Optimization of multi-enzyme production by fungi isolated from palm kernel expeller using response surface methodology

Chen, W. L., Liang, J. B., Jahromi, M. F., Ho, Y. W., and Abdullah, N. (2013). "Optimization of multi-enzyme production by fungi isolated from palm kernel expeller using response surface methodology," BioRes. 8(3), 3844-3857.

Abstract

Response surface methodology (RSM) was used to optimize the co-production of a mixture of crude cellulosic and hemicellulosic enzymes (endoglucanase, xylanase, and mannanase) by Aspergillus terreus K1 in solid-state fermentation (SSF) using palm kernel expeller (PKE) as the sole carbon source. These enzymes have gained renewed interest due to their efficacy to improve the digestibility of PKE for use in diets of mono-gastric animals (poultry, pigs, and fish). The results showed that temperature, moisture, inoculum concentration, and initial pH had significant (P< 0.05) effects on the enzymes production. Using PKE as a solid substrate, maximum endoglucanase, mannanase, and xylanase (17.37, 41.24, and 265.57 U/g DM, respectively) were obtained at 30.5 °C, 62.7% moisture, 6% inoculum, and pH 5.8. The enzyme activities recorded were close to the predicted values (19.97, 44.12, and 262.01 U/g DM, respectively).

Download PDF

Full Article

Optimization of Multi-enzyme Production by Fungi Isolated from Palm Kernel Expeller using Response Surface Methodology

Wei Li Chen,^a Juan Boo Liang,^a,* Mohammed Faseleh Jahromi,^a Yin Wan Ho,^b and Norhani Abdullah ^a,c

Keywords: Solid-state fermentation; Aspergillus terreus; Palm kernel expeller; Response surface methodology

Contact information: a: Institute of Tropical Agriculture, University Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia; b: Institute of Bioscience, University Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia; c: Department of Biochemistry, Faculty of Biotechnology and Biomolecular Science, University Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia;

* Corresponding author: jbliang@upm.edu.my

INTRODUCTION

Escalating demand for traditional feed ingredients such as corn, soybean, and other grains for animal feed over the last two decades has resulted in their scarcity and created competition for food with the human population (Vasta et al. 2008). Thus, the search for alternative feed ingredients, including agro-industrial byproducts such as palm kernel expeller (PKE), has been given considerable attention in recent years. PKE, a by-product of the palm oil industry obtained through the screw-pressing process, is a good source of energy and protein for ruminant animals, but it is sparingly used in poultry feed (Soltan 2009) because it contains a high level of non-starch polysaccharides (NSPs), mainly mannans and some cellulose, xylose, and other polysaccharides (Sundu et al. 2006) that monogastric animals (poultry, pigs, and fish) cannot digest. This suggests that at least three main cellulolytic and hemicellulolytic enzymes are needed to digest the glycosidic bonds of mannan, cellulose, and xylan to improve the nutritive value of PKE if it is to be used more efficiently in diets for poultry and other monogastric animals. Previous studies in Malaysia (Ng et al. 2002; Saenphoom et al. 2011) showed that commercial enzymes could be used to pre-treat PKE to degrade these fibrous compounds to usable monosaccharides to enhance their metabolizable energy before feeding them to fish and poultry. However, most of the enzymes tested were imported and not specifically designed to break down the lignocellulosic components in PKE (Ibrahim 2008).

Discovery of more robust local fungi for the production of enzymes to specifically digest PKE may be more effective and economical. Filamentous fungi have been widely used to produce hydrolytic enzymes for industrial applications. Those belonging to the genus Aspergillus are commonly used for the production of cellulase (Bakir et al. 2001), mannanase (Kurakake and Komaki 2001; Lin and Chen 2004; Puchart et al. 2004), and xylanase (Lu et al. 2003).

Solid-state fermentation (SSF) has gained renewed interest in recent years for the production of enzymes due to its lower operating costs and energy requirements, as well as requiring simpler plant equipment compared to submerged fermentation (Mitchell and Losane 1992; Pandey 2003). Nevertheless, the efficacy of SSF depends on several factors, such as initial pH, temperature, moisture content, and inoculum size (Mitchelle and Losane 1992; Baysal et al. 2003). Response surface methodology (RSM) is a collective statistical technique that has recently been used to model and optimize several bioprocess, including fermentation, enzymatic reactions, product recovery, and enzyme immobilization techniques (Ismail 2005; Levin et al. 2008; Bonugli-Santos et al. 2010; Su et al. 2011; Zhang et al. 2011). A second-order model like the central composite design (CCD) is widely used in RSM, because it can take on a wide variety of functional forms, and this flexibility allows it to predict the true response surface more closely. This approach has been successfully employed to maximize production of enzyme in SSF (Abdeshahian et al. 2010; Abdeshahian et al. 2011; Coman and Bahrim 2011). However, most of the above studies focused on the optimization of single enzyme production. In this study, CCD-based RSM was used to optimize the fermentation parameters for co-production of endoglucanase, mannanase, and xylanase using A. terreus K1 (which was isolated in this study) in SSF using PKE as the sole carbon source.

EXPERIMENTAL

Sample Collection

PKE was collected from two commercial kernel oil extraction mills in Malaysia; Klang in Selangor and Kuantan in Pahang from the west- and east-coast of Peninsula Malaysia, respectively. The fresh samples were divided into two equal portions; one portion was immediately packed and stored at 4 °C for the isolation of fungi, and the other was ground and passed through a sieve of 2.5 mm and stored at 4 °C to be used for SSF studies (Saenphoom et al. 2011).

Isolation, Screening, and Identification of Potential Isolates

Serial dilution technique was used for the isolation of effective fungal strains. For this technique, 0.1 mL diluent was pipetted onto potato dextrose agar plates, spread with a glass spreader, and incubated at 30 °C for 5 to 7 days for observation. Each formed colony was transferred onto a fresh PDA plate, sub-cultured, and maintained on PDA slant at 4 °C with periodic (30 days) sub-culturing.

Spore suspensions were prepared by adding Tween-80 (0.1%) to 5-day-old cultures grown on PDA slant at 30 °C and gently brushing the mycelium with a sterile wire loop. Spores were counted using a hemocytometer, and the concentration of the spore suspension was adjusted to a ﬁnal spore count of 1.0 × 10⁷ spores/mL.

To screen for the best enzyme producer, each isolate was grown in SSF at 30 °C for 7 days, using PKE as the sole carbon source. The activities of endoglucanase, mannanase, and xylanase for each isolate were assessed. The best enzyme producer was identified by the analysis of its genomic internal transcribed spacer, ITS-region, using the standard methodology of White et al. (1990) and subsequently compared with sequences in the public databases of GenBank.

Optimization of Enzyme Production

Response surface methodology was used to optimize the SSF process and enhance the endoglucanase, mannanase, and xylanase production. Design-Expert® software (version 8.0) was used for the statistical design of experiments and data analysis. A CCD with four factors and five levels using six replicated center points was employed. The range and center point values of the four independent variables are presented in Table 1. The full experimental design with respect to the real value of the independent variables and attained values for the response (endoglucanase, mannanase, and xylanase activity) is presented in Table 2. The experiment was carried out in duplicate, and the mean enzyme activity was taken as response Y. Data from the CCD (Table 2) were analyzed by the least squares method to fit the following second-order polynomial equation,

Y = B_o+ B₁X₁+ B₂X₂ + B₃X₃ + B₄X₄ + B₁₁X₁₂ + B₂₂X₂₂ + B₃₃X₃₂ + B₄₄X₄₂ + B₁₂X₁X₂ + B₁₃X₁X₃ + B₁₄X₁X₄+ B₂₃X₂X₃ + B₂₄X₂X₄ + B₃₄X₃X₄ (1)

where Y is the measured response; B₀ is the intercept term; B₁, B₂, B₃, and B₄ are linear coefficients; B₁₁, B₂₂, B₃₃, and B₁₄ are quadratic coefficients; B₁₂, B₁₃, B₂₃, and B₂₄ are interaction coefficients; and X₁, X₂, X₃, and X₄ are coded independent variables.

The statistical analysis of the model was performed using an analysis of variance (ANOVA) generated by the Design-Expert software.

Table 1. Coded Values of Variables Used in Central Composite Design

Solid-state Fermentation

Ground PKE (2.5 mm) was moistened with different volumes of distilled water to achieve the moisture content (weight of liquid/total weight of liquid plus solid)and adjusted to the initial pH as shown in Table 2. This medium was sterilized by autoclaving prior to treatment. SSF was carried out in 500-mL Erlenmeyer flasks containing 30 g of PKE, which was inoculated with different concentrations of inocula and incubated for 7 days at different temperatures, according to the experimental design.

Table 2. Central Composite Design with Experimental and Predicted Values of Enzyme Produced by Aspergillus terreus K1

Expt. = experimental

Pred. = predicted

Enzyme Extraction and Enzyme Assays

Enzymes were extracted by shaking the PKE in 50 mM citrate buffer (pH 5) at 4 °C for 24 h, centrifuging at 10,000 rpm for 10 min, and filtering through Whatman No. 1 filter paper. The filtrate was used for the analysis of endoglucanase, xylanase, and mannanase.

Endoglucanase (carboxymethylcellulase, endo-1,4-b-D-glucanase; EC 3.2.1.4) activity was determined according to Grajek’s method (1987), whereas xylanase activity was estimated using the method of Bailey et al. (1992). The concentration of free carboxymethyl glucose and xylose units that reacted with dinitrosalicylic acid reagent was estimated using the DNS method (Miller 1959). Endoglucanase and xylanase activities were expressed in international units (IU), where one IU is the amount of enzyme required to release 1 µmol of reducing sugar (glucose or xylose) equivalent in 1 mL of enzyme solution in one minute.

The β-mannanase assay was performed using the procedure specified by Megazyme (Ireland) with a slight modification. About 0.2 mL of the previously prepared PKE filtrate was added to 0.2 mL of the substrate (Azo-Carob galactomannan) solution, stirred for 5 s on a vortex stirrer, and incubated at 40 °C for 10 min. After that, 1 mL of ethanol (~95%) was added to the mixture, which was stirred continuously for another 10 s on the vortex stirrer. The mixture was allowed to equilibrate to room temperature for 10 min and was then centrifuged at 3,000 rpm for 10 min. The supernatant solution was poured directly from the centrifuge tube into a cuvette, and the absorbance was measured using a spectrophotometer (Barnstead Turner SP-380 plus, USA) at 590 nm. Different concentrations of pure endo-1,4-β-mannanase (Megazyme, Ireland) were used for the standard curve, following the same procedure as previously described.

Enzyme activity assays were carried out in triplicate, where the average enzyme activity obtained was used as the response.

RESULTS AND DISCUSSION

Isolation and Identification of Lignocellulosic-Degrading Enzyme Producers

Lignocellulosic-degrading enzymes are necessary for the degradation of the biomass cell wall. The ability of different fungi to produce these enzymes varies, depending on the carbon source and the microorganisms used. In this study, palm kernel expeller (PKE) was used as the sole carbon source. Ten fungi were initially isolated from PKE, which was obtained from commercial kernel oil extraction mills, using potato dextrose agar. Table 3 shows the enzyme activities (endoglucanase, mannanase, and xylanase) of the ten fungal isolates, with isolates F4, F3, and K1 produced the highest endoglucanase (10.32 U/g), mannanase (43.37 U/g), and xylanase (81.54 U/g) activity, respectively. However, because the endoglucanase and mannanase activities of K1 were the same as (P < 0.05) that of isolate F4 and isolate F3, respectively, isolate K1 was selected for the subsequent optimization process. The analysis of the ITS-region of Isolate F3 showed similarity to the ITS region of known Paecilomyces variotii while both ITS sequences of Isolate F4 and Isolate K1 showed similarity to Aspergillus terreus sequence deposited in the Genbank.

Table 3. Enzyme Production in Fungi Isolated from Palm Kernel Expeller

*Results are mean values ± SD (n=3)

** ^a-gValues on the same column with different superscript differ significantly (P < 0.05)

Optimization of Enzyme production

To obtain the optimum production of lignocellulosic enzymes by Aspergillus terreus K1, the four parameters that most significantly affected enzyme production were statistically optimized using RSM. The parameters were as follows: temperature (X1), moisture (X2), medium pH (X3), and inoculum concentration (X4). The maximum and minimum levels of these parameters for the tests in the CCD are shown in Table 1. To improve the accuracy of the regression model, the center point was replicated six times. A total of 30 experiments were performed following the experimental design. The experimental results and predicted activities for each enzyme as estimated from the model equations are shown in Table 2. This approach was chosen to preserve the significance of the interaction effects, which would have been lost if variables were examined one at a time while keeping the other variables constant.

Optimization of endoglucanase production

The ANOVA summary for endoglucanase production is presented in Table 4. The model validity was estimated as a function of its coefficients of determination (R²), which provided a measure of variability in the observed response values that could be explained by the experimental factors and their interactions. In this experiment, a R² value of 0.974 indicated that the model was appropriate and could be used for quantitative prediction of endoglucanase production. In addition, the large model’s F-value (39.97) implied that the model was significant (P < 0.01), and the lack of fit test result of 1.76 implied that it was insignificant relative to pure error (P > 0.05).

Table 4. Analysis of Variance (ANOVA) Table for Endoglucanase Production

Analysis of the P-values was used to check the signiﬁcance of each coefficient. This analysis was required to understand the pattern of the mutual interactions between the independent variables. The smaller the magnitude of the P, the more signiﬁcant is the corresponding coefficient. This implies that all of the first order main effects, and the interaction terms X₁X₂, X₁X₄, and X₂X₃, are highly significant (P < 0.01). On the other hand, the interaction terms between temperature and pH (X₁X₃), moisture and inoculum (X₂X₄), and pH and inoculum (X₃X₄) are not significant; there is thus no correlation between each of these variables, and their interactions did not contribute to endoglucanase production. The contour plot of these interactions shows a relatively broad plateau region (Fig. 1), indicating only small changes in endoglucanase activity when these factors were varied.

Fig. 1. Contour plot showing the effect of (a) temperature and pH, (b) moisture and inoculum, and (c) pH and inoculum on the production of endoglucanase

By applying multiple regression analysis to the experimental data, a second-order polynomial equation was found to explain endoglucanase production, regardless of the significance of coefficients (Table 5).

Table 5. Predictive Second-Order Polynomial Equation Describing the Relationship between Various Enzyme Activities

The results predicted by the model equation showed that adjusting the fermen-tation conditions to 30.4 °C, 60.5% moisture, pH 5.3, and 7.5% inoculum would favor maximum endoglucanase yield (18.05 U/g), which was close to the experimental endoglucanase activity of 20.12 U/g. The coefficients for temperature were larger than the coefficients for other factors, indicating that temperature had the most significant effect on endoglucanase production. In Table 5, X₁, X₂, X₃, and X₄ are the coded values for temperature, moisture, pH, and inoculum, respectively.

Optimization of mannanase production

The R² value of 0.9884 indicated that 98.84% of the total variability in the response could be explained by the second-order polynomial equation (Table 5). The large model F-value (85.18) indicated that the model was highly significant (P < 0.01). In addition, the small value obtained from the lack of fit test (1.52) implied that it was insignificant relative to pure error, and the small CV (5.84%) indicated the reliability of the experiment performed. All of these statistical results (Table 6) showed good agree-ment between the experimental and predicted values and implied that the mathematical models were suitable for the simulation of mannanase production in the present study. Based on the statistical analysis, only the interaction between temperature and pH had no significant effect (P < 0.05) on mannanase production. The plot (Fig. 2) was used to represent the interaction effect of all independent variable. A non-perfectly ellipse contour plot means that there were fewer interaction effect between the independent variable (Fig. 2b, 2c, 2e, and 2f) on mannanase production (Muralidhar et al. 2001). The results predicted by the model equation showed that the optimal values for mannanase production of the four variables in un-coded units were 31.2 °C, 60.8% moisture, pH 6.4, and 6.00% inoculum. Under the optimum conditions, the predicted maximum mannanase production was 42.03 U/g, which was lower than the actual experimental mannanase activity of 46.07 U/g.

Table 6. Analysis of Variance (ANOVA) Table for Mannanase Production

Fig. 2. Contour plot showing the effect of (a) temperature and moisture, (b) temperature and pH, (c) temperature and inoculum, (d) moisure and pH, (e) moisture and inoculum, and (f) pH and inoculum on the production of mannanse

Optimization of xylanase production

The statistical signiﬁcance of the ﬁtted model, essential for determining patterns of interaction between experimental variables, was evaluated (Table 7). The computed model’s F-value (44.85) with a probability value of P < 0.01 indicated that the selected quadratic regression model ﬁt well to the experimental data. The lack of ﬁt F-value (0.61) also implied that the model provided a good ﬁt to the data.

Yields of xylanase produced by Aspergillus terreus K1 are shown in Table 2. The most xylanase (339.80 U/g) was produced when the fungus was cultured at 30 °C, initial moisture of 70%, pH 4.5, and inoculum size of 6% (run 11), whereas the minimum xylanase activity (101.50 U/g) was produced when the fermentation process was conducted at incubation temperature, moisture, pH, and inoculum of 30 °C, 70%, 7.5, and 12%, respectively (run 27). By applying multiple regression analysis to the test results, the second-order polynomial equation representing xylanase production was obtained (Table 5). Using the Design-Expert software, the optimal conditions for xylanase production were predicted to be 29.3 °C, 69.6% moisture, pH 4.6, and 7.7% inoculum, with a yield of 339.68 U/g xylanase, which was close to the actual xylanase activity of 343.07 U/g.

Table 7. Analysis of Variance (ANOVA) Table for Xylanase Production

Optimization of multi-enzyme production

Co-production of endoglucanase, mannanase, and xylanase can be found in many enzyme production systems when microorganisms are grown in agro-wastes. Neverthe-less, the efficient production of this enzyme mixture is dependent on various factors such as temperature, moisture, pH, and inocula (Abdel-Sater and El-Said 2001; Facchini et al. 2011). Thus, it is important to optimize the production system to enhance the enzyme production capability of the isolated fungal strain. The results of this study showed that each individual enzyme is typically produced under a different set of conditions. To find the best environment for the co-production of these enzymes (with an emphasis on mannanase production), optimization was carried out using Design-Expert software, and the activities of all three enzymes were used as responses. It was predicted that by incubating PKE at 30.5 °C, 62.7% moisture, pH 5.8, and 6% A. terreus K1 spores (1.0 × 10^-7 spores/mL), maximum endoglucanase, mannanase, and xylanase could be obtained (17.37, 41.24, and 265.57 U/g DM, respectively). Verification of this predicted condition was conducted in triplicate, and the enzyme activities obtained (19.97, 44.12, and 262.01 U/g DM, respectively) were close to the predicted values.

It has been reported that enzyme production is subject to induction or catabolic repression (Abdel-Sater and El-Said 2001). Because PKE constitutes mainly mannan polymers, mannanase was expected to be the major enzyme produced (Lee 2007). However, our results showed otherwise. The higher xylanase activity obtained in this study could have been induced by both xylan and cellulose present in the PKE (Biely 1985; Royer and Nakas 1989) or through induction by the end-products of mannanase during fermentation of PKE, which will release glucose, mannose, or xylose (Sachslehner et al. 1998). Alternatively, the production of mannanase is growth-dependent, and it is an induced enzyme (Feng et al. 2003). That is, in the presence of an appropriate inducer, mannanase will be produced, but once depleted or when the cell is in the stationary phase, production will cease immediately.

Temperature is one of the main factors affecting the growth of fungi, although these microorganisms have been shown to be able to tolerate a wide range of tempera-tures, typically from 30 to 40 °C, and some are able to survive in extreme temperatures (≥ 50 °C) (Michael 1972; Smith and Wood 1991; Lin and Chen 2004; Wang et al. 2006; Chellapandi and Jani 2009; Sohail et al. 2009; Facchini et al. 2011). Despite the wide temperature tolerance, the optimum temperature predicted was 30 °C, a temperature that is close to the natural habitat where this fungus was isolated. Similar to the production of endoglucanase, the production of mannanase was more affected by changes in tempera-ture (a temperature above 32 °C marked a decrease in mannanase productivity). It has been proposed that the mRNA involved in mannanase synthesis is only stable within a certain temperature range. Thus, a decrease in temperature will gradually stabilize and prolong the production of this enzyme, but production will cease with further drops in temperature due to decreased biochemical processes (Feng et al. 2003).

Unlike mannanase and endoglucanase production, xylanase production was more significantly affected by changes in pH (as shown by the larger value of coefficient estimation, X4). Based on the time course of enzyme production (Lee 2007), it was observed that the production of mannanase occurred during the fungal growth stage, whereas the optimum production of endoglucanase and xylanase occurred later. During the growth of fungi, the pH will initially decrease and then increases slightly with incubation time due to the accumulation of organic acid and soluble sugars (Kurakake and Komaki 2001). Thus, too low of an initial pH will affect the production of mannanase, but the reverse will lead to an environment that is too alkaline and might deter subsequent endoglucanase or xylanase production due to the increase in pH along incubation period.

In this study, a local fungus isolated from PKE was used for the production of multiple enzymes using SSF. Though several fungal strains have been isolated and characterized for the production of enzyme (Abdeshahian et al., 2010; Abdeshahian et al., 2011; Coman and Bahrim, 2011), study on the co-production of enzymes specifically for hydrolysis of PKE using PKE as the sole substrate is limited. Through optimization, 19.97, 44.12, and 262.01 U/g of endoglucanase, mannanase, and xylanase activity were obtained. These values are higher than those (1.20, 24.0, and 1.90 U/g of endoglucanase, mannanase and xylanase activity, respectively) reported by Lee (2007) using A. wentii TISTR 3075 to ferment Palm Kernel Meal (a by-product of palm kernel oil extraction using solvent extraction method). Nevertheless, direct comparison of enzyme activities among studies is not always possible due to the lack of standardized enzyme assay conditions and variation in source of substrate used as shown in the Table 8.

Table 8. Enzyme Production by Different Fungi in SSF

CONCLUSIONS

Production of multi-enzymes (endoglucanase, mannanase, and xylanase) by locally isolated A. terreus K1 using PKE as the sole substrate was successfully optimized. It was demonstrated that temperature, moisture, pH, and inoculum affect the production of each enzyme. Thus, by controlling these variables at an optimal level, the enzyme yield can be increased
Results of the study provide a viable option for production of an economical enzyme mixture using indigenous fungi isolated from the target substrate (PKE). The enzyme mixture, specifically designed to digest PKE, can be used to enhance the digestibility of PKE to be used more efficiently in feed for poultry and other monogastric animals.

ACKNOWLEDGMENTS

The present study was supported by the LRGS Fasa 1/2012 grant UPM/700-1/3/LRGS. The authors wish to thank FELDA Kernel Products (Kilang Isisawit Semambu, Malaysia) for supplying the PKE used for the study.

REFERENCES CITED

Abdel-Sater, M. A., and El-Said, A. H. M. (2001). “Xylan-decomposing fungi and xylanolytic activity in agricultural and industrial wastes,” Intl. Biodeterioration Biodegrad. 47(1), 15-21.

Abdeshahian, P., Samat, N., Abdul Hamid, A and Wan Yusoff, W. M. (2010), “Utilization of palm kernel cake for production of β-mannanase by Aspergillus niger FTCC 5003 in solid substrate fermentation using an aerated column bioreactor,” J Ind Microbiol Biotechnol. 37(1), 103-109

Abdeshahian, P., Samat, N., Abdul Hamid, A., and Wan Yusoff, W. M. (2011). “Solid substrate fermentation for cellulase production using palm kernel cake as a renewable lignocellulosic source in packed-bed bioreactor,” Biotechnol. Bioproc. E. 16(2), 238-244.

Bailey, M. J., Biely, P., and Poutanen, K. (1992). “Interlaboratory testing of methods for assay of xylanase activity,” J. Biotech. 23(3), 257-270.

Bakir, U., Yavascaoglu, S., Guvenc, F., and Ersayin, A. (2001). “An endo-β-1,4-xylanase from Rhizopus oryzae: Production, partial purification and biochemical characterization,” Enzyme Microb. Technol. 29(6-7), 328-334.

Baysal, Z., Uyar, F., and Aytekin, C. (2003). “Solid state fermentation for production of α-amylase by a thermotolerant Bacillus subtilis from hot-spring water,” Process Biochem. 38(12), 1665-1668.

Biely, P. (1985). “Microbial xylanolytic systems,” Trends Biotechnol. 3(11), 286-290.

Bonugli-Santos, R. C., Durrant, L. R., Silva, M., and Sette, L. D. (2010). “Production of laccase, manganese peroxidase and lignin peroxidase by Brazilian marine-derived fungi,” Enzyme Microb. Technol. 46(1), 32-37.

Chellapandi, P., and Jani, A. A. (2009). “Enhanced endoglucanase production by soil isolates of Fusarium sp. and Aspergillus sp. through submerged fermentation process,” Turk. J. Biochem. 34(4), 209-214.

Coman, G., and Bahrim, G., (2011). “Optimization of xylanase production by Streptomyces sp. P12-137 using response surface methodology and central composite design,” Ann. Microbiol. 61(4), 774-779.

Facchini, F. D. A., Vici, A. C., Reis, V. R. A., Jorge, J. A., Terenzi, H. F., Reis, R. A., Lourdes, M., and Moraes Polizeli, T. (2011). “Production of fibrolytic enzyme by Aspergillus japonicas C03 using agro-industrial residues with potential application as additives in animal feed,” Bioprocess Biosyst. Eng. 34(3), 347-355.

Feng, Y., He, Z., Ong, S. L., Hu, J., Zhang, Z., and Ng, W. J. (2003). “Optimization of agitation, aeration and temperature condition for maximum β-mannanase production,” Enzyme Microb. Technol. 32, 282-289.

Grajek, W. (1987). “Comparative studies on the production of cellulases by thermophilic fungi in submerged and solid-state fermentation,” Appl. Microbiol. Biotechnol. 26, 126-129.

Ibrahim, C. O. (2008). “Development of applications of industrial enzymes from Malaysian indigenous microbial sources,” Bioresour. Technol. 99(11), 4572-4582.

Ismail, H. B. (2005). “A new approach for determination of enzyme kinetic constant using response surface methodology,” Biochem. Eng. J. 25(1), 55-62.

Kurakake, M., and Komaki, T. (2001). “Production of β-mannanase and β-mannosidase from Aspergillus awamori K4 and their properties,” Curr. Microbiol. 42(6), 377-380.

Lee, N. S. (2007). “The production of fungal mannanase, cellulase, and xylanase using palm kernel meal as substrate,” Walailak J. Sci. Tech. 4(1), 67-82.

Levin, L., Herrmann, C., and Papinutti, V. L. (2008). “Optimization of lignocellulolytic enzyme production by the white-rot fungus Trametes trogii in solid-state fermentation using response surface methodology,” Biochem. Eng. J. 39(1), 207-214.

Lin, T. C., and Chen, C. (2004). “Enhanced mannanase production by submerged culture of Aspergillus niger NCH-189 using defatted copra based media,” Process Biochem. 39(9), 1103-1109.

Lu, W., Li, D., and Wu, Y. (2003). “Influence of water activity and temperature on xylanase biosynthesis in pilot-scale solid-state fermentation by Aspergillus sulphureus,” Enzyme Microb. Technol. 32(2), 305-311.

Michael, R. T. (1972). “Effect of temperature on growth rate and development of the thermophilic fungus Chaetomium thermophile,” Mycologia 64(6), 1290-1299.

Miller, G. L. (1959). “Use of dinitrosalicylic acid reagent for determination of reducing sugar,” Anal. Chem. 31(3), 426-428.

Mitchell, D. A., and Losane, B. K. (1992). “Definition, characterization and potential,” In Solid Substrate Cultivation, H. W. Doelle, D. A. Mitchell, and C. E. Rolz (eds.), Elsevier Science Publishers Ltd., 1-16.

Ng, W. K., Lim, H. A., Lim, S. L., and Ibrahim, C. O. (2002). “Nutritive value of palm kernel meal pretreated with enzyme or fermented with Trichoderma koningii (Oudemans) as a dietary ingredient for red hybrid tilapia (Oreochromis sp.),” Aquacult. Res. 33(15), 1199-1207.

Ong, L. G. A., Abd-Aziz, S., Noraini., S., Karim, M. I. A., and Hassan, M. A. (2004). “Enzyme production and profile by Aspergillus niger during solid substrate fermentation using palm kernel cake as substrate,” Appl. Biochem. Biotechnol.118(1-3), 73-79.

Pandey, A. (2003). “Solid-state fermentation,” Biochem. Eng. J. 13(2-3), 81-84.

Puchart, V., Vrsanska, M., Svoboda, P., Pohl, J., Ogel, Z. B., and Biely, P. (2004). “Purification and characterization of two forms of endo-β-1,4-mannanase from a thermotolerant fungus, Aspergillus fumigatus IMI 385708 (formerly Thermomyces lanuginosus IMI 158749),” Biochem. Biophys. Acta 1674(3), 239-250.

Rashid, S. A., Ibrahim, D., and Omar, I. C., (2012), “Mannanase production by Aspergillus niger USM F4 via solid substrate fermentation in a shallow tray using palm kernel cake as a substrate,” Malaysian Journal of Microbiology 8(4), 273-279.

Royer, J. C., and Nakas, J. P. (1989). “Xylanase production by Trichoderma longibrachiatum,” Enzyme Microb. Technol. 11(7), 405-410.

Sachslehner, A., Nidetzky, B., Kulbe, K. D., and ltrich, D. (1998). “Induction of mannanase, xylanase, and endoglucanase activities in Sclerotium rolfsii,” Appl. Environ. Microbiol. 64(2), 594-600.

Saenphoom, P., Liang, J. B., Ho, Y. W., Loh, T. C., and Rosfarizan, M. (2011). “Effect of enzyme treatment on chemical composition and production of reducing sugar in palm (Elais guineenis) kernel expeller,” Afr. J. Biotechnol. 10, 15372-15377.

Smith, D. C., and Wood, T. M. (1991). “Xylanase production by Aspergillus awamori. Development of a medium and optimization of the fermentation parameters for the production of extracellular xylanase and β-xylosidase while maintaining low protease production,” Biotechnol. Bioeng. 38(8), 883-890.

Sohail, M., Siddiqi, R., Ahmad, A., and Khan, S. A. (2009). “Cellulase production from Aspergillus niger MS82: Effect of temperature and pH,” New Biotech. 25(6), 437-441.

Soltan, M. A. (2009). “Growth performance, immune response and carcass traits of broiler chicks fed on graded level of palm kernel cake without or with enzyme supplementation,” Livestock Research for Rural Development 21(3), 37.

Su, Y., Zhang, X., Hou, Z., Zhu, X., Guo, X., and Ling, P. (2011). “Improvement of xylanase production by thermophilic fungus Thermomyces lanuginosus SDYKY-1 using response surface methodology,” New Biotech. 28(1), 40-46.

Sundu, B., Kumar, A., and Dingle, J. (2006). “Palm kernel meal in broiler diets: Effect on chicken performance and health,” World Poultry Science Association 62(2), 316-325.

Vasta, V., Nuddha, A., Cannas, A., Lanza, M., and Priolo, A. (2008). “Alternative feed resources and their effects on the quality of meat and milk from small ruminants,” Anim. Feed Sci. Technol. 147(1-3), 223-246.

Wang, X. J., Bai, J. G., and Liang, Y. X. (2006). “Optimization of multi-enzyme production by two mixed strains in solid-state fermentation,” Appl. Microbiol. Biotechnol. 73(3), 533-540.

White, T., Bruns, T., Lee, S., and Taylor, J. (1990). “Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics,” PCR Protocols: A Guide to Methods and Applications, 315-322.

Zhang, J., Jia, S., Liu, Y., Wu, S., and Ran, J. (2011). “Optimization of enzyme-assisted extraction of the Lycium barbarum polysaccharides using response surface methodology,” Carbohyd. Polymers 86(2), 1089-1092.

Article submitted: April 4, 2013; Peer review completed: May 21, 2013; Revised version received and accepted: May 29, 2103; Published: June 3, 2013.