Journal of Biotechnology, 13 (1990) 305-314
305
Elsevier BlOTEC 00494
Immobilization of enzymes with polyaziridines Louis L. Wood, Carrington S. Cobbs, Leon Lantz II, Lin Peng and Gary 1. Calton Rhone Poulenc Research Center, Savage, MD 20763, U.S.A.
(Received 30 August 1989; accepted 7 November 1989)
Summary
A novel method of enzyme immobilization using a low molecular weight prepolymer of tri-functional aziridines which can immobilize enzymes both by covalent attachment and entrapment within a gel matrix is described. The enzymes are immobilized on a solid support and exhibit an excellent retention of enzymatic activity. The immobilization procedure is essentially a single step process which can be easily performed at room temperature or 4 0 C in either aqueous solution or in an inert organic solvent. The polyaziridines used in the immobilization are nontoxic, available in bulk at low cost and completely miscible with water and many organic solvents, thus providing one of the most satisfactory methods of immobilization available. Enzyme immobilization; Polyaziridine
Introduction The use of immobilized enzymes for the production of specialty chemicals has not achieved a high rate of commercial acceptance, as only a small number of these processes are utilized. These include the high volume production of high fructose corn syrup (Antrim et al., 1979); the production of the amino acids, aspartic acid (Tosa et al., 1973) and alanine (Takamatsu et al., 1981); the production of 6-aminopenicillanic acid (Mauz et al., 1984); the production of malic acid (Takato et al., 1980); and the recently announced production of (R)-2-chloropropionic acid (Klibanov and Kirchner, 1986). Other chemicals, such as phenylalanine have been
Correspondence to: G.J. Calton, [S][R]CHEM, 5331 Landing Road, Elkridge, MD 21227, U.S.A.
0168-1656/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
306
made by immobilized enzymes on a large scale in the past (Wood and Calton, 1986). There are also a number of chemicals which are being investigated on a small scale, such as glycidol (Walts et aI., 1987), but as yet, these have not moved to a commercial scale. Several reasons have been advanced for this situation. Foremost among these is a lack of suitable targets (Akiyama et aI., 1988). Since the production of new chemicals has not ceased, it can be assumed that enzymatic production of specialty chemicals is simply not economically competitive. The reasons for this lack of competition may lie in three areas: the ability of an enzyme to carry out a suitable chemical reaction; the cost of the enzyme; and the cost of using the enzyme. Recombinant DNA technology presents an obvious solution to the first of these problems by providing a method for overproduction of any enzyme desired. Enzymes are already known which will carry out all major types of chemical reactions; however, they are not suitable for all substrates at useful levels of activity (Kieslich, 1976). Some investigators believe that an enzyme can be found to carry out any chemical reaction (Calton, 1988). The third major variable preventing commercial exploitation is the cost of using the enzyme which may be ameliorated by immobilization of the enzyme. Methods of immobilization which are applicable to commercial processes are rarely described in the literature. This is not due to lack of effort as the literature contains extensive reviews of both methods for immobilization and enzymes which have been immobilized (Mosbach, 1987). Nevertheless, relatively few of these methods are suitable for industrial use. Again, there are three main problems. First, there is a lack of suitable stability of the immobilized enzyme. In the current major application of immobilized enzymes, the production of high fructose corn syrup, the standard is a minimum of 2000 kg of product per kg of enzyme (Antrim et aI., 1979), whereas in the immobilized cell production of aspartic acid, over 200,000 kg of product per kg of immobilized microbe can be obtained (Wood and Calton, 1984). Second, the characteristics of immobilized catalysts in general are not suitable for industrial use. The particles are either soft, irregularly shaped or unstable in the process stream. Processes for both microbial cells (Calton et aI., 1987) and enzymes (Mauz et aI., 1984; Noetzel et aI., 1984) are known which have overcome this problem. Third, the conditions necessary to immobilize the enzyme require conditions that are deleterious to the enzyme, resulting either in higher costs, or require conditions that are too hazardous to carry out on a large scale at reasonable cost. We now report a novel and extremely useful method of immobilization which obviates the above problems.
Materials and Methods Materials
Prepolymers of the triaziridine (TA), trimethylolpropane-tris-(,B-( N-aziridinyl) propionate), and the hydroxylated triaziridine (HTA), pentaerythritol-tris-(,B-(N-
307
aziridinyl)propionate), are commercially available under the trade names, XAMA2™ and XAMA-7™, respectively, from Hoechst Celanese (Richmond, VA). Lipase MY was obtained from Meito Sangyo (Tokyo, Japan). D-amino acid oxidase (DAAO), horse radish peroxidase, alkaline phosphatase and ,8-galactosidase were purchased from Sigma Chemical Co. (St. Louis, MO). Candida rugosa lipase was purchased from Biocatalysts Ltd. (Pontypridd, u.K.). DP-l ion exchange resin was obtained from Rohm and Haas (Philadelphia, PA) and cellulose beads were obtained from the Savannah Sugar Co. (Savannah, GA). Bradford protein assay reagent was purchased from BioRad (Richmond, CA). The C-18 RAC™ column was purchased from Whatman, Hillsboro, OR. All other chemicals and reagents were obtained from Fisher Scientific (Silver Spring, MD), and were of analytical grade or higher purity. Polyaziridine immobilization of D-amino acid oxidase (DAAO): Method 1 One gram of crude enzyme extract from pig kidney (10 mg ml- J protein) containing DAAO was mixed with 3 g of H 20. HTA, 2 g, was thoroughly mixed with the aqueous enzyme solution for 2 min. The prepolymer enzyme mixture was then added to 1.5 g of dry cellulose sponge which served as a support medium, and mixed thoroughly. The polymer was allowed to cure at 4°C overnight in a sealed container. Following polymerization, the immobilized enzyme was washed with 10 volumes of phosphate buffered saline (PBS) and tested for substrate conversion. DAAO immobilization: Method 2 A mixture of 1.0 g of aqueous DAAO prepared as above and 0.1 g of the HTA prepolymer were added to 0.25 g of dry, cellulose sponge. After overnight polymerization and washing as in method 1, the enzyme catalyst was tested for activity. DAAO immobilization: Method 3 One gram of DAAO solution was mixed with 100 mg of HTA. This solution was then mixed gently with 2.0 g of cellulose beads, previously dried to 45% moisture, and allowed to cure overnight in a sealed container at 4°C. DAAO immobilization: Method 4 To 30 ml of cellulose beads dried to 42% moisture was added 3 ml of crude enzyme solution containing 30 mg ml- J of DAAO and 2.5 g of HTA. The mixture was stirred gently with a stainless steel spatula and stored at 4°C in a capped glass beaker until assay. HTA optimization for DAA 0 immobilization HTA in varying quantities was added to cellulose beads according to method 3 and assayed. Immobilized DAAO batch assay Initial tests were carried out in a batch mode in 50 ml flasks. The substrate was 10 mM D,L-phenylalanine adjusted to pH 8.3 with NH 4 0H. The production of
308
phenylpyruvate was monitored as a function of time by removing samples and diluting 1: 100 into 1 N NaOH. The absorbance of the resultant solution was measured at 320 nm. Immobilized DAAO column assay A 1.5 X 30 cm jacketed column was filled with 30 ml of catalyst from the polyaziridine immobilization procedure, method 3. A solution of 1 mM D-Phe containing 1 mM sodium phosphate, pH 8.3, and 100 pJ 1- 1 of catalase was passed over the column at 20.5 ml h -1. The substrate solution was kept at 4DC to prevent the breakdown of the catalase and the D-Phe before passage over the column. The activity was monitored spectrophotometrically by absorbance at 320 nm of a 1: 10 dilution of the column effluent in 1 N NaOH. The column was maintained at 20 DC by a circulating water bath. Samples were obtained over a 30 min period daily while simultaneously determining the flow rate. Immobilization of lipase MY in aqueous phase Lipase MY, 0.15 g, dehydrated Amberlite DP-1, 0.5 g, and 0.5 ml of 0.01 M phosphate buffer, pH 6.5, were mixed thoroughly with 100 mg of HTA at 4 DC for 20 min. The mixture was allowed to stand for 4 h at room temperature after which the immobilized enzyme was collected and washed with deionized water. Lipase assay The immobilized enzyme was shaken with 10 ml of 0.5 M phosphate buffer, pH 6.5, and 5 ml of toluene containing 1 g of methyl 2-(4-hydroxyphenoxy)propionate (HPPA-Me) for 4 h. The remaining ester and the acid produced by hydrolysis of HPPA-Me in aqueous solution was determined using a Whatman C18 RAC column. The mobile phase used was 25% methanol, 0.5 g 1- 1 K 2 HP04 , pH 3.0, at a flowrate of 1.0 ml min -1 and UV detection at 220 nm. All dilutions were made into deionized water. Under these conditions the HPPA-Me eluted at 11.0 min, and HPPA eluted at 5.0 min. Immobilization of lipase MY in toluene Lipase MY, 0.15 g, dehydrated Amberlite DP-1, 0.5 g, and HTA, 50 mg, were mixed in 10 ml of toluene. The mixture was shaken (100 rpm) at room temperature for 4 h and the toluene was then allowed to evaporate in a fume hood at room temperature. The immobilized enzyme was collected, washed, and assayed as above. This preparation was also repeated using TA instead of HTA. Immobilization of lipase MY in trichloroethane Lipase MY, 0.15 g, dehydrated Amberlite DP-1, 0.5 g, and HTA, 75 mg, were mixed in 10 ml of 1,1,1-trichloroethane. The mixture was shaken at room temperature for 4 h and the solvent was evaporated, the immobilized enzyme collected, washed, and assayed as above. Immobilization of tyrosinase Mushroom tyrosinase, 5 mg, was dissolved in 0.6 ml of water and mixed well with 500 mg of HTA. This was homogeneously dispersed over dehydrated Amberlite
309
DP-1 and allowed to cure for 16 h at 4°C. After extensive washing with PBS, the immobilized enzyme was mixed with 1.5 mM tyrosine and 15 mM ascorbic acid in 25 ml of 0.1 M potassium phosphate buffer, pH 7.0. The assay mixture was incubated at 25°C with continuous aeration. L-DOPA formation was monitored by HPLC using a C18 column with a mobile phase of 0.1 % trifluoroacetic acid, 10% acetonitrile and 90% water. Under these conditions, tyrosine elutes at 6.13 min and L-DOPA elutes at 5.58 min. The control used soluble enzyme. Immobilization of horseradish peroxidase in dioxane Horseradish peroxidase, 12.5 mg (1000 units), and dehydrated Amberlite DP-1, 1.0 g, were lyophilized in 10 ml of 0.01 M phosphate buffer, pH 6.0. The lyophilized mixture was shaken with 50 mg of TA in 10 ml of dioxane at room temperature for 4 h and the dioxane was then allowed to evaporate. The immobilized enzyme, approximately 2 g, was collected and washed with phosphate buffered saline. The immobilized enzyme, 0.3 g, was assayed by following the conversion of pyrogallol, 25 mg, to purpurogallin (f: = 2.47 cm2 mmol- 1 ) spectrophotometrically at 420 nm in the presence of 50 }L1 of 30% hydrogen peroxide in 10 ml of phosphate buffer, pH 6.5. One unit will form 1.0 mg of purpurogallin in 20 s at pH 6.0, at 20°e. Immobilization of horseradish peroxidase in buffer HTA, 0.2 g, was mixed with horseradish peroxidase, 2.4 mg, 192 units, in 100 }LI of either (2-[N-morpholino]ethanesulfonic acid buffer, pH 5.5, phosphate buffer, pH 7.2, or N-tris-(hydroxymethyl)methyl-3-aminopropanesulfonic acid buffer, pH 9.0, for 5 min and then mixed with an additional 800 }LI of the same buffer and immediately mixed with 0.5 g (0.75 ml) of cellulose beads. The beads were allowed to cure at 4°C for 18 h, washed with 50 volumes of PBS and assayed by the phenolaminoantipyrine method. A solution of 0.17 M phenol and 0.0025 M 4-aminoantipyrine, 1.4 ml, was added to 1.5 ml of 0.0017 M hydrogen peroxide in 0.1 M phosphate buffer, pH 7.0, and 8 mg of the buffer immobilized horseradish peroxidase. The solution was stirred for 1 min and the absorbance at 510 nm was determined. Alkaline phosphatase immobilization Alkaline phosphatase, 0.9 mg, was lyophilized with 1 g of dehydrated Amberlite DP-1 in 10 ml of PBS. The mixture was shaken with 50 mg of HTA in 10 ml of toluene at room temperature for 4 h before toluene was evaporated. Approximately 2 g of immobilized enzyme was collected. The immobilized enzyme, 0.5 g, was assayed spectrophotometrically at 405 nm with 25 mg of p-nitrophenyl phosphate in 10 ml of phosphate buffer, pH 8.2, 0.5 M phosphate, 0.2 M ethanolamine). One unit will hydrolyze 1.0 }Lmol of p-nitrophenyl phosphate per min at pH 10.4 at 37°C to p-nitrophenol (f: = 13,200 cm2 mmol- 1 ). Immobilization of f3-galactosidase in toluene Galactosidase, 0.5 mg, was lyophilized with 1 g of Amberlite DP-1 in 5 ml of PBS. The enzyme was immobilized with 50 mg of HTA in 10 ml of toluene. Approximately 2 g of immobilized enzyme was collected after work-up.
310
The immobilized enzyme was assayed spectrophotometrically at 405 nm with 25 mg of p-nitrophenyl ,B-D-galactopyranoside in 10 ml of 0.5 M phosphate buffer, pH 7.0. One unit will hydrolyze 1.0 !Lmol of o-nitrophenyl-,B-D-galactoside per min to give o-nitrophenol ({ = 3100 cm2 mol-I).
Results and Discussion
Watanabe and Royer (1983) used aziridines as crosslinking agents to form polyethyleneimine bonded coatings on various matrices for chromatography or as enzyme supports (Watanabe and Royer, 1983); however, they did not use aziridines for the purpose of immobilizing proteins to an insoluble matrix. Porath and Axen had suggested earlier that aziridine groups might be used for coupling enzymes to a matrix (Porath and Axen, 1976); however, they presented no data for aziridines and were unsuccessful in their attempts to use three membered thio analogs. Thus, the value of the use of aziridines for the immobilization of enzymes lay unanswered, although the materials necessary for its investigation had been available for many years. The aziridine prepolymer system, described here, is readily prepared by the Michael type addition of an activated bond to aziridines as shown in Fig. 1 (Heissberger 1964; Mark et aI., 1964). The monomer can react with itself to form low molecular weight prepolymers. The commercially available prepolymers in the polyaziridine system vary in degree of solubility in organic solvents and range in size from less than 500 mol. wt. to considerably greater (Sello and Stevens, 1980). The polyaziridines, as purchased, may be stored for long periods of time with minimal precautions and are quite inexpensive. The order of reaction of the aziridine moiety with other functional groups is: COOH » SH > OH > NH 2 • The reactions of the aziridine and these functional groups is given in Fig. 2. The outstanding value of the polyaziridine system is the mild conditions under which aziridines will react and the simple method by which the prepolymer can be used to immobilize enzymes. Hazardous chemicals are not used, nor is a fume hood required unless the immobilization is to be carried out in the presence of an organic solvent. The availability of a wide range of pH values to immobilize is often an important factor in optimization of the retention of activity and stability of an enzyme. Polyaziridines may be crosslinked at temperatures below 40°C in the presence of relatively large volumes of water while still maintaining the activity of
o
RCH2Cf"2oi',";CH~3
• 3 [ONH ]
RCH2C~H20CCH2CH2-N13
•
R = CH 3 . TA = OH. HTA
Fig. 1. Preparation of aziridine prepolymers.
311
R ,-N
1
--J
o
0 II
11
+R Z-C-O-H ----- R ,-N-CH Z-CH Z-O-C-R Z I
H
R
, -N1+RZ-O-H~R --J
Z 1-N-CHZ-CHZ-O-R , H Fig. 2. Reactions of polyaziridine with other functional groups.
the enzyme. The pH of the polyaziridine immobilization reaction may range from 2 to 10 but polymerization is more rapid as the pH is lowered or at high polyaziridine concentrations (Table 1). At concentrations of 1 part HTA in 10 parts water or at below pH 8.0, a gel network was not obtained. The enzyme can be covalently attached to the polymer network while the network will also crosslink with functional groups on other surfaces, thus allowing the covalent attachment of enzymes to a wide variety of supports. The networks formed contain the immobilized enzymes which still have useful levels of their original activities. In general, the resultant gel composites have sufficient openness and hydrophilicity to allow free migration of chemical substrates for interaction with the immobilized species, and thus permit high levels of transformations to take place. Low molecular weight substrates can freely diffuse to the enzyme; however, for enzymes with a high turnover number, such as horseradish peroxidase, high flow rates are needed to overcome the diffusional limitation. The immobilized enzyme can also be taken to dryness by evaporation of solvent or lyophilization, while retaining most of its
TABLE 1 EFFECT OF HTA CONCENTRATION AND pH ON HTA POLYMERIZATION HTA (g)
H 2O (g)
pH
Gel time
Comments
10
10
10
45 s
2
10
9.5
26 min
2
10
8.5
13 min
2
10
8.0
6 min
2
10
7.0
2 min
1
10
9.5
4h
Exotherm 55 ° C Rigid gel Exotherm 30 D C Rigid gel pH adjust with 2 g KH 2 P04 Soft gel HCl pH adjust with rigid gel, exotherm to 40 D C HCl pH adjust with rigid gel, exotherm to 40°C Rigid gel
312 TABLE 2 EFFECT OF VARIATION IN HTA, PROTEIN AND SUPPORT CONCENTRATION ON THE IMMOBILIZATION AND RETENTION OF ACTIVITY OF DAAO BY HTA Method (mg)
Protein (mg)
HTA (mg)
Support Cellulose
Quantity (mg)
1 2 3 4
10 10 10 30
2000 100 100 2500
sponge sponge bead bead
1500 250 2000 15000
Retention of activity (%)
Ratio HTA/matrix
Protein/ HTA
35 43 48 53
1.33 0.4 0.05 0.116
0.005 0.1 0.1 0.012
original activity. The drying process has the advantage of providing strong, well crosslinked, insoluble compositions in the form of coatings as described here or membranes (data not shown) having high concentrations of active, immobilized enzyme. The preparation of a homogeneous solution of the enzyme and the prepolymer is easily obtained by stirring a mixture of the two in a small container. The solution may then be poured onto a suitable glass or teflon tray to form membranes or onto a suitable support, such as an ion exchange resin, silica, glass beads or other rigid or semi-rigid matrix. After addition of the solution to a support, additional stirring to obtain a homogeneous dispersion is carried out. The mixture or membrane may then be set aside at either 4°C or room temperature and allowed to cure overnight. The immobilization of the enzyme is complete and appropriate washing steps may then be carried out. The presence of the aziridine moiety and its nitrogen atom complicates the normal analyses of protein by colorimetric or absorbance readings and thus, the determination of the quantity of protein bound cannot be easily made. TABLE 3 RETENTION OF ACTIVITY AFTER ENZYME IMMOBILIZATION BY HTA AND TA Enzyme
Carrier
Solvent
Retention of activity
DAAO DAAO Lipase MY Lipase MY Tyrosinase
DP-1 Cellulose DP-1 DP-1 DP-1 DP-1 DP-1 DP-1 Cellulose Cellulose Cellulose DP-1 DP-l
Buffer Buffer Buffer Toluene Buffer Toluene Trichloroethane Dioxane pH 5.5 buffer pH 7.2 buffer pH 9.0 buffer Toluene Toluene
48
Polyaziridine
(%)
Candida rugosa lipase Lipase MY Horseradish peroxidase Horseradish peroxidase Horseradish peroxidase Horseradish peroxidase Alkaline phosphatase ,B-Galactosidase
48 65 68 20 70 80 2 18 1 1 2 10
HTA HTA HTA TAor HTA HTA HTA HTA TA HTA HTA HTA HTA HTA
313
As shown in Table 2, preliminary experiments using vastly different ratios of HTA to matrix or to protein were effective in the immobilization of DAAO and resulted in high retained activity. The stability of the DAAO on cellulose beads was examined using 10 ml D,L-Phe pH 8.0, at 20 0 e and 30°C. The stability of the enzyme was excellent at these temperatures for a 24 h period. At 20 o e, 96% of the original activity was present, while at 30 o e, 92% of the activity remained. The immobilization of DAAO in aqueous systems gave high yields of retained activity (Table 3). Excellent yields of retained activity were obtained when Lipase MY was immobilized in aqueous solution. When Lipase MY was immobilized in toluene, the yield of retained activity was equal to that obtained in buffer. The use of trichloroethane to immobilize Lipase MY gave even higher retained activity (80%) than immobilization in buffer (Table 3). The use of organic solvents allowed the preparation of a more homogeneous dispersion of the polymer/enzyme mix. In buffer systems, large aggregates were observed with this enzyme which were undesirable in a fixed-bed reactor, especially in large scale preparations. The retained activity of Lipase MY did not differ greatly whether the immobilization took place in buffer or organic solvents, including both toluene and trichloroethane (Table 3). Thus, the prepolymer may be used either in water or inert organic solvents without adversely affecting the immobilization of an individual enzyme. It should be noted that TA has a greater solubility in organic solvents than HTA; conversely TA is not very soluble in aqueous systems. The immobilization of ,B-galactosidase in an organic solvent provided a catalyst which retained 10% of its original activity (Table 3), however, immobilized horseradish peroxidase retained only 2% of its activity and alkaline phosphatase retained only 2% of its initial activity. Further investigation of horseradish peroxidase showed that it did not immobilize well at an initial pH of 7.2 or 9.0 in aqueous systems; however, at an initial pH of 5.5, the retention of activity was 18%. The complete optimization of horseradish peroxidase was not attempted, but it can be seen that the proper selection of the pH of the enzyme solution before immobilization can be quite valuable for maximum activity retention in this system. The polyaziridine system demonstrates wide applicability to the immobilization of enzymes under varying conditions. The details of these immobilizations, their stability and their commercial utility will be reported in due course.
References Akiyama, A., Bednarski, M., Kim, N-J., Simon, E.S., Waldmann, H. and Whitesides, G.M. (1988) Enzymes in organic synthesis. Chemtech 18, 627-634. Antrim, R.L., Colilla, W. and Schnyder, BJ. (1979) Glucose isomerase production of high-fructose syrups. App\. Biochem. Bioeng. 2, 97-155. Calton, GJ. (1988) Towards the Golden Age of Enzymatic Chemistry. Biofutur. Fevrier, 41-46. Calton, G.J., Wood, L.L. and Campbell, M.L. (1987) Phenylalanine production via polyazetidine immobilized E. coli: optimization of cell loading. Methods Enzymo\. 136, 497-503.
314 Heissberger, A. (1964) Heterocyclic Compounds with Three-and Four-Membered Rings, vol. 1, J. Wiley and Sons, New York, pp. 542-550. Kieslich, K. (1976) Microbial Transformations of Non-Steroid Cyclic Compounds, John Wiley and Sons, New York. Klibanov, A.M. and Kirchner, G. (1986) Enzymatic production of optical isomers of 2-halopropionic acids. US Patent 4,601,987. Mark, H.P., Gaylord, N.G. and Bikales, N.M. Eds. (1964) Encyclopedia of Polymer Science and Technology, vol. 1, J. Wiley & Sons, New York, pp. 735-736. Mauz, 0., Noetzel, S. and Sauber K. (1984) New synthetic carriers for enzymes. Ann. N.Y. Acad. Sci. 434, 251-253. Mosbach K., Ed. (1987) Immobilized enzymes and cells. Methods Enzymol 135-138. Noetzel, S., Mauz, O. and Sauber, K. (1984) Macroporous polymer beads and their use for enzyme immobilization. Patent German Offen. DE 3404021. Porath, J. and Axen, R. (1976) Immobilization of enzymes to agar, agarose and sephadex supports. Methods Enzymol. 44, 19-45. Sello, S.B. and Stevens, c.Y. (1980) Aziridine-terminated polymers as wool shrink-proofing agents. Quinquenn. Int. Wool Text. Res. Conf., 6th Fiche 6/G/9 [CA94:85-565b]. Takamatsu, S., Yamamoto, K., Tosa, T. and Chibata, I. (1981) Stabilization of L-aspartic .a-decarboxylase activity of Pseudomonas dacunhae immobilized with carrageenan. J. Ferment. Technol. 59, 489-493. Takata, I., Yamamoto, K., Tosa, T. and Chibata, I. (1980) Immobilization of Brevibacteriumflavum with carrageenan and its application for continuous production of L-malic. Enzyme Microb. Technol. 2, 30-36. Tosa, T., Sato, T., Mori, T., Matsuo, Y., and Chibata, I. (1973) Continuous production of L-aspartic acid by immobilized aspartase. Biotech. Bioeng. 15, 69-84. Walts, A.E., Fox, E.M. and Jackson, C.B. (1987) (R)-Glycidyl butyrate: evolution of a laboratory procedure to an industrial process. Biotech USA 1987, Online International Inc., London, pp. 91-98. Watanabe, K. and Royer, G.P. (1983) Polyethylenimine/silica gel as an enzyme support. J. Mol. Catal. 22, 145-152. Watanabe, K., Chow, W.S. and Royer, G.P. (1982) Column chromatography on polyethylenimine-silica: rapid resolution of nucleotides and proteins with short columns and low pressures. Anal. Biochem. 127, 155-158. Wood, L.L. and Calton, GJ. (1984) A Novel Method of Immobilization and Its Use in Aspartic Acid Production. Bio/Techn. 2, 1081-1084. Wood, L.L. and Calton, G.J. (1986) The production of L-phenylalanine by polyazetidine immobilized microbes. Bio/Techn. 4, 317-320.
305
Elsevier BlOTEC 00494
Immobilization of enzymes with polyaziridines Louis L. Wood, Carrington S. Cobbs, Leon Lantz II, Lin Peng and Gary 1. Calton Rhone Poulenc Research Center, Savage, MD 20763, U.S.A.
(Received 30 August 1989; accepted 7 November 1989)
Summary
A novel method of enzyme immobilization using a low molecular weight prepolymer of tri-functional aziridines which can immobilize enzymes both by covalent attachment and entrapment within a gel matrix is described. The enzymes are immobilized on a solid support and exhibit an excellent retention of enzymatic activity. The immobilization procedure is essentially a single step process which can be easily performed at room temperature or 4 0 C in either aqueous solution or in an inert organic solvent. The polyaziridines used in the immobilization are nontoxic, available in bulk at low cost and completely miscible with water and many organic solvents, thus providing one of the most satisfactory methods of immobilization available. Enzyme immobilization; Polyaziridine
Introduction The use of immobilized enzymes for the production of specialty chemicals has not achieved a high rate of commercial acceptance, as only a small number of these processes are utilized. These include the high volume production of high fructose corn syrup (Antrim et al., 1979); the production of the amino acids, aspartic acid (Tosa et al., 1973) and alanine (Takamatsu et al., 1981); the production of 6-aminopenicillanic acid (Mauz et al., 1984); the production of malic acid (Takato et al., 1980); and the recently announced production of (R)-2-chloropropionic acid (Klibanov and Kirchner, 1986). Other chemicals, such as phenylalanine have been
Correspondence to: G.J. Calton, [S][R]CHEM, 5331 Landing Road, Elkridge, MD 21227, U.S.A.
0168-1656/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
306
made by immobilized enzymes on a large scale in the past (Wood and Calton, 1986). There are also a number of chemicals which are being investigated on a small scale, such as glycidol (Walts et aI., 1987), but as yet, these have not moved to a commercial scale. Several reasons have been advanced for this situation. Foremost among these is a lack of suitable targets (Akiyama et aI., 1988). Since the production of new chemicals has not ceased, it can be assumed that enzymatic production of specialty chemicals is simply not economically competitive. The reasons for this lack of competition may lie in three areas: the ability of an enzyme to carry out a suitable chemical reaction; the cost of the enzyme; and the cost of using the enzyme. Recombinant DNA technology presents an obvious solution to the first of these problems by providing a method for overproduction of any enzyme desired. Enzymes are already known which will carry out all major types of chemical reactions; however, they are not suitable for all substrates at useful levels of activity (Kieslich, 1976). Some investigators believe that an enzyme can be found to carry out any chemical reaction (Calton, 1988). The third major variable preventing commercial exploitation is the cost of using the enzyme which may be ameliorated by immobilization of the enzyme. Methods of immobilization which are applicable to commercial processes are rarely described in the literature. This is not due to lack of effort as the literature contains extensive reviews of both methods for immobilization and enzymes which have been immobilized (Mosbach, 1987). Nevertheless, relatively few of these methods are suitable for industrial use. Again, there are three main problems. First, there is a lack of suitable stability of the immobilized enzyme. In the current major application of immobilized enzymes, the production of high fructose corn syrup, the standard is a minimum of 2000 kg of product per kg of enzyme (Antrim et aI., 1979), whereas in the immobilized cell production of aspartic acid, over 200,000 kg of product per kg of immobilized microbe can be obtained (Wood and Calton, 1984). Second, the characteristics of immobilized catalysts in general are not suitable for industrial use. The particles are either soft, irregularly shaped or unstable in the process stream. Processes for both microbial cells (Calton et aI., 1987) and enzymes (Mauz et aI., 1984; Noetzel et aI., 1984) are known which have overcome this problem. Third, the conditions necessary to immobilize the enzyme require conditions that are deleterious to the enzyme, resulting either in higher costs, or require conditions that are too hazardous to carry out on a large scale at reasonable cost. We now report a novel and extremely useful method of immobilization which obviates the above problems.
Materials and Methods Materials
Prepolymers of the triaziridine (TA), trimethylolpropane-tris-(,B-( N-aziridinyl) propionate), and the hydroxylated triaziridine (HTA), pentaerythritol-tris-(,B-(N-
307
aziridinyl)propionate), are commercially available under the trade names, XAMA2™ and XAMA-7™, respectively, from Hoechst Celanese (Richmond, VA). Lipase MY was obtained from Meito Sangyo (Tokyo, Japan). D-amino acid oxidase (DAAO), horse radish peroxidase, alkaline phosphatase and ,8-galactosidase were purchased from Sigma Chemical Co. (St. Louis, MO). Candida rugosa lipase was purchased from Biocatalysts Ltd. (Pontypridd, u.K.). DP-l ion exchange resin was obtained from Rohm and Haas (Philadelphia, PA) and cellulose beads were obtained from the Savannah Sugar Co. (Savannah, GA). Bradford protein assay reagent was purchased from BioRad (Richmond, CA). The C-18 RAC™ column was purchased from Whatman, Hillsboro, OR. All other chemicals and reagents were obtained from Fisher Scientific (Silver Spring, MD), and were of analytical grade or higher purity. Polyaziridine immobilization of D-amino acid oxidase (DAAO): Method 1 One gram of crude enzyme extract from pig kidney (10 mg ml- J protein) containing DAAO was mixed with 3 g of H 20. HTA, 2 g, was thoroughly mixed with the aqueous enzyme solution for 2 min. The prepolymer enzyme mixture was then added to 1.5 g of dry cellulose sponge which served as a support medium, and mixed thoroughly. The polymer was allowed to cure at 4°C overnight in a sealed container. Following polymerization, the immobilized enzyme was washed with 10 volumes of phosphate buffered saline (PBS) and tested for substrate conversion. DAAO immobilization: Method 2 A mixture of 1.0 g of aqueous DAAO prepared as above and 0.1 g of the HTA prepolymer were added to 0.25 g of dry, cellulose sponge. After overnight polymerization and washing as in method 1, the enzyme catalyst was tested for activity. DAAO immobilization: Method 3 One gram of DAAO solution was mixed with 100 mg of HTA. This solution was then mixed gently with 2.0 g of cellulose beads, previously dried to 45% moisture, and allowed to cure overnight in a sealed container at 4°C. DAAO immobilization: Method 4 To 30 ml of cellulose beads dried to 42% moisture was added 3 ml of crude enzyme solution containing 30 mg ml- J of DAAO and 2.5 g of HTA. The mixture was stirred gently with a stainless steel spatula and stored at 4°C in a capped glass beaker until assay. HTA optimization for DAA 0 immobilization HTA in varying quantities was added to cellulose beads according to method 3 and assayed. Immobilized DAAO batch assay Initial tests were carried out in a batch mode in 50 ml flasks. The substrate was 10 mM D,L-phenylalanine adjusted to pH 8.3 with NH 4 0H. The production of
308
phenylpyruvate was monitored as a function of time by removing samples and diluting 1: 100 into 1 N NaOH. The absorbance of the resultant solution was measured at 320 nm. Immobilized DAAO column assay A 1.5 X 30 cm jacketed column was filled with 30 ml of catalyst from the polyaziridine immobilization procedure, method 3. A solution of 1 mM D-Phe containing 1 mM sodium phosphate, pH 8.3, and 100 pJ 1- 1 of catalase was passed over the column at 20.5 ml h -1. The substrate solution was kept at 4DC to prevent the breakdown of the catalase and the D-Phe before passage over the column. The activity was monitored spectrophotometrically by absorbance at 320 nm of a 1: 10 dilution of the column effluent in 1 N NaOH. The column was maintained at 20 DC by a circulating water bath. Samples were obtained over a 30 min period daily while simultaneously determining the flow rate. Immobilization of lipase MY in aqueous phase Lipase MY, 0.15 g, dehydrated Amberlite DP-1, 0.5 g, and 0.5 ml of 0.01 M phosphate buffer, pH 6.5, were mixed thoroughly with 100 mg of HTA at 4 DC for 20 min. The mixture was allowed to stand for 4 h at room temperature after which the immobilized enzyme was collected and washed with deionized water. Lipase assay The immobilized enzyme was shaken with 10 ml of 0.5 M phosphate buffer, pH 6.5, and 5 ml of toluene containing 1 g of methyl 2-(4-hydroxyphenoxy)propionate (HPPA-Me) for 4 h. The remaining ester and the acid produced by hydrolysis of HPPA-Me in aqueous solution was determined using a Whatman C18 RAC column. The mobile phase used was 25% methanol, 0.5 g 1- 1 K 2 HP04 , pH 3.0, at a flowrate of 1.0 ml min -1 and UV detection at 220 nm. All dilutions were made into deionized water. Under these conditions the HPPA-Me eluted at 11.0 min, and HPPA eluted at 5.0 min. Immobilization of lipase MY in toluene Lipase MY, 0.15 g, dehydrated Amberlite DP-1, 0.5 g, and HTA, 50 mg, were mixed in 10 ml of toluene. The mixture was shaken (100 rpm) at room temperature for 4 h and the toluene was then allowed to evaporate in a fume hood at room temperature. The immobilized enzyme was collected, washed, and assayed as above. This preparation was also repeated using TA instead of HTA. Immobilization of lipase MY in trichloroethane Lipase MY, 0.15 g, dehydrated Amberlite DP-1, 0.5 g, and HTA, 75 mg, were mixed in 10 ml of 1,1,1-trichloroethane. The mixture was shaken at room temperature for 4 h and the solvent was evaporated, the immobilized enzyme collected, washed, and assayed as above. Immobilization of tyrosinase Mushroom tyrosinase, 5 mg, was dissolved in 0.6 ml of water and mixed well with 500 mg of HTA. This was homogeneously dispersed over dehydrated Amberlite
309
DP-1 and allowed to cure for 16 h at 4°C. After extensive washing with PBS, the immobilized enzyme was mixed with 1.5 mM tyrosine and 15 mM ascorbic acid in 25 ml of 0.1 M potassium phosphate buffer, pH 7.0. The assay mixture was incubated at 25°C with continuous aeration. L-DOPA formation was monitored by HPLC using a C18 column with a mobile phase of 0.1 % trifluoroacetic acid, 10% acetonitrile and 90% water. Under these conditions, tyrosine elutes at 6.13 min and L-DOPA elutes at 5.58 min. The control used soluble enzyme. Immobilization of horseradish peroxidase in dioxane Horseradish peroxidase, 12.5 mg (1000 units), and dehydrated Amberlite DP-1, 1.0 g, were lyophilized in 10 ml of 0.01 M phosphate buffer, pH 6.0. The lyophilized mixture was shaken with 50 mg of TA in 10 ml of dioxane at room temperature for 4 h and the dioxane was then allowed to evaporate. The immobilized enzyme, approximately 2 g, was collected and washed with phosphate buffered saline. The immobilized enzyme, 0.3 g, was assayed by following the conversion of pyrogallol, 25 mg, to purpurogallin (f: = 2.47 cm2 mmol- 1 ) spectrophotometrically at 420 nm in the presence of 50 }L1 of 30% hydrogen peroxide in 10 ml of phosphate buffer, pH 6.5. One unit will form 1.0 mg of purpurogallin in 20 s at pH 6.0, at 20°e. Immobilization of horseradish peroxidase in buffer HTA, 0.2 g, was mixed with horseradish peroxidase, 2.4 mg, 192 units, in 100 }LI of either (2-[N-morpholino]ethanesulfonic acid buffer, pH 5.5, phosphate buffer, pH 7.2, or N-tris-(hydroxymethyl)methyl-3-aminopropanesulfonic acid buffer, pH 9.0, for 5 min and then mixed with an additional 800 }LI of the same buffer and immediately mixed with 0.5 g (0.75 ml) of cellulose beads. The beads were allowed to cure at 4°C for 18 h, washed with 50 volumes of PBS and assayed by the phenolaminoantipyrine method. A solution of 0.17 M phenol and 0.0025 M 4-aminoantipyrine, 1.4 ml, was added to 1.5 ml of 0.0017 M hydrogen peroxide in 0.1 M phosphate buffer, pH 7.0, and 8 mg of the buffer immobilized horseradish peroxidase. The solution was stirred for 1 min and the absorbance at 510 nm was determined. Alkaline phosphatase immobilization Alkaline phosphatase, 0.9 mg, was lyophilized with 1 g of dehydrated Amberlite DP-1 in 10 ml of PBS. The mixture was shaken with 50 mg of HTA in 10 ml of toluene at room temperature for 4 h before toluene was evaporated. Approximately 2 g of immobilized enzyme was collected. The immobilized enzyme, 0.5 g, was assayed spectrophotometrically at 405 nm with 25 mg of p-nitrophenyl phosphate in 10 ml of phosphate buffer, pH 8.2, 0.5 M phosphate, 0.2 M ethanolamine). One unit will hydrolyze 1.0 }Lmol of p-nitrophenyl phosphate per min at pH 10.4 at 37°C to p-nitrophenol (f: = 13,200 cm2 mmol- 1 ). Immobilization of f3-galactosidase in toluene Galactosidase, 0.5 mg, was lyophilized with 1 g of Amberlite DP-1 in 5 ml of PBS. The enzyme was immobilized with 50 mg of HTA in 10 ml of toluene. Approximately 2 g of immobilized enzyme was collected after work-up.
310
The immobilized enzyme was assayed spectrophotometrically at 405 nm with 25 mg of p-nitrophenyl ,B-D-galactopyranoside in 10 ml of 0.5 M phosphate buffer, pH 7.0. One unit will hydrolyze 1.0 !Lmol of o-nitrophenyl-,B-D-galactoside per min to give o-nitrophenol ({ = 3100 cm2 mol-I).
Results and Discussion
Watanabe and Royer (1983) used aziridines as crosslinking agents to form polyethyleneimine bonded coatings on various matrices for chromatography or as enzyme supports (Watanabe and Royer, 1983); however, they did not use aziridines for the purpose of immobilizing proteins to an insoluble matrix. Porath and Axen had suggested earlier that aziridine groups might be used for coupling enzymes to a matrix (Porath and Axen, 1976); however, they presented no data for aziridines and were unsuccessful in their attempts to use three membered thio analogs. Thus, the value of the use of aziridines for the immobilization of enzymes lay unanswered, although the materials necessary for its investigation had been available for many years. The aziridine prepolymer system, described here, is readily prepared by the Michael type addition of an activated bond to aziridines as shown in Fig. 1 (Heissberger 1964; Mark et aI., 1964). The monomer can react with itself to form low molecular weight prepolymers. The commercially available prepolymers in the polyaziridine system vary in degree of solubility in organic solvents and range in size from less than 500 mol. wt. to considerably greater (Sello and Stevens, 1980). The polyaziridines, as purchased, may be stored for long periods of time with minimal precautions and are quite inexpensive. The order of reaction of the aziridine moiety with other functional groups is: COOH » SH > OH > NH 2 • The reactions of the aziridine and these functional groups is given in Fig. 2. The outstanding value of the polyaziridine system is the mild conditions under which aziridines will react and the simple method by which the prepolymer can be used to immobilize enzymes. Hazardous chemicals are not used, nor is a fume hood required unless the immobilization is to be carried out in the presence of an organic solvent. The availability of a wide range of pH values to immobilize is often an important factor in optimization of the retention of activity and stability of an enzyme. Polyaziridines may be crosslinked at temperatures below 40°C in the presence of relatively large volumes of water while still maintaining the activity of
o
RCH2Cf"2oi',";CH~3
• 3 [ONH ]
RCH2C~H20CCH2CH2-N13
•
R = CH 3 . TA = OH. HTA
Fig. 1. Preparation of aziridine prepolymers.
311
R ,-N
1
--J
o
0 II
11
+R Z-C-O-H ----- R ,-N-CH Z-CH Z-O-C-R Z I
H
R
, -N1+RZ-O-H~R --J
Z 1-N-CHZ-CHZ-O-R , H Fig. 2. Reactions of polyaziridine with other functional groups.
the enzyme. The pH of the polyaziridine immobilization reaction may range from 2 to 10 but polymerization is more rapid as the pH is lowered or at high polyaziridine concentrations (Table 1). At concentrations of 1 part HTA in 10 parts water or at below pH 8.0, a gel network was not obtained. The enzyme can be covalently attached to the polymer network while the network will also crosslink with functional groups on other surfaces, thus allowing the covalent attachment of enzymes to a wide variety of supports. The networks formed contain the immobilized enzymes which still have useful levels of their original activities. In general, the resultant gel composites have sufficient openness and hydrophilicity to allow free migration of chemical substrates for interaction with the immobilized species, and thus permit high levels of transformations to take place. Low molecular weight substrates can freely diffuse to the enzyme; however, for enzymes with a high turnover number, such as horseradish peroxidase, high flow rates are needed to overcome the diffusional limitation. The immobilized enzyme can also be taken to dryness by evaporation of solvent or lyophilization, while retaining most of its
TABLE 1 EFFECT OF HTA CONCENTRATION AND pH ON HTA POLYMERIZATION HTA (g)
H 2O (g)
pH
Gel time
Comments
10
10
10
45 s
2
10
9.5
26 min
2
10
8.5
13 min
2
10
8.0
6 min
2
10
7.0
2 min
1
10
9.5
4h
Exotherm 55 ° C Rigid gel Exotherm 30 D C Rigid gel pH adjust with 2 g KH 2 P04 Soft gel HCl pH adjust with rigid gel, exotherm to 40 D C HCl pH adjust with rigid gel, exotherm to 40°C Rigid gel
312 TABLE 2 EFFECT OF VARIATION IN HTA, PROTEIN AND SUPPORT CONCENTRATION ON THE IMMOBILIZATION AND RETENTION OF ACTIVITY OF DAAO BY HTA Method (mg)
Protein (mg)
HTA (mg)
Support Cellulose
Quantity (mg)
1 2 3 4
10 10 10 30
2000 100 100 2500
sponge sponge bead bead
1500 250 2000 15000
Retention of activity (%)
Ratio HTA/matrix
Protein/ HTA
35 43 48 53
1.33 0.4 0.05 0.116
0.005 0.1 0.1 0.012
original activity. The drying process has the advantage of providing strong, well crosslinked, insoluble compositions in the form of coatings as described here or membranes (data not shown) having high concentrations of active, immobilized enzyme. The preparation of a homogeneous solution of the enzyme and the prepolymer is easily obtained by stirring a mixture of the two in a small container. The solution may then be poured onto a suitable glass or teflon tray to form membranes or onto a suitable support, such as an ion exchange resin, silica, glass beads or other rigid or semi-rigid matrix. After addition of the solution to a support, additional stirring to obtain a homogeneous dispersion is carried out. The mixture or membrane may then be set aside at either 4°C or room temperature and allowed to cure overnight. The immobilization of the enzyme is complete and appropriate washing steps may then be carried out. The presence of the aziridine moiety and its nitrogen atom complicates the normal analyses of protein by colorimetric or absorbance readings and thus, the determination of the quantity of protein bound cannot be easily made. TABLE 3 RETENTION OF ACTIVITY AFTER ENZYME IMMOBILIZATION BY HTA AND TA Enzyme
Carrier
Solvent
Retention of activity
DAAO DAAO Lipase MY Lipase MY Tyrosinase
DP-1 Cellulose DP-1 DP-1 DP-1 DP-1 DP-1 DP-1 Cellulose Cellulose Cellulose DP-1 DP-l
Buffer Buffer Buffer Toluene Buffer Toluene Trichloroethane Dioxane pH 5.5 buffer pH 7.2 buffer pH 9.0 buffer Toluene Toluene
48
Polyaziridine
(%)
Candida rugosa lipase Lipase MY Horseradish peroxidase Horseradish peroxidase Horseradish peroxidase Horseradish peroxidase Alkaline phosphatase ,B-Galactosidase
48 65 68 20 70 80 2 18 1 1 2 10
HTA HTA HTA TAor HTA HTA HTA HTA TA HTA HTA HTA HTA HTA
313
As shown in Table 2, preliminary experiments using vastly different ratios of HTA to matrix or to protein were effective in the immobilization of DAAO and resulted in high retained activity. The stability of the DAAO on cellulose beads was examined using 10 ml D,L-Phe pH 8.0, at 20 0 e and 30°C. The stability of the enzyme was excellent at these temperatures for a 24 h period. At 20 o e, 96% of the original activity was present, while at 30 o e, 92% of the activity remained. The immobilization of DAAO in aqueous systems gave high yields of retained activity (Table 3). Excellent yields of retained activity were obtained when Lipase MY was immobilized in aqueous solution. When Lipase MY was immobilized in toluene, the yield of retained activity was equal to that obtained in buffer. The use of trichloroethane to immobilize Lipase MY gave even higher retained activity (80%) than immobilization in buffer (Table 3). The use of organic solvents allowed the preparation of a more homogeneous dispersion of the polymer/enzyme mix. In buffer systems, large aggregates were observed with this enzyme which were undesirable in a fixed-bed reactor, especially in large scale preparations. The retained activity of Lipase MY did not differ greatly whether the immobilization took place in buffer or organic solvents, including both toluene and trichloroethane (Table 3). Thus, the prepolymer may be used either in water or inert organic solvents without adversely affecting the immobilization of an individual enzyme. It should be noted that TA has a greater solubility in organic solvents than HTA; conversely TA is not very soluble in aqueous systems. The immobilization of ,B-galactosidase in an organic solvent provided a catalyst which retained 10% of its original activity (Table 3), however, immobilized horseradish peroxidase retained only 2% of its activity and alkaline phosphatase retained only 2% of its initial activity. Further investigation of horseradish peroxidase showed that it did not immobilize well at an initial pH of 7.2 or 9.0 in aqueous systems; however, at an initial pH of 5.5, the retention of activity was 18%. The complete optimization of horseradish peroxidase was not attempted, but it can be seen that the proper selection of the pH of the enzyme solution before immobilization can be quite valuable for maximum activity retention in this system. The polyaziridine system demonstrates wide applicability to the immobilization of enzymes under varying conditions. The details of these immobilizations, their stability and their commercial utility will be reported in due course.
References Akiyama, A., Bednarski, M., Kim, N-J., Simon, E.S., Waldmann, H. and Whitesides, G.M. (1988) Enzymes in organic synthesis. Chemtech 18, 627-634. Antrim, R.L., Colilla, W. and Schnyder, BJ. (1979) Glucose isomerase production of high-fructose syrups. App\. Biochem. Bioeng. 2, 97-155. Calton, GJ. (1988) Towards the Golden Age of Enzymatic Chemistry. Biofutur. Fevrier, 41-46. Calton, G.J., Wood, L.L. and Campbell, M.L. (1987) Phenylalanine production via polyazetidine immobilized E. coli: optimization of cell loading. Methods Enzymo\. 136, 497-503.
314 Heissberger, A. (1964) Heterocyclic Compounds with Three-and Four-Membered Rings, vol. 1, J. Wiley and Sons, New York, pp. 542-550. Kieslich, K. (1976) Microbial Transformations of Non-Steroid Cyclic Compounds, John Wiley and Sons, New York. Klibanov, A.M. and Kirchner, G. (1986) Enzymatic production of optical isomers of 2-halopropionic acids. US Patent 4,601,987. Mark, H.P., Gaylord, N.G. and Bikales, N.M. Eds. (1964) Encyclopedia of Polymer Science and Technology, vol. 1, J. Wiley & Sons, New York, pp. 735-736. Mauz, 0., Noetzel, S. and Sauber K. (1984) New synthetic carriers for enzymes. Ann. N.Y. Acad. Sci. 434, 251-253. Mosbach K., Ed. (1987) Immobilized enzymes and cells. Methods Enzymol 135-138. Noetzel, S., Mauz, O. and Sauber, K. (1984) Macroporous polymer beads and their use for enzyme immobilization. Patent German Offen. DE 3404021. Porath, J. and Axen, R. (1976) Immobilization of enzymes to agar, agarose and sephadex supports. Methods Enzymol. 44, 19-45. Sello, S.B. and Stevens, c.Y. (1980) Aziridine-terminated polymers as wool shrink-proofing agents. Quinquenn. Int. Wool Text. Res. Conf., 6th Fiche 6/G/9 [CA94:85-565b]. Takamatsu, S., Yamamoto, K., Tosa, T. and Chibata, I. (1981) Stabilization of L-aspartic .a-decarboxylase activity of Pseudomonas dacunhae immobilized with carrageenan. J. Ferment. Technol. 59, 489-493. Takata, I., Yamamoto, K., Tosa, T. and Chibata, I. (1980) Immobilization of Brevibacteriumflavum with carrageenan and its application for continuous production of L-malic. Enzyme Microb. Technol. 2, 30-36. Tosa, T., Sato, T., Mori, T., Matsuo, Y., and Chibata, I. (1973) Continuous production of L-aspartic acid by immobilized aspartase. Biotech. Bioeng. 15, 69-84. Walts, A.E., Fox, E.M. and Jackson, C.B. (1987) (R)-Glycidyl butyrate: evolution of a laboratory procedure to an industrial process. Biotech USA 1987, Online International Inc., London, pp. 91-98. Watanabe, K. and Royer, G.P. (1983) Polyethylenimine/silica gel as an enzyme support. J. Mol. Catal. 22, 145-152. Watanabe, K., Chow, W.S. and Royer, G.P. (1982) Column chromatography on polyethylenimine-silica: rapid resolution of nucleotides and proteins with short columns and low pressures. Anal. Biochem. 127, 155-158. Wood, L.L. and Calton, GJ. (1984) A Novel Method of Immobilization and Its Use in Aspartic Acid Production. Bio/Techn. 2, 1081-1084. Wood, L.L. and Calton, G.J. (1986) The production of L-phenylalanine by polyazetidine immobilized microbes. Bio/Techn. 4, 317-320.
