The inhibition specificity of recombinant Penicillium funiculosum xylanase B towards wheat proteinaceous inhibitors

Biochimica et Biophysica Acta 1701 (2004) 121 – 128 www.bba-direct.com

The inhibition specificity of recombinant Penicillium funiculosum xylanase B towards wheat proteinaceous inhibitors Alexandre Brutusa, Claude Villarda, Anne Durandb, Tariq Tahirb, Caroline Furnissb, Antoine Puigservera, Nathalie Jugea,b,*, Thierry Giardinaa a

Laboratoire de Biochimie et Biologie de la Nutrition, Institut Me´diterrane´en de Recherche en Nutrition, UMR Universite´ Aix Marseille III-INRA 1111, Service 342, Faculte´ des Sciences et Techniques Saint-Je´roˆme, 13397 Marseille Cedex 20, France b Institute of Food Research, Norwich Research Park, Norwich NR4 7UA, United Kingdom Received 16 April 2004; received in revised form 14 June 2004; accepted 24 June 2004 Available online 27 July 2004

Abstract The filamentous fungus Penicillium funiculosum produces a mixture of modular and non-modular xylanases belonging to different glycoside hydrolase (GH) families. In the present study, we heterologously expressed the cDNA encoding GH11 xylanase B (XYNB) and studied the enzymatic properties of the recombinant enzyme. Expression in Escherichia coli led to the partial purification of a glutathione fusion protein from the soluble fraction whereas the recombinant protein produced in Pichia pastoris was successfully purified using a onestep chromatography. Despite O-glycosylation heterogeneity, the purified enzyme efficiently degraded low viscosity xylan [K m=40F3 g l 1, V max=16.1F0.8 Amol xylose min 1 and k cat=5405F150 s 1 at pH 4.2 and 45 8C] and medium viscosity xylan [K m=34.5F3.2 g l 1, V max=14.9F1.0 Amol xylose min 1 k cat=4966F333 s 1 at pH 4.2 and 45 8C]. XYNB was further tested for its ability to interact with wheat xylanase inhibitors. The xylanase activity of XYNB produced in P. pastoris was strongly inhibited by both XIP-I and TAXI-I in a competitive manner, with a K i of 89.7F8.5 and 2.9F0.3 nM, respectively, whereas no inhibition was detected with TAXI-II. Physical interaction of both TAXI-I and XIP-I with XYNB was observed using titration curves across a pH range 3–9. D 2004 Elsevier B.V. All rights reserved. EC classification: xylanase B, endo-(1,4)-h-xylanase (EC 3.2.1.8) Keywords: Penicillium funiculosum; Heterologous expression; Xylanase; Proteinaceous inhibitor; Enzyme inhibition; Wheat xylanase inhibitors; Glycoside hydrolase family 11

1. Introduction Endo-1,4-g-xylanases (EC 3.2.1.8) hydrolyse h-(1,4)linked d-xylosyl residues releasing xylo-oligosaccharides Abbreviations: CBM, carbohydrate binding module; GH, glycoside hydrolase; E-XYNB, XYNB produced in Escherichia coli; GST, glutathione-S-transferase; P-XYNB, XYNB produced in Pichia pastoris; IPTG, isopropyl-1-thio-h-d-galactopyranoside; LVX, low viscosity xylan; MVX, medium viscosity xylan; PAS, periodic acid Schiff; PBS, phosphatebuffered saline; TFA, trifluoroacetic acid * Corresponding author. Institute of Food Research, Norwich Research Park, Norwich NR4 7UA, UK. Tel.: +44 1603 255 068; fax: +44 1603 255 038. E-mail address: [email protected] (N. Juge). 1570-9639/$ - see front matter D 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.bbapap.2004.06.010

[1]. According to the sequence-based glycoside hydrolase (GH) classification, they have been mainly grouped into families 10 (GH10) and 11 (GH11) (http://afmb.cnrs-mrs.fr/ CAZY/index.html) [2]. The filamentous fungus Penicillium funiculosum secretes three different types of xylanases for which the genes have been cloned; two have a modular structure, XYNA (GH7) and XYNB (GH11) [3], whereas the third one is non-modular (XYNC, GH11) [4]. Xylanases are used in many food and feed animal industrial processes such as additives for broilers feed, bread making, aiding in separation of wheat or other cereal gluten from starch, production of juice from fruits or vegetables, extracting more fermentable sugar from barley for making beer, etc. [1]. However, proteinaceous xylanase inhibitors, recently

122

A. Brutus et al. / Biochimica et Biophysica Acta 1701 (2004) 121–128

identified in cereals, might hamper the efficacy of the xylanases used in industrial applications (for a review, see Ref. [5]). Two distinct types of xylanase inhibitors have so far been characterized (for a review, see Refs. [6,7]). XIP-I isolated from wheat (Triticum aestivum) is a monomeric and glycosylated protein with a pI of 8.7–8.9 and a molecular mass of 31.0 kDa [8]. The inhibitor shows high sequence homology with GH18 chitinases [9] and has the typical TIM-barrel (h/a)8-fold characteristic of this family [10]. XIP-I inhibits all GH10 and GH11 fungal xylanases tested so far, except for the Aspergillus aculeatus xylanase but bacterial xylanases are not inhibited [11]. The other type of xylanase inhibitor, TAXI (T. aestivum xylanase inhibitor), was also purified from wheat [12,13]. It is a mixture of two proteins, TAXI I and TAXI II, which show structural homology but differ from one another in pI and xylanase specificity. TAXI-like proteins occur as two molecular forms, A and B. Form A is a non-proteolytically processed protein consisting of one 40-kDa polypeptide chain, while form B is its proteolytically processed counterpart which consists of two polypeptide chains of 30 and 10 kDa, respectively, which are associated by disulfide bonds [13]. In contrast to XIP, TAXI inhibitors appear specific for GH11 xylanases from both bacterial and fungal origin [14]. Both classes of inhibitors also occur in other cereals such as barley, rye and durum wheat [15–17] (for a review, see Ref. [18]). In order to test whether the P. funiculosum xylanase B was affected by the presence of these inhibitors, the enzyme was produced by heterologous expression, purified and characterised in terms of enzymatic properties and inhibition by the three wheat inhibitors, XIP-I, TAXI-I and TAXI-II.

2. Materials and methods 2.1. Materials The Pichia pastoris expression kit including the P. pastoris strain (his4)/GS115 and the pCRR2.1-TOPOR TA cloning vector was from Invitrogen (Groningen, Netherlands). Restriction endonucleases and DNA modifying enzymes were from Promega (Madison, WI) and used according to the manufacturer’s recommendations. Pfu polymerase for polymerase chain reaction (PCR) was from Stratagene (Amsterdam Zuidoost, The Netherlands). Escherichia coli DH5 (supE44, hsdR17, recA1, endA1, gyrA96, thi-1, relA1) was used for DNA manipulation. Lysozyme, dinitrosalicylic acid and (1,4)-h-xylose were from Sigma; azo-wheat-arabinoxylan, wheat arabinoxylan low viscosity (LVX) and wheat arabinoxylan medium viscosity (MVX) were from Megazyme (International Ireland Ltd.). Oligonucleotides were synthesized by Eurogentec (Belgium). The expression vector pGEX-1ET, GSTrap glutathione-Agarose beads, DEAE-Sepharose and

thrombin protease were from Amersham-Pharmacia Biotech (Uppsala, Sweden); DyNAzyme was from Finnzymes (Espoo, Finland) and BugBusterR, E. coli BL21 (DE3) strain ( F , ompT, hsdS (r B , m B ), gal) and benzonaseR from Novagen. XynB gene (accession number AJ489605) was from in-house collection [3]. 2.2. Cloning, expression, and purification of E-XYNB in E. coli The XynB gene was used for expression of the P. funiculosum xylanase cDNA in E. coli. One intron was eliminated by overlap extension method [19] using the following primers: SOE1, 5V-AGTGCCCTTGCTGAGCGAAATTTGTTCAGAGCCGCT-3Vand its reverse complement antisense AOE2, 5VAGCGGCTCT GGACAAATTTCQ GCTCAGC-3Vand primers hybridizing with the extremities of the XynB cDNA: P1 5V-GAATTCGCTGAGGCGATCAACTACAACCAA-3V and P2 5V-GGATCCCTATTGGCACTGGCTGTAGTAAGCGTT-3V. XynB cDNA (200 ng) was used as a template in two independent PCR runs using DyNaZyme polymerase (1.5 units) and 0.375 mM dNTP. For the first round of PCR, P1 (80 pmol) and SOE1 primers were used for PCR I and equimolar amounts of P2 and AOE2 primers for PCR II. The annealing step was carried out at 55 8C for 2 min and the extension step at 72 8C for 3 min, in an overall volume of 100 Al and for a total number of 25 cycles in a MastercyclerR gradient thermocycler (Eppendorf). The resulting PCR products of 155 bp (PCR I) and 589 bp (PCR II) were gel-purified using the QIAQuick Gel Extraction Kit (Qiagen, Chatsworth, CA, USA) and used as templates (150 ng) in a second PCR run along with 1.5 unit Pfu and 0.375 mM dNTP. After five cycles of denaturation (1 min at 94 8C), annealing (2 min at 55 8C) and extension (3 min at 72 8C), 80 pmol of the forward P1 and reverse P2 primers was added and the reaction subjected to 25 cycles of denaturation (1 min at 94 8C), annealing (2 min at 55 8C) and extension (3 min at 72 8C). The final PCR product (744 bp) was gel-purified and ligated into pCRR2.1-TOPOR TA cloning vector for sequencing in order to confirm the integrity of the sequence (Genome Express, Meylan, France). The pGEX-1ET-derived expression plasmid with the cDNA insert encoding XYNB was constructed following standard procedures [20]. The cDNA fragment was isolated from the cloning vector by restriction endonuclease digestion with EcoRI, gel-purified using the Qiaquick purification kit (Qiagen), and ligated into EcoRI-pGEX-1ET expression vector to give pGEX-1ET/XynB. E. coli BL21 strain was transformed with pGEX-1ET/XynB and grown at 37 8C in Luria Bertani (LB) culture medium. Once the OD600nm value had reached 0.6, 1 mM isopropyl-1-thio-h-d-galactopyranoside (IPTG) was added to the culture medium to induce protein expression. After incubation at 200 rpm and 17 8C for 18 h, the cells were centrifuged at 3000 rpm for 20 min at 4 8C and stored at 80 8C until use.

A. Brutus et al. / Biochimica et Biophysica Acta 1701 (2004) 121–128

Bacterial lysis was performed in a BugBusterR buffer (5 ml g 1 of pellet) with benzonaseR (25 units ml 1 BugBusterR) and lysozyme (200 Ag ml 1 BugBusterR). The digestion was carried out at room temperature for 15 min. After a 12 000-rpm centrifugation for 20 min at 4 8C, the supernatant was mixed with glutathione-agarose beads and incubated for 15 min with 1 mM DTT. The beads were washed three times with 50 mM Tris–HCl at pH 8.0 and the fusion protein was released from the gel using the above buffer containing 40 mM reduced glutathione. Proteolysis of GST–XYNB was performed by incubation of the supernatant with thrombin (5 units ml 1) for 24 h at 22 8C in 20 mM Tris–HCl buffer pH 8.4, 150 mM NaCl, 2.5 mM CaCl2 and 1 mM DTT. 2.3. Cloning, expression, and purification of P-XYNB in P. pastoris The pHIL-S1-derived expression plasmid with the cDNA insert encoding XYNB was constructed following standard procedures [20]. The cDNA fragment was isolated from the cloning vector by restriction endonuclease digestion with BamHI/EcoRI, gel-purified using the Qiaquick purification kit (Qiagen), and ligated into pHIL-S1 EcoRI restriction site located at the start of P. pastoris phosphatase signal sequence to give pHIL-S1/XynB. E. coli strain DH5 was transformed according to the procedures described [20]. Transformants were grown in liquid bacterial cultures and recombinant plasmids isolated using Qiagen columns (MidiPrep kit, Qiagen) and checked for single insertion. The transformation of the P. pastoris strain (his4) GS115 [21] was achieved using the spheroplast method [22], as previously described [23]. Large-scale expression was achieved by growing the cells at 28 8C and 300 rpm in 500-ml BMGY until saturation followed by induction for 4 days in 100-ml BMM at 28 8C and 250 rpm, as previously described in Ref. [23]. The culture supernatant was dialyzed against the 20 mM Tris–HCl buffer at pH 7 overnight. The dialyzed sample was then loaded onto a DEAE-Sepharose column connected to FPLCR equipment (Pharmacia Biotech) and eluted with a linear gradient of 0–1 M NaCl in 20 mM Tris–HCl buffer at pH 7 and at a flow rate of 2 ml.min 1. Fractions containing xylanase activity were pooled. The protein concentration was determined using Bradford’s method with bovine serum albumin as the standard [24]. 2.4. Gel electrophoresis SDS-PAGE was performed in 12% (w/v) polyacrylamide gel as described by Laemmli [25] using a Pharmacia LMW electrophoresis calibration kit. For glycosylation determination, SDS-PAGE was stained for carbohydrate with periodic acid-Schiff (PAS) stain using a SIGMA glycoprotein staining kitR according to the manufacturer’s protocol.

123

2.5. Electrophoretic titration Titration curves of the P-XYNB alone or in combination with XIP-I or TAXI-I were produced using the Phast system as previously described [11]. Isoelectric focusing (IEF) 3–9 gels (Amersham Biosciences) were used according to the manufacturer’s instructions. Briefly, the carrier ampholytes contained in the IEF gel were prefocused at 2000 V for 150 V h to generate a stationary pH gradient (3 to 9). Prior to loading, XIP-I or TAXI-I (2.7 mg ml 1, 1.5 Al) was preincubated with P-XYNB (3.3 mg ml 1, 1.5 Al) in McIlvaine’s buffer, pH 5.5, for 10 min at room temperature. The samples were then applied across the middle of the gel, perpendicular to the pH gradient, using a single-well applicator. Electrophoresis perpendicular to the first dimension axis was run at 1000 V for 60 V h. For staining, the gels were fixed for 20 min in trichloroacetic acid (20%) and rinsed in phosphoric acid solution (3%) followed by staining with Serva Violet 17 stain (Serva Electrophoresis GmbH, Heidelberg, Germany) (1:1 ratio of 0.2% Serva Violet: 20% phosphoric acid) for 10 min. The gels were destained using phosphoric acid (3%). 2.6. N-terminal sequencing and molecular mass determination N-terminal amino acid sequencing of the Ponceau redstained protein after electro-transfer on a polyvinylidene difluoride membrane was performed by Edman degradation on an Applied Biosystems Model 494 protein gas-phase sequencer [26]. Phenylthiohydantoin amino acids were separated on a C18 reverse phase column (2.1250 mm) [27]. Molecular mass determination was performed by electrospray ionisation-mass spectrometry (ESI-MS) using a Quattro II instrument (Micromass, Marseille, France) and apomyoglobin as standard. 2.7. Purification of XIP-I, TAXI I and TAXI II XIP-I, TAXI I and TAXI II were purified as previously described [11,13]. 2.8. Enzyme and enzyme inhibition assays Routine assays during screening for the highest xylanase producer were performed using a colorimetric assay from Megazyme as described previously [28]. Xylanase activity was measured using the dinitrosalicylic acid (DNS) assay [8]. Aliquot of the xylanase (E-XYNB or P-XYNB) (20 Al) was mixed with 10 mg ml 1 LVX (Megazyme) in McIlvaine’s buffer, pH 5.5 (total volume reaction volume 200 Al) at 30 8C for 5 min. The reaction was terminated with the addition of 300-Al DNS reagent and boiled for 5 min. The reactions were cooled and centrifuged for 5 min at 12 000g and 200 Al was transferred to a microtitre plate. One unit of xylanase

124

A. Brutus et al. / Biochimica et Biophysica Acta 1701 (2004) 121–128

activity was defined as the amount of protein that released 1 Amol of xylose min 1. Optimal pH for xylanase activity was estimated using the DNS assay with LVX (10 mg ml 1) in McIlvaine’s buffer in a pH range of 1.0 to 7.0. Optimal temperature was also estimated using the same xylanase assay with LVX (10 mg ml 1) in McIlvaine’s buffer, pH 4.2, at temperatures ranging from 30 to 70 8C. Thermal stability was performed at pH 4.2 and 45 or 50 8C. For determination of apparent Michaelis–Menten constants, the initial velocities of the P-XYNB (50 nM) were measured at two conditions: pH 4.2/45 8C and pH 5.5/30 8C in McIlvaine’s buffer with LVX and MVX ranging from 0.75% to 1.8% (w/v). The effect of xylanase inhibitors from wheat (XIP-I, TAXI-I, and TAXI-II) on the activity of P-XYNB or E-XYNB was determined on LVX using the DNS activity assay at 45 8C, pH 4.2. The inhibition constant, K i, was calculated for the interaction between P-XYNB and the inhibitors (XIP-I and TAXI-I) using different LVX concentrations (0.75%, 1% and 1.5% for XIP-I and 1%, 1.25% and 1.5% for TAXI-I) and increasing amounts of inhibitor.

3. Results and discussion 3.1. Production of recombinant XYNB The XynB cDNA is predicted to encode a modular protein of 282 aa residues, which includes a 180-aa catalytic module, a 29-aa linker and a 34-aa binding module with significant similarity to the carbohydrate binding module characteristic of fungal CBM 1 [3]. The full-length XynB coding sequence was expressed in E. coli as a GST-XYNB fusion of 50 kDa, consistent with the sum of the calculated molecular masses for GST (26 kDa) and XYNB (25.5 kDa). The recombinant protein was induced in E. coli cells containing plasmid pGEX-XynB (Fig. 1A, lane 1) but not in E. coli cells containing plasmid pGEX-1ET (Fig. 1A, lane 2) nor in untransformed cell (Fig. 1A, lane 3). Optimum yields were obtained at 17 8C after 18-h induction. The GST-XYNB fusion protein was active, as measured in the soluble fraction of IPTG-induced cells (13F5 U ml 1). The GST-XYNB protein was purified from the supernatant by affinity chromatography. Proteins bound to the glutathioneagarose beads were eluted from the matrix using 40 mM reduced glutathione. A major band of 50 kDa, corresponding to GST-XYNB, was observed on SDS-PAGE (Fig. 1A lane 4). Attempt to remove the GST part of the fusion protein was performed by proteolytic digestion with thrombin. Following proteolysis, one major band of about 26 kDa was seen on SDS-PAGE corresponding to both free GST protein and recombinant XYNB (E-XYNB) (Fig. 1A lane 5). The digestion mixture was further subjected to anion-exchange chromatography or hydrophobic chromatography to isolate the recombinant protein but pure EXYNB could not be obtained, as confirmed by ESMS (not

Fig. 1. Electrophoresis analysis of recombinant XYNB. (A) Expression of XynB in E. coli and purification of GST-XYNB fusion protein. SDS-PAGE analysis of samples from cell extracts of E. coli BL21:pGEX-1ET/XynB (lane 1), BL21:pGEX-1ET (lane 2), untransformed E. coli BL21 (lane 3), GST-XYNB fusion protein after elution from GSH-agarose beads (lane 4), and thrombin cleaved fusion protein after 24 h digestion (lane 5). (B) Expression of XynB in P. pastoris and purification of P-XYNB. SDSPAGE analysis of culture supernatant (20 Al) of GS115:pHIL-S1/XynB at different days after induction (lanes 1–4), untransformed P. pastoris GS115 (lane 5), purified P-XYNB after DEAE-chromatography (lane 6), and purified P-XYNB stained using PAS reaction (lane 7).

shown), probably due to unspecific interactions between GST and E-XYNB. P. pastoris was thus used as an alternative host for the production of XYNB. The full-length XynB coding sequence was expressed using P. pastoris phosphatase signal sequence under the control of the AOX1 promoter. Highest secretion yields (up to 80 mg l 1) were obtained in buffered complex medium after 4 days of culture. The use of rich medium increased the activity/secretion of recombinant XYNB (P-XYNB) by a factor of 10 compared to the use of synthetic medium (data not shown). The medium composition, probably through a better control of pH and protease activity, thus greatly influenced the production of xylanase in the culture supernatant, as previously found with the A. niger xylanase also produced in P. pastoris [28]. After induction, a major specific protein band of about 43 kDa was observed on SDS-PAGE. This protein was absent from the supernatant of untransformed cells (Fig. 1B, lane 5). The amount of the 43-kDa protein increased with time of induction (Fig. 1B, lanes 1–4). Only trace amounts of other proteins were present in the culture supernatant. Highly purified P-XYNB was obtained using a single chromatography step (DEAE) and 70% of the activity was recovered. The N-terminal sequence Arg-Glu-Phe-Ala-Glu-Ala-Ile was in complete agreement with that expected from the

A. Brutus et al. / Biochimica et Biophysica Acta 1701 (2004) 121–128

Fig. 2. Thermostability of the recombinant XYNB. Relative activity of the E-XYNB (E) and P-XYNB (x) after incubation at 45 8C (—) and 50 8C (- - -).

recombinant protein, Arg corresponding to the C-terminal end of the signal peptide followed by Glu-Phe derived from the restriction site and Ala-Glu-Ala-Ile from XYNB Nterminal mature sequence. The purified recombinant enzyme migrated in SDS-PAGE as a single band (Fig. 1B, lane 6). The estimated size was around 43 kDa, thus much higher than the deduced size of 25902 Da, suggesting that the recombinant P-XYNB was glycosylated. The presence of sugars linked to the protein was confirmed by periodic acid Schiff staining (Fig. 1B, lane 7). No N-glycosylation site is present in XYNB amino acid sequence, suggesting that O-glycosylation is responsible for the higher molecular mass. Indeed, after reaction with trifluoroacetic acid [29], HPLC analysis showed that P-XYNB contains O-glycans composed of linked mannose structures (data not shown). The heterogeneity of the recombinant P-XYNB was further confirmed by ESI-MS (data not shown). The molecular mass of the recombinant P-XYNB protein ranged from 29 949 to 34 912 Da with incremented spikes of 162.6 Da corresponding to a mannose unit. XYNB shows a modular structure and high level of O-glycosylation is expected from the Ser/Thr rich linker separating the catalytic domain from the CBM [3].

125

the pH range 2.5 to 4.5 whereas the activity was lost at pH 1 and 6 (data not shown). At pH 4, the optimal temperature was observed between 35 and 45 8C for E-XYNB and between 45 and 65 8C for P-XYNB whereas the activity drastically decreased above 65 8C (data not shown). Both recombinant enzymes were stable at 45 8C for 60 min but lost activity at 50 8C (Fig. 2). GH11 xylanases exhibit pH optima spanning a remarkable range from approximately 2 to 11 [30]. From sequence comparisons of XYNB and the related GH11 xylanases, the likely proton donor and nucleophile are Glu117 and Glu206, respectively. There is a strong correlation in that the residue hydrogen-bonded to the general acid/base catalyst is Asn in the so-called balkalineQ xylanases (N5), whereas it is an Asp in those with a more acidic pH optimum (b5) [30,31]. Asp37 is conserved within the XYNB sequence, in agreement with the measured pH optimum range of the recombinant enzyme. The apparent kinetic parameters of P-XYNB were determined using LVX and MVX at pH and temperature optima (pH 4.2 and 45 8C) as well as 30 8C and pH 5.5. PXYNB demonstrated the same rate of hydrolysis using both conditions with a K m of 34.5F3.2 and 40F3 g l 1, and a k cat of 4966F333 and 5405F150 s 1 on MVX and LVX, respectively. 3.3. Enzyme inhibition by wheat xylanase inhibitors

3.2. Enzymatic properties of recombinant E-XYNB and P-XYNB

XYNB activity produced in E. coli or P. pastoris was inhibited by XIP-I and TAXI-I but not by TAXI-II (Fig. 3 and Fig. 4). The reduction in the rate of XYNB activity by XIP-I and TAXI-I occurred rapidly and preincubation of the P. funiculosum enzyme with the inhibitors did not lead to an increase in the observed reaction rate (not shown), indicating that neither XIP-I nor TAXI-I acts as a slow-binding inhibitor of this enzyme. Total inhibition of E-XYNB activity on LVX was achieved with 135 ng of XIP-I and 4.5 ng of TAXI-I (Fig. 3), respectively, whereas 44.4 ng of XIP-I and 1.5 ng of TAXI-I were required to completely abolish the activity of P-XYNB under the same conditions.

The enzymatic activity of recombinant E-XYNB using LVX as substrate was 13F5 U ml 1 whereas P-XYNB, produced in P. pastoris, showed an activity of 225F25 U ml 1, corresponding to a specific activity of 180 U mg 1. Although E-XYNB activity was very low, the enzyme was expressed in the soluble fraction, indicating that the recombinant enzyme was at least partially folded. XYNB shows a typical modular structure with the Ser/Thr-rich region separating the catalytic domain from the CBM. The present data suggest that glycosylation is involved in stabilizing the activity of XYNB, and that the linker region requires high glycosylation level to be functional. The effects of pH and temperature on the stability of the enzyme were investigated on both recombinant enzymes. Both E-XYNB and P-XYNB displayed optimum activity in

Fig. 3. Inhibition curve of E-XYNB xylanase. The xylanase activity of EXYNB (0.2 U) was determined on LVX as substrate in the presence of increasing concentrations of XIP-I (n), TAXI-I (E), and TAXI-II (.). Error bars represent FS.D. (n=3).

126

A. Brutus et al. / Biochimica et Biophysica Acta 1701 (2004) 121–128

Fig. 4. Kinetic analysis of the inhibition of P-XYNB by XIP-I and TAXI-I. The xylanase activity of P-XYNB was determined over a range of substrate (LVX) concentrations in the absence and presence of known concentrations of (A) XIP-I at 37.9 nM (n), 113.7 nM (E) and 151.6 nM (), and (B) TAXIQI at 11 nM (n), 16.5 nM (E) and 27.5 nM (). The data are presented as a double reciprocal plot with no inhibitor as control ( S ). The inset shows a secondary reciprocal plot of the slopes from the primary reciprocal plot versus [XIP-I] or [TAXIQI]. Reactions were performed in triplicate.

The kinetics of inhibition of P-XYNB by XIP-I and TAXI-I on LVX were further determined. In the presence of inhibitor, the V max did not change but the K m increased with increasing inhibitor concentrations of inhibitor protein (Fig. 4). Hence, the lines on the double reciprocal plot converge at the y axis, demonstrating that the inhibition is competitive irrespective of the inhibitor used, XIP-I or TAXI-I. The secondary plot of slope against inhibitor concentration gave an inhibition constant K i of 89.7F8.5 nM for XIP-I and 2.9F0.3 nM for TAXI-I (Fig. 4). The greater sensitivity of the enzyme to TAXI-I was in good agreement with the 30-fold difference observed for the inhibition of E-XYNB by XIP-I and TAXI-I (see above and

Fig. 3). These results are consistent with the general specificity reported for these inhibitors. XIP-type inhibitors are specific for fungal xylanases; the inhibition is competitive with K i values ranging from 3.4 to 610 nM and does not show any preference for GH10 or GH11 xylanases [11]. In contrast, TAXI-type inhibitors are specific for fungal or bacterial GH11 xylanases and the specificity towards these enzymes varies between TAXI-I and TAXI-II. For example, although TAXI-I inhibits A. niger xylanase to a greater extent than B. subtilis enzyme, of these two enzymes TAXIII inhibits only B. subtilis xylanase [13]. In contrast to the present results, another GH11 xylanase from P. funiculosum, XYNC, was previously reported to be inhibited by the three

A. Brutus et al. / Biochimica et Biophysica Acta 1701 (2004) 121–128

127

Fig. 5. Electrophoretic titration curves of P-XYNB in the presence of XIP-I or TAXI-I. (A) P-XYNB (4 Ag), (B) XIP-I (4 Ag), (C) a mixture containing PXYNB (5 Ag) and XIP-I (4 Ag) (4 Ag), (D) TAXI-I (6 Ag), (E) a mixture containing P-XYNB (5 Ag) and TAXI-I (6 Ag). The pH mobility curves of the proteins, free or in complex, are indicated with an arrow.

wheat inhibitors, XIP-I, TAXI-I and TAXI-II [4]. The inhibition was competitive, with K i of 3.5, 16 and 17 nM, for XIP-I, TAXI-I and TAXI-II, respectively. In the present study, competitive inhibition was also observed between XYNB and TAXI-I and the low K i (3.2 nM) indicates very tight binding. However, XYNB, as previously reported for A. niger xylanase, was not inhibited by TAXI-II. Sequence comparisons between GH11 xylanases from A. niger, B. subtilis, and P. funiculosum (XYNB and XYNC) did not reveal any striking features explaining the lack of inhibition of A. niger and P. funiculosum XYNB xylanases by TAXIII. Among all the enzymes tested with TAXI-II, XYNB is the only one possessing a modular structure, but this does not seem to influence the interaction since A. niger xylanase which displays a similar spectrum of inhibition is a single domain protein. Alternatively, it might be that the high level of glycosylation of the recombinant enzyme indirectly influences the specificity of the interaction. However, EXYNB, produced in E. coli, thus unglycosylated, showed the same inhibition specificity as P-XYNB, which tends to suggest that glycosylation might affect the strength but not the specificity of the inhibition. The relative affinities and pH dependencies of the interaction of XIP-I and TAXI-I with P-XYNB were further studied using titration curves (Fig. 5). The complex between P-XYNB and the inhibitors was observed over the entire pH range (3–9; the range covered on the gel), as previously observed with A. nidulans xylanases (pI 3.4) with XIP-I [11] and B. subtilis xylanase (pI 9.3) with TAXI-I [32]. In contrast, it was previously shown that A. niger xylanase, which showed a weaker binding to XIP-I, only interacted with the inhibitor at pH 4–7 [11]. These results demonstrate the stability of the interaction between P-XYNB and the wheat inhibitors in this range of pH. In conclusion, the present work demonstrates that P. funiculosum produces enzymes with different specificities

towards wheat xylanase inhibitors. XIP-I, TAXI-I and TAXI-II might have evolved to counteract the many xylanases secreted by bacteria and fungi. This multispecificity is also observed with other plant proteinaceous inhibitors of carbohydrate-active enzymes such as polygalacturonase-inhibiting proteins from monocots and dicots, which have evolved a structural variability leading to a diversification of their specificity of recognition against the multiple polygalacturonase isozymes secreted by fungi [33]. This diversification is likely to ensure to a plant a higher level of protection and to confer a selective advantage against pathogens.

Acknowledgements The authors gratefully acknowledge the support of Neville Fish (Genencor International (formely Rhodia Food UK) and the financial support of the Commission of the European Communities, specific RTD program bQuality of Life and Management of Living ResourcesQ, Key Action 1Health Food and Environment, Project no. QLRT-2000-811 GEMINI bSolving the problem of glycosidase inhibitors in food processingQ.

References [1] M.P. Coughlan, G.P. Hazlewood, Beta-1,4-d-xylan-degrading enzyme systems: biochemistry, molecular biology and applications, Biotechnol. Appl. Biochem. 17 (1993) 259 – 289. [2] P.M. Coutinho, B. Henrissat, in: H.J. Gilbert, G.J. Davies, B. Henrissat, B. Svensson (Eds.), Recent Advances in Carbohydrate Bioengineering, The Royal Society, Cambridge, 1999, pp. 3 – 12. [3] M.J. Alcocer, C.S. Furniss, P.A. Kroon, M. Campbell, D.B. Archer, Comparison of modular and non-modular xylanases as carrier proteins for the efficient secretion of heterologous proteins from Penicillium funiculosum, Appl. Microbiol. Biotechnol. 60 (2003) 726 – 732.

128

A. Brutus et al. / Biochimica et Biophysica Acta 1701 (2004) 121–128

[4] C.S Furniss, N.J. Belshaw, M.J. Alcocer, G. Williamson, G.O. Elliott, K. Gebruers, N.P. Haigh, N.M. Fish, P.A. Kroon, A family 11 xylanase from Penicillium funiculosum is strongly inhibited by three wheat xylanase inhibitors, Biochim. Biophys. Acta 1598 (2002) 24 – 29. [5] J.F. Sbrensen, K.M. Kragh, O. Sibbesen, J.A. Delcour, H. Goesaert, B. Svensson, T.A. Tahir, J. Brufau, A.M. Perez-Vendrell, D. Bellincampi, R. D’Ovidio, L. Camardella, A. Giovane, E. Bonnin, N. Juge, Potential role of glycosidase inhibitors in industrial biotechnological applications, Biochim. Biophys. Acta 1696 (2004) 275 – 287. [6] N. Juge, F. Payan, G. Williamson, XIP-I, a xylanase inhibitor from wheat: a novel protein function, Biochim. Biophys. Acta 1696 (2004) 203 – 211. [7] K. Gebruers, K. Brijs, C.M. Courtin, K. Fierens, H. Goesaert, A. Rabijns, G. Raedschelders, J. Robben, S. Sansen, J.F. Sbrensen, S. Van Campenhout, J.A. Delcour, Properties of TAXI-type endoxylanase inhibitors, Biochim. Biophys. Acta 1696 (2004) 213 – 221. [8] W.R. McLauchlan, M.T. Garcia-Conesa, G. Williamson, M. Roza, P. Ranestein, M. Jan, A novel class of protein from wheat which inhibits xylanases, Biochem. J. 338 (1999) 441 – 446. [9] G.O. Elliott, R.K. Hughes, N. Juge, P.A. Kroon, G. Williamson, Functional identification of the cDNA coding for a wheat endo-1,4beta-d-xylanase inhibitor, FEBS Lett. 519 (2002) 66 – 70. [10] F. Payan, R. Flatman, S. Porciero, G. Williamson, N. Juge, A. Roussel, Structural analysis of xylanase inhibitor protein I (XIP-I), a proteinaceous xylanase inhibitor from wheat (Triticum aestivum, var Soisson), Biochem. J. 372 (2003) 399 – 405. [11] R. Flatman, W.R. McLauchlan, N. Juge, C.S. Furniss, J.G. Berrin, R.K. Hughes, P. Manzanares, J.E. Ladbury, R. O’Brien, G. Williamson, Interactions defining the specificity between fungal xylanases and the wheat proteinaceous inhibitor, XIP-I, Biochem. J. 365 (2002) 773 – 781. [12] W. Debyser, W.J. Peumans, E.J.M. Van Damme, J.A. Delcour, Triticum aestivum Xylanase Inhibitor (TAXI), a new class of enzyme inhibitor affecting breadmaking performance, J. Cereal Sci. 30 (1999) 39 – 43. [13] K. Gebruers, W. Debyser, H. Goesaert, P. Proost, J. Van Damme, J.A. Delcour, Triticum aestivum L. endoxylanase inhibitor (TAXI) consist of two inhibitors, TAXI 1 and TAXI 2, with different specificities, Biochem. J. 353 (2001) 239 – 244. [14] K. Gebruers, K. Brijs, C.M. Courtin, H. Goesaert, P. Proost, J. Van Damme, J.A. Delcour, Affinity chromatography with immobilised endoxylanases separates TAXI- and XIP-type endoxylanase inhibitors from wheat (Triticum aestivum L.), J. Cereal Sci. 36 (2002) 367 – 375. [15] G.O. Eliott, W.R. McLauchlan, G. Williamson, P.A. Kroon, A wheat xylanase inhibitor protein (XIP-I) accumulates in the grain and has homologues in other cereals, J. Cereal Sci. 37 (2003) 187 – 194. [16] H. Goesaert, K. Gebruers, K. Brijs, C.M. Courtin, J.A. Delcour, TAXI type endoxylanase inhibitors in different cereals, J. Agric. Food Chem. 51 (2003) 3770 – 3775. [17] H. Goesaert, K. Gebruers, K. Brijs, C.M. Courtin, J.A. Delcour, dXIPT-type endoxylanase inhibitors in different cereals, J. Cereal Sci. 38 (2003) 317 – 324.

[18] H. Goesaert, G. Elliott, K. Gebruers, C.M. Courtin, J.A. Delcour, N. Juge, Occurrence of proteinaceous endoxylanase inhibitors in cereals, Biochim. Biophys. Acta 1696 (2004) 193 – 202. [19] S.N. Ho, H.D. Hunt, R.M. Horton, J.K. Pullen, L.R. Pease, Sitedirected mutagenesis by overlap extension using the polymerase chain reaction, Gene 77 (1989) 51 – 59. [20] J. Sambrook, E.F. Fritsch, T. Maniatis, Cloning: A Laboratory Manual, 2nd ed., Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, 1989. [21] J.M. Cregg, K.J. Barringer, A.Y. Hessler, K.R. Madden, Pichia pastoris as a host system for transformations, Mol. Cell. Biol. 5 (1985) 3376 – 3385. [22] A. Hinnen, J.B. Hicks, G.R. Fink, Transformation of yeast, Proc. Natl. Acad. Sci. U. S. A. 75 (1978) 1929 – 1933. [23] N. Juge, G. Williamson, A. Puigserver, N.J. Cummings, I.F. Connerton, C.B. Faulds, High-level production of recombinant Aspergillus niger cinnamoyl esterase (FAEA) in the methylotrophic yeast Pichia pastoris, FEMS Yeast Res. 1 (2001) 127 – 132. [24] M.M. Bradford, A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding, Anal. Biochem. 72 (1976) 248 – 254. [25] U.K. Laemmli, Cleavage of structural proteins during the assembly of the head of bacteriophage T4, Nature 227 (1970) 680 – 685. [26] P. Matsudaira, Sequence from picomole quantities of proteins electroblotted onto polyvinylidene difluoride membranes, J. Biol. Chem. 262 (1987) 10035 – 10038. [27] R.M. Rewick, M.W. Hunkapiller, L.E. Hood, W.J. Dreyer, A gas– liquid solid phase peptide and protein sequenator, J. Biol. Chem. 256 (1981) 7990 – 7997. [28] J.G. Berrin, G. Williamson, A. Puigserver, J.-C. Chaix, W.R. McLauchlan, N. Juge, High-level production of recombinant fungal endo-b-1,4-xylanase in the methylotrophic yeast of Pichia pastoris, Protein Expr. Purif. 19 (2000) 179 – 187. [29] M.R. Hardy, R.R. Townsend, Y.C. Lee, Monosaccharide analysis of glycoconjugates by anion exchange chromatography with pulsed amperometric detection, Anal. Biochem. 170 (1988) 54 – 62. [30] M.D. Joshi, G. Sidhu, I. Pot, G.D. Brayer, S.G. Withers, L.P. McIntosh, Hydrogen bonding and catalysis: a novel explanation for how a single amino acid substitution can change the pH optimum of a glycosidase, J. Mol. Biol. 2999 (2000) 255 – 279. [31] A. Sapag, J. Wouters, C. Lambert, P. de Ioannes, J. Eyzaguirre, E. Depiereux, The endoxylanases from family 11: computer analysis of protein sequences reveals important structural and phylogenetic relationships, J. Biotechnol. 95 (2002) 109 – 131. [32] K. Gebruers, H. Goesaert, K. Brijs, C.M. Courtin, J.A. Delcour, Purification of TAXI-like endoxylanase inhibitors from wheat (Triticum aestivum L.) whole meal reveals a family of iso-forms, J. Enzyme Inhib. Med. Chem. 17 (2002) 61 – 68. [33] R. D’Ovidio, B. Mattei, S. Roberti, D. Bellincampi, Polygalacturonases, polygalacturonase inhibiting proteins and pectic oligomers in plant–pathogen interactions, Biochim. Biophys. Acta 1696 (2004) 237 – 244.