Continuous Conversion of Starch to Glucose with Immobilized Glucoamylase*

Transcription

1 BIOTECHNOLOGY AND BIOENGINEERING VOL. XIII, PAGES (1971) Purchased by U. S. Dept. of Agriculture for Officiai Use Continuous Conversion of Starch to Glucose with Immobilized Glucoamylase* K L. SMILEY, Northern Regional Research Laboratory, t Peoria, Illinois Summary Partially purified glucoamylase from Aspergillus awamori NRRL 3112 was immobilized on diethylaminoethyl cellulose in the presence of low ionic-strength acetate buffers at ph 4.2. The active enzyme-cellulose complex was used to convert starch substrates continuously to glucose in stirred reactors. Substrate concentrations as high as 30% could be quantitatively converted to glucose at a rate of more than 25 mg/min/liter at 55 C for periods of 3 to 4 weeks in a 4-liter reactor. Shutdowns were due to mechanical problems and not to loss of enzymes, which could be recovered with no appreciable loss of specific activity. Transfer products, such as isomaltose and panose, were present in immobilized enzymeproduced syrups but to no greater degree than in soluble glucoamylase digests of starch. INTRODUCTION Large-scale industrial use of enzymes is hampered by high cost and instability. Of the several thousand enzymes knowl1, only a few hydrolases have large-scale industrial uses. Recent developments in enzyme technology have shown that enzymes can be immobilized on an insoluble carrier and still retain significant activity.! In addition, immobilized enzymes are generally more stable than their counterparts in solution. Because immobilized enzymes can be recovered, they have the potential of circumventing both the highcost and unstable aspects of conventional enzymes. *Presented at the Meeting of the American Chemical Society, Chicago, Ill., September 13-18, tthis is a laboratory of the Northern Marketing and Nutrition Research Division, Agricultural Research Service, U.S. Department of Agriculture by John Wiley & Sons, Inc. 309

2 310 SMILEY A simple method of binding an enzyme to a carrier is ionic bonding. The enzyme is bound to a charged carrier such as diethylaminoethyl (DEAE) cellulose or other ion exchange polymer. As long as certain physical conditions are maintained, such as ph and ionic strength, the enzyme will remain tightly bound indefinitely even though large volumes of substrate solution are passed over the immobilized enzyme. Tosa et al. 2 have demonstrated that amino acid acylase can be successfully bound to either DEAE cellulose or Sephadex,* and racemic mixtures of acylaminoacids can be resolved on columns of the immobilized enzyme. Bachler et al. 3 showed that glucoamylase can be tightly bound to DEAE cellulose. They continuously converted dilute starch solutions to glucose using columns of immobilized glucoamylase. Columns have some disadvantages. To get sufficient residence time for the substrate they must be quite long and as a result the cellulose-enzyme complex packs very tightly, restricting flow rate. Starch solutions, unless filtered after liquefaction, tend to plug the columns after a short operation time. This paper indicates that DEAE cellulose-bound glucoamylase can convert high concentrations of starch substrate nearly quantitatively to glucose in stirred reactors. METHODS Preparation of DEAE Cellulose-Glucoamylase Complex Five grams of Diazyme, a commercial glucoamylase preparation, was slurried in about 40 ml of 0.02 M Na acetate buffer at ph 4.2. The slurry,vas stirred 10 min and filtered to remove the carrier. After the carrier was washed with the same buffer, the filtrate and washings were combined and made to 100 ml. This solution was then added to 10 g of DEAE cellulose that had previously been prepared by cycling between 0.5 M HCl and 0..5 M NaOH and finally equilibrated with 0.02 ivi Na acetate buffer, ph 4.2. The cellulose Diazyme mixture \vas stirred at room temperature for 30 min and then filtered. The insoluble complex was reslurried in 500 ml of the 0.02 AI acetate buffer and filtered again. This process was repeated *The mention of firm names or trade' products is for identification only and does not imply endorsement by the U.S. Department of Agriculture. BIOTECHNOLOGY AND BIOENGINEERING, VOL. XIII, ISSUE 2

3 CONTINUOUS GLUCOSE CONVERSION TO STARCH 311 until the filtrate was free of protein as judged by lack of light absorption at 280 mj.l. The washed cellulose-diazyme complex was finally suspended in 100 ml of 0.02 ivi buffer and stored at 4 C. The DEAE cellulose was obtained from Schleicher-Schuell Co. Preparation of Substrate The starch substrate was prepared by slurrying 1200 g of foodgrade starch in 2800 ml of water. One gram of Rhozyme H-39, a starch-liquefying enzyme, was added and the temperature raised to about 75 C. Vigorous agitation was employed until the mixture thinned (about 15 min at 75 G). The solution was maintained at this temperature for another hour and then raised to 85 C and held there for 1 hr. The liquefied starch solution was used,-vithout further treatment even though a precipitate formed on standing. IVIaltodextrins of about 15 dextrose equivalents (DE) were prepared by simply dissolving them in sufficient water to give a 30% (wiv) concentration. Components of the Stirred Reaction Vessel The stirred reactor consisted of an open-topped 8 in. X 12 in. Pyrex jar which holds about 8 liter. A stainless-steel lid was fabricated which had a sleeve bearing to accommodate an agitator and five I-in. holes spaced around the bearing 2! in. from center. The lid was fitted with an agitator having a 1 X 4 in. paddle, a thermoregulator, thermometer, constant level substrate reservoir, and a porous stone filter, through which the finished sugar solution was "ithdrawn. A peristaltic pump was used to,-vithdraw the sugar solution and feed it to a I-in. column containing 10 g of Darco G-50 charcoal. A Bron-Will thermoregulator was used to control the temperature. The temperature was maintained by means of a 4-ft. heating tape wrapped around the Pyrex jar. The tape was plugged into a variable voltage transformer and adjusted so that the current remained on about 75% of the time at normal laboratory room temperature. Figure 1 diagrammatically illustrates the reaction vessel. Cylindrical 2 in. X 2 in. "Aloxite" porous stone air spargers served as filters to "ithdraw the sugar solution while retaining the immobilized glucoamylase in the reaction vessel. Soxhlet thimbles also

4 312 SMILEY rconstant level Substrate Reservoir rperistaltic Pump ~'-V~~ \ta;nl~s~ St~el lid ~ _Charcoal Column " DecoloTIZer Pyrex Jar S1oo',,,t,, tl'"''''.,,,,,,, ~~:' et:::j.t-agitalor Paddle Fig. 1. Diagrammatic sketch of stirred reactor: Not shown is a thermoregulator and a thermometer fitted through the lid. proved successful as "ithdrawal filters. Darco G-60 was put in the discharge solution. The column containing 10 g line to decolorize the sugar The reactor was charged \\ith 4 liters of starch or dextrin substrate. The desired quantity of immobilized glucoamylase was added and the reactor allowed to run batchwise for about 48 hr or until the contents were hydrolyzed sufficiently to give a negative starch-iodine reaction. For 30% corn starch solutions the starch-i 2 color test must be run on a filtrate since an insoluble starch fragment is starchiodine positive. The substrate reservoir was then put in place and the pump set to deliver 50 to 60 nil/hr. The flow rate was set by the physical limitations of the filter employed. No attempt was made in this work to establish the optimum conditions for the stirred reactor. However, engineering aspects of the reactor are now being studied. The temperature was maintained at 55 C. Glucose Analysis Glucose was analyzed on a Technicon Autoanalyzer according to the alkaline ferricyanide procedure. 4 Glucoamylase Assays Glucoamylase was assayed by a modification of the procedure described in Miles Laboratories Technical Bulletin Five milliliters of 4% starch substrate was added to a ;jo-ml Erlenmeyer flask. The flask was equilibrated to 60 C on a shaking water bath. BIOTECHNOLOGY AND BIOENGINEERING, VOL. XIII, ISSUE 2

5 CONTINUOUS GLUCOSE CONVERSION TO STARCH 313 A suitable amount of DEAE-bound glucoamylase was added to the flask. The flask was stoppered tightly and incubated 1 hi'. The reaction was stopped by adding t,vo drops of 10 N NaOH and diluting to the proper volume to give between 0.5 and 2.0 mg/ml of reducing sugar. The shaking bath is necessary to keep the immobilized enzyme in suspension during the reaction. Considerable care must be exercised to get a representative sample of the celluloseenzyme slurry since it settles out rapidly. One glucoamylase unit is the amount of enzyme required to form 1 g of glucose under the conditions stated. Paper Chromatography Samples from the reactors were chromatographed on vvhatman No. 1 paper with butanol: pyridine: water 6:4:3 as irrigant. To separate components adequately the multiple development technique was followed. 6 Spots were visualized by spraying the chromatogram with AgN0 3 in acetone followed by spraying,vith ethanolic NaOH. RESULTS Run 1 was started with 5% 16 DE dextrin as substrate. Approximately 43 immobilized glucoamylase units were added to 4 liters of the dextrin solution and the mixture was incubated for 30 hi' "ith slow stirring. At this time the starch-i 2 reaction was negative and feeding of substrate "ith concomitant withdrawal of product was started. The reactor ran for 7 days under these conditions and the concentration of reducing sugar in the product remained constant at about 4.8% as glucose. Although the flow rate varied over this time from about 60 to 100 ml/hr, the production of glucose remained constant (Table I). It was therefore decided to try a more concentrated substrate feed. Accordingly, 30% 16 DE dextrin was substituted for the 5% solution. It took about 210 hi' for the reactor to equilibrate to the higher dextrin concentration. The flow rate tended to decrease somewhat due to the higher viscosity of the sugar solution. Glucose production increased to about 27.0% but was still independent of flow rate over a range of 36 to 75 ml/hr (Table I). Soluble glucoamylase produced 4.8% reducing sugar as glucose from the 5% 16 DE dextrin. This amount is 94% of theory after

6 314 SMILEY TABLE I Glucose Production from Corn Dextrin Substrate in a Stirred Reactor Using Glucoamylase Immobilized on Diethylaminoethyl (DEAE) Cellulose G % Glucose Age, % Liters % G glucose/ % glucose hr Interval Dextrin collected Glucose glucose min in control' _b % Glucose produced by soluble glucoamylase in 5% dextrin; 28.3% glucose produced by soluble glucoamylase in 30% dextrin. b Time at which reactor reached equilibrium after 30% dextrin feed started at 160 hr. correcting for a moisture content of 7.7% in the dextrin. The 30% dextrin yielded 28.3% reducing sugar with soluble glucoamylase, or 92% of theory. At the lower dextrin level the reactor equaled the soluble enzyme in gluose production. However, at the 30% level the reactor averaged about 95% of the yield obtained by soluble glucoamylase. Run 2 was made,vith 30% starch solution thinned with a-amylase. The reactor was charged with 4 liters of substrate and 35 immobilized glucoamylase units. The reactor was allowed to run batch,vise for 3 days. A positive starch-iodine reaction persisted. However, a filtrate of the material in the reactor was starch-iznegative so feeding and withdrawal were started. A..ithdrawal rate of about 60 mljhr was established as the maximum since higher pump speeds caused cavitation in the..ithdrawalline with no increase in flow. Table II shows that the reactor made approximately the same amount of BIOTECHNOLOGY AND BIOENGINEERING, VOL. XIII, ISSUE 2

7 CONTINUOUS GLUCOSE CONVERSION TO STARCH 315 reducing sugar as was obtained by soluble enzyme for the first 10 days, \vhich corresponded to 13 days running time. After this time the percentage of reducing sugar slowly declined until it was only about 80% as high as soluble enzyme control. However, starch per se did not come through since the starch-i 2 reaction remained negative throughout 28 days of operation. The decline in efficiency is probably due to accumulation of an insoluble starch fraction that was not digested by glucoamylase but coated the cellulose-glucoamylase complex and interfered \\ith the contact between the enzyme and starch. Paper chromatographs of the sugar samples also indicated a decline in efficiency \vith time in run 2. Since maltose was evident in the product during the third and fourth weeks of operation, its presence indicated the rate of glucose formation was low. During the first 2 weeks no maltose was detected though isomaltose was found throughout the entire run. Small amounts of panose appeared in samples taken the fourth week. In run 1 where maltodextrin was used, no significant sugar other than glusose appeared with the exception of trace amounts of isomaltose which was evident from the start. Since the original dextrin solution did not have detectable isomaltose present, it must have been formed as a result of the action of the immobilized glucoamylase. After about 3 weeks of operation unhydrolyzed dextrin appeared in the product as evidenced by a positive starch-i 2 reaction. TABLE II Glucose Production from 30% Starch Substrate in a Stirred Reactor Using Glucoamylase Immobilized on DEAE Cellulose G % Glucose Time, Liters % G glucose/ % glucose hr Interval collected Glucose glucose min in control H2O Soluble glucoamylase control gave 26.7% glucose.

8 316 SMILEY In both runs when the reactor started to fail, accumulation of an insoluble fraction of the substrate was noticed which may have been interfering with the enzymatic action. The enzyme itself did not appear to be irreversibly inactivated since it could be eluted in active form from the DEAE cellulose-diazyme complex by molar sodium acetate. DISCUSSION It may be practical to convert starch substrates continuously to glucose by glucoamylase immobilized on DEAE cellulose. The decreased rate of reducing sugar production after 3 to 4 weeks appears to be due to mechanical difficulties involving insoluble starch fractions. This trouble can perhaps be corrected by filtration of the liquefied starch. Wilson and Lilly7 insolubilized glucoamylase by covalently linking the enzyme to an s-triazinyl derivative of cellulose. They showed that a column of this enzyme could increase the glucose content of a dextrin syrup. The enzyme was fully active after 28 days of operation. It is difficult to compare the performance of the s-triazinyj cellulose-glucoamylase with the DEAE cellulose-glucoamylase since they presented no data showing the apparent activity of the triazinyl derivative in enzyme units. They did show that for a 100-hr period the s-triazinyl derivative produced glucose at the rate of 0.63 g/min. In the current work a rate of 0.26 g/min was maintained for 6 days with 35 l\iiiles units of glucoamylase. However, flow rate \vas limited by the filter and not by the rate of conversion of starch to glucose. Glucose concentration in the product was independent of feed rate in the range studied, and no substrate was found in the product. This absence indicates that the dilution rate could be increased beyond the range studied and thereby increase productivity. Mechanical limitations of the present equipment prevented investigations at higher dilution rates. Some concern has been voiced that enzymes bound by charge would gradually leach off the support. No evidence of leaching was found when samples of the glucose-containing product were dialyzed and the contents of the dialysis bag checked for glucoamylase activity. Tosa et au have shown that amino acid acylase is tightly bound to either DEAE cellulose or DEAE Sephadex. Columns of DEAE BIOTECHNOLOGY AND BIONEGINEERING, VOL. XIII, ISSUE 2

9 CONTINUOUS GLUCOSE CONVERSION TO STARCH 317 Sephadex-amino acylase were operated continuously for 32 days and still had more than 60% of the original activity. There was no evidence that the loss in activity was due to wash out of the enzyme. One advantage charge-bound enzymes have over covalently bound enzymes is ease of preparation. Presumably the cost of making immobilized enzymes of comparable activity would be much less with ion exchangers than by chemical bonding. If their active life is comparable to covalently bound enzymes, they should be carefully considered as potential sources of useful immobilized enzymes. The technical assistance of Mr. D. E. Hensley is greatly appreciated. References 1. L. Goldstein, Fermentation Advances, Academic Press, Inc., New York, 1969, p T. Tosa, T. Mori, N. Fuse, and 1. Chibata, Agr. Bioi. Chem., 33,1047 (1969). 3. M. J. Bachler, G. W. Strandberg, and K. L. Smiley, Biotechnol. Bioeng., 12, 85 (1970). 4. Technicon Instruments Corp., Technicon Autoanalyzer, Methodology N-2A, Chauncey, N.Y. (1963). 5. Miles Laboratories. Tech. Bull , Elkhart, Ind. (1962). 6. A. Jeanes, C. S. Wise, and R. J. Dimler, Anal. Chem., 23, 415 (1951). 7. R. J. H. Wilson and M. D. Lilly, Biotechnol. Bioeng., 11,349 (1969). Received February 8, 1971