ap

lerAvense,

Fra

protein was a glycosylated dimer with a molecular weight of approximately 250 kDa. The specic enzy-matic activity of the protein was determined for different substrates: sucrose, inulin, Agave tequilana fruc-tan, levan and Actilight and compared with the activity of Fructozyme. The hydrolysis prole of the

complex structures combining fructose moieties linked byb (2? 1) and b (2? 6) bonds which have a degree of polymeriza-tion ranging from 3 to 29 units (Lopez et al., 2003; Mancilla-Margalli & Lopez, 2006). It has been found in ATF that the proportionof free and polymerized fructose as well as the branching degreechange in function of plant age (Arrizon et al., 2010). ATF are veryimportant for the Mexican agro-industry as they constitute the

lus niger) compared to inulin hydrolysis. They found a lower enzy-matic specicity on ATF (Vmax = 32,10 U/ml, Km = 27 mM) thaninulin (Vmax = 34,10 U/ml, Km = 7.2 mM). This behavior was attrib-uted to the complex structure of ATF with high degree of branchingas well as different ratio of b (2? 1)/b (2? 6) linkages. Therefore,the selection of enzymes with higher specicity on ATF is neces-sary for industrial application. Kluyveromyces sp. yeasts are goodinulinase (EC3.2.1.80) producers (Singh et al., 2007). The commoncharacteristic of these enzymes is a high ratio of enzymatic activityon sucrose versus inulin (S/I ratio). In the case of Agave beverages,Kluyveromyces sp. yeast strains isolated from Aguamiel (Agave sap)

Corresponding author. Tel.: +52 33 33 45 52 00; fax: +52 33 33 45 52 45.E-mail addresses: agschaedler@ciatej.net.mx, gschaedlera@hotmail.com

Bioresource Technology 102 (2011) 32983303

Contents lists availab

T

els(A. Gschaedler).Fructans are plant reserve carbohydrate polymers with fructoseas the repeat unit and a single glucose moiety. Normally they canbe found both in monocotyledons and dicotyledons, and are knownas inulin owing to the fact that they were rst isolated from Inulahelenium. Inulin is a b (2? 1) linear fructan which is built from1-kestose. However, other type of fructans such as those found ingrasses are b (2? 6) macromolecules with building units basedon 6-kestose (Toriz et al., 2007). Some of them are longer b(2? 6) polymers know as levans and are normally produced frombacteria like Erwinia herbicola and Streptococcus salivarum (Stivalaand Khorramian, 1982). In the case of Agave tequilana fructans(ATF) they are based on neo-inulin structure (glucose moiety insidethe fructose chain). They normally are branched polymers with

with incomplete hydrolysis (Waleckx et al., 2008), which resultsin a fructose rich syrup (Pinal et al., 2009) with some Maillardcompounds important in the aroma quality of tequila (Mancilla-Margalli and Lopez, 2002). Nowadays, in order to improve theefciency of the hydrolysis process, autoclaves are used, and morerecently diffusers have replaced the use of autoclaves (Lamas-Robleset al., 2004). The fructose syrup obtained after hydrolysis is fer-mented, distilled, and sometimes aged to obtain the nal product(Pinal et al., 2009). For the optimization of ATF hydrolysis, somestudies have been performed developing enzymatic processes toreduce energy consumption and to increase sugar recovery.Muoz-Gutierrez et al. (2009) have characterized ATF hydrolysiswith Fructozyme (endo and exo-inulinases produced by Aspergill-Accepted 15 October 2010Available online 20 October 2010

Keywords:Kluyveromyces marxianusAgave tequilana fructansFructanaseTequilaMezcal

1. Introduction0960-8524/$ - see front matter 2010 Elsevier Ltd. Adoi:10.1016/j.biortech.2010.10.071different substrates analyzed by HPAEC-PAD showed that the enzyme has different afnities over thesubstrates tested with a sucrose/inulin enzymatic activity ratio (S/I) of 125. For the hydrolysis of Agavetequilana fructans, the enzyme also showed a higher enzymatic activity and specicity than Fructozyme,which is important for its potential application in the tequila industry.

2010 Elsevier Ltd. All rights reserved.

principal raw material for tequila elaboration (Pinal et al., 2009).In the traditional tequila process, ATF are hydrolyzed by cookingReceived 22 June 2010Received in revised form 13 October 2010

A fructanase, produced by a Kluyveromyces marxianus strain isolated during the fermentation step of theelaboration process of Mezcal de Guerrero was puried and biochemically characterized. The activePurication and substrate specicities ofmarxianus isolated from the fermentation

Javier Arrizon a,b, Sandrine Morel b,c,d, Anne GschaedaCentro de Investigacin y Asistencia en Tecnologa y Diseo del Estado de Jalisco, A.C.,bUniversit de Toulouse, INSA, UPS, INP, LISBP, 135 Avenue de Rangueil, F-31077 TouloucCNRS, UMR5504, F-31400 Toulouse, Franced INRA, UMR792 Ingnierie des Systmes Biologiques et des Procds, F-31400 Toulouse,e Institut Universitaire de France, 103 boulevard Saint-Michel, F-75005 Paris, France

a r t i c l e i n f o

Article history:

a b s t r a c t

Bioresource

journal homepage: www.ll rights reserved.fructanase from Kluyveromycesrocess of Mezcala,, Pierre Monsan b,c,d,eida Normalistas # 800, Col. Colinas de la Normal, 44270 Guadalajara, Jalisco, MexicoFrance

nce

le at ScienceDirect

echnology

evier .com/locate /bior tech

Estado de Jalisco (Guadalajara, Mexico) was used. Substrates used

echninclude sucrose (Sigmaaldrich), inulin from Dalhia tubers (Fluka),levan from E. herbicola (Sigmaaldrich), nystose (Fluka) and 1-kestose (Fluka). Actilight is a commercial mixture composed of4.8% w/v of free sugars (sucrose, glucose and fructose), 35.1% w/vof 1-kestose, 53% w/v of 1-nystose and 7.1% w/v of 1-fructofurano-syl-nystose and it was provided by Beghin Meiji (Marckolsheim,France). ATF was provided by BUSTAR Alimentos (Guadalajara,Mexico). Fructozyme was purchased from Novozymes A/S(Denmark).

2.2. Enzyme production

Yeast propagation was carried out by an overnight culture(30 C, 250 rpm) in liquid medium (yeast extract 10 g/l, peptone20 g/l and glucose 20 g/l, pH 5). Then, 2 106 cells were inoculatedin an optimized medium (yeast extract 0.5 g/l, K2HPO4 3 g/l, CaCl2and MnCl2 0.002 M, pH 5.0 and sterilized by 15 min at 121 C),then 50 g/l of ATF was added after ltration (0.45 lm). Fermenta-tion was carried out during 72 h (30 C, 250 rpm) and PMSF (phen-ylmethylsulfonyl uoride) was added (0.001 M) as proteaseinhibitor. The cells were discarded by centrifugation (5000 rpm,30 min, 4 C) and the extract was dialyzed during 24 h at 4 C inTrisHCl (20 mM, pH 7.0), then the dialyzed solution was lyophi-lized and the powder was conserved at 20 C.

2.3. Enzyme puricationand pulque are inulinase-hyper-producing strains (Cruz-Guerreroet al., 1995, 2006) capable of simultaneous production of lactases,pectinases and inulinases (Espinoza et al., 1992). In these studies,inulin was used as substrate for inulinases characterization in Kluy-veromyces sp. strains (Cruz-Guerrero et al., 1995; Zhang et al.,2005; Singh et al., 2007). Other substrates like sucrose have beenextensively used (Cruz-Guerrero et al., 1995; Zhang et al., 2005;Singh et al., 2007; Treichel et al., 2009). To our knowledge onlyone report has been published over the use of a Kluyveromycesmarxianus enzymatic extract on Agave fructo-oligosaccharides toobtain fructose syrup (Garca-Aguirre et al., 2009), but there areno reports about the biochemical characterization of a puriedfructanase from K. marxianus on ATF as substrate. In this study, apuried enzyme from a K. marxianus yeast strain (isolated fromthe fermentation of Mezcal in the State of Guerrero, Mexico) wasbiochemically characterized. As ATF is a complex mixture of shortand long fructose polymers with levan b-(2? 6) and inulin b-(2? 1) type linkages, the hydrolysis was characterized on shortfructooligosaccharides like sucrose and Actilight, and long fruc-tooligosaccharides like inulin and levan. Actilight is a commercialmixture of lowmolecular weight fructooligosaccharides. This is therst time that a fructanase from K. marxianus is evaluated on ATF,Actilight and levan substrates. The hydrolysis of these substratesis useful to study how the different fructose linkage types presentin ATF are hydrolyzed. The same tests were carried out withFructozyme for comparing the enzymatic capacity of the puriedenzyme with industrial enzymes.

2. Methods

2.1. Yeast strain and chemicals

K. marxianus strain belonging to the culture collection of theCentro de Investigacin y Asistencia en Tecnologa y Diseo del

J. Arrizon et al. / Bioresource TThe lyophilized extract (1.89 g) was dissolved in 1.0 l of buffer A(TrisHCl 20 mM, pH 7.8) and concentrated to 30 ml by ultraltra-tion (100 kDa ultraltration lter). It was applied to an anionexchange column (monoQ Sepharose, GE Healthcare, UK) previ-ously equilibrated with 10 column volumes (buffer A). Then thesample was loaded and eluted with buffer B (1 M NaCl inTrisHCl 20 mM, pH 7.8) by three isocratic ows (0.06, 0.15 and1.0 M of NaCl). Protein concentration was followed by UV(280 nm) and the enzyme activity was determined by incubationwith 2% (w/v) of substrate in 100 mM acetate buffer, pH 5.0 at50 C, the rate of glucose and fructose released was determinedas described below. Purication of the protein was veried by anal-ysis of the fractions with a 38% acrilamide gradient gel by SDSPAGE and Coomassie staining. A molecular weight characterizationof the puried protein was carried out with four different treat-ments; protein control sample (1). Monomer separation (2) byheating during 10 min at 65 C. Deglycosylation (3) with Endo Hf(New England Biolabs) at 37 C during 24 h. Comparative control(4) with a commercial enzyme (Fructozyme, Novozymes). The foursamples were charged in two replicates to a NuPAGE 38% acryl-amide gradient gel. Electrophoresis was carried out with TrisAce-tate buffer (150 V during 80 min). Coomassie staining and azymogram according to Gabriel and Wang (1969) were appliedwith triphenyltetrazolium chloride (TTC) as revelator. This zymo-gram was carried on sucrose, inulin and ATF in order to detectactivity on substrates with different fructose linkages.

2.4. Activity assays

Enzymatic activity was determined on different substrates:50 ll of the puried enzyme solution (0.63763.7 mg/l) in100 mM of acetate buffer were mixed with 50 ll of substrate solu-tion (1% w/v in 100 mM of acetate buffer), at pH 5.0 during 15 min,50 C. The reaction was stopped by addition of 100 ll of dinitrosal-icylic acid (DNS) and boiling for 5 min at100 C, then the mixturewas placed on ice. The blank was obtained by inactivation of50 ll of diluted protein solution (0.6376.37 mg/l) with 100 ll ofDNS during 15 min, then 50 ll of substrate solution was added(1% w/v). The absorbance was measured with a microplate readerat 540 nm. One unit of enzyme activity was dened as the amountof enzyme liberating 1 lmol of reducing sugars per minute.

2.5. Determination of optimal temperature, pH and thermal stabilityon different substrates

Optimal pH was determined by measuring initial enzymaticactivity at 50 C in 100 mM acetate buffer adjusted at differentpH (3.0, 3.5, 4.0, 4.5, 5.0, 5.5, 6.0), and 100 mM TrisHCl bufferfor solutions at pH 6.0, 6.5, 7.0, 7.5 and 8.0. For each pH value, sub-strate solutions (1% w/v) were prepared for all the substrates. Opti-mal temperature was determined by measuring enzymatic activityas previously described at a constant pH (5.0) and it was carriedout at 30, 40, 50, 55, 60, 70 and 80 C. For the thermal stability, en-zyme solutions with a substrate concentration of 1% w/v (100 mMbuffer adjusted to optimal pH depending of the substrate) and thereactions were maintained at different constant temperatures (40,50 and 60 C) during 72 h.

2.6. Effect of substrates concentration on fructanase activity

Solutions (100 mM buffer at optimal pH depending on the sub-strate) were prepared as follows: for inulin and levan, 0.25%, 0.5%,1.0%, 1.5% and 2% (w/v) of substrate were used, while for sucrose,Actilight and ATF, 1%, 2%, 3%, 4%, 5% and 10% (w/v) of substratewere used. The enzymatic activity was carried out as described be-fore at optimal temperature depending on substrate. Km (mM) and

ology 102 (2011) 32983303 3299Vmax (U/ml) were calculated from the MichaelisMenten equation.Catalytic constant kcat (s1) was calculated considering the concen-tration of enzyme and molecular weight of the protein (250 kDa).

fructanase and Fructozyme

sisted in a Dionex ASI 100 (Sunnyvalle, CA, USA) with a Shodex101 RI detector. The column was a Biorad HPLC Carbohydrate Anal-

Separation of the products of the hydrolysis reactions was per-formed by HPAEC with pulsed amperometric detection (HPAEC-

As ATF is the substrate of interest for industrial application, the

3. Results and discussion

3.1. Enzyme purication and partial molecular characterization of thefructanase

The fraction eluted at 0.15 M (NaCl in TrisHCl 20 mM) showedthe highest enzymatic activity on ATF and corresponded to thepuried protein. A yield recovery of 51% and a purication factor

3.2. Biochemical characterization

3 3876 78.24 49.5 51 12.3

chnology 102 (2011) 32983303possible effect of ions over the enzymatic activity of the puriedenzyme was evaluated by preparing a solution of ATF (2% w/v in100 mM acetate buffer adjusted to pH 5.0) at 50 C in presence ofEDTA and different ions (Hg2+, Ag+, Cu2+, Cd2+, Ca2+, Mn2+, Mg2+,Fe2+, Fe3+ and Zn2+) at 20 mM, using HgCl2, AgNO3, Cd(NO3)2, CaCl2,PAD, Bio-LC50 system, detector ED40, Dionex, Sunnyvalle, CA,USA) using an analytical CarboPac PA-100 column (4 250 mm;Dionex, Sunnyvalle, CA, USA). The column temperature was35 C, and a sodium acetate gradient in 150 mM NaOH was usedat a ow rate of 1 ml min1. The elution program consisted of6 mM sodium acetate (010 min), 6500 mM (10190 min), and6 mM (190200 min). As standards, inulin, glucose, fructose,sucrose, 1-kestose and nystose were used. The samples of thehydrolysis reactions were diluted in distilled water and ltered(0.45 lm membrane) before injection.

2.10. Effect of ions on A. tequilana fructans hydrolysisysis column: Aminex HPX-87 C column (300 7.8 mm; Biorad,Hercules, CA, USA) at 80 C (elution with degassed ultrapure waterat a ow rate of 0.5 ml min1). As standards, inulin, glucose, fruc-tose and sucrose were utilized. All the samples of substrate hydro-lysis reactions were diluted in distilled water and then ltered(membrane with 0.45 lm) before injection.

2.9. Hydrolysis prole evolution determined by high performanceanionic exchange chromatography with pulsed amperometric detection(HPAEC-PAD)The protein concentration of the puried enzyme andFructozyme corresponding to 63.7 mg/l and 8800 mg/l respec-tively, were determined by the Bradford method (Bradford,1976). Then hydrolysis reactions were performed with solutionsof sucrose, Actilight and ATF (5.0% w/v of substrate in 100 mMacetate buffer), and with inulin and levan solutions (2.0% w/v ofsubstrate in 100 mM acetate buffer). The reactions were performedat optimal pH and temperature depending on the substrate. Ahydrolysis rate of 0.22 U/ml was chosen for performing the reac-tions during 48 h.

2.8. Quantication of hydrolysis products by high performance liquidchromatography (HPLC)

High performance liquid chromatography analysis device con-The values of the molecular weight of A. tequilana fructans, inulinand levan were taken from the reports of Toriz et al. (2007),Haraguchi et al. (2003) and Stivala and Khorramian (1982) respec-tively. The molecular mass of Actilight was calculated from thecomposition reported by Barthomeuf et al. (1997).

2.7. Comparison of enzymatic hydrolysis of different substrates by the

3300 J. Arrizon et al. / Bioresource TeMnCl2, MgCl2, FeSO4, FeCl3 and ZnCl2. A control reaction wascarried out only with ATF in order to evaluate the effect of thepresence of cations.3.2.1. Effect of pH and temperature on enzymatic activity on differentsubstrates

A maximum enzymatic activity occurred at pH 5.0 for inulin,Actilight and sucrose (Table 2). For other K. marxianus, differencesin the values of the optimal pH depending on the substrate andyeast strain were observed (Table 3). The optimal pH for inulinand sucrose observed for this fructanase were similar (Table 2).For Actilight, there is no report on enzymatic hydrolysis with inu-linases from Kluyveromyces sp. The optimal pH was the same asinulin, possibly due to b (2? 1) linkages. In the case of levan,the maximum enzymatic activity was found between 4.5 and 5.0(Table 2) and it is also the rst time reported. For ATF, the enzy-matic activity was constant from pH 4.0 to 6.5. Thus, the effect ofthe enzyme concentration on enzymatic activity on ATF was veri-ed at a pH range from 3.0 to 7.5 (data not shown). Low variationswith a small increase in a pH range from 4.5 to 5.0 were observed(Table 2). The optimal temperature for levan and ATF hydrolysiswas 50 C, while for inulin, Actilight and sucrose it was 55 C(Table 2). As well as for the optimal pH, the optimal temperaturefor other Kluyveromyces fructanases varied in function of strainand substrate (Table 3). For Actilight, levan and ATF, there are

Table 1Fructanase purication.

Step Totalactivity(U)

Totalprotein(mg)

Specicactivity(U/mg)

Yieldrecovery(%)

Puricationfactor

1 7600 1890 4.02 100 12 6460 229.5 28.14 85 7of 12.3 were obtained (Table 1). The enzyme showed in Coomassieblue gel only one glycosylated dispersed band with a molecularweight of approximately 250 kDa (data not shown). It has beenfound that K. marxianus produces glycosylated inulinases with amolecular weight from 200 to 250 kDa (Pessoa and Vitolo, 1997),thus the puried fructanase is within this range. The puried fruc-tanase was separated in two monomers by simple heating with amolecular weight of 125 kDa (data not shown). After deglycosyla-tion, a clearer band with a molecular weight approximately of65 kDa was observed (data not shown). In literature, inulinasesfrom Kluyveromyces sp. yeasts are close to 60 kDa (Wen et al.,2003). The protein showed a typical pink band of fructanase activ-ity in the zymogram (Gabriel and Wang, 1969) at 250 kDa. Then azymogramwas performed in the same gel with different substrates(sucrose, inulin and ATF) and it showed enzymatic activity on thesedifferent fructose linkages. The fructanase lost the activity when itwas deglycosylated (data not shown). It has been found that thedeglycosylation decrease the thermal stability and the specicityas well as tolerance to metal ions of a b-fructofuranosidase fromAureobasidium sp ATCC-20524 (Hayashi et al., 1992, 1994). There-fore, it is possible that glycosylation is important for the enzymaticactivity of the puried fructanase.1: Crude extract. 2: Filtration with membrane 100 kDa. 3: Mono Q Sepharoseanionic column.

Table 2Optimal temperature, pH and thermal stability of K. marxianus fructanase usingdifferent substrates.

Substrate Optimal pH Optimal T(C)

Thermal stability(C)

Levan 4.55 (16.6)a 50 (15.8)a 40 (100)b

Inulin 5 (52.4)a 55 (41.2)a 40 (84.2)b

Agave tequilanafructans

4.55.5(1253.1)a

50 (1406.2)a 40 (98.6)b

Actilight 5 (6000)a 5055(3750)a

40 (92.9)b

Sucrose 5 (7656.2)a 55 (7187.5)a 40 (100)b

ATCC16045 Sucrose 4.75 55 13.4 Kushi et al. (2000)

J. Arrizon et al. / Bioresource TechnInulin 55 17.3NRRL-Y-7571 Sucrose 4.44.8 50 13 Treichel et al.

(2009)CDBB-L-278 Sucrose 5 70 40.2 Cruz-Guerrero et al.

(1995)Inulin 5 50 3.04

YS-1 Inulin 5.5 50 3.4 Singh et al. (2007)KWO Inulin 4.5 55 5.88 Zhang et al. (2005)Puried

fructanaseSucrose 5 55 29.1 This work

Inulin 5 55 7.26 This worka Maximum enzymatic activity of the fructanase (U/mg).b Percentage of residual enzymatic activity (%) at 72 h of reaction.

Table 3Comparison of the optimal pH, temperature and kinetic parameters of differentinulinases from Kluyveromyces sp. yeast strains versus the puried protein.

Yeast strain Substrate pH T(C)

Km(mM)

Referenceno reports of optimal temperature for a fructanase from K. marxi-anus. The optimal temperatures were the same for Actilight andinulin, possibly due to the presence of b (2? 1) linkages, as wellas for levan and ATF, possibly due to the presence of b (2? 6) link-ages. Therefore, it seems that for this fructanase, linkage type couldcorrelate with optimal temperature. Thermal stability of the pro-tein was also determined. The enzyme was stable at 40 C for allthe substrates (Table 2). Even though the different pH and temper-atures tested, the highest and lowest hydrolysis efciency corre-sponds to sucrose and levan respectively (Table 2).

3.2.2. Effect of substrate concentration on the enzymatic activityThe fructanase followed MichaelisMenten kinetics for all the

substrates tested. The regression coefcient of the linear adjust-ment was close to 1.0 (Table 4). As the molecular weight of thesubstrate tested was increased, Km and kcat were decreased (Table4). The values of kcat/Km showed that the catalytic efciency wassimilar for sucrose, Actilight and ATF and was higher than kcat/Km for inulin and levan. It means that, the catalytic efciency is

Table 4Kinetic parameters of the K. marxianus fructanase on different substrates.

Substrate Molecularweight(kDa)

Km (mM) kcat (s1) kcat/Km(s1 mM1)

R2b

Levan 26,570a 1.8 0.21 0.33 0.102 0.18 0.01 0.98Inulin 5000a 7.16 0.34 0.70 0.007 0.1 0.00 0.99A. tequilana

fructans2690a 12.9 0.39 11.7 0.627 0.90 0.03 0.98

Actilight 617.6a 27.8 3.98 30.2 0.817 1.09 0.13 0.99Sucrose 342 29.1 3.47 37.05 1.234 1.27 0.11 0.98

a Molecular weight average.b Regression coefcient of Michaelis Menten linear adjustment.higher for low molecular weight substrate molecules (sucrose)and lower for high molecular weight molecules (levan).Thisbehavior can be explained in terms of the access of substrate tothe active site of the enzyme. The Km values for inulinases fromother K. marxianus showed a high variability for sucrose and inulin(Table 3). The Km values found for the fructanase were within therange reported (Table 4). For ATF, Kmwas 12.9 mM, which was dif-ferent from the value reported by Muoz-Gutierrez et al., (2009)using Fructozyme. In the case of levan and Actilight, this is therst time that they are calculated with a fructanase from K.marxianus.

3.3. Enzymatic hydrolysis prole on different substrates

3.3.1. Comparison of enzymatic activity between the fructanase andFructozyme on different substrates

In general, the fructanase had a higher specic enzymatic activ-ity than Fructozyme on all the substrates (data not shown). Forboth systems, the highest enzymatic activity was observed onsucrose and Actilight (data not shown). In the case of ATF and inu-lin hydrolysis, Fructozyme had 1.7-fold more activity on inulinthan ATF (data not shown), while the fructanase had 20-fold moreactivity on ATF than inulin (data not shown). Therefore, the fruc-tanase has a technological advantage for ATF hydrolysis. Muoz-Gutierrez et al. (2009) have compared the enzymatic hydrolysisof inulin and ATF with Fructozyme. A higher specicity over inu-lin than ATF was also observed. They explain this behavior on thebasis of the complex structure of the branched b (2? 1) and b(2? 6) molecules of ATF. As the K. marxianus yeast strain usedcames from an Agave fermenting step, it could be possible thatby evolution, it has developed specic enzymes for ATF consump-tion. The protein has a S/I ratio of 125, which means that theenzyme acts preferably by an exo-inulinase mechanism (data notshown). In the case of levan, Fructozyme showed a very low enzy-matic activity (data not shown), which was lower for the fructan-ase (data not shown). Normally, the b (2? 6) linkages arepreferentially degraded by levanases, most of them produced bybacteria. For example, Thermotoga maritima has an exo-inulinasecapable of b (2? 1) and b (2? 6) fructose linkages hydrolysis(Muoz-Gutierrez et al., 2009). As can be seen, both types of enzy-matic systems have differences in specicity for the substratestested. For a better characterization of the hydrolysis of sucrose,Actilight, ATF, inulin and levan, an enzymatic hydrolysis was car-ried out with solutions of these substrates at controlled rate(0.22 U/ml) with different dilutions of the fructanase and Fructo-zyme. The prole of substrate hydrolysis was followed by HPLCand HPAEC-PAD. In the case of levan, it was not possible to deter-mine the hydrolysis rate for the concentrated solution due to thevery low enzymatic activity of the fructanase on this substrate.

3.3.2. Comparison of product hydrolysis between the fructanase andFructozyme on different substrates by HPLC and HPAEC-PAD

The hydrolysis rate and the total sugar hydrolyzed (TSH) fol-lowed by HPLC on the different substrates varied in function ofthe enzyme used (Fig. 1). For sucrose, the hydrolysis rate was fasterfor the Fructozyme than for the fructanase with a TSH of 99% and82% respectively (Fig. 1). The same behavior was observed for Acti-light hydrolysis, with a TSH of 97% and 78.8% respectively (Fig. 1).The same behavior was corroborated by HPAEC-PAD (data notshown). Normally, the exo-inulinases have a higher specicity onshort molecules like sucrose (Singh and Gill, 2006). Thus, it couldbe possible that the exo-inulinases from Fructozyme are moreactive than the fructanase. This could explain why the hydrolysis

ology 102 (2011) 32983303 3301rate and TSH were higher on sucrose and Actilight with Fructo-zyme than the fructanase. For inulin hydrolysis followed by HPLC,the rate and the TSH were higher with Fructozyme than with the

follo

chnfructanase with 100% and 74.5% respectively (Fiure 1). This couldbe caused by the synergetic action of endo and exo-inulinases ofthe enzymes mixture of Fructozyme. When the hydrolysis rateof inulin was followed by HPAEC-PAD, for both enzymatic systems

Fig. 1. Comparison of the products of enzymatic hydrolysis on different substratessymbols) and glucose (empty symbols).3302 J. Arrizon et al. / Bioresource Teonly fructose was observed without smaller fructans produced bybreak down of fructose linkages inside the polymer chain (datanot shown). Therefore, it could be possible that the fructanase actsmore like an exo-inulinase enzyme, which is in agreement with theobserved S/I ratio. Conversely, for ATF hydrolysis followed by HPLC,the hydrolysis rate and the TSH were higher for the fructanase thanfor Fructozyme with 80% and 63.8% respectively (Fig. 1). Never-theless for both enzymatic systems only fructose was observedby HPAEC-PAD as the principal hydrolysis product (data notshown). Even though the complex structure of ATF (Lopez et al.,2003; Arrizon et al., 2010), one more time the higher specicityof the puried fructanase over ATF is conrmed. It has been foundthat other K. marxianus yeasts isolated from agave beverages likepulque (Agave fermented beverage) and Aguamiel (Agave sap)are inulinase hyper-producing strains (Cruz-Guerrero et al.,2006). The enzymatic extract of this type of yeast strains has beenused for the elaboration of high-fructose syrup from ATF(Garca-Aguirre et al., 2009). Therefore, Kluyeromyces yeasts arewell adapted to degrade this type of fructans. In the case of levanhydrolysis, only the level of fructose hydrolyzed was followed byHPLC.A very slow hydrolysis rate for Fructozyme was observed, corre-sponding to small quantities of fructose observed (data notshown), while for the fructanase the fructose concentration wasnot detected (data not shown).

3.4. Effect of ions in the hydrolysis of A. tequilana fructans

As ATF is the substrate of industrial interest, a comparison ofthe enzymatic activity with the fructanase was carried out withand without the presence of different ions. It can be observed thata great reduction was observed in the enzymatic activity with Hg2+,Cu2+, Zn2+ and Mn2+ addition, with less than 50% of residualenzymatic activity. The enzymatic activity on ATF was less affectedby Fe2+, Fe3+, Ag+ and Ba2+ with more than 50% of residual enzy-matic activity. Only Ca2+ was capable of increasing the enzymatic

wed by HPLC with K. marxianus fructanase (h) and Fructozyme (s), fructose (full

ology 102 (2011) 32983303activity by 10%. This result was congruent with the reduction ofenzymatic activity in presence of EDTA. It therefore seems thatCa2+ has a signicative role on the enzymatic activity. It can be no-ticed that A. tequilana juice has naturally a high concentration ofCa2+.

4. Conclusions

This is the rst time that a puried fructanase fromK. marxianusis characterized for the enzymatic hydrolysis over short fructooli-gosaccharides (Actilight), ATF and levan. An antagonist correla-tion between substrate specicity and molecular weight ofsubstrates was established for the fructanase. The hydrolysis pro-le analyzed by HPLC and HPAEC-PAD showed that possibly thefructanase acts by an exo-inulinase mechanism, and it has morehydrolysis capacity over ATF than Fructozyme. The fructanase isactivated by Ca2+, which is naturally present in A. tequilana. There-fore, the protein has a higher potential application in the tequilaindustry than Fructozyme.

Acknowledgements

This work has been sponsored by the projects SEP-CONACYT24556 and SAGARPA-CONACYT 109799, Mxico. We thank thecompanies Tequilera el Triangulo, Cd. Guzman, Mxico andBustar Alimentos, Zapopan, Mxico for supplying the vegetalmaterial.

References

Arrizon, G., Morel, S., Gschaedler, A., Monsan, P., 2010. Comparison of the water-soluble carbohydrate composition and fructan structures of Agave tequilanaplants of different ages. Food Chem. 122, 123130.

Barthomeuf, C., Grizard, D., Teulade, J.C., 1997. Assay and structural determinationof fructo-oligosaccharides synthesized by an enzymatic system from Peniciliumrugulosum. Biotechnol. Tech. 11, 845848.

Bradford, M., 1976. A rapid and sensitive method for the quantitation of microgramquantities of protein utilizing the principle of protein-dye binding. Anal.Biochem. 72, 248254.

Cruz-Guerrero, A., Garca-Pea, I., Barzana, E., Garca-Garibay, M., Gomez-Ruiz, L.,1995. Kluyveromyces marxianus CDBB-L-278: a wild inulinase hyperproducingstrain. J. Ferment. Bioeng. 80, 159163.

Cruz-Guerrero, A.E., Olvera, J.L., Garca-Garibay, M., Gmez-Ruiz, L., 2006. Inulinase-hyperproducing strains of Kluyveromyces sp. isolated from aguamiel (Agave sap)and pulque. World J. Microbiol. Biotechnol. 22, 115117.

Espinoza, P., Brzana, E., Garca-Garibay, M., Gmez-Ruiz, L., 1992. Evaluation ofKluyveromyces marxianus for the producion of lactase simultaneously topectinase or inulinase. Biotechnol. Lett. 14, 10531058.

Gabriel, O., Wang, S.F., 1969. Determination of enzymatic activity in polyacrylamidegels. I. Enzymes catalysing the conversion of non-reducing substrates toreducing products. Anal. Biochem. 27, 545554.

Garca-Aguirre, M., Senz-lvaro, V.A., Rodrguez-Soto, M.A., Vicente-Magueyal, F.J.,Botello-lvarez, E., Jmenez-Islas, H., Crdenas-Manriquez, M., Rico-Martnez,R., Navarrete-Bolaos, J.L., 2009. Strategy for biotechnological process designapplied to the enzymatic hydrolysis of Agave fructo-oligosaccharides to obtainfructose-rich syrups. J. Agric. Food Chem. 57, 1020510210.

Haraguchi, K., Yoshida, M., Otsubo, K., 2003. Purication and properties of a heat-stable inulin fructotransferase from Arthrobacter ureafaciens. Biotechnol. Lett.25, 10491053.

Hayashi, S., Shimokawa, Y., Tubouchi, M., Takasaki, Y., Imada, K., 1992. Properties ofthe deglycosylated beta-fructofuranosidase P-2 from Aureobasidium sp ATCC-20524. J. Ind. Microbiol. 10 (34), 191194.

Hayashi, S., Tubouchi, M., Takasaki, Y., Imada, K., 1994. The effect of chemicals onthe activity of deglycosylated aureobasidium beta-fructofuranosidases. Lett.Appl. Microbiol. 18 (2), 105106.

Kushi, R.T., Monti, R., Contiero, J., 2000. Production, purication andcharacterization of an extracellular inulinase from Kluyveromyces marxianusvar. Bulgaricus. J. Ind. Microbiol. Biotechnol. 25, 6369.

Lamas-Robles, R., Sandoval-Fabian, G.C., Osuna-Tenes, A.A., Prado-Ramrez,Gschaedler-Mathis, A.C., 2004. Cocimiento y molienda, in: Gschaedler-Mathis,A.C. (Ed), Ciencia y Tecnologia del Tequila, Guadalajara, pp. 4160.

Lopez, M.G., Mancilla-Margalli, N.A., Mendoza-Diaz, G., 2003. Molecular structuresof fructans from Agave tequilanaWeber var. azul. J. Agric. Food Chem. 51, 78357840.

Mancilla-Margalli, N.A., Lopez, M.G., 2002. Generation of Maillard compounds frominulin during the thermal processing of Agave tequilanaWeber Var. azul. J. Agric.Food Chem. 50, 806812.

Mancilla-Margalli, N.A., Lopez, M.G., 2006. Water-soluble carbohydrates andfructan structure patterns from Agave and Dasylirion species. J. Agric. FoodChem. 54, 78327839.

Muoz-Gutierrez, I., Rodrguez-Alegra, M.E., Lopez-Mungua, A., 2009. Kineticbehavior and specicity of b-fructosidase in the hydrolysis of plant andmicrobial fructans. Process Biochem. 44, 891898.

Pessoa, A., Vitolo, M., 1997. Separation of inulinase from Kluyveromyces marxianususing reversed micellar extraction. Biotechnol. Tech. 11, 421422.

Pinal, L., Cornejo, E., Arellano, M., Herrera, E., Nuez, L., Arrizon, J., Gschaedler, A.,2009. Effect of Agave tequilana age, cultivation eld location and yeast strain ontequila fermentation process. J. Ind. Microbiol. Biotechnol. 36, 655661.

Singh, P., Gill, P.K., 2006. Production of inulinases: recent advances. Food Technol.Biotechnol. 44 (2), 151162.

Singh, R.S., Dhaliwal, R., Puri, M., 2007. Partial purication and characterization ofexoinulinase from Kluyveromyces marxianus YS-1 for preparation of high-frutose syrup. J. Microbiol. Biotechnol. 17, 733738.

Stivala, S.S., Khorramian, B.A., 1982. Assesment of branching in S. salivarum levanfrom small-angle X-ray scattering. Carbohydr. Res. 101, 111.

Toriz, G., Delgado, E., Zuiga, V., 2007. A proposed chemical structure for fructansfrom blue agave plant (Tequilana Weber var. azul). E-Gnosis 5, 15.

Treichel, H., Mazutti, M.A., Filho, F.M., Rodrigues, M.I., 2009. Technical viability ofthe production, partial purication and characterization of inulinase usingpretreated agroindustrial residues. Bioprocess Biosyst. Eng. 39, 425433.

Waleckx, E., Gschaedler, A., Colonna-Ceccaldi, B., Monsan, P., 2008. Hydrolysis offructans from Agave tequilana Weber var. azul during the cooking step in atraditional tequila elaboration process. Food Chem. 108, 4048.

Wen, T., Liu, F., Huo, K., Li, Y.Y., 2003. Cloning and analysis of the inulinase genefrom Klyveromyces cicerisporus CBS4857. World J. Microbiol. Biotechnol. 19,423426.

Zhang, L., Zhao, C., Ohta, W.Y., Wang, Y., 2005. Inhibition of glucose on anexoinulinase from Kluyveromyces marxianus expressed in Pichia pastoris. ProcessBiochem. 40, 15411545.

J. Arrizon et al. / Bioresource Technology 102 (2011) 32983303 3303

Purification and substrate specificities of a fructanase from Kluyveromyces marxianus isolated from the fermentation process of MezcalIntroductionMethodsYeast strain and chemicalsEnzyme productionEnzyme purificationActivity assaysDetermination of optimal temperature, pH and thermal stability on different substratesEffect of substrates concentration on fructanase activityComparison of enzymatic hydrolysis of different substrates by the fructanase and FructozymeQuantification of hydrolysis products by high performance liquid chromatography (HPLC)Hydrolysis profile evolution determined by high performance anionic exchange chromatography with pulsed amperometric detection (HPAEC-PAD)Effect of ions on A. tequilana fructans hydrolysis

Results and discussionEnzyme purification and partial molecular characterization of the fructanaseBiochemical characterizationEffect of pH and temperature on enzymatic activity on different substratesEffect of substrate concentration on the enzymatic activity

Enzymatic hydrolysis profile on different substratesComparison of enzymatic activity between the fructanase and Fructozyme on different substratesComparison of product hydrolysis between the fructanase and Fructozyme on different substrates by HPLC and HPAEC-PAD

Effect of ions in the hydrolysis of A. tequilana fructans

ConclusionsAcknowledgementsReferences