Process Biochemistry 38 (2002) 727 /731 www.elsevier.com/locate/procbio
Optimization of xylanase production by Bacillus circulans D1 in submerged fermentation using response surface methodology D.A. Bocchini a,b, H.F. Alves-Prado a, L.C. Baida a, I.C. Roberto c, E. Gomes a, R. Da Silva a,b,* a
Biochemistry and Applied Microbiology Laboratory, IBILCE / UNESP, Sao Jose do Rio Preto, SP, Brazil b Chemistry Institute, UNESP, Araraquara, SP, Brazil c Faculty of Chemical Engineering of Lorena, Faenquil, Lorena, SP, Brazil Received 26 October 2001; accepted 21 June 2002
Abstract In this work, a 33 factorial design was performed with the aim of optimizing the culture conditions for xylanase production by an alkalophilic thermophilic strain of Bacillus circulans , using response surface methodology. The variables involved in this study were xylan concentration (X1), pH (X2) and cultivation time (X3). The optimal response region was approached without using paths of steepest ascent. Statistical analysis of results showed that, in the range studied, only pH did not have a significant effect on xylanase production. A second-order model was proposed to represent the enzymic activity as a function of xylan concentration (X1) and cultivation time (X3). The optimum xylan concentration and cultivation time were 5 g/l and 48 h, respectively. Under these conditions, the model predicted a xylanase activity of 19.1 U/ml. # 2002 Elsevier Science Ltd. All rights reserved. Keywords: Bacillus circulans ; Xylanase; Experimental design; Optimization
1. Introduction Xylan is the main hemicellulosic polysaccharide found in plant cell walls and is composed of a backbone chain of b-1,4-linked xylosyl residues and short side chains of arabinosyl, glucoronosyl and acetyl residues . Xylan represents a significant resource of renewable biomass and comprises up to 20/35% dry weight of wood and agricultural wastes. For most bioconversion process, xylan must first be converted to xylose or xylooligosaccharides . Xylanase (endo-1,4-b-xylanase) and b-xylosidase (b-D-xyloside xylohydrolase) the main constituents of the xylanolytic enzyme system, convert this polysaccharide into a more readily fermentable form . Interest in xylanases has increased over the last years mainly due to the application of these enzymes * Corresponding author. Address: IBILCE / UNESP, Chemistry and Geoscience Department, Cristo˜va˜o Colombo Street Number 2265, Jardim Nazare´, 15054-000 Sao Jose do Rio Preto, SP, Brazil. Tel.: / 55-17-221-2354; fax: /55-17-221-2356 E-mail address: [email protected]
(R. Da Silva).
in paper industries for pulp treatment, improving the effectiveness of conventional bleaching chemicals [4,5]. Xylanases can also be used in fruit juice and wine clarification [6,7]. Most of the reports concerning xylanases deal with the purification and characterization of these enzymes. On the other hand, studies regarding to optimization of culture medium and culture conditions for the production of xylanases are still few in the scientific literature. For development of good fermentation some parameters should be optimized according to the limits of the process, such as pH, temperature, substrate concentration, cultivation time and aeration. Much information can be obtained using experimental designs based on statistical principles, performing a minimum number of experiments . Different statistical designs for medium optimization regarding xylanase production have been reported, among which factorial experiments and response surface methodology (RSM) are included [9 /11]. These are useful techniques for approaching the region of optimum response . RSM is a model, consisting of mathematical and statistical techniques, widely used to study the effect of
0032-9592/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 0 3 2 - 9 5 9 2 ( 0 2 ) 0 0 2 0 7 - 8
D.A. Bocchini et al. / Process Biochemistry 38 (2002) 727 /731
several variables and to optimize different biotechnological process [13,14]. Successful application of factorial design experiments and RSM to enhance xylanase production by several microorganisms, mainly fungi such as Sclerotium rolfsii , Thermomyces lanuginosus , Melanocarpus albomyces  and Trichoderma longibrachiatum [17,18] has been reported. Studies regarding optimization of xylanase production by Bacillus species are less available in the literature  and some of these reports do not cite the use of factorial designs .The present work describes the optimization of extracellular xylanase production by the alkalophilic and thermophilic bacteria Bacillus circulans D1 in submerged fermentation, with the aim of reducing the costs of this enzyme production.
is defined as the amount of enzyme releasing 1 mmol of reducing sugar equivalent per minute under the assay conditions. 2.5. Experimental design A 33 factorial design was performed in order to determine the optimal conditions for the production of xylanase by B. circulans D1. The dependent variable selected for this study was the enzymic activity, expressed in U/ml, and the independent variables chosen were xylan concentration, medium pH and cultivation time. The range and the levels of these variables are given in Table 1. 2.6. Statistical analysis
2. Materials and methods
STATISTICA, (version 5.0) software from StatSoft Inc., was used for regression and graphical analysis.
2.1. Microorganism Bacillus circulans D1, isolated from decayed wood, was used in this study. Stock cultures were maintained on culture medium, which was used for the submerged culture, added of 1.5% agar, at 4 8C. 2.2. Cultivation medium The initial submerged cultivation medium was composed of xylan 10.0 g/l, beef extract 10.0 g/l, peptone 10.0 g/l, NaCl 10.0 g/l, KH2PO4 1.0 g/l and Na2CO3 5.0 g/l (separately sterilized) . For the experiments of optimization, the cultivation medium was composed of different concentrations of xylan, also varying the medium pH and the time of cultivation, according to the experimental design. 2.3. Shake flask cultivation The optimization experiments were carried out in 125ml Erlenmeyer flasks containing 20 ml of cultivation medium. Shake flasks were seeded with inocula, at initial concentration of 4.0 /106 cells/ml, which were incubated for 12 h at 45 8C under 150 rev/min on the initial cultivation medium cited above. Shake flasks cultures were maintained at 45 8C under 150 rev/min for 24, 48 or 72 h, according to the experimental design. 2.4. Enzyme assays Xylanase activity was determined by the dinitrosalicylic acid method , measuring reducing sugars released as xylose after incubation of 0.1 ml of the enzyme solution with 0.9 ml of a birchwood xylan (Sigma) solution (5.0 g/l) in acetate buffer (pH 5.0, 0.1 M) at 60 8C for 10 min. One unit (U) of enzyme activity
3. Results Based on earlier studies, xylan concentration, medium pH and cultivation time were identified as the major factors affecting xylanase production by B. circulans D1. In the present work, these variables were statistically optimized with the help of a 33 factorial design using RSM. The experimental design and the results are shown in Table 2. The highest xylanase activity (22.45 U/ml) was observed at run number 12, where the factors xylan concentration, pH and cultivation time were used at their levels intermediary (7.5 g/l), smaller (8.0) and higher (72 h), respectively. This activity was 5.5-fold higher than that observed at run number 25, where the related factors were used at highest levels for X1 and X2 and lowest for X3. Table 3 shows the regression analysis for this experiment, presenting the estimates and hypothesis tests to the coefficients of regression. At the 5% probability level, the linear and quadratic coefficients of X3 (cultivation time factor), and the coefficient of the interaction X1X3 (xylan concentration and cultivation time) were found to be significant. The mathematical model representing the xylanase activity (/y) ˆ in the experimental region studied can be expressed by Eq. (1). Table 1 Values of independent variables a different levels of the 33 factorial design Independent variables
Xylan (g/l) pH Cultivation time (h)
X1 X2 X3
5 8.0 24
7.5 8.5 48
10 9.0 72
D.A. Bocchini et al. / Process Biochemistry 38 (2002) 727 /731
The response surface described by the model equation ˆ to estimate xylanase activity over independent (/y) variables xylan concentration (X1) and cultivation time (X3) is shown in Fig. 1. There is a rather broad plateau region over which the enzyme activity changes relatively little when xylan concentration is varied. Xylanase activities practically identical were observed in experimental runs when the factor cultivation time was used as 48 or 72 h, independently of the xylan concentration used. The production of xylanase reaches a peak in the region of levels ‘0’ and ‘/1’ for the factor cultivation time and the level ‘0’ for the factor xylan concentration, respectively (Fig. 1). As observed, the pH in the range studied (8.0 /9.0), did not influence xylanase production and it was maintained around 9.0. Based on the model obtained, the optimal working conditions were defined to attain high xylanase activity minimizing the xylan concentration and the cultivation time. Thus, the point assigned as optimum corresponded to 5.0 g/l of xylan concentration and 48 h of cultivation time. Under this condition, the model predicted a xylanase activity of 19.1 U/ml being possible a variation of 17.5 /21.1 U/ml.
Table 2 Experimental design and results of the 33 factorial design Run number
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
Xylanase activity (U/ml)
1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1
1 1 1 0 0 0 1 1 1 1 1 1 0 0 0 1 1 1 1 1 1 0 0 0 1 1 1
1 0 1 1 0 1 1 0 1 1 0 1 1 0 1 1 0 1 1 0 1 1 0 1 1 0 1
11.11 16.20 15.81 8.17 17.04 16.12 6.75 21.54 18.65 6.75 18.63 22.45 8.10 18.73 17.04 4.70 18.36 18.97 5.59 18.75 19.12 4.44 17.64 16.60 4.05 17.14 17.93
8.62 18.45 16.54 7.62 17.96 16.56 8.26 19.11 18.22 8.31 19.39 18.73 6.63 18.22 18.07 6.58 18.69 19.05 6.39 18.69 19.28 3.99 16.84 17.94 3.27 16.63 18.24
Table 3 Results of regression analysis of the 33 factorial design Term
Intercept X1 X1X1 X2 X2X2 X3 X3X3 X1X2 X1X3 X2X3
18.225 0.562 0.823 0.351 0.818 5.724 5.873 0.680 1.251 0.509
20.943 1.396 1.179 0.870 1.172 14.207 8.416 1.379 2.535 1.032
0.000 0.180 0.254 0.396 0.257 0.000* 0.000* 0.186 0.027** 0.316
* Statistically significant at 99% of confidence level. ** Statistically significant at 95% of confidence level.
y ˆ 18:220:56x1 5:72x3 1:25x1 x3 5:87x23
The statistical significance of a second-order model equation was evaluated by the F-test analysis of variance (ANOVA), which showed that this regression is statistically significant at 99% confidence level. The model did not show lack-of-fit and presented a high determination coefficient (R2 /0.92), explaining 92% of the variability in the response (Table 4).
Over the last few decades, even though several papers regarding optimization of xylanase production have been reported, little information about the optimization of this enzyme production by submerged fermentation, using Bacillus species, is available in the scientific literature. Some of these studies with bacteria have not reported the use of statistical designs [23,24]. B. circulans D1 is a good producer of thermostable cellulase-free xylanase. From the 33 factorial experiment a response surface was given and indicated that, after a period of incubation, xylanase production does not increase, even though a high xylan concentration is used. The highest enzyme production (22.45 U/ml) achieved with the 33 factorial design was approximately 315% higher than production on the initial medium. Different statistical experimental designs were combined by Pham et al.  with the aim of optimizing a culture medium for the production of xylanase by Bacillus sp I-1018. The xylan, casein hydrolysate and NaCl concentration were optimized by means of surface response. The optimum composition of the nutrient medium was determined and found to be 3.16 g/l xylan, 1.94 g/l casein hydrolysate and 0.8 g/l NaCl. The highest xylanase production observed was, approximately, 151 U/ml. The cultivation of Bacillus sp I-1018 in the optimized medium resulted in an increase of 135% on xylanase production, compared to the initial medium. Surface response methods were also used to optimize a culture medium with regard to xylanase production by
D.A. Bocchini et al. / Process Biochemistry 38 (2002) 727 /731
Table 4 Analysis of variance (ANOVA) for the model regression representing xylanase activity Source
Model Residual Total (corr.)
821.17 68.7005 889.871
4 22 26
R2 0.92; SS, sum of squares; DF, degrees of freedom; MS, mean square.
and a decrease of 50% of the xylan amount initially used. These facts are important in making the whole process economically more feasible, in view of the high cost of pure xylan and the difficulty of its extraction from vegetable material, as has been proposed from some agricultural residues. The alkalophilic pH of the cultivation medium and the high temperature used also reduce the chances of contamination by opportunist microorganisms.
Acknowledgements Fig. 1. Response surface described by the model y; ˆ which represents xylanase activity (U/ml) as a function of xylan concentration and cultivation time.
a Schizophyllum commune strain . The optimum concentrations of avicel, yeast extract and NH4NO3 (73.4, 55.5 and 1.38 g/l, respectively) were determined by a central composite design. The highest xylanase production (5.74 U/ml), resulted in an increase of 330% compared to the initial medium and was reached within 11 days, a very long period of cultivation. The cultural conditions for thermostable xylanase production of Thermomyces lanuginosus were studied by Purkarthofer et al. . An experiment using central composition was performed to optimize the medium components concentrations. When the fungus was grown in the optimized medium (31.2 g/l corn cob, 30.2 g/l yeast extract and 5.0 g/l KH2PO4), the highest xylanase production (2.70 U/ml) was observed within 7 days. In the present work, the increase in xylanase production is comparable with those observed by Haltrich et al.  and higher than that observed by Pham et al. . In addition, the highest xylanase activity was reached in a shorter time period in comparison to those observed by Haltrich et al.  and Purkarthofer et al. . With the present study it was possible to determine the parameters of 5.0 g/l and 48 h of cultivation time as the better conditions for enzyme production. pH did not influence this production and it was fixed around 9.0. It can be concluded that the study of optimization of culture conditions for xylanase production by B. circulans D1 a to an increase of 315% in enzyme production
We thank FAPESP for financial support.
References  Kubata BK, Takamizawa K, Kawai K, Suzuki T, Horitsu H. Xylanase IV, an endoxylanase of Aeromonas caviae ME-1 which produces xylotetraose as the only low molecular weight oligosaccharide from xylan. App Environ Microbiol 1995;61:1666 /8.  Chen C, Chen JL, Lin TY. Purification and characterization of a xylanase from Trichoderma longibrachiatum for xylooligosaccharide production. Enzyme Microbiol Technol 1997;21:91 /6.  Ghosh M, Amitabha D, Mishra AK, Nanda G. Aspergillus sydowii MG 49 is a strong producer of thermostable xylanolytic enzymes. Enzyme Microbiol Technol 1993;15:703 /9.  Buchert J, Tenkanen M, Kantelinen A, Viikari L. Application of xylanases in the pulp and paper industry. Bioresource Technol 1994;50:65 /72.  Da Silva R, Yim DK, Park YK. Application of thermostable xylanases from Humicola sp for pulp improvement. J Ferment Bioeng 1994;77(1):109 /11.  Biely P. Microbial xylanolytic systems. Trends Biotechnol 1985;3(11):286 /90.  Da Silva R, Franco CML, Gomes E. Substrato, Ac¸a˜o, Produc¸a˜o e Aplicac¸a˜o de pectinases, hemicelulases e celulases no processamento de alimentos: revisa˜o. Bol SBCTA 1997;31(2):249 /60.  Neto BB, Scarminio SI., Bruns ER. Planejamento e Otimizac¸a˜o de Experimentos. 2nd ed., Editora da Universidade Estadual de Campinas-SP, 1996.  Royer JC, Nakas JP. Xylanase production by Trichoderma longibrachiatum . Enzyme Microbiol Technol 1998;11:405 /10.  Purkarthofer H, Sinner M, Steiner W. Cellulase-free xylanase from Thermomyces lanuginosus : optimization of production in submerged and solid-state culture. Enzyme Microbiol Technol 1993;15:677 /82.  Adamsen AK, Lindhagen J, Ahring BK. Optimization of extracellular xylanase production by Dictyoglomus sp. B1 in continuous culture. Appl Microbiol Biotechnol 1995;44:327 /32.
D.A. Bocchini et al. / Process Biochemistry 38 (2002) 727 /731  Haltrich D, Preiss M, Steiner W. Optimization of a culture medium for increased xylanase production by a wild strain of Schizophyllum commune . Enzyme Microbiol Technol 1993;15:854 /60.  Hamsaveni DR, Prapulla SG, Divakar S. Response surface methodological approach for the synthesis of isobutyl isobutyrate. Process Biochem 2001;36:1103 /9.  Silva CJSM, Roberto IC. Optimization of xylitol production by Candida guilliermondii FTI 20037 using response surface methodology. Process Biochem 2001;36:1119 /24.  Haltrich D, Laussamayer B, Steiner W, Nidetzky B, Kulbe KD. Cellulolytic and hemicellulolytic enzymes of Sclerotium rolfsii */ optimization of the culture medium and enzymatic hydrolysis of lignocellulosic material. Bioresource Technol 1994;50(1):43 /50.  Narang S, Sahai V, Bisaria VS. Optimization of xylanase production by Melanocarpus albomyces IIS68 in solid state fermentation using response surface methodology. J Biosc Bioeng 2001;91(4):425 /7.  Ridder ER, Nokes SE, Knutson BL. Optimization of solid state fermentation parameters for the production of xylanase by Trichoderma longibrachiatum on wheat bran. Trans ASAE 1998;41(5):1453 /9.
 Ridder ER, Nokes SE, Knutson BL. Optimization of solid-state fermentation for the production of xylanase by Trichoderma longibrachiatum on wheat bran in a forced aeration system. Trans ASAE 1999;42(6):1785 /90.  Pham PL, Taillandier P, Delmas M, Strehaiano P. Optimization of a culture medium for xylanase production by Bacillus sp. Using statistical experimental designs. W J Microbiol Biotechnol 1998;14:185 /90.  Subramaniyan S, Prema P. Optimization of cultural parameters for the synthesis of endo-xylanases from Bacillus SSP-34. J Scientific Ind Res 1998;57:611 /6.  Horikoshi K. Production of alkaline enzymes by alkalophilic microorganisms. Agr Biol Chem 1971;35(11):1783 /91.  Miller GL. Use of dinitrossalicylic acid reagent for determination of reducing sugars. Anal Chem 1959;31:426 /8.  Brodel B, Samain E, Debeire P. Regulation and optimization of xylanase production in Clostridium thermolacticum . Biotechnol Lett 1990;12(1):65 /70.  Subramaniyan S, Prema P. Optimization of cultural parameters for the synthesis of endo-xylanases from Bacillus SSP-34. J Scientific Ind Res 1998;57:611 /6.