Screening of Xylanolytic Aspergillus fumigatus for Prebiotic Xylooligosaccharide Production Using Bagasse

Xylans present in lignocellulosic materials have been studied for obtaining xylooligosaccharides (XOS) from waste materials such as corncobs, rice hulls, olive pits, barley straw (2), tobacco stalk, cott on stalk, sunfl ower stalk, wheat straw (3) and sugarcane bagasse (4–6). Sugarcane bagasse is inexpensive, renewable and abundant source of XOS especially in the countries that produce ethanol and sugar from sugarcane. However, more research is necessary to improve the use of this residue.


Introduction
Currently, lignocellulosic waste is a topic of global studies, given that fossil fuel reserves are diminishing, the new agricultural frontiers are limited and the demand for food and biofuels is increasing by the growing world population (1).For these reasons, technology must advance to improve the use of agricultural and agroindustrial residues to obtain food and biofuel.
Xylans present in lignocellulosic materials have been studied for obtaining xylooligosaccharides (XOS) from waste materials such as corncobs, rice hulls, olive pits, barley straw (2), tobacco stalk, cott on stalk, sunfl ower stalk, wheat straw (3) and sugarcane bagasse (4)(5)(6).Sugarcane bagasse is inexpensive, renewable and abundant source of XOS especially in the countries that produce ethanol and sugar from sugarcane.However, more research is necessary to improve the use of this residue.
Xylooligosaccharides are oligomers obtained from the hydrolysis of xylan extracted from lignocellulosic materials.These oligosaccharides are considered a new soluble dietary fi bre due to their low caloric value and prebiotic eff ect (2).They behave as nondigestible oligosaccharides, i.e. they are not degraded in the stomach and thus reach the colon intact.XOS can benefi cially aff ect humans by modulating the colon microbiota, especially bifi dobacteria and lactobacilli (7).The addition of XOS to food has excellent physiological eff ects on the organism, including improvement of bowel function, calcium absorption, prevention of dental caries, protection against cardiovascular disease and reduction of colon cancer risks due to the formation of smaller chain fatt y acids (8,9).In addition, they contribute to benefi cial eff ects related to skin and blood, immunological action, antioxidant activities, anti-infl ammatory and antiallergenic eff ects (1,10).
High quality XOS can be produced enzymatically using xylanases from a variety of microorganisms including: Aspergillus, Thermoascus, Trichoderma, Streptomycetes, Phanerochaetes, Chytridiomycetes, Ruminoccocus, Penicillium, Fibrobacter, Clostridium, Pichia and Bacillus (1,(10)(11)(12)(13).However, the search for more effi cient xylanase-producing strains is necessary considering the production costs and low yields of production.These are the major problems of the use of enzymes for industrial applications.
Xylanases have att racted considerable att ention not only for their potential application in lignocellulose hydrolysis and their bioconversion into sugars (14), but also for juice clarifi cation, vegetable oil extraction, improvement of animal digestion, fl our for baked goods, and bleaching paper pulp (15)(16)(17).
The use of residues like bagasse for growth of xylanase-producing microorganisms or substrate for enzymatic XOS production could decrease the costs of XOS (14,18).Furthermore, the bioconversion of these substrates can help in the reduction of the environmental impact caused by the accumulation of waste (19).
The conditions for maximum production of xylanase are highly dependent on the microorganism, bioprocess and culture medium.Cellulosimicrobium cellulans produced only 0.7 U/mL of xylanase aft er 3 days in a culture medium containing sugarcane bagasse (18).Penicillium janczewskii cultured at 28 °C for 7 days in a medium with nine diff erent agroindustrial wastes confi rmed the infl uence of the substrate on xylanase production because the best substrate (15.4U/mL) was wheat bran, followed by oat bran (5.8 U/mL), corn cobs (5.3 U/mL), barley grain (4.9 U/mL) and cane bagasse (3.1 U/mL) (20).
In this work, a method of fungal screening using sugarcane bagasse as the sole carbon source is developed focusing on the production of xylanases and xylooligosaccharides.The enzymes of the most promising cultures were evaluated in enzymatic reaction for XOS production using xylan from bagasse.

Isolation and selection of xylanase-producing strains
Samples of sugarcane bagasse were collected in Assis municipality in the western state of São Paulo, Brazil.The collected material was resuspended in peptone water 0.1 % (by mass per volume) and aseptically plated on the cul-ture medium.The optimal temperature for fungal growth on solid medium in Petri dishes was determined by incubation at 28, 35 and 40 °C for 120 h, in an incubator with a humidifi cation system.The diameters of colonies were measured for assessment of growth aft er 1 to 3 days of cultivation.
The strains were isolated from sugarcane bagasse and maintained as stock cultures at 7 °C on potato dextrose agar (Acumidea, Lansing, MI, USA).Only the microorganisms that were the best producers of xylanases were identifi ed in Chemical Biological and Agricultural Pluridisciplinary Researcher Center (CPQBA), Campinas State University, SP, Brazil.

Methodology of strain identifi cation
The strains were identifi ed in four steps: (i) DNA extraction according to the protocol described by Raeder and Broda ( 21); (ii) amplifi cation of the region ITS1-5.8S--ITS2and the identifi ed calmodulin gene was done directly by DNA extraction from the samples using PCR.The primers (synthetic oligonucleotides) used in PCR reactions were ITS-1 and ITS-4 (for the amplifi cation of ITS region) and CF-1 and CF-4 (for the amplifi cation of calmo dulin gene); (iii) the amplifi ed fragments (from primers ITS-1, ITS-4, CF-1 and CF-4) were purifi ed and sequenced (automatic sequencer 3500XL series; Applied Biosystems, São Paulo, Brazil); (iv) the partial sequences of ITS and calmo dulin gene obtained from diff erent primers were assembled in a contig and compared with the sequences of GeneBank (Bethesda, MD, USA) and CBS (Utrecht, The Netherlands).The sequences were aligned using Clustal W program (22) and philogenetic analysis was performed by MEGA program (23).The matrices of evolutionary distance were calculated using the model developed by Kimura (24) and the phylogenetic tree was made by applying neighbor-joining method (25).
The strains M51 and U2370 were identifi ed as Aspergillus fumigatus and Aspergillus fumigatus Fresenius 1863, respectively.The strain Trichoderma reesei CCT 2768 was acquired from the culture collection named Tropical Culture Collection (CCT), André Tosello Foundation, Campinas, SP, Brazil.Aspergillus fumigatus M51 was deposited in CCT with the code CCT 7732.

Culture medium
The culture medium for the isolation of xylanolytic microorganisms was prepared with (in mass %): sugarcane bagasse 1.0, (NH 4 ) 2 SO 4 0.

Cultivation and enzyme production
Xylanases were produced by submerged fermentation (SmF) in Erlenmeyer fl asks (250 mL) containing 50 mL of medium composed of sugarcane bagasse 3.0 % (by mass per volume) and (in mass %): and ZnSO 4 0.02) at pH=5.0.The culture medium was inoculated with 10 6 spores per mL counted by microscopy in a Neubauer chamber.The fl asks were incubated at the optimal growth temperature with orbital shaking (model TE421; Tecnal, São Paulo, Brazil) at 180 rpm for 144 h.The biomass was separated by fi ltering through gauze and fi lter paper.The fi ltrate was used as a crude xylanase complex.Fermentations were carried out in triplicate.

Enzymatic activities
Two tests were performed to determine xylanase and β-xylosidase activities.Enzyme activity was assayed at 50 °C in a reaction mixture containing 0.1 mL of diluted crude enzyme and 0.65 mL of substrate solution in 0.25 M sodium acetate buff er, pH=5.0.The used substrate was 0.5 % (by mass per volume) birchwood xylan (Sigma-Aldrich, Darmstadt, Germany).The amount of reducing sugars was quantifi ed by the dinitrosalicylic acid method (26).One unit (U) of xylanase activity was defi ned as the amount of enzyme that releases carbohydrates having a reducing power corresponding to 1 μmol of d-xylose from birch xylan per minute under assay conditions.When 4-ni trophenyl-β-d-xylopyranoside (PNPX) was used as a substrate, the β-xylosidase activity was measured in a mixture containing 0.25 mL of 100 mM sodium acetate, pH= 5.0, 0.25 mL of 4 mM substrate solution, and 0.05 mL of crude enzyme.Aft er 10 min of incubation at 50 °C the reaction was stopped by adding 2 mL of 2 M sodium carbonate, and the released p-nitrophenol was quantifi ed spectrophotometrically at 410 nm.One unit (U) of β-xy losidase activity was defi ned as the amount of enzyme that releases 1 μmol of p-nitrophenol per minute in the reaction mixture (27).

Enzyme characterisation
Optimum pH and temperature for enzyme activity A 2 2 full factorial design with four replicates at the midpoint was used to evaluate the infl uence of two diff erent variables, temperature and pH.They were studied to determine their eff ect on xylanolytic activity of fungal crude enzymes.The xylanase activities (U/mL) were taken as dependent variables or response of the experimental design.To fi t an empirical second-order polynomial model, a central composite design was used.The results were analysed by the response surface analysis using STATISTICA v. 6.0 soft ware (StatSoft Inc., Tulsa, OK, USA).

Thermostability
The crude enzyme solution was incubated at various temperatures (40-95 °C) for 1 and 3 h at pH=7.0 in sealed tubes to prevent evaporation.Water was used instead of crude enzyme as a control.In both assays, an aliquot was removed and placed on ice before assaying for residual enzyme activity at the optimal pH and temperature.

Extraction and chemical characterisation of hemicellulose
The method used for the extraction of hemicellulose was described by Zilliox and Debeire (28) and Akpinar et al. (29), with slight modifi cations.A sample of 20 g of sugarcane bagasse was swollen in water at 60 °C for 16 h.Then it was treated for 3 h at 35 °C with 24 % KOH including 1 % (by mass per volume) NaBH 4 .The extract was fi ltered through a gauze, and the xylan present in the supernatant was precipitated in 2 volumes of cold ethanol and 0.2 volume of acetic acid, then washed tree times with 70 % ethanol and centrifuged at 5300×g for 20 min.Aft er centrifugation, the solid was dried using an air circulation oven at 45 °C.The yield of crude hemicellulose (precipitated material extracted from bagasse) was 25.4 %, but only 71.9 % of this material was pure hemicellulose.Therefore, the yield of the pure hemicellulose extracted from bagasse was 18.3 %.

Production of XOS by enzymatic reaction of xylanases
The enzymatic reaction was performed using the mixture of 60-500 U of xylanases (NS 22083; Novozymes, Bagsvaerd, Denmark) or crude enzymes from selected fungi per g of substrate in a total volume of 10 mL of 100 mM acetate buff er (pH=5.0) in a 20-mL glass test tube.The hemicellulose 1-2 % (by mass per volume) extracted from bagasse (as previously described) was used as substrate.According to Novozyme, the commercial xylanase (NS22083) was a purifi ed endoxylanase with optimal conditions of 35-55 °C and pH=4.5-6.0.The mixture was incubated at (50±1) °C for 1-96 h in a shaker (model TE 405; Tecnal) at 130 rpm.The hydrolysis was stopped by boiling in a water bath for 10 min.The xylooligosaccharides released during the reaction were analysed by HPLC (ICS-5000; Dionex, Sunnyvale, CA, USA).The experiment was performed in triplicate.Statistical analysis was performed to compare XOS production by the xylanases of diff erent fungi and with the commercial xylanase using the ANOVA and Tukey's tests in the STATISTICA v. 6.0 soft ware.

Determination and yield of xylooligosaccharides
The carbohydrates formed during enzymatic hydrolysis were quantifi ed using an anion exchange column, CarboPac PA100 on a HPLC (ICS-5000; Dionex).The elution was done with 0.2 M sodium hydroxide and 0.5 M sodium acetate with linear gradient (0-20 %) for 10 min, followed by a wash step with 0.5 M sodium acetate for 5 min.Finally, the solution was stabilized with 0.2 M sodium hydroxide for 7 min at a fl ow rate of 1 mL/min.The

Isolation and selection of xylanolytic fungi
Xylanolytic fungi were selected from 138 strains isolated from agar sugarcane culture medium by plating from samples.Only nine strains (Table 1) showed colonies with diameters of 0.8 cm or larger, and they were selected for the production of xylanase by submerged fermentation (SmF), showing activity from 8.9 to 43.7 U/mL.

Production of xylanases in submerged fermentation
The best producers of xylanases were identifi ed as Trichoderma reesei CCT 2768 (43.7 U/mL), Aspergillus fumigatus M51 (35.6 U/mL) and A. fumigatus U2370 (28.5 U/ mL) using sugarcane bagasse as the sole carbon source for 5 days of incubation (Table 1).The production of β --xylosidase by these fungi was very low when compared with the xylanase activity.The maximum β-xylosidase activity was 0.05 U/mL (Table 1).

Eff ects of temperature and pH of the reaction on the xylanase activity
The enzymatic reaction of selected fungi was carried out at 50 °C and pH=5.0;however, the optimisation of these parameters was determined to obtain higher xylanolytic activity for each crude enzyme.
The values obtained from 2 2 full factorial design, with coded and uncoded values, and results for xylanase activity of T. reesei CCT 2768, A. fumigatus M51 and A. fumigatus U2370 are shown in Table 2.
The optimal pH and temperature values were calculated by deriving equations from the second order uncoded model, which describes the relationship of these dependent variables and the enzyme activity (independent variable).The following equations refer to the xylanase activity in the extracts of T. reesei CCT 2768, A. fumigatus U2370 and A. fumigatus M51, respectively: Xylanase activity=46.7-1.7•t-19.0•pH 2 -17.9•t 2 R 2 = 9.0 % /2/ Xylanase activity=26.5-10.9•pH 2  2768, A. fumigatus U2370 and A. fumigatus M51.On the other hand, at pH below 3.5 the enzyme activity decreased considerably, obtaining less than 50 % of the maximum enzyme activity (Fig. 1).This result indicates that extremely acidic conditions are not recommended for these enzymes.Under slightly alkaline conditions (pH=7.5), the xylanase activity in the crude enzyme from A. fumigatus U2370 retained 71 % of the maximum activity (Fig. 1b), but less than 55 % of the xylanolytic activity of the crude enzymes from T. reesei CCT 2768 and A. fumigatus M51 was obtained (Figs.1a and c).
The thermostability of the crude enzymes from T. reesei CCT2768, A. fumigatus M51 and A. fumigatus U2370 was similar, and 100 % of the xylanase activity at 10 °C was maintained for 1 to 3 h at 25 to 40 °C.At 50 °C for 1 h, 64, 60 and 61 % of the xylanolytic activity of the enzymes from these three fungi at 10 °C were maintained, respectively.When they were incubated at 60 °C for 1 h, only 3, 7 and 11 % of their activities at 10 °C remained, respectively.Finally, at higher temperatures (70-95 °C) the enzymes from the three fungi had completely lost their xylanolytic activity in 1 h of incubation (Fig. 2a).However, aft er 3 h at 50 °C, only the crude enzymes from T. reesei CCT 2768 showed a higher decrease (28 %) of the original activity at 10 °C.The enzymes from A. fumigatus M51 and A. fumigatus U2370 maintained 54 and 56 % of their original activity, respectively.In addition, at 40 °C the three enzymes remained 100 % stable during 3 h of incubation (Fig. 2b).These results demonstrated a higher thermostability of the xylanases of A. fumigatus than the enzyme of T. reesei CCT 2768.
The pH stability of enzymes from T. reesei CCT 2768, A. fumigatus U2370 and A. fumigatus M51 showed more than 87 % stability under acidic conditions in the pH range from 4 to 6.Even in the range of pH=7 to 9, between 70 and 100 % of the original activity remained stable.However, at pH=10 the stability was greatly reduced (more than 82 %) (Fig. 2c).

Production of XOS by enzymatic hydrolysis
The enzymes from the three selected fungi were compared to the commercial endoxylanase (NS 22083) for XOS production by the enzymatic hydrolysis of hemicellulose extracted from sugarcane bagasse.The total XOS production of the xylanase from A. fumigatus M51 (1.04 mg/mL) and T. reesei CCT 2768 (0.88 mg/mL) was significantly higher (p<0.05)than the xylanase from A. fumigatus U2370 (0.54 mg/mL).The commercial xylanase also had lower activity (0.50 mg/mL).
Another important aspect was the type of xylooligosaccharide produced.The reactions with the enzyme from A. fumigatus M51 showed a signifi cantly higher concentrations (p<0.05) of xylobiose (0.59 mg/mL) and xylotriose (0.45 mg/mL) than of xylotetraose, xylopentaose and xylohexaose (less than 0.01 mg/mL each).
On the other hand, the commercial xylanase Novozymes NS 22083 produced signifi cantly lower amount (p<0.05) of xylobiose (0.51 mg/mL) and xylotriose (0.08 mg/mL) than the xylanases from A. fumigatus M51.The smallest fraction of the sugars obtained from the hemicellulose was again of xylose; only 0.04 mg/mL was produced by A. fumigatus M51 and 0.10 mg/mL by T. reesei CCT 2768 (Table 3).
The highest and signifi cant (p<0.05)yield of XOS was obtained from the crude enzyme from A. fumigatus M51 (14.7 %), followed by T. reesei CCT 2768 (11.9 %) in just 3 h of reaction time.The commercial enzyme had much lower XOS yield when compared to the enzymes from A. fumigatus M51.According to these results, the enzymes from A. fumigatus M51 showed great potential for use in the production of short-chain XOS, since this crude enzyme was superior to the purifi ed Novozymes xylanase (NS 22083) and other fungal enzymes evaluated under the same conditions and enzyme amount.The improvement in XOS yield was obtained in another assay (Fig. 3) performed using diff erent reaction times and dosages of enzymes from A. fumigatus M51.There was no signifi cant diff erence (p<0.05) in XOS production with 120 and 500 U/g hemicellulose, and the maximum level of XOS yield was obtained (35-37.6 %) in 48-72 h with 2 % hemicelullose.

Discussion
Microorganisms selected in this work exhibited weak β-xylosidase activity, which is favourable for XOS pro-duction, since lower xylose production means weaker hydro lysis of XOS released by xylanases.According to Vázquez et al. (31) xylanases are inhibited by xylose at higher concentrations.
The crude xylanases from A. fumigatus M51 were stable in a wide range of pH=4-9, in which they maintained between 70 and 100 % of the original activity.Therefore, these enzymes can be used in industry under diff erent conditions including acidic, neutral and alkaline environment.A lower range of optimal pH values (4.0-6.0) for xylanase activity was found in other fungi, e.g.Schizophyllum commune ATCC 38548, A. awamori and Aspergillus sp.(33,34).Xylanases from A. caespitosus remained stable in the pH range of 5-7 (35).On the other hand, the best xylanase activity produced by A. foetidus MTCC 4898 was at pH=5.3, but decreased by 34 and 50 % at pH=4.5 and 6.0, respectively (36).The xylanases from A. fumigatus M51 and A. fumigatus U2370 showed higher thermostability at 10 to 50 °C for 1 to 3 h of incubation than the xylanase from T. reesei CCT 2768 (10 to 40 °C).The same result was obtained for Aspergillus sp.FP-470 (33).However, an inferior thermostability (10-40 °C) was found for xylanases of Schizophyllum commune ATCC 38548 (34).
The XOS yield obtained by enzymatic reaction can be infl uenced by several factors including the type and concentration of xylanase, source and concentration of hemicellulose, xylan composition, type of pretreatment and reaction time, among others (1,3).In the present work, the eff ects of enzyme concentration and reaction time were important to increase XOS yield up to 31-36 % in 48 h with 60-500 U of xylanase per g of substrate (Fig. 3).A similar XOS yield (37.1 %) was obtained by crude enzyme from Thermoascus aurantiacus with 2 % hemicellulose from cane bagasse, 60 U per g of substrate, but with longer reaction time (96 h) at 50 °C and pH=5.0 (11).Another similar XOS yield (36.8 %) was obtained with the reaction of hemicellulose from Populus tomentosa and 25 U of crude xylanase from Pichia stipites per g of substrate at 50 °C and pH=5.4,but in only 14 h (13).However, a low yield of XOS (8.6 %) was obtained when only 13.3 U of endoxylanase from Trichoderma viridae (Sigma-Aldrich) per g of substrate were added (37).In addition, a XOS yield of only 11.4 % was obtained with hemicellulose extracted from tobacco stalk, aft er 24 h of hydrolysis using 20 U of xylanase from A. niger per g of substrate at 40 °C (3).
From the yield of hemicellulose extracted from bagasse (18.3 %), and yield of XOS from hemicellulose (37.6 %; Fig. 3), is possible to obtain 68.8 kg of prebiotic XOS per metric tonne of sugarcane bagasse by the above mentioned enzymatic reaction.In addition, in a biorefi nery mo del, aft er extraction of hemicellulose used for XOS pro duction, the residual cellulose can be used for the production of second-generation ethanol in the same industrial plant.
The enzymatic method resulted in the production of XOS with degree of polymerisation (DP) of 2-3, while chemical hydrolysis (autohydrolysis or acid hydrolysis) produced these, but also other XOS with higher DP (2,(38)(39)(40), without prebiotic eff ect.Commercial XOS (Xylooligo 95P; Suntory, Osaka, Japan), produced by autohydrolysis, containing 83 % of xylobiose and xylotriose were used in the culture medium as carbon source.These XOS were responsible for higher growth of Bifi dobacterium strains (B.adolescentis and B. longum) than the culture with the medium formulated with XOS with higher DP and containing only 24-41 % of xylobiose and xylotriose.Furthermore, XOS with DP=5-6 reduced the degree of consumption of these oligosaccharides by the bacteria (41).Therefore, the quality of the produced XOS is important when considering the prebiotic eff ect.In the present work, xylobiose and xylotriose were predominant XOS in the culture madium, which made it more suitable for selective Bifi dobacterium growth.

Conclusion
The strain Aspergillus fumigatus M51 was selected by a screening method developed to obtain fungal xylanases with the ability to produce xylooligosaccharides (XOS) from sugarcane bagasse.A signifi cant level of xylanase was obtained by submerged fermentation.This crude enzyme is thermostable at 40-50 °C and can be used in a wide range of pH.The A. fumigatus M51 xylanase produced more prebiotic XOS (xylobiose and xylotriose) than a commercial xylanase.The research also showed the potential of using sugarcane bagasse as feedstock for the production of xylanases and prebiotic xylooligosaccharides.

Table 2 .
Factorial design used for evaluating the activity of xylanolytic strains Trichoderma reesei CCT 2768, Aspergillus fumigatus M51 and A. fumigatus U2370

Table 3 .
Production of xylooligosacharides (XOS) by enzymatic hydrolysis of selected fungal enzymes using hemicellulose extracted from sugarcane bagasse Reaction time was measured at 50 °C, pH=5.0 and 500 U of xylanases per g of substrate.Total XOS=sum of xylobiose and xylotriose