Anal. Chem. 1997, 69, 2217-2220
Simultaneous Separation of Soya Bean and Animal Whey Proteins by Reversed-Phase High-Performance Liquid Chromatography. Quantitative Analysis in Edible Samples Ma Concepcio´n Garcı´a,†,‡ Ma Luisa Marina,†,‡ and Mercedes Torre*,‡
Centro de Tecnologı´a de los Alimentos y Servicios Biosanitarios, Universidad de Alcala´ de Henares, Ctra. Madrid-Barcelona Km. 33.600. 28871 Alcala´ de Henares, Madrid, Spain, and Departamento de Quı´mica Analı´tica, Facultad de Ciencias, Universidad de Alcala´ de Henares, Ctra. Madrid-Barcelona Km. 33.600. 28871 Alcala´ de Henares, Madrid, Spain
A reversed-phase high-performance liquid chromatography method was developed successfully for the simultaneous and rapid separation of the main soya bean proteins (globulins) and animal whey proteins (r-lactalbumin and β-lactoglobulin). This method consisted of a linear gradient of two mobile phases (0.1% trifluoroacetic acid in water and 0.1% trifluoroacetic acid in acetonitrile). For the first time, soya bean globulins were separated from animal whey proteins. The analysis time for this separation was 20 min, eluting soya bean globulins in the first 10 min. The method was applied to quantify soya bean proteins in soya bean commercial milks with the possibility of detecting the presence of animal whey proteins. Adulteration of a powdered soya bean milk by animal whey proteins was detected. The possibility of detecting the presence of soya bean proteins in animal milks was also studied. The need to look for new sources of proteins as a result of the increasing demand for animal proteins such as milk and meat by the rising world population has extended the use of soya bean proteins for human consumption. These vegetable proteins constitute an interesting alternative to animal proteins due to their nutritional value (48-50% of proteins) and low cost.1-3 Another factor that increases interest in these vegetable proteins is that they have been used to replace the consumption of milk proteins from animal species for those individuals, especially infants, who are allergic to these animal proteins.4 All these properties have promoted the appearance of a great number of commercial products derived from soya bean such as milk (liquid and powder), infant formulas, creams, shake, and yogurtlike in addition to the traditional soya bean seeds, textured or flour. Also, soya bean proteins have been used as additives to enhance the quality of †
Centro de Tecnologı´a de los Alimentos y Servicios Biosanitarios. Departamento de Quı´mica Analı´tica. (1) Steinke, F. H. In New Protein Foods in Human Health: Nutrition, Prevention, and Therapy; Waggle, D. H., Steinke, F. H., Volgarev, M. N., Eds.; CRC Press: Boca Raton, FL, 1992; pp 59-66. (2) Henley, E. C.; Steinke, F. H.; Waggle, D. H. In Proceedings of the World Conference on Oilseed Technology and Utilization, 1992; Applewhite, T. H., Ed.; Am. Oil Chem. Soc.: Champaign, IL, 1993; pp 248-256. (3) Garcı´a, M. C.; Torre, M.; Marina, M. L.; Laborda, F. CRC Crit. Rev. Food Sci. Nutr., in press. (4) Ladodo, K. S.; Borovik, T. E. In New Protein Foods in Human Health: Nutrition, Prevention, and Therapy; Waggle, D. H., Steinke, F. H., Volgarev, M. N., Eds.; CRC Press: Boca Raton, FL, 1992; pp 85-89. ‡
S0003-2700(96)00843-8 CCC: $14.00
© 1997 American Chemical Society
other non vegetable foods, such as dairy and bakery products, fish, and meat.5-10 The use of soya bean proteins as supplement in bovine milk is forbidden in many countries, and in some of them government regulations specify maximum allowance levels of these derivatives in meat products. On the other hand, interest has been shown in detecting the presence of animal whey proteins (R-lactalbumin and β-lactoglobulin) that cause allergy in some individuals. Thus, to detect possible adulteration, the development of analytical methods for the detection of low levels of animal whey proteins in dairylike soya bean products or vice versa is essential. The existing analytical methods for soya bean proteins were developed in order to separate and analyze the storage proteins (globulins) of soya bean. A soya bean protein isolate contains ∼90% globulins. Their major components are glycinin (11S) (isoelectric point 6.4), and β- and γ-conglycinin (7S) (isoelectric point 4.8), which represent about 40, 28, and 3% of the total storage proteins, respectively.11 Characterization of these proteins was initially performed by inmunoelectrophoresis, disc electrophoresis, gel filtration chromatography, ion-exchange chromatography, and sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE).12-16 More recently, high-performance liquid chromatography (HPLC) has also been used. Reversed-phase (RPHPLC)17-20 and size exclusion (SE-HPLC)19,21,22 modes have been generally used. These techniques enable the separation of soya bean proteins in analysis times ranging from 35 to 90 min. These (5) Chronakis, I. S.; Kasapis, S. Food Hydrocolloids 1993, 7, 459-478. (6) Keeton, J. T. Food, Fats and Health; Council for Agricultural Science and Technology (CAST). Task Force Report No. 118, 1991; pp 42-51. (7) Camire, M. E.; King, C. C. J. Food Sci. 1991, 56, 760-763. (8) Keeton, J. T. Meat Sci. 1994, 36, 261-276. (9) Kashlan, N. B.; Hassan, S. A.; Srivastava, V. P.; Mohana, N. A.; Shubber, K. M. Food Chem. 1991, 42, 57-64. (10) Yadav, V. D.; Jha, Y. K.; Garg, S. K.; Mital, B. K. Aust. J. Dairy Technol. 1994, 49, 34-38. (11) Fukushima, D. Food Rev. Int. 1991, 7, 323-351. (12) Dreau, D.; Larre, C.; Lalles, J. P. J. Food Sci. Technol. 1994, 31, 489-493. (13) Thanh, V. H.; Okubo, K.; Shibashaki, K. Plant Physiol. 1975, 56, 19-22. (14) Koshiyama, I. Agric. Biol. Chem. 1972, 36, 2255-2257. (15) Catsimpoolas, N.; Meyer, E. W. Arch. Biochem. Biophys. 1968, 125, 742750. (16) Petruccelli, S.; An ˜o´n, M. C. J. Agric. Food Chem. 1995, 43, 1762-1767. (17) Peterson, R. E.; Wolf, W. J. J. Chromatogr. 1988, 444, 263-268. (18) Peterson, R. E.; Wolf, W. J. Cereal Chem. 1992, 69, 101-104. (19) Oomah, B. D.; Voldeng, H.; Fregeau-Reid, J. A. Plant Foods Hum. Nutr. 1994, 45, 251-263. (20) Ashoor, S. H.; Stiles, P. G. J. Chromatogr. 1987, 393, 321-328. (21) Cole, K. D.; Cousin, S. L., Jr. J. Agric. Food Chem. 1994, 42, 2713-2720. (22) Cattaneo, T. M. P.; Feroldi, A.; Toppino, P. M.; Olieman, C. Neth. Milk Dairy J. 1994, 40, 225-234.
Analytical Chemistry, Vol. 69, No. 11, June 1, 1997 2217
separations have been achieved after isolation of soya bean proteins generally from soya beans or soya bean flour by different fractionation methods.13,23,24 Applications include development of a quantitative method with potential for assessing soya bean cultivars based on protein content19 and determination of soya bean proteins in unheated meats20 and cheeses.22 However, as far as we know, there is no reference in the literature on the application of chromatographic methods either to the separation of soya bean proteins in vegetable milks or to the simultaneous separation of these proteins from animal whey proteins. This work is aimed at developing a RP-HPLC method for the rapid and simultaneous separation of soya bean and animal whey proteins. This method would allow both the quantitation of soya bean proteins in vegetable milks and the detection of the presence of animal whey proteins in these samples, thus establishing a rapid and suitable method to detect adulterations in commercial vegetable milks. One other goal is the utilization of this method to discover the presence of soya bean proteins in animal milks. EXPERIMENTAL SECTION Chemicals and Samples. Acetonitrile (ACN) (HPLC grade; Scharlau, Barcelona, Spain), trifluoroacetic acid (TFA) (HPLC grade; Pierce Europe, Ond Beijerland, Netherlands), and HPLC grade water (Milli-Q system; Millipore, Bedford, MA) were used in the preparation of mobile phases. The soya bean protein isolate and the β-lactoglobulin (β-LG) (A + B) from bovine milk were obtained from ICN (Aurora, OH), and R-lactalbumin (R-LA) from bovine milk was from Sigma (St. Louis, MO). Standard protein solutions were freshly prepared in water. The bovine and soya bean milks were purchased from local markets in Alcala´ de Henares, Madrid, Spain. Total protein content of soya bean milks was determined by an AOAC (1990) method.25 Samples of approximately 26 and 4 mg/mL (in the case of powdered soya bean milks) were prepared by diluting the appropriate amount of dairy product in water, without further treatment. Wheys from animal and soya bean milks were prepared by acidic precipitation at pH 4.6 (2 mol/L HCl) of caseins (bovine milk) and the major fraction of globulins (soya bean milks) and subsequent centrifugation to remove the supernatant.26 Centrifugation was performed at 2000g for 20 min in a Sorvall RC-5B refrigerated superspeed centrifuge (Du Pont Instruments, Newtown, CT). In the case of the powdered soya bean milks, the precipitation was carried out in a solution freshly prepared corresponding to a glass of milk (∼12 g of soya bean milk in 250 mL of water).26 The wheys were injected directly without dilution. All sample solutions were filtered through 0.22 µm disposable sterile polysulfone filters (Alltech; Alltech Associates, Deerfield, IL) before injection. Samples and standards were stored at 3 °C or frozen at -70 °C, when appropriate, until use. Sample solutions were prepared in the day and kept on ice until use. High-Performance Liquid Chromatography. A HewlettPackard 1090 Series II liquid chromatograph (Hewlett-Packard, Pittsburgh, PA) equipped with a diode array detector and a HP 9153C data acquisition system was used. Proteins were detected at 245 nm. The injection volume was 20 µL. The separation was (23) Koshiyama, I. Agric. Biol. Chem. 1965, 29, 885-887. (24) Wolf, W. J.; Babcock, G. E.; Smith, A. K. Arch. Biochem. Biophys. 1962, 99, 265-274. (25) AOAC Official Methods of Analysis, 15th ed.; Association of Official Analytical Chemists, Arlington, VA, 1990. (26) International Dairy Federation, Questionnaire 3293/E, October 1993.
2218
Analytical Chemistry, Vol. 69, No. 11, June 1, 1997
carried out with a PLRP-S column (150 × 4.6 mm i.d.) from Polymer Laboratories LTD (Church Strutton, U.K.), packed with polystyrene divinylbenzene beads (300 Å, 8 µm particle size). The column’s dead time (1.75 min) was determined by using uracyl as nonretained solute. The RP-HPLC method was performed with a linear binary gradient of six steps: 20% B for 1 min, 20-42% B in 15 min, 42-46% B in 4 min, 46-100% B in 0.5 min, 100% B for 0.5 min, and finally from 100 to 20% B in 0.5 min in order to re-equilibrate the column to starting conditions. The flow rate was 1 mL/min and 50 °C temperature was used. The mobile phases used were as follows: phase A, 0.1% TFA in water; phase B, 0.1% TFA in ACN. The mobile phases were filtered through 0.45 µm nylon filters and degassed by sparing with helium. Calibration. Using the external standard method, the chromatographic system was calibrated individually with solutions that contained soya bean protein isolate (0.5-6 mg/mL), R-LA (0.020.30 mg/mL), and β-LG (0.07-2.20 mg/mL). The standard solutions were prepared by taking aliquots of a stock solution of each protein and diluting to 1 mL with water. Data Treatment. The peak areas of the standard proteins were plotted against the injected concentrations. The linearity in this relationship was obtained by least-squares regression analysis carried out with a Univariate Linear Calibration program.27 The linear model was validated by means of the analysis of residual and the analysis of variance.28 RESULTS AND DISCUSSION Chromatographic Separation. In order to achieve the simultaneous separation of soya bean and animal whey proteins, a preliminar linear binary gradient (20-46% B in 20 min) of ACN/ water/0.1% TFA (employed for the separation of soya bean proteins by RP-HPLC29) was first used but no separation of animal whey proteins from soya bean proteins was obtained. Thus, the gradient solvent strength and gradient range were increased until simultaneous separation of animal whey proteins and soya bean proteins, without impairing the separation of the latter. The chosen gradient consisted of a six-step linear binary gradient: 20% B for 1 min, 20-42% B in 15 min, 42-46% B in 4 min, 46-100% B in 0.5 min, 100% B for 0.5 min, and finally from 100 to 20% B in 0.5 min using mobile phases 0.1% TFA in water (A) and 0.1% TFA in ACN (B) and a gradient range of 3.47%/min. This gradient was used to separate a synthetic sample composed by soya bean protein isolate, R-LA standard, and β-LG (A + B) standard. Figure 1 shows the resulting chromatogram for this sample. It can be observed that the method enables the clear separation of the two different groups of proteins in only 20 min, soya bean proteins eluting within the first 10 min and bovine whey proteins eluting from 16 to 20 min. Soya bean proteins appear in five peaks, which were assigned in previous work29 to the main soya bean globulins (7S and 11S) by comparing their retention times with those of the 7S and 11S globulin fractions isolated from soya bean products by means of a fractionation procedure.13 Assignment of the chromatographic peaks for bovine whey proteins was accom(27) Blanco, M.; Boque´, R.; Cela, R.; Coello, J.; Maspoch, S.; Ortı´z, M. C.; Riba, J.; Rius, F. X.; Ruı´z, A.; Sarabia, L. A.; Toma´s, X. Avances en Quimiometrı´a Pra´ ctica; Servicio de Publicacio´n e Intercambio Cientı´fico: Santiago de Compostela, 1994. (28) Massart, D. L.; Vandeginste, B. G. M.; Deming, S. N.; Michotte, Y.; Kaufman, L. Chemometrics: A Textbook; Vandeginste, B. G. M., Michotte, Y., Eds.; Data Handling in Science and Technology 2; Elsevier: Amsterdam, 1988. (29) Garcı´a, M. C.; Torre, M.; Laborda, F.; Marina, M. L. J. Chromatogr., A 1997, 758, 75-83.2.
Table 1. Calibration by the External Standard Method of Soya Bean Proteins and Bovine Whey Proteins (r-LA and β-LG) by RP-HPLCa protein
linear concn range (mg/mL)
nb
slopec
interceptc
rd
detection limit (mg/mL)
soya bean protein isolate R-LA β-LG
1.0049-4.9166 0.0302-0.3182 0.5462-2.2250
5 6 4
36.48 (1.02) 505.70 (11.21) 143.50 (3.25)
0.57 (3.27) 3.37 (2.18) 5.76 (4.74)
0.999 0.999 0.999
0.361 0.017 0.086
a Experimental conditions as in Figure 1. b n, number of points considered for the regression. Each point represents the average of three consecutive injections of each standard solution. c Errors in the slope and intercept of the regression line are given in parentheses. d r, correlation coefficient.
Table 2. Repeatability, Expressed as Relative Standard Deviation, for the Retention Times and Peak Area Corresponding to the RP-HPLC Simultaneous Separation of Proteins in a Synthetic Mixture of Soya Bean Protein Isolate and Standards of Animal Whey Proteinsa soya bean proteins peak 1
peak 2
peak 3
β-LG (A + B)
R-LA peak 4
peak 5
peak 6
peak 7
peak 8
assay
n
t (%)
A (%)
t (%)
A (%)
t (%)
A (%)
t (%)
A (%)
t (%)
A (%)
t (%)
A (%)
t (%)
A (%)
t (%)
A (%)
same day between daysb
7
0.36 2.75
7.03 3.19
0.19 2.84
5.19 5.08
0.29 0.25
13.47 9.65
0.15 2.57
0.68 1.00
0.12 2.31
2.92 6.08
0.03 0.92
1.71 3.11
0.04 1.11
2.29 3.54
0.03 1.26
1.93 1.40
b
a Protein concentration (mg/mL) in the synthetic sample injected: 1.2341 soya bean protein isolate + 0.0734 R-LA + 1.4729 β-LG (A + B). Two consecutive days.
Figure 1. Chromatogram corresponding to a simultaneous separation of soya bean and animal whey proteins. Conditions: temperature 50 °C; flow rate 1 mL/min; gradient: 20% B, 1 min, 20-42% B in 15 min, 42-46% B in 4 min, and 46-100% B in 0.5 min; soya bean protein isolate, 1.3241 mg/mL, R-LA 0.1297 mg/mL; β-LG 0.5991 mg/ mL; mobile phase A 0.1% TFA in water, mobile phase B 0.1% TFA in ACN. Peak identification: 1-5, soya bean proteins (7S and 11S globulins); 6, R-LA; 7 and 8, β-LG (A + B).
plished by comparing their retention times with those of the individual standards injected under the same chromatographic conditions. The ratio (polar + ionic amino acids)/(nonpolar amino acids) in the different proteins investigated should explain the shorter retention times of soya bean proteins in relation to those observed for the animal whey poteins. In fact, this ratio is higher for soya bean proteins (∼1.75) than for whey proteins (∼1.2). Figure 1 also shows a good resolution between peaks from soya bean proteins and between peaks from bovine whey (R-LA and β-LG). Quantitative Analysis of Soya bean and Animal Whey Proteins. Calibration of soya bean proteins (external standard method) was performed by adding the areas of the five peaks corresponding to soya bean protein isolate (peaks 1-5). Calibration of R-LA and β-LG by the external standard method was achieved by using standards prepared as indicated in the Experi-
mental Section. Calibration of β-LG was carried out by adding the areas of the two peaks corresponding to β-LG (A) and β-LG (B). Values of the slope, intercept, correlation coefficient, and detection limit of every calibration in a linear concentration range are shown in Table 1. In all cases, a good fit to a straight line is observed within the range of concentrations studied. From the values for the detection limits, the possibility of detecting up to 50 µg of R-LA and 240 µg of β-LG, per milligram of soya bean protein isolate can be noted. The precision of the method was determined using the soya bean protein isolate and the R-LA and β-LG standards. The results obtained expressed as relative standard deviation (RSD) for the measurement of the retention time and the peak area in the same day and between two consecutive days are given in Table 2. The precision obtained was better for animal whey proteins than for soya bean proteins, especially for peak area. Application to Edible Samples. The method described above was used to quantify soya bean proteins and to investigate the presence of animal whey proteins in samples of commercially available liquid and powdered soya bean milks. Determination of soya bean protein concentration in these samples of vegetable milks was achieved by injecting into the chromatographic system a quantity of these samples diluted in water. The concentration of soya bean proteins in soya bean milks was calculated using the calibration equation given in Table 1. Under these conditions, animal whey proteins were not detected in any of the soya bean milks tested, at least at the concentration at which the samples were injected. In order to confirm the absence of animal whey proteins in these samples, the wheys obtained from an acidic precipitation of the soya bean milks (see Experimental Section) were also injected in the chromatographic system under the same experimental conditions. Figure 2 shows the chromatogram resulting from the injection of the whey of a powdered soya bean milk in which animal whey proteins were found. As the whey was injected without dilution, the minor fraction of soya bean Analytical Chemistry, Vol. 69, No. 11, June 1, 1997
2219
Table 3. Concentration (mg/100 mg of Milk) of Soya Bean Protein, r-LA, and β-LG in Three Soya Bean Milks Determined by the RP-HPLC Methoda Kjeldahl analysis sample soya bean milk powdered soya bean milk A powdered soya bean milk B
protein content (%) 4.51 25.03 34.90
RP-HPLC method soya bean proteinsb
SD 0.02 0.27 0.19
6.39 ( 1.25 27.39 ( 4.86 29.51 ( 4.85
R-LAc ndd nd 0.38 ( 0.02
β-LG (A + B)c nd nd 0.28 ( 0.30
a Experimental conditions as in Figure 1. Results expressed as the mean value of three determinations ( standard deviation. b Results obtained by direct injection of the sample into the chromatographic system (diluted in water). c Results obtained by direct injection of the acid whey in the chromatographic system (without dilution). d nd, not detected.
Figure 2. Chromatogram of whey obtained by acidic precipitation from a solution of powdered soya bean milk corresponding to a glass of milk (∼48.92 mg/mL). Experimental conditions and peak identification as in Figure 1.
globulins that does not precipitate at the precipitation pH appears in the chromatogram. The two groups of different proteins appeared in this commercial milk expended as a soya bean milk in a public market. The concentrations of soya bean proteins found in three of the analyzed samples (one liquid soya bean milk and two powdered soya bean milks) are given in Table 3 together with the concentrations of R-LA and β-LG found in one of the samples investigated. As expected, results obtained indicated a higher proportion of soya bean proteins in powdered soya bean milks. As for the β-LG, the error in its measurement is especially high, due to the low content of this protein in powdered soya bean milk B. Despite that, this method is very useful from a practical viewpoint since it enables detection of possible adulterations with low levels of β-LG proteins and it could be a valuable method to quantify these proteins in soya bean milks where β-LG (A + B) has been added at higher proportion. The presence of R-LA in soya bean milks can be quantified with a low error even at a low concentration. The values of total protein obtained by the Kjeldahl method are also included in Table 3. Good agreement was observed when results obtained by both methods were compared. In fact, the values obtained by the Kjeldahl method are always near or inside the confidence interval of the protein content determined by the RP-HPLC method.
2220 Analytical Chemistry, Vol. 69, No. 11, June 1, 1997
In order to show the application of the optimized gradient to the determination of soya bean proteins in animal milks, a simultaneous separation of soya bean and animal whey proteins was performed by injecting into the chromatographic system a synthetic sample prepared by adding soya bean proteins from a soya bean protein isolate to an animal whey (obtained by acidic precipitation) from an animal milk available commercially. The chromatogram obtained was similar to that shown in Figure 2, proving the usefulness of the method presented for the detection of adulterations of animal milks by the addition of soya bean proteins. Soya bean proteins were not found in any of the commercial bovine milks analyzed. CONCLUSIONS The RP-HPLC method presented in this work enables the rapid and simultaneous separation of soya bean and animal whey proteins. Quantitation of soya bean proteins in commercial soya bean milks (liquid and powder) is achieved by direct injection of the diluted sample into the chromatographic system. Detection of animal whey proteins at low concentration in these vegetable samples requires an acidic precipitation in order to concentrate the animal whey proteins in the sample previous to the injection into the chromatographic system. The usefulness of the method developed to detect soya bean proteins in commercial bovine milks is also shown. From a practical viewpoint, the RP-HPLC method presented could be of great interest in quality control in order to detect adulteration of soya bean milks by addition of animal whey proteins or adulteration of animal dairy products by soya bean proteins. ACKNOWLEDGMENT The authors thank the Comunidad Auto´noma de Madrid (Spain) for Project COR 0035/94 and C. Marina for linguistic assistance. M.T. thanks the University of Alcala´ de Henares (Madrid, Spain) for Project 031/96. Received for review August 19, 1996. Accepted February 21, 1997.X AC9608432 X
Abstract published in Advance ACS Abstracts, April 1, 1997.
