- Email: [email protected]
Enzyme-Based Flow Injection Analysis System for Glutamine and Glutamate in Mammalian Cell Culture Media
Analytical Biochemistry 268, 110 –116 (1999) Article ID abio.1998.3044, available online at http://www.idealibrary.com on
Enzyme-Based Flow Injection Analysis System for Glutamine and Glutamate in Mammalian Cell Culture Media Christian Mayer,* Alexandra Frauer,* Thomas Schalkhammer,† and Fritz Pittner* ¨ sterreichische Akademie der Wissenschaften-APART, *Institut fu¨r Biochemie und Molekulare Zellbiologie der †O Universita¨t Wien und Ludwig Boltzmann Forschungsstelle fu¨r Biochemie, Dr. Bohrgasse 9, A-1030 Vienna, Austria
Received September 16, 1998
We present the setup of a flow injection analysis system designed for on-line monitoring of glutamate and glutamine. These amino acids represent a major energy source in mammalian cell culture. A cycling assay consisting of glutamate dehydrogenase and aspartate aminotransferase produces NADH proportional to the glutamate concentration in the sample. NADH is then measured spectrophotometrically. Glutamine is determined by conversion to glutamate which is fed into the cycling assay. The conversion of glutamine to glutamate is catalyzed by asparaginase. Asparaginase was used in place of glutaminase due to its relatively high reactivity with glutamine and a pH optimum similar to that of glutamate dehydrogenase. The enzymes were immobilized covalently to activated controlled pore glass beads and integrated into the flow injection analysis system. The application of the immobilized enzymes and the technical setup are presented in this paper. © 1999 Academic Press
Control of the composition of the culture media in which mammalian cells are grown is important. Nutritional depletion, accumulation of side products, and pH may change significantly during the incubation period thereby reducing cell viability as well as interfering with the production of specific proteins (1). The major energy sources in cultivation media for mammalian cells are glucose (2) and L-glutamine (3, 4). Several off-line methods to determine the concentration of such substrates are time-consuming and labor-intensive. Methods for rapid on-line measurement provide data which can be used to optimize fermentation conditions. Analysis systems for glutamine based on amperometric biosensors involving glutaminase and glutamate oxidase have been described (5–10). Nevertheless, op110
tical detection offers the advantage of a multianalyte detection based on a variety of commercially available NAD 1-converting enzymes and avoids electrochemical electrodes sensitive to fouling. In this report an on-line flow injection analysis system for glutamine and glutamate is presented. Glutamate dehydrogenase and aspartate aminotransferase have previously been used in a cycling assay to measure glutamate (11). In combination with Erwinia crysanthemii asparaginase (which has a high glutaminase activity) we were able to monitor glutamine during a fermentation process. The enzymes were immobilized onto controlled pore glass (CPG), 1 packed into columns, and integrated into a flow injection analysis (FIA) system. MATERIALS AND METHODS
Materials NAD 1 (free acid, grade II, ca. 98%) and L-glutamate dehydrogenase (EC 1.4.1.3) from bovine liver (1260 units/mg protein, 20 mg protein/ml) were obtained from Boehringer Mannheim GmbH. (Mannheim, Germany). Aspartate aminotransferase (EC 2.6.1.1) type I from porcine heart (470 units/mg protein, 6.5 mg protein/ml suspension in 3 M (NH 4) 2SO 4), asparaginase (EC 3.5.1.1) from E. chrysanthemii (320 units/mg protein, 140 units/mg lyophilized powder) and CPG (120 – 200 mesh, mean pore diameter, 239 Å, pore distance 65.5%, 0.82 cm 3/g pore volume, 81.6 m 2/g surface area) were obtained from Sigma Chemical Co. (St. Louis, MO). Plastic columns (Mobicols, 1 ml, order No. M1002) were purchased from MoBi Tec (Go¨ttingen, Germany). 1 Abbreviations used: CPG, controlled pore glass; FIA, flow injection analysis; BSA, bovine serum albumin.
0003-2697/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
ENZYME-BASED GLUTAMINE AND GLUTAMATE FLOW INJECTION ANALYSIS
111
FIG. 1. Method of enzyme immobilization on controlled pore glass (CPG).
All other chemicals and reagents were obtained from one of the following sources: Boehringer Mannheim GmbH. (Germany), Sigma Chemical Co., Fluka Chemie (Buchs, Switzerland), and E. Merck (Darmstadt, Germany). Immobilization of Proteins Controlled pore glass was degassed in 10% (v/v) 3-aminopropylmethyldiethoxysilane in toluene and subsequently incubated at 60°C. Beads were filtered, rinsed with toluene, acetone, and ethanol, and dried at 110°C (Fig. 1). Silanized glass beads were degassed in 1 M cyanuric chloride and 2 M triethanolamine in acetone and shaken at 40°C. Subsequently, the beads were filtered and rinsed with acetone. Activated beads were dried at 60°C overnight. Activated glass beads were degassed in 20 mM imidazole/HCl, pH 7.5. The washing buffer was replaced by 20 mM imidazole/HCl, pH 7.5 (53 CPG volume) with 10 mg of protein per 1 mg of controlled pore glass. For the immobilization of L-glutamate dehydrogenase and aspartate aminotransferase, the activator ADP (12) was added (3 mM). This allosteric stabilization of L-glutamate dehydrogenase during immobilization caused a more than threefold increase in bound activity.
The enzymes were coupled by gentle shaking at 4°C overnight. Three types of immobilized protein columns were prepared: BSA was used for the blanks, asparaginase was used for the reaction with glutamine, and a combination of L-glutamate dehydrogenase and aspartate aminotransferase was used for the oxidation of glutamate. A molar ratio of 7 parts L-glutamate dehydrogenase to 1 part aspartate aminotransferase was found to be optimal. The immobilized enzymes were packed into columns, rinsed with 20 mM imidazole/ HCl, pH 7.5, and stored at 4°C until used. The enzyme columns showed excellent stability at room temperature without significant loss of activity for 4 months. An initial loss of about 30% activity within the first 2 to 6 h was due to the leakage of noncovalently bound enzymes. Determination of the Activity of Native Enzymes L-Glutamate dehydrogenase and aspartate aminotransferase. The formation of NADH after addition of a defined amount of a mixture of L-glutamate dehydrogenase and aspartate aminotransferase (molar ratio 7:1) to a glutamate standard solution was monitored at 345 nm. Asparaginase. A defined amount of asparaginase was mixed and incubated with a glutamine standard solution, the reaction was stopped by heating to 95°C,
112
MAYER ET AL.
FIG. 2. Glutamate and glutamine calibration plot and glutamine calibration in the presence of 1 and 2 mM glutamate showing that the slope of the glutamine curves remains unchanged in the presence of glutamate.
and the glutamate generated was determined with the colorimetric test of Boehringer Mannheim (reference methods below). Determination of the Activity of Immobilized Enzymes L-Glutamate dehydrogenase and aspartate aminotransferase. Glutamate standard and NAD 1 were incubated with the immobilized enzymes. The absorbance of the supernatant was measured at 345 nm. Asparaginase. Glutamine standard was incubated with the immobilized enzyme. The glutamate generated in the supernatant was determined with the colorimetric test of Boehringer Mannheim (reference methods below). The activities of the immobilized enzymes were compared to those of the same amounts of native enzymes assayed under identical reaction conditions.
Quantification of Immobilized Proteins Immobilized enzymes were hydrolyzed with 6 M HCl at 110°C in a tightly closed vial. Amino acids were quantified using a ninhydrin-based assay and BSA as reference protein. Under optimized conditions, 3.5 mg of protein was bound per milligram of porous glass
beads (yield 35%). The specific activity of the immobilized asparaginase was about 20% that of the native enzyme. The combination of immobilized L-glutamate dehydrogenase and aspartate aminotransferase had a specific activity about 20% relative to that of the native enzyme combination. Procedure for Measurement of Glutamate and Glutamine The columns (7 mm diameter, 19 mm length), packed with proteins immobilized onto glass beads, were integrated into a flow injection analysis system constructed in the authors’ laboratory with an HP 89072A autosampler (Hewlett Packard), an MV7 motor injection valve (Pharmacia Biotech Europe GmbH), and a UV-M II spectrophotometer (Pharmacia Biotech Europe GmbH, Uppsala, Sweden). The three-way valves (three-way Stopcock from Bio-Rad Laboratories GmbH Wien, Austria) had to be handled manually. For the flow buffer we used 100 mM imidazole/HCl (recommended by Boehringer Mannheim (Biochemica Information) for L-glutamate dehydrogenase assays), 100 mM aspartate, 1 g/liter (1.5 mM) NAD 1 free acid, pH 8.5 with a flow rate of 2 ml/min. ADP (activator of L-glutamate dehydrogenase (13)) was not added to the
ENZYME-BASED GLUTAMINE AND GLUTAMATE FLOW INJECTION ANALYSIS
113
FIG. 3. Typical flow injection analysis chart of glutamate calibration showing high reproducibility.
flow buffer because the signal increase of about 15% did not justify the costs. The flow system was run at room temperature (25°C). The injection volume was 25 ml. Calibration curves were made using glutamine or glutamate in the same buffer. The solutions were prepared just prior to the flow analysis experiments. Although the maximal absorption of NADH is at 340 nm, the wavelength used in the present procedure was 365 nm because there was no 340-nm filter provided for our flow spectrophotometer. The signal at 365 nm was about 70% of that at 340 nm. The spectrophotometer signal was transduced to a PCL-714 (14 bit A/D D/A) Super Lab card. Data acquisition was performed using a real-time Borland Pascal 7.0 program (called ZEUS 3.1) for MS-DOS 6.0 and MS-Windows which was developed and programmed for the applications in our laboratory by Dr. T. Schalkhammer. The program uses a DOS-based real time device driver with a time resolution of 15.6 ms which is coupled to a Windows-based surface by an asynchronous protocol. Reference Methods Glutamate analysis was carried out by the colorimetric test of Boehringer Mannheim (L-glutamic acid, food analysis, colorimetric method, Cat. No. 139 092). The
concentration of glutamine was analyzed by combining the glutamate colorimetric assay with asparaginase from E. chrysanthemii. RESULTS AND DISCUSSION
Asparaginase from E. chrysanthemii catalyzes the conversion of asparagine to aspartate and ammonia. Nevertheless, due to a slightly larger binding cavity than that required for the conversion of asparagine, the enzyme is also able to catalyze the conversion of glutamine to glutamate (Eq. [1]). Glutamine 1 H2 O 7 glutamate 1 NH3
[1]
Glutamate is then deaminated oxidatively by glutamate dehydrogenase to 2-oxoglutarate. The reaction requires NAD 1 which is reduced to NADH (Eq. [2]). Glutamate 1 NAD 1 1 H2 O 7 2-oxoglutarate 1 NADH 1 NH 1 4
[2]
NADH can be detected spectrophotometrically. Since the equilibrium of the reaction is far to the left, end products must be removed to drive the reaction. Using
114
MAYER ET AL.
FIG. 4. Typical flow injection analysis chart of glutamine measurement in the absence or presence of glutamate background showing reproducibility and the possibility of difference measurement. Note the effect of low pass filtering of the detector signal compared to Fig. 3. Low pass filtering enhances sensitivity about three times.
an alkaline buffer (to decrease the H 1 concentration) and a method to remove 2-oxoglutarate, quantitative measurement of glutamate can be achieved. For this purpose, we combined glutamate dehydrogenase with a second enzyme. Aspartate aminotransferase catalyzes a transamination of aspartate (added to the flow buffer) with 2-oxoglutarate forming glutamate and oxalacetate (Eq. [3]). 2-Oxoglutarate 1 aspartate 7 glutamate 1 oxalacetate
[3]
This reaction not only removes 2-oxoglutarate but also leads to the recycling of glutamate (11, 13). The combination of both enzymes results in an amplification factor for the NADH production of about 60 (100 s incubation time, 1 mM glutamate). Multienzyme immobilisates, prepared as described under Materials and Methods, were packed into columns. Three combinations of columns were necessary to determine glutamate, glutamine, and nonspecific background. A blank was obtained using two BSA columns. The glutamate in the medium can be deter-
mined by combining the first BSA column with the column containing glutamate dehydrogenase and aspartate aminotransferase. The combination of the asparaginase column and the column containing immobilized glutamate dehydrogenase and aspartate aminotransferase provides a signal corresponding to glutamate plus glutamine in the medium. Figure 2 shows the calibration curves for glutamine and glutamate. To ensure that glutamine can be accurately determined in the presence of glutamate, glutamine calibration curves were also measured with a background of 1 and 2 mM glutamate. The glutamate level in cell culture media is approximately 2 mM. The slope of the glutamine calibration curves remained unchanged in the presence of glutamate. The linear range of the system spans 0 to 3 mM for glutamate and 0 to 7 mM for glutamine with a glutamate background up to 2 mM, which is more than sufficient for this field of application. Typical flow injection analysis charts of glutamate and glutamine calibration curves are given in Figs. 3 and 4 demonstrating that at least 20 samples per hour can be assayed due to the short response time of the system.
115
ENZYME-BASED GLUTAMINE AND GLUTAMATE FLOW INJECTION ANALYSIS
FIG. 5. Accuracy of the measurement of glutamate in fermentation media. The concentration of glutamate in the medium was determined using a standard calibration curve. The medium spiked with different concentrations of glutamate was then measured. The value for the unspiked medium was subtracted from that of the spiked medium, and the calculated concentration was plotted against the glutamate concentration added to the medium.
Figure 3 shows the high reproducibility of glutamate measurements. In this experimental setup only the column with the enzyme combination was used (no BSA column). The peaks are sharper than those in Fig. 4 where glutamine was measured (requires 2 columns). Figure 4 shows reproducibility of the glutamine measurement in the absence and presence of glutamate background and the possibility of difference measurement. The measurement of glutamine and glutamate in fermentation media, containing other amino acids, carbohydrates, and other nutrients, is shown in Figs. 5 and 6. The concentration of glutamine and glutamate in the medium was determined using a standard cali-
FIG. 6. Accuracy of the measurement of glutamine in fermentation media. The concentration of glutamine in the medium was determined using a standard calibration curve. The medium spiked with different concentrations of glutamine was then measured. The value for the unspiked medium was subtracted from that of the spiked medium, and the calculated concentration was plotted against the glutamine concentration added to the medium.
bration curve. The medium spiked with different concentrations of glutamate or glutamine was then measured. The value for the unspiked medium was subtracted from the value for the spiked medium, and the calculated concentration was plotted against the added concentration. The 45° slopes of the resulting curves show that glutamate and glutamine concentrations in media can be determined with high accuracy in the required range. Samples of the cell culture media were analyzed at different stages of cell cultivation: fresh medium, 5 h after cell division, just before cell division and harvest-
TABLE 1
Comparison of Glutamine and Glutamate Concentrations of Cell Culture Media at Different Stages of Cell Cultivation Obtained with both the Flow Injection Analysis System and the Reference Method Showing a Deviation of Less Than 3%
Fresh medium 5 h after cell division Before cell division Before harvesting
Glutamine (flow system)
Concn (mM) (reference method)
Glutamate (flow system)
Concn (mM) (reference method)
5.12 3.75 1.98 1.33
5.24 3.67 2.05 1.29
1.05 1.24 1.45 1.66
1.02 1.27 1.49 1.61
Note. The measurements were repeated 6 times, the listed concentrations are average values.
116
MAYER ET AL.
ing. Glutamine and glutamate concentrations were determined by both the flow injection analysis system and the reference methods referred to under Materials and Methods. A good correlation for both methods was found (Table 1). Additionally other amino acids in the cultivation media did not interfere with the cycling process. Finally it was shown that the method could readily be applied to measure glutamate and glutamine in cell culture. ACKNOWLEDGMENTS This work was supported by Bender & Co. Ges.m.b.H. Vienna, Austria and the Grant 3/10982/1317 of the “Forschungsfo¨rderungsfond fu¨r die gewerbliche Wirtschaft.” The cell culture samples were provided by Bender & Co. Ges.m.b.H. Vienna, Austria. The authors thank Mag. D. Anrather for careful correction of the manuscript.
REFERENCES 1. Butler, M., and Jenkins, H. (1989) J. Biotechnol. 12, 97–110. 2. Nahapetian, A. T., Thomas, J. N., and Thilly, W. G. (1986) J. Cell Sci. 81, 65–103. 3. Miller, W. M., Wilke, C. R., and Blanch, H. W. (1989) Biotech. Bioeng. 33, 487– 499.
4. Schneider, Y-J., and Lavoix, A. (1990) J. Immunol. Methods 129, 251–268. 5. Renneberg, R., Trott-Kriegeskorte, G., Lietz, M., Ja¨ger, V., Pawlowa, M., Kaiser, G., Wollenberger, U., Schubert, F., Wagner, R., Schmid, R. D., and Scheller, F. W. (1991) J. Biotechnol. 21, 173–186. 6. Cattaneo, M. V., Luong, J. H. T., and Mercille, S. (1992) Biosensors Bioelectronics 7, 329 –334. 7. Schalkhammer, Th., Lobmaier, Ch., Ecker, B., Wakolbinger, W., Kynclova, E., Hawa, G., and Pittner, F. (1994) Sensors Actuators B 18 –19, 587–591. 8. Moser, I., Jobst, G., Aschauer, E., Svasek, P., Varahram, M., and Urban, G., Zanin, V. A., Tjoutrina, G. Y., Zharikova, A. V., and Berezov, T. T. (1995) Biosensors Bioelectronics 10, 527–532. 9. White, S. F., and Turner, A. P. F., Bilitewski, U., Bradley, J., and Schmid, R. D. (1995) Biosensors Bioelectronics 10, 543–551. 10. Lobmaier, Ch., Schalkhammer, Th., Hawa, G., Ecker, B., and Pittner, F. (1995) J. Mol. Recogn. 8, 146 –150. 11. Lowry, O. H., and Passonneua, J. V. (1972) A Flexible System of Enzymatic Analysis, pp. 143–144, Academic Press, New York. 12. Schmidt, E., and Schmidt, F. W. (1983) in Methods of Enzymatic Analysis, pp. 216 –227, VCH, Weinheim, W. Germany/Derrfield Beach, FL. 13. Wollenberger, U., Schubert, F., Pfeiffer, D., and Scheller, F. (1993) Trends Biotechnol. 11, 255–262.
Enzyme-Based Flow Injection Analysis System for Glutamine and Glutamate in Mammalian Cell Culture Media Christian Mayer,* Alexandra Frauer,* Thomas Schalkhammer,† and Fritz Pittner* ¨ sterreichische Akademie der Wissenschaften-APART, *Institut fu¨r Biochemie und Molekulare Zellbiologie der †O Universita¨t Wien und Ludwig Boltzmann Forschungsstelle fu¨r Biochemie, Dr. Bohrgasse 9, A-1030 Vienna, Austria
Received September 16, 1998
We present the setup of a flow injection analysis system designed for on-line monitoring of glutamate and glutamine. These amino acids represent a major energy source in mammalian cell culture. A cycling assay consisting of glutamate dehydrogenase and aspartate aminotransferase produces NADH proportional to the glutamate concentration in the sample. NADH is then measured spectrophotometrically. Glutamine is determined by conversion to glutamate which is fed into the cycling assay. The conversion of glutamine to glutamate is catalyzed by asparaginase. Asparaginase was used in place of glutaminase due to its relatively high reactivity with glutamine and a pH optimum similar to that of glutamate dehydrogenase. The enzymes were immobilized covalently to activated controlled pore glass beads and integrated into the flow injection analysis system. The application of the immobilized enzymes and the technical setup are presented in this paper. © 1999 Academic Press
Control of the composition of the culture media in which mammalian cells are grown is important. Nutritional depletion, accumulation of side products, and pH may change significantly during the incubation period thereby reducing cell viability as well as interfering with the production of specific proteins (1). The major energy sources in cultivation media for mammalian cells are glucose (2) and L-glutamine (3, 4). Several off-line methods to determine the concentration of such substrates are time-consuming and labor-intensive. Methods for rapid on-line measurement provide data which can be used to optimize fermentation conditions. Analysis systems for glutamine based on amperometric biosensors involving glutaminase and glutamate oxidase have been described (5–10). Nevertheless, op110
tical detection offers the advantage of a multianalyte detection based on a variety of commercially available NAD 1-converting enzymes and avoids electrochemical electrodes sensitive to fouling. In this report an on-line flow injection analysis system for glutamine and glutamate is presented. Glutamate dehydrogenase and aspartate aminotransferase have previously been used in a cycling assay to measure glutamate (11). In combination with Erwinia crysanthemii asparaginase (which has a high glutaminase activity) we were able to monitor glutamine during a fermentation process. The enzymes were immobilized onto controlled pore glass (CPG), 1 packed into columns, and integrated into a flow injection analysis (FIA) system. MATERIALS AND METHODS
Materials NAD 1 (free acid, grade II, ca. 98%) and L-glutamate dehydrogenase (EC 1.4.1.3) from bovine liver (1260 units/mg protein, 20 mg protein/ml) were obtained from Boehringer Mannheim GmbH. (Mannheim, Germany). Aspartate aminotransferase (EC 2.6.1.1) type I from porcine heart (470 units/mg protein, 6.5 mg protein/ml suspension in 3 M (NH 4) 2SO 4), asparaginase (EC 3.5.1.1) from E. chrysanthemii (320 units/mg protein, 140 units/mg lyophilized powder) and CPG (120 – 200 mesh, mean pore diameter, 239 Å, pore distance 65.5%, 0.82 cm 3/g pore volume, 81.6 m 2/g surface area) were obtained from Sigma Chemical Co. (St. Louis, MO). Plastic columns (Mobicols, 1 ml, order No. M1002) were purchased from MoBi Tec (Go¨ttingen, Germany). 1 Abbreviations used: CPG, controlled pore glass; FIA, flow injection analysis; BSA, bovine serum albumin.
0003-2697/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
ENZYME-BASED GLUTAMINE AND GLUTAMATE FLOW INJECTION ANALYSIS
111
FIG. 1. Method of enzyme immobilization on controlled pore glass (CPG).
All other chemicals and reagents were obtained from one of the following sources: Boehringer Mannheim GmbH. (Germany), Sigma Chemical Co., Fluka Chemie (Buchs, Switzerland), and E. Merck (Darmstadt, Germany). Immobilization of Proteins Controlled pore glass was degassed in 10% (v/v) 3-aminopropylmethyldiethoxysilane in toluene and subsequently incubated at 60°C. Beads were filtered, rinsed with toluene, acetone, and ethanol, and dried at 110°C (Fig. 1). Silanized glass beads were degassed in 1 M cyanuric chloride and 2 M triethanolamine in acetone and shaken at 40°C. Subsequently, the beads were filtered and rinsed with acetone. Activated beads were dried at 60°C overnight. Activated glass beads were degassed in 20 mM imidazole/HCl, pH 7.5. The washing buffer was replaced by 20 mM imidazole/HCl, pH 7.5 (53 CPG volume) with 10 mg of protein per 1 mg of controlled pore glass. For the immobilization of L-glutamate dehydrogenase and aspartate aminotransferase, the activator ADP (12) was added (3 mM). This allosteric stabilization of L-glutamate dehydrogenase during immobilization caused a more than threefold increase in bound activity.
The enzymes were coupled by gentle shaking at 4°C overnight. Three types of immobilized protein columns were prepared: BSA was used for the blanks, asparaginase was used for the reaction with glutamine, and a combination of L-glutamate dehydrogenase and aspartate aminotransferase was used for the oxidation of glutamate. A molar ratio of 7 parts L-glutamate dehydrogenase to 1 part aspartate aminotransferase was found to be optimal. The immobilized enzymes were packed into columns, rinsed with 20 mM imidazole/ HCl, pH 7.5, and stored at 4°C until used. The enzyme columns showed excellent stability at room temperature without significant loss of activity for 4 months. An initial loss of about 30% activity within the first 2 to 6 h was due to the leakage of noncovalently bound enzymes. Determination of the Activity of Native Enzymes L-Glutamate dehydrogenase and aspartate aminotransferase. The formation of NADH after addition of a defined amount of a mixture of L-glutamate dehydrogenase and aspartate aminotransferase (molar ratio 7:1) to a glutamate standard solution was monitored at 345 nm. Asparaginase. A defined amount of asparaginase was mixed and incubated with a glutamine standard solution, the reaction was stopped by heating to 95°C,
112
MAYER ET AL.
FIG. 2. Glutamate and glutamine calibration plot and glutamine calibration in the presence of 1 and 2 mM glutamate showing that the slope of the glutamine curves remains unchanged in the presence of glutamate.
and the glutamate generated was determined with the colorimetric test of Boehringer Mannheim (reference methods below). Determination of the Activity of Immobilized Enzymes L-Glutamate dehydrogenase and aspartate aminotransferase. Glutamate standard and NAD 1 were incubated with the immobilized enzymes. The absorbance of the supernatant was measured at 345 nm. Asparaginase. Glutamine standard was incubated with the immobilized enzyme. The glutamate generated in the supernatant was determined with the colorimetric test of Boehringer Mannheim (reference methods below). The activities of the immobilized enzymes were compared to those of the same amounts of native enzymes assayed under identical reaction conditions.
Quantification of Immobilized Proteins Immobilized enzymes were hydrolyzed with 6 M HCl at 110°C in a tightly closed vial. Amino acids were quantified using a ninhydrin-based assay and BSA as reference protein. Under optimized conditions, 3.5 mg of protein was bound per milligram of porous glass
beads (yield 35%). The specific activity of the immobilized asparaginase was about 20% that of the native enzyme. The combination of immobilized L-glutamate dehydrogenase and aspartate aminotransferase had a specific activity about 20% relative to that of the native enzyme combination. Procedure for Measurement of Glutamate and Glutamine The columns (7 mm diameter, 19 mm length), packed with proteins immobilized onto glass beads, were integrated into a flow injection analysis system constructed in the authors’ laboratory with an HP 89072A autosampler (Hewlett Packard), an MV7 motor injection valve (Pharmacia Biotech Europe GmbH), and a UV-M II spectrophotometer (Pharmacia Biotech Europe GmbH, Uppsala, Sweden). The three-way valves (three-way Stopcock from Bio-Rad Laboratories GmbH Wien, Austria) had to be handled manually. For the flow buffer we used 100 mM imidazole/HCl (recommended by Boehringer Mannheim (Biochemica Information) for L-glutamate dehydrogenase assays), 100 mM aspartate, 1 g/liter (1.5 mM) NAD 1 free acid, pH 8.5 with a flow rate of 2 ml/min. ADP (activator of L-glutamate dehydrogenase (13)) was not added to the
ENZYME-BASED GLUTAMINE AND GLUTAMATE FLOW INJECTION ANALYSIS
113
FIG. 3. Typical flow injection analysis chart of glutamate calibration showing high reproducibility.
flow buffer because the signal increase of about 15% did not justify the costs. The flow system was run at room temperature (25°C). The injection volume was 25 ml. Calibration curves were made using glutamine or glutamate in the same buffer. The solutions were prepared just prior to the flow analysis experiments. Although the maximal absorption of NADH is at 340 nm, the wavelength used in the present procedure was 365 nm because there was no 340-nm filter provided for our flow spectrophotometer. The signal at 365 nm was about 70% of that at 340 nm. The spectrophotometer signal was transduced to a PCL-714 (14 bit A/D D/A) Super Lab card. Data acquisition was performed using a real-time Borland Pascal 7.0 program (called ZEUS 3.1) for MS-DOS 6.0 and MS-Windows which was developed and programmed for the applications in our laboratory by Dr. T. Schalkhammer. The program uses a DOS-based real time device driver with a time resolution of 15.6 ms which is coupled to a Windows-based surface by an asynchronous protocol. Reference Methods Glutamate analysis was carried out by the colorimetric test of Boehringer Mannheim (L-glutamic acid, food analysis, colorimetric method, Cat. No. 139 092). The
concentration of glutamine was analyzed by combining the glutamate colorimetric assay with asparaginase from E. chrysanthemii. RESULTS AND DISCUSSION
Asparaginase from E. chrysanthemii catalyzes the conversion of asparagine to aspartate and ammonia. Nevertheless, due to a slightly larger binding cavity than that required for the conversion of asparagine, the enzyme is also able to catalyze the conversion of glutamine to glutamate (Eq. [1]). Glutamine 1 H2 O 7 glutamate 1 NH3
[1]
Glutamate is then deaminated oxidatively by glutamate dehydrogenase to 2-oxoglutarate. The reaction requires NAD 1 which is reduced to NADH (Eq. [2]). Glutamate 1 NAD 1 1 H2 O 7 2-oxoglutarate 1 NADH 1 NH 1 4
[2]
NADH can be detected spectrophotometrically. Since the equilibrium of the reaction is far to the left, end products must be removed to drive the reaction. Using
114
MAYER ET AL.
FIG. 4. Typical flow injection analysis chart of glutamine measurement in the absence or presence of glutamate background showing reproducibility and the possibility of difference measurement. Note the effect of low pass filtering of the detector signal compared to Fig. 3. Low pass filtering enhances sensitivity about three times.
an alkaline buffer (to decrease the H 1 concentration) and a method to remove 2-oxoglutarate, quantitative measurement of glutamate can be achieved. For this purpose, we combined glutamate dehydrogenase with a second enzyme. Aspartate aminotransferase catalyzes a transamination of aspartate (added to the flow buffer) with 2-oxoglutarate forming glutamate and oxalacetate (Eq. [3]). 2-Oxoglutarate 1 aspartate 7 glutamate 1 oxalacetate
[3]
This reaction not only removes 2-oxoglutarate but also leads to the recycling of glutamate (11, 13). The combination of both enzymes results in an amplification factor for the NADH production of about 60 (100 s incubation time, 1 mM glutamate). Multienzyme immobilisates, prepared as described under Materials and Methods, were packed into columns. Three combinations of columns were necessary to determine glutamate, glutamine, and nonspecific background. A blank was obtained using two BSA columns. The glutamate in the medium can be deter-
mined by combining the first BSA column with the column containing glutamate dehydrogenase and aspartate aminotransferase. The combination of the asparaginase column and the column containing immobilized glutamate dehydrogenase and aspartate aminotransferase provides a signal corresponding to glutamate plus glutamine in the medium. Figure 2 shows the calibration curves for glutamine and glutamate. To ensure that glutamine can be accurately determined in the presence of glutamate, glutamine calibration curves were also measured with a background of 1 and 2 mM glutamate. The glutamate level in cell culture media is approximately 2 mM. The slope of the glutamine calibration curves remained unchanged in the presence of glutamate. The linear range of the system spans 0 to 3 mM for glutamate and 0 to 7 mM for glutamine with a glutamate background up to 2 mM, which is more than sufficient for this field of application. Typical flow injection analysis charts of glutamate and glutamine calibration curves are given in Figs. 3 and 4 demonstrating that at least 20 samples per hour can be assayed due to the short response time of the system.
115
ENZYME-BASED GLUTAMINE AND GLUTAMATE FLOW INJECTION ANALYSIS
FIG. 5. Accuracy of the measurement of glutamate in fermentation media. The concentration of glutamate in the medium was determined using a standard calibration curve. The medium spiked with different concentrations of glutamate was then measured. The value for the unspiked medium was subtracted from that of the spiked medium, and the calculated concentration was plotted against the glutamate concentration added to the medium.
Figure 3 shows the high reproducibility of glutamate measurements. In this experimental setup only the column with the enzyme combination was used (no BSA column). The peaks are sharper than those in Fig. 4 where glutamine was measured (requires 2 columns). Figure 4 shows reproducibility of the glutamine measurement in the absence and presence of glutamate background and the possibility of difference measurement. The measurement of glutamine and glutamate in fermentation media, containing other amino acids, carbohydrates, and other nutrients, is shown in Figs. 5 and 6. The concentration of glutamine and glutamate in the medium was determined using a standard cali-
FIG. 6. Accuracy of the measurement of glutamine in fermentation media. The concentration of glutamine in the medium was determined using a standard calibration curve. The medium spiked with different concentrations of glutamine was then measured. The value for the unspiked medium was subtracted from that of the spiked medium, and the calculated concentration was plotted against the glutamine concentration added to the medium.
bration curve. The medium spiked with different concentrations of glutamate or glutamine was then measured. The value for the unspiked medium was subtracted from the value for the spiked medium, and the calculated concentration was plotted against the added concentration. The 45° slopes of the resulting curves show that glutamate and glutamine concentrations in media can be determined with high accuracy in the required range. Samples of the cell culture media were analyzed at different stages of cell cultivation: fresh medium, 5 h after cell division, just before cell division and harvest-
TABLE 1
Comparison of Glutamine and Glutamate Concentrations of Cell Culture Media at Different Stages of Cell Cultivation Obtained with both the Flow Injection Analysis System and the Reference Method Showing a Deviation of Less Than 3%
Fresh medium 5 h after cell division Before cell division Before harvesting
Glutamine (flow system)
Concn (mM) (reference method)
Glutamate (flow system)
Concn (mM) (reference method)
5.12 3.75 1.98 1.33
5.24 3.67 2.05 1.29
1.05 1.24 1.45 1.66
1.02 1.27 1.49 1.61
Note. The measurements were repeated 6 times, the listed concentrations are average values.
116
MAYER ET AL.
ing. Glutamine and glutamate concentrations were determined by both the flow injection analysis system and the reference methods referred to under Materials and Methods. A good correlation for both methods was found (Table 1). Additionally other amino acids in the cultivation media did not interfere with the cycling process. Finally it was shown that the method could readily be applied to measure glutamate and glutamine in cell culture. ACKNOWLEDGMENTS This work was supported by Bender & Co. Ges.m.b.H. Vienna, Austria and the Grant 3/10982/1317 of the “Forschungsfo¨rderungsfond fu¨r die gewerbliche Wirtschaft.” The cell culture samples were provided by Bender & Co. Ges.m.b.H. Vienna, Austria. The authors thank Mag. D. Anrather for careful correction of the manuscript.
REFERENCES 1. Butler, M., and Jenkins, H. (1989) J. Biotechnol. 12, 97–110. 2. Nahapetian, A. T., Thomas, J. N., and Thilly, W. G. (1986) J. Cell Sci. 81, 65–103. 3. Miller, W. M., Wilke, C. R., and Blanch, H. W. (1989) Biotech. Bioeng. 33, 487– 499.
4. Schneider, Y-J., and Lavoix, A. (1990) J. Immunol. Methods 129, 251–268. 5. Renneberg, R., Trott-Kriegeskorte, G., Lietz, M., Ja¨ger, V., Pawlowa, M., Kaiser, G., Wollenberger, U., Schubert, F., Wagner, R., Schmid, R. D., and Scheller, F. W. (1991) J. Biotechnol. 21, 173–186. 6. Cattaneo, M. V., Luong, J. H. T., and Mercille, S. (1992) Biosensors Bioelectronics 7, 329 –334. 7. Schalkhammer, Th., Lobmaier, Ch., Ecker, B., Wakolbinger, W., Kynclova, E., Hawa, G., and Pittner, F. (1994) Sensors Actuators B 18 –19, 587–591. 8. Moser, I., Jobst, G., Aschauer, E., Svasek, P., Varahram, M., and Urban, G., Zanin, V. A., Tjoutrina, G. Y., Zharikova, A. V., and Berezov, T. T. (1995) Biosensors Bioelectronics 10, 527–532. 9. White, S. F., and Turner, A. P. F., Bilitewski, U., Bradley, J., and Schmid, R. D. (1995) Biosensors Bioelectronics 10, 543–551. 10. Lobmaier, Ch., Schalkhammer, Th., Hawa, G., Ecker, B., and Pittner, F. (1995) J. Mol. Recogn. 8, 146 –150. 11. Lowry, O. H., and Passonneua, J. V. (1972) A Flexible System of Enzymatic Analysis, pp. 143–144, Academic Press, New York. 12. Schmidt, E., and Schmidt, F. W. (1983) in Methods of Enzymatic Analysis, pp. 216 –227, VCH, Weinheim, W. Germany/Derrfield Beach, FL. 13. Wollenberger, U., Schubert, F., Pfeiffer, D., and Scheller, F. (1993) Trends Biotechnol. 11, 255–262.