Optimization of ligninolytic enzymes production through response surface methodology

Abstract

There is an increasing demand for green chemistry technologies that can cope with environmental waste management challenges. Agro-industrial residues are primarily composed of complex polysaccharides that support microbial growth for the production of industrially important enzymes such as ligninolytic enzymes. Schyzophyllum commune and Ganoderma lucidum were used alone, as well as mixed/co-culture, to produce crude ligninolytic enzymes extracts using corn stover and banana stalk as a substrate during solid state fermentation (SSF). In the initial screening, the extracted ligninolytic enzymes from S. commune produced using corn stover as the substrate showed higher activities of lignin peroxidase (1007.39 U/mL), manganese peroxidase (614.23 U/mL), and laccase (97.47 U/mL) as compared to G. lucidum and the mixed culture. To improve the production of ligninolytic enzymes by S. commune with solid state fermentation (SSF), physical factors such as pH, temperature, moisture, inoculum size, and incubation time were optimized by varying them simultaneously using response surface methodology (RSM) with a central composite design (CCD). The optimum SSF conditions were (for a 5 g corn stover substrate size): pH = 4.5; temperature = 35°C; inoculum size = 4 mL; and moisture content = 60%. Under optimum conditions, the activities of lignin peroxidase (LiP), manganese peroxidase (MnP), and laccase were 1270.40, 715.08, and 130.80 IU/mL, respectively.

Full Article

Optimization of Ligninolytic Enzymes Production through Response Surface Methodology

QamarYasmeen,^a Muhammad Asgher,^a,* Munir A. Sheikh,^a and Haq Nawaz ^b

There is an increasing demand for green chemistry technologies that can cope with environmental waste management challenges. Agro-industrial residues are primarily composed of complex polysaccharides that support microbial growth for the production of industrially important enzymes such as ligninolytic enzymes. Schyzophyllum commune and Ganoderma lucidum were used alone, as well as mixed/co-culture, to produce crude ligninolytic enzymes extracts using corn stover and banana stalk as a substrate during solid state fermentation (SSF). In the initial screening, the extracted ligninolytic enzymes from S. commune produced using corn stover as the substrate showed higher activities of lignin peroxidase (1007.39 U/mL), manganese peroxidase (614.23 U/mL), and laccase (97.47 U/mL) as compared to G. lucidum and the mixed culture. To improve the production of ligninolytic enzymes by S. commune with solid state fermentation (SSF), physical factors such as pH, temperature, moisture, inoculum size, and incubation time were optimized by varying them simultaneously using response surface methodology (RSM) with a central composite design (CCD). The optimum SSF conditions were (for a 5 g corn stover substrate size): pH = 4.5; temperature = 35°C; inoculum size = 4 mL; and moisture content = 60%. Under optimum conditions, the activities of lignin peroxidase (LiP), manganese peroxidase (MnP), and laccase were 1270.40, 715.08, and 130.80 IU/mL, respectively.

Keywords: Schyzophyllum commune; Ganoderma lucidum; Corn stover; Ligninolytic enzymes; Optimization; Response Surface Methodology (RSM)

Contact information: a: Industrial Biotechnology Laboratory, Department of Chemistry and Biochemistry, University of Agriculture, Faisalabad, Pakistan; b: Institute of Animal Nutrition and Feed Technology, University of Agriculture, Faisalabad, Pakistan;

* Corresponding author: Phone: +92-41-9200161/3312 E-mail:mabajwapk@yahoo.com

INTRODUCTION

Industry and biotechnology demands for ligninolytic enzymes complexes from white-rot fungi (WRF) are increasing due to their versatile catalytic effects on various industrial processes. Ligninolytic enzymes have potential applications in a large number of fields, including bioremediation, biofuels, food, agriculture, paper & pulp, textile finishing, denim stone washing, cosmetics, biosensors, and many others (Wesenberg et al. 2003; Pazarlioglu et al. 2005; Asgher et al. 2008; Levin et al. 2008; Sadhasivam et al.2008; Stoilova et al. 2010; Asgher et al. 2011, 2012 a, b, c). The robust non-specific and extracellular lignin-degrading enzymes, such as lignin peroxidase (LiP), laccase (Lac), and manganese peroxidase (MnP), are responsible for lignin degradation and bioremediation capabilities of WRF (Yang et al. 2011; Martorell et al. 2012). However, problems may arise associated with the direct application of the fungi, such as difficulties in satisfying the growth requirements on a large scale, long incubation times, and adsorption of the pollutants on fungal mycelia (Casas et al. 2009; Kamei et al. 2009). In contrast, in vitro treatment with ligninolytic enzymes produced through WRF fermenta-tion can minimize these problems (Wang et al. 2008; Torres-Duarte et al. 2009). However, the use of these purified and/or immobilized enzymes increases the cost of industrial processes.

In the recent past, some efforts have been made to investigate the potential of cell-free culture extracts of WRF for in vitro large-scale industrial and bioremediation treatment of recalcitrant and xenobiotic compounds (Rubilar et al. 2008). The process control of these enzymatic applications is simplified by the absence of the living bacterium, and has advantages, such as the ability to operate over a wider range of pollutant concentrations, pH, and temperatures. In addition, the biodegradative capacity of the enzymes may be unaffected by toxic compounds that could otherwise inhibit fungal cultures (Tsutsumi et al. 2001). Moreover, using an unpurified crude enzyme extract may also provide the enzyme with substances that mediate the catalytic cycle of WRF peroxidases and laccases, which are eliminated during enzyme isolation and purification.

However, the fungus must secrete high activities of ligninolytic and associated enzymes in the enzyme extract. To achieve this, enzyme production must be optimized (Champagne and Ramsay 2005). The ligninolytic machinery in most basidiomycetes is highly regulated by nutrients, such as nitrogen, copper, and manganese. Their production is also affected by fermentation factors, such as medium composition, nature of carbon source, concentration of carbon source, pH of fermentation broth, fermentation temperature, amount and nature of nitrogen source, as well as the presence of inducers, mediators and organic acids, such as citric, oxalic, and tartaric acids (Ryan et al. 2007; Wen et al. 2009; Iqbal et al. 2011).

Reducing the costs of enzyme production by using cheaper raw materials and optimizing the fermentation process for industrial applications is the ultimate target of basic research (Lee et al. 2011; Soni et al. 2012). The conventional classical optimization strategy (COS) of varying one-factor-at-a-time is time consuming and cannot guarantee the optimum physical and nutritional conditions, since the strategy does not consider the interactions among different variables/parameters (Lotfy et al. 2007; Hye et al. 2008; Kammoun et al. 2008; Gadhe et al. 2011; Tijani et al. 2011). In contrast, response surface methodology (RSM) can examine varying more than one-factor-at-a-time, at several different levels, to determine the interactions between two or more factors. This testing methodology can provide reliable optimization results (Benzina et al. 2012).

Among the various processes used for enzyme production, solid-state fermenta-tion (SSF),which uses lignocellulosic biomass, appears promising because it has many advantages for fungal cultivations (Pointing 2001). SSF is an attractive option for bioconversion of lignocellulosic biomass and production of lignocellulolytic enzymes. The enzymes can be isolated and purified to different extents for diverse industrial applications, and the residual biomass can be utilized as animal feed (Elisashvili et al. 2009; Sharma and Arora 2010). Since SSF operates under low moisture conditions, bacterial contamination chances are also minimized (Basu et al. 2002; Li et al. 2006). SSF reproduces the natural conditions for the growth of WRF and has been shown to be more suitable for the production of industrial enzymes that do not need purification.

The present study was aimed at producing enzyme extracts by growing single and co-cultures of G. lucidum and S. commune utilizing SSF of lignocellulosic substrates. A further goal was to optimize the production of ligninolytic enzyme extracts by selected fungal culture using the selected lignocellulosic substrate in RSM.

EXPERIMENTAL

All experimental and analytical work was carried out at the Industrial Biotechnology Laboratory (IBL) of the Department of Chemistry and Biochemistry at the University of Agriculture, Faisalabad (UAF) in Pakistan. Two indigenous white rot fungi (WRF), G. lucidum and S. commune, were isolated and used as individual cultures, as well as in co-cultures, to produce extracts of ligninolytic enzymes. The fungal strains/cultures were screened on different substrates in a SSF to select the hyper-producing fungal culture on the basis of higher ligninolytic enzymes activities in the crude enzyme extract. The physical and nutritional parameters were studied with response surface methodology (RSM) to optimize the production of ligninolytic enzymes by the selected fungal culture in a SSF on selected substrates.

Production of Crude Ligninolytic Enzymes Extracts

Lignocellulosic substrates

Corn stover and banana stalk were collected from local vegetable and fruit markets of Faisalabad, Pakistan. Substrates were sliced into pieces, oven dried at 50ºC, ground to 40 mm mesh particle size, and stored in airtight plastic jars to avoid moisture re-adsorption.

Ligninolytic enzymes producing white rot fungi

Pure cultures of indigenous strains of G. lucidum and S. commune were obtained from the Industrial Biotechnology Laboratory of the Department of Chemistry & Biochemistry at the University of Agriculture in Faisalabad, Pakistan. The fungi were raised on potato dextrose agar (PDA) slants at pH 4.5. The inoculated slants were incubated for five days at 30ºC for spore multiplication and stored in a refrigerator at 4ºC for subsequent use in production of crude enzyme extracts.

Inoculum preparation

Kirk’s basal nutrient media (100 mL) was the inoculum medium used (Tien and Kirk 1988). It was prepared in three separate labeled Erlenmeyer flasks (500 mL) and adjusted at pH 4.5 with 1 M NaOH or 1 M HCl. After sterilization for 15 min. at 121ºC, the medium was supplemented with Millipore filtered (0.3 µm) 1% glucose. Spores of G. lucidum and S. commune were added to the respective sterilized inoculum media from slant cultures, and the flasks were incubated in a temperature-controlled still culture incubator (EYLA SLI-600ND, Japan) at 30ºC for 5 to 7 days to obtain a homogenous spore suspensions of fungi (1×10⁶-10⁸spores/mL) to use as inoculum. Fresh inoculum was prepared for each experiment (Asgher et al. 2006).

Screening of fungi for production of enzyme extracts on different substrates in SSF

Two separate sets of flasks for respective lignocellulosic substrates were prepared in triplicate using 5 g of the respective substrate moistened with Kirk’s basal medium (Tien and Kirk 1988) at pH 4.5 and 60% (w/w) moisture. All the reactor flasks were autoclaved in a laboratory scale autoclave (Sanyo, Japan) and inoculated with 5 mL homogenous suspension of each strain. The co-culture flasks received 2.5 mL inoculum each of the two fungi. The inoculated flasks were incubated at 30ºC for 10 days. The culture flasks were harvested at different time intervals (48, 96, 144, 192, 240 h). To the fermented SSF culture flasks, 100 mL of 100 mM sodium succinate buffer (pH 4.5) was added; afterwards, the flasks were shaken (120 rpm) in an orbital shaker (Sanyo-Gallemkemp, UK) for half an hour. The biomass was filtered through a Whatman No.1 filter paper, and the resulting filtrates were centrifuged at 3000 g for 10 min. at 4ºC to remove fungal mycelia and cell debris. The clear supernatants were used to determine the activities of LiP, MnP, and laccase.

Optimization of Physical Factors though Response Surface Methodology

The physical parameters including inoculum size, temperature, pH, moisture content, and incubation time were optimized through response surface methodology (RSM), using a five factor-five level central composite design (CCD) with six center points (=0.5) and six replicates for each center point. A total of 32 runs were employed (Table 1). The center point replicates were chosen to verify any change in the estimation procedure, as a measure of precision property.

Once the experiments were performed, a second order polynomial equation (1) shown below was used to describe the effect of variables in terms of linear, quadratic, and cross product terms

where i and j are linear and quadratic coefficients, respectively, while ‘b’ symbols represent regression coefficients, Y is the ligninase yield, k the number of factors studied and optimized in the experiment, and ‘e’ is random error. When developing the regression equation, the test factors were coded according to the following equation (2),

where x_i is the dimensionless value of an independent variable, X_iis the real value of an independent variable, X_o is the real value of the independent variable at the center point, and ∆X_i is the step change value.

Table 1. Five Factor-Five Levels Central Composite Design (CCD) Using RSM for the Optimization of Crude Ligninolytic Enzymes by S. commune in a SSF from Corn Stover

Ligninolytic Enzyme Assays

Lignin peroxidase activity

LiP was assayed by the method of Tien and Kirk (1983). The LiP assay was performed by using 2.6 mL of reaction mixture containing 1 mL buffer of pH 3, 1 mL of 4 mM veratryl alcohol (3,4-dimethoxybenzylalcohol), 500 µL of 1 mM H₂O₂, and 100 µL of the enzyme aliquot. A blank contained 100 µL of distilled water instead of enzyme aliquot. The UV/Vis absorbance was read after a 10-minute reaction interval at 310 nm (₃₁₀ =9300 M^-1cm^-1). Enzyme activity was defined as µM of veratraldehyde formed per min.

Manganese peroxidase Assay

The activity of MnP was measured by the method of Wariishi et al. (1992). The assay mixture (2.6 mL) contained 1 mL of 1 mM MnSO₄, 1 mL of 50 mM sodium malonate buffer of pH 4.5, and 100 µL of the culture supernatant. Five hundred microliters of 0.1 mM H₂O₂ was added as an oxidizing agent. Manganic ions Mn⁺³form a complex with malonate that absorbs at 270 nm (₂₇₀ = 11,590 M^-1cm^-1). MnP activity was defined as µM of MnSO₄oxidized per min.

Laccase assay

Laccase activity in the crude enzyme extracts was measured by the method of Shin and Lee (2000) by monitoring the oxidation 2,2-azinobis(3-ethylbenzthiazoline-6 sulphonic acid) (ABTS) in a reaction mixture containing 1 mL of 1 mM ABTS in 1 mL of 50 Mm malonate buffer (pH 4.5) and 100 µL of the culture supernatant. The reaction mixture was incubated at 25ºC and absorbance was taken at 436 nm after a10 min interval (₄₃₆ = 36,000 M^-1cm^-1).

A blank contained 100 µL of distilled water instead of enzyme solution or culture supernatant. Laccase activity was calculated as change in absorbance of the assay mixture after a ten-minute reaction interval.

RESULTS

Screening of WRF Cultures for Production of Ligninolytic Enzymes Extracts

Two white rot fungi, S. commune and G. lucidum, and their co-culture were used to produce ligninolytic enzymes extracts in the SSF of corn stover or banana stalks over a 10-day fermentation period. In case of S. commune, the maximum activities of LiP (1007.39 U/mL), MnP (614.23 U/mL), and laccase (97.47 U/mL) were noted on the ninth day of incubation in corn stover medium. G. lucidum also showed optimum activities of ligninolytic enzyme after nine days of SSF with banana stalk (Table 2).

However, the fungal co-culture of both strains showed lower activities of ligninolytic enzymes as compared to the individual cultures. The lower ligninolytic activities produced by the co-culture were unusual because both fungi have almost similar growth conditions (Irshad and Asgher 2011). The individual fungi may have secreted some metabolites that are antagonistic to each other leading to their growth inhibition.

Table 2. Production of Ligninolytic Enzymes by Single and Co-culture of S. commune and G. lucidum on Different Lignocellulosic Substrates in a SSF *

*pH 4.5; Temperature, 30°C

The genetic variation among the fungi, and the nature and composition of the substrates used may be responsible for the better growth of fungi on different substrates (Giardina et al. 2000; Patel et al. 2009). Overall, the crude enzyme extract produced by S. commune in the SSF of corn stover had maximum activities of ligninolytic enzymes. On the basis of these results, S. communewas selected for the optimization of ligninolytic enzymes produced by the SSF of banana stalks and corn stover.

Optimization of Parameters for Enhanced Production of Ligninolytic Enzymes by S. commume

Optmization of physical factors through response surface methodology (RSM)

RSM was employed with a central composite design (CCD). The independent variables (i.e. the main effects) were: inoculum size (A), temperature (B), pH (C), moisture (D), and incubation time (E). The optimum ligninase activities were obtained with: inoculum size, 4 mL; temperature, 35°C; pH 4.5; moisture, 60%; and incubation time, 144 h. The activities of lignin peroxidase (1270.40 U/mL), mangenese peroxidase (715U/mL), and laccase (130.8 U/mL) were substantially enhanced by optimizing the process conditions (Table 3). LiP, MnP, and laccase are secreted by different fungi in different activity profiles even under the same growth conditions. Some fungi secrete more LiP and MnP, and low laccase activity, and some even lack laccase altogether. Others are better laccase producers with low or even negligible LiP or MnP activities. The production of all three enzymes by a particular white rot fungus is optimum under similar physical growth conditions such as pH, moisture, and temperature, but their activities are different. However, the optimum conditions for synthesis of individual enzymes might be variable under different nutritional conditions.

The effects of interactions among the variables on ligninases activities (dependant variables) were also studied. The model’s large F-values of 633.56, 178.83, and 11.74 for LiP, MnP, and laccase, respectively, indicated the model’s significance. There was statistically 0.01% chance that this large F-value of model could occur due to noise (Table 4 A, B, & C). Linear (A, B, C, D, E), interaction (AB, AC, AD, BC, BD, BE, CD, CE, DE), and quadratic (A²,C²,D²) terms were statistically significant with p-values< 0.0001. Non-significant F-values for lack-of-fit for LiP (1.50), MnP (0.46), and laccase (0.0090), relative to the pure error, confirmed the model’s predictability. A positive value of the t-statistic in the case of inoculum size, temperature, moisture, and incubation time indicated a positive linear effect, whereas it was found to be negative for pH. The coefficient for the interaction between temperature and incubation time (BE) was significant with p-value of 0.003, while all other interactions were non-significant with p-values higher than 0.05. The negative coefficient value for pH in the linear term showed that the production of laccase decreased with an increase in pH. The positive quadratic term indicated the existence of a minimum for these activities.

Justification of the variability in observed response values by the experimental factors (variables) and their interactions was measured by R² (coefficient of determina-tion). The predicted R²values of 0.8158, 0.8062, and 0.4921 by the model for LiP, MnP, and laccase were in close agreement with the actual R² values of 0.9976, 0.9909, and 0.4921, respectively. The adjusted R² was very close to the actual R² value. LiP, MnP, and laccase R² values of 0.9991, 0.9968, and 0.9950, respectively, implied that the fitted linear, interaction, and quadratic terms could elucidate 99, 99.68, and 99.50% of variation, showing satisfactory representation of the process by the model.

The preciseness and reliability of conducted experiments were confirmed by lower values of coefficient of variation (CV) of 0.54, 0.96, and 6.86% for LiP, MnP, and laccase, respectively. The measured signal-to-noise ratio indicated adequate precision. A desirable ratio should be larger than 4. All the model ratios (106.81, 58.626, and 15.124 for LiP, MnP, and laccase, respectively) were greater than 4, indicating adequate signals that can be used for design space navigation. The standard deviation values of 5.36, 4.08, and 5.32 for LiP, MnP, and laccase indicated that the model showed strong compliance with predicted response.

Table 3. Optimization of Fermentation Parameters using the RSM with a Central Composite Design (CCD)

Interaction among variables

Comparative effects of any two variables were explained by contour plots while holding the other factors fixed at their central point values. From response surface (3D) and contour plots (2D), the interactive effects of experimental factors on ligninolytic enzymes synthesis were determined.

Inoculum size vs.temperature

The response surface curves (Fig. 1) showed how LiP, MnP, and laccase produc-tion was a function of inoculum size and temperature by keeping the levels of pH, moisture, and incubation time at 4.50, 60%, and 144 h, respectively.

Table 4A. Analysis of Variance (ANOVA) for the Quadratic Polynomial Model for Lignin Peroxidase Production in SSF

Table 4B. ANOVA of CCD for the Quadratic Polynomial Model for Manganese Peroxidase Production in SSF

Table 4C. ANOVA of CCD for the Quadratic Polynomial Model for Laccase Production in SSF

The production of lignin-modifying enzymes was affected by both factors and was maximized by the combination of the level of inoculum size and temperature. It was observed that inoculum size and temperature had strong interactive effect on production of lignin-modifying enzymes (LMEs) by S. commune. LiP, MnP, and laccase showed analogous results with maximum activities with 4 mL of inoculum at 35°C. The interaction of temperature and inoculum size showed that there is an increase in MnP and laccase production with an initial increase in temperature and inoculum size; however, high levels of these factors had inhibitory effects.

Inoculum size vs. pH

Countour and response surface plots (Fig. 2) showed that there was high production of LiP, MnP, and laccase at 4 mL of inoculum size and pH 4.5, but low yields at higher levels of these parameters. The contour plot for all lignin-modifying enzymes predicted that a pH range of 4 to 4.5 and inculum level of 2 to 4 mL was optimum. Response surface plot predicted that a pH range of 5 to 5.5 decreased the synthesis of the enzymes.

Inoculum size vs.moisture

At a moderate level of moisture (60%) and inoculum size (4 mL), S. commune showed optimum production of LMEs (Fig. 3). A further increase in both variables decreased the growth of fungus, and hence, the production of enzymes in the SSF. At their initial levels, both inoculum size (2 to 4 mL) and moisture (56 to 64%) showed a profound effect on crude enzymes production. Conversely, at higher levels of these independent variables, the net LMEs yield decreased. Interaction of inoculum size and moisture had a significant effect on LiP, MnP, and laccase production in SSF by S. commune using corn stover.

Fig. 1. Response surface plots showing the interactive effect of inoculum size and temperature on (a) LiP, (b) MnP, and (c) laccase production (hold value: pH, 4.5; moisture, 60%; and incubation time, 144 h)

Fig. 2. Response surface plots showing the interactive effect of inoculum size and pH on (a) LiP, (b) MnP, and (c) laccase production (hold values: temperature, 35°C; moisture, 60%; incubation time, 144 h)

Fig. 3. Response surface plots showing the interactive effect of inoculum size and moisture on (a) LiP, (b) MnP, and (c) laccase production (hold value: temperature, 35°C; pH, 4.5; incubation time, 144 h)