Cf

AOa

b

a

ARR1A

KEHLPTW

1

tofbllcsep

cab

dG

((o

0d

Enzyme and Microbial Technology 49 (2011) 485– 491

Contents lists available at SciVerse ScienceDirect

Enzyme and Microbial Technology

jou rn al h om epage: www.elsev ier .com/ locate /emt

loning of novel cellulases from cellulolytic fungi: Heterologous expression of aamily 5 glycoside hydrolase from Trametes versicolor in Pichia pastoris

lejandro Salinasa, Marcela Vegaa, María Elena Lienqueoa, Alejandro Garciab, Rene Carmonab,riana Salazara,∗

Centre for Biochemical Engineering and Biotechnology, Department of Chemical Engineering and Biotechnology, University of Chile, Beauchef 850, Santiago P.O. 8370448, ChileDepartment of Wood Engineering and their Biomaterials, Faculty of Forest Sciences and Nature Conservation, University of Chile, Santa Rosa 11315, Santiago Box 9206, Chile

r t i c l e i n f o

rticle history:eceived 15 February 2011eceived in revised form5 September 2011ccepted 11 October 2011

a b s t r a c t

Total cDNA isolated from cellulolytic fungi cultured in cellulose was examined for the presence ofsequences encoding for endoglucanases. Novel sequences encoding for glycoside hydrolases (GHs) wereidentified in Fusarium oxysporum, Ganoderma applanatum and Trametes versicolor. The cDNA encodingfor partial sequences of GH family 61 cellulases from F. oxysporum and G. applanatum shares 58 and 68%identity with endoglucanases from Glomerella graminicola and Laccaria bicolor, respectively. A new GHfamily 5 endoglucanase from T. versicolor was also identified. The cDNA encoding for the mature protein

eywords:ndoglucanaseeterologous expressionignocellulose hydrolysisichia pastorisrametes versicolor

was completely sequenced. This enzyme shares 96% identity with Trametes hirsuta endoglucanase and22% with Trichoderma reesei endoglucanase II (EGII). The enzyme, named TvEG, has N-terminal family1 carbohydrate binding module (CBM1). The full length cDNA was cloned into the pPICZ�B vector andexpressed as an active, extracellular enzyme in the methylotrophic yeast Pichia pastoris. Preliminarystudies suggest that T. versicolor could be useful for lignocellulose degradation.

hite rot fungi

. Introduction

The study of microbial cellulases has recovered attention inhe last years, mainly because of the increased interest in theptimization of processes of manufacture of bio-products derivedrom lignocellulosic residues, e.g. bioethanol [1]. The cellulosicioethanol process requires basically three steps: pretreatment to

oosen up the lignin-cellulose fiber entanglement, hydrolysis of cel-ulose and fermentation to ethanol. The cellulose hydrolysis is aomplex and expensive process, due mainly to the lignocellulosetructure and recalcitrance. Developments in the field are still nec-ssary in order to decrease the impact of this stage in the wholerocess’s economy.

Cellulose hydrolysis is carried out mostly by the action of

ellulases, enzymes acting in concert to release soluble sug-rs from cellulose [2]. These enzymes are glycoside hydrolasesreaking �-1,4-linkages in cellulose with different specificities.

Abbreviations: CBM, carbohydrate binding module; cDNA, complementaryeoxyribonucleic acid; CMC, carboxymethyl cellulose; DNS, dinitrosalicylic acid;H, glycoside hydrolase; RT-PCR, reverse transcription polymerase chain reaction.∗ Corresponding author. Tel.: +56 2 9784711; fax: +56 2 6991084.

E-mail addresses: alsalina@ing.uchile.cl (A. Salinas), marcvega@ing.uchile.clM. Vega), mlienque@ing.uchile.cl (M.E. Lienqueo), agarcia@uchile.clA. Garcia), recarmon@uchile.cl (R. Carmona), orsalaza@ing.uchile.cl,riana.orsalaza3@gmail.com (O. Salazar).

141-0229/$ – see front matter © 2011 Elsevier Inc. All rights reserved.oi:10.1016/j.enzmictec.2011.10.003

© 2011 Elsevier Inc. All rights reserved.

Endo-�-1,4-glucanase (EG; EC 3.2.1.4) hydrolyzes internal glyco-sidic bonds in cellulose, while exo-�-1,4-glucanase, or cellobiohy-drolase (CBH; EC 3.2.1.91), releases cellobiose from the reducingor non-reducing ends of the cellulose chain; finally, cellobiase (EC3.2.1.21) hydrolyzes cellobiose to glucose, which can be consumedby fermenting microorganisms. Cellulases are typically multido-main proteins consisting of a catalytic core domain linked to acarbohydrate-binding module (CBM) via a flexible linker region[3]. Catalytic domains of glycoside hydrolases (GHs) are classifiedaccording to their sequence in more than 120 families [4], while61 CBMs families have been described up to now (Carbohydrate-Active enZYmes Database, http://www.cazy.org/) [5]. Now it is wellunderstood that other enzymes are also necessary components tocomplete and efficient hydrolysis of lignocellulose [6].

Cellulases are produced by many different microorganisms innature. Lignocellulolytic enzyme-producing fungi are widespread,and include species from phylum Ascomycota (e.g. Trichoderma ree-sei), Basidiomycota including white-rot fungi (e.g. Phanerochaetechrysosporium), brown-rot fungi (e.g. Fomitopsis palustris) andother anaerobic species. Historically cellulases have been derivedfrom Trichoderma species, due to the great capacity of productionof this fungus. T. reesei produces two major cellobiohydrolases

belonging to GH families 6 and 7 (Cel6A and Cel7A, or CBHI andCBHII, respectively) and at least five endoglucanases classified asGH families 7, 5, 12, 61 and 45 (Cel7B, Cel5A, Cel12A, Cel61A, andCel45A or EGI, EGII, EGIII, EGIV, and EGV) [7]; these enzymes, which

4 crobia

as

nftlaAaevoAbtit

nIatlicusNa

2

2

fipiMwr

2

tNtPRtAa

2

sft13[Ecr

2

nsh

86 A. Salinas et al. / Enzyme and Mi

re produced exclusively when cellulose is available as carbonource [8], act synergistically to hydrolyze the cellulose [9,10].

Investigations addressed the search and characterization ofovel and more efficient cellulases have driven scientists toward

ungi responsible of white rot in woods, specifically basidiomyce-ous fungi [6]. In this group, Trametes versicolor has been studiedargely because its ligninolytic properties, which have potentialpplications in bioremediation of pulp and paper mill effluents.ctivities of selected wood-degrading enzymes such as cellobi-se and cellulase have been monitored in T. versicolor [9,10]. Thexistence of an endoglucanase activity in the supernatant of a T.ersicolor culture was reported some years ago [11] but no cloningr recombinant expression of the enzyme has been informed yet.lso, a cellobiohydrolase-encoding cDNA from T. versicolor haseen cloned and expressed in Aspergillus niger [9]. In summary,he potential of T. versicolor as producer of cellulolytic enzymess widely recognized, but a systematic study of the functionality ofhe T. versicolor cellulases is still incomplete.

The aim of this study was the identification and cloning ofovel endoglucanase genes from native Chilean cellulolytic fungi.

n particular, a new T. versicolor endoglucanase was expressed as recombinant protein in Pichia pastoris, a first step addressed tohe enzyme overproduction and future characterization of the cel-ulolytic properties of this enzyme. The long term purpose is thesolation of cellulases having efficient hydrolytic action on woodhips of trees that are abundant in Chile. In this work we havesed Nothofagus pumilio (Lenga beech or Lenga) as lignocellulosicubstrate. Lenga, a native hardwood tree belonging to the familyothofagaceae, comprises about 25% of native forest surface and isn important source of lignocellulosic waste in Chile.

. Materials and methods

.1. Nucleic acid manipulations

DNA manipulations were carried out as described [12]. PCR products were puri-ed from agarose gel after electrophoresis using the QIAEXII kit (QIAGEN). PCRroducts were cloned into the pGEM-T Easy vector (Promega) and transformed

nto Escherichia coli DH5� cells. Plasmid DNA was isolated using the QIAprep Spininiprep kit (QIAGEN). Primers, restriction enzymes and Taq DNA polymeraseere supplied by Integrated DNA Technologies, New England Biolabs and Promega,

espectively.

.2. Fungal growth conditions

A battery of seven Chilean native cellulolytic fungi was supplied by the Labora-ory of Biodegradation and Wood Preservation of the Faculty of Forest Sciences andature Conservation of the University of Chile. Cultures of T. versicolor, G. applana-

um, Poria placenta, Fusarium oxysporum, Pleurotus ostreatus, Lentinus edodes andeniophora gigantea were maintained on potato dextrose agar (PDA) at 4 ◦C. ForNA isolation, the fungi were cultured in agar plates at 28 ◦C for 10 days on induc-ion media [13] including 0.5% (w/v) carboxymethyl cellulose (CMC; Megazyme) orvicel (Merck). For induction in liquid medium the same medium was used withoutgar.

.3. Evaluation of the cellulase production by cellulolytic fungi

The production of carboxymethycellulase activity (CMCase) in liquid mediumupplemented with cellulose was evaluated by measuring reducing sugars releasedrom CMC, after six days inoculation with 1 cm2 PDA culture. The enzymatic reac-ion was performed as follow: a mixture of 50 �L of fungal culture supernatant and00 �L of 1% (w/v) CMC in 50 mM sodium citrate pH 4.8 was incubated for 30 min at7 ◦C; reducing sugars were quantified by reaction with dinitrosalicylic acid (DNS)14]. Absorbance at 550 nm was measured in a microplate reader (Asys UVM 340,ugendorf, Austria) and values were converted to enzyme units by use of a glu-ose calibration curve. One enzyme unit was defined as the amount of enzyme thateleases 1 �mol of glucose equivalent per minute.

.4. Hydrolysis of N. pumilio wood chips

Supernatant of fungi cultured in induction medium was concentrated by ammo-ium sulfate precipitation at 80% saturation. Pellets were suspended in 50 mModium acetate buffer pH 6 and dialyzed over night against the same buffer. Theydrolysis assays were conducted in 50 mM sodium acetate pH 5, using 50 mg mL−1

l Technology 49 (2011) 485– 491

ionic liquid-pretreated N. pumilio wood chips (0.1–0.2 mm) [15], 0.25 paper fil-ter units (FPU) per mL of concentrated fungal cellulase or Celluclast (Novozyme),equivalent to 5 FPU g−1 substrate, and 2.5 g L−1 Tween 20. Hydrolysis was conductedat 50 ◦C with orbital agitation (300 rpm) for 1.5, 25 or 50 h. Reducing sugars weremeasured as indicated in point 2.3. FPU activity was measured as described [16].

2.5. Primers design and bioinformatic analyses

cDNA sequences of fungal GH family 5, 7 and 61 endoglucanase (EC 3.2.1.4)were obtained from EMBL-EBI database and aligned using ClustalX 2.0.9 [17].The accession numbers of the sequences used to design the degenerated primerswere Q75UV6, Q5W7K4, Q9C3Z8, O74706, Q4WN62, O59951, Q5TKT6, Q2VRK9,Q8WZD7, Q12638, Q12624, Q12637 and Q04469 for GH family 5; P07981, Q12714,Q4WCM9, O13455, Q8NK01, Q4WAJ6, Q12622, P56680, P46237 and Q9HGT3 for GHfamily 7 and O14405, Q4WF08, A2QJX0, Q6MYM8, Q7Z9M7 and A2R5N0 for GH fam-ily 61. For identification of the partial cellulase sequences, consensus-degenerateprimers were designed based on these alignments, using the CODEHOP strategy[18]. Specific primers were designed by hand in order to complete the TvEG cDNAsequence. The primer sequences are shown in Table 1.

Homology analysis was carried out using the BLAST software [19] from theNational Center for Biotechnology Information (NCBI); putative functional domainsidentification was carried out by search the Conserved Domain Database (CDD) [20].

2.6. RNA isolation and cDNA synthesis

Total RNA was isolated from fungal mycelium grown in solid induction media.The mycelium was harvested from the plate and lysed with TRI reagent (Ambion) inliquid nitrogen using a mortar and pestle, according to the manufacturer’s directions.Single strand cDNA (sscDNA) was synthesized with the Oligo(dT)15 primer usingthe Reverse Transcription System kit (Promega) according to the manufacturer’sdirections. PCR was performed in a reaction mixture containing: 1.4 mM MgCl2,0.2 mM dNTPs, 1 �M reverse primer, 1 �M forward primer, buffer Taq 1×, 1 U TaqDNA polymerase and 0.5 �L of template sscDNA in a final volume of 25 �L. Reactionswere carried out in an Eppendorf Master Cycler Gradient; gel-purified PCR productswere cloned into pGEM-T Easy vector, and sequenced by Macrogen (Korea).

2.7. Phylogenetic analysis

Multiple alignments of sequences of glycoside hydrolases from families 5, 7 and61 obtained from the EMBL-EBI data base were performed using ClustalX 2.0.9. Infor-mation on the accession number of the sequences used as templates are availablein Supplementary Material 1. Construction and visualization of a radial neighbor-joining phylogenetic tree were done with the TreeView version 1.6.6 program [21].

2.8. Expression of the endoglucanase gene in P. pastoris X-33

The cloning and recombinant expression of the full length T. versicolor endoglu-canase gene was carried out using the Easy-Select Pichia expression kit (Invitrogen).Specific primers TvEGPP1 For and TvEGPP1 Rev (Table 1) were designed in order toadd XhoI and XbaI restriction sites at the ends of the endoglucanase gene. The PCRproduct was digested with these enzymes and the fragment was cloned into thepPICZ�B vector. The forward primer contained a XhoI recognition site, a Kex2 signalcleavage and the sequence encoding for the five first amino acids of the mature TvEGenzyme. The reverse primer contained the sequence encoding for the last four aminoacids of the protein, followed by six histidine codons and a stop codon. Recombinantplasmid was introduced into P. pastoris X-33 by electroporation, following the sup-plier recommendations. Transformants were selected in YPD plates supplementedwith 1 M sorbitol and 100 �g mL−1 Zeocin. Clones having multiple insertions in theyeast genome were selected in YPD plates supplemented with 500 �g mL−1 Zeocin.The correct sequence of the construction was confirmed by automatic sequencingat Macrogen. Recombinant protein production was carried out according to the kitprovider directions. The confirmed clone was cultured in BMGY (Buffered complexmedium containing glycerol) until OD600 4. For the induction phase, cells were trans-ferred to BMMY medium (Buffered complex medium with 0.5% (v/v) methanol),repeating the addition of methanol every 24 h. Recombinant expression was moni-tored every 24 h in aliquots of the culture supernatant by measuring activity on CMCand by electrophoresis. For this, a four mL aliquot of each supernatant was concen-trated by ultracentrifugation using a Centricon centrifugal device YM-3 (Millipore).The concentrated fractions were analyzed by SDS–PAGE and CMC-zymogram. Forthe CMCase activity assays, aliquots of supernatant were analyzed directly, while theelectrophoretic analysis was carried out using the concentrated fractions. SDS–PAGEwas carried out in 12.5% polyacrylamide gel in Miniprotean II (BioRad), as rec-ommended. For the zymogram analysis the protein separation was carried out ina similar protocol except that the gel contained 0.2% (w/v) CMC and the proteinsamples were not heated before the electrophoresis. SDS removal for protein renat-

uration in the gel was carried out by incubating with 0.25% (v/v) Triton X-100 for onehour at room temperature, and then the gel was incubated for one hour at 37 ◦C in50 mM sodium acetate pH 5.0. The activity bands were revealed by staining with 0.1%(w/v) Congo Red and distaining with 1 M NaCl. In all the experiments P. pastoris X-33transformed with the vector pPICZ�B was analyzed in parallel and used as control.

A. Salinas et al. / Enzyme and Microbial Technology 49 (2011) 485– 491 487

Table 1PCR primers used for cDNA amplification and plasmid construction.

Primer name Sequence (5′ → 3′)a Target/use

F5fwd TGATCTTCGACATCATGAACGARYMHCAYRA GH family 5F5rev CCCCCACCAGGAGCCDKYDKCCCA GH family 5F7fwd CGACGAGGAGACCTGCGSNVANAAYTG GH family 7F7rev CACGTAGCACTGGGCGTCRCARTANCC GH family 7F61fwd TGGACAAGACCACCAACAAGTKBKTNAARAT GH family 61F61rev ACTTCTAGATGTTGATGAAGATGCCNGSRTCNGT GH family 61106EN5For GGACCCGAACAACAACATTGCTA Specific primer for TvEG cDNA cloning106EN5Rev GCGAGGATGAGCTGGCTGGT Specific primer for TvEG cDNA cloningTvEGPP1 For CTCGAGAAAAGAGTCTGGCGGCAGTG Cloning full length TvEG in pPICZ�B

GGAA

3

3

TLLoadcltttctlrtTltsfh

Fa(asv

TvEGPP1 Rev TCTAGATCAATGATGATGATGATGATGCT

a The IUPAC nomenclature for degenerate primers is used.

. Results and discussion

.1. Evaluation of the cellulase activity produced by fungi

The production of cellulase activity by native Chilean strains of. versicolor, G. applanatum, P. placenta, F. oxysporum, P. ostreatus,. edodes and P. gigantea was evaluated. P. placenta, P. ostreatus,. edodes and P. gigantea did not produce any significant activityf CMC hydrolysis (data not shown). T. versicolor, G. applanatumnd F. oxysporum were selected based on the cellulase activity pro-uced under conditions of induction with CMC, microcrystallineellulose and N. pumilio lignocellulose. As shown in Fig. 1, regard-ess of the type of inducer present in the medium, production ofhe enzymes needed for CMC hydrolysis was highest in G. applana-um, followed by T. versicolor and F. oxysporum. Considering thathe polymer substrate must be previously hydrolyzed to be used aarbon source and also to be effective for cellulase gene induction,he growth in lignocellulose can be a good marker of the cellu-ases functionality on lignocellulose. This notion was confirmed byesults in Fig. 2, where the hydrolysis of N. pumilio wood chips byhe crude cellulase from T. versicolor and G. applanatum is shown.hese fungi exhibit significant hydrolyzing activity on lignocellu-ose. Despite in both cases the hydrolytic activity is lower thanhe activity exhibited by the commercial preparation of T. ree-

ei, these results are helpful as they indicate that these white rotungi are good producers of competent enzymes for lignocelluloseydrolysis.

ig. 1. Production of CMCase activity by Trametes versicolor, Fusarium oxysporumnd Ganoderma applanatum. The fungi were cultured for 6 days at 28 ◦C using CMChatched bars); Avicel (white bars); N. pumilio (black bars) as carbon sources. CMCasectivity was measured in fungi culture supernatants using 1% (w/v) CMC as sub-trate, at 37 ◦C for 60 min. All points represent averages of at least three experimentalalues. Standard deviation was always less of 10%.

CGCCTTG Cloning full length TvEG in pPICZ�B

3.2. Cloning of the fungal cellulase genes

Total RNA was obtained from the fungi cultured in solid mediumsupplemented with CMC. Afterward cDNA was amplified by PCRwith degenerate primers designed for targeting GH families 5,7 and 61. These primers were designed based on the alignmentof cellulases belonging to the IUBMB classification EC 3.2.1.4,which contains most of the fungal endo-�-1,4-glucanases (seeSupplementary Material 1 for details on the accession number ofthe sequences used as templates). Five fragments were identified(see Supplementary Material 2 for details of the cDNA sequences),and the BLAST homology analysis revealed that the amino acidsequence of these fragments effectively shares important identitywith cellulases from families 5, 7 and 61, respectively (Table 2).One of the fragments isolated from F. oxysporum resulted to beidentical to the gene encoding for the EG I already described bySheppard et al. [22]. The endoglucanase I (EG I) from the F. oxys-porum consists of a 411 amino acid catalytic core and it lacks aCBM. The enzyme belongs to glycoside hydrolase family 7, whichalso includes the cellobiohydrolase I from T. reesei. The second frag-ment from F. oxysporum resulted to be a non described sequence;the highest alignment was to a putative, uncharacterized GH family61 cellulase from G. graminicola.

Analysis of the cDNA isolated from G. applanatum revealed a

non-described GH with 58% identity in 86 amino acids to the GHfamily 61 of the fungus Laccaria bicolor, specifically the sequencecomprising the catalytic domain of the enzyme. This is the firstdescription of a GH61 from G. applanatum. GH61 are enigmatic

Fig. 2. Hydrolysis of N. pumilio lignocellulose by T. versicolor, G. applanatum and T.reesei cellulases. The crude enzymes from the fungal culture supernatant, concen-trated by ammonium sulfate precipitation, were incubated with 5% N. pumilio woodchips at 50 ◦C for 1.5 h (hatched bars), 25 h (white bars) or 50 h (black bars) andreducing sugars were determined by the DNS method. All points represent averagesof at least three experimental values. Standard deviation was always less of 10%.

488 A. Salinas et al. / Enzyme and Microbial Technology 49 (2011) 485– 491

Table 2Analysis of the cDNA sequences encoding for the cellulases identified in this work.

Fungus cDNA Highest alignmentb GH family (CDD e-value)d

Code Fraction sequenced (%)a Microorganism Accession codec Identity (%)

Fusarium oxysporum Fof7 23 Fusarium oxysporum AAG09047 100 7 (10−38)Fusarium oxysporum Fof61 28 Glomerella graminicola EFQ31130 68 61 (10−30)Ganoderma applanatum Gaf61 26 Laccaria bicolor XP 001883194 58 61 (10−27)Trametes versicolor TvEG 95 Trametes hirsuta BAD01163 96 5 (10−14)Trametes versicolor Tvf7 22 Polyporus arcularius BAF80326 82 7 (10−44)

a Assuming as 100% the size of the enzyme with the highest alignment in the BLASTX analysis [19].

pbawtolpo

ic5isv

Terwffttdfpcc(7snpa

tGolarob

3

aas

Fig. 3. Phylogenetic analysis of the new cellulases. The alignments were constructedusing CLUSTAL X program [17]. The tree was visualized by TREEVIEW [21]. Glyco-side hydrolase families are surrounded by ellipses. Arrows highlight the sequencesidentified in this work. Sequences were obtained from the Europe Bioinformatics

b According to the results of the BLASTX analysis.c EMBL database.d Results of the Conserved Domain Database search [20].

roteins that lack measurable hydrolytic activity by themselves,ut can accomplish an auxiliary role during the hydrolysis cat-lyzed by classical cellulases [23]. As an example, Harris et al.ere able to reduce by 2-fold the total protein loading required

o hydrolyze lignocellulosic biomass by incorporating the gene forne GH61 protein into a commercial T. reesei strain producing highevels of cellulolytic enzymes [24]. In consequence, these GH61roteins are attractive enzymes to be considered at the momentf producing optimized cellulolytic mixtures.

Analysis of the cDNA obtained from T. versicolor allowed thesolation of sequences encoding for putative GH family 7 and 5.DNA cloning of a GH family 7 cellobiohydrolase from T. versicolor2P was recently reported [25], but the sequence is not available

n public data bases; in consequence, is not possible to know if theequence identified in this work corresponds to cel7A present in T.ersicolor 52P.

Similitude analysis of the second fragment from T. versicolor,vf5, suggests that the gene encodes for a novel GH family 5ndoglucanase [4]. The complete cDNA for the mature protein waseconstructed using RT-PCR with specific primers and the proteinas named TvEG. The sequence of a GH family 5 endoglucanase

rom T. versicolor has not been reported so far. An endoglucanaserom this fungus was described in 1963 by Pettersson [11] buthe molecular weight reported was significantly lower (11 kDa)han the 38.6 kDa estimated for TvEG, suggesting that the enzymeescribed by Pettersson is either a different enzyme or a proteolyticragment of TvEG. Similarly, Idogaki and Kitamoto reported theurification and brief characterization of an endoglucanase from aulture of Coriolus versicolor (formerly T. versicolor) strain TD-7 [26];omparison of the molecular weight estimated for that CMCase29 kDa) suggests that TvEG and the endoglucanase from strain TD-

could be the same enzyme. Information regarding the amino acidequence of this or any other endoglucanase from T. versicolor isot available so far. Comparison of the biochemical and molecularroperties should be very useful to clearly identify these enzymesnd to define the complete set of T. versicolor endocellulases.

A phylogenetic tree was constructed based on the alignment ofhe amino acid sequences of fungal known endoglucanases fromHs families 5, 7 and 61. Three main branches were obtained, eachne corresponding to each GH family. As shown in Fig. 3, the cellu-ase sequences identified in this work are distributed in the treeccording to the primers used in the cDNA screen. From theseesults is clear that the use of degenerate primers designed basedn homology regions is a strong strategy to identify novel enzymes,eing highly specific for the targeted glycoside hydrolase family.

.3. Analysis of the TvEG amino acid sequence

TvEG full length cDNA encoded a 364 amino acid protein, with predicted molecular weight of 38.6 kDa. The BLAST homologynalysis of the deduced amino acid sequence of TvEG in Fig. 4howed the highest overall identity with endoglucanases from

Institute database. The length of the braches indicates the divergence among theamino acid sequences. Bar, 0.1 substitutions per site. The accession numbers andspecies used to generate the tree are available in Supplementary Material 1.

Trametes hirsuta ThEG (96% identity, accession Q75UV6) [27], Irpexlacteus En-1 (65%, Q5W7K4) [28], Volvariella volvacea EGI (57%,Q9C3Z8) [29], Humicola insolens EGL2 (58%, Q12624) [30] and T.reesei EG II (22%, P079829) [31]. All these sequences are membersof the GH family 5 and have a family 1 CBM.

TvEG has a multi-domain structure composed of a N-terminalCBM, a Ser-/Thr-rich linker, and a catalytic domain (Fig. 4). TheCBM of T. versicolor EG belongs to the CBM family 1, which ischaracteristic of fungi [32]. It contains four cysteines which formtwo disulfide bridges (Cys31-Cys48 and Cys42-Cys58, according tonumbering in Fig. 4), four aromatic amino acid residues (Trp28,Phe36, Tyr54 and Tyr55), and two glutamines (Gln30 and Gln57);all these amino acid residues are highly conserved in this familyand are essential for the adsorption capacity [33]. In the catalyticdomain is possible to distinguish the conserved motifs Asn-Glu-Proand Glu-X-Gly, corresponding to the acid catalyst (Glu229) and sta-bilizing anion/nucleophile (Glu339), respectively [34]. Also, Glu229and Glu339 are the only Glu residues that are conserved through-

out the catalytic domain, suggesting that they are involved in thecatalysis.

The BLAST analysis showed that the T. hirsuta and I. lac-teus endoglucanases are the most similar to TvEG (Fig. 4). It is

A. Salinas et al. / Enzyme and Microbial Technology 49 (2011) 485– 491 489

F ck andB esidut rsicol

iicafFcTocc

ig. 4. Sequence alignment of T. versicolor EG and other fungal endoglucanases. Bla define the CBM and the catalytic domain, respectively. The black circles indicate rhe residues that are predicted to be critical in the catalysis. Abbreviations: T.v., T. ve

nteresting that T. hirsuta endoglucanase activity ratio CMC/Avicels significantly lower than the same index for I. lacteus endoglu-anase [26,37]. This effect was attributed by the authors to a higherccessibility of the T. hirsuta catalytic domain to the substrate,acilitated by a shorter linker [27]. As shown in the alignment inig. 4, this structural characteristic is also present in TvEG; theonservation of this feature and the high identity level between

vEG and ThEG suggest that TvEG also could be highly reactiven microcrystalline cellulose. Considering that this is a desirableharacteristic in endoglucanases used for lignocellulosic biomassonversion, this possibility must be investigated by extensive

grey boxes indicate identical and similar amino acid, respectively. Brackets A andes that are predicted to participate in disulfide bridges. The black triangles indicateor; T.h., T. hirsuta; I.l., I. lacteus; V.v., V. volvacea; and H.i., H. insolens; T.r., T. reesei.

characterization and study of the real potential of TvEG forapplication in insoluble cellulose breakdown.

3.4. Recombinant expression of TvEG in P. pastoris

The sequence corresponding to the mature TvEG (from aminoacid 27 to 405, Fig. 4), was cloned into the expression vector

pPICZ�B and transformed into P. pastoris X-33. The TvEG gene wascloned downstream the Saccharomyces cerevisiae � factor signalsequence and the Kex2 cleavage sequence. The Kex2 signal cleavagewas conveniently introduced in the forward primer TvEGPP1 For,

4 crobia

irtiotpsa

ibtt(cCtCmtt

FCifD(c(a

90 A. Salinas et al. / Enzyme and Mi

n order to promote the correct N-terminal processing of TvEG andeduce the possibility of non-native amino acids in the enzyme. Forhe same reason, a sequence encoding for six histidine codons wasncluded in the reverse primer, TvEGPP1 Rev, just after the 3′ endf the gene and before the stop codon. This design should avoidhe presence of the C-myc epitope at the C-terminal of TvEG, stillermitting the purification by affinity to the polyhistidine tail. Con-equently, TvEG should be secreted to the extracellular space, with

native N-terminal and a six-histidine tag at the C-terminal.The kinetic of the recombinant enzyme production was stud-

ed following the appearance of the CMCase activity in the cultureroth and in extracts of the intracellular fraction. Only the results ofhe extracellular content are shown. Estimates showed that morehat the 95% of the activity was released to the extracellular spacedata not shown), suggesting that the secretion signal from the S.erevisiae �-factor correctly signaled the exportation of TvEG. TheMCase activity increased consistently in the supernatant until thehird day (72 h), reaching a plateau after that. The production of

MCase activity was 18 U L−1 at the end of the monitoring experi-ent. SDS–PAGE analysis of the extracellular proteins produced by

he control (P. pastoris X-33/pPICZ�B) and the recombinant cul-ure (P. pastoris X-33/pPICZ�B-TvEG) during the induction with

ig. 5. Time course of recombinant full length TvEG production by P. pastoris X-33.ells were induced in BMMY (buffered methanol-complex medium) for the times

ndicated. (A) CMCase activity measured in culture broth of P. pastoris X-33 trans-ormed with pPICZ�B-TvEG (white circles) or pPICZ�B (black circles). The Opticalensity at 600 nm (OD600) of the recombinant culture during the induction is shown

black diamonds). SDS–PAGE (B) and zymogram (C) of the extracellular fractionsoncentrated by ultrafiltration from the experiment shown in A. Induction timesh) are indicated on top of the wells; lane M, molecular mass markers (kDa). Therrow head indicates the recombinant enzyme.

l Technology 49 (2011) 485– 491

methanol revealed in P. pastoris X-33/pPICZ�B-TvEG, the appear-ance of a faint band around 35 kDa, consistent with the 38.6 kDaestimated for the recombinant TvEG (Fig. 5B); the zymographicanalysis (Fig. 5C) showed an activity signal consistent con this pro-tein, demonstrating the functional expression of TvEG in P. pastoris.

The activity gel also showed a high molecular weight proteinhaving activity on CMC, present in both cultures. This activity isprobably responsible of the background observed in supernatantsof the control culture in Fig. 5A. Genomic sequences encoding forglycoside hydrolases have been described in P. pastoris GS115, aclose variant of the X-33 strain. These sequences belong to fam-ilies 2, 3, 10, 20, 45, 53 and 88 [35]; among them, families 20and 45 have members with putative hydrolytic activity on endo-�-1,4-glycosidic bonds (EC 3.2.1.4), that could be responsible ofthe background activity observed in our experiments. Not enoughexperimental information is available in literature for clearly iden-tify the origin of these bands.

The zymogram in Fig. 5C shows that TvEG in the supernatant isnot a homogeneous population of protein molecules; several bandsaround the expected molecular weight are visible. The origin of thisheterogeneity in the protein size could be on potential proteolyticcleavage of the endoglucanase in the linker region, which, by sep-arating the CBD could create active forms of the enzyme slightlysmaller. A second possibility is the modification of the enzyme byaddition of a variable number of carbohydrate molecules (glycosy-lation). It is recognized that P. pastoris is capable of adding both O-and N-linked carbohydrate moieties to secreted proteins [36]. Atthis point is not possible to discard any of these possibilities.

Preliminary evaluation of the endoglucanase activity on CMCin the IMAC-purified recombinant TvEG (not shown) gave a spe-cific activity of 35–40 U mg−1. Using this value a simple estimategives approximately 0.5 mg of recombinant TvEG per L of culturebroth. This production level is significantly lower than the expres-sion level of V. volvacea and A. niger endoglucanases [37,38] inP. pastoris; these enzymes were produced at 65 and 40 mg L−1,respectively. We are currently studying experimental alternativesfor improvement of the TvEG expression level by optimization ofthe fermentation conditions.

4. Conclusions

1. This study demonstrates that the Chilean native strains of G.applanatum and T. versicolor analyzed in this work produceenzymes able of degrading the cellulose present in lignocellulosefrom N. pumilio.

2. The screening of cellulolytic fungi carried out by reverse tran-scription and PCR with degenerate primers resulted to be afruitful strategy. Despite the primers were designed based onconsensus sequences, results reported in this work indicate thatis possible to identify sequences that diverge of the establishedconsensus and still maintain the targeted specificity.

3. TvEG, an endoglucanase from T. versicolor, has been cloned andexpressed as a functional recombinant enzyme. The enzymeshares high identity with ThEG, an endoglucanase from T. hir-suta with high activity on crystalline cellulose. This fact andthe presence of a CBM in the enzyme suggest that TvEG couldhave interesting applications for hydrolysis of insoluble cellu-losic substrates.

4. The yeast P. pastoris produces the fungal endoglucanase as anactive enzyme, but with low yield. The system needs to be opti-mized in order to improve the production of recombinant TvEG.

Acknowledgements

This work was supported by the Domeyko Research Programme,University of Chile; CONICYT through its Bicentennial Programme

robia

C(moi

A

t

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

A. Salinas et al. / Enzyme and Mic

CF05 and The Institute for Cell Dynamics and BiotechnologyICDB). We thank the Vice Presidency of Research and Develop-

ent, University of Chile and the Academy Direction of the Facultyf Physical Sciences and Mathematics for travel support for attend-ng to the 14th IBS Congress.

ppendix A. Supplementary data

Supplementary data associated with this article can be found, inhe online version, at doi:10.1016/j.enzmictec.2011.10.003.

eferences

[1] Dashtban M, Schraft H, Qin W. Fungal bioconversion of lignocellulosic residues;opportunities & perspectives. Int J Biol Sci 2009;5:578–95.

[2] Bhat MK, Bhat S. Cellulose degrading enzymes and their potential industrialapplications. Biotechnol Adv 1997;15:583–620.

[3] Gilkes NR, Henrissat B, Kilburn DG, Miller Jr RC, Warren RA. Domains inmicrobial beta-1,4-glycanases: sequence conservation, function, and enzymefamilies. Microbiol Rev 1991;55:303–15.

[4] Henrissat B, Davies GJ. Structural and sequence-based classification of glycosidehydrolases. Curr Opin Struct Biol 1997;7:637–44.

[5] Boraston AB, Bolam DN, Gilbert HJ, Davies GJ. Carbohydrate-binding modules:fine-tuning polysaccharide recognition. Biochem J 2004;382:769–81.

[6] Baldrian P, Valaskova V. Degradation of cellulose by basidiomycetous fungi.FEMS Microbiol Rev 2008;32:501–21.

[7] Vinzant TB, Adney WS, Decker SR, Baker JO, Kinter MT, Sherman NE, et al.Fingerprinting Trichoderma reesei hydrolases in a commercial cellulase prepa-ration. Appl Biochem Biotechnol 2001;91–93:99–107.

[8] Schmoll M, Kubicek CP. Regulation of Trichoderma cellulase formation: lessonsin molecular biology from an industrial fungus. A review. Acta MicrobiolImmunol Hung 2003;50:125–45.

[9] Henrissat B, Driguez H, Viet C, Schülein M. Synergism of cellulases fromTrichoderma reesei in the degradation of cellulose. Biotechnology 1985;3:722–6.

10] Nidetzky B, Steiner W, Hayn M, Claeyssens M. Cellulose hydrolysis by the cellu-lases from Trichoderma reesei: a new model for synergistic interaction. BiochemJ 1994;298(Pt 3):705–10.

11] Pettersson G, Cowling EB, Porath J. Studies on celluloytic enzymes. I. Isolation ofa low-molecular-weight cellulase from Polyporus versicolor. Biochim BiophysActa 1963;67:1–8.

12] Sambrook J, Russell DW. Molecular cloning. A laboratory manual. Cold SpringHarbor: Cold Spring Harbor Laboratory Press; 2001.

13] Lee CC, Wong DW, Robertson GH. Cloning and characterization of two cellulasegenes from Lentinula edodes. FEMS Microbiol Lett 2001;205:355–60.

14] Miller GL. Use of dinitrosalicylic acid reagent for determination of reducingsugar. Anal Chem 1959;31:426–8.

15] Pezoa R, Cortinez V, Hyvarinen S, Reunanen M, Hemming J, Lienqueo ME,et al. Use of ionic liquids in the pretreatment of forest and agriculturalresidues for the production of bioethanol. Cellul Chem Technol 2010;44:

165–72.

16] Xiao Z, Storms R, Tsang A. Microplate-based filter paper assay to measure totalcellulase activity. Biotechnol Bioeng 2004;88:832–7.

17] Larkin MA, Blackshields G, Brown NP, Chenna R, McGettigan PA, McWilliam H,et al. ClustalW and ClustalX version 2. Bioinformatics 2007;23:2947–8.

[

l Technology 49 (2011) 485– 491 491

18] Rose TM, Henikoff JG, Henikoff S. CODEHOP (COnsensus-DEgenerate HybridOligonucleotide Primer) PCR primer design. Nucleic Acids Res 2003;31:3763–6.

19] Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W, et al. GappedBLAST and PSI-BLAST: a new generation of protein database search programs.Nucleic Acids Res 1997;25:3389–402.

20] Marchler-Bauer A, Lu S, Anderson JB, Chitsaz F, Derbyshire MK, DeWeese-ScottC, et al. CDD: a Conserved Domain Database for the functional annotation ofproteins. Nucleic Acids Res 2011;39:D225–9.

21] Page RDM. TREEVIEW: an application to display phylogenetic trees on personalcomputers. Comput Appl Biosci 1996;12:357–8.

22] Sheppard PO, Grant FJ, Oort PJ, Sprecher CA, Foster DC, Hagen FS, et al. The useof conserved cellulase family-specific sequences to clone cellulase homologuecDNAs from Fusarium oxysporum. Gene 1994;150:163–7.

23] Merino ST, Cherry J. Progress and challenges in enzyme development forbiomass utilization. Adv Biochem Eng Biotechnol 2007;108:95–120.

24] Harris PV, Welner D, McFarland KC, Re E, Navarro Poulsen J-C, Brown K,et al. Stimulation of lignocellulosic biomass hydrolysis by proteins of glyco-side hydrolase family 61: structure and function of a large, enigmatic family.Biochemistry 2010;49:3305–16.

25] Lahjouji K, Storms R, Xiao Z, Joung KB, Zheng Y, Powlowski J, et al. Biochemicaland molecular characterization of a cellobiohydrolase from Trametes versicolor.Appl Microbiol Biotechnol 2007;75:337–46.

26] Idogaki H, Kitamoto Y. Purification and some properties of a carboxymethylcellulase from Coriolus versicolor. Biosci Biotechnol Biochem 1992;56:970–1.

27] Nozaki K, Seki T, Matsui K, Mizuno M, Kanda T, Amano Y. Structure andcharacteristics of an endo-beta-1,4-glucanase, isolated from Trametes hir-suta, with high degradation to crystalline cellulose. Biosci Biotechnol Biochem2007;71:2375–82.

28] Toda H, Takada S, Oda M, Amano Y, Kanda T, Okazaki M, et al. Gene cloning of anendoglucanase from the basidiomycete Irpex lacteus and its cDNA expressionin Saccharomyces cerevisiae. Biosci Biotechnol Biochem 2005;69:1262–9.

29] Ding SJ, Ge W, Buswell JA. Endoglucanase I from the edible straw mushroom,Volvariella volvacea. Purification, characterization, cloning and expression. EurJ Biochem 2001;268:5687–95.

30] Takashima S, Nakamura A, Masaki H, Uozumi T. Cloning, sequencing, andexpression of a thermostable cellulase gene of Humicola grisea. Biosci Biotech-nol Biochem 1997;61:245–50.

31] Saloheimo M, Lehtovaara P, Penttila M, Teeri TT, Stahlberg J, Johansson G, et al.EGIII, a new endoglucanase from Trichoderma reesei: the characterization ofboth gene and enzyme. Gene 1988;63:11–22.

32] Linder M, Teeri T. The roles and function of cellulose-binding domains. JBiotechnol 1997;57:15–28.

33] Linder M, Mattinen ML, Kontteli M, Lindeberg G, Stahlberg J, Drakenberg T, et al.Identification of functionally important amino acids in the cellulose-bindingdomain of Trichoderma reesei cellobiohydrolase I. Protein Sci 1995;4:1056–64.

34] Clarke AJ, Drummelsmith J, Yaguchi M. Identification of the catalytic nucle-ophile in the cellulase from Schizophyllum commune and assignment of theenzyme to Family 5, subtype 5 of the glycosidases. FEBS Lett 1997;414:359–61.

35] De Schutter K, Lin Y-C, Tiels P, Van Hecke A, Glinka S, Weber-Lehmann J, et al.Genome sequence of the recombinant protein production host Pichia pastoris.Nat Biotechnol 2009;27:561–6.

36] Macauley-Patrick S, Fazenda ML, McNeil B, Harvey LM. Heterologous proteinproduction using the Pichia pastoris expression system. Yeast 2005;22:249–70.

37] Quay DHX, Bakar FDA, Rabu A, Said M, Illias RM, Mahadi NM, et al. Overexpres-

sion, purification and characterization of the Aspergillus niger endoglucanase,EglA, in Pichia pastoris. Afr J Biotechnol 2011;10:2101–11.

38] Ding S-J, Ge W, Buswell JA. Secretion, purification and characterisation of arecombinant Volvariella volvacea endoglucanase expressed in the yeast Pichiapastoris. Enzyme Microb Technol 2002;31:621–6.