[44)
`
`
`
`IMMOBILIZED AMINOTRANSFERASES
`
`479
`
`
`
`lized pH-and glutaraldehyde-treated cells. The substrate
`P. dacunhae
`
`
`
`
`solution (adjusted to pH 8.5 with ammonia) consisting of 1.5 M ammo­
`
`
`immo­and 1 mM Mg2+ is applied to the column containing
`nium fumarate
`
`
`
`
`= 1.0/hr. After addition bilized at a flow rate of space velocity E. coli cells
`
`
`
`of PLP and pyruvic acid (these concentrations are 0.1 and 1 mM, respec­
`
`
`
`
`
`of is adjusted to 6.0 by addition tively) into the effluent, pH of the solution
`
`
`
`
`acetic acid. The solution is passed through the immobilized
`P. dacunhae
`
`
`
`
`closed column at a flow rate of space velocity = 0.06/hr and a pressure of
`
`about 8 kg/cm2 achieved by plunger pump.
`The effluent of
`
`
`Crystallization of L-Alanine from Column Effluent.
`
`
`
`appropriate volume is concentrated to about one-fourth of its original
`
`volume and cooled to 15°. L-Alanine
`
`
`crystallized is collected by centrifu­
`
`
`
`
`gation or by filtration and washed with 80% aqueous ethanol. The yield of
`
`
`
`L-alanine from ammonium fumarate is about 90% (theoretical). [a]if =
`+ 14.8 (c == 10 in 6 N HCI).
`
`Conclusion
`
`In 1982, Tanabe Seiyaku Co. Ltd. successfully industrialized a contin­
`
`
`
`
`
`
`
`uous production system of L-alanine from ammonium fumarate, using a
`
`
`
`
`column reactor containing immobilized pH-treated E. coli cells and a
`
`
`
`
`closed column reactor containing immobilized pH-and glutaraldehyde­
`treated
`cells.
`P. dacunhae
`By this system, L-alanine has been produced at low cost. This is
`
`
`
`
`
`
`
`
`considered to be the first industrial application of sequential enzyme reac­
`
`
`tions using two immobilized microbial cells.
`
`
`
`for[44]Immobilized Aminotransferases
`Amino Acid Production
`
`By J. DA YID ROZZELL
`
`Aminotransferases
`
`Background
`
`Aminotransferases (more commonly called transaminases, EC
`
`
`
`
`
`
`
`2.6. l ,_) are a widely distributed class of enzymes. These enzymes cata­
`
`
`lyze the synthesis and breakdown of amino acids in microorganisms,
`
`METHODS IN ENZYMOLOGY, VOL. 136
`
`
`
`Copyright© 1987 by Academic Press, Inc.
`
`
`All rights of reproduction in any form reserved.
`
`Reactive Surfaces Ltd. LLP
`Ex. 1051 (Rozzell Attachment D)
`Reactive Surfaces Ltd. LLP v. Toyota Motor Corp.
`IPR2016-01914
`
`

`

`480 ENZYME ENGINEERING (ENZYME TECHNOLOGY) [44] H2N ~'H + 0 0 + H 2 N~'%H SCHEME 1. plants, and animals by the transfer of an amino group from an a-amino acid to a 2-ketoacid as shown in Scheme 1. The first evidence for aminotransferases was published by Needham j and Szent-Gy6rgyi and co-workers 2 who noticed a relationship between the L-glutamic acid, L-aspartic acid, and oxaloacetic acid levels in pigeon breast muscle. Banga and Szent-Gy6rgyi 3 demonstrated the reversibility of glutamic-pyruvic transaminase (EC 2.6.1.2, alanine aminotransferase) by chemically isolating the amino acid products L-glutamate and L- alanine. Since that time, a large number of aminotransferases have been discovered and characterized. One feature of aminotransferases is the requirement for the small molecule, pyridoxal Y-phosphate, for catalytic activity, this cofactor being bound through a Schiff base linkage to the e- amino group of an active-site lysine. Although the binding of pyridoxal 5'- phosphate to the enzyme is reversible, most aminotransferases show maximal catalytic activity at cofactor concentrations of 100/~M or less. Such low saturating concentrations of pyridoxal phosphate are an impor- tant property of aminotransferases; at concentrations of 100/.~M or less, the cost of the cofactor in biocatalytic transamination processes is a rela- tively minor component of the total cost. The mechanism of transamination is well known, and has been re- viewed previously.* The reaction catalyzed by aminotransferases occurs as the result of two distinct half-reactions: the first involves transfer of the amino group of the L-amino acid donor to pyridoxal 5'-phosphate to yield a 2-ketoacid product which is released from the enzyme and an enzyme- bound pyridoxamine Y-phosphate; the second is the binding of the 2- ketoacid to be transaminated to the enzyme and the transfer of the amino group from pyridoxamine 5'-phosphate to this 2-ketoacid to produce the desired L-amino acid and regenerate the pyridoxal 5'-phosphate. As a result, aminotransferases characteristically exhibit Ping-Pong kinetics. D. M. Needham, Biochem. J. 24, 208 (1930). 2 E. Annau, I. Banga, A. Blazo, V. Bruckner, K. Laki, F. B. Staub, and A. Szent-Gy6rgyi, Z. Physiol. Chem. 224, 105 (1936). I. Banga and A. Szent-Gy6rgyi, Z. Physiol. Chem. 245, 118 (1937). A. E. Braunstein, "The Enzymes IX" (P. D, Boyer, ed.), Part B, pp. 379-481. Academic Press, New York, 1973.
`
`

`

`44 IMMOBILIZED AMINOTRANSFERASES 481 Advantages and Disadvantages for Use in Biocatalysis Although aminotransferases have been known for decades, these en- zymes have seen little use as biocatalysts until now. However, since the normal function of aminotransferases is the biosynthesis and metabolism of amino acids, it is natural to look to these enzymes as potentially useful catalysts for the production of amino acids. In principle, almost any de- sired amino acid can be produced from the appropriate 2-ketoacid using an inexpensive amino acid as the amino donor. There are a number of advantages to the use of this kind of technology. (1) The aminotransferase enzymes catalyze the stereoselective synthesis of only L-amino acids from their corresponding 2-ketoacids. No D isomer is produced, and no resolution is required. (2) Aminotransferases have uniformly high cata- lytic rates, capable of converting up to 400/~mol of substrate/min per milligram of protein. (3) Many of the required 2-ketoacid precursors can be conveniently prepared by chemical synthesis at low cost. (4) The capi- tal investment for an immobilized enzyme process using aminotrans- ferases is much lower than for a fermentation process, and the productiv- ity of the bioreactor is more than an order of magnitude higher. (5) The technology is generally applicable to a broad range of L-amino acids be- cause aminotransferases exist with varying specificities. For example, there are enzymes specific for the transamination of amino acids with acidic side chains, aromatic side chains, branched alkyl side chains, etc. Such broad scope allows a number of different L-amino acids to be pro- duced with the same equipment and often the same biocatalyst. We have demonstrated laboratory-scale processes for the production of a variety of L-amino acids including L-alanine, L-phenylalanine, L-tyrosine, L-tryptophan, and several others. As an example, we have immobilized the commercially available glutamic-pyruvic aminotrans- ferase from porcine heart on porous glass by covalent attachment, and obtained a stable biocatalyst with an activity of 400 International Units per gram. A column packed with 500 mg of this immobilized enzyme was operated continuously for 6 months and produced 160 mg L-ala- nine/hr from pyruvic acid as a starting material. This example illustrates the potential of immobilized aminotransferases applied to the production of L-amino acids. There is one inherent disadvantage to the practice of this technology as described so far; as a catalyst, the aminotransferase can only acceler- ate the approach to equilibrium between the L-amino acid and 2-ketoacid precursors on one side of the equation and the 2-ketoacid and L-amino acid products on the other. Thus, the equilibrium constant for the generic transamination reaction as written in Scheme 1 is near unity, and the
`
`

`

`482 ENZYME ENGINEERING (ENZYME TECHNOLOGY) 44 HO2CCH2 \C/C02H HO2CCH2~(CO2H H2N ~' %H 0 L-Aspartic Acid Oxaloacetic Acid transamtnase + + 0 R. C02H R C02H 2-KetoacJd L-Amino Acid SCHEME 2. H3C C02H oxaloacetate "~( decarboxylase ~ 0 Pyruvic Acid conversion of a 2-ketoacid feedstock to a desired L-amino acid will not proceed to completion in most cases. The key to the development of a general and commercially successful transamination process for the pro- duction of L-amino acids lies in overcoming this problem of incomplete conversion of a 2-ketoacid to the desired L-amino acid. Driving the Reaction to Completion Solving the problem of incomplete conversion of 2-ketoacid starting material to a desired L-amino acid required one important observation regarding the substrate specificity of aminotransferases. Although L-glu- tamic acid is generally considered to be the amino donor for aminotrans- ferases that catalyze the transamination of a broad range of 2-ketoacids to L-amino acids, we have found that L-aspartic acid can also function com- petently as a general donor of an amino group with certain enzymes. We have worked extensively with an aminotransferase capable of using L-aspartic acid isolated from Escherichia coli. 5 When L-aspartic acid is used as the amino group donor for the trans- amination of a given 2-ketoacid, oxaloacetic acid is coproduced along with the desired L-amino acid. Oxaloacetate, unlike 2-ketoglutarate, is a /3-ketoacid, and as such can facilely be converted to pyruvic acid via an essentially irreversible decarboxylation step. This may be accomplished chemically by the use of certain metal ions or amines, thermally, or most preferably, enzymatically using the enzyme oxaloacetate decarboxylase. The coupled two-enzyme reaction is illustrated in Scheme 2. The important feature of this process is the decarboxylation of ox- aloacetate to pyruvate. It is this essentially irreversible decarboxylation that drives the entire process to completion to produce L-amino acids in quantitative yields from the appropriate 2-ketoacid precursors. The 5 C. Mavrides and W. Orr, Biochim. Biophys. Acta 336, 70 (1974). + C02
`
`

`

`44 IMMOBILIZED AMINOTRANSFERASES 483 pyruvic acid by-product is easily separated from the product mixture by crystallization of the L-amino acid or by ion-exchange methods. We have investigated several methods for decarboxylating oxaloace- tate, including catalysis by primary amines and divalent metal ions such as Mg 2+ , Mn z+ , and Zn z+ and the enzymatic decarboxylation by oxaloace- tate decarboxylase (OAD, EC 4.1.1.3). This chapter will focus on driving the overall reaction by the OAD-catalyzed decarboxylation of oxaloace- tate to produce pyruvate and the desired L-amino acid. Methodology Sources and Production of Enzymes Aminotransferases can be isolated from virtually any microbial, plant, or animal source. The most easily obtained enzymes are from porcine heart, yeast, and E. coli. However, the usefulness of the individual aminotransferases for amino acid production varies. The glutamic-ox- aloacetic aminotransferase from porcine heart (EC 2.6.1.1, aspartate aminotransferase), although very stable and commercially available, is of limited utility for the production of amino acids because of its high speci- ficity for L-glutamic acid, L-aspartic acid, and the corresponding 2-ke- toacids as substrates. 6 Other substrates are not transaminated at reason- able rates. Similarly, the commercially available porcine glutamic-pyruvic amino- transferase (EC 2.6.1.2, alanine aminotransferase) also exhibits the desir- able properties of high stability, high specific activity, and lack of severe inhibition even at substrate concentrations up to 0.4 M, but the enzyme cannot use L-aspartic acid as the amino group donor. Thus, a highly productive immobilized biocatalyst can be prepared using this amino- transferase, and it can be used for the production of e-alanine from pyruvic acid and L-glutamate, but the reaction cannot be driven to com- pletion by the coupling of oxaloacetate decarboxylase. These readily available enzymes have nonetheless been useful as model aminotrans- ferases in the development and design of biocatalytic transamination pro- cesses, and data for the immobilization and use of these enzymes in bioreactors will be presented. The microorganism E coli. is one of the most useful sources of amino- transferases. Of the four aminotransferases from this microorganism characterized to date, v we have found the so-called glutamic-oxaloacetic 6 I. W. Sizer and W. T. Jenkins, this series, Vol. 5, p. 677. 7 j. T. Powell and J. F. Morrison, Eur. J. Biochem. 87, 391 (1978).
`
`

`

`484 ENZYME ENGINEERING (ENZYME TECHNOLOGY) 44 TABLE I RELATIVE RATES OF PRODUCTION OF L-AMINO ACIDS USING ASPARTIC AMINOTRANSFERASE FROM E. coli a L-Amino acid Precursor Relative rate L-Glutamic acid 2-Ketoglutarate 225 L-Phenylalanine Phenylpyruvate 100 L-Tyrosine p-Hydroxyphenylpyruvate 130 L-Tryptophan lndolyl-3-pyruvate 150 L-2-Aminoadipic acid 2-Ketoadipate 22 L-4-Phenyl-2-aminobutanoic acid 4-Phenyl-2-ketobutyrate 17 L-Histidine Imidazole 3-pyruvate 3 a Amino group donor in all cases: L-aspartic acid. aminotransferase (EC 2.6.1.1) to be an extremely useful catalyst for e-amino acid production. This enzyme is the most stable of the E. coli aminotransferases, and is produced constitutively. 8,9 The enzyme is capa- ble of catalyzing the transamination of a large number of 2-ketoacids to e-amino acids using either L-glutamate or L-aspartate as cosubstrate. Ta- ble I lists the relative rate of transamination for the production of a variety of amino acids. Interestingly, for use as a biocatalyst in the production of aromatic amino acids such as L-phenylalanine, the E. coli glutamic-oxaloacetic aminotransferase (GOA) is superior to the enzyme which carries out this function metabolically, the so-called aromatic-amino-acid aminotrans- ferase (AA) (EC 2.6.1.5). The former enzyme is significantly more ther- mostable than the latter. In addition, the glutamic-oxaloacetic amino- transferase has catalytic rate constants for the transamination of phenylpyruvate to L-phenyalanine, p-hydroxypyruvate to L-tyrosine, and indole 3-pyruvate to L-tryptophan comparable to those for the aromatic aminotransferase. The K,~ for these aromatic amino acids is approxi- mately an order of magnitude higher for the GOA enzyme, but given that high concentrations of substrates are generally used in a biocatalytic pro- cess, the enzyme is functioning at its maximal catalytic rate. The enzyme oxaloacetate decarboxylase (OAD) has been isolated from three different sources: Pseudomonas putida ATCC 950,1° Micro- 8 C. Mavrides and W. On', J. Biol. Chem. 250, 4128 (1975). 9 S. Chesne, N. Monnier, and J. Pelmont, Biochimie 60, 403 (1978). ~0 A. A. Horton and H. L. Kornberg, Biochim. Biophys. Acta 89, 381 (1964).
`
`

`

`44 IMMOBILIZED AMINOTRANSFERASES 485 coccus luteus ATCC 4698,11 and Azotobacter vinelandii ATCC 478.12,13 We have purified all three of these enzymes, and have found the OADs isolated from P. putida and M. luteus as the most suitable for a biocata- lytic process based on the criteria of specific activity, stability to opera- tional conditions, activity after immobilization, and the ease of produc- tion and isolation. The enzymes may be obtained by modifications of published procedures, l°,ll Details of these purifications will be published elsewhere. 14 Immobilization of Aminotransferases and Assays for Activity A large number of immobilization methods have been examined for the immobilization of aminotransferases. Adsorption methods such as the binding of enzyme to ion-exchange resins failed to yield a stable immobi- lized enzyme preparation due to desorption of enzyme from the support. Entrapment in polymeric gels such as polyacrylamide produced active enzyme, but the activity was lower than that obtained using other meth- ods. Also, this method suffered from the slow loss of activity due to diffusion of the enzyme out of the polymeric matrix. Covalent coupling of the enzyme to inert supports has proved to be the method of choice for the aminotransferases we have examined to this point, providing immobi- lized enzymes with high retention of activity, long-term operational stabil- ity, and good mechanical properties. The successful immobilization of aminotransferases is highly depen- dent on the chemistry of the immobilization technique. Because of the requirement for the binding of the cofactor pyridoxal 5'-phosphate to the 8-amino group of a lysine residue, reagents such as glutaraldehyde, p-nitrophenyl esters, N-hydroxysuccinimidyl esters, and the like, which react with amines on the protein, can deactivate aminotransferases. For this reason, it is absolutely essential that the active site be protected during the immobilization by including pyridoxal 5'-phosphate and a 2- ketoacid (e.g., 2-ketoglutarate) in the reaction mixture. Alternatively, the coupling can be accomplished through carboxyl groups on the enzyme to amino groups on the support using a water-soluble carbodiimide. Methods for the immobilization of aminotransferases involving the covalent binding of enzyme through its carboxyl groups to primary H O. L. Krampitz and C. H. Werkman, Biochem. J. 35, 595 (1941). x2 S. S. Lee, R. H. Burris, and P. W. Wilson, Proc. Soc. Exp. Biol. Med. 50, 96 (1942). 13 G. W. E. Plaut and H. A. Lardy, J. Biol. Chem. 180, 13 (1949). 14 G. Edwards, J. Heier, and J. D. Rozzell, in preparation.
`
`

`

`486 ENZYME ENGINEERING (ENZYME TECHNOLOGY) 44 amines on the surface of an inert support gave immobilized aminotrans- ferases of high activity. Typical supports include porous glasses, porous ceramics, and porous diatomaceous earths. Oxaloacetate decarboxylase can be immobilized using similar procedures, also with good retention of catalytic activity. The following are sample procedures for the immobilization of various preparations of aminotransferase and oxaloacetate decarboxylase and procedures for assaying the immobilized enzymes. Immobilization of Glutamic-Oxaloacetic Aminotransferase from E. coli for the Production of L-Phenylalanine. Controlled-pore glass with an average pore size of 500 ,~ is converted to the aminopropyl form by reaction with triethoxy-3-aminopropylsilane using the aqueous procedure of Weetall. ~5 Aminopropyl glass (2.0 g) is added to 10 ml of a solution of sodium borate (5 mM) containing pyridoxal 5'-phosphate (0.5 mM). The pH is adjusted to 7.0, and 50 mg of a lyophilized glutamic-oxaloacetic aminotransferase from E. coli, which has been partially purified to a specific activity of approximately 2.5 units/mg (when assayed for the transamination of phenylpyruvate to L-phenylalanine) is added. After dis- solution of the enzyme, ethyl dimethylaminopropylcarbodiimide hydro- chloride (100 mg) is added, and the reaction mixture is shaken at room temperature for 45 min on a rotary shaker. At the end of this time the suspension is poured into a funnel with a glass frit (coarse porosity) and suction filtered. The support is washed successively with three portions of potassium phosphate buffer (50 mM, pH 7.0), three portions of 200 mM NaCI, and three more times with phosphate buffer. The immobilized bio- catalyst has an activity of 120 units (60 units/g) when assayed for the transamination of phenylpyruvate to L-phenylalanine with L-aspartic acid as the amino donor. The activity retained on immobilization is approxi- mately 95%. Assays for the activity of free aspartic aminotransferase are carried out by a method similar to that described by Mavrides and Orr. 8 In a typical assay procedure, 0.700 ml potassium phosphate buffer (50 mM, pH 7.0), 0.100 ml L-aspartate (pH 7.0, 200 mM), 0.100 ml 2-ketoacid solution (pH 7.0, 100 mM), 0.030 ml NADH solution (5 mg/ml in H20), 0.050 ml malate dehydrogenase solution (1000 U/ml in 50 mM potassium phosphate buffer, pH 7.0), and 0.010 ml pyridoxal 5'-phosphate (pH 7.0, 10 raM) are pipetted into a cuvette. A background rate is determined by monitoring the change in the absorbance at 340 nm as a function of time, and the reaction is initiated by the addition of the aminotransferase sam- 15 H. H. Weetall, this series, Vol. 34, p. 59.
`
`

`

`44 IMMOBILIZED AMINOTRANSFERASES 487 pie (0.010 ml). Activity of the enzyme is calculated by the following equation: Activity (International Units/ml) = (AOD340/min)(100/6.2) To assay the immobilized enzyme for activity, a weighed amount of immobilized enzyme (approximately 20-40 mg) is added to 1.0 ml of a solution containing potassium phosphate buffer (50 mM, pH 7.0), L-aspar- tic acid monosodium salt (100 mM), phenylpyruvic acid sodium salt (100 mM), and pyridoxal 5'-phosphate (0.1 mM). Aliquots of 10/zl are taken at 1-min time intervals and diluted into 990/zl of 2.5% NaOH (w/v). After mixing, 100/xl of this solution is added to a cuvette containing 900/xl of 2.5% NaOH, and the absorbance at 320 nm is read. The concentration of phenylpyruvate remaining at 1-min time intervals is calculated (e320 = 17.5 mM -1 cm -1) and from this data an aminotransferase activity can be calcu- lated using the following equation: Activity (International Units)/g AOD320 1000 min 17.5 x grams of immobilized enzyme An alternative assay is based on the quantitation of total oxaloacetate + pyruvate produced as a result of transamination (a small amount of pyruvate is produced by the spontaneous decarboxylation of oxaloace- tate) by stoichiometric reduction with NADH using a combination of malic dehydrogenase and lactic dehydrogenase. The incubation of immo- bilized aminotransferase and substrates is carried out as above, and 20-/zl aliquots are removed at 2-min time intervals and diluted into a cuvette containing 850/zl of 50 mM potassium phosphate buffer (pH 7.0), 50/xl of a 1000 unit/ml solution of lactate dehydrogenase, 50/zl of a 1000 unit/ml solution of malate dehydrogenase, and 30 /~l of a 5 mg/ml solution of NADH. The net change in the absorbance at 340 nm is a measure of the consumption of NADH and therefore the total oxaloacetate + pyruvate produced (8340 = 6.2 mM -~ cm-~). Activity of the immobilized enzyme is calculated by the following equation: Activity (International Units)/g AOD340 50 min 6.2 x grams of immobilized enzyme A small background correction must be made for the reduction of phenyl- pyruvate, which is a poor substrate for lactate dehydrogenase.
`
`

`

`488 ENZYME ENGINEERING (ENZYME TECHNOLOGY) 44 Immobilization of Glutamic-Oxaloacetic Aminotransferase from Por- cine Heart. Aminopropyl controlled-pore glass prepared as described above (0.500 g) is suspended in 5 ml of sodium borate solution (5 mM) containing pyridoxal 5'-phosphate (0.5 raM) and the pH is adjusted to approximately 7. Glutamic-oxaloacetic aminotransferase (14 rag, specific activity 18.2 units/nag) is dissolved in the solution, followed by the addi- tion of ethyl dimethylaminopropylcarbodiimide hydrochloride (50 rag). The reaction mixture is agitated at room temperature on a rotary shaker for 60 rain, after which time assays have shown negligible aminotrans- ferase activity in the solution. The controlled-pore glass particles are transferred to a coarse frit funnel and washed repetitively with 10-ml portions of water, potassium phosphate buffer (pH 7.0), three times with 0.2 M NaCI, and finally with phosphate buffer. The combined washings contain 8.8 mg of protein. The immobilized aminotransferase activity is 50 units when assayed for the transamination of L-aspartate and 2-ketoglu- tarate. The enzymatic activity retained after immobilization is 53%. The assay for aminotransferase activity is carried out as described earlier for the quantitation of oxaloacetate + pyruvate produced using NADH, lactate dehydrogenase, and malate dehydrogenase. Immobilization of Glutamic-Pyruvic Aminotransferase from Porcine Heart. Aminopropyl controlled-pore glass prepared as described above (0.500 g) is suspended in 5 rnl of sodium borate solution (5 raM) contain- ing pyridoxal 5'-phosphate (0.5 mM). The pH is adjusted to 7, and 30 mg of glutamic-pyruvic aminotransferase having a specific activity of 51 units/rag is dissolved in the solution. Ethyl dimethylaminopropylcarbo- diimide hydrochloride (50 rag) is added, and the reaction mixture is shaken on a rotary shaker for 60 min at room temperature. At the end of this time, the reaction mixture is transferred to a glass frit funnel (coarse porosity) and the porous glass particles are washed with water, potassium phosphate buffer (50 mM, pH 7.0), three times with NaC1 (200 mM), and again with potassium phosphate buffer. The combined washings contain 20 mg of protein (assayed by the method of Bradford'6), indicating that 10 mg has been retained on the support. The activity of the immobilized enzyme was 200 units/0.5 g of support. The enzymatic activity retained after immobilization is 40% of the activity of the same quantity of native enzyme in solution. Activity of the immobilized enzyme can be determined by quantitating the amount of pyruvate remaining in the reaction mixture after fixed times. A weighed amount of immobilized enzyme (approximately 20 rag) is suspended in a solution containing sodium pyruvate (100 raM), mono- 16 M. Bradford, Anal. Biochem. 72, 248 (1976).
`
`

`

`44 IMMOBILIZED AMINOTRANSFEILASES 489 sodium L-glutamate (100 mM), pyridoxal 5'-phosphate (0.1 mM), and potassium phosphate (50 mM). The pH of the solution is 7. Aliquots of 0.020 ml are removed at 1.0-min intervals and diluted into 0.180 ml of potassium phosphate buffer (50 mM). Ten microliters of this solution is added to a cuvette containing 0.900 ml potassium phosphate (pH 7.0, 50 raM), 0.050 ml of NADH solution (5 mg/ml in H20), and 0.040 ml lactate dehydrogenase (1000 U/ml in potassium phosphate buffer, pH 7.0, 50 mM). The net change in the absorbance at 340 nm is measured, and the average AOD340 per minute is calculated. Activity of immobilized glutamic-pyruvic aminotransferase is calculated by the following equa- tion: Activity (International Units)/g AOD340 500 min 6.2 × grams of immobilized enzyme An alternative assay for the activity of immobilized glutamic-pyruvic aminotransferase is described below which measures the amount of 2- ketoglutarate produced as a function of time. Approximately 10 mg of immobilized enzyme is suspended in 1.0 ml of a solution containing mono- sodium L-glutamate (I00 mM), sodium pyruvate (100 mM), potassium phosphate buffer (pH 7.0, 50 mM), and pyridoxal 5'-phosphate (0.1 mM). Aliquots of 10/zl are withdrawn at 1-min time intervals and diluted into a cuvette containing 800/xl of potassium phosphate buffer (pH 7.0), 50/xl of a 200 unit/ml solution of saccharopine dehydrogenase (Sigma Chemical Co., St. Louis, MO), 100/zl of a 100 mM solution of L-lysine, and 40/~1 of a 5 mg/ml solution of NADH. The net change in the absorbance at 340 nm is determined (e340 = 6.2 mM -1 cm-1). Immobilization of Oxaloacetate Decarboxylase from Pseudomonas putida. Cells of Pseudomonas putida are grown, harvested, and the en- zyme purified as described elsewhere, l: Aminopropyl glass (0.500 g) pre- pared as described earlier is suspended in 5 ml of 5 mM sodium borate buffer containing 10 mM MgC12, and 45 mg of a partially purified ox- aloacetate decarboxylase having a specific activity of 30 units/mg is added. Ethyl dimethylaminopropylcarbodiimide hydrochloride (125 mg) is added and the reaction mixture is agitated on a rotary shaker for 60 min at room temperature. Assays have shown that by this time the decline of oxaloacetate decarboxylase activity in the supernatant (due to immobili- zation of the enzyme on the support) has slowed significantly. Indepen- dent controls show a negligible decrease in activity of the OAD due to chemical modification by the carbodiimide. The reaction mixture is trans- ferred to a funnel with a glass frit (coarse porosity), and the immobilized
`
`

`

`490 ENZYME ENGINEERING (ENZYME TECHNOLOGY) 44 enzyme is washed with water, three times with Tris-HCl buffer (50 mM, pH 8.0), containing MgCI2 (10 mM), three times with NaCI solution (200 mM), and finally with Tris buffer. The combined washings contain 14 mg protein. The immobilized oxaloacetate decarboxylase contain 275 units/ 500 mg support. The enzymatic activity retained after immobilization is 30%. To determine the activity of immobilized oxaloacetate decarboxylase, a weighed amount of immobilized enzyme (approximately 10 mg) is added to 2.0 ml of a solution containing oxaloacetate (100 mM), Tris-HCl buffer (50 mM) with pH adjusted to 8.0, and MgCI2 (10 mM). The mixture is shaken at 25 °, 10-/zl aliquots are withdrawn at 1- to 2-min intervals and diluted into a cuvette containing 990 tzl of water, and the absorbance at 262 nm is measured. The net change in the absorbance at 262 nm corre- sponds to the consumption of oxaloacetate (B262 = 0.78 mM -1 cm-~). Activity of the immobilized enzyme is calculated by the following equation: Activity (International Units)/g AOD262 100 rain 0.78 x grams of immobilized enzyme Immobilization of Glutamic-Pyruoic Aminotransferase on Porous Diatomaceous Earth. Porous diatomaceous earth from Johns-Manville (R 640, 50/100 mesh) is boiled for 12 hr in 5% nitric acid and washed with deionized water; the fine particles are decanted, and dried for 12 hr in an oven at 110 °. The support is then converted to the aminopropyl derivative by the aqueous activation procedure of Weetall. j5 Aminopropyl support (5.0 g) is suspended in sodium borate (5 mM) containing 2-ketoglutarate (0.5 mM), pyridoxal phosphate (0.5 mM), and 110 mg glutamic-pyruvic aminotransferase (specific activity 49 units/mg, Lee Scientific). Ethyl di- methylaminopropylcarbodiimide-HC1 (250 mg) is added and the reaction mixture is placed on a rotary shaker for 2 hr at room temperature. At the end of this time, the support is transferred to a glass frit funnel (coarse porosity) and washed repeatedly with water and then 50 mM potassium phosphate buffer containing 0.1 mM pyridoxal phosphate. Assay for pro- tein in the combined washings ~6 has indicated that 80 mg of protein is bound to the support. The support contains 850 units of activity. The activity retained after immobilization is 22%. Immobilization of Glutamic-Oxaloacetic Aminotransferase on Acti- vated Porous Diatomaceous Earth. Aminopropyl porous diatomaceous earth (500 rag) prepared as described above is suspended in 2 ml of potas-
`
`

`

`44 IMMOBILIZED AMINOTRANSFERASES 491 sium phosphate buffer (50 mM, pH 7.0) containing 2-ketoglutarate (2 mM), pyridoxal phosphate (0.1 mM), and 26 mg partially purified glu- tamic-oxaloacetic aminotransferase from porcine heart (specific activity of 14 units/mg). Ethyl dimethylaminopropylcarbodiimide hydrochloride (I0 mg) is added and the reaction mixture is agitated on a rotary shaker for 1.5 hr. After washes as described above on a glass frit funnel with 50 mM phosphate buffer containing 0.5 M NaC1 and finally phosphate buffer, 3.3 mg of protein is found to be bound to the support by quantitation of the recovered protein. Assay of the bound aminotransferase as described earlier gives 19 units of activity on the support. The specific activity of the immobilized enzyme is 41%. Summary of Results. Using the carbodiimide method for immobiliza- tion, the three different aminotransferase (glutamic-oxaloacetic from por- cine heart, glutamic-oxaloacetic from E. coli, and glutamic-pyruvic from porcine heart) could all be immobilized to porous supports bearing pri- mary amine functional groups with good retention of activity. The results are summarized in Table II. The stability to operational conditions is also quite high. For GPA immobilized on porous glass, the immobilized enzyme showed little loss of activity over 6 months of operation as described below. Measurement of the Long-Term Stability of the Immobilized En- zymes. Measurements to determine the operational stability of immobi- lized enzyme were carried out by pumping substrate mixtures through a packed bed of biocatalyst at 25 ° and quantitating the amount of product in the effluent stream as outlined previously. In the case of glutamic-pyruvic aminotransferase, a solution of L-glutamate monosodium salt (200 mM), sodium pyruvate (400 mM), and pyridoxal phosphate (0.1 mM) with a pH between 7.0 and 7.5 was used as the substrate mixture. Activity of the TABLE II IMMOBILIZATION ON POROUS SUPPORTS Activity of immobilized Loading enzyme Enzyme Support (mg/g) (units/g) GOA (E. coli) Glass 25 60 GOA (porcine) Glass 10 100 GPA Glass 20 400 OAD Glass 30 550 GPA Diatomaceous earth 16 170 GOA Diatomaceous earth 6 38
`
`

`

`492 ENZYME ENGINEERING (ENZYME TECHNOLOGY) 44 IOO 80 % Activity 60 / 40 20 I I I I I I I 5 I0 20 30 40 50 60 Time (days) FIG. 1. Long-term operational stability of immobilized glutamic-pyruvic aminotrans- ferase. biocatalyst was determined by quantitating both the amount of 2-ketoglu- tarate produced and the amount of pyruvate consumed by the methods described earlier. The column was operated intermittently over 2 months for periods varying between 1