Amino Acid Analysis by Capillary Electrophoresis Electrospray

A method for the determination of underivatized amino acids based on capillary electrophoresis coupled to electrospray ionization mass spectrometry ...
0 downloads 0 Views 105KB Size
Download PDF
Anal. Chem. 2000, 72, 1236-1241

Amino Acid Analysis by Capillary Electrophoresis Electrospray Ionization Mass Spectrometry Tomoyoshi Soga*

Yokogawa Analytical Systems Inc., 2-11-13 Nakacho, Musashino-shi, Tokyo 180-0006, Japan David N. Heiger

Agilent Technologies, Little Falls Site, 2850 Centerville Road, Wilmington, Delaware 19808

A method for the determination of underivatized amino acids based on capillary electrophoresis coupled to electrospray ionization mass spectrometry (CE-ESI-MS) is described. To analyze free amino acids simultaneously a low acidic pH condition was used to confer positive charge on whole amino acids. The choice of the electrolyte and its concentration influenced resolution and peak shape of the amino acids, and 1 M formic acid was selected as the optimal electrolyte. Meanwhile, the sheath liquid composition had a significant effect on sensitivity and the highest sensitivity was obtained when 5 mM ammonium acetate in 50% (v/v) methanol-water was used. Protonated amino acids were roughly separated by CE and selectively detected by a quadrupole mass spectrometer with a sheath flow electrospray ionization interface. Under the optimized conditions, 19 free amino acids normally found in proteins and several physiological amino acids were well determined in less than 17 min. The detection limits for basic amino acids were between 0.3 and 1.1 µmol/L and for acidic and low molecular weight amino acids were less than 6.0 µmol/L with pressure injection of 50 mbar for 3 s (3 nL) at a signal-to-noise ratio of 3. This method is simple, rapid, and selective compared with conventional techniques and could be readily applied to the analysis of free amino acids in soy sauce. Amino acids are very important in many fields. Their compositional analysis in proteins and peptides is essential for the study of the primary structure in biochemistry. Determination of amino acids in body fluids such as urine and blood can help in the diagnosis and treatment of diseases. While in food science, amino acids are measured to determine the quality of the final products. Since amino acids are nonvolatile compounds and most of them have little UV absorbance, amino acids have been commonly analyzed by high-performance liquid chromatography (HPLC) methods with precolumn or postcolumn derivatization using UV chromophore or fluorophore reagents. Initially, free amino acids were separated by cation-exchange chromatography and treated with ninhydrin in a postcolumn derivatization step, and the derivatives were detected with a UV * Corresponding author: (tel) (+81) 422 52 5645; (fax) (+81) 422 52 5966; (e-mail) [email protected].

1236 Analytical Chemistry, Vol. 72, No. 6, March 15, 2000

detector.1,2 This approach was reliable; however, it needs a long analysis time, has a high running cost, and is a complicated system. Alternatives to the postcolumn HPLC method were developed with precolumn derivatization using reagents such as phenyl isothiocyanate (PITC),3 o-phthalaldehyde (OPA),4 4-(dimethylamino)azobenzene-4-sulfonyl (DABSYL) chloride,5 dimethylaminonaphthalene-1-sulfonyl (DANSYL) chloride,6 or 9-fluorenylmethyl chloroformate (FMOC),7 followed by separation and detection. Although these precolumn methods can improve sensitivity and reduce the analysis time, each method has some drawbacks: such as manual derivatization procedures are necessary or excess reagent must be removed to prevent interference with other amino acids. Another approach to amino acid analysis is liquid chromatography coupled to mass spectrometry (LC-MS). Although it has demonstrated outstanding performance, suitable LC-MS conditions for free amino acids are not established. This is because the mobile phase is limited in LC-MS; it is difficult to obtain appropriate separation, analysis time and sensitivity. Therefore, most of the LC-MS methods have been developed for derivatized amino acids.8,9 Capillary electrophoresis (CE) is a new separation technique that can provide high resolution efficiency, and a great number of CE methods have been developed for the analysis of amino acids. The analysis of amino acids by CE has been mainly performed with precolumn, postcolumn, and on-column derivatization techniques with UV chromophore10,11 or fluorophore reagents12-15 to provide better resolution and sensitivity. On the (1) Moore, S.; Stein, W. H. J. Biol. Chem. 1951, 192, 663-681. (2) Spackman, D. H.; Stein, W. H.; Moore S. Anal. Chem.1958, 30, 11901206. (3) Heinrikson, R. L.; Meredith, S. C. Anal. Biochem. 1984, 136, 65-74. (4) Schuster, R. J. Chromatogr. 1988, 431, 271-284. (5) Lim, J. K.; Wang, C. H. Clin. Chem. 1980, 28, 579-583. (6) Schmidt, G. J.; Olson, D. C.; Slavin, W. J. Liq. Chromatogr. 1979, 2, 10311045. (7) Einarson, S.; Josefsson, B.; Lagerkvist, S. J. Chromatogr. 1983, 282, 609618. (8) van Leuken, R. G. J.; Duchateau, A. L. L.; Kwakkenbos, G. T. C. J. Pharm. Biomed. Anal. 1995, 13, 1459-1464. (9) Fujii, K.; Shimoya, T.; Ikai, Y.; Oka, H. Harada, K. Tetrahedron Lett. 1998, 39, 2579-2582. (10) Skocir, E.; Vindevogel, J.; Sandra, P. Chromatographia 1994, 39, 7-10. (11) Terabe, S.; Ishihama, Y.; Nishi, H.; Fukuyama, T.; Otsuka, K. J. Chromatogr. 1991, 545, 359-368. 10.1021/ac990976y CCC: $19.00

© 2000 American Chemical Society Published on Web 02/18/2000

other hand, several CE methods with indirect UV detection for the analysis of underivatized amino acids were reported.16,17 These methods are simple; however, their drawback is less sensitivity and selectivity. Recently, capillary electrophoresis electrospray ionization mass spectrometry (CE-ESI-MS) has rapidly developed as a powerful analytical tool for charged species ranged from small moleculars such as carboxylic acids,18 phenolic compounds,19 metal species,20 tetramines,21 herbicides,22 and drugs and drug metabolites23 to peptides and proteins.24,25 While CE confers rapid analysis and efficient resolution, MS provides high selectivity and sensitivity. The ESI mode has proven to be sensitive, versatile, and relatively easy to use in combination with CE. Very few studies applied to amino acid analysis have been performed by CE-ESI-MS. Garcia and Henion reported a gelfilled CE-MS method for the determination of DANSYL-amino acids.26 Lu and co-workers determined four amino acids by CEESI-MS.27 In this paper, we developed a CE-ESI-MS method for the determination of 19 underivatized normally found amino acids and several physiological amino acids. The method was optimized and then applied to the analysis of amino acids in soy sauce. EXPERIMENTAL SECTION Chemicals. A 17-amino acid standard mixture at a concentration of 2.5 mmol/L each, except for Cys at 1.25 mmol/L, in 0.1 N HCl was purchased from Pierce (Rockford, IL). Asp and Trp were obtained from Wako (Osaka, Japan) and prepared at a concentration of 2.5 mmol/L in 0.1 N NaOH just before making the mixture standard. The working mixture standard was prepared by diluting these stock solutions with Milli-Q water. All other reagents were from Wako. The chemicals used were of analytical or reagent grade. Water was purified with a Milli-Q purification system (Millipore, Bedford, MA). Instrumentation. All CE-ESI-MS experiments were performed using an Agilent CE capillary electrophoresis system with built-in diode-array detector, an Agilent 1100 series MSD mass spectrometer, an Agilent1100 series isocratic HPLC pump, a G1603A Agilent CE-MS adapter kit, and a G1607A Agilent CE(12) Taga, A.; Honda, S. J. Chromatogr., A 1996, 742, 243-250. (13) Albin, M.; Weinberger, R.; Sapp, E.; Moring, S. Anal. Chem. 1991, 63, 417422. (14) Ueda, T.; Mitchell, R.; Kitamura, F.; Metcalf, T.; Kuwana, T.; Nakamoto, A. J. Chromatogr. 1992, 593, 265-274. (15) Arriaga, E. A.; Zhang, Y.; Dovichi, N. J. Anal. Chim. Acta 1995, 299, 319326. (16) Lee, Y. H.; Lin, T.-I. J. Chromatogr., A 1994, 680, 287-297. (17) Soga, T.; Ross, G. A. J. Chromatogr., A 1999, 837, 231-239. (18) Johnson, S. K.; Houk, L. L.; Johnson, D. C.; Houk, R. S. Anal. Chim. Acta 1999, 389, 1-8. (19) Lafont, F.; Aramendia. M. A.; Garcı´a, I.; Borau, V.; Jime´nez, C.; Marinas, J. M.; Urbano, F. J. Rapid Commun. Mass Spectrom. 1999, 13, 562-567. (20) Schramal, O.; Michalke, B.; Kettrup, A. J. Chromatogr., A 1998, 819, 231242. (21) Zhao, J.; Thibault, P.; Tazawa, T.; Quilliam M. A. J. Chromatogr., A 1997, 781, 555-564. (22) Song, X.; Budde, W. L. J. Am. Soc. Mass Spectrom.1996, 7, 981-986. (23) Lu, W.; Poon, G. K.; Carmichael, P. L.; Cole, R. B. Anal. Chem. 1996, 68, 668-674. (24) Cao, P.; Moini, M.J. Am. Soc. Mass Spectrom. 1998, 9, 1081-1088. (25) Kelly, J. F.; Locke, S.J.; Ramaley, L.; Thibault, P. J. Chromatogr., A 1996, 720, 409-427. (26) Garcia, F.; Henion, J. D. Anal. Chem. 1992, 64, 985-990. (27) Lu, W.; Yang, G.; Cole, R. B. Electrophoresis 1995, 16, 487-492.

ESI-MS sprayer kit (all Agilent Technologies, Waldbronn, Germany). All system control, data acquisition, and MSD data evaluation were performed with G2201AA Agilent ChemStation software for CE-MSD. The CE-MS adapter kit includes a capillary cassette, which facilitates thermostating of the capillary, and the CE-ESI-MS sprayer kit, which simplifies coupling the CE system with MS systems equipped with an electrospray source. The sprayer was designed as orthogonal flow to reduce the detrimental effects caused by the charged particles or droplets as Voyksner and Lee described before.28 CE-ESI-MS Conditions. Separations were carried out on fused-silica capillaries with 50 µm i.d. × 100 cm total length. The electrolyte for the CE separation was 1 M formic acid solution. Prior to first use, a new capillary was pretreated with the run electrolyte for 20 min. Before each injection, the capillary was preconditioned for 4 min by flushing with the running electrolyte. Sample was inserted with a pressure injection of 50 mbar for 3.0 s (∼3 nL). The applied voltage was set at 30 kV, and the capillary temperature was thermostated to 20 °C. The Agilent 1100 series pump equipped with a 1:100 splitter was used to deliver 10 µL/ min 5 mM ammonium acetate in 50% (v/v) methanol-water to the CE interface where it is used as a sheath liquid around the outside of the CE capillary to provide a stable electrical connection between the tip of the capillary and ground. ESI-MS was conducted in the positive ion mode, and the capillary voltage was set at 4000 V. A flow rate of heated dry nitrogen gas (heater temperature 300 °C) was maintained at 10 L/min. The spectrometer was scanned from m/z 50 to 350 at 1.3 s/scan during the separation and detection. In selective ion monitoring mode, protonated [M + H]+ ions were monitored for amino acids with 44 ms sampling time for each. RESULTS AND DISCUSSION Choice of Electrolyte. In CE, ionic species are separated on the basis of their charge and size; therefore, basic amino acids and acidic amino acids migrate in opposite directions at a medium pH range. To achieve the simultaneous determination of amino acids, a pH lower than 2.77 was employed. Since isoelectric points (pI) of these amino acids range from 2.77 to 10.76,29 every amino acid was charged positively below a pH value of 2.77 and thus migrated toward the cathode, which is the direction of the electrospray interface. In cases where a low pH is necessary for ESI-MS, volatile acids such as formic acid or acetic acid are commonly used. Therefore, both acids were investigated in this study. First, the effect of formic acid concentration on the 19amino acid separation was investigated. Figure 1 shows the relationship between formic acid concentration and the mobility of several amino acids. The apparent mobility, µa, for each amino acid was calculated using the following equation:

µa ) L2/tV (cm2 V-1 s-1)

where L is the length of the capillary to the detector, t is the migration time of the amino acid, and V is the applied potential.30 (28) Voyksner, R. D.; Lee, H. Anal. Chem. 1999, 71, 1441-1447. (29) Lide, D. R., Ed. CRC Handbook of Chemistry and Physics, 75th ed.; CRC Press: Boca Raton, FL, 1994; p 7-1.

Analytical Chemistry, Vol. 72, No. 6, March 15, 2000

1237

Figure 1. Effect of formic acid concentration on amino acid mobility.

Figure 2. Effect of formic acid concentration on theoretical plates.

Although mobilities were increased at a higher concentration, changing the electrolyte concentration had a negligible effect on mobility orders except for a few amino acids such as Pro and Phe. However, peak shapes were significantly influenced by changing formic acid concentration. Figure 2 illustrates the effect of formic acid concentration on the theoretical plates of some amino acids. At lower than 100 mM (pH 2.3), the peak shape of every amino acid was broadened and their theoretical plates were dramatically decreased. Especially, considerable band broadening of basic amino acids such as Lys, Arg, and His were observed at a (30) Heiger, D. N. High Performance Capillary Electrophoresis-An Introduction, 2nd ed.; Hewlett-Packard primer 12-5091-6199E, 1992; p 23.

1238

Analytical Chemistry, Vol. 72, No. 6, March 15, 2000

concentration of 25 mM (pH 2.6). Li described that silanol groups on the fused-silica capillary wall can exist in anionic form (SiO-) at pH higher than 2.31 We therefore presumed that the broadening peaks were due to some adsorption of protonated amino acids onto the capillary wall by ion exchange. On the other hand, at higher concentration, theoretical plates of most amino acids were increased, and complete resolution between Ile and Leu was obtained at 1 M (pH 1.8). Second, we investigated acetic acid as the electrolyte for the determination of 19 amino acids. Lu et al. used 10% acetic acid (pH 3.1) for the separation of four amino acids.27 However, under this condition, basic amino acids such as Arg (pI 10.76)29 and Lys (pI 9.74)29 were detected as severe broadened and split peaks. This is because silanol groups were more ionized at pH 3.1, so that basic amino acids adsorbed on the silanol groups. Furthermore, 35 min was needed to detect all the amino acids with the 10% acetic acid electrolyte. When a higher concentration of acetic acid was used to improve peak shapes by decreasing pH, analysis time were much increased as Lu described.27 These results indicated formic acid is the suitable electrolyte for amino acids because of its strong acidity, and 1 M formic acid provided better resolution and shorter analysis time. For these reasons, 1 M formic acid was used for all subsequent experiments. Evaluation of Sheath Liquid Parameters. The choice of the sheath liquid parameters is also very important in developing a method employing CE-ESI-MS. Volatile reagents such as formic acid or ammonium acetate in methanol-water solutions have been proven to be well suited sheath liquids in CE-ESI-MS.32-34 Three kinds of volatile salts, 5 mM each of formic acid, ammonium formate, and ammonium acetate, were dissolved in 50% (v/v) methanol-water, respectively, and the each sheath liquid was investigated. The signal-to-noise ratio for 100 µM each of 19 amino acids obtained by CE-ESI-MS is listed in Table 1. When 5 mM ammonium acetate was used, the highest sensitivities were obtained for 15 amino acids compared with other liquids. Especially regarding Pro, ammonium acetate provided more than 4 times better sensitivity than the others. Therefore, 5 mM ammonium acetate was selected as the volatile salt in the sheath liquid. With a constant concentration of 5 mM ammonium acetate, the influence of methanol composition in the sheath liquid was also studied at 20, 50, and 70%. As listed in Table 2, higher sensitivities for most amino acids were obtained at 50% methanol composition. Also, the effect of flow rate of the sheath liquid with 5 mM ammonium acetate in 50% (v/v) methanol-water was investigated over the range 4-10 µL/min. Although sensitivity was increased, a unexpected current drop was often observed at lower flow rate. As Foret and co-workers35 described in for a CE-ESI-MS system with limited or zero electroosmotic flow (EOF), the electric charge (31) Li, S. F. Y. Capillary Electrophoresis: Principles, practice, and applications; Journal of Chromatography Library, Elsevier Scientific Publishers: Amsterdam, 1992; p 6. (32) Deterding, L. J.; Parker, C. E.; Perkins, J. R.; Mosely, M. A.; Jorgenson, J. W.; Tomer, K. B. J. Chromatogr. 1991, 554, 329-338. (33) Pleasance, S.; Thibault, P.; Kelly, J. J. Chromatogr. 1992, 591, 325-339. (34) Sheppard, R. L.; Tong, X.; Cai, J.; Henion, J. D. Anal. Chem. 1995, 67, 2054-2058. (35) Foret, F.; Thompson, T. J.; Vouros, P.; Karger, B. L. Anal. Chem. 1994, 66, 4450-4458.

Table 1. Effect of Volatile Salt in the Sheath Liquid on the Signal-to-Noise Ratio volatile salt composition in 50% (v/v) methanol-water amino acid

5 mM formic acid

5 mM ammonium formate

5mM ammonium acetate

Gly Ala Ser Pro Val Thr Ile Leu Asn Asp Lys Glu Met His Phe Arg Tyr Trp CysCys

4 5 6 36 15 82 23 23 9 6 14 17 17 101 56 113 27 27 40

5 7 6 19 9 79 20 20 7 3 21 19 3 80 38 122 29 29 12

10 12 8 160 14 172 27 27 21 14 37 20 9 170 67 240 28 28 14

Figure 3. Positive ion ESI mass spectrum of Arg.

Table 2. Effect of Methanol Composition in the Sheath Liquid on the Signal-to-Noise Ratio

amino acid Gly Ala Ser Pro Val Thr Ile Leu Asn Asp Lys Glu Met His Phe Arg Tyr Trp CysCys a

methanol composition with constant concentration of 5 mM ammonium acetate 20%

50%

70%

nda

10 12 8 160 14 172 27 27 21 14 37 20 9 170 67 240 28 28 14

11 10 6 148 13 126 38 38 11 8 27 21 5 135 55 169 24 6 12

nd 12 8 10 36 14 14 3 10 11 9 6 27 9 20 12 6 2

nd, not determined.

transported by the ions existing one end of the separation capillary must be substituted by a charge carried by ions of the sign entering the end of the capillary opposite from the sheath liquid in order to maintain electroneutrality. Since in this CE-ESI-MS with 1 M formic acid electrolyte system, the EOF is fully diminished and formic acid in the capillary migrates toward the anode (opposite MS direction), the unexpected current drop was caused by a zone where no anionic ions existed at the capillary end. In this experiment, the higher sheath liquid flow rate provided longer term stability without current drop because anions in the sheath liquid can easily enter the capillary at higher flow rate. These results indicate 5 mM ammonium acetate in 50% (v/v) methanol-water at a flow rate of 10 µL/min was the optimum

Figure 4. CE-ESI-MS selective ion electropherograms for a standard mixture of 19 amino acids. Experimental conditions: sample concentration, Cys 125 µmol/L and others 250 µmol/L each; capillary, fused silica 50 µm i.d. × 100 cm; electrolyte, 1 M formic acid; applied potential, 30 kV; injection, 3 s at 50 mbar; temperature, 20 °C; sheath liquid, 10 µL/min of 5 mM ammonium acetate in 50% (v/v) methanolwater.

sheath liquid condition for amino acid analysis by CE-ESI-MS. Other optimized conditions were described in the Experimental Section. Determination of Amino Acids by CE-ESI-MS. Figure 3 shows a mass spectrum of Arg (molecular weight 174.2) which was acquired in the positive ion mode scanning from m/z 50 to 350. The protonated molecular ion dominated the mass spectrum, and the obtained mass spectra of other amino acids were almost similar. Figure 4 shows selected ion electropherograms of the 19amino acid standard mixture obtained by using the optimized conditions with CE-ESI-MS. Although all of the amino acids were not electrophoretically separated, they could be selectively detected at their protonated molecules on the mass spectrometer. In this study, an ion at m/z 122 for Cys was not observed but an Analytical Chemistry, Vol. 72, No. 6, March 15, 2000

1239

Table 3. Reproducibility, Linearity, and Sensitivity RSD (n ) 8) (%) amino acid

migration time

peak area

linearity correlation

detection limit (µmol/L)

Lys Arg His Gly Ala Ser Val Ile Leu Asn Met Pro CysCys Glu Trp Thr Phe Tyr Asp

0.6 0.7 0.7 0.8 0.9 1.0 1.1 1.0 1.0 1.1 1.1 1.1 1.1 1.1 1.2 1.2 1.2 1.2 1.2

2.9 2.5 3.2 3.9 3.1 2.8 4.3 2.0 2.0 3.9 6.5 2.4 3.4 3.5 4.7 2.4 2.7 3.0 3.8

0.9992 0.9962 0.9978 0.9982 0.9992 0.9993 0.9996 0.9992 0.9992 0.9988 0.9987 0.9964 0.9980 0.9990 0.9981 0.9991 0.9990 0.9993 0.9999

1.1 0.3 0.3 10 5.9 5.6 3.4 1.3 1.3 3.8 3.0 0.4 0.7 6.0 3.8 0.8 0.8 1.6 11

ion at m/z 241 for Cys-Cys appeared. As Cys is readily oxidized to Cys-Cys by air,36 it is presumed that the disappearance of Cys and appearance of Cys-Cys was caused by oxidation. The reproducibility, linearity, and sensitivity of the method for 19 amino acids were tested and the results are listed in Table 3. Practical reproducibilities were obtained for all amino acids with RSD values (n ) 8) for migration times better than 1.2% and for peak areas between 2.0 and 4.7% except for Met. Only Met showed decreasing peak areas during the runs because Met tends to be oxidized. The calibration curves for all species were linear at 10, 25, 50, 100, 250, and 500 µmol/L with correlation coefficients between 0.9962 and 0.9999. The signals of several amino acids such as Arg, His, and Pro were a little bit saturated at higher concentrations; thus, their calibration curves were slightly nonlinear at higher concentrations. Sensitivity differences were observed between basic and acidic amino acids. The detection limits for basic amino acids such as Arg, His, and Lys were between 0.3 and 1.1 µmol/L, while those for acidic Glu and Asp were 6.0 and 11 µmol/L, respectively, with pressure injection of 50 mbar for 3 s (3 nL) at a signal-to-noise ratio of 3. These results were reasonable because basic compounds are more sensitive in the positive ion mode. Also, the sensitivities of low molecular weight compounds such as Gly and Ala were lower compared to other amino acids. The use of this CE-ESI-MS method resulted in ∼100-fold increase in sensitivity against a CE method with indirect UV detection.17 On the other hand, compared with HPLC derivatization techniques, this method was equal to or better in sensitivity than UV detection, but 10100-fold less sensitive than fluorescence detection (FLD) with OPA reagent.4 Although sensitivity in the CE-ESI-MS method is inferior to that in HPLC with the OPA method, CE-ESI-MS enables us to analyze amino acids without derivatization. Figure 5 illustrates the result from several physiological amino acids obtained by this method. Every physiological amino acid including β-Ala, sarcosine (Sar), γ-aminobutyric acid (γ-ABA), hydroxyproline (Hypro), ornithine (Orn), and R-aminoadipic acid (R-AAA), (36) Windholz, M. The Merck Index, 10th ed.; Merck: Rahway, NJ, 1983; p 2773.

1240 Analytical Chemistry, Vol. 72, No. 6, March 15, 2000

Figure 5. CE-ESI-MS selected ion electropherograms for a standard mixture of physiological amino acids. Experimental conditions: sample concentration, Sar 625 µmol/L, a-AAA, Cysthi each 125 µmol/L and others 250 µmol/L each; other conditions are the same as in Figure 4.

Figure 6. Selected ion electropherograms for amino acids in soy sauce. Experimental conditions are the same as in Figure 4.

plus allo-δ-hydroxylysine (Allo hylys), 1-methylhistidine (Mehis), citrulline (Cit), cystathionine (Cysthi), carnosine (Car), and anserine (Ans) could be determined, and every isobaric amino acid such as β-Ala, Ala, and sarcosine (Sar) was well separated at its protonated molecules. Since this CE-ESI-MS method is considerably selective, it can be useful for free amino acid analysis in food and physiological samples, which contain various kinds of matrix compounds. Application to the Analysis of Soy Sauce. The developed method was applied to the determination of amino acids in soy sauce. Soy sauce is a very complex mixture that contains many kinds of amino acids, and their analysis is important because

measurement of their concentrations can help track metabolic products of fermentation and correlate flavor trends. Figure 6 shows selected ion electropherograms of a soy sauce analysis obtained by the CE-ESI-MS method. Although say sauce contains a great number of organic compounds, well-defined selected ion electropherograms were obtained without interference from other matrix compounds. With the HPLC method, since amino functional groups of every component in soy sauce are derivatized and detected, their identification and resolution are problematic. However, in the CE-ESI-MS method, primarily the only the amino acid of interest or the amino acid with a few additional peaks could be detected. Therefore, the reliability of identification of the method is superior to other techniques. In addition, sample preparation was simple and consisted only of diluting the soy sauce 1:100 with Milli-Q water and centrifugal filtering through a Millipore 30-kDa-cutoff filter to remove proteins and peptides. Satisfactory reproducibilities were obtained for all amino acids with RSD values (n ) 5) for migration times better than 0.4% and for peak areas between 1.1 and 6.0% except for Met. Additional samples including beer, sake, and human urine were determined by this method, and most amino acids could be determined with good selectivity and sensitivity as well as soy sauce (data not shown). More than 300 samples were analyzed by this method over a one-month period without any instrument maintenance and with no decrease in sensitivity.

CONCLUSIONS A simple, rapid, and selective CE-ESI-MS method for the determination of underivatized amino acids has been developed. Compared with other developed techniques, this method has a big advantage: (1) every type of amino acid can be analyzed without derivatization in a short analysis time, (2) sensitivity is relatively high, (3) amino acids can be selectively determined without other matrix interference, and (4) sample preparation is minimal. The methodology provides excellent reproducibility, good linearity, and excellent long-term stability. Its utility was demonstrated by the analysis of soy sauce, beer, sake, and urine. These results indicate that the proposed method can be useful for not only the 19 naturally occurring amino acids but also for physiological amino acids. Therefore it is expected that the method can be applied to wide areas such as the analysis of amino acids in foods, their composition analysis of hydrolyzed protein, and the analysis of physiological fluids to diagnose amino acid metabolism abnormalities.

Received for review August 25, 1999. Accepted December 30, 1999. AC990976Y

Analytical Chemistry, Vol. 72, No. 6, March 15, 2000

1241