Improvement of Chitosan Derivatization for the Immobilization ofBacillus circulans β‑Galactosidase and Its Further Application inGalacto-oligosaccharide SynthesisPaulina Urrutia, Claudia Bernal, Lorena Wilson, and Andres Illanes*

School of Biochemical Engineering, Pontificia Universidad Catolica de Valparaíso, Avenida Brasil, 2147 Valparaıso, Chile

*S Supporting Information

ABSTRACT: Chitosan was derivatized by two methodologies to design a robust biocatalyst of immobilized Bacillus circulans β-galactosidase from a low-cost support for its further application in the synthesis of galacto-oligosaccharides (GOS). In the firstone, chitosan was derivatized by cross-linking with glutaraldehyde and activated with epichlorohydrin; in the second one, cross-linking and activation were done with epichlorohydrin in a two-step process, favoring first support cross-linking and then supportfunctionalization (C-EPI-EPI). Epoxy groups were hydrolyzed and oxidized, obtaining two supports activated with differentaldehyde concentrations (100−250 μmol/g). The expressed activity and stability of the immobilized biocatalysts varied accordingto the derivatization methodology, showing that both the cross-linking agent and the activation degree are key parameters in thefinal biocatalyst performance. The best compromise between expressed activity and thermal stability was obtained using C-EPI-EPI with 200 μmol of aldehyde groups per gram of support. The immobilization conditions were optimized, obtaining abiocatalyst with 280 IU/g, immobilization yields in terms of activity and protein of 17.3 ± 0.4 and 61.5 ± 3.9%, respectively, anda high thermal stability, with a half-life of 449 times the value of the soluble enzyme. The biocatalyst was applied to the synthesisof GOS in repeated batch operation without affecting the product composition. Four successive batches were required forobtaining a cumulative specific productivity higher than the one obtained with the soluble enzyme.

KEYWORDS: chitosan, β-galactosidase, immobilization, galacto-oligosaccharides

■ INTRODUCTION

β-Galactosidase (β-D-galactoside galactohydrolase, EC 3.2.1.23)is a widely utilized enzyme in the dairy industry mainly due toits ability to hydrolyze lactose from milk.1 A newer and moreprofitable application of this enzyme is the catalysis oftransgalactosylation reactions in which a galactose moiety istransferred to a nucleophilic acceptor other than water,potentially any sugar present in the reaction medium, leadingthen to the synthesis of oligosaccharides.2−4 When lactose isused as substrate, galacto-oligosaccharides (GOS) are synthe-sized. GOS are products with a recognized prebiotic capacitycomposed by two, three, or even more sugar units, of whichglucose is the terminal unit and galactose the rest of them.5,6

The synthesis of GOS is a kinetically controlled reaction wherethe product concentration increases to a maximum value andthen decreases as a result of the competition between reactionsof hydrolysis and synthesis. The prevalence of synthesis overhydrolysis varies according to the β-galactosidase origin,7,8

initial lactose concentration,9−11 and the water thermodynamicactivity.12 The use of soluble or immobilized enzymes may alsoaffect GOS yield, because the diffusional restrictions of theimmobilized biocatalyst may limit the maximum productconcentration attainable13 and immobilization may alter theenzyme catalytic properties.14

Among the different β-galactosidases that have been utilizedin GOS synthesis, the one from Bacillus circulans is oftenpreferred due to the higher GOS yield attained.15,16 Severalisoenzymes have been identified in commercial preparations ofB. circulans β-galactosidase: Mozaffar et al.17 reported the

presence of two monomeric isoenzymes with molecular weightsof 240 and 160 kDa and similar isoelectric points of 4.5; later,Vetere and Paoletti18 reported the presence of threemonomeric isoenzymes (212, 145, and 86 kDa), and Song etal.19 isolated a fourth monomeric isoform, establishing thatisoenzymes molecular weights were 189, 154, 135, and 92 kDa.The synthesis of GOS was evaluated with each isoenzyme at 40°C, pH 6, and 5% of lactose, reporting that maximum GOSconcentration varied according to the β-galactosidase isoformused.19 Warmerdam et al.20 evaluated the effect of initial lactoseconcentration on the synthesis of GOS with the isolatedisoenzymes, finding that the increase of lactose concentrationresulted in a smaller difference among the GOS yields obtainedwith each isoform.Biocatalysts with high thermal stability are desirable because

high reaction temperature will increase lactose concentrationand favor synthesis over hydrolysis. Enzyme immobilization is amature technology that may allow enzyme stabilization with theadditional advantage of obtaining an insoluble biocatalyst thatcan be easily recovered, extending its lifespan, facilitating thecontrol of the reaction, and improving product quality.21,22

Additionally, the immobilization of enzymes may modify theircatalytic properties.23,24 Among the different methods forenzyme immobilization, immobilization by covalent bonding

Received: January 22, 2014Revised: September 4, 2014Accepted: September 4, 2014

Article

pubs.acs.org/JAFC

© XXXX American Chemical Society A dx.doi.org/10.1021/jf500351j | J. Agric. Food Chem. XXXX, XXX, XXX−XXX

usually offers the highest degree of rigidification25 and,consequently, enzyme stabilization, being normally selectedwhen the objective is to improve the enzyme stability and/orwhen the immobilized biocatalyst will be used in an aqueousmedia where immobilization by adsorption is in most casesprecluded.26 However, the strength of the enzyme−supportlinkage makes it difficult or impossible to recover the supportafter enzyme inactivation; therefore, support cost may have ahigh impact on the economics of the process employing theimmobilized biocatalyst.Chitosan is a natural biopolymer outstanding for its relatively

low cost, innocuousness, biocompatibility, antimicrobialactivity, and the presence of functional groups that may bechemically modified.27,28 It is a linear polysaccharide composedof β-(1−4)-linked D-glucosamine and N-acetyl-D-glucosamineunits, which is produced by partial deacetylation of chitin that isobtained from the exoskeletons of crustaceans, particularly from

wastes of the seafood-processing industry.29 The use ofchitosan as a carrier for the covalent immobilization of enzymesrequires the introduction of functional groups that can reactwith the amino acid residues of the protein surface. Addition-ally, polymer chains must be cross-linked to improve thesupport mechanical resistance and avoid its solubilization inaqueous acidic media due to its cationic nature (pKa of theamino groups ∼ 6.5).27 One typical derivatization method isthe cross-linking and activation of chitosan with the bifunc-tional reagent glutaraldehyde (GA). The chemistry of thereaction between GA and amino groups is not fully understood,and according to the pH, the reaction mechanism may involveSchiff bases, nucleophilic substitution of amino groups, andMichael addition.30,31 The reaction of GA with chitosan allowsthe cross-linking of different polymer chains and the activationof the support with aldehyde groups that may react with theenzyme through its amino groups (lysine residues and terminal

Figure 1. Possible structures after cross-linking (1) and activation (2) of C-EPI-EPI (a) and C-GA-EPI (b).

Journal of Agricultural and Food Chemistry Article

dx.doi.org/10.1021/jf500351j | J. Agric. Food Chem. XXXX, XXX, XXX−XXXB

amino group)32,33 and eventually with other surface functionalgroups (thiols, phenols, and imidazoles).30 Epichlorohydrin(EPI) has been also utilized in the derivatization of chitosanbecause it may react with amino and hydroxyl groups of thepolymer, cross-linking different chitosan chains and activatingthe support with epoxide groups. The reaction occurs mainlythrough the chitosan hydroxyl groups when high temperatureand alkaline media are utilized.34,35 Because part of the residualepoxy groups may be spontaneously hydrolyzed, a heterofunc-tional support is obtained, presenting epoxide, diols, and aminogroups. The residual epoxy groups may be also hydrolyzed, anddiols could be then oxidized with periodate to aldehyde groupsthat form Schiff bases with the ε-amino group of protein lysineresidues. To avoid the opening of the chitosan pyranosidic ringsdue to the oxidation of vicinal diols with periodate,36 a limitedquantity of periodate should be added and chitosan chainsshould be previously cross-linked. This final step of supportmodification presents the advantage of obtaining a morehomogeneous support, and aldehydes may favor a more intensemultipoint covalent attachment than when epoxy groups areutilized because the instability of Schiff bases should favorimmobilization through the enzyme surface area with thehighest density of lysine residues.37

It has been reported that according to the methodologyutilized for chitosan derivatization, the immobilization yield andthermal stability of the biocatalyst may be modified.33,38,39

These results may be related to the modification of chitosanfunctional groups, the activation degree, and/or the mod-ification of the structure of the support by varying the cross-linking process. The objective of this investigation was todesign a methodology for chitosan derivatization that offers thebest compromise of expressed activity and thermal stability ofan immobilized biocatalyst of B. circulans β-galactosidase for itsfurther application in GOS synthesis in repeated batchoperation.

■ MATERIALS AND METHODSMaterials. A commercial preparation of β-galactosidase from B.

circulans (Biolactasa-NTL CONC X2) in the form of a liquid solutionwas purchased from BIOCON (Barcelona, Spain), presenting 31.3 ±0.5 mg of protein per mL and 1353 ± 11 IU per mL (determination ofenzyme activity is in section Analysis). o-Nitrophenyl-β-D-galactopyr-anoside (o-NPG) was from Carbosynth Limited (Bershire, UK).Lactose monohydrate was from Quimatic S.A. (Santiago, Chile).Sodium metaperiodate and glycerol were purchased from Merck(Darmstadt, Germany). EPI, GA, sodium borohydride, o-nitrophenol(o-NP), and chitosan (≥75% deacetylated; MW 190,000−375,000;batch 061M0046 V) were from Sigma-Aldrich (St. Louis, MO, USA).All other reagents were of the highest available purity and used aspurchased.Analysis. Enzymatic activity of β-galactosidase was assessed

spectrophotometrically by measuring the production of o-NP at 420nm. The reaction was initiated by adding 50 μL of β-galactosidasesolution or suspension to 2 mL of 45 mM o-NPG solution, using atemperature-controlled cell with constant magnetic stirring. Oneinternational unit of β-galactosidase activity (IU) was defined as theamount of enzyme producing 1 μmol of o-NP per minute from a 45mM o-NPG solution in 0.1 M citrate−phosphate buffer, pH 6, at 25°C. The extinction molar coefficient of o-NP under the assayconditions was 560 M−1 cm−1.Protein concentration was determined according to the Bradford

methodology, using bovine serum albumin as standard.40

Lactose and products of GOS synthesis (galactose, glucose, andGOS) were measured as previously reported, using a Jasco RI 2031HPLC delivery system, provided with a refractive index detector, anisocratic pump (Jasco PU2080), an autosampler (Jasco AS 2055), and

BP-100 Ag columns (300 mm × 7.8 mm) for carbohydrate analysis(Benson Polymerics, Reno, NV, USA).41

Support Preparation. Chitosan particles were produced aspreviously reported.39 Chitosan was solubilized in acetic acid solution(3% v/v) to achieve a concentration of 1.5% (w/v) and stirred for 12h. This solution was fed by gravity and sprayed into 1 M NaOH in40% (v/v) ethanol solution under stirring. The particles wererecovered and washed with distilled water and phosphate buffer(0.05 M, pH 7.0). The particles were derivatized by two differentmethodologies, and the possible chemistry involved after cross-linkingand activation of chitosan is presented in Figure 1.

Cross-Linking with Glutaraldehyde and Activation with Epi-chlorohydrin (C-GA-EPI). Chitosan particles were cross-linked withGA as previously reported,39,42 modifying the quantity of GA utilized.To favor the polymer cross-linking, 1 mol of GA was added per 4equiv mol of chitosan amino groups, assuming 75% deacetylation. Theimines formed by the reaction of GA with the amino groups ofchitosan were reduced to secondary amines with NaBH4 to form stablebonds. Finally, the cross-linked chitosan was activated with excess EPIto favor chitosan activation. A typical cross-linking and activationprocess was performed as follows: 1 g of chitosan particles wassuspended in 10 mL of a solution of 0.78 g/L of GA prepared in 5 mMphosphate buffer, pH 7, and maintained under stirring for 30 min at 4°C. The cross-linked particles were washed with distilled water. Forthe reduction of imines, 1 g of cross-linked particles was suspended in10 mL of a solution of 3.8 g/L of NaBH4 in 0.1 M bicarbonate buffer,pH 10, and maintained under stirring for 2 h at 25 °C. Afterward, theparticles were washed with distilled water. The support was activatedby suspending 1 g of support in 6 mL of an aqueous solutioncontaining 40 g/L of NaOH, 37% (v/v) acetone, and 3.8 g/L ofNaBH4 at 4 °C. Then, EPI was added to reach a final concentration of185 g/L. The mixture was stirred at 25 °C for 18 h, and the supportwas carefully washed with 20% (v/v) aqueous acetone solution, withdistilled water, and finally with 50 mM phosphate buffer, pH 7.0. Thesupport was finally contacted with a 0.5 M H2SO4 solution during 1 hto hydrolyze all unreacted epoxy groups, and then the diol groups wereoxidized during 2 h with an aqueous solution containing NaIO4 (aratio of 1 g of support per 10 mL of solution was used in both steps).Because 1 mol of NaIO4 oxidizes 1 equiv mol of diol, the activationdegree (moles of aldehyde groups per gram of support) was modifiedby varying the concentration of NaIO4 utilized during the oxidationstep (10−25 mM).

Cross-Linking and Activation with Epichlorohydrin (C-EPI-EPI).Chitosan particles were derivatized in a two-step process with EPI aspreviously reported.24 For support cross-linking, 1 mol of EPI wascontacted per 4 equiv mol of primary hydroxyl groups of chitosan,assuming a 75% deacetylation. For this process, the chitosan particles(1 g) were resuspended in 2 mL of mixture with 4.8 g/L of EPI and2.7 g/L of NaOH at 45 °C during 2 h. The cross-linked support wasthen activated with EPI following the same procedure as the activationof C-GA-EPI. For both supports the quantification of aldehyde groupswas done spectrophotometrically by back-titration with NaHCO3/KI.38

Enzyme Immobilization. The enzyme was immobilized in C-GA-EPI and C-EPI-EPI contacting 10 mL of the enzyme solution per 1 gof support at 25 °C. The β-galactosidase solution was prepared in 0.1M bicarbonate buffer, pH 10, with 20% (v/v) of glycerol, which wasused to avoid enzyme inactivation during immobilization. Finally, toobtain stable bonds between the enzyme and the support, the Schiffbases formed were reduced by adding NaBH4 to reach a finalconcentration of 1 mg/mL. The suspension was maintained understirring during 30 min at 25 °C, and the biocatalyst was thoroughlywashed with distilled water.

Immobilization was followed by measuring the enzyme activity inthe suspension and in the supernatant. The activity of the suspensionwas determined according to the same methodology utilized for thesoluble enzyme. Immobilization yield in terms of the bound protein(IYP) was defined as the mass ratio of immobilized protein (differencebetween contacted protein and unbound protein in the supernatant)to contacted protein. Immobilization yield in terms of expressed

Journal of Agricultural and Food Chemistry Article

dx.doi.org/10.1021/jf500351j | J. Agric. Food Chem. XXXX, XXX, XXX−XXXC

activity (IYA) was defined as the ratio of the activity expressed by theimmobilized biocatalyst to the contacted activity.Selection of the Methodology of Chitosan Derivatization

and Optimization of Immobilization Conditions. To select amethodology of chitosan derivatization, an enzymatic solution of 1.3mg protein/mL was immobilized in C-GA-EPI and C-EPI-EPI withdifferent activation degrees for 10 h, under the conditions previouslymentioned. The performance of the biocatalysts was evaluated interms of their immobilization yield and thermal stability. The thermalstability was evaluated through the stability factor (SF), defined as theratio of the half-lives of immobilized and soluble biocatalyst at 60 °Cand pH 6 (0.1 M citrate−phosphate buffer). Half-life was calculatedafter modeling biocatalyst inactivation kinetics by a biphasicmechanism:

α α= + ×−

× − × − ×−

× − ×

⎡⎣⎢

⎤⎦⎥

⎡⎣⎢

⎤⎦⎥

AA

kk k

k tk

k k

k t

1 exp( )

exp( )

I

I0

1

2 11

1

2 1

2 (1)

AI/AI0 represents the residual activity at time t, α is the specific activityratio of the intermediate species with respect to the native enzymespecies, and k1 and k2 are the transition rate constants from the nativeto the intermediate and from the intermediate to the final enzymespecies, respectively.To consider both the activity of the immobilized biocatalyst and its

thermal stability, the lumped parameter maximum catalytic potential(MCP) was also utilized for selecting a chitosan derivatizationmethodology.39,41 MCP integrates the variation of expressed activity(AI) along biocatalyst inactivation (eq 2), considering the time ofbiocatalyst utilization (tf) as the one at which residual activity is 35%.

∫= A tMCP dt

0I

f

(2)

After selecting the chitosan derivatization methodology, thecontacted protein and immobilization time were optimized using acentral composite rotatable design of two variables at two levels withfive replicas at the central point.

Galacto-oligosaccharide Synthesis. GOS synthesis with thesoluble and immobilized β-galactosidase was carried out using 50%(w/w) lactose solution at pH 6 (0.1 M citrate−phosphate buffer) and60 °C. β-Galactosidase activity in the reaction mixture was adjusted to40 IU/g lactose, as previously reported.41 To evaluate the impact ofreusing the immobilized enzyme on GOS synthesis and measure thebiocatalyst thermal stability under reactive conditions, sequentialbatches of GOS synthesis were carried out. In this case, the ratio ofbiocatalyst mass to lactose mass was maintained constant and equal tothe one used in the first batch.

The following parameters were used for evaluating GOS synthesis:Yield (YGOS): maximum GOS mass (MGOS) per unit mass of initial

lactose mass (ML0)

= ×YMM

100GOSGOS

L0 (3)

Volumetric productivity (π): total GOS mass per unit of reactionvolume (V) and unit of reaction time (t). For the purposes of thiswork, π was evaluated at the time when maximum GOS concentrationwas achieved.

π =×

MV t

GOS(4)

Cumulative specific productivity (πC): total GOS mass obtainedover the whole operation of sequential batches (n batches) per unitmass of biocatalyst protein (MBP) and unit of reaction time. For thepurposes of this work, total GOS mass of each batch was evaluated atthe moment when maximum GOS concentration was achieved.

Figure 2. Effect of the activation degree of C-GA-EPI (gray bars) and C-EPI-EPI (black bars) on the performance of immobilized B. circulans β-galactosidase: (a) immobilization yield in terms of expressed activity (IYA); (b) immobilization yield in terms of protein (IYP); (c) stability factor(SF); (d) maximum catalytic potential (MCP). Immobilization conditions: 0.1 M citrate−phosphate buffer, pH 10, 20% glycerol, and 25 °C.Inactivation conditions: 60 °C and 0.1 M citrate−phosphate buffer, pH 6.

Journal of Agricultural and Food Chemistry Article

dx.doi.org/10.1021/jf500351j | J. Agric. Food Chem. XXXX, XXX, XXX−XXXD

π =∑

× ∑

M

M t

n

nCbatch1batch

GOS

BP batch1batch

(5)

In the case of the immobilized enzyme, πC is expressed in terms ofthe mass of protein that was contacted with the support for biocatalystimmobilization. MB represents the mass of biocatalyst utilized in thefirst batch of GOS synthesis.

π =∑

× × ∑

M

M P t

n

nCbatch1batch

GOS

B c batch1batch

(6)

■ RESULTS AND DISCUSSIONSelection of Chitosan Derivatization Method. Chitosan

particles were derivatized either using glutaraldehyde for cross-linking and epichlorohydrin for support activation (C-GA-EPI)or else using epichlorohydrin for both cross-linking andactivation (C-EPI-EPI). The derivatization of chitosan withepichlorohydrin was done in a two-step process because it waspreviously reported39 that this methodology allows a higherdegree of activation than when chitosan is derivatized in justone step.38 This difference may be related to the fact that thefirst step of cross-linking should produce some sterichindrances that limit the additional cross-linking of chitosanchains when a high concentration of epichlorohydrin iscontacted during the second step of derivatization, favoringthe support activation. The characterization of C-GA-EPI ispresented as Supporting Information. In the case of C-EPI-EPI,its physical and chemical characterization was reportedpreviously,39 showing that the used methodology allows themodification of the amino and hydroxyl groups of chitosan. Inthe case of C-GA-EPI, the modification was tested with FTIR,showing that the GA and EPI groups have reacted with thechitosan matrix and new aldehyde groups were introduced asreflected by the best resolution bands related with the vibrationof CN of amine and CO of aldehydes (1570 and 1650cm−1, respectively) (Figure SM3, Supporting Information).The concentration of aldehyde groups of C-GA-EPI and C-

EPI-EPI was modified by varying the concentration of NaIO4utilized for the oxidation of the support. After oxidation, allNaIO4 was reacted, obtaining supports with activation degreesin the range from 100 to 250 μmol of aldehyde per gram. Theeffect of chitosan derivatization method over IYA, IYP, SF, andMCP of the immobilized enzyme is presented in Figure 2. Theinactivation kinetics of each biocatalyst was modeled by abiphasic mechanism (eq 1), and the values of thecorresponding inactivation parameters are summarized inTable 1.The activation degree of both supports affected the

biocatalysts’ expressed activity; however, immobilized proteinvaried significantly only in the case of C-EPI-EPI. In general, C-GA-EPI produced the higher values of IYA and IYP independentof the aldehyde concentration in the support. With bothsupports IYP was higher than IYA, which is explained by theimmobilization of proteins different from the β-galactosidase(the commercial enzyme preparation is not pure) and also bythe presence of diffusional restrictions and/or inactivation ofthe catalyst due to the immobilization process. The stability(SF) of the biocatalysts immobilized in both supports alsovaried according to the activation degree. The maximum valuesof IYA and SF were obtained when the C-GA-EPI and C-EPI-EPI were activated with 150 and 200 μmol of aldehyde pergram of support, respectively. The increase in the concentration

of aldehyde groups in the support should allow theimmobilization of a higher proportion of the contacted enzymeand favor multipoint interaction between the protein and thesupport, which may explain the increase of expressed activityand stability until 150 and 200 μmol of aldehyde per gram of C-GA-EPI and C-EPI-EPI, respectively. On the other hand, thedecrease of IYA in the supports with a higher activation degreemay be explained by the formation of covalent bonds close tothe active site or by an excessive stiffening of the protein,reducing mobility below the minimum level required forcatalysis. In the case of SF, an excessive stiffening of the proteinmay expose areas more prone to denaturation, explaining thelower thermal stability of the biocatalysts. In general,independent of the activation degree, best results were obtainedwith C-GA-EPI in terms of IYA: in the case of C-GA-EPI150 IYAwas almost twice that obtained with C-EPI-EPI150. Theopposite effect was observed for SF, a better performancebeing obtained with the β-galactosidase immobilized in C-EPI-EPI, irrespective of the activation degree. The highest stabilitywas obtained with C-EPI-EPI200, for which the SF was 4 timesthe value reported for the biocatalyst of B. circulans β-galactosidase immobilized in glyoxyl-agarose under optimizedconditions and 7 times the value reported for the enzymeimmobilized in silica support.41,43 Because C-GA-EPI and C-EPI-EPI were activated using the same methodology, thedifferent performance obtained with both supports may beexplained by a difference in support structure as a consequenceof their cross-linking process. When the combined effect ofactivity and stability was assessed using MCP as lumpedparameter, the support C-EPI-EPI200 offered the best perform-ance due to the prevalence of the effect of thermal stabilization.This support was selected for the further optimization ofprotein loading and immobilization time.The leaching of β-galactosidase from the derivatized supports

was not evaluated in this work because Bernal et al.44 showedthat the covalent immobilization of this enzyme in a supportactivated with aldehyde groups avoids the leaching of theenzyme.

Optimization of Immobilization Conditions. Biocatalystspecific activity and thermal stability are important parametersfor the industrial application of immobilized enzymes becausethe first one will define the quantity of catalyst required for aspecific reaction and the latter will determine its lifespan.Biocatalyst specific activity varies according to the quantity ofenzyme contacted during the immobilization process, whereasit has been reported that the thermal stability of an immobilizedenzyme may be improved through the increase of immobiliza-

Table 1. Inactivation Parameters of B. circulans β-Galactosidase Immobilized in Different Chitosan Supports,Determined According to a Series Biphasic Model (Equation1)a

support k1 (h−1) k2 (h

−1) α R2

C-GA-EPI100 1.35 0.03 0.28 0.998C-GA-EPI150 0.77 0.02 0.29 0.988C-GA-EPI200 0.94 0.03 0.37 0.999C-GA-EPI250 1.37 0.05 0.28 0.996C-EPI-EPI100 1.14 0.03 0.53 0.996C-EPI-EPI150 1.83 0.03 0.60 0.999C-EPI-EPI200 0.63 0.02 0.58 0.996C-EPI-EPI250 2.54 0.03 0.63 0.993

aR2, coefficient of determination.

Journal of Agricultural and Food Chemistry Article

dx.doi.org/10.1021/jf500351j | J. Agric. Food Chem. XXXX, XXX, XXX−XXXE

tion time.45 To set the range of contacted protein to beconsidered in the optimization of immobilization conditions,the maximum protein loading capacity of C-EPI-EPI wasassessed (Figure 3), obtaining a maximum of 23.8 mg of

immobilized protein per gram of support. This maximumloading capacity was 45 and 20% lower than the values reportedfor the immobilization of B. circulans β-galactosidase in glyoxyl-agarose41 and acrylic resins,46 respectively. Such a differencemay be due to the lower total surface area. The range ofimmobilization time considered for the optimization of theimmobilization process was established after evaluating thestability of the biocatalysts at maximum protein loading and at10% of that maximum. Two different immobilization timeswere considered: the minimum time required for obtaining aconstant activity in the supernatant (tmin) and tmin plus 24 h. Asshown in Figure 4, thermal stability of the biocatalyst loadedwith the maximum quantity of protein was independent ofimmobilization time, in contrast to the biocatalyst loaded with alower quantity of protein. This effect may be explained by adifference in the multipoint interaction between the enzymeand support due to the nonhomogeneous surface of thesupport, obtaining higher interaction intensity when theenzyme is first linked to the most reactive areas and a lowerinteraction intensity when the enzyme is immobilized in theless reactive areas. This effect was also reported for glyoxyl-agarose.41

Considering the interaction observed between the quantityof protein contacted and the immobilization time, bothvariables were optimized by a central composite design, usingMCP as response parameter (Figure 5). The response surfacewas modeled by a second-order polynomial equation (eq 7),obtaining the maximum value of MPC when 25.8 mg/g wascontacted for tmin plus 14.1 h (total immobilization time = 38.1h).

= × + × − ×

× − × − × −

P t t P

P t

MCP 35010.8 17136.5

127.9 640.5 491.4 190011c add add c

c2

add2

(7)

The ANOVA test was performed, obtaining a p value of0.0005, an R2 value of 0.9366, and an adjusted R value of0.8913. The optimized conditions were validated experimen-

tally, obtaining a MCP of 367,032 μmol/g (4.5% lower than thepredicted value). The optimized biocatalyst expressed 280 ± 7IU/g, with an IYA of 17.3 ± 0.4% and an IYP of 61.5 ± 3.9%.The expressed activity was 53% lower than the value reportedfor the enzyme immobilized under optimized conditions inglyoxyl-agarose.41 However, the SF obtained with the enzymeimmobilized in chitosan was 450% higher than the correspond-ing value obtained with the agarose biocatalyst.41 Additionally,the use of chitosan presents an economic advantage, becauseone of the limiting factors in the use of agarose is its highcost.47

Galacto-oligosaccharide Synthesis. The optimized bio-catalyst was utilized in the synthesis of GOS in repeated batchoperation. A comparison of YGOS and π values obtained in onebatch of GOS synthesis catalyzed by the soluble andimmobilized β-galactosidase is presented in Table 2. The

Figure 3. Loading capacity of C-EPI-EPI200 expressed in terms of B.circulans β-galactosidase protein content.

Figure 4. Thermal inactivation at 60 °C and pH 6 (0.1 M citrate−phosphate buffer) of the soluble enzyme (closed circles) andderivatives of β-galactosidase immobilized in C-EPI-EPI200 atmaximum protein loading capacity (solid symbols) and at 10% ofthat maximum (open symbols) at different immobilization times:(rhombuses) tmin; (squares) tmin + 48 h. tmin is the minimum timerequired to obtain a constant activity in the supernatant.

Figure 5. Optimization of B. circulans β-galactosidase immobilizationin C-EPI-EPI200, considering additional immobilization time (tadd) andcontacted protein as variables and maximum catalytic potential (MCP)as objective function. Immobilization conditions: 0.1 M citrate−phosphate buffer, pH 10, 20% glycerol, and 25 °C.

Journal of Agricultural and Food Chemistry Article

dx.doi.org/10.1021/jf500351j | J. Agric. Food Chem. XXXX, XXX, XXX−XXXF

production of GOS was not affected by the immobilization ofthe enzyme, as reflected by the lack of a significant variation inYGOS and π. These results indicate that diffusional restrictions inthe immobilized enzyme are not significant and theimmobilization process did not produce conformationalchanges or steric hindrances that affect the activity of thebiocatalyst. In contrast, Huerta et al.42 and Sheu et al.48

reported that the synthesis of GOS with Aspergillus oryzae β-galactosidase immobilized in chitosan derivatized with gluta-raldehyde resulted in a lower YGOS than the one obtained withthe soluble enzyme, a result explained by the presence ofdiffusional restrictions. The insignificant effect of diffusionalrestrictions in the biocatalyst of B. circulans β-galactosidaseimmobilized in C-EPI-EPI200 may be explained by the use of alower enzyme loading and/or by differences in the support sizeand structure. Considering that several isoenzymes of B.circulans β-galactosidase have been identified in the commercialsoluble enzyme preparation,18−20 the absence of variation inYGOS and π also indicates that the relative activity of eachisoform was not significantly altered upon immobilization. Theprofile of GOS obtained with β-galactosidase immobilized in C-EPI-EPI200 was similar to the one obtained with the solubleenzyme (Figure 6). This pattern remained unchanged along 10repeated batches, as may be observed in Figure 7.

To make a fair comparison between the soluble andimmobilized β-galactosidase, the parameters cumulative specificproductivity (πc) and total mass of GOS per mass of contactedprotein (RGOS) were evaluated. The expression of eachparameter in terms of the contacted protein takes intoconsideration the immobilization yield; therefore, both thesoluble and immobilized enzymes are compared considering

Table 2. Comparison of Synthesis of GOS with Soluble β-Galactosidase and β-Galactosidase Immobilized in C-EPI-EPI200

a

parameter soluble β-gal β-gal/C-EPI-EPI200

YGOS (%) 38.3 39.1π (gGOS/h·L) 462.2 473.7

aYGOS, GOS yield; π, volumetric productivity.

Figure 6. Profiles of GOS synthesized by B. circulans β-galactosidase: (a) total GOS; (b) trisaccharides; (c) tetrasaccharides; (d) penta- andhexasaccharides; (solid symbols) soluble enzyme; (open symbols) enzyme immobilized under optimized conditions in C-EPI-EPI200. Reactionconditions: 50% (w/w) initial lactose concentration, 0.1 M citrate−phosphate buffer, pH 6, 60 °C, and 40 IU/g initial lactose.

Figure 7. Product distribution of GOS synthesized by the β-galactosidase from B. circulans immobilized in C-EPI-EPI200 alongbatches in repeated batch operation: (black bars) trisaccharides; (graybars) tetrasaccharides; (white bars) penta- and hexasaccharides.Reaction conditions: 50% (w/w) initial lactose concentration, 0.1 Mcitrate−phosphate buffer, pH 6, and 60 °C. Biocatalyst added in thefirst batch was adjusted to 40 IU/g initial lactose, and for thesubsequent batches the ratio of biocatalyst mass to lactose mass wasmaintained constant and equal to the one used in the first batch.

Journal of Agricultural and Food Chemistry Article

dx.doi.org/10.1021/jf500351j | J. Agric. Food Chem. XXXX, XXX, XXX−XXXG

the total mass of protein expended for the production of GOS.The values of πc along 10 successive batches are presented inFigure 8a. The πC of the first batch of synthesis was lower for

the immobilized enzyme due to the loss of activity duringimmobilization; however, the reutilization of the biocatalystcompensates for the loss of activity, requiring only foursequential batches to surpass the πC of the soluble enzyme.After 10 successive batches, the πC of the immobilized enzymewas 2.9 times the value obtained with the soluble enzyme. Thisvalue is lower than the πC reported for the enzyme immobilizedin glyoxyl-agarose41 due to the lower expressed activity of thebiocatalyst. The advantage associated with the high thermalstability of the enzyme immobilized in C-EPI-EPI200 should beobserved with a higher number of batches or extending thetime of each batch. Values of RGOS are presented in Figure 8b,observing that after 10 batches almost 1 kg of GOS per gram ofcontacted protein is produced when the immobilizedbiocatalyst is utilized. The thermal stability of the immobilizedenzyme was compared under reactive and nonreactiveconditions (Figure 9), observing a positive modulation by thereaction media. Under reactive conditions, <10% of the initial

biocatalyst hydrolytic activity was lost after 10 successivebatches.

In conclusion, the cross-linking agent utilized for chitosanderivatization and the activation degree of the support had ahigh impact on the performance of the immobilized B. circulansβ-galactosidase. The best compromise between expressedactivity and thermal stability was obtained with the enzymeimmobilized in cross-linked chitosan and activated withepichlorohydrin at a degree of activation of 200 μmol ofaldehydes per gram of support. The optimization ofimmobilization conditions (protein load and immobilizationtime) increased biocatalyst expressed activity and thermalstability. The optimized biocatalyst was applied to the synthesisof GOS, without affecting the product profile obtained with thesoluble enzyme. Four successive batches were required forobtaining a cumulative specific productivity higher than the oneobtained with the soluble enzyme. The appropriate design of alow-cost support, such as chitosan, is a good strategy for thedevelopment of a highly stable biocatalyst adequate for GOSsynthesis.

■ ASSOCIATED CONTENT*S Supporting InformationPhysical and chemical characterization of chitosan cross-linkedwith glutaraldehyde and activated with epichlorohydrin (C-GA-EPI). This material is available free of charge via the Internet athttp://pubs.acs.org.

■ AUTHOR INFORMATIONCorresponding Author*(A.I.) E-mail: aillanes@ucv.cl.FundingWork financed by Chilean Fondecyt Grant 1100050. Doctoralfellowship to P.U. from Conicyt-Chile is acknowledged.NotesThe authors declare no competing financial interest.

Figure 8. Comparison of soluble β-galactosidase (gray bars) and β-galactosidase immobilized under optimized conditions in C-EPI-EPI200(black bars): (a) cumulative GOS productivity in terms of contactedprotein (πc); (b) accumulated mass of GOS per unit mass of contactedprotein (RGOS). Reaction conditions: 50% (w/w) initial lactoseconcentration, 0.1 M citrate−phosphate buffer, pH 6, and 60 °C.Biocatalyst added in the first batch was adjusted to 40 IU/g initiallactose, and for the subsequent batches the ratio of biocatalyst mass tolactose mass was maintained constant and equal to the one used in thefirst batch.

Figure 9. Comparison of thermal stability of β-galactosidaseimmobilized under optimized conditions in C-EPI-EPI200 undernonreactive (circles) and reactive (squares) conditions. Nonreactiveconditions: 0.1 M citrate−phosphate buffer, pH 6, 60 °C. Reactiveconditions: 50% (w/w) initial lactose concentration, 0.1 M citrate−phosphate buffer, pH 6, 60 °C.

Journal of Agricultural and Food Chemistry Article

dx.doi.org/10.1021/jf500351j | J. Agric. Food Chem. XXXX, XXX, XXX−XXXH

■ REFERENCES(1) Adam, A. C.; Rubio-Texeira, M.; Polaina, J. Lactose: the milksugar from a biotechnological perspective. Crit. Rev. Food Sci. Nutr.2004, 44, 553−557.(2) Playne, M. J.; Crittenden, R. G. Galacto-oligosaccharides andother products derived from lactose. In Advanced Dairy Chemistry;McSweeney, P. L. H., Fox, P. F., Eds.; Springer Science&BusinessMedia: New York, 2009; pp 121−201.(3) Guerrero, C.; Vera, C.; Illanes, A. Optimisation of synthesis ofoligosaccharides derived from lactulose (fructosyl-galacto-oligosac-charides) with β-galactosidases of different origin. Food Chem. 2013,138, 2225−2232.(4) Irazoqui, G.; Giacomini, C.; Batista-Viera, F.; Brena, B. M.;Cardelle-Cobas, A.; Corzo, N.; Jimeno, M. L. Characterization ofgalactosyl derivatives obtained by transgalactosylation of lactose anddifferent polyols using immobilized β-galactosidase from Aspergillusoryzae. J. Agric. Food Chem. 2009, 57, 11302−11307.(5) Roberfroid, M. Prebiotics: the concept revisited. J. Nutr. 2007,137, 830−837S.(6) Tzortzis, G.; Vulevic, J. Galacto-oligosaccharide prebiotics. InPrebiotic and Probiotic Science and Technology; Charalampopoulos, D.,Rastall, R. A., Eds.; Springer Science&Bussiness Media: New York,2009; pp 207−244.(7) Boon, M. A.; Janssen, A. E. M.; van’t Riet, K. Effect oftemperature and enzyme origin on the enzymatic synthesis ofgalactooligosaccharides. Enzyme Microb. Technol. 2000, 26, 271−281.(8) Guerrero, C.; Vera, C.; Plou, F.; Illanes, A. Influence of reactionconditions on the selectivity of the synthesis of lactulose withmicrobial β-galactosidases. J. Mol. Catal. B: Enzym. 2011, 72, 206−212.(9) Huber, R. E.; Kurz, G.; Wallenfels, K. A. A quantitation of thefactors which affect the hydrolase and transgalactosylase activities of β-galactosidase (E. coli) on lactose. Biochemistry 1976, 15, 1994−2001.(10) Splechtna, B.; Nguyen, T. H.; Steinbo, M.; Kulbe, K. D.; Lorenz,W.; Haltrich, D. Production of prebiotic galacto-oligosaccharides fromlactose using β-galactosidases from Lactobacillus reuteri. J. Agric. FoodChem. 2006, 54, 4999−5006.(11) Gosling, A.; Stevens, G. W.; Barber, A. R.; Kentish, S. E.; Gras,S. L. Effect of the substrate concentration and water activity on theyield and rate of the transfer reaction of β-galactosidase from Bacilluscirculans. J. Agric. Food Chem. 2011, 59, 3366−3372.(12) Chen, S. X.; Wei, D. Z.; Hu, Z. H. Synthesis of galacto-oligosaccharides in AOT/isooctane reverse micelles by β-galactosidase.J. Mol. Catal. B−Enzym. 2001, 16, 109−114.(13) Gaur, R.; Pant, H.; Jain, R.; Khare, S. K. Galacto-oligosaccharidesynthesis by immobilized Aspergillus oryzae β-galactosidase. FoodChem. 2006, 97, 426−430.(14) Pessela, B. C. C.; Dellamora-Ortiz, G.; Betancor, L.; Fuentes,M.; Guisan, J. M.; Fernandez-Lafuente, R. Modulation of the catalyticproperties of multimeric β-galactosidase from E. coli by using differentimmobilization protocols. Enzyme Microb. Technol. 2007, 40, 310−315.(15) Boon, M. A.; Janssen, A. E. M.; van’t Riet, K. Effect oftemperature and enzyme origin on the enzymatic synthesis ofgalactooligosaccharides. Enzyme Microb. Technol. 2000, 26, 271−281.(16) Urrutia, P.; Rodríguez-Colinas, B.; Fernandez-Arrojo, L.;Ballesteros, A.; Wilson, L.; Illanes, A.; Plou, F. A detailed analysis ofgalactooligosaccharides synthesis with β-galactosidase from Aspergillusoryzae. J. Agric. Food Chem. 2013, 61, 1081−1087.(17) Mozaffar, Z.; Nakanishi, K.; Matsuno, R.; Kamikuba, T.Production and properties of β-galactosidases from Bacillus circulans.Agric. Biol. Chem. 1984, 48, 3053−3061.(18) Vetere, A.; Paoletti, S. Separation and characterization of threeβ-galactosidases from Bacillus circulans. Biochim. Biophys. Acta 1998,1380, 223−232.(19) Song, J.; Abe, K.; Imanaka, H.; Imamura, K.; Minoda, M.;Yamaguchi, S.; Nakanishi, K. Causes of the production of multipleforms of β-galactosidase by Bacillus circulans. Biosci., Biotechnol.,Biochem. 2011, 75, 268−271.(20) Warmerdam, A.; Paudel, E.; Jia, W.; Boom, R. M.; Janssen, A. E.M. Characterization of β-galactosidase isoforms from Bacillus circulans

and their contribution to GOS production. Appl. Biochem. Biotechnol.2013, 170, 340−358.(21) Stepankova, V.; Bidmanova, S.; Koudelakova, T.; Prokop, Z.;Chaloupkova, R.; Damborsky, J. Strategies for stabilization of enzymesin organic solvents. ACS Catal. 2013, 3, 2823−2836.(22) Verma, M. L.; Barrow, C. J.; Puri, M. Nanobiotechnology as anovel paradigm for enzyme immobilisation and stabilisation withpotential applications in biodiesel production. Appl. Microbiol.Biotechnol. 2013, 97, 23−39.(23) Rodrigues, R. C.; Ortiz, C.; Berenguer-Murcia, A.; Torres, R.;Fernandez-Lafuente, R. Modifying enzyme activity and selectivity byimmobilization. Chem. Soc. Rev. 2013, 42, 6290−6307.(24) Hernandez, K.; Fernandez-Lafuente, R. Control of proteinimmobilization: coupling immobilization and site-directed mutagenesisto improve biocatalyst or biosensor performance. Enzyme Microb.Technol. 2011, 48, 107−122.(25) Garcia-Galan, C.; Berenguer-Murcia, A.; Fernandez-Lafuente,R.; Rodrigues, R. C. Potential of different enzyme immobilizationstrategies to improve enzyme performance. Adv. Synth. Catal. 2011,353, 2885−2904.(26) Hanefeld, U.; Gardossi, L.; Magner, E. Understanding enzymeimmobilisation. Chem. Soc. Rev. 2008, 38, 453−468.(27) Krajewska, B. Application of chitin- and chitosan-based materialsfor enzyme immobilization: a review. Enzyme Microb. Technol. 2004,35, 126−139.(28) Wang, K. Enzyme immobilization on chitosan-based supports.In Chitosan-Based Hydrogels; Yao, K., Yao, J., Li, F., Yin, Y., Eds.;Functions and Applications; CRC Press: Boca Raton, FL, USA, 2012;pp 339−406.(29) Gorecka, E.; Jastrzebska, M. Immobilization techniques andbiopolymer carriers. Food Sci. 2011, 75, 65−68.(30) Barbosa, O.; Ortiz, C.; Berenguer-Murcia, A.; Torres, R.;Rodrigues, R. C.; Fernandez-Lafuente, R. Glutaraldehyde in bio-catalysts design: a useful crosslinker and a versatile tool in enzymeimmobilization. RSC Adv. 2014, 4, 1583−1600.(31) Migneault, I.; Dartiguenave, C.; Bertrand, M. J.; Waldron, K. C.Glutaraldehyde: behavior in aqueous solution, reaction with proteins,and application to enzyme crosslinking. BioTechniques 2004, 37, 790−802.(32) Singh, A. N.; Singh, S.; Suthar, N.; Dubey, V. K. Glutaraldehyde-activated chitosan matrix for immobilization of a novel cysteineprotease, procerain B. J. Agric. Food Chem. 2011, 59, 6256−6262.(33) Manrich, A.; Galvao, C. M. A.; Jesus, C. D. F.; Giordano, R. C.;Giordano, R. L. C. Immobilization of trypsin on chitosan gels: use ofdifferent activation protocols and comparison with other supports. J.Biol. Macromol. 2008, 43, 54−61.(34) Cestari, A. R.; Vieira, E. F.; Mota, J. A. The removal of ananionic red dye from aqueous solutions using chitosan beads − therole of experimental factors on adsorption using a full factorial design.J. Hazard. Mater. 2008, 160, 337−343.(35) Vieira, R. S.; Beppu, M. M. Chitosan as adsorbent for heavymetal ions: performance and adsorption mechanism. In WaterResources Research Progress; Robinson, L. N., Ed.; Nova SciencePublishers: New York, 2008; pp 149−175.(36) Feng, Y.; Yang, L.; Li, F. A novel sensing platform based onperiodate-oxidized chitosan. Anal. Methods. 2010, 2011−2016.(37) Mateo, C.; Palomo, J. M.; Fernandez-Lorente, G.; Guisan, J. M.;Fernandez-Lafuente, R. Improvement of enzyme activity, stability andselectivity via immobilization techniques. Enzyme Microb. Technol.2007, 40, 1451−1463.(38) Rodrigues, D. S.; Mendes, A. A.; Adriano, W. S.; Goncalves, L.R. B.; Giordano, R. L. C. Multipoint covalent immobilization ofmicrobial lipase on chitosan and agarose activated by differentmethods. J. Mol. Catal. B−Enzym. 2008, 51, 100−109.(39) Urrutia, P.; Bernal, C.; Escobar, S.; Santa, C.; Mesa, M.; Wilson,L.; Illanes, A. Influence of chitosan derivatization on its physicochem-ical characteristics and its use as enzyme support. J. Appl. Polym. Sci.2014, DOI: 10.1002/APP.40701.

Journal of Agricultural and Food Chemistry Article

dx.doi.org/10.1021/jf500351j | J. Agric. Food Chem. XXXX, XXX, XXX−XXXI

(40) Bradford, M. M. A rapid and sensitive method for thequantitation of microgram quantities of protein utilizing the principleof protein-dye binding. Anal. Biochem. 1976, 72, 248−254.(41) Urrutia, P.; Mateo, C.; Guisan, J. M.; Wilson, L.; Illanes, A.Immobilization of Bacillus circulans β-galactosidase and its applicationin the synthesis of galacto-oligosaccharides under repeated-batchoperation. Biochem. Eng. J. 2013, 77, 41−48.(42) Huerta, L. M.; Vera, C.; Guerrero, C.; Wilson, L.; Illanes, A.Synthesis of galacto-oligosaccharides at very high lactose concen-trations with immobilized β-galactosidases from Aspergillus oryzae.Process Biochem. 2011, 46, 245−252.(43) Bernal, C.; Sierra, L.; Mesa, M. Improvement of thermal stabilityof β-galactosidase from Bacillus circulans by multipoint covalentimmobilization in hierarchical macro-mesoporous silica. J. Mol. Catal.B−Enzym. 2012, 84, 166−172.(44) Bernal, C.; Mesa, M.; Sierra, L. Design of β-galactosidase/silicabiocatalysts: impact of the enzyme properties and immobilizationpathways on their catalytic performance. Eng. Life Sci. 2014, 14, 85−94.(45) Pedroche, J.; Yust, M.; Mateo, C.; Fernandez-Lafuente, R.;Giron-Calle, J.; Alaiz, M.; Vioque, J.; Guisan, J. M.; Millan, F. Effect ofsupport and experimental conditions in the intensity of the multipointcovalent attachment of proteins on glyoxyl-agarose supports:correlation between enzyme-support linkages and thermal stability.Enzyme Microb. Technol. 2007, 40, 1160−1166.(46) Torres, P.; Batista-Viera, F. Improved biocatalysts based onBacillus circulans β-galactosidase immobilized onto epoxy-activatedacrylic supports: applications in whey processing. J. Mol. Catal. B−Enzym. 2012, 83, 57−64.(47) Brena, B.; Batista-Viera, F. Immobilization of cnzymes: aliterature survey. In Immobilization of Enzymes and Cells; Guisan, J. M.,Ed.; Humana Press: Totowa, NJ, USA, 2006; pp 15−30.(48) Sheu, D. C.; Li, S. Y.; Duan, K. J.; Chen, C. W. Production ofgalactooligosaccharides by β-galactosidase immobilized on glutaralde-hyde-treated chitosan beads. Biotechnol. Technol. 1998, 12, 273−276.

Journal of Agricultural and Food Chemistry Article

dx.doi.org/10.1021/jf500351j | J. Agric. Food Chem. XXXX, XXX, XXX−XXXJ