A

aapvcwis©

K

1

dsthnsu

sscta

h

1d

Biochemical Engineering Journal 34 (2007) 13–19

Influence of the functional activating agent on the biochemical andkinetic properties of Candida rugosa lipase immobilized

on chemically modified cellulignin

Victor H. Perez ∗, Grazielle S. da Silva, Fabrıcio M. Gomes, Heizir F. de CastroEngineering School of Lorena, University of Sao Paulo, PO Box 116, 12600-970 Lorena, SP, Brazil

Received 8 December 2005; received in revised form 8 November 2006; accepted 13 November 2006

bstract

Candida rugosa lipase was immobilized by covalent binding on wood cellulignin (Eucaliptus grandis) chemically modified with differentctivating agents as carbonyldiimidazole, glutaraldehyde and sodium metaperiodate. The resulted immobilized derivatives were evaluated in bothqueous (hydrolysis) and organic (ester synthesis) media. In aqueous media a comparative study between free and immobilized derivatives wasrovided in terms of pH, temperature and kinetic constants (Vmax and Km) following the hydrolysis of p-nitrophenyl palmitate, in which new optimaalues were established. The experimental results suggested that functional activating agents render different interactions between enzyme andellulignin, producing consequently alterations in the optimal reaction conditions. Different behavior was found when the immobilized derivatives

ere tested in organic media, under these conditions similar esterification activities were observed, independent on the agent used to active the

mmobilizing support. Reasons for this are discussed on the light of the interactions among the support, functional activating agent and lipasetructure.

2006 Elsevier B.V. All rights reserved.

Immo

embaf

s[feofp

eywords: Candida rugosa lipase; Cellulignin; Functional activating reagent;

. Introduction

Enzymes are versatile biocatalysts, capable of catalyzingiverse and unique reactions that are highly specific, often stereopecific, in their catalytic mechanisms, enabling simplified stepsoward structurally specific product formation and making themighly desirable for targeted reactions [1]. Advances in biotech-ology, in recent years, have made more efficient generation ofpecific enzymes available, expanding their potential, practicalse in large-scale conversion of chemicals and materials [2–4].

Enzymes immobilized by water-insoluble supports canerve as reusable and removable catalysts, which often pos-ess improved storage and operational stability [5]. Both

hemical and physical methods have been developed forhe purpose of immobilizing enzymes. Enzymes can bedsorbed onto inert solids, ion-exchange resins, or physically

∗ Corresponding author. Tel.: +55 12 31595149.E-mail addresses: victorh@dequi.faenquil.br (V.H. Perez),

eizir@dequi.faenquil.br (H.F. de Castro).

a

tooc–s

369-703X/$ – see front matter © 2006 Elsevier B.V. All rights reserved.oi:10.1016/j.bej.2006.11.012

bilization

ntrapped/encapsulated in solids, such as crosslinked gels,icrocapsules, and hollow fibers. Enzymes can be covalently

onded to solids via various chemical bonding methods, suchs cross linking, multi-functional reagents, or surface reactiveunctional groups [5,6].

Among these methods, chemical covalent bonds offer thetrongest links, and thus the most stable enzyme-solid complexes7]. To chemically bond enzymes to a solid, the structures andunctions of both the enzymes and the solids should be consid-red. It is of utmost importance to consider the functional groupsn the enzyme proteins through which the covalent bonds areormed and the physical and chemical characteristic of the sup-ort material onto which chemically reactive groups are to bettached [7,8].

Enzymes are protein molecules with wide ranging composi-ions and structural complexities. Enzyme proteins may consistf more than 20 types of amino acids. The functional groups

n enzyme proteins that can be utilized, in principle, for theovalent binding include amino –NH2 (lysine), carboxylic acidCOOH (aspartic, glutamic) and hydroxyl –OH (serine, tyro-ine) and cysteine groups. These reactive functional groups on

1 ngine

tse

fcisTthsts

cmlcwlmfOt

bispifmes

eocfdibd

2

2

p(fefla

RiCda(acasp

2

(abt1a[d

2

msbbsba1w

[zwDawewaif

2

a

4 V.H. Perez et al. / Biochemical E

he proteins, when targeted for covalent bonding attachment toolids, should be nonessential for the catalytic activity of thenzymes [7].

The characteristics of solid supports that are desirableor biomolecular attachment include large surface area, goodhemical, mechanical and thermal stability, hydrophilicity andnsolubility. Nonporous materials possess no diffusion con-traints, but have very low surface areas for protein binding.he high surface areas of porous materials provide higher pro-

ein loading capacity. If most of the surfaces are internal surfaces,owever, inefficient diffusion of solutions and significant pres-ure drop can present major drawbacks. With porous solids,herefore, pore structures must be engineered for efficient diffu-ion of solutions and minimal pressure drop [5].

The versatile chemical compositions and physical propertiesoupled with widely available structural forms of polymers haveade them excellent candidates as supports for enzyme immobi-

ization. Natural polymers including polysaccharides (cellulose,ellulose derivatives, dextran and chitonsan) and proteins asell as synthetic polymers, such as polystyrene and polyacry-

ates, have been studied to immobilize enzymes [8–11]. Withost polymers, the major barrier is the lack of highly reactive

unctional groups on the surfaces for direct covalent bonding.ften, surface modification and reactions are needed to fulfill

his particular task.Recently, we have proposed, an alternative matrix for immo-

ilizing catalysts, a product designed as cellulignin [13], whichs obtained from biomass acidic prehydrolysis carried out in ateel reactor lined with titanium metal [14]. Cellulignin has higholymerization degree (35% lignin + 65% cellulose) and due tots physical and chemical properties, such as porosity and sur-ace area, showed compatible affinity to be used as immobilizingatrix for lipases, especially for applications under non-aqueous

nvironment [13]. The specific lipase used was from microbialource (Candida rugosa).

Enzyme immobilization is often accompanied by changes innzymatic activity, optimum pH, and affinity to substrate, amongthers. The extent of these changes depends on the enzyme andarrier support and on the immobilization conditions. There-ore, the current focus of this work is to access the influence ofifferent approaches to activate cellulignin surfaces for optimiz-ng the formation of solid-supported catalysts and compared theiochemical and kinetics properties of the obtained immobilizederivatives in relation to the free enzyme.

. Materials and methods

.1. Materials

Commercial C. rugosa lipase (Type VII) and p-nitrophenylalmitate (p-NPP) were purchased from Sigma Chemical Co.St. Louis, MO, USA). The lipase was supplied in lyophilizedorm with a declared activity of 974 U mg−1 solid using Sigma

mulsified olive oil as substrate. This lipase is substantiallyree of protease, and contains lactose as an extender. Polyethy-ene glycol-1500 (Reagen, SP, Brazil) was used as stabilizinggent of the enzyme. Gum arabic and Triton X-100 came from

rapT

ering Journal 34 (2007) 13–19

eagen (SP, Brazil). Glutaraldehyde (25% solution), carbonyldi-midazole and dimethyl sulfoxide (DMSO) were from Aldrichhemical Co. (Milwaukee, WI, USA) and sodium metaperio-ate was from Nuclear (Sao Paulo, SP, Brazil). Solvents suchs hexane, heptane and 2-propanol were purchased from SynthSao Paulo, SP, Brazil). Heptane was dried with metallic sodiumnd used as solvent for all experiments. Substrates for esterifi-ation reactions (n-butanol and butyric acid came from Merck)nd were dehydrated, with 0.32 cm molecular sieves (aluminumodium silicate, type 13 X-BHD Chemicals, Toronto, Canada),reviously activated in an oven at 350 ◦C for 6 h.

.2. Support

Wood cellulignin from Eucalyptus grandis with 3.5%w/w) moisture content was kindly supplied by RM Materi-is Refratarios Ltda. (Sao Paulo, Brazil) in the form of a darkrown powder, having the following properties: porous struc-ure, 35% of lignin; 65% of cellulose, medium particle diameter.161 mm; density of 0.35 g cm−3. Due to its acid characteristic,limiting factor to be used as a support for immobilizing lipases

15], cellulignin was initially neutralized according to procedureescribed by Gomes et al. [13].

.3. Support activation procedures

Cellulignin activation with glutaraldehyde was based on theethodology described by de Castro et al. [16]. Initially, the

upport was submitted to vacuum for 10 min. Under vacuum,uffered glutaraldehyde solution (2.5%, v/v, 0.1 M, phosphateuffer pH 8.0) was slowly added in order to reach a completeolid immersion. Then, the material was transferred to a 100 mLeaker and 4.6 mL of glutaraldehyde solution (2.5%, v/v) wasdded. The reaction was carried out at room temperature forh. The activated support was filtered and washed with distilledater to eliminate excess glutaraldehyde.The methodology described by Carneiro-da-Cunha et al.

10] was used for cellulignin activation with carbonyldiimida-ole and sodium metaperiodate. Support sample (4 g, dry wt)as immersed in a solution containing carbonyldiimidazole inMSO (20 mg mL−1) in closed flasks for 2 h at room temper-

ture. Afterward, the support was thoroughly washed with aater:DMSO (1:1) solution and then with water to eliminate

xcess carbonyldiimidazole. While, for cellulignin activationith sodium metaperiodate, the support sample was immersed insodium metaperiodate solution 0.5 M under agitation for 1.5 h

n dark place. Later, the active support was transfer to Buchnerunnel and washed with distilled water until neutral pH.

.4. Immobilization procedure

Lipase was immobilized by covalent binding in celluligninctivated with glutaraldehyde, carbonyldiimidazole and metape-

iodate in the presence of polyethylene glycol (PEG, MM 1500)s stabilizing agent [13]. Cellulignin samples (4 g, dry wt) werereviously soaked in hexane under agitation (100 rpm) for 1 h.he excess solvent was discharged, and the amount of enzyme

V.H. Perez et al. / Biochemical Engine

Table 1Hydrolytic activities and activity coupling yield of the immobilized derivativesobtained with cellulignin activated with different chemical agents

Activation agent Hydrolyticactivity (U mg−1)

Coupling yield (%)

Glutaraldehyde 154.67 50.49Carbonyldiimidazole 193.27 62.58S

L

nwpw(taphtt[m

2

atv(oopo7telupoep

2

(lmadCAt

Ftnprl

2a

acpotpw

ip3srEart

3

datpiCo

lsFidiormwenT

odium metaperiodate 216.97 70.00

ipase loading: 306 U g−1 support.

ecessary to give a lipase loading of 300 U mg−1 of dry supportas dissolved in 20 mL of distilled water and added to the sup-ort under low stirring for 2 h at room temperature. PEG-1500as added together with the enzyme solution at a fixed amount

5 mg g−1 of support). Next, 10 mL of hexane was added tohe enzyme-support mixtures and coupling took place overnightt 4 ◦C. Immobilized derivatives were filtered (Whatman filteraper 41) and thoroughly rinsed with hexane. Analyses of theydrolytic activities carried out on initial and spent lipase solu-ions and immobilized preparations showed lipase recovery onhe support on the range from 50 to 70% as shown in Table 113]. The highest coupling yield was obtained when sodiumetaperiodate was used as functional activating agent.

.5. Activity assay of the free and immobilized lipase

The hydrolytic activity was measured with emulsified p-NPPccording to Kordel et al. [17]. One volume of a 16.5 mM solu-ion of p-NPP in 2-propanol was a mixed just before use with 9olumes of 100 mM phosphate buffer pH 7.0 containing 0.4%w/v) Triton X-100 and 0.1% (w/v) gum Arabic. Then, 2.7 mLf this mixture was pre-equilibrated at 37 ◦C in a 1 mL cuvettef an UV-vis spectrophotometer (Varian UV-Carry, Varian Cor-oration). The reaction was started by addition of either 0.3 mLf enzyme solution (4 mg mL−1) in 50 mM phosphate buffer pH.0 or 25–50 mg of the immobilized derivative. The variation ofhe absorbance at 410 nm of the assay against a blank withoutnzyme was monitored for 2–5 min. Reaction rate was calcu-ated from the slope of the absorbance curve versus time bysing a molar extinction coefficient of 13 × 106 cm2 mol−1 for-nitrophenol. This value was determined from the absorbancef standard solutions of p-NPP in the reaction mixture. Onenzyme unit was the amount of enzyme liberating 1 �mol of-nitrophenol per minute in the above conditions.

.6. Esterification reaction

Reaction systems consisted of heptane (20 mL), n-butanol250 mM), butyric acid (280 mM) and either free or immobi-ized lipase derivatives (500 U of activity/mL of substrate). The

ixture was incubated at 37 ◦C for 24 h with continuous shakingt 150 rpm. The consumed butanol and the formed product were

etermined by gas chromatography using a 6 ft 5% DEGS onhromosorb W HP, 80/10 mesh column (Hewlett Packard, Palolto, CA) and hexanol as an internal standard. Water concen-

rations in liquid and solid phases were measured by the Karl

otoi

ering Journal 34 (2007) 13–19 15

ischer method (Mettler DL 18). Fatty acid concentrations wereitrated with 0.02 M potassium hydroxide solutions with phe-olphthalein as an indicator. The molar conversion (X, %), theroductivity (P, g L−1 butyl butyrate h−1) and initial reactionates (A, �M butyl butyrate min−1 mg−1 catalyst) were calcu-ated based on Foresti and Ferreira [18].

.7. Catalytic properties of free and immobilized lipase inqueous medium

The estimation of free and immobilized hydrolytic activitiest different pH values were carried out with reaction mixturesontaining 16.5 mM solution of p-NPP and 50 mM of sodiumhosphate buffer at pH in the range of 6.5–8.5 at 37 ◦C. The effectf temperature on both lipase activities was determined from 37o 60 ◦C under the assay conditions. Results for the influence ofH and temperature on the lipase activity were plotted in graphshere the maximum activity was taken as 100%.The influence of substrate concentration on hydrolytic activ-

ties was also analyzed for free and immobilized derivatives in-NPP solutions varying from 100 to 1000 �M at pH 7.0 and7 ◦C. Michaelis–Menten constant (Km) as the concentration ofubstrate at which half of the maximum reaction rate (Vmax) iseached was calculated with aided by computational programnzfitter version 1.05 published by Elsevier-Biosoft, 1987. Inll cases, enzyme activity was measured as the initial reactionate (0–5% hydrolysis) to avoid the possible inhibition that mightake place due to the appearance of reaction products.

. Results and discussion

Immobilization of enzymes on charged supports often causesisplacements in the pH activity profile, ascribed to an imbal-nce in the partitioning of H+ and OH− concentrations betweenhe microenvironment of the immobilized enzyme and the bulkhase due to electrostatic interactions with the support. To ver-fy the effect of immobilization upon the intrinsic activity of. rugosa, a set of experiments was carried out in which newptima values were established as shown in Figs. 1–3.

The relative activity of the immobilized derivatives on cel-ulignin activated with glutaraldehyde, carbonyldiimidazole andodium metaperiodate as a function of substrate pH is shown inig. 1. Comparing with the pH profile for the free lipase as

nserted in the small box in Fig. 1, all immobilized derivativesisplayed optimal pH for more alkaline side. When carbonyldi-midazole and glutaraldehyde were used as activating agents, theptima pH values shifted 0.5 and 1.0 U to the alkaline region,espectively. Whereas for cellulignin activated with sodiumetaperiodate a broader pH range was determined (pH 7.0–8.5)ith an optimal value at pH 8.0. Usually for lipases this is an

xpected behavior in the view of the established serine-baseducleophilic attack aided by a histidine working as a base [15].his observation is expected because at least partial opening

f the lid upon immobilization is likely to expose the His athe active site more directly to solution hydrogen ions, and sonly less acidic conditions yield the unprotonated form of themidazolium ring required for a basic behavior [2,13].

16 V.H. Perez et al. / Biochemical Engineering Journal 34 (2007) 13–19

Fig. 1. Effect of pH on the hydrolytic activity (p-nitrophenyl palmitate at 37 ◦C)of Candida rugosa lipase immobilized on chemically modified cellulignin withcarbonyldiimidazole (�), glutaraldehyde (�) and sodium metaperiodate (�).Inserted small box shows the pH profile for the free lipase. Initial activi-twa

fifailme

Fafi(awa

Fig. 3. Effect of the substrate concentration on hydrolytic activity of C. rugosalipase immobilized on cellulignin chemically modified by carbonyldiimidazole(sw

apmtotmo

ies (free lipase: 1800 U mg−1); lipase immobilized on activated celluligninith carbonyldiimidazole (255.83 U mg−1), glutaraldehyde (282.76 U mg−1)

nd sodium metaperiodate (311.66 U mg−1) were defined as 100%.

Fig. 2 shows the dependence of the temperature incubationor free lipase (inserted box) and immobilized derivatives. Max-mum activity for free lipase occurred at 37 ◦C (1800 U mg−1),ollowing by a sharp decline on the activity while the temper-ture incubation was raised. Similar behavior was found for themmobilized derivative prepared in glutaraldehyde active cel-

ulignin; although a slight higher temperature (40 ◦C) to attain aaximum activity was observed. This profile indicated that the

xpected interaction between support and lipase which provided

ig. 2. Effect of the temperature on hydrolytic activity (p-nitrophenyl palmitatet pH 8.0) of C. rugosa lipase immobilized on cellulignin chemically modi-ed by carbonyldiimidazole (�), glutaraldehyde (�) and sodium metaperiodate�). Inserted small box shows the temperature profile for the free lipase. Initialctivities (free lipase: 1800 U mg−1; lipase immobilized on activated celluligninith carbonyldiimidazole (207.24 U mg−1), glutaraldehyde (154.67 U mg−1)

nd sodium metaperiodate (271.20 U mg−1) were defined as 100%.

pssaa

tiaMocarci

lAsdatcrl

�), glutaraldehyde (�) and sodium metaperiodate (�). Inserted small boxhows the concentration profile for the free lipase. Reactions were carried outith p-NPP emulsion at different concentrations, pH 7.0 and 37 ◦C.

more rigid external backbone for lipase molecules did not takelace [2], 1996. However, both carbonyldiimidazole and sodiumetaperiodate active cellulignin provided more rigid interac-

ions that became less notorious the effect of raising temperaturen the catalytically active structure of the lipase. Accordingo data showed in Fig. 2, upon lipase immobilization on these

odified supports a shift toward higher temperature (50 ◦C) wasbserved.

To determine the kinetic parameters, hydrolysis reactions of-nitrophenyl palmitate were carried out at 37 ◦C and pH 8.0 atubstrate concentration ranging from 100 to 1000 �M. Fig. 3hows the influence of substrate concentration on hydrolyticctivities for free and immobilized lipases and values for Kmnd Vmax are showed in Table 2.

Results showed that an increase on the substrate concen-ration (p-NPP) from 100 to 600 �M, promoted a significantncremental on the hydrolytic activity values for both freend immobilized lipases. The enzymatic reaction obeys theichaelis–Menten equation and no product inhibition was

bserved. With a subsequent increase on the substrate con-entration, the activity became essentially independent andpproached asymptotically to a constant rate (Vmax). A slighteduction on the hydrolytic activity was verified for substrateoncentrations higher than 800 �M, suggesting a substratenhibition.

The kinetic parameters Km and Vmax, determined for freeipase were 0.170 mM and 1928.66 U mg−1, respectively.ccording to these results, the apparent affinity toward the

ubstrate for the immobilized derivatives changed and this wasepended on the kind of functional chemical reagent used toctive the cellulignin. The experimental results suggested that

hese reagents modified the interactions between enzyme andellulignin, producing consequently alterations in the reactionate (Table 2). From Km and Vmax values it is possible to calcu-ate the ratio of initial rate, i.e. (Vmax/Km) for all immobilized

V.H. Perez et al. / Biochemical Engineering Journal 34 (2007) 13–19 17

Table 2Biochemical and kinetic parameters for free Candida rugosa lipase and immobilized in cellulignin activated with carbonyldiimidazole (CDI), glutaraldehyde (GA)and sodium metaperiodate (META), determined by enzymatic hydrolysis of p-nitrophenyl palmitate

Parameter Free lipase Immobilized lipase on activated cellulignin

CDI GA META

Optimum pH 7.0 7.5–8.0 8.0 7.0–8.5Optimum temperature (◦C) 37 50 40 50Km (mM)a 0.170 0.274 0.136 0.158Vmax (U mg−1)a 1928.66 162.34 148.86 250.64Vmax/Km 11345.06 592.48 1094 1586.32

a Estimated parameters by computational program Enzfitter version 1.05 published by Elsevier-Biosoft, 1987.

F ctivat(

dcmcw

aobbibafm

octas

tgroups of the activating agents were successfully inserted withinthe cellulignin. This also provided an appropriate conditions forimmobilizing C. rugosa lipase in the active surface supports asconfirmed by the bands between 1000 and 1200 cm−1 (Fig. 6).

ig. 4. Possible mechanism reactions for cellulignin with different functional ac) glutaraldehyde.

erivatives. The initial rate for the lipase immobilized onellulignin activated with either glutaraldehyde and sodiumetaperiodate was similar while for the lipase immobilized in

ellulignin activated with carbonyldiimidazole the initial rateas almost the three times lower than the other two cases.Some insight into the reason for the influence of the functional

ctivating reagent on the biochemical and kinetic propertiesf C. rugosa lipase immobilized on modified cellulignin maye gained by considering the possible interaction mechanismsetween the activated supports and C. rugosa lipase as shownn Fig. 4(a)–(c). For activated sodium metaperiodate and car-onyldiimidazole cellulignin (Fig. 4(a) and (b)) the interactionsre straight and sustained by published data [2,19], howeveror glutaraldehyde, which displays a complex chemistry theechanism is represented in a generic manner (Fig. 4(c)).Nevertheless, comparing the infrared spectroscopy spectra

btained for pure and activated cellulignin samples (Fig. 5)

hemical alterations are observed for all active cellulignin. Spec-ra for activated cellulignin samples exhibited two bands at 1500nd 2800 cm−1, which were not present in the pure celluligninample. These bands can be attributed to the axial deformation of

Fl

ing chemical reagents: (a) carbonyldiimidazole, (b) sodium metaperiodate and

he carbonyl and hydroxyl groups, suggesting that the functional

ig. 5. FTIR spectra for pure cellulignin (inserted small box) and active cel-ulignin with carbonyldiimidazole, glutaraldehyde and sodium metaperiodate.

18 V.H. Perez et al. / Biochemical Engineering Journal 34 (2007) 13–19

Fds

afilfst

cttahhatts(d

ineHcehhigccdwiocac

Fig. 7. Synthesis of butyl butyrate from substrate containing 0.25 M butanol and0os1

rtipia

nadtsaogsr

(lgstca0d(iId

−1

ig. 6. FTIR spectra for the free lipase (inserted small box) and immobilizederivatives in activate cellulignin with glutaraldehyde, carbonyldiimidazole andodium metaperiodate.

In relation to the carbonyldiimidazole (CDI), as a functionalctivating agent, in presence of aprotic solvents, such as DMSO,rst the imidazolyl carbamate complex is formed with the cel-

ulose hydroxyl group, which may be displaced by binding theree amino group of lipase. The reaction is an N-nucleophilicubstitution, resulted in a stable N-alkylcarbamate linkage ofhe ligand to the cellulose under working conditions [20].

On the other hand, the sodium metaperiodate oxidation ofellulose is a complex process because higher reagent concen-rations should be necessary to access into the inner region ofhe cellulose. This reaction is highly specific, cleaved the C2nd C3 bond to open the glucose molecules and converting theydroxyl groups in two aldehydic groups, resulting in dialde-yde cellulose. The lipase was then covalently linked to theldehyde groups, in similar form as discussed above. The exis-ence of an imidazolyl carbamate and aldehyde complexes withhe cellulose was confirmed by infrared analysis because bothpectra present carbonyl group absorption in the same region1690–1760 cm−1). In addition, the aldehyde group presents aistention band of formyl group at approximately 2720 cm−1.

For glutaraldehyde, the interaction mechanism with celluloses unclear. It is a complex mechanism since the cellulose wasot previously functionalized with reactive functional groups,.g. amino group, capable of reacting with aldehyde groups.owever, the chemical reaction between glutaraldehyde and

ellulignin matrix was also confirmed from infrared spectralvidence. According to Walt and Agayn [21] several studiesave shown that commercial aqueous solution of glutaralde-yde (25 or 70%) represent multicomponent mixtures whose pHs 3.1, probably due to the partial oxidation of some aldehyderoups to carboxylic acids. Under this condition, glutaraldehydeould cause otherwise unspecific binding effects, remain afteroupling with cellulose. On the other hand, under alkaline con-itions as used in this work, the cellulose chemical activationith glutaraldehyde cause glutaraldehyde to undergo rapidly

ntermolecular aldol condensation, resulting in �-�-unsaturated

ligomeric aldehyde which the internal aldehyde groups exist inonjugation with carbon double bonds and Schiff bases formedt this position. If glutaraldehyde, on the one side interacts withellulose and for another reacts with the lipase amino groups,

Ubat

.28 M butyric acid in heptane using free lipase (©) and lipase immobilizedn activated cellulignin with carbonyldiimidazole (�), glutaraldehyde (�) andodium metaperiodate (�). Reactions were carried out at 37 ◦C with shaking at50 rpm.

esulting in an imine stabilized conjugate. In addition, reticula-ion with polyfunctional agents such as glutaraldehyde insertsntra molecular cross-linkages into one or more regions of therotein molecule. These additional linkages increase the rigid-ty of the protein structure and enhanced the enzyme stabilitygainst several denaturing factors [6].

To verify if the same behavior could be taken place inon-aqueous media, additional information on the catalyticctivity was obtained by testing both free lipase and immobilizederivatives in synthetic applications, that is, in the esterifica-ion reaction of n-butanol with butyric acid. This reaction waselected because it gave measurable results with the greatestccuracy in a short span of time and with a minimum amountf lipase. In addition, this reaction system has been used by ourroup as a standard reaction system for lipases immobilized ineveral supports [12,22–24]. The data plot for initial reactionate is shown in Fig. 7 and results are summarized in Table 3.

Although the initial reaction rate for free lipase1.66 �M min−1 mg−1) was higher than the immobilizedipases, the esterification progress was limited by the waterradually formed that encompassed the enzyme, separatingubstrate and enzyme from each other by an interface. Thus,he reaction stopped before reaching the desired extent ofonversion that is only 40.47% molar conversion was reachedfter 24 h incubation, which corresponded to a productivity.72 g L−1 butyl butyrate h−1. Contrarily for all immobilizederivatives tested, the reaction was driven towards to completionmolar conversion higher than 88%) although a slight lowernitial reaction rates were found in relation to the free lipase.nitial reaction rates varied from 0.58 to 1.16 �M min−1 mg−1

epending on the functional agent used to activate the support.

nder these conditions, productivities (1.30–1.39 g L butylutyrate h−1) were analogous for all immobilized derivativesnd were found to be independent on the functional agent usedo activate the support (Table 3).

V.H. Perez et al. / Biochemical Engineering Journal 34 (2007) 13–19 19

Table 3Values for initial reaction rate, productivity (P) and molar conversion in the synthesis of butyl butyrate by free and immobilized C. rugosa lipase at 37 ◦C

Lipase system Reaction rate(�M min−1 mg−1)

Pa (g L−1 h−1) Molara

conversion (%)

Free C. rugosa 1.66 0.72 40.47

Immobilized lipase on celluligninGlutaraldehyde 0.93 1.38 89.28Carbonyldiimidazole 1.16 1.39 91.12

ewisTpwtt(i

4

lmimio

A

ACn

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

Sodium metaperiodate 0.58

a Calculated with 24 h reaction.

This effect is probably owing to the dispersing effects of thenzyme molecules that facilitate mass transfer when the lipaseas used in the immobilized form. It is also probable that the

mmobilization improved the esterification activity by better pre-erving the native structure of the enzyme in organic media.he results also suggested that the effect of the immobilizationrocedure on the free lipase for using in non-aqueous mediaas exceptionally large, as compared to the effect exhibited on

he hydrolysis. In addition, previously work [13] indicated thathe recycling potential of the lipase immobilized in cellulignint1/2 = 302 h) is favorable compared to data reported for lipasemmobilized on chitosan and chitin [9,23].

. Conclusion

Cellulignin is a novel support and can be used for immobi-izing enzymes, in particular lipases. The results here presented

ay be important to the industrial application of C. rugosa lipasemmobilized in cellulignin in both aqueous and non-aqueous

edia. Therefore, efforts are being direct towards to test themmobilized lipase on cellulignin in typical biotransformationf fat and oils.

cknowledgments

This work was financially supported by Coordenacao deperfeicoamento de Pessoal de Nıvel Superior (CAPES) andonselho Nacional de Desenvolvimento Cientıfico e Tec-ologico (CNPq), Brazil.

eferences

[1] A.L. Paiva, V.M. Balcao, F.X. Malcata, Kinetics and mechanisms of reac-tions catalyzed by immobilized lipases, Enzyme Microb. Technol. 27(2000) 187–204.

[2] V.M. Balcao, A.L. Paiva, F.X. Malcata, Bioreactors with immobilizedlipases: state of the art, Enzyme Microb. Technol. 18 (1996) 392–416.

[3] A. Houde, A. Kademi, D. Leblanc, Lipases and their industrial applications.An overview, Appl. Biochem. Biotechnol. 118 (2004) 155–170.

[4] Z. Liu, R. Weis, A. Gleider, Enzymes from higher eukaryotes for industrialbiocatalysis, Food Technol. Biotechnol. 42 (2004) 237–249.

[5] L. Cao, R.D. Schmid, Carrier-bound immobilized enzymes: principles, in:

Application and Design, John Wiley & Sons, New York, 2005.

[6] L. Gianfreda, M.R. Scarfi, Enzyme stabilization: state of the art, Mol. Cell.Biochem. 100 (1991) 97–128.

[7] J.M. Moreno, M.J. Hernaiz, J.M. Sanchez-Montero, J.V. Sinisterra, M.T.Bustos, M.E. Sanchez, J.F. Bello, Covalent immobilization of pure lipases

[

1.30 88.00

A and B from Candida rugosa, J. Mol. Catal. B: Enzyme 2 (1997) 177–184.

[8] B. Krajewska, Application of chitin and chitosan-based materials forenzyme immobilizations: a review, Enzyme Microb. Technol. 35 (2004)126–139.

[9] E.B. Pereira, H.F. de Castro, F.F. Moraes, G.M. Zanin, Esterification activityand stability of Candida rugosa lipase immobilized into chitosan, Appl.Biochem. Biotechnol. 98–100 (2002) 977–986.

10] M.G. Carneiro-da-Cunha, J.M.S. Rocha, F.A.P. Garcia, M.H. Gil, Lipaseimmobilisation on to polymeric membranes, Biotechnol. Technol. 13(1999) 403–409.

11] S.A. Costa, R.L. Reis, Immobilisation of catalase on the surface ofbiodegradable starch-based polymers as a way to change its surface char-acteristics, J. Mater. Sci. 15 (2004) 335–342.

12] P.C. de Oliveira, G.M. Alves, H.F. de Castro, Immobilisation studiesand catalytic properties of microbial lipase onto styrene–divinylbenzenecopolymer, Biochem. Eng. J. 5 (2000) 63–71.

13] F.M. Gomes, G.S. Santos, R.M. Conte, D.G. Pinatti, H.F. de Castro, Woodcellulignin as an alternative matrix for enzyme immobilization, Appl.Biochem. Biotechnol. 121–124 (2005) 255–268.

14] D.G. Pinatti, A.G. Soares, C.A. Vieira, D. Rodrigues, R.A. Conte, Proceed-ings of the Sixth Brazilian Symposium of Chemistry Lignins and WoodComponents, vol. 1, 2001, p. 26.

15] P. Villeneuve, J.M. Muderhwa, J. Graille, M.J. Haas, Customizing lipasesfor biocatalysis: a survey of chemical, physical and molecular biologicalapproaches, J. Mol. Catal. B: Enzyme 9 (2000) 113–148.

16] H.F. de Castro, R. de Lima, I.C. Roberto, Rice straw as a support for immo-bilization of microbial lipase, Biotechnol. Prog. 17 (2001) 1061–1064.

17] M. Kordel, B. Hofmann, D. Schomburg, R.D. Schmid, Extracellular lipaseof Pseudomonas sp. strain ATCC 21808: purification, characterization,crystallization, and preliminary X-ray diffraction data, J. Bacteriol. 173(1991) 4836–4841.

18] D. Stollner, F.W. Scheller, A. Warsinke, Activation of cellulose membraneswith 1,1-carbonyldiimidazole or 1-cyano-4-dimethylaminopyridiniumtetrafluoroborate as a basis for the development of immunosensors, Anal.Biochem. 304 (2002) 157–165.

19] M.L. Foresti, M.L. Ferreira, Frequent analytical/experimental problems inlipase-mediated synthesis in solvent-free systems and how to avoid them,Anal. Bioanal. Chem. 381 (2005) 1408–1425.

20] M.T.W. Hearn, 1,1-Carbonyldiimidazole-mediated immobilization ofenzymes and affinity ligands, in: Y.C. Lee, R.T. Lee (Eds.), Methods inEnzymology, Academic Press, San Diego, 1987.

21] D.R. Walt, V.I. Agayn, The chemistry of enzyme and protein immobiliza-tion with glutaraldehyde, Trends Anal. Chem. 13 (1994) 425–430.

22] C.M.F. Soares, O.A. dos Santos, H.F. de Castro, F.F. de Moraes, G.M. Zanin,Studies on immobilized lipase in hydrophobic sol–gel, Appl. Biochem.Biotechnol. 113–116 (2004) 307–319.

23] F.M. Gomes, E.B. Pereira, H.F. de Castro, Immobilization of lipase on

chitin and its use in nonconventional biocatalysis, Biomacromolecules 5(2004) 17–23.

24] C.M.F. Soares, H.F. de Castro, F.F. de Moraes, G.M. Zanin, Characteriza-tion and utilization of Candida rugosa lipase immobilized on controlledpore silica, Appl. Biochem. Biotechnol. 77–79 (1999) 745–757.