Abstract: The optimization of fermentation medium for the production of chitinase by Bacillus licheniformis TH-1 was carried out using Response Surface Methodology (RSM) based on the two level factorial design. This procedure limited the number of actual experiments performed while allowing for possible interactions between 5 components. RSM was adopted to derive a statistical model for the effect of chitin, Yeast Extract (YE), peptone, NaNO3 and K2HPO4 on chitinase production. The p-value of the coefficient for linear effects of chitin, peptone and YE was 0.0001, suggesting that this was the principal experiment variable, having the greatest effect on the production of chitinase. The optimal combinations of media constituent for maximum chitinase production are determined as 10 g L-1 chitin, 0.5 g L-1 YE, 0.5 g L-1 peptone, 2.55 g L-1 NaNO3 and 1.55 g L-1 K2HPO4. The optimization of the fermentation medium resulted not only in a 5.4 fold increase of enzyme activity compared to unoptimized medium but also a reduced amount of the required medium constituents. The response surface analysis provided a useful tool for the optimization of a low cost enzyme producing medium for potential use on an industrial scale.
INTRODUCTION
Chitinases are the enzyme responsible for biological hydrolysis of chitin to its monomer N-acetyl D-glucosamine (Sahai and Monacha, 1993) and have been found to be produced by a number of microorganisms. Microbial chitinases attracted the attention as one of the potential enzyme for applications in agriculture, pharmaceutical, waste management, biotechnology and industry. Their high demand and wide potential uses, has led to the discovery of new strains of microorganisms that capable to produce enzymes with novel properties and the development of low cost industrial media formulations. Microorganisms such as Paenibacillus sp. CHE-N1 (Kao et al., 2007), Penicillium chrysogenum (Patidar et al., 2005), Serratia marcescens (Nawani and Kapadnis, 2001), Bacillus cereus (Pleban et al., 1997), Aspergillus carneus (Sherief et al., 1991) and Aeromonas sp. (Ueda et al., 2003) are capable to produce chitinase.
Studies on medium optimization for chitinases production are the worthwhile technique for multifactor experiments because it is less time consuming and capable of detecting the true optimum of the factor. In addition, medium compositions greatly influence the microbial production of extracellular chitinase and their interaction play an important role in the synthesis of this enzyme. On the other hand, medium optimization is very important not only to maximize the yield and productivity, but also to minimize the production cost (Park et al., 2005). In most cases, chitin (colloidal chitin, chitin flakes or chitin powder) was utilized as a carbon source in the production of chitinase (Gohel et al., 2006; Nawani and Kapadnis, 2005; Vaidya et al., 2003; Huang et al., 1996; McCormack et al., 1991). Studies on the medium optimization for chitinase production using the statistical approach have been done by Gohel et al. (2005, 2006), Nawani and Kapadnis (2005) and Andrade et al. (2003). The commercial interest of utilizing chitin and its derivatives to produce various products lead to the need of inexpensive, reliable source of active and stable chitinase preparations. Moreover, there is a growing interest in the production of monomers, such as N-acetyl-D-glucosamine (GlcNAc) and D-glucosamine (GlcN) from chitin hydrolysis (Ramirez-Coutiho et al., 2006). Hence, the objective of this study was to optimize the fermentation medium by applying response surface methodology for the production of chitinase from laboratory to the pilot scale. In addition, response surface methodology using the fractional factorial central composite design experiment can be used to develop a mathematical correlation between chitin, nitrogen and mineral salt for the optimum chitinase production by B. licheniformis TH-1.
MATERIALS AND METHODS
Microorganisms and culture conditions: The bacterium used in this study,
Bacillus licheniformis TH-1 was supplied by research collaborator from
UTM Skudai; Johore, Malaysia. The strain was kept as glycerol stock at –81°C.
The microbe was grown on modified chitinase-detection agar (CHDA) prior to their
use for inoculum preparation. Basically, the medium used in this study consisted
of colloidal chitin as a carbon source and a mixture of yeast extract and peptone
as a nitrogen source. The medium 4 consisted of 2 g L-1 colloidal
chitin, 3.5 g L-1 bacteriological peptone (Oxoid), 1.5 g L-1
yeast extract (Oxoid), 1.6 g L-1 NaNO3 (Unilab), 1 g L-1
K2HPO4 (R and M), 0.5 g L-1 KCl (Univar), 0.5
g L-1 MgSO4.7H2O (Riedel de Haen), 0.01 g L-1
FeSO4.7H2O (Univar) (Kawachi et al., 2001). The
final pH of the fermentation medium was adjusted to 7.0 before the inoculation.
The 10% (v/v) culture with 0.5 OD600 was used as an inoculum. The
fermentation process was carried at 150 rpm and temperature at 45°C for
24 h. Preliminary study using optimized culture condition was conducted in 500
mL of shake flask to investigate the performance of chitinase enzyme production
by Bacillus licheniformis TH-1. Samples were taken at certain time intervals
and centrifuged at 7000 rpm for 5 min. The supernatant was then used for residual
NAG (N-acetyl D-glucosamine) and chitinase enzyme assays.
| Table 1: | Variables in real values, for screening by the 2 level fractional factorial design |
Analysis: Enzyme activity was determined by an adaptation of the chitinase assay established by Rojaz-Avelizapa et al. (1999) by detecting the amount of reducing sugar liberated from the hydrolysis of the chitin polymer to the simpler forms of N-acetyl-D-glucosamine monomers (NAG). One unit of chitinase activity is defined as the amount of enzyme required to form 1 μmol of NAG in an hour at 50°C. The reducing sugar released (as NAG) was measured by the DNS method (Miller, 1959) and the amount of monomer released was extrapolated from the standard graph of NAG.
Experimental design: A two-level fractional factorial design was employed to determine the maximum chitinase production at optimum levels for chitin, yeast extract, peptone, NaNO3 and K2HPO4. Five variables, which were expected to have a significant effect on chitinase production, were identified by preliminary experiments. The minimum and maximum range of variables considered for the design are shown in the Table 1. The full experimental plan with respect to their actual and coded forms which contains a total of 20 experimental trials involving 4 replicates at the centre points are shown in the Table 2. The design was employed by selecting chitin, yeast extract, peptone, NaNO3, K2HPO4 and 1 response of chitinase activity. Each independent variable was investigated at three different levels of a high (+1), center points (0) and a low (-1) level. Runs of center points were included in the matrix and statistical analysis was used to identify the effect of each variable on chitinase production. The runs were randomized for statistical reasons (based on fractional factorial design results obtained using the statistical software package Design Expert 6.0.1, Stat-Ease Inc., Minneapolis, USA). The variables having major effects on chitinase production were identified for the isolates on the basis of confidence levels above 95% (p<0.05). The response surface graphs were obtained using the statistical software to understand the effect of variables individually and in combination and to determine their optimum level for maximal chitinase production. The optimize medium compositions will be further used in fermentation process to scale-up the chitinase production.
RESULTS AND DISCUSSION
Chitinase production: Initially, the basal medium for chitinase production containing (in g L-1) chitin, 2; yeast extract, 1.5; peptone, 3.5; NaNO3, 1.6, K2HPO4, 1.0; KCl, 0.5; MgSO4.7H2O, 0.5 and FeSO4, 0.01 were used for the production of chitinase. From the shake flask fermentation, B. licheniformis TH-1 was found to produce maximum chitinase activity 0.215 U mL-1 after 10 h of fermentation time.
Effect of medium components on chitinase production: For optimization of the chitinase production, the combinations of design experiments with the observed responses of 20 formulations including four center points were determined. A functional factorial design was applied to derive a statistical model for the effects of the medium formulations on chitinase production by B. licheniformis TH-1 and to identify the combination of factors that would lead to the enhancement of chitinase yield. The concentration ranges of the medium components were established on the basis of the data reported by Gohel et al. (2006) and Nawani and Kapadnis (2005). The results showed that the chitinase yields varied within the range of 0.242 to 1.163 U mL-1. The highest chitinase activity (1.163 U mL-1) was obtained from medium formulation No. 8 (Table 2) which contained (in g L-1): chitin, 10; peptone, 0.5; yeast extract, 0.5; NaNO3, 5 and K2HPO4, 3, while the lowest chitinase concentration (0.242 U mL-1) was obtained in medium formulation no. 10 which contained (in g L-1): chitin, 10; peptone, 0.5; yeast extract, 0.5; NaNO3, 5 and K2HPO4, 0.1.
Medium supplemented with low level of yeast extract and peptone and high level of NaNO3 and K2HPO4 produced the highest chitinase activity. After treatments combinations, the response data for chitin, yeast extract, peptone, NaNO3 and K2HPO4 yielded significant terms. Therefore, the level of each of five factors needs to be optimized for maximum response.
Once the variables statistically showed significant influence towards chitinase activity, the responses were identified and Central Composite Design (CCD) was performed to determine the optimal level of medium constituents and their interaction (Box and Wilson, 1951). A CCD with 6 replicates at the centre points leading to a total of 22 experiments was employed. Table 3 shows the design and results of experiments carried out by the CCD design. According to the response surface of five variables, peptone and yeast extract gave the most significant effect toward chitinase production. Yeast extract contains nitrogenous compounds, several growth factors and oligomers of NAG, so its addition in low concentrations can have a stimulating effect on cell growth (Nawani and Kapadnis, 2005).
All the linear and quadratic terms of chitin were included in the model since
these were significant terms based on the value of p<0.01. Sequential F-tests
were performed, starting with a linear model and adding terms (2FI and quadratic).
The F-statistic is calculated for each type of model and the highest order model
with significant terms normally would be chosen. The chitinase production by
B. licheniformis TH-1 can be expressed in terms of the following regression
equation (coded factors):
| Table 2: | Two level fractional factorial design for the optimisation of chitin and four nutrients for maximum chitinase activity during fermentation of B. licheniformis TH-1, as well as the experimental values of chitinase activity |
| A: Chitin, B: Peptone, C: Yeast extract, D: NaNO3, E: K2HPO4 | |
| Table 3: | Central composite design for the optimization of chitin and four nutrients for maximum chitinase production during fermentation of B. licheniformis TH-1, as well as the experimental values of chitinase activity |
A: Chitin, B: Peptone, C: Yeast extract, D: NaNO3, E: K2HPO4 |
|
| Table 4: | Analysis of variance for the reduced quadratic model |
SS: Sum of Squares, df: Degree of freedom |
|
Y =0.14+0.14A+0.0287B+0.0208C–0.00499A2–
0.00582A*B – 0.0076A*C |
(1) |
where, A is the chitin, B is the peptone and C is the yeast extract.
The quadratic model in Eq.1 with 6 terms contains 3 linear terms, 1 quadratic term and 2 two factorial interactions.
The ANOVA analysis of the optimization study indicated that the model terms
A, A2, AB and AC are significant model in terms of chitinase production
(prob>F is less than 0.05). The results of variance analysis are shown in
Table 4. The quadratic models derived from RSM can be adequately
used to describe the medium concentrations and the chitinase yield (Y) under
a wide range of operating conditions. Thus, a reduced quadratic model was selected
for the analysis. The model F-value is 15.79 and lack of fit F-value is 1.68,
(the lack of fit is not significant relative to the pure error). The fitness
between developed model and experimental data can be determined based on coefficient
value (R2).
| Fig. 1: | Effects of (a) chitin and peptone (A and B) and (b) chitin and yeast extract (A and C) and their interactive effect on chitinase production |
High F-value and non-significant lack of fit indicate the model is a good fit.). In this case, the R2 value (multiple correlation coefficient) is 0.863 (a value >0.75 indicates aptness of the model). It is an estimate of the fraction of overall variation in the data accounted for by the model can explain 90.42% variation in the response. The value of adjusted R2 and predicted R2 values are 0.809 and 0.685, respectively. For a good statistical model, R2 value should be close to 1.0 and all the five factors should be positive and close to each other. Also, the model has an adequate precision value of 13.3, which suggests that the model can be used to navigate the design space.
The fitted response for the above regression model was shown in Fig. 1. The 3D response surface curve shown the variation of chitinase activity, as a function of concentrations of two medium components (peptone and yeast extract) with the other two (NaNO3 and K2HPO4) being at their constant levels (obtained through analysis of variance). It is easy and convenient to understand the interactions between two nutrients and also to locate their optimum levels. Combination of high chitin concentration and low yeast extract concentration are the key factors that influence chitinase activity. An increase in chitin concentration in medium supplemented with low yeast extract can further enhance chitinase production was also suggested by Nawani and Kapadnis (2005).
The predicted optimum levels of tested variables (chitin (A), peptone (B), yeast extract (C), NaNO3 and K2HPO4) were obtained by using regression analysis of Eq. 1. The optimal levels for the variables were as follows: A = 10 g L-1, B = 0.5 g L-1 and C = 0.501 g L-1 with the corresponding Y = 0.864 U mL-1. To validate this model, an experiment was conducted using optimal medium with an addition of 2.55 g L-1 NaNO3 and 1.55 g L-1 K2HPO4 and the activity values were measured. The maximum chitinase activity was found to be 0.844 U mL-1 where the model predicted a value of 0.996 U mL-1 under the same conditions. This result corroborated the validity and the effectiveness of this model.
CONCLUSION
A highly significant quadratic polynomial obtained from CCD was very useful for determination the optimal concentrations of constituents that gave significant effects on chitinase production. Under the optimal condition, 0.864 and 0.844 U mL-1 of chitinase activity could be produced in theory and practical experiment, respectively. Medium No. 8 which formulated was superior as compared to other medium in terms of original compositions for enhancing chitinase production. Linear model obtained from experimental data showed chitin, peptone, yeast extract, NaNO3 and K2HPO4 gave positive effect on chitinase production. The methodology used in this work proved to be adequate for the design and optimization of fermentation process for production of potential valuable product such as chitinase from chitinous waste generated from aquaculture waste.