Selenomethionine chiral speciation in yeast and parenteral solutions by chiral phase capillary gas chromatography-ICP-MS† S. Pe�rez Me�ndez, M. Montes Bayo�n, E. Blanco Gonza�lez and A. Sanz-Medel* Department of Physical and Analytical Chemistry. University of Oviedo, C/ Julia�n Claverý�a 8, 33006-Oviedo, Spain. E-mail: asm@sauron.quimica.uniovi.es Received 29th March 1999, Accepted 2nd June 1999 Chiral resolution and speciation of DL-selenomethionine enantiomers by capillary gas chromatography (GC) as Ntrifluoroacetyl (TFA)-O-isopropyl derivatives using an L-valine-tert-butylamide modified polydimethylsiloxane chiral stationary phase (Chirasil-L-Val ) were investigated.Good resolution was achieved using a temperature program from 100 °C (held for 3 min) to 160 °C at 1.5 °Cmin-1 and He as carrier gas. Very good selectivity and excellent detection limits of 0.25 mg L-1 (0.1 mg L-1 as Se) for each enantiomer were obtained by on-line coupling of the chiral capillary column with selenium-specific detection by inductively coupled plasma mass spectrometry (ICP-MS).Experimental parameter optimization is described in detail. This optimised GC-ICP-MS method was successfully applied to the determination of the optical purity of L-selenomethionine in commercial samples, to the determination of the enantiomeric ratio of selenomethionine in parenteral solutions used for human use and also to chiral Se speciation in selenized yeast. metabolism of Se by microorganisms (e.g., yeast) are topics Introduction of great analytical interest.Selenium is nowdays widely recognized as both a toxic and However, very few papers on this problem have appeared an essential element depending on its concentration.1 The gap so far. The enantiomeric resolution of selenomethionine by between toxic and essential levels of Se in humans is narrow2 reversed phase HPLC with UV detection after derivatization because when ingested at levels only 3–5 times higher than with a chiral reagent has been reported by Hansen and those required for optimum nutrition, detrimental eVects on Poulsen,21 while Vespale and co-workers22,23 used vancomycin health are noticed.3 Diseases related to Se deficiency such as as a chiral selector for the optical resolution of diVerent seleno- Keshan and Kaschin–Beck diseases, hypertension, infertility, containing derivatized amino acids by capillary electrophoresis arthritis, ageing and cataracts4–7 seem to have caused more (CE), again using UV detection.Both methods suVer from problems than selenium toxicity.8,9 Therefore, dietary Se sup- inadequate sensitivity and selectivity for application to complex plementation has become very popular for the prevention of real samples. In fact, no applications of chiral speciation to such diseases.5–7 Prophylactic Se supplementation has also real samples have been reported.21–23 In a previous paper,24 been recommended for clinical cancer chemoprevention.10,11 we described a method for the chiral separation of However, the nutritional bioavailability,7,12–14 toxicity7,13,15 DL-selenomethionine enantiomers derivatized with naphand cancer chemopreventive activity14,16 of Se are not only thalene-2,3-dicarboxaldehyde (NDA) by reversed phase influenced by the total trace element concentration but also HPLC on a chiral b-cyclodextrin column.Detection was by its chemical form. Among the numerous forms of Se able accomplished by molecular fluorescence of the NDA derivato be used for Se supplementation, selenomethinonine has tives and compared with Se-specific detection by on-line been suggested to be less toxic than inorganic Se forms.17 microwave digestion assisted-hydride generation-ICP-MS Selenomethionine also has an increased absorption compared (HG-ICP-MS). Fluorimetric detection proved to be superior with the inorganic salts selenite and selenate.18 In addition, in terms of sensitivity but it was unselective for real complex selenium ingested as free selenomethionine or selenized yeast samples such as selenium enriched yeast in which more than has been shown to be more bioavailable than selenite in the 20 selenium-containing species have been reported to be prelactating rats.19 Hence selenomethinonine is commonly used sent.25,26 HG-ICP-MS, however, provided suYcient selectivity as a source of selenium in several nutritional supplements and to allow the detection of DL-selenomethionine in such samples.parenteral solutions. Unfortunately, precise quantification of the enantiomers of On the other hand, selenomethionine is chiral in nature and DL-selenomethionine in these samples was not possible owing it is known that individual enantiomers of chiral compounds to the poor detection limit achieved. Therefore, there is still a usually play diVerent roles in living organisms. In fact, it has lack of adequate methodologies to deal with the separation been reported that L-selenocystine is much more toxic in rats and determination of DL-selenomethionine enantiomers in real that the D-form.20 Therefore, optical resolution of selenome- complex samples.thionine into its D- and L-enantiomers, the determination of Over the last two decades, chiral stationary phases in enantiomeric ratios of DL-selenomethionine in diVerent sel- capillary gas chromatography (GC) have undergone exciting enium solutions used for human comsumption and chiral developments,27,28 particularly for ‘optical purity’ determispeciation of selenium in selenoamino acids produced by the nations in amino acids.Fused silica capillary columns coated with L-valine-tert-butylamide (Chirasil-L-Val ) have been shown to be very suitable for the optical resolution of N- †Presented at the 1999 European Winter Conference on Plasma Spectrochemistry, Pau, France, January 10–15, 1999. trifluoroacetyl (TFA)-O-alkyl derivatives of several amino acid J.Anal. At. Spectrom., 1999, 14, 1333–1337 1333enantiomers.29 Unfortunately, an unselective flame ionization Derivatization procedures detector (FID) is the usual detector used.29 N-Trifluoroacetyl-O-isopropylselenomethionine esters of In the present work, we investigated for the first time the standards were synthesized as follows. Amounts of 0.3 mg of analytical potential of coupling chiral GC, on a Chirasil-Lracemic selenomethionine or L-selenomethionine samples and Val column, with selenium-specific detection by ICP-MS for 250 mL of propan-2-ol–acetyl chloride (8+2) were heated in a the enantiomeric resolution and determination of DL-selenomeclosed vial at 100 °C for 1 h.After the esterification reaction, thionine optical isomers. The Se specificity and high sensitivity the solvents were removed in a stream of nitrogen, 200 mL of of ICP-MS combined with the high chiral resolution power dichloromethane and 100 mL of trifluoroacetic anhydride were aVorded by the capillary column allowed the determination of added and the mixture was heated at 100 °C for 1 h.The the optical purity of L-selenomethionine in commercial solvents were then removed in a stream of nitrogen and the samples, the determination of the enantiomeric ratio of selenoresidue was dissolved in 2 mL of dichloromethane. Aliquots methionine in parenteral solutions for human use and the of 1 mL of such solution were injected into the GC column.chiral speciation of Se in more complex nutritional samples Similarly, aliquots of 200 mL of the solutions obtained as such as ‘selenized yeast’. described above from the selenium-enriched yeast and serum infusion samples were mixed with 300 mL of concentrated HCl and 500 mL of propan-2-ol and the mixture was heated in a Experimental closed vial at 100 °C for 1 h. The solution was evaporated in a stream of nitrogen, 200 mL of dichloromethane and 100 mL Apparatus of trifluoroacetic anhydride were added and the mixture was GC-FID analysis was performed using a Hewlett-Packard heated at 100 °C for 1 h.Solvents were removed again in a (Avondale, PA, USA) Model 5890 gas chromatograph stream of nitrogen and the residue was dissolved in 2 mL of equipped with an FID and a fused silica Chirasil-L-Val colmethane and used for the GC analysis as above. (25 m×0.25 mm id) (Macherey–Nagel, Du� ren, Germany).Data were acquired with Hewlett-Packard 3365 Chemstation Results and discussion software. GC-ICP-MS analysis was carried out with the same Enantiomeric separation chromatographic column in a Hewlett-Packard Model 6890 Careful optimization of the relevant operating conditions for gas chromatograph coupled to a Hewlett-Packard 4500 inducthe chiral phase capillary GC separation of DL-selenomethion- tively coupled plasma mass spectrometer (Hewlett-Packard, ine enantiomers using conventional FID was first carried out.Yokogawa Analytical Systems, Tokyo, Japan) througth a Parameters such as the temperature program of the column, laboratory-built interface which has been described in detail the injector temperature, the injection mode and the carrier elsewhere.30 It mainly consist in a metallic T-piece connected gas (He) flow rate were investigated. Then the parameters to a copper tube where the chromatographic column is placed. aVecting the performance of interface between the GC and This copper tube is inserted in a metallic block heated by ICP-MS systems were also optimized, including the length of means of an electric heater and controlled by the GC.The Tthe transfer line, the interface temperature and the make-up piece allows the insertion of an Ar gas sheathing flow (make-up sheathing gas flow (Ar). Finally, the optimum conditions for flow) orthogonal to the GC eZuent. This flow transports the specific on-line detection of Se by GC-ICP-MS were estab- analytes from the end of the GC column, via a PTFE tube lished.All these optimizations were aimed at achieving simul- (1.5 mm id) of about 40 cm length, to the nebulizer of the taneously the satisfactory optical resolution of the DL- ICP-MS for sample introduction of volatile analytes. selenomethionine derivatives and also maximum sensitivity of selenium detection. In addition, the performance of the ICP- Reagents and samples MS instrument was optimized independently of the GC system for ion lens voltages using a standard solution containing Li Racemic DL-selenomethionine and L-selenomethionine were (m/z=7), Y (m/z=89) and Tl (m/z=205) at 10 ng g-1 to purchased from Sigma (St.Louis, MO, USA), protease from obtain maximum sensitivity on the Y signal. Merck (Darmstdt, Germany), acetyl chloride from Aldrich The operating conditions selected after optimization for the (St. Louis, MO, USA), trifluoroacetic anhydride from Fluka enantiomeric GC separation were as follows: an initial oven (Buchs, Switzerland) and propan-2-ol and dichloromethane temperature of 100 °C was maintained for 3 min, then increased from Teknockroma (Barcelona, Spain).to 160 °C at a rate of 1.5 °Cmin-1 and the final oven tempera- All other chemicals and solvents were of analytical-reagent ture was maintained for 1 min. Injections were made in the grade and ultrapure Milli-Q water (18 MV cm) (Millipore, split/splitless mode (splitting ratio 1540, valve time 0.3 min) Bedford, MA, USA) was used throughout. Selenium-enriched and the injector temperature was 250 °C, the carrier gas being yeast and parenteral solution samples were kindly provided He at a pressure of 140 kPa.A typical GC-FID trace obtained from Dr. M. Potin-Gautier (University of Pau, France) and for a standard racemic mixture of DL-selenomethionine (total Dr. P. Bra�tter (Hahn-Meitner-Institut, Berlin, Germany), Se content 20 mg L-1), derivatized as described and analyzed respectively.as detailed above (Table 1), is shown in Fig. 1(A). As can be seen, very good resolution was achieved between the two Sample preparation enantiomers (Rs=2.5). Fig. 1(B) shows the chromatogram obtained when the same Selenium-enriched yeast samples were treated by enzymatic hydrolysis using the method reported by Gilon et al.:31 yeast chiral GC separation was coupled to Se-specific ICP-MS detection, through an interface design proposed by our (200 mg) and protease (20 mg) were added to 5 mL of water in a polypropylene centrifuge tube and shaken in the dark for research group.30 As can be seen, using the chromatographic conditions detailed in Table 1 and a temperature of the metallic 24 h; the solution was then centrifuged for 30 min at 3000g.The supernatant solution was removed and filtered through a block of 250 °C, band broadening took place, ruining the resolution between the D- and L-selenomethionine derivatives, 0.45 mm membrane. Aliquots of such filtrate solutions were used for derivatization.but without any significant variation of the previously observed values for retention times. In order to decrease this post- The parenteral solution sample was diluted (1540 v/v) with ultrapure water before taking an aliquot to be derivatized. column band broadening interface, the argon make-up flow 1334 J. Anal. At. Spectrom., 1999, 14, 1333–1337Table 1 Operating conditions for the hybrid GC-ICP-MS system for the enantiomeric separation of DL-selenomethionine Injector parameters— Injector port Split/splitless Injection volume 1 mL Splitting ratio 1540 Injection temperature 250 °C GC parameters— Column Chirasil-L-Val (25 m×0.25 mm id) Temperature program 100 °C (3 min); 1.5 °Cmin-1 to 160 °C Carrier gas and inlet pressure He, 130 kPa Interface temperature 250 °C Transfer line length 45 cm ICP-MS parameters— Isotopes 78Se, 82Se Rf power 1300W Fig. 2 Optimised GC-ICP-MS of a racemic mixture of DL-selenome- Sampling depth 5.8 mm thionine (as N-TFA-O-isopropyl derivatives). Experimental conditions Make-up gas flow rate 1.5 L min-1 are given in Table 1. Intermediate gas flow rate 1.0 L min-1 External gas flow rate 15 L min-1 Dwell time 0.1 s per mass ation could be almost completely achieved with this hybrid system (Rs=0.96). rate and the length of the PTFE tubing connecting the T- Analytical performance characteristics piece, where the GC column is inserted, to the ICP-MS were The precision of the total analytical method (derivatization optimized.It was observed that an increase in the Ar make-up procedure+chromatographic determination) in terms of RSD flow rate provided a noticeable improvement in the observed (n=5) was 15% for the D-selenomethionine enantiomer and band broadening while optimization of the length of the tubing 12% for the L-selenomethionine enantiomer at a concentration was critical.The optimum operating conditions finally selected of 50 mg L-1 for each enantiomer. These precisions seems to for the GC-ICP-MS system are summarized in Table 1. A be adequate for the application of the method to real samples. typical chromatogram obtained under such optimum working The calibration graphs, obtained from GC-ICP-MS analysis conditions for a standard racemic mixture of DL-selenomeof standard solutions of racemic DL-selenomethionine deriva- thionine derivatives (total Se content 6 mg L-1) is presented tives of increasing concentration showed good linearity over in Fig. 2. The results show that selenomethionine chiral specithe concentration range studied (0–100 mg L-1 of selenomethionine). The calibration curve for the D-enantiomer was best described by the equation y=32461x+1525 and that for the L-enantiomer by y=28884x+4548 ( y being the measured peak area and x the selenomethionine concentration in mg L-1). Good correlation coeYcients (r2=0.998) for both enantiomers were observed.The developed GC-ICP-MS method proved to be very sensitive as the detection limits (calculated as the concentration for a net signal equivalent to three times the background noise in the chromatogram) were found to be 0.25 mg L-1 of selenomethionine (0.1 mg L-1 as Se) for each enantiomer, with an injection volume of 1 mL. Detection limits using the GC-FID method were only estimated (owing to the elevated drift of the baseline) for comparison and turned out to be around 2–3 mg L-1.It is worth nothing that the detection limits obtained by the proposed GC-ICP-MS method were about 20 times lower than the values obtained by chiral HPLC with fluorescence detection and around 700 times lower than those observed by chiral HPG-ICP-MS methods previously developed in our laboratory.24 Chiral speciation of Se in selenomethionine in some real samples The applicability of the developed chiral phase capillary GC-ICP-MS methodology for the enantiomeric resolution and determination of D- and L-selenomethionine enantiomers in real samples was investigated.The following samples were analyzed: a commercial ‘pure’ L-selenomethionine sample, a parenteral solution used for human nutrition and a seleniumenriched yeast used as a nutritional supplement for human consumption. All these samples were treated and derivatized before analysis as described in the Experimental section. In all Fig. 1 GC of a racemic mixture of DL-selenomethionine (20 mg L-1, cases, standards together with reagent blanks were run in as Se) injected as N-TFA-O-isopropyl derivatives: (A) FID; (B) ICP-MS detection. Experimental conditions are given in the text. parallel. The signals from the blanks were always negligible, J. Anal. At. Spectrom., 1999, 14, 1333–1337 1335chloride (3.6 mg), copper DL-hydrogenaspartate (5.05 mg), manganese DL-hydrogenaspartate (2.18 mg), sodium fluoride (1.26 mg), selenomethionine (0.12 mg) and sodium molybdate (43.5 mg).Fig. 4 clearly indicates that selenomethionine in the analyzed parenteral solution sample is present as a racemate (information not given on the commercial label ). In our previous work using HPLC, this chiral speciation of Se was not possible owing to the elevated content of hydrogenaspartate in the parenteral solution sample (in HPLC with fluorescence detection, the peak corresponding to the derivatized aminoacid hydrogenaspartate overlapped the NDA selenomethionine derivative peak,24 and using HG-ICP-MS for the low sensitivity of this hypenated HPLC-HG-ICP-MS methodology provided the detection of Se in the 1540 diluted pareteral solution sample24).Chiral speciation of Se in selenized yeast. The enantiomeric resolution and determination of selenomethionine enantiomers Fig. 3 GC-ICP-MS of a commercial sample of ‘pure’ L-selenomethionby the proposed methodology was finally applied to a complex ine derivatized as above.sample of commercial selenium enriched yeast. This sample is a Saccharomyces cerevisiae yeast brought up in a sodium so the amount of DL-selenomethionine in the sample was selenite-rich medium. After pasteurization, the yeast is used as estimated from the respective peak areas of the corresponding a source of selenium and sold in pharmacies as a nutritional analyte in the sample and standard. supplement for human consumption. Recents reports31 have suggested that selenomethionine is Commercial L-selenomethionine purity.Fig. 3 shows the the predominant selenium species in selenium-enriched yeast GC-ICP-MS trace obtained for the derivatized ‘commercially (about 40% of the total selenium). The problem, however, is pure L-selenomethionine’ sample. As can be seen the presence open to debate because, as stated before, more than 20 of D-selenomethionine in the sample is apparent, in agreement selenium-containing species have been reported to appear in with our previous analysis of this sample by chiral HPLC with selenized yeast, including selenomethionine, inorganic selboth fluorimetric and HG-ICP-MS detection.24 The relative enium, selenocysteine and metylselenocysteine.25,26 Fig. 5 level of D-selenomethionine in such a commercial sample was shows the GC-ICP-MS trace obtained for an enzymatic calculated to be ca. 8% of the total DL-selenomethionine hydrolyzate of the enriched yeast sample provided by Dr.M. content, which is assumed to be 100%. Simple calculations Potin-Gautier (enzymatic hydrolysis extraction eYciency based on the extreme selectivity and sensitivity of the developed 92%31). As can be seen, a selenium peak eluting at the expected methodology demonstrate that impurity levels of the D-enanti- retention time (32.5 min) of the L-selenomethionine enantiomer omers of less than 0.1% could be detected in the commercial (Fig. 2) is apparent. However, another selenium peak eluting product.at the retention time (31 min) of the D-selenomethionine enantiomer is also visible. From the areas of the selenomethion- Serum infusion selenomethionine: chiral discrimination. The ine peaks, the latter amounts to about 15% of the total DLchromatogram for the GC-ICP-MS chiral analysis of a par- selenomethionine content in the sample. Although selenoamino enteral solution sample (diluted 1540 with ultrapure water acids of natural origin should possess the L-configuration,20 it before derivatization) is presented in Fig. 4. This sample was seems that D-amino acids are ubiquitous and common constitulabeled as containing (per 30 mL) magnesium L-hydrogenas- ents of fermented foods and beverages32 as a consequence of partate (4.87 g), zinc DL-hydrogenaspartate (15.82 mg), iron Fig. 5 Direct GC-ICP-MS of a ‘selenized yeast’ (after enzymatic Fig. 4 GC-ICP-MS of a serum infusion solution (after dilution 1540 and derivatization as above). hydrolysis and derivatization as above). 1336 J. Anal. At. Spectrom., 1999, 14, 1333–1337the action of microorganisms (bacteria, yeast). Moreover, it References has been reported that dietary commercial yeast can be a 1 M. SimonoV and G. SimonoV, Le Selenium et la Vie, Masson, potential source of D-amino acids in foodstuVs.32 Paris, 1991. In the light of the above results, it is not surprising to 2 W. N. Choy, P. R. Henika, C. C. Willhite and A. F. Tarantal, observe the presence of D-selenomethionine in the selenium- Environ.Mol. Mutagen., 1993, 21, 73. enriched yeast analyzed for Se species in this work (Fig. 5). 3 G. Yang, L. Gu, L. Zhou and R. Yin, in Selenium in Biology and Fig. 5 also shows the presence of many other selenium- Medicine, ed. A.Wendel, Springer-Verlag, Berlin, 1988, p. 223. 4 L. Fisbein, in Metals and Their Compounds in the Environmental. containing species in the enriched yeast. Occurrence, Analysis and Biological Relevance, ed.E. Merian, VCH Publishers, New York, 1991, p. 1153. 5 J. Arnaud, V. Imbault-Huart and A. Favier, in Selenium in Conclusion Medicine and Biology, ed. J. E. Neve, Walter de Gruyter and Co., GC with a fused-silica capillary column coated with L-valine- Berlin, 1988, p. 125. 6 A. D. Salbe and O. A. Levander, J. Nutr., 1990, 120, 200. tert-butylamide (Chirasil-L-Val ), as the chiral stationary 7 L. H. Foster and S. Sumar, Crit. Rev. Food Sci. Nutr., 1997, phase, has been shown to allow the enantiomeric resolution 37, 211.of N-TFA-O-isopropyl derivatives of DL-selenomethionine 8 M. T. Lo and E. J. Sandi, Environ. Pathol. Toxicol., 1980, 4, 193. enantiomers. 9 O. E. Olson, J. Am. Coll. Toxicol., 1986, 5, 45. The use of a simple new interface developed in our 10 G. F. Combs and S. B. Combs, The Role of Selenium in Nutrition, laboratory30 to couple such chiral GC separation to ICP-MS Academic Press, Orlando, FL, 1986. 11 N. V. Dimtrow and D.E. Velrey, J. Am. Coll. Toxicol., 1986, 5, 95. selenium-specific detection has proved to provide a very selec- 12 S. J. Fairweather-Tait, Eur. J. Clin. Nutr., 1997, 51, S20. tive and sensitive method allowing the enantiomeric separation 13 C. C. Willhite, W. C. Hawkes, S. T. Omaye, W. N. Choy, and chiral speciation of selenium in selenomethionine. D. N. Cox and M. Cukierski, J. Food Chem. Toxicol., 1992, 30, Detection limits around 250 ng L-1 (ppt) (100 ng L-1 as 903.selenium) can be achieved. This means a detectability for Se 14 A. J. Butler, C. D. Thomson, P. D. Whanger and M. F. Robinson, chiral speciation of around 700 times better than that obtained Am. J. Clin. Nutr., 1991, 53, 748. 15 G. H. Heinz, L. J. HoVman, L. J. LeCaptain, Arch. Environ. in our previously work by chiral HPLC-HG-ICP-MS.24 Such Contam. Toxicol., 1996, 30, 93. detection limits are also better than those reported by several 16 C. Ip and H. Gonter, in Cancer Chemoprevention, ed. workers using conventional (achiral ) HPLC-ICP-MS33–35 and L.Wattenberg, M. Lipkin, C. W. Brone and G. F. Kellof, CRC those reported by De la Calle-Guntin�as et al.36 using conven- Press, Boca Raton, FL, 1992, p. 479. tional (achiral ) GC-MIP-AES d GC-MS. It should be 17 M. A. Beilstein and P. D. Whanger, J. Nutr., 1986, 116, 1701. pointed out that chiral separation is a diYcult separation 18 P. B. Moser-Veillon, A. R. Mangels, K. Y. Patterson and C. Veillon, Analyst, 1992, 117, 3.process achieved in this work which was not aimed at in the 19 A. M. Smith and M. F. Picciano, J. Nutr., 1987, 117, 725. other work. 20 P. A. Mc Adam and O. A. Levander, Nutr. Res., 1987, 7, 601. The method has also a precision in the range reported so 21 S. M. Hansen and M. N. Poulsen, Acta Pharm. Nord., 1991, 3, 95. far using GC with a derivatization step.36 Therefore, to our 22 R. Vespale, H. Corstein, H. A. H. Billiet, J. Frank and knowledge, this hybrid GC-ICP-MS technique seems to pro- K.Ch. A. M. Luyben, Anal. Chem., 1995, 19, 3223. vide one of the more sensitive and selective methods reported 23 R. Vespale, H. A. H. Billiet, J. Frank and K. Ch. A. M. Luyben, J. High Resolut. Chromatogr., 1996, 19, 137. for Se speciation in amino acids and an excellent approach to 24 S. Pe�rez Me�ndez, E. Blanco Gonza�lez, M. L. Ferna�ndez Sa�nchez chiral selenium speciation both in parenteral solutions and, and A. J. Sanz-Medel, J.Anal. At. Spectrom., 1998, 13, 893. what is more diYcult, in complex biological samples. 25 S. Bird, G. Honghong, P. C. Uden, J. F. Tyson, E. Block and Application of this coupled method to diVerent Se E. J. Denoyer, J. Chromatogr. A, 1997, 789, 349. compounds has been demonstrated with the determination of 26 S. Bird, P. C. Uden, J. F. Tyson, E. Block and E. J. Denoyer, the ‘optical purity’ (percentage of D-enantiomer relative to the J. Anal. At. Spectrom., 1997, 12, 785. 27 V.Schuring, J. Chromatogr. A, 1994, 666, 111. total amount of amino acid) of L-selenomethionine in a 28 I. Abe, N. Fijimoto, T. Nishiyama, K. Terada and T. Nakahara, commercial ‘pure’ L-selenomethionine sample and in a par- J. Chromatogr. A, 1996, 722, 221. enteral solution used for Se supplementation in humans. 29 H. Bru�ckner and M. Lu�pke, Chromatographia, 1991, 31, 123. The applicability of the proposed technique appears to be 30 M. Montes Bayo� n, M. Gutie�rrez Camblor, J. I. Garcý�a Alonso particularly important for tackling the diYcult problem of and A. Sanz-Medel, J. Anal. Atom. Spectrom., 1999, 14, 1317. chiral speciation of Se in real biological matrices; in fact, the 31 N. Gilon, M. Potin Gautier and M. Astruc, J. Chromatogr. A, 1996, 750, 327. very high sensitivity and the specificity for Se aVorded by 32 H. Bru�ckner, M. Langer, M. Lu�pke, H. T. Westhauser, ICP-MS detection allows the direct chiral speciation of this J. Chromatogr. A, 1995, 697, 229. semi-metal in the extremely complex mixture of aminoacids 33 M. A. Quijano, A. Gutierrez, M. C. Pe�rez-Conde, C. Ca�mara, and selenoamino acids25,26 resulting from hydrolysis of J. Anal. At. Spectrom., 1996, 11, 407. ‘selenized yeast’ (Fig. 5). 34 H. M. Crews, P. A. Clarke, J. Lewis, L. M. Owen, P. R. Strutt and A. Izquierdo, J. Anal. At. Spectrom., 1996, 11, 1171. 35 R. Olivas, O. F. X. Donard, N. Gilon and M. Potin-Gautier, Acknowledgements J. Anal. At. Spectrom., 1996, 11, 1177. 36 B. De la Calle-Guntin� as, C. Brunori, R. Scerbo, S. Chiavarini, Support via a grant to S. Pe�rez Me�ndez from the Ministerio P. Quevauviller, F. Adams and R. Morabito, J. Anal. At. de Educacio�n y Cultura (Spain) is gratefully acknowledged. Spectrom., 1997, 12, 1047. Thanks are extended to Dr. M. Potin-Gautier (University of Pau, France) and to Dr. P. Bra�tter (Hahn-Meitner-Institut, Berlin, Germany) for providing the real samples. Paper 9/02524C J. Anal. At. Spe