Development and Evaluation of a GC/FID Method for the Analysis of Free Amino Acids in Quince Fruit and Jam

Transcription

1 ANALYTICAL SCIENCES SEPTEMBER 2003, VOL The Japan Society for Analytical Chemistry 1285 Development and Evaluation of a GC/FID Method for the Analysis of Free Amino Acids in Quince Fruit and Jam Branca M. SILVA,* Susana CASAL,** Paula B. ANDRADE,* Rosa M. SEABRA,* M. Beatriz OLIVEIRA,** and Margarida A. FERREIRA** *REQUIMTE, Serviço de Farmacognosia, Faculdade de Farmácia, Universidade do Porto, R. Aníbal Cunha, Porto, Portugal **REQUIMTE, Serviço de Bromatologia, Faculdade de Farmácia, Universidade do Porto, R. Aníbal Cunha, Porto, Portugal A GC/FID methodology for determination of twenty-one free amino acids in quince fruit (pulp and peel) and jam is described. The sample preparation was simple, involving a SCX Solid-Phase Extraction (SPE) purification step and a fast derivatization with ethyl chloroformate for gas chromatographic analysis. The chromatographic separation was achieved using a CP-Sil 19 CB wcot fused-silica capillary column. Under the chosen conditions, with temperature and pressure programming, this capillary column was able to separate all the amino acids not only in a short time but also with good separation. The GC/FID procedure is rapid, sensitive, reproducible and accurate. The detection limit values for amino acids were low, between and µg/ml, and the method was precise. As a general rule, the recovery values were high. Due to its rapidity and low cost, this technique can be useful in the quality control of quince products. (Received April 21, 2003; Accepted June 11, 2003) s are present in almost all foods, so a foodstuff may often be characterized by its relative amounts (fruit juices, syrups, jams and jellies; meat and meat products; wines). 1 In plants and plant materials, in fruit and their products such as juices, jams and jellies, they are found in the free form, mainly in the L-configuration. The applications of the amino acids analysis are nearly unlimited and may roughly be divided into three main groups: 1 determination of the chemical score of food proteins; influence of technological processes and additives on that score; detection of fraudulent manipulations or economically profitable falsifications, especially in fruit products. Generally, the analysis of amino acids in food matrices is done by liquid chromatography. 2 Gas chromatography (GC) has, however, some advantages, given its capillary flexibility, higher resolution, speed of analysis and even its instrumental cost that is about one third of that of HPLC. Nevertheless, GC analysis of amino acids require a derivatization step, which commonly involves laborious and multi-step procedures, and this limitation greatly diminishes the above-cited advantages of GC. 3 This problem was solved by Husek, 3 6 who developed rapid derivatization procedures based on the treatment of amino acids with chloroformates. Quince is the fruit of a deciduous tree of the Rosaceae family, Cydonia oblonga Miller. This fruit is too acid, astringent, and tough to be consumed fresh. However, it can be consumed when cooked or processed as jam or jelly, two food products with long tradition in Portugal. Quince jam is homemade or industrially manufactured, by boiling a mixture of sugar and quince puree until a convenient texture is obtained (usually to To whom correspondence should be addressed. rseabra@ff.up.pt reach Brix). When quince production is scarce, industry manufacturers are tempted to adulterate quince jam by adding apples and/or pears due to their low cost. For quince fruit and its derivatives, few studies have been developed. In 1979 and 1986, the volatile constituents of quince fruit were analyzed by GC/MS. 7,8 Later, the usefulness of phenolic compounds in the determination of genuineness of quince puree, 9 jam 10,11 and jelly 12 has been reported. Glucosides of procyanidin polymers have been previously identified in this fruit 13,14 and, recently, it was possible to discriminate quince pulp and peel by the analysis of phenolic compounds in quince fruit. 15 In 2002, an HPLC/UV method was developed for the determination of organic acids in quince fruit and its jam. 16 As far as we know, there is no study about amino acids composition of quince fruit and jam. So, once all fruits show a typical free amino acids pattern, 1 the aim of the present study was to develop an GC analytical technique for the qualitative and quantitative analysis of free amino acids in quince fruit (pulp and peel) and jam. Experimental Samples and standards The L-amino acids and the internal standard (L-pchlorophenylalanine) were all from Sigma (St. Louis, MO, USA). Ethyl chloroformate (ECF) was from Aldrich (Steinheim, Germany) and pyridine from Fluka (Neu-Ulm, Germany). All other chemicals were of analytical grade from several suppliers. Healthy quince fruit samples were collected in Northern Portugal. The fruits were separated into pulp and peel. Each part of the fruit was cut into thin slices and freeze-dried.

2 1286 ANALYTICAL SCIENCES SEPTEMBER 2003, VOL. 19 Lyophilizations were carried out using a Labconco 4.5 apparatus (Kansas City, MO). Quince jam samples were purchased on the Portuguese market. SPE cartridges The benzenesulfonic SCX Spe-ed SPE cartridges (200 mg; 3 ml) were obtained from Applied Separations (Allentown, USA). Extraction of free amino acids Each sample (ca. 1.5 g for freeze-dried quince pulps and peels and 5 g for quince jams) was thoroughly mixed with 3 25 ml of acid water (ph 2.2 with HCl 0.1 M), at room temperature, with magnetic stirring for 3 10 min. The extracts were gathered, filtered and passed through a SCX cartridge, previously conditioned with 10 ml of methanol and 10 ml of HCl 5 mm. The amino acids were eluted with a mixture of ammonia (4 M) and methanol (50:50 v/v) (3 500 µl). To each extract, 150 µl of L-p-chlorophenylalanine solution (10 µl/ml) (internal standard) were added. The obtained solutions were dried under N 2 stream and kept below 0 C until derivatization. Derivatization procedure The derivatization was achieved as previously reported. 3 6 Each dried residue was dissolved in 60 µl of water and 40 µl of ethanol/pyridine (4:1); 5 µl ECF were added and the sample was vortex mixed (3 5 s). Gas evolution (carbon dioxide) usually occurs. Five minutes later, 150 µl of dichloromethane and ca g of NaCl were added and the vial was thoroughly shaken for extraction of the derivatives into the organic layer. This phase was transferred into a 200 µl insert adjustable to the liquid sampler vials. About 1.5 µl was injected into the gas chromatographic system. GC analysis Separation of L-amino acids was achieved by gas chromatography, carried out with a Chrompack CP 9001 instrument (Chrompack, Middelburg, The Netherlands) equipped with a flame ionization detector (FID), and an automatic liquid sampler (CP-9050, Chrompack). The injector was kept at 250 C and the detector at 280 C. The GC was equipped with electronic pressure control allowing programmable gas pressure during the chromatographic run. Helium as carrier gas was used with the following pressure program: increase from initial 50 kpa to 70 kpa. A CP-Sil 19 CB (10 m 0.25 mm i.d.) wcot fused-silica capillary column (Varian) was used with the following temperature program: increase from 140 C (1 min hold) to 280 C, at 40 C/min. The compounds were identified by their retention times and chromatographic comparison with authentic standards. Quantification was based on the internal standard method using L-p-chlorophenylalanine. Results and Discussion Extraction method and SPE purification The extraction method was studied, based on published methodologies applied to other matrices 1,2,17,18 (fruit juices and jams, honeys and coffee beans), in order to obtain the highest recoveries and cleaner chromatograms. Acid water 1,2,17 (ph 2.2 with HCl 0.1 M), hydrochloric acid 0.1 M, 1,2 methanol 70% 1,2 and sulfosalicylic acid 2% 2,18 were tested on quince fruit (pulp and peel) and jam. Acid water (ph 2.2 with HCl 0.1 M) Table 1 Concentration range of linearity and detection limits for amino acids Linearity/ Detection limit/ µg ml 1 µg ml 1 Alanine Glycine Valine Leucine Isoleucine Proline Threonine Serine Glutamic acid Asparagine Aspartic acid Methionine Hydroxyproline Phenylalanine Cysteine Glutamine Ornithine Lysine Histidine Tyrosine Tryptophan provided the best recoveries (data not shown). In order to remove interference compounds, a strong cation exchange (SCX) Solid-Phase Extraction (SPE) purification step was needed. With the purpose of eliminating any disturbance compounds, after charging the sample solution to the cartridge, we have tried to wash the SCX cartridge with water and with acid water (ph 2.2). We have observed that better recoveries were obtained without washing the cartridge. It seems that some losses of amino acids occurred during the washing step. The addition of internal standard was tested before and after the SCX step. When the internal standard was added after SPE purification, more accurate results were obtained. Derivatization procedure Since none of the amino acids is volatile enough for direct GC analysis, it is necessary to transform them into volatile derivatives. The chosen amino acids derivatization procedure with ethyl chloroformate is unique in rapidity, although one drawback is the inability to determine arginine, as already described by Husek. 3 6 Not all reactive groups in the amino acids are altered by action of the reagent. The imino group of arginine remains untouched, which is the reason for absorption of this amino acid derivative in the column. 3 6 For the determination of arginine, an additional reaction step would be necessary or, alternatively, its conversion into ornithine by arginase before derivatization. 18 GC analysis Separation of twenty-one amino acids was achieved by gas chromatography, with an analysis time of only 6 min, with good resolution, and low reagent and instrumentation costs. Analytical curves and detection limits Calibration curves were determined after subjecting standards to the same total procedure in order to compensate the losses during extraction, SPE clean-up and derivatization steps. Under the assay conditions described, a linear relationship between the

3 ANALYTICAL SCIENCES SEPTEMBER 2003, VOL Table 2 Free amino acids composition of quince fruits (pulp and peel) and quince jams (µg kg 1 ) a (quantification by internal standard technique) Fruit Jam Pulp Peel A B A B A B Alanine 12.9(1.05) 62.3(2.05) 14.8(0.51) 70.1(1.33) 68.1(1.05) 10.1(0.61) Glycine 67.8(3.18) 177.9(5.64) 13.6(0.34) 355.0(6.32) 65.3(1.92) 45.2(2.57) Valine 13.8(0.64) 47.9(1.21) 13.6(0.45) 63.0(1.21) 19.5(0.70) 6.4(0.16) Leucine 4.8(0.16) 15.8(1.08) 5.0(0.08) 18.3(0.28) 5.9(0.11) 3.6(0.17) Isoleucine 13.7(0.26) 70.2(2.97) 15.4(0.70) 125.4(2.20) 23.4(0.45) 12.5(0.48) Proline 5.0(0.21) 15.6(0.59) 5.8(0.25) 24.9(0.43) 6.2(0.12) 3.1(0.02) Threonine 2.6(0.22) 37.6(0.70) 7.3(0.29) 51.8(0.68) 7.7(0.23) 9.9(0.39) Serine 9.8(0.76) 77.9(6.49) 14.7(1.40) 145.6(2.10) 16.3(0.27) 10.3(0.53) Glutamic acid 41.7(2.58) 84.7(1.29) 51.7(1.81) 231.7(5.17) 25.0(1.13) 16.0(0.79) Asparagine 90.7(2.72) 130.6(3.56) 139.1(12.5) 112.6(2.15) 219.0(7.71) 65.9(2.31) Aspartic acid 79.2(5.20) 163.7(5.72) 77.4(2.58) 252.7(6.12) 51.0(1.13) 33.7(0.94) Methionine 0.4(0.03) 1.4(0.02) 2.2(0.06) 0.4(0.02) 0.4(0.02) 0.3(0.01) Hydroxyproline 16.7(1.33) 24.8(1.73) 29.7(1.90) 173.0(3.73) 41.5(0.65) 27.9(0.74) Phenylalanine 8.0(0.09) 13.3(0.05) 11.0(0.25) 13.7(0.26) 4.0(0.12) 6.7(0.05) Cysteine 5.8(0.22) 27.6(1.37) 30.3(1.01) 34.6(0.45) 0.4(0.01) 7.6(0.68) Glutamine 2.9(0.21) 20.3(0.40) 21.9(0.97) 27.2(0.63) 5.8(0.23) 4.8(0.32) Ornithine 1.2(0.09) 8.2(0.96) 0.9(0.08) 6.0(0.09) 4.8(0.30) 3.7(0.33) Lysine 12.5(0.33) 38.0(2.05) 22.8(0.94) 49.6(0.98) 14.9(0.35) 8.9(0.38) Histidine 115.6(7.21) 15.6(0.22) 85.4(3.61) 54.5(0.98) 10.7(0.61) 27.0(0.61) Tyrosine 1.1(0.06) 3.4(0.11) 0.5(0.04) 5.6(0.09) 1.9(0.10) 1.1(0.06) Tryptophan 20.1(0.86) 7.9(0.24) 53.3(1.01) 4.2(0.09) 1.3(0.05) 4.6(0.24) a. Values are expressed as mean (SD) of three determinations. Fig. 1 Free amino acids profile of a quince pulp: (1) alanine, (2) glycine, (3) valine, (4) leucine, (5) isoleucine, (6) proline, (7) threonine, (8) serine, (9) glutamic acid, (10) asparagine, (11) aspartic acid, (12) methionine, (13) hydroxyproline, (14) phenylalanine, (15) cysteine, (IS) internal standard (L-p-chlorophenylalanine), (16) glutamine, (17) ornithine, (18) lysine, (19) histidine, (20) tyrosine and (21) tryptophan. Fig. 2 Free amino acids profile of a quince peel. (1) (12) are the same as in Fig. 1. concentration of amino acids and the FID response was obtained in the tested range. The correlation coefficient for the standard curves invariably exceeded 0.99, for all the compounds. The detection limit values were calculated as the concentration corresponding to three times the standard deviation of the background noise, and the values obtained ranged from to µg/ml (Table 1). Validation of the method This method was specially developed for the determination of amino acids in quince jams, fruit products with high amounts of sugar and other interfering compounds, some of them produced during thermal processing. Since quince pulps and peels are less complex matrices, the developed technique was also applied to them (we only tested the accuracy of the procedure Fig. 3 Free amino acids profile of a quince jam. (1) (12) are the same as in Fig. 1. for these matrices). The free amino acids from quince fruit (pulp and peel) and quince jam samples were analyzed by the proposed technique in order to validate this procedure and to assess its applicability to the routine free amino acid analysis of these food products (Table 2). Typical chromatograms obtained with quince pulp, peel and jam samples are represented in Figs. 1, 2 and 3, respectively. The retention times (RT) obtained for amino acids

4 1288 ANALYTICAL SCIENCES SEPTEMBER 2003, VOL. 19 Table 3 Retention times of the amino acids (n = 3) Retention time a /min Alanine Glycine Valine Leucine Isoleucine Proline Threonine Serine Glutamic acid Asparagine Aspartic acid Methionine Hydroxyproline Phenylalanine Cysteine Internal standard Glutamine Ornithine Lysine Histidine Tyrosine Tryptophan a. Mean of the retention times; SD, standard deviation. SD/min Table 5 Evaluation of the analytical method reproducibility (n = 6) (quantification by internal standard technique) SD/µg kg 1 CV, % Alanine Glycine Valine Leucine Isoleucine Proline Threonine Serine Glutamic acid Asparagine Aspartic acid Methionine Hydroxyproline Phenylalanine Cysteine Glutamine Ornithine Lysine Histidine Tyrosine Tryptophan Table 4 Evaluation of the analytical method precision (n = 6) (quantification by internal standard technique) SD/µg kg 1 CV, % Alanine Glycine Valine Leucine Isoleucine Proline Threonine Serine Glutamic acid Asparagine Aspartic acid Methionine Hydroxyproline Phenylalanine Cysteine Glutamine Ornithine Lysine Histidine Tyrosine Tryptophan are shown in Table 3. The precision of the analytical method was evaluated by measuring the peak chromatographic area of amino acids six times on the same quince jam sample. The analytical method is precise, once the coefficients of variation of free amino acids were low (between 0.25 and 5.22%) (Table 4). The reproducibility of the method was evaluated by the calculation of the coefficient of variation for each free amino acid of six repeated extractions of the same quince jam sample. Considering that we are dealing with food matrices with low free amino acids content, in a general way, we found that the coefficient of variation values were low (Table 5) and the method is reproducible. For the cases where coefficients were higher, they, generally, correspond to the amino acids that were present in very low amounts, such as methionine. However, no linear relation between the content and the coefficient of variation was observed, but in a general way coefficients higher than 6.5 were found for amino acids in amounts below 5 µg/kg. The accuracy (% recovery) of the procedure was evaluated in triplicate using the same quince fruit (pulp and peel) and quince jam samples spiked with known standard amounts. Generally, the recovery values were high (Tables 6, 7 and 8), which demonstrates the effectiveness of the extraction and the accuracy of the method. In conclusion, the proposed GC/FID procedure for free amino acid profile determination is simple, sensitive, reproducible and accurate. This method has the main advantage of being very rapid, especially in terms of derivatization process and run time, which makes it suitable for routine analysis of amino acids in quality control determinations of quince jam and fruit. Its only limitation is the inability to determine arginine. In future studies, we will apply this method in the determination of free amino acid profiles of quince jams of several homemade and commercially available samples, quince fruits (pulps and peels) from several geographical origins of Portugal, during three consecutive years, in order to test if this pattern can be useful in the evaluation of quince jam authenticity. Acknowledgements Branca M. Silva is grateful to Fundação para a Ciência e a Tecnologia for a grant (PRAXIS XXI/BD/21339/99).

5 ANALYTICAL SCIENCES SEPTEMBER 2003, VOL Table 6 sample a Recoveries of free amino acids from a spiked quince jam Table 7 pulp a Recoveries of free amino acids from a spiked quince Present/ Added/ Found/ SD/ µg kg 1 µg kg 1 µg kg 1 µg kg 1 CV, % Recovery, % Alanine Glycine Valine Leucine Isoleucine Proline Threonine Serine Glutamic acid Asparagine Aspartic acid Methionine Hydroxyproline Phenylalanine Cysteine Glutamine Ornithine Lysine Histidine Tyrosine Tryptophan a. Mean value found for three assays for each studied concentration; Present/ Added/ Found/ SD/ CV, % Recovery, µg kg 1 µg kg 1 µg kg 1 µg kg 1 % Alanine Glycine Valine Leucine Isoleucine Proline Threonine Serine Glutamic acid Asparagine Aspartic acid Methionine Hydroxyproline Phenylalanine Cysteine Glutamine Ornithine Lysine Histidine Tyrosine Tryptophan a. Mean value found for three assays for each studied concentration; Table 8 pulp a References Recoveries of free amino acids from a spiked quince Present/ Added/ Found/ SD/ CV, % Recovery, µg kg 1 µg kg 1 µg kg 1 µg kg 1 % Alanine Glycine Valine Leucine Isoleucine Proline Threonine Serine Glutamic acid Asparagine Aspartic acid Methionine Hydroxyproline Phenylalanine Cysteine Glutamine Ornithine Lysine Histidine Tyrosine Tryptophan a. Mean value found for three assays for each studied concentration; 1. W. Ooghe, Authenticity and Adulteration of Food The Analytical Approach-Proceedings of Eurofood Chem. IX, 1997, Switzerland, J. H. Baxter, in Handbook of Food Analysis, ed. L. M. L.

6 1290 ANALYTICAL SCIENCES SEPTEMBER 2003, VOL. 19 Mollet, 1996, Ghent, Belgium, P. Husek, J. Chromatogr., 1991, 552, P. Husek, Febs Lett., 1991, 280, P. Husek, LC-GC INTL, 1992, 5, P. Husek, J. Chromatogr. B, 1998, 717, L. Schreyen, P. Dirinck, P. Sandra, and N. Schamp, J. Agric. Food Chem., 1979, 27, K. Umano, A. Shoji, Y. Hagi, and T. Shibamoto, J. Agric. Food Chem., 1986, 34, P. B. Andrade, A. R. F. Carvalho, R. M. Seabra, and M. A. Ferreira, J. Agric. Food Chem., 1998, 46, B. M. Silva, P. B. Andrade, G. C. Mendes, P. Valentão, R. M. Seabra, and M. A. Ferreira, J. Agric. Food Chem., 2000, 48, B. M. Silva, P. B. Andrade, R. M. Seabra, and M. A. Ferreira, J. Liq. Chromatogr. Relat. Technol., 2001, 24, B. M. Silva, P. B. Andrade, P. Valentão, G. C. Mendes, R. M. Seabra, and M. A. Ferreira, Food Chem., 2000, 71, J.-J. Macheix, A. Fleuriet, and J. Billot, in Fruit phenolics, ed. CRC Press, Inc., 1990, Florida, USA, L. J. Porter, L. Y. Foo, and R. H. Furneaux, Phytochemistry, 1985, 24, B. M. Silva, P. B. Andrade, F. Ferreres, A. L. Domingues, R. M. Seabra, and M. A. Ferreira, J. Agric. Food Chem., 2002, 50, B. M. Silva, P. B. Andrade, G. C. Mendes, R. M. Seabra, and M. A. Ferreira, J. Agric. Food Chem., 2002, 50, A. Pririni, L. Conte, O. Francioso, and G. Lercker, J. High Resol. Chromatogr., 1992, 15, S. Casal, M. B. Oliveira, and M. A. Ferreira, J. Chromatogr. A, 2000, 866, 221.