Analysis of selenized yeast for selenium speciation by size-exclusion chromatography and capillary zone electrophoresis with inductively coupled plasma mass spectrometric detection (SEC-CZE-ICP-MS)

Abstract

A two-dimensional separation approach based on size-exclusion chromatography (SEC) followed by capillary zone electrophoresis (CZE) is proposed for the mapping of seleno-compounds in aqueous extracts of selenized yeast. The coupling of CZE with ICP-MS via a self-aspirating total consumption micronebulizer was optimized for the separation of Se species. Selenate, selenite, selenocystine, selenomethionine and selenoethionine could be baseline separated at pH 10.5 using a 10 mM phosphate buffer containing 0.8 mM cetyltriammonium bromide. Detection limits were 7–18 ng mL⁻¹ for a 20 nL injection. The CZE-ICP-MS analysis of a yeast extract demonstrated the presence of many Se species, which all migrated less rapidly than any of the standards. The following difficulties occurred during the CZE-ICP-MS analysis of the SEC fractions of the extract: the recovery of the high-molecular Se-species from the electrophoretic capillary, the presence of a large number of compounds in the medium-molecular weight fraction and the presence of a single intense peak for the low-molecular weight fraction. Proteolysis of the high- and medium-molecular weight fractions dramatically improved the recovery of Se species from the capillary, resulting in several peaks in the CZE-ICP-MS electropherograms.

---

Introduction

Pending the outcome of the ongoing human cancer prevention trials, selenium supplementation is likely to be officially recognized as a means of lowering cancer risk.^1–3 The search for an economical source of organic nutritional forms of Se has resulted in the choice of yeast grown in Se-enriched media, because of the presence of highly bioactive organic Se forms and the ease of production in quantity under controlled conditions.⁴ The wider use of Se supplements raises a number of questions as to how their quality and safety should be controlled and how fraud can be detected and prevented. Therefore a growing interest is observed in analytical methodology allowing the characterization (fingerprinting) of selenized yeast during the production process, and allowing the comparison of different products available on the market.

Despite a considerable number of reports on the analytical chemistry of selenium in yeast during the last few years, the speciation of this element still remains unknown, except in samples containing only selenomethionine. In the majority of yeast samples investigated, the water-soluble selenium seems to be present in a number of compounds,^7–20 only a few of which have been identified. Moreover, the major seleno-compounds in some yeast, e.g., Se-adenosylhomocysteine, may not contain selenomethionine. The results have so far been obtained by HPLC-ICP-MS,^5–11 but the limited peak capacity of this technique and the complexity of the Se speciation leads to interest in two dimensional HPLC or higher resolution separation methods prior to ICP-MS detection. One such method is capillary zone electrophoresis (CZE) that is known for its high efficiency (up to 100 000 theoretical plates) and for its capacity for simultaneous separation of differently charged compounds. CZE with UV detection was recently proposed for the quality control analysis of Se supplements containing selenomethionine as the only seleno-compound.¹² The applicability of the method to a larger variety of samples, as emphasized by the authors, was critically dependent on the use of a sensitive, Se-specific detector, e.g., ICP-MS.

The coupling of CZE and ICP-MS, pioneered by Olesik et al.¹³ and Lu et al.,¹⁴ has undergone a number of developments in recent years. Several interface designs, usually employing a pneumatic (micro)nebulizer^14–24 but also an ultrasonic²⁵ or a direct injection nebulizer (DIN),^26,27 have been proposed. The latest trends in CZE-ICP-MS interfacing include the use of a total consumption self-aspirating modified MCN-100 micronebulizer^28,29 and a modified direct injection high efficiency nebulizer (DIHEN).³⁰

Despite the constantly increasing number of reports on CZE-ICP-MS interface designs, the use of CZE-ICP-MS for the analysis of biological samples has been scarce. Michalke and Schramel observed a number of peaks at concentrations close to the detection limits (10–50 ng mL⁻¹) when monitoring Se in human body fluids.³¹ Other applications have included speciation of iodine in human serum³² and of metal complexes with metallothionein in rat organs³³ and in human brain.³⁴ The likely reason for this paucity is the effect of the biological matrix on the separation for custom-designed sample preparation methods.

The objective of this research was the evaluation of the potential of CZE-ICP-MS for mapping of water-soluble Se metabolites in yeast. The interface used was based on a total consumption self-aspirating micronebulizer, described in detail elsewhere^28,29 and recently successfully applied to the determination of metal complexes with metallothionein isoforms.^33,34 The achievement of this objective turned out to depend critically on the optimization of a preliminary fractionation of the aqueous extract (by size-exclusion chromatography), and on the verification of the compound's stability in the presence of proteolytic enzymes.

Experimental

Instrumentation

CZE experiments were carried out with a Beckmann P/ACE 2200 (Beckmann, Fullerton, CA). The ICP-MS instrument used was an Agilent Model 7500 (Yokogawa, Yamanashi-Ken, Japan). The instrument was automatically tuned by the Chemstation software using the ion signal of ⁸⁹Y in the make-up flow. CZE separations were carried out in an uncoated fused silica (ca. 1 m, 75 µm id) capillary.

The CZE-ICP-MS interface was based on a microconcentric nebulizer (MCN-100, CETAC, Omaha, NE), in which the original nebulizer capillary was replaced by one with a narrower diameter, and a small volume (ca. 5 ml) spray chamber (CETAC). The interface was described in detail elsewhere.^28,29 In brief, the make-up liquid, grounded by a Pt electrode, was mixed with the CZE buffer at the end of the CZE capillary. It was transported to the nebulizer by self-aspiration at 6 µL min⁻¹.

Low-pressure LC was carried out by means of a model Minipuls 3 peristaltic pump (Gilson, France). For size-exclusion chromatography, a 700 × 16 mm column (Pharmacia, Uppsala, Sweden) was filled with G-15 Sephadex gel (Pharmacia) (exclusion M_r 1.5 × 10³) and calibrated according to the manufacturer's protocol. For anion-exchange chromatography a Hamilton PRP X-100 column (250 mm × 4.6 mm × 5 µm) (Hamilton, Reno, NV, USA) was used. Fractions for off-line analyses were collected using a Model FC-2 automatic fraction collector (Dynamax, France). A Model LP3 lyophilizer (Jouan, France) was used for freeze-drying of sample extracts and eluates. A Universal Model 16 centrifuge (Hettich, Germany) was used.

Electrospray MS experiments were performed using a PE-SCIEX API 300 ion-spray triple-quadrupole mass spectrometer (Thornhill, ON, Canada). Samples were introduced using a syringe pump (Harvard Apparatus, South Natic, MA).

Standards, samples and reagents

DL-Selenocystine, DL-selenomethionine, and selenoethionine were purchased from Sigma (St. Quentin Fallavier, France) and were used without further purification (90% purity for selenocystine). Stock solutions containing 1 mg mL⁻¹ of each compound in deionized water were stored in the dark at 4 °C. Hydrochloric acid (3%) was used to dissolve selenocystine. Working standards were prepared daily by dilution in deionized water.

A sample of industrially produced (Lasafre, France) selenium-enriched yeast was used. Saccharomyces cerevisiae was grown in the presence of sodium selenite and organic seleno-compounds were naturally synthesized. It was then pasteurized and dried (to avoid further growth and facilitate handling).

Analytical-reagent grade chemicals, including Pronase E (Protease XIV type) were obtained from Sigma-Aldrich (St. Quentin Fallavier, France). Water purified to 18.2 MΩ cm resistivity using a Milli-Q water purification system (Millipore, Bedford, MA) was used. The CZE buffer solutions were prepared by dissolving 10 mM Na₂HPO₄ in water and adjusting the pH with ammonia.

Procedures

A 0.2 g yeast sample was extracted with 5 mL of water in an ultrasonic bath. The supernatant was separated by centrifugation (4000 rev min⁻¹ for 5 min at room temperature) and freeze-dried (overnight). The residue was dissolved in 500 µL of water and analyzed directly by CZE-ICP-MS. For size-exclusion chromatography, a sample of 4 mL of extract was fractionated using a 1% (v/v) solution of acetic acid solution as the mobile phase. The fractions corresponding to peaks of Se compounds were collected, lyophilized and redissolved in 500 µL of CZE buffer prior to CZE-ICP-MS analysis. For anion-exchange chromatography, the low molecular mass fraction after SEC was collected, lyophilized, redissolved in 1 m of 20 mM ammonium acetate∶acetic acid (mixture 1 + 1, pH 4.7) and fractionated using a gradient: 0–5 min 20 mM, 5–30 min up to 200 mM and 200 mM of the same buffer remaining constant after 30 min. The Se-containing fractions within each of the peaks were combined, lyophilized and redissolved in 500 µL of CZE buffer prior to CZE-ICP-MS analysis. For proteolysis, a Se-containing fraction was lyophilized, dissolved in 5 mL of phosphate buffer (pH 7.5) containing 20 mg of Pronase E and incubated for 16 h at 37 °C. The CZE-ICP-MS experimental conditions are summarized in Table 1. The ES MS/MS experimental conditions were as specified elsewhere.³⁵

Table 1 Optimum CZE-ICP-MS conditions

CZE separation parameters
Capillary length	80 to 100 cm
Running buffer	10 mM Na2HPO4 with 0.88 mM cetyltriammonium bromide
Injection (hydrodynamic)	0.5 psi for 3 s
Voltage (current)	20 kV (20 µA)
Pressure during separation	None
Temperature	25 °C
Interface conditions
Sheath liquid	10 mM Na2HPO4
Sheath liquid flow rate	6 µL min^−1
ICP-MS conditions
Sampler	Ni (1.0 mm orifice)
Skimmer	Ni (0.4 mm orifice)
Rf Power	1050 kW
Plasma gas flow rate	15 L min^−1
Auxiliary gas flow rate	1 L min^−1
Nebulizer gas flow rate	1.13 L min^−1
Isotopes monitored	^89Y, ^80Se, ^82Se
Dwell time	100 ms

---

Results and discussion

Optimization of the CZE-ICP-MS conditions

The basics of the separation of seleno-compounds by CZE were set by Albert et al.³⁶ and Gilon and Potin-Gautier³⁷ and further developed by Zheng et al.¹² Inorganic and organic Se species can be separated at negative potentials provided that the polarity of the capillary surface covered by silanol groups is decreased or reversed by the addition of a cationic surfactant (adsorbed on the capillary inner wall). In such a system the electrophoretic mobilities of anions are in the same direction as the electro-osmotic flow, which allows a fast separation even of ions with considerable negative charge density. The principal parameters to be optimized include pH, the choice and concentration of the cationic surfactant, the potential value and the buffer concentration.

Alkaline pH is necessary to achieve the separation of selenoamino acids, most of which remain zwitterionic (uncharged) at pH values lower than 7. Therefore, buffers at three different pH values, all of which allowed the suppression of the ionization of the amine functions, were investigated. Fig. 1 shows electropherograms obtained at the different pH values, the other parameters being set at optimal. It shows that the choice of pH is critical; the best separation efficiency being obtained at pH 10.5 ± 0.05. At lower pH values the electro-osmotic flow is too weak, leading to long separations and peak broadening. At higher pH values the baseline separation between the standards is not achieved. The effects of other parameters were negligible in the investigated ranges: 0.2–1.5 mM for cetyltriammonium bromide concentration, 10–30 mM for the buffer and 10–30 V for the separation voltage.

Fig. 1 CZE-ICP-MS electropherograms obtained for a mixture of seleno-compounds at different pH of the running buffer: a) pH 9.7; b) pH 10.5; c) pH 11.3. 1 - selenate, 2 - selenite, 3 - selenocystine, 4 - selenomethione, 5 - selenoethionine. 10 µg mL⁻¹ of each compound (as Se). See Table 1 for experimental conditions.

The detection limits are dependent on the dilution factor, the sheath flow and the nebulization efficiency. The dilution factor is controlled by the aspiration rate of the make-up liquid. The latter remains constant at 6 µL min⁻¹ for the argon nebulizer gas flow rates above 1 L min⁻¹, which induces a dilution factor of about five. The signal intensity is the highest for a nebulizer gas flow rate of 1.13 L min⁻¹, which was therefore applied in further work. The ICP-MS sensitivity under these conditions reached 10⁵ counts s⁻¹ for 10 µg L⁻¹ of Y. The signal stability was within 3–4%. The nebulization efficiency was determined to be between 75 and 85% by collecting the aerosol leaving the spray chamber onto an absorbant paper for 10 min and comparing the quantity of absorbed Se with that introduced. Since the system does not have a drain, a small part of the aerosol remains on the wall of the spray chamber, which requires a daily cleaning to reduce the baseline intensity and noise.

The electropherogram of the five Se standards at 10 times the detection limit level is shown in Fig. 2a. The figures of merit are summarized in Table 2. The detection limits are identical with those reported earlier for inorganic Se species using the same interface but with detection by a sector-field instrument.²⁸ The detection limits for organic species are a factor of three lower than those reported by Michalke and Schramel.³⁸ These values compare unfavourably with the 0.1 ng mL⁻¹ reported by Liu et al.²⁶ and the 0.2 µg L⁻¹ recently reported by Bendahl et al.,³⁰ which were obtained with direct injection nebulization where no spray chamber related peak broadening occurs. Indeed, the peak width at the baseline in our work was ca. 15 s compared to a value of 3 s reported by Bendahl et al.³⁰

Fig. 2 CZE-ICP-MS electropherograms obtained under optimum conditions (cf., Table 1) of: a) a mixture of selenium standards at 90 ng mL⁻¹; and b) an aqueous extract (diluted 100 times) of selenized yeast. 1 - Selenate, 2 - selenite, 3 - selenocystine, 4- selenomethione, 5 - selenoethionine.

Table 2 Figures of merit for the determination of Se species by CZE-ICP-MS

	Migration time/min	Peak intensity reproducibility (%)	Detection limits/ng mL^−1	Linearity coefficient R^2 (10–4000 µg L^−1 range)
Se^VI	5.0 ± 0.2	3	9	0.9992
Se^IV	5.2 ± 0.1	9	7	0.9947
Selenocystine	7.2 ± 0.2	8	18	0.9995
Selenomethionine	7.7 ± 0.3	10	14	0.9999
Selenoethionine	7.8 ± 0.8	11	16	0.9995

---

CZE-ICP-MS of an aqueous extract of yeast

Fig. 2b shows an electropherogram of an aqueous extract of a selenized yeast sample. A number of peaks can be seen, none of which corresponds to the migration time of standard species. Note that the dilution factor (100) was sufficiently high to prevent any matrix-related shift in migration times, as verified by a spiking experiment. The presence in the sample of Se compounds other than the standards required a re-optimization of the separation conditions. A comparison of the electropherograms of the extract obtained at different pH (9.7, 10.5 and 11.3, data not shown) indicated that the highest separation efficiency was obtained at pH 10.5, which is similar to standards. However, under these conditions no baseline separation was obtained for a number of compounds. Moreover, ca. 25–30% of selenium did not leave the capillary within 30 min. A deeper insight into these phenomena can be obtained by a preliminary fractionation of seleno-compounds in the extract obtained by SEC followed by CZE-ICP-MS.

Size-exclusion LC-CZE-ICP-MS of an aqueous extract of yeast

A size exclusion chromatogram of the yeast aqueous extract (insets in Fig. 3) shows the presence of three rather well separated fractions as a function of the molecular weight (hydrodynamic volume): one excluded from the column (M_r > 20 kDa); one in the 5–10 kDa range; and one in the total volume of the column. The fractions account for 98.7 ± 1.5% of total Se. The fractions were collected as indicated in the insets in Fig. 3a, b and c, respectively, preconcentrated as described in the Procedures section, and analyzed by CZE-ICP-MS.

Fig. 3 Analysis of a yeast aqueous extract by SEC-CZE-ICP-MS. The SEC-ICP-MS chromatograms (⁸⁰Se, off-line detection) are shown in the insets. Panels a, b and c show CZE-ICP-MS electropherograms of the SEC fractions shaded in the insets. The dotted line marks the end of an electropherogram after which pressure (5 psi) was applied to remove the rest of the sample from the capillary.

Fig. 3a shows that none of the Se species excluded from the size-exclusion column migrates fast enough to leave the capillary within 25 min. However, when a pressure is applied after this time period, seleno-compounds are pushed out and separated to a certain degree. This behavior suggests that the compounds are likely to be large selenopolypeptides. Indeed, when the excluded fraction is submitted to enzymolysis with lipase-pronase³⁹ followed by CZE-ICP-MS, a number of seleno-compounds can be seen (Fig. 4a). At this stage the compounds cannot be identified but they are likely to be selenopeptides, products of the enzymolysis of a selenoprotein. Note that, after proteolysis, more than 95% of Se leaves the capillary without the applied pressure.

Fig. 4 CZE-ICP-MS electropherograms of proteolytic digests of SEC fractions: a) fraction shaded in the inset in Fig. 3a; and b) fraction shaded in the inset in Fig. 3b.

The CZE-ICP-MS electropherogram of the medium-molecular weight SEC fraction (Fig. 3b) shows the presence of a number of compounds with a considerable fraction of selenium unable to leave the capillary within 25 min. The baseline is very noisy, which suggests the presence of many poorly resolved seleno-species. An increase in the CTAB concentration does not improve the separation and the presence of a significant number of compounds was confirmed by chromatography. The proteolysis does not affect the bulk of selenospecies (Fig. 4b).

However, four new peaks between 18–20 min followed by an intense peak at 24.9 min, which are apparently degradation products of larger selenopeptides, can be seen.

The CZE-ICP-MS electropherogram of the low-molecular weight SEC fraction shows one very intense peak (Fig. 3c), which suggests the presence of a single selenium species. Above 80% of the peak intensity is recovered after proteolysis. In order to validate the hypothesis regarding the peak purity, the low-molecular weight SEC fraction was analyzed by anion-exchange HPLC-ICP-MS.

Size-exclusion-anion exchange HPLC-CZE-ICP-MS of an aqueous extract of yeast

The results of the three dimensional separation (SEC-anion exchange HPLC-CZE)-ICP-MS analysis are shown in Fig. 5. Anion exchange HPLC-ICP-MS chromatograms (shown in the insets) confirm that the vast majority of selenium (80.4 ± 1.0%) elutes as a strongly anionic compound; the rest of the selenium being distributed in a compound eluting in the void (16.7 ± 1.0%) and a small species (2.9 ± 0.8%) eluting in the middle of the chromatogram.

Fig. 5 Anion exchange HPLC-CZE- CP-MS of the low-molecular weight size-exclusion fraction of yeast extract. The insets show the anion exchange HPLC-ICP-MS chromatograms (off-line detection) and corresponding ES MS/MS results obtained for the most intense ES MS signals (containing ⁸⁰Se); selenium-containing fragments are marked in bold. Panels a, b and c show sets of two CZE-ICP-MS electropherograms of the anion exchange HPLC fractions shaded in the insets. The top panel in each set corresponds to an analysis of the untreated HPLC fraction and the bottom panel corresponds to an analysis of the proteolytic digest of this fraction.

The major anion exchange HPLC fraction submitted to CZE-ICP-MS (Fig. 5a, top panel) indicates the presence of one single seleno-compound, which is resistant to proteolysis (Fig. 5a, bottom panel). Electrospray MS indicates a cluster with the isotopic pattern of selenium centered at m/z 433, which corresponds to Se-adenosylhomocysteine (M_r = 432), confirmed by ES MS/MS (spectrum shown in the inset).

The minor peak in the SEC-anion exchange HPLC-ICP-MS chromatogram corresponds to a compound that was unable to leave the capillary within 30 min (Fig. 5b, top panel). The enzymolysis of this compound leads to a major peak (Fig. 5b, bottom panel) at the migration time of 24.9 min, which is close to the peak appearing after proteolysis of the medium-molecular weight SEC fraction (c.f., Fig. 4b). No selenium cluster could be observed for this species using ES MS.

Finally, the CZE-ICP-MS electropherogram (Fig. 5c, top panel) of the SEC-anion exchange HPLC void indicates the presence of two compounds; the major one being three times more intense than the minor one. The proteolysis (Fig. 5c, bottom panel) converts part of the major compound into the minor. Electrospray MS indicates the presence in this fraction of a compound with M_r 430 (m/z of protonated ion at 431). The collision-induced fragmentation of this compound (shown in the inset) indicates the presence of the adenosyl group (fragments at m/z 136 and 250). However, the presence of a signal at m/z 180, as well as of intense adenine and adenosine peaks, indicate the presence of unsaturated residue on the opposite side to adenosine.

Conclusions

Capillary zone electrophoresis with ICP-MS detection is a promising technique for the screening of yeast extracts that allows a fingerprint of selenium speciation. Despite the high separation efficiency of CZE, baseline resolution of all the seleno-species present is not achieved, which makes a preliminary chromatographic fractionation step necessary. At the optimum separation conditions for inorganic selenium and selenoaminoacids, larger molecules, such as selenopolypeptides, do not migrate fast enough to leave the column in a reasonable time. Enzymatic degradation of selenoprotein fractions may be necessary prior to CZE-ICP-MS. Three dimensional chromatography (electrophoresis), SEC-anion exchange HPLC-CZE-ICP-MS, allows one to obtain maps of the individual metal species that may be correlated with ES MS/MS data for the purpose of identification, but considerable work in this direction is still required.

References

1. L. C. Clark, J. G. Combs, B. W. Turnbill, E. H. Slate, D. K. Chalker, J. Chow, K. S. Davis, R. A. Glover, G. F. Graham, E. G. Gross, A. Krongrad, J. L. Lesher, K. Park, B. B. Sanders, C. L. Smith and J. R. Taylor, J. Am. Med. Assoc., 1996, 276, 1957.
2. M. P. Rayman, Br. Med. J., 1997, 314, 387.
3. M. P. Rayman, Lancet, 2000, 356, 233.
4. J. Neve, in Bioavailability and safety of selenium supplements, ed. P. Collery, P. Bratter, V. Negrettii de Bratter, I. Khassanova and J. C. Etienne, John Libbey Eurotext, Paris, 1998.
5. J. Zheng, W. Goessler and W. Kosmus, Trace Elements Electrol., 1998, 15, 70.
6. P. C. Uden, S. M. Bird, M. Kotrebai, P. Nolibos, J. F. Tyson, E. Block and E. Denoyer, Fresenius' J. Anal. Chem., 1998, 362, 447.
7. M. Kotrebai, M. Birringer, J. F. Tyson, E. Block and P. C. Uden, Analyst, 2000, 125, 71.
8. M. Kotrebai, M. Biringer, J. F. Tyson, E. Block and P. C. Uden, Anal. Commun., 1999, 36, 249.
9. M. Kotrebai, J. F. Tyson, E. Block and P. C. Uden, J. Chromatogr. A, 2000, 866, 51.
10. S. M. Bird, H. Ge, P. C. Uden, J. F. Tyson, E. Block and E. Denoyer, J. Chromatogr, A., 1997, 789, 349.
11. S. M. Bird, P. C. Uden, J. F. Tyson, E. Block and E. Denoyer, J. Anal. At. Spectrom., 1997, 12, 785.
12. J. Zheng, H. Greschonig, F. Liu and W. Kosmus, Trace Elements Electrol., 2000, 17, 40.
13. J. W. Olesik, J. A. Kinzer and S. V. Olesik, Anal. Chem., 1995, 67, 1.
14. Q. Lu, S. M. Bird and R. M. Barnes, Anal. Chem., 1995, 67, 2949.
15. M. Van Holderbeke, Y. N. Zhao, F. Vanhaecke, L. Moens, R. Dams and P. Sandra, J. Anal. At. .Spectrom., 1999, 14, 229.
16. S. A. Baker and N. J. Miller-Ihli, Appl. Spectrosc., 1999, 53, 471.
17. B. Michalke and P. Schramel, Analusis, 1998, 26, M51.
18. K. A. Taylor, B. L. Sharp, D. J. Lewis and H. M. Crews, J. Anal. At. Spectrom., 1998, 13, 1095.
19. A. Tangen and W. Lund, J. Chromatogr., 2000, 891, 129.
20. J. A. Day, J. A. Caruso, J. S. Becker and H. J. Dietze, J. Anal. At. Spectrom., 2000, 15, 1343.
21. J. M. Costa-Fernandez, N. H. Bings, A. M. Leach and G. M. Hieftje, J. Anal. At. Spectrom., 2000, 15, 1063.
22. S. Mounicou, K. Polec, H. Chassaigne, M. Potin-Gautier and R. Lobinski, J. Anal. At. Spectrom., 2000, 15, 635.
23. C. B'Hymer, J. A. Day and J. A. Caruso, Appl. Spectrosc., 2000, 54, 1040.
24. J. A. Day, K. L. Sutton, R. S. Soman and J. A. Caruso, Analyst, 2000, 125, 819.
25. Q. H. Lu and R. M. Barnes, Microchem. J., 1996, 54, 129.
26. Y. Liu, V. Lopez Avila, J. J. Zhu, D. R. Wiederin and W. F. Beckert, Anal. Chem., 1995, 67, 2020.
27. A. Tangen, W. Lund, B. Josefsson and H. Borg, J. Chromatogr., A., 1998, 826, 87.
28. A. Prange and D. Schaumlöffel, J. Anal. At. Spectrom., 1999, 14, 1329.
29. D. Schaumlöffel and A. Prange, Fresenius' J. Anal. Chem., 1999, 364, 452.
30. L. Bendahl, B. Gammelgaard, O. Jons, O. Farver and S. H. Hansen, J. Anal. At. Spectrom., 2001, 16, 38.
31. B. Michalke and P. Schramel, J. Chromatogr., A., 1998, 807, 71.
32. B. Michalke and P. Schramel, Electrophoresis, 1999, 20, 2547.
33. K. Polec, M. Peréz-Calvo, O. Garcia-Arribas, J. Szpunar, B. Ribas-Ozonas and R. Lobinski, J. Inorg. Biochem., 2001, in the press.
34. A. Prange, D. Schaumlöffel, P. Brätter, A. Richarz and C. Wolf, Fresenius' J. Anal. Chem., 2001, 371, 764.
35. C. Casiot, V. Vacchina, H. Chassaigne, J. Szpunar, M. Potin-Gautier and R. Lobinski, Anal. Commun., 1999, 36, 77.
36. M. Albert, C. Demesmay and J. L. Rocca, Analusis, 1993, 21, 403.
37. N. Gilon and M. Potin-Gautier, J. Chromatogr., 1996, 732, 369.
38. B. Michalke and P. Schramel, Electrophoresis, 1998, 19, 270.
39. N. Gilon, A. Astruc, M. Astruc and M. Potin-Gautier, Appl. Organomet. Chem., 1995, 9, 623.

---