z = 209 is not signiﬁcant since
this fragment is dominated by siloxane fragmentation
originating from the Antifoam agent or from the GC column.
Besides this, we could show that hepatoma cells are able to
methylate CBS and bismuth cysteine but not bismuth
glutathione. These results show that human hepatoma cells
have the potential to methylate bismuth and that the
permethylated species is not generated within the observed
period of time. If methylation was not observed as in the case
of bismuth glutathione, this might result from the low uptake
of this compound into the hepatoma cells. In conclusion, this
study shows that bismuth is methylated by human hepatic cells
in vitro. It appears from the result that both the intestinal
microﬂora and the liver could be involved in bismuth bio-
transformation in the human body. In future studies, after
suitable standards for method validation are available, further
investigations concerning the transformation process and the
cellular distribution of bismuth compounds should be done.
We would like to thank Prof. A. W. Rettenmeier (Institute of
Hygiene and Occupational Medicine, University of Duisburg-
Essen, Germany) for providing the facilities to perform the cell
Fig. 2 (a) Fragmentation of peak b indicating diethyl methyl
bismuth. (b) Fragmentation of peak d indicating triethyl bismuth.
Fig. 3 Ethylation of inorganic bismuth compounds in HEPES showing
no methylated species. Peak b at t
= 11.564 min is not observed.
This journal is
The Royal Society of Chemistry 2010 Metallomics, 2010, 2, 52–56 | 55
culture experiments. We would also like to thank the team of
technical assistants, especially Mrs Zimmer for helping us in
diﬀerent ways while performing the cell culture experiments.
This work was ﬁnanced by the German Research Foundation
(Deutsche Forschungsgemeinschaft; DFG) ‘‘Synthesis and
analysis of bismuth species with biological relevance’’ (‘‘Synthese
und Analyse biologisch relevanter Bismutspezies’’).
1 Y. Palmieri, Bismuth: The amazingly ‘‘green’’ environmentally-
minded element, Bulletin Bismuth Institute, Special Issue, Brussels,
2 E. Wiberg, N. Wiberg and A. F. Holleman, Inorganic Chemistry,
Academic Press, De Gruyter, San Diego, Berlin, New York, 2001.
3 H. Z. Sun, L. Zhang and K. Y. Szeto, Met. Ions Biol. Syst., 2004,
4 E. M. Grandjean, E. Ducommun, G. Gauthier and B. Courvoisier,
Schweizerische Medizinische Wochenschrift, 1976, 106, 1006–1011.
5 T. Sollmann and J. Seifter, J. Pharmacol. Exp. Ther., 1939, 67,
6 R. J. Burns, D. W. Thomas and V. J. Barron, Aust. N. Z. J. Med.,
1974, 4, 109–110.
7 A. Buge and C. Rancurel, Sem. Hop.: Ther., 1976, 52, 514–514.
8 K. Michalke, J. Meyer, A. V. Hirner and R. Hensel, Appl.
Organomet. Chem., 2002, 16, 221–227.
9 K. Michalke, A. Schmidt, B. Huber, J. Meyer, M. Sulkowski,
A. V. Hirner, J. Boertz, F. Mosel, P. Dammann, G. Hilken,
H. J. Hedrich, M. Dorsch, A. W. Rettenmeier and R. Hensel,
Appl. Environ. Microbiol., 2008, 74, 3069–3075.
10 J. Boertz, L. M. Hartmann, M. Sulkowski, J. Hippler, F. Mosel,
R. A. Diaz-Bone, K. Michalke, A. W. Rettenmeier and
A. V. Hirner, Drug Metab. Dispos., 2009, 37, 352–358.
11 U. von Recklinghausen, L. M. Hartmann, S. Rabieh, J. Hippler,
A. V. Hirner, A. W. Rettenmeier and E. Dopp, Chem. Res.
Toxicol., 2008, 21, 1219–1228.
12 M. Styblo, L. M. Del Razo, E. L. LeCluyse, G. A. Hamilton,
C. Q. Wang, W. R. Cullen and D. J. Thomas, Chem. Res. Toxicol.,
1999, 12, 560–565.
13 C. C. Bridges and R. K. Zalups, Chem. Res. Toxicol., 2006, 19,
14 E. M. Krupp, B. F. Milne, A. Mestrot, A. A. Meharg and
J. Feldmann, Anal. Bioanal. Chem., 2008, 390, 1753–1764.
15 H. H. Harris, I. J. Pickering and G. N. George, Science, 2003, 301,
16 G. G. Briand, N. Burford, M. D. Eelman, N. Aumeerally, L. Chen,
T. S. Cameron and K. N. Robertson, Inorg. Chem., 2004, 43,
17 N. Burford, M. D. Eelman, D. E. Mahony and M. Morash, Chem.
Commun., 2003, 146–147.
18 J. Hippler, M. Hollmann, H. Juerling and A. V. Hirner, Chem.
19 J. Feldmann, E. M. Krupp, D. Glindemann, A. V. Hirner and
W. R. Cullen, Appl. Organomet. Chem., 1999, 13, 739–748.
20 U. M. Gruter, J. Kresimon and A. V. Hirner, Fresenius J. Anal.
Chem., 2000, 368, 67–72.
21 J. Kresimon, U. M. Gruter and A. V. Hirner, Fresenius J. Anal.
Chem., 2001, 371, 586–590.
22 E. M. Krupp, R. Grumping, U. R. R. Furchtbar and A. V. Hirner,
Fresenius J. Anal. Chem., 1996, 354, 546–549.
23 M. V. Tagle, M. R. F. delaCampa and A. SanzMedel, Anales De
Quimica, 1996, 92, 213–218.
24 A. V. Hirner et al.,inOrganic Metal and Metalloid Species in the
Environment, ed. A. V. Hirner and H. Emons, Springer, Berlin,
1st edn, 2004, ch. 4, pp. 181–204.
25 R. A. Diaz-Bone, M. Hollmann, O. Wu
rfel and D. Pieper, J. Anal.
At. Spectrom., 2009.
26 E. Asato, K. Katsura, M. Mikuriya, U. Turpeinen, I. Mutikainen
and J. Reedijk, Inorg. Chem., 1995, 34, 2447–2454.
27 J. Kosters, J. Hippler, R. A. Diaz-Bone and A. V. Hirner, J. Anal.
At. Spectrom., 2005, 20, 996–999.
28 A. V. Hirner, Anal. Bioanal. Chem., 2006, 385, 555–567.
29 C. Elschenbroich, Organometallics, Wiley-VCH, Weinheim, 2006.
30 T. W. Clarkson, J. B. Vyas and N. Ballatorl, Am. J. Ind. Med.,
2007, 50, 757–764.
Table 2 Conversion rates of inorganic to methylated bismuth species
Compound Conversion rate
[%] Absolute amount of methylated Bi
CBS E2 E4
cells were used.
Conversion of BiGSH
was not observed.
56 | Metallomics, 2010, 2, 52–56 This journal is
The Royal Society of Chemistry 2010