Analyst, January 1996, Vol. 121 (71-73) 71 Amperometric Biosensor for the Determination of the Artificial Sweetener Aspartame With an Immobilized Bienzyme System Shu-Fen Chou Department of Food Health, Chia-Nan College of Pharmacy, Tainan Hsien, Taiwan A biosensor using an immobilized bienzyme system was developed to determine the artificial sweetener aspartame in foodstuffs. The bienzyme system, consisting of a-chymotrypsin and alcohol oxidase, was co-immobilized by a covalent cross-linking method with glutaraldehyde on a dissolved oxygen electrode. The optimum operating conditions for the biosensor were pH 6.5-8.5 and 30 "C. A linear relationship was observed between the decrease in dissolved oxygen and the aspartame concentration in the range 0.33-2 mmol l-1. When the enzyme electrode was stored in dithiothreitol solution, the biosensor was stable for more than 50 d and 450 assays.An amperometric biosensor method was established for the assay of aspartame in commercial products without the need for sample pre-treatment or special reagents. Such a development should facilitate rapid quality control testing in the food industry. Keywords: Aspartame; amperometric biosensor; covalent cross-linking method Introduction Aspartame (L-aspartyl-L-phenylalanine methyl ester; Nutra- sweet) is a low calorie artificial sweetener with a pronounced sucrose-like taste and which, organoleptically, is about 180 times sweeter than sucrose. Recently, it has been approved for use in various foods such as breakfast cereals, table-top sweeteners and soft drinks as well as other low calorie and therapeutic diets.It is also rapidly replacing saccharin which has a bitter aftertaste and has been implicated in causing cancer in laboratory animals. l ~ 2 Various chromatographic methods for the determination of aspartame in foodstuffs such as HPLC, GC, gel-permeation chromatography and TLG6 have been devel- oped. These methods, although having been properly optimized for routine applications, suffer from lengthy procedures. In 1985, a simpler method, viz., a microbial amperometric sensor for measuring the change in respiration rate (oxygen change) caused by uptake of substrate, was described.7~8 The device achieved reasonable sensitivity and stability; however, glucose and the amino acid constitutents of aspartame interfered. An attempt has been made to establish an enzyme sensor which exploits immobilized L-aspartase to convert the aspartame molecule into an electrode-detectable species (ammonium).9 Despite requiring no sample pre-treatment or special reagents, the sensor system was adversely affected by L-aspartate (usually present in beverages).In order to overcome this drawback, an amperometric enzyme sensor for aspartame, composed of a- chymotrypsin and alcohol oxidase, was developed. lo The bienzyme system was immobilized by an entrapping method with a dialysis membrane, and was combined with dissolved oxygen electrode equipment. The bienzyme system was able to convert the aspartame molecule into an oxygen electrode- detectable species, but is was only stable for about 7 d.In this work, the aforementioned bienzyme system was co- immobilized by a covalent cross-linking method with glutar- aldehyde and combined with electrochemical equipment in order to improve the operational stability of the biosensor. The results obtained indicate that the biosensor approach should be an attractive alternative for aspartame testing in the food industry. Experimental Reagents and Standards a-Chyrnotrypsin (EC 3.4.2 1.1 ., from bovine pancreas), alcohol oxidase (EC 1.1.3.13., from Hansenula sp.) and dithiothreitol were obtained from Sigma (St. Louis, MO, USA). Aspartame was obtained from Tokyo Kasei Chemical (Tokyo, Japan). Standard solutions of aspartame were prepared weekly and stored at 5 "C. All chemicals used were of analytical-reagent grade.Apparatus The sensor system (Fig. 1) consisted of a standard bath assembly [Yellow Springs Instruments (YSI) 5301, Yellow Springs, OH, USA], a standard oxygen probe (YSI 5357,5331), a biological oxygen monitor (YSI 5300), a circulator (Type BL20, Yihder Instruments, Taipei, Taiwan) and a recorder (Type BD40, Gilson, The Netherlands). Preparation of Immobilized Enzyme Electrode Enzymes were immobilized by a covalent cross-linking method with glutaraldehyde.9 No bovine serum albumin was used, because of the high protein content of a-chymotrypsin com- bined with alcohol oxidase. The lyophilized enzymes {a- chymotrypsin [250 mU (1 VU = 16.67 nkat)] and alcohol oxidase (250 mu)} were deposited on a dialysis membrane, and 20 yl of phosphate buffer were added in order to solubilize the two enzymes. Next, 2 pl of 6.25% m/v glutaraldehyde solution were gently added and mixed.The resulting membrane was allowed to dry at room temperature for 2 h, and was mounted against a gas-selective membrane, taking care to avoid bubble entrapment. The electrode was equilibrated in distilled water for at least 5 h at 30 "C before use. The electrode was stored in 10 mmol 1- dithiothreitol solution. All measurements were taken at 30 & 1 "C. Samples All samples were purchased from local supermarkets, Measure- ments were made after direct dissolution of the dry powdered72 Analyst, January 1996, Vol. 121 samples or dilution of Diet Coke with de-ionized water or phosphate buffer (pH 6.5-8.5). Results and Discussion Several enzymic systems might be able to convert the aspartame molecule into an electrode-detectable species,g for example: ( 1) with L-aspartase, liberation of ammonium ions from aspartame can be monitored potentiometrically with an ammonium ion- selective electrode.However, aspartate, which is a breakdown product of aspartame in stored soft drinks, can interfere. (2) Use of a direct process involving L-amino acid oxidase which might recognize aspartame. This enzyme might be able to convert the aspartame molecule into an oxygen electrode-detectable spe- cies. The limitations of this approach are low enzymic activity and poor selectivity. (3) Aspartame can be cleaved by a- chymotrypsin to give methanol, which is detectable with a second enzyme, alcohol oxidase. This bienzyme system can convert the aspartame molecule into an oxygen electrode- detectable species.A study was undertaken of the bienzyme system (3) in order to overcome the disadvantages of systems (1) and (2) and establish an aspartame biosensor. It has been demonstrated that aspartame can hydrolyse slowly in the low pH range used in soft drinks; hence the aspartame products become less sweet on prolonged storage.3 Thus, this biosensor might be used to monitor the deterioration in the quality of aspartame in stored soft drinks. However, the bienzyme system, which has reasonable sensitivity and high enzymic activity, could be used to analyse several types of foodstuffs. C Fig. 1 Schematic diagram of enzyme sensor system. A, Oxygen electron, B, standard bath assembly (1, exit port; 2, inlet port; 3, magnetic stirrer; 4, sample chamber; 5 , brake); C , biological oxygen monitor; D, recorder; and E, circulator. c z T l , , l , , l 3 4 5 6 7 8 9 PH Fig.2 Effect of pH on the aspartame biosensor. Circles, 0.05 mol 1-l acetate buffer; and triangles, 0.05 mol 1-1 phosphate buffer. Reaction temperature 30 "C. Optimization of Operating Conditions for the Biosensor Analytical parameters such as pH, temperature and linear range of substrate concentration were studied. Fig. 2 displays the optimum pH values of the biosensor with the immobilized bienzyme system [containing a-chymotrypsin (250 mu) and alcohol oxidase (250 mu)], and clearly indicates that the optimum pH values are in the range 6.5-8.5. Moreover, because the optimum pH values of this system covered a wide range (pH 6.5-8.5), de-ionized water could be substituted for buffer solution in the reaction (this was demonstrated experimentally).Thus, experimental procedures could be simplified. In order to study the effect of temperature, the decrease in dissolved oxygen (DO) was measured at 25,30,40,50 and 60 "C 9. The results demonstrated that the higher the temperature the larger the decrease in DO; however, the decrease in DO appeared to be slower above 50 "C. Hence, 30 "C was selected as the optimum temperature for reasons of economy and to improve the stability of the biosensor. Fig. 3 shows that under the optimum conditions, a linear relationship is observed between the decrease in DO and the aspartame concentration in the range 0.33-2 mmol 1-1.Interference Study The effects of some interferents which might be present in commercial low calorie foods on the aspartame biosensor were tested. The results demonstrated that, at a concentration of 0.66 mmol 1- l , most of the ingredients in foodstuffs such as sucrose, lactose, citric acid, succinic acid, silicon dioxide, sodium benzoate, dextrin and dextrose, and the amino acid breakdown products of aspartame, did not interfere; however, TRIS buffer did interfere because the propane-1,3-diol part of the molecule was also oxidized by alcohol oxidase." Hence, TRIS buffer could not be used as the buffer system. Biosensor Stability A prerequisite consideration for the possible application of the biosensor in industry is its operational stability. The operational 0 ' ' ' I r I 0.5 1.0 1.5 2.0 Concentration/rnrnol I-' Fig.3 Calibration curve of aspartame concentration for the biosensor. a a 150 .- - G o 1 ; ; ; : ; ;j ; ' 2 Ti rn e/d ays Fig. 4 Operational stabilities of the aspartame biosensor. Stored in (a) dithiothreitol solution; and (b) buffer solution. Reaction temperature 30 "C.Analyst, January 1996, Vol. 121 73 Table 1 Determination of aspartame in commercial products Aspartame content (%) Sample* Nominali Previous method* Proposed methods A 3.8 3.70 B 3.5 3.56 C 2.7 2.58 3.82 3.60 2.65 compared with a calibration graph for aspartame. The results obtained with the proposed biosensor closely corresponded to the labelled values of aspartame for the commercially available products (storage time of the commercial products was less than 1 month).Relative errors were within 3%. Because primary alcohols interfere,12 the determination of aspartame in alcoholic foodstuffs cannot be carried out with the proposed biosensor. Conclusions A biosensor composed of an immobilized bienzyme system was developed for the efficient assay of aspartame in several low calorie foodstuffs (excluding alcoholic products). The biosensor might also be used to monitor the deterioration in the quality of aspartame in stored soft drinks. The device exhibited reasonable sensitivity and excellent stability of operations. * A: Aspartame + lactose + silicon dioxide; B: aspartame + dextrin; C: t Labelled values provided by the manufacturer. * The previous enzyme sensor was composed of a bienzyme system (containing cu-chymotrypsin and alcohol oxidase) immobilized by an entrapping method with a dialysis membrane and was combined with a dissolved oxygen electrode. l o Diet Coke.3 n = 3, standard deviation: 1-2%. References stability of the biosensor was tested by storing the enzyme electrode in 10 mmol 1-l dithiothreitol solution, i.e., an antioxidant which can prevent the -SH group of the enzymes from being oxidized. Fig. 4 shows the stability of the biosensor with and without storage in the dithiothreitol solution. The biosensor was stable for more than 50 d when stored in the dithiothreitol solution and could be used for 450 assays. Without storage in the dithiothreitol solution, the biosensor was only stable for up to 2 d. Determination of Aspartame in Various Commercial Products Aspartame was determined in three types of commercial low calorie foods using the proposed biosensor and the previously described aspartame enzyme sensor.10 Table 1 summarizes the data obtained for these samples.Two types of dry powdered mixtures were analysed: the first was a mixture of aspartame and dextrin, while the second was a mixture of lactose, aspartame and silicon dioxide. In addition, Diet Coke, which contains citrate, carbonate, sodium benzoate, caffeine, caramel and aspartame, was also analysed. Measurements were taken after direct dissolution of the samples. No pre-treatment or special reagents were required, and the electrode response was 1 2 3 4 5 6 7 8 9 10 1 1 12 Cloninger, M. R., and Baldwin, R. E., J . Food Sci., 1974,39, 347. Cloninger, M. R., and Baldwin, R. E., Science, (Washington, D.C., Tsang, W.-S., Clarke, M. A., and Parrish, F. W., J . Agric. Food Chern., 1985,33,734. Hussein, M. M., D’Amelia, R. P., Manz, A. L., Jacin, H., and Chen, W.-T.C., J . Food Sci., 1984, 49, 520. Homler, B. E., Food Technol., 1984,38(5-8), 50. Issaq, H. J., Weiss, D., Ridlon, C., Fox, S. D., and Muschik, J. B., J . Liq. Chromatogr., 1986, 9, 1791. Renneberg, R., Riedel, K., and Scheller, F., Appl. Microbiol. Biotechnol., 1985, 21, 180. Rechnitz, G. A., Science, (Washington, D.C., 1883-), 1981, 214, 287. Guilbault, G. G., Lubrane, G. J., Kauffmann, J.-M., and Patriarche, G. J., Anal. Chim. Acta, 1988, 206, 369. Chou, S. F., Chen, J. H., Chou, L. W., Fan, J. J., and Chen, C. Y., J . Food Drug Anal., 1995, 3(2), 121. Janssen, F. W., and Ruelius, H. W., Biochim. Biophys. Acta, 1968, 151, 330. Bergmeyer, H. U., Methods of Enzymatic Analysis, Verlag Chemie, Weinheim, 3rd edn., 1984, vol. 2, p. 143. Paper 5f04889C Received July 24, 1995 Accepted September 11,1995 1883-), 1970, 170, 81.