Indi an Journal of Biochemi stry & Biophysics Vol. 39, August 2002, pp. 253-258

Depolymerization of starch and pectin using superporous matrix supported enzymes

Arvind Lali*, Kushal Manudhane, Nuzhat Motlekar and Priti Karandikar

Chem ica l Eng ineering Di vision, UICT, Mumbai 4000 19, Indi a

Received 23 Jlln e 2002; revised alld accepted 28 Jlllle 2002

Immobi li zed enzyme catalyzed biotransformat ions in volving macromolec ul ar substrates and/or products are greatl y re­tarded due to slow d iffusion of large substrate mo lecules in and o ut of the typical enzyme supports. Slow diffusion of mac­romo lecules into the matri x pores can be speeded up by usc of macroporous upports as enzyme carriers. Depolymerization reactions of polysaccharides like starch, pectin , and dextran to their respective low molecular weight products are some of the reactions that can benefit from use of such superporous matri ces. In the present work, an indigeno usly prepared ri gid cross-linked cellulose matrix (ca lled CELB EA DS) has been used as support fo r immobili zing alpha amylase (I,4-a-D­glucan glucanohydrolase, EC 3.2 . I . 1.) and pectinase (endo-PG: po ly( I ,4-a-galactouronide) glycanohydrolase, EC 3.2. 1.1 5). The immobilized enzymes were used for starch and pectin hydrolysis respecti ve ly, in batch, packed bed and expanded bed modes. The macroporos ity of CELBEADS was fo und to permit th rough-fl ow and easy diffusion of substrates pectin and starch to enzyme sites in the porous supports and gave reaction rates comparable to the rates obtained using soluble en ­zy mes.

Introduction Immobilized enzymes offer several advantages

over soluble enzymes, including easy recovery and reuse of the enzyme, possibility of continuous and controlled operation, and many a times increased en­zyme stability. Despite a very large amount of work on immobilized enzymes, large-scale successful in­dustrial applications have been limited to a few in­stances. One of the major examples of large-scale en­zy me catalyzed biotransformations is depolymeriza­tion of macromolecular substrates like starch and pectin. Not surpri singly , these reactions have not found use of immobilized enzy mes as advantageous. Typical matri ces or supports used fo r enzyme immo­bilization are porous, but have a pore size that is small to allow easy access fo r macromolecular substrates to the immobili zed enzyme sites. These matrices, like porous sili ca, glass or polymers li ke po lystyrene, have a mean pore size of a few hundred angstroms. In such cases, while the enzyme is trapped inside the support , pore di ffusional res istance [0 macromolecular sub­strates limits their accessibility to catalytic sites. For example, it has been reported that immobi lized alpha amy lase gives low reac ti on rates!') . It is also known

*Au thor for correspo ndence Te l: 9122 4 1456 16; Fax : 91 22 4145614 E-mail : arvi ml@l;dcLerneLin

that immobilized pectinases do not give adeq uate re­action rates due to high molecular weight of the sub­strate4

•

There have been some attempts to overcome this problem and faci litate th e use of immobilized en­zymes for reactions involving macro-substrates. One way has been to covalently conjugate the enzyme al­pha-amylase on to reversibl y soluble polymers5

.6

. In such a case the enzy matic reaction is carried out with the polymer-enzyme complex in soluble state, and the complex is precipitated for separation and reuse after the reacti on. Another way to overcome the diffusional resistance in a porous matri x has been the use of sup­ports with very large pores for enzyme immobiliza­ti on 7.8. K vesitadze and Dvali I indicated that immobi­li zed amylase action was best when immobilized on a matrix with largest pore size (760 A. in their case). The work of Kminkova and Kucera4 showed thaI presence of fine suspended particles in fruit juices res tricted use of immobili zed enzymes and the per­fonnance was poor due to inabi li ty of pectin fO have access to the enzyme in small pores.

Use of macroporOlls matrices has been found to have important bearing on purification of protein s~. It is known that presence of large pores in a matri x can lead to through-flow in pores thereby reduci ng the di ffusiona l resistance for macromolecules like prn­tei ns 10. In the present work, we study macromolecular

254 INDI AN 1. BIOCHEM . BIOPHYS. , VOL. 39, AUGUST 2002

biotransformations using enzy mes immobili zed on an indigenously prepared rigid , superporous cross- linked cellulose matrix ca ll ed CELBEADS. Alpha amylase and pectinase were chosen for the study as they ac t on macro molecular substrates starch and pectin, respec­tively.

Materials and Methods

CELBEADS, a rigid cross- lin ked cellulose beaded matrix, was prepared indigenously under another project and was kind ly made avai lab le fo r thi s work. Native cross- lin ked agarose beaded matrix was a kind gift from UpFront Chromatography AIS, Denmark . Dinitrosalicylic acid (DNSA) fo r red ucing sugar as­say, and sod ium polypectate were obtained from Sigma Aldrich Co., USA. Soluble starch was obtai ned from Loba Chemicals, India. The enzymes alpha amy lase (Termamyl 120 L) and pectinase (O KL-IOO) preparations were generous gifts from ovo Nordi sk, De nmark and Biocon India Ltd, India, respectively. All the other chemicals were of ana lytical grade and were obtained from S. D. Fine Chemicals Ltd, India .

Determination of alpha amylase activity

The activity of alpha amy lase was measured in terms of rate of red ucing sugars produced using the dinitrosa licy lic acid (DNSA) method II; I % so luble starch in 0.2 M acetate buffer (P H 5.6) was used as substrate. For estimation of the enzy me activi ty , 0.5 ml of the substrate sol uti on was mixed with 1.5 ml of acetate buffer (0.2 M, pH 5.6) containing sui table enzyme aliquot. The solution was then heated at 55°C for to min . The reaction was stopped by additi on of 1 ml DNSA reagent. The resulting solution was heated in a boiling water bath for 10 min. The reduc­ing sugar produced was measured spectrophotometri­cal ly at 540 nm using Jasco V-530 UV-VIS spectro­photometer, using a standard glucose calibration curve. One enzyme unit of enzyme acti vity was de­fined as that requi red to liberate one micromole of glucose per min from starch under the assay condi­tions.

Determination of pectinase activity Pectinase activ ity , as endo-polygalacturonase ac­

tivity, was determined as polygalacturonase activity with sodium polypectate (sodi um salt of polygalactu­ronic acid) as substrate l2

. The amount of reducing sugars produced by action of enzyme on sodium polypectate was estimated by the standard DNSA

method II . Thus, I % sod ium polypectate in 0.1 M acetate buffer (pH 5) was used as substrate. After in­cubat ion at 37°C for 15 min, the amoun t of reducing sugars was determined spectrophotomet ri cally at 540 nm agains t a ca librati on chart prepared using glu ­cose as reducing sugar. One unit of enzyme activity was defined as that required to liberate one mi cromole of glucose per min under the assay cond itions.

Protein Estill/atioll

Protei n content of the so luble enzyme was deter­mined using Bradford's method l 3

. Standard assay procedure using BioRad protein assay kit was em­ployed with BSA as the calibration protein.

1II/II1Obifizalioll of ell zymes on CELBEADS

The soluble enzymes, alpha amy lase or pectinase, were immobili zed onto CELBEA DS by covalent at­tachment through epichlorohydrin acti vation of the hydroxy l groups on the CELBEADS surface 14 . Ethyl­enediamine was attached as a spacer arm to prevent possible steri c res istance to a macromolecul ar sub­strate's approach to the immobili zed enzy me si tes . 10 ml of suction dried CELBEADS was washed well with distilled water on a sintered glass funne l and agai n suction dried. The matrix was then added to a con ical flask containing a mixture of 34.5 ml 2 M NaOH, 0.1275 g sodium borohydride and 3.75 ml epi chlorhydrin . To thi s fl ask another 34.5 ml NaO H was added wi th 17 ml epichl ohydri n in small portions over a peri od of 2 hr. The mixture was shaken on an orbital shaker overnight at room temperature. The matrix was then fi ltered on a sintered glass funnel and washed thoroughly with 0.1 M acetic acid, 0.2 M so­di l1m bicarbonate, and fina lly with di stilled water. Thus, epoxy activated and suction dried CELBEADS was then added to a mixture of 22.5 ml of 0.2 M so­di um carbonate containing IS ml ethylenediamine, and shaken in the conical fl ask for 24 hr at 50°C. The beads were then fi ltered and washed with 0. 1 M acetic ac id, 0.2 M sodium carbonate and fina lly with dis­tilled water. The resultant thylenediamine­CELBEADS was again activated in 30 ml alkaline solution of 12.5% aqueous glutaraldehyde under overnight shaking conditi ons at room temperature. The ac tivated matrix was fi ltered and washed well wi th distilled water to remove residual glutaralde­hyde. Suitable dilutions of the industri al enzyme preparations (pectinase or amylase) were prepared in 0.1 M phosphate buffer, pH 7.5. For immobilization

LALI el af.: DEPOL YMERIZATION OF STARCH AND PECTIN 255

the prepared enzy me solution was added to the acti­vated CELBEADS and kept overnight at 4°C. The immobilized matrix was washed next day with 0.1 M phosphate buffer, pH 7.5 to remove the soluble en­zyme. The total amount of unadsorbed protein and enzyme was assayed. The amount of enzy me (units) and protein immobilized was estimated as the differ­ence between the total loaded and unbound enzyme and protein . The immobilized enzy me was stored 111

0. 1 M phosphate buffer at 4°C.

Biocoll version using free and immobilized enzymes

Bioconvers ion of starch and pectin were carri ed out in three modes: Batch, packed bed, and fluidized or expanded bed. All the batch depolymerization ex­periments with starch or pectin were carried out in stoppered conical flasks at constant temperature (30°C for pectin hydrolys is, and 55°C for starch hy­drolysis) on an orbital shaker- incubator. The initial starch concentration used was 3% w/v, and pectin concentration was 0.5% w/v, all th rough the experi­ments. Samples were drawn at suitable time intervals and the reaction mixture assayed for concentration of reducing sugars. Runs were made with both soluble and immobilized enzymes in equival ent enzy me units.

Since the feed solutions of starch and pectin are turbid, it can be imagined that these wi ll tend to clog typical packed beds with conven tional support matr i­ces. However, it has been found that CELBEADS all ows th rough -flow of not only macromolecul es but also of small particulate matter li ke cell debris (A mritkar et al. , 2002 unpu bli shed work ). Thus packed bed flow experiments were conducted with plain CELBEADS (without immobilized enzyme) with both starch and pectin solutions to check if there was any retention of starch or pectin in the column. The experiments consisted of pass ing a fixed volume of 15% w/v starch or 0.5 % w/v pectin so lution up and out through the packed bed at a fixed fl ow rate, and then measuring the total amount of starch or pectin eluted from the column.

Column bioconversion experiments were carried out by packing the immobi li zed enzy me (amylase or pectinase) in a 20 cm lo ng, I cm interna l diameter jacketed glass column eq uipped at the two ends with two adjustable 14 cm BioRad Econo flow adapters provided with modified flow di stributors cons isti ng of a muslin cloth piece in place of the polymeric flow distributors . The lower adapter was connected to an Alitea peri staltic pump that pumped the feed solution

,

•• n Stirred starch or pectin Peri staltic sllspension in water bath pump

t-

Jacketed packed bed! fluidized bed reactor

Fig . I-Schematic diagram of the experi me ntal set-up for packed bed and n uidized bed depo lymerizatio n of starch or pectin

(initial concentration 3% w/v starch or 0.5% w/v pec­tin) from a stirred beaker up through the column, and recirculated back to the feed beaker. While pectin hy­drolysis was calTled out at room temperature (30°C), starch hydrolysis was carri ed out at 55°C by keeping the feed solution in a controlled temperature water bath and circulating the hot water through the col umn jacket. The schematic di agram of the assembly is shown in Fig. I . For packed bed experiments the flow adapters were inserted to touch the matri x bed from both ends. In case of fluidized bed experiments, the lOp act<.lpter was moved up to provide free board over the settled bed whi ch could be fluidized by the up­flowin g feed solution.

Results and Discussion

CELBEADS used in the work is an indigenously prepared rigid superporous cross-linked cellu lose based beaded matrix. The propert ies of CELBEADS. as determined by Manudhane l 5

, are given in Table I . Gel fi ltration studies 15 on CELB EADS using two marker mOlecu les, bovine serum albumin (MW 66,000) and blue dextran (MW 200,000), have shown that the pore structure in CELBEADS is nearly 1l10nodisperse while pore vol ume is in the range 55-57%. Scanning electron micro copy of the beads

256 INDIAN J. BIOCHEM. BIOPHYS., VOL. 39, AUGUST 2002

Table 1- Properties of CELI3 EADS (taken from Ref. 15)

Mean Bead size (range)

Sphericity

Nature

Average pore size

Total voluille poros ity

Bulk density (water)

pH stab ility

Temperature st:1bility

Surface hydroxy l density

250 pm ( 150-300 pm)

0.7-0.9

ri gid aeroge l

>5-7 pill

-57"1c

1800 kg/m.1

2- 14

Testcd safe up to 125°C

-9 mM

indicated a mean pore size of the beads to be in th e range 5-7 J..un IS

. It has been reported that with large pore size (i n excess of 0.5 J.,lm, for exam ple) there is a fini te through-flow of so lu tion in the bead pores when the beads are packed in a column and the flow ex­ceeds a certain th resho ld flow rate t6 This situation is shown in Fi g. 2. The flow through the pores of the beads can make immobilized enzy me sites on matri x surface more accessible to substrates, especially to high molecu lar weight substrates that have consider­able difficulty in diffusing into the pores of conven­tiona l enzyme supports. In additi on to being superpo­rous, a matrix should be fairly ri gid to withstand mul ­tipl e and prolonged usage. Most macroporous sup­ports used for protein adsorption are ge l matri ces based on agarose, cellulose or dextran. CELBEADS offers a matrix that is based on cellulose but is a ri gid matrix that can withstand harsher environment (e.g. treatment with I M HCI or I M NaOH) and is auto­clavable. It was, therefore, dec ided to use CEL­BEADS as enzyme support in systems where the sub­strates are macromolecular. CELBEADS has de­ri vatizable hydroxyl groups on the surface (surface OH density - 9 mM ts

) that can be used for a wide va­ri ety of app li cations from ligand lin kage to enzyme immobilization. Amylase and pectinase were immo­bilized on CELBEADS for the present work.

Enzyme immobilizatioll Termamy l 120 L had an es timated amylase activity

of 60000 EU/ ml , with a protein content of 39 mg/ ml. A I: 100 dilution of thi s preparation was used for im­mobilization onto activated CELBEADS in a 1:2 ratio of beads to solution. The immobi li zation procedure gave 36% enzy me binding amounting to 436 enzyme units (EU) of alpha amylase immobi lized per ml of CELBEADS. Protein binding was 46% (0.359 mg protein bound per ml of matrix). Since the dry

Fig. 2- Compari son o f diffusion of the solution in bead pores of the normal and macroporous matri x in packed bed mode

CELBEADS density is half of wet beads IS, the immo­bili zed amylase acti vity was about 900 EU amylase per gram of dry beads.

Pectinase OKL- IOO had an estimated activity of 1900 EUI ml , and a protein con tent of 13 mg/ ml. A 1:2 dilution of the enzyme was used for immobili za­tion onto CELBEADS in a 1:2 ratio of beads to en­zy me solution. The immobili zation procedure gave by enzyme balance 4 I 2 EU of pectinase immobilized per ml of CELBEADS (43 % immobili za tion), while nearly all of the protein was estimated as bound on the matri x.

Temperature alld pH optima

Temperature and p H optimum of both so luble and immobilized alpha amylase was found to be 55°C and 5.6, respectively. On the other hand the p H optimum for the immobilized pectinase was 2 compared to the pH optimum of 5 for soluble pectinase. The tempera­ture optimum was however unchanged at 50°C. All further experiments with immobilized enzymes were carried out at their respective optimum pH. The reac­tion temperature used was 55°C for starch hydro lysis, and room temperature (30°C) for pectin hydro lysis , since industrial use of pectinase is preferably done at low temperatures to prevent colour development in processed products.

Good activity of immobili zed pectinase at a pH as low as 2 is an important and serendipitous outcome of the work. Pectinase essentiall y is a group of enzymes. The shift in the pH optima of immobi lized pectinase may be attributed to the selective im mobi lizat ion of one or the other enzy me from the group on to CELBEADS and it may be possible that thi s enzyme is active at lower pH.

LA LI el al.: DEPOLYMERIZATION OF STARCH AND PECTIN 257

Effect of macroporosity of CELBEADS on substrate . flo w in packed bed

To study the th rough-fl ow of substrates in the CELBEADS, 15% so lu ble starch and 0.5 % pecti n solu tions were passed through a packed bed of CELBEADS. The elut ing fractions were co llected and analyzed fo r starch or pectin content. It was observed that the percent recovery of starch was around 95 %, while recovery was 65% using pect in. Corresponding recoveries with a commercial agarose gel based ad­sorbent (UF Agarose from UpFront Chromatography A/S, De nmark) were 70% and 34%, respectively. Thus, whil e CELB EA DS was proven to be less than 100% permeable to the two substrates, it d id perfo rm better than the typical agarose gel based adsorbent. Loss of both starch and pectin is due to phys ical en­trapment of the polymeri c substrates in the tortuous pores of the matrices used. CELBEADS is a rig id matrix with near mono-di sperse large pore size l5 and can be expected to trap less of the suspended or mac­romolecular matte r. The macroporous agarose matri x on the other hand is a continuous gel phase, and al­though there is evidence that cell fragents can pass through a gel matrix 17

, such a matri x is like ly to retain suspended matter and polymeric substances by phys i­cal obstructi ons. In addi tion, a gel structure li ke the one tested in this work has been show n to possess di f­fe rent acti ve pore vo lumes fo r di ffe rent molecular weight solu tes (e.g. BSA and Blue Dex tran), thereby

.c ~ III ...... <Jl

OJ E o o ..­.... Q) Q.

'0 Q) U ::J

U e Q.

Q) <Jl o U ::J

OJ OJ E

50

45

40

.. .............

a

0000 •

o 10 20 30

Time in hours

Fig. 3-Performance of alpha amy lase immobili zed on CELBEA DS using 3% starch, in different modes [(0) Soluble enzy me; (. ), Batch; (~), Packed bed; ( ... ), Expanded bed]

indicat ing a pore size distri bution whi ch is more like ly to trap suspended matter and polymeric sub­stances than a matri x th at is macroporous and has monodi sperse pore size d istribu tion (A mritkar et al. , 2002 unpubli shed work). Further, it can be expected that g lobul ar macro molecules or rig id suspended par­ticles will show different retention behav iour than li near or branched po lymer chains in a macroporous matri x of any kind.

Bioconversioll llsing immobilized enzymes Perfo rmance of alpha amy lase and pectinase Im­

mobili zed on CELB EA DS, in batch, packed bed and fl uidi zed bed modes is shown in Fig. 3 and Fig. 4 , respec tively. The runs were carried out using equiva­lent units of enzymes in each case. Fig. 3 shows that fas tes t convers ion of starch to reducing sLl gars was obtained with soluble enzy me. However the batch and packed bed runs with immobilized enzyme gave simil ar results, and 85% conversion (o f that with soluble enzy me) was achieved in about 20 hr. Flu id­ized bed run gave lesser (64%) conversion in same time. On the other hand, pectin hydro lys is showed same results with soluble, immobilized enzy me in batch, and immobili zed enzy me in packed bed gave nearl y same conversion of pectin to reducing sugars. Fluidized bed conversion however was far slower than the other th ree cases. Reuse of the immobilized enzyme over 5 rLlns was possible without any signifi-

80 0> E

70 0 0

0

• r- 60 "-Q) 0.. 50 "0 C Q) :0::;

E u 40 "- Q) 0 0..

'+--30 Q)

(/)

0 20 u :J

0 t:.

t:. • ~ 0

... ... ... t:. ...

... • ... • ...

0> 10 0>

E 0

0 5 10 Time in hours

Fig. 4-Performance of pectinase immobili zed on CELBEADS using 0.5% pectin, in diffe rent modes [(0), Soluble enzyme; (. ), Batch; (~), Packed bed; ( ... ), Expanded bed]

258 INDIAN J. BIOCHEM. BIOPHYS., VOL. 39, AUGUST 2002

cant loss of activity in each of the batch, packed bed and expanded bed cases.

Fluidized conversion in all cases is likely to be poorer since the through-flow of the solution and sub­strates is not possible when the beads are fluidized and the pressure differential across each bead is lower than it is in a packed bed at the same flow rate. Fur­ther, the flowing fluid does not provide enough time for the substrate to diffuse into the matrix and use all the available enzyme sites. Catalytic si te utilization can be expected to be far more in packed bed when there is a through-flow in the bead pores even if the residence time in the bed is low. On the other hand, in a batch system the large pores do not offer any appre­ciable resistance to concentration driven diffusion of the substrates, thereby giving comparable conversions to free soluble enzyme runs. It is to be noted that the equivalent enzyme activities used are on material bal­ance basis and not on the basis of the actual intrinsic activity of the enzyme after immobilization.

Conclusions The present work was carried out specifically to

evaluate the possible advantage of the macroporosity of a beaded matrix like CELBEADS when used as an enzyme support to carry out bioconversion of macro­molecular substrates that have been reported to per­form poorly on conventional immobilized enzyme supports. It was expected that with a pore size of the order of a few microns, both enhanced effective diffu­sion (due to very large pores) and facilitated diffusion (due to through-flow in pores in packed bed) would yield bioconversion results comparable to those ob­tained with soluble enzymes. Thus in cases where conventional enzyme supports have been reported to give undesirable results, use of CELBEADS immobi­lized amylase and pectinase were shown to give satis­factory results. Use of higher volume of immobilized matrix per unit volume of feed can be used to achieve faster conversions with both pectinase and amylase. Evaluation of rate parameters with appropriate reactor

design for scale up of bioconversion with CELBEADS is required and is currently under prog­ress.

Acknowledgement The authors are grateful to Prof. M N Gupta, De­

partment of Chemistry, Indian Institute of Techno­logy, New Delhi for useful suggestions to this work. This work was carried out under the project sponsored by Department of Science and Technology, Govern­ment of India.

References I Kvesitadze G I & Dvali M S H (1982) Biotechnol Bioeng,

24, 1765-1772 2 Tanyolac D, Yuruksoy B I & Ozdural A R (1998) Biochem

Eng J 2,179-186 3 Tumturk H, Aksoy S & Hasirci N (2000) Food Chem 68,

259-266 4 Kiminkova M & Kucera J (1982) Enzyme Microb Technol 5,

204-208 5 Hoshino K, Katag iri M Taniguchi M, Sasakura T & Fuji M

(1994) J Fennent Bioeng 77, 407-412 6 Chen J P, Chu DH & Sun Y M (1997) J Chem Tech Biotech­

nol 69, 421-428 7 Siso MIG, Grabber M, Condoret J-S & Combes D (1990) J

Chem Tech Biotechnol 48, 185-200 8 Romero C, Manjon A & Iborra 1 L (1988) Biotechnol Letters

10(2),97-100 9 Pai A, Gondkar S & Lali A (2000) J Chromatogr 867(1-2) ,

113-130 10 Gondkar S B, Manudhane K S, Amritkar N S, Pai A U &

Lali A M (200 I) Biotecllllol Progress 17(3), 522-529 II Miller G L (1959) Anal Chem 31 , 426-428 12 Raxova-Benkova L. Omelkova 1, Veruovic B & Kubanek V

(1989) Biotechnol Bioeng 34, 79-85 13 Bradford M A (1976) Anal Biochem 72, 248-254 14 Hermanson G T, Mallia A K & Smith P K (1992) Immobi­

lized Affinity Ligands and Techniques, Chapt. 2, pp 78-79. Chapt. 3, pp. 182-183, Academic Press Inc

15 Manudhane K S (2002) Designer polymers in biotransfor­mations and bioseparations, PhD thesis, Mumbai Universi ty, Mumbai

16 Rodrigues A E, Chenou C & Vega M R (1996) Chem Eng J 61,191-201

17 Gustavsson P, Axelsson A & Larsson P (1998) J Chroma­togr A 795, 199-210