Journal of Genetic Engineering and Biotechnology (2013) 11, 111-116

Academy of Scientific Research & Technology and National Research Center, Egypt

Journal of Genetic Engineering and Biotechnology

www.elsevier.com/locate/jgeb

ARTICLE

Statistical optimization of critical medium components for lipase production from Yarrowia lipolytica (MTCC 35)

G. Kishan a, P. Gopalakannan a, C. Muthukumaran b,

K. Thirumalai Muthukumaresan c, M. Dharmendira Kumar d, K. Tamilarasan a'*

a Department of Biotechnology, Madha Engineering College, Kundrathur 600069, Tamilnadu, India b Department of Industrial Biotechnology, Government College of Technology, Coimbatore 641013, Tamilnadu, India c Department of Biotechnology, Indian Institute of Technology Madras, Chennai 603203, Tamilnadu, India d Department of Applied Science and Technology, Anna University, Chennai 600025, Tamilnadu, India

Received 25 January 2013; revised 18 April 2013; accepted 2 June 2013 Available online 26 June 2013

KEYWORDS

Yarrowia lipolytica; Lipase; Mahua cake; Plackett-Burman design; Central composite design;

Abstract Statistical optimization is an effective technique for the investigation of complex processes with minimal number of experimental runs. In this study, statistical approach was used to study the optimization of media components for lipase production from Yarrowia lipolytica MTCC 35. Mahua cake, glucose, MnCl2 and KH2PO4 were screened to be the most significant variables among the nine medium variables that were tested to determine influence on lipase production by Plackett-Burman design. Central Composite Design was used for further optimization of these screened variables for enhanced lipase production. The determination coefficient (R2) value of 0.922 showed that the regression models adequately explain the data variation and represent the actual relationships between the variables and response. The optimum values of investigated variables for the maximum lipase production were 6.0% Mahua cake, 2.0% glucose, 0.2% MnCl2 and 0.2% KH2PO4. The maximum lipase production (9.40 U mL_1) was obtained under optimal condition.

© 2013 Production and hosting by Elsevier B.V. on behalf of Academy of Scientific Research &

Technology.

* Corresponding author. Present address: Department of Chemical Engineering, SRM University, Kattankulathur 603203, Tamilnadu, India Tel.: +91 98 40017648.

E-mail address: tamilbio@gmail.com (K. Tamilarasan).

Peer review under responsibility of National Research Center, Egypt.

1. Introduction

Lipases are enzymes that catalyze the hydrolysis of triglycerides to fatty acids and glycerol which have wide uses in modification of fats and oils [8,9,19]. Lipases also catalyze several industrially significant catalytic biotransformation reactions namely esteri-fication, transesterification, alcoholysis and acidolysis [3,13,18]. Lipases are the choice of biocatalyst as they show unique che-mo-, regio-, enantioselectives, which enable the production of novel drugs, agrochemicals and fine products [5,6,7,14,16].

1687-157X © 2013 Production and hosting by Elsevier B.V. on behalf of Academy of Scientific Research & Technology. http://dx.doi.org/10.1016/jjgeb.2013.06.001

Recently, many microorganisms are known to be good producers of extracellular lipases [21]. Increased productivity of lipase during the fermentation process is of great importance since lower costs of production could promote new industrial applications. The productivity of lipase is affected by different factors such as temperature, pH, medium composition and presence ofinducers among others. The traditional one-variable at a time method is simple, but the interaction effects of variables are not considered. Also, this method requires more experimental runs to determine the optimum levels, which are incorrect. These drawbacks of one-variable at a time optimization process can be eliminated by optimizing all the affecting parameters collectively by using response surface methodology (RSM) [4,11,15,20]. Statistical model is widely used to study an aggregate effect of several variables and to seek optimum conditions for a multivariable system [10]. The purpose of this study is to investigate the effect of medium components and their interaction on lipase production by Yarrowia lipolytica (MTCC 35) using statistical methods.

2. Materials and methods

2.1. Chemicals and micro organism

All the media components were of analytical grade and were purchased from HiMedia Laboratories Pvt Ltd., Mumbai, India. The solvents were purchased from Merck Specialties Pvt Ltd., Mumbai, India. The strain (Y. lipolytica, MTCC 35) was obtained from Microbial Type Culture Collection, Institute of Microbial Technology, Chandigarh, India.

2.2. Culture maintenance and media

The culture was maintained at deep freeze condition by storing in media containing glycerol in equal proportions. Slants were grown at 30 0C for 24 h and stored in refrigerator. The cells were cultivated in the inoculum media consisted of 1% glucose, 0.5% peptone and 0.3% yeast extract and incubated at 30 0C for 24 h. The production media used for lipase production consisted of following components (per liter) Mahua oil cake 20 g, ammonium chloride 1 g, glucose 1 g, gelatin 0.5 g, KH2PO4 0.2 g, MgSO4 0.1 g, NaCl 0.1 g, CaCl2 0.1 g, and MnCl2 0.1 g and the pH of the medium was adjusted to 7.5.

2.3. Enzyme assay

Enzyme activity in the supernatant of the culture was determined using p-nitro phenol palmitate (pNPP) as substrate In brief, a stock solution (50 mM) of pNPP was prepared in the ratio of 1:1 of acetonitrile and n-butanol. The reaction mixture contained 20 iL of pNPP stock solution, 200 iL of cell-free culture supernatant (crude lipase), and 1.8 mL of Tris-buffer (0.1 M, pH 8.0).The reaction mixture was incubated at 35 0C for 10 min. The amount of liberated 4-nitrophenol was recorded at 400 nm. One unit of lipase activity was defined as micro moles of para nitro phenol released from pNPPmL-1 min-1 under standard assay conditions [12].

2.4. Experimental designs

The first step in the optimization strategy was to identify medium components that have a significant effect on lipase production [2]. For screening purpose, various medium components have been evaluated using Plackett-Burman (PB) statistical design, which is a two level factorial design [17]. A set of 12 experiments was constructed using Minitab 14 software for nine medium components namely Mahua cake (X1), ammonium chloride (X2), gelatin (X3), KH2PO4 (X4), glucose (X5), NaCl (X6), MgSO4 (X7), CaCl2 (X8) and MnCl2 (X9). Each component was tested at two different concentration levels. The experiments were carried out in 250 mL Erlenmeyer flasks containing 100 mL of production media kept in the temperature controlled rotary shaker at 120 rpm and 35 oc. The response was measured as lipase activity. The variables with confidence level greater than 90% were considered to be more significant for lipase production. Pareto chart displays the magnitude of each factor estimate, and it is convenient way to view the results of PB design. Subsequently another statistical approach, Central Composite Design (CCD) was employed to study the interaction between the four parameters that proved to be significant from PB design [1]. Each parameter was studied at five different levels (-2, -1,0, +1, and +2). The minimum and maximum ranges of parameters were investigated and the full experimental plan with respect to their values. A matrix of thirty-one experiments with four factors was generated using the software package Minitab14. The lipase activity was taken as the dependent variable or response (Y).

Table 1 Plackett-Burman experimental design for screening of media components for lipase production.

Exp. No. Mahua NH4Cl Gelatin KH2PO4 Glucose NaCl MgSO4 CaCl2 MnCl2 Lipase activity

cake (Xj) (X2) (X3) (X4) (X5) (X6) (X7) (X8) (X9) (UmL-1)

1 8 (+1) 0.5(-1) 0.6 (+ 1) 0.1(-1) 0.5(- 1) 0.05(- 1) 0.15 (+ 1) 0.15 (+ 1) 0.15 (+ 1) 9.0

2 8 (+1) 1.5 (+1) 0.4(- 1) 0.3(+1) 0.5(- 1) 0.05(- 1) 0.05 (- 1) 0.15(+ 1) 0.15 (+ 1) 8.6

3 4(-1) 1.5 (+1) 0.6 (+ 1) 0.1(-1) 1.5 ( 1) 0.05(- 1) 0.05 (- 1) 0.05(- 1) 0.15 (+ 1) 7.0

4 8 (+1) 0.5(-1) 0.6 (+ 1) 0.3 (+1) 0.5(- 1) 0.15(+ 1) 0.05 (- 1) 0.05(- 1) 0.05(- 1) 8.1

5 8 (+1) 1.5 (+1) 0.4(- 1) 0.3 (+1) 1.5 (+ 1) 0.05 ( -1) 0.15 (+ 1) 0.05(- 1) 0.05(- 1) 7.1

6 8 (+1) 1.5 (+1) 0.6 (+ 1) 0.1(-1) 1.5 (+ 1) 0.15(+ 1) 0.05 (- 1) 0.15 (+ 1) 0.05(- 1) 7.0

7 4 (-1) 1.5 (+1) 0.6 (+ 1) 0.3 (+1) 0.5(- 1) 0.15(+ 1) 0.15 (+ 1) 0.05(- 1) 0.15 (+ 1) 7.1

8 4(-1) 0.5(-1) 0.6 (+ 1) 0.3 (+1) 1.5 (+ 1) 0.05 ( -1) 0.15 (+ 1) 0.15 (+ 1) 0.05(- 1) 5.1

9 4(-1) 0.5(-1) 0.4(- 1) 0.3 (+1) 1.5 (+ 1) 0.15 (+ 1) 0.05 (- 1) 0.15 (+ 1) 0.15 (+ 1) 5.8

10 8 (+1) 0.5(-1) 0.4(- 1) 0.1(-1) 1.5 (+ 1) 0.15 (+ 1) 0.15 (+ 1) 0.05(- 1) 0.15 (+ 1) 7.7

11 4(-1) 1.5 (+1) 0.4(- 1) 0.1(-1) 0.5(- 1) 0.15 (+ 1) 0.15 (+ 1) 0.15 (+ 1) 0.05(- 1) 6.9

12 4(-1) 0.5(-1) 0.4(- 1) 0.1(-1) 0.5(- 1) 0.05(- 1) 0.05 (- 1) 0.05 ( 1) 0.05(- 1) 6.6

Table 3 Experimental range and levels of the independent variables used in CCD.

Figure 1 Pareto chart representing the order of the significant medium variables on lipase production by Yarrowia lipolytica.

This resulted in an empirical model (Eq. (1)) that related the measured response and the independent factors:

y=bo+£ bx+£ bx2+£ E

j=1 j=1

where Y is the response, bo, bj, bjj, bj are regression coefficients for the intercept, linear, quadratic and interaction effects respectively and X and Xj are coded independent variables.

3. Results and discussion

3.1. Screening of significant variables using Plackett-Burman design

Plackett-Burman experimental design was employed to evaluate factors which significantly affect the lipase production by Y. lipolytica MTCC 35. The Plackett-Burman design for 12 trials with two levels of concentrations for nine different variables were carried out according to the experimental matrix as shown in Table 1 and lipase activity was measured as response. The maximum lipase activity (9.0 U mL_1) was observed in trial number 1, while the minimum lipase activity (5.1 U mL_1) was observed in trial number 8. Main effects of the examined variables on lipase production were calculated and presented graphically in Pareto chart (Fig. 1). Significance of each variable was determined using Student's t-test. The components were screened at the confidence level of 90% on the basis of their effects. Table 2 represents the result of Plackett-Burman

Factor Variables Range of levels

+ 2 +1 0 -1 -2

X1 Mahua cake (%, w/v) 14 10 6 2 0

X2 Glucose (%, w/v) 4 3 2 1 0

X3 MnCl2 (%, w/v) 0.4 0.3 0.2 0.1 0

X4 KH2PO4 (%, w/v) 0.4 0.3 0.2 0.1 0

trials with respect to t-value, p-value and confidence level of each component. Out of nine variables studied, the lipase production was highly influenced by an increase in the concentration of Mahua cake. The positive effect of glucose and MnCl2 was very similar. Based on the results from PB design, further optimization study was carried out by CCD.

3.2. Optimization of media components using central composite design

Central composite design was mainly employed to estimate the optimum level of each variable along with their interactions on lipase activity. Based on the results of Plackett-Burman design, Mahua cake (X0, glucose (X2), MnCl2 (X3) and KH2PO4 (X4) were selected and experimental range is shown in Table 3. Central composite design matrix with observed and predicted results is given in Table 4. The experimental results were fitted with a second-order polynomial equation. The values of regression coefficients (in coded units) were calculated and the fitted equations for predicting enzyme activity are as follows:

Lipase activity(U mL-1) = 9.113 + 0.908 X0.280 X2

- 0.106 X3 + 0.167 X4

- 1.601 X2 - 0.789 X2

- 0.426 X2 - 0.514 X4

+ 0.034 X X2 + 0.302 X X3

- 0.422 X X4 - 0.382 X2 X3

- 0.466 X2 X4 + 0.074 X3 X4.

The significance of each coefficient was determined by stu-dent's't' test and 'p' values, which are listed in Table 5. The larger the magnitude of t' value and the smaller the p' value,

Table 2 Statistical analysis of Plackett-Burman design results on lipase production.

Variable Main effect Coefficients t-Value p-Value Confidence level %

Constant 7.166 61.74 0.000 100

Mahua cake (%, w/v) 1.500 0.750 6.46 0.023 97.7

Ammonium chloride (%, w/v) 0.233 0.116 1.01 0.421 57.9

Gelatin (%, w/v) 0.100 0.050 0.43 0.709 29.1

KH2PO4 (%, w/v) -0.400 -0.200 -1.72 0.227 77.3

Glucose (%, w/v) -1.100 -0.550 -4.74 0.042 95.8

NaCl (%, w/v) -0.133 -0.066 -0.57 0.624 37.6

MgSO4 (%, w/v) -0.033 -0.016 -0.14 0.899 10.1

CaCl2 (%,w/v) -0.200 -0.100 -0.86 0.480 52.0

MnCl2 (%, w/v) 0.733 0.366 3.16 0.087 91.3

Table 4 CCD matrix of independent variables used in RSM with corresponding experimental and predicted values of lipase activity.

Std order X1 X2 X3 X4 Lipase activity (U mL-1) Experimental Predicted

1 -1 -1 -1 -1 3.00 4.32

2 + 1 -1 -1 -1 6.43 5.86

3 -1 +1 -1 -1 5.45 5.89

4 +1 +1 -1 -1 8.45 8.13

5 -1 -1 +1 -1 3.12 3.50

6 +1 -1 +1 -1 6.45 5.81

7 -1 +1 +1 -1 3.54 4.61

8 +1 +1 +1 -1 7.25 7.62

9 -1 -1 -1 +1 5.65 6.11

10 +1 -1 -1 +1 6.24 6.03

11 -1 +1 -1 +1 5.89 5.38

12 +1 +1 -1 +1 6.55 6.00

13 -1 -1 +1 +1 4.98 6.16

14 +1 -1 +1 +1 7.45 6.84

15 -1 +1 +1 +1 3.56 4.96

16 +1 +1 +1 +1 6.56 6.35

17 -2 0 0 0 2.55 0.65

18 +2 0 0 0 3.35 3.57

19 0 -2 0 0 5.50 6.32

20 0 +2 0 0 6.90 7.40

21 0 0 -2 0 7.10 7.54

22 0 0 +2 0 8.20 7.08

23 0 0 0 -2 7.10 6.85

24 0 0 0 +2 7.50 7.37

25 0 0 0 0 9.20 9.11

26 0 0 0 0 9.10 9.11

27 0 0 0 0 9.50 9.11

28 0 0 0 0 8.99 9.11

29 0 0 0 0 8.50 9.11

30 0 0 0 0 9.40 9.11

31 0 0 0 0 9.10 9.11

Table 5 Estimated regression coefficients , t and p values for the variables used in the RSM experiment.

Term Coefficients t-Value p-Value

X1 0.907 5.580 0.000

X2 0.280 1.723 0.104

X3 -0.106 -0.653 0.523

X4 0.166 1.022 0.322

X1X1 -1.601 -10.743 0.000

X2 X2 -0.788 -5.293 0.000

X3 X3 -0.426 -2.861 0.011

X4 X4 -0.513 -3.448 0.003

X1 X2 0.034 0.172 0.865

X1 X3 0.301 1.515 0.149

X1 X4 -0.421 -2.117 0.050

X2 X3 -0.381 -1.916 0.073

X2 X4 -0.465 -2.337 0.033

X3 X4 0.074 0.373 0.714

the more significant is the corresponding coefficient. Model coefficients were estimated by multiple linear regression and the p-value was used for checking the significance of each of the coefficients. In the present work, individual effect of X^ square effects of X1, X2, X3, X4 and interaction effect of X2 X4 were found significant for lipase production (Table 5). The statistical significance of the model was also determined

by the F-test for the analysis of variance (ANOVA), and the residuals analysis was performed to validate the model at 90% confidence level. ANOVA results showed that the model was significant with coefficient of determination (R2) of 0.922. The closer the R2 value is to 1.00, the stronger the model is and the better it predicts the response. ANOVA statistics given in Table 6 indicate that the linear and square terms in the second order polynomial Model (Eq. (2)) were highly significant (p < 0.005), and adequate to represent the relationship between lipase activity and Mahua cake, glucose, MnCl2 and KH2PO4. The optimum concentrations of the four factors of interaction for maximum lipase activity (9.40 UmL-1) were obtained with 6.0% Mahua cake, 2.0% glucose, 0.2% MnCl2 and 0.2% KH2PO4.

Response surface plots as function of two factors at a time, maintaining all other factors at a constant level, more helpful in understanding both the main and interaction effects of these factors. Fig. 2 shows the response surface plot effects of MnCl2 and KH2PO4 concentration, while other two variables are kept constant at zero level. It was observed that the lipase activity was high at middle level of both MnCl2 and KH2PO4. But li-pase activity was decreased at the higher level of KH2PO4, and less at the low and high level of manganese chloride. Fig 3 shows the response surface plot effects of MnCl2 and glucose concentration and from the figure, activity increases at the middle level of glucose and low level of MnCl2. Decrease in

Table 6 Analysis of variance of second order Polynomial model for effect of variable on lipase production.

Factors Degrees of freedom Sum of squares Mean square F-Value p-Value

Regression 14 120.448 8.6034 13.54 <0.001

Linear 4 22.605 5.6512 8.89 <0.001

Square 4 87.628 21.9069 34.48 <0.001

Interaction 6 10.215 1.7025 2.68 0.054

Residual Error 16 10.167 0.6354

Lack-of-fit 10 9.536 0.9536 9.07 0.007

Pure error 6 0.631 0.1052

Total 30 130.615

Figure 2 Three dimensional response surface plot showing the interaction effect of MnCl2 and KH2PO4 on lipase activity.

Figure 5 Three dimensional response surface plot showing the interaction effect of mahua cake and KH2PO4 on lipase activity.

Figure 3 Three dimensional response surface plot showing the Figure 6 Three dimensional response surface plot showing the interaction effect of glucose and MnCl2 on lipase activity. interaction effect of mahua cake and MnCl2 on lipase activity.

the lipase activity was observed at low, high level of glucose and also at high level of MnCl2. Fig. 4 shows the effect of glucose and KH2PO4 concentration, while two other variables were kept fixed at zero levels. It was observed from the figure

that activity increases at the middle level of glucose and KH2PO4 and the lipase activity was seem to be decreased at lower and higher level of glucose. Fig. 5 shows the response surface plot effects of Mahua cake and KH2PO4 on lipase

Figure 4 Three dimensional response surface plot showing the interaction effect of glucose and KH2PO4 on lipase activity.

Figure 7 Three dimensional response surface plot showing the interaction effect of mahua cake and glucose on lipase activity.

activity, while two other variables are kept fixed at zero levels. The increase in lipase activity was found at middle level of both Mahua cake and KH2PO4 and the decrease in lipase activity was observed in lower and higher level of Mahua cake. Fig. 6 shows the effect of Mahua cake and MnCl2 on lipase activity and maximum lipase activity was found in the middle levels of Mahua cake and MnCl2. Fig 7 shows the combined effect of Mahua cake and glucose concentration and increase in the lipase activity was found at the middle levels of both Mahua cake and glucose.

4. Conclusion

The use of statistical models to optimize medium components has increased in present bioprocess industry, due to its easy applicability and reliability. The present work, PB design was used to identify the effect of media components and their interaction on lipase production by Y. lipolytica. It is evident that various process parameters like Mahua cake, glucose, MnCl2 and KH2PO4 are the most significant factors influencing lipase production. From further optimization studies, using CCD the optimized values of the variables for lipase production were as follows: 6.0% Mahua cake, 2.0% glucose, 0.2% MnCl2 and 0.2% KH2PO4. Using the optimized conditions, the experimentally obtained lipase activity reaches 9.40 UmL"1.

Acknowledgement

Authors wish to thank the Management and Department of Biotechnology, Madha Engineering College for providing facilities to carry out this research.

References

[1] G. Baskar, S. Renganathan, Chem. Pap. 65 (2011) 798-804.

[2] R. Gaur, A. Gupta, S.K. Khare, Bioresour. Technol. 99 (2008) 4796-4802.

[3] V. Gotor, Bioorg. Med. Chem. 7 (1999) 2189-2197.

[4] N. Gupta, G. Mehra, R. Gupta, Can. J. Microbiol. 50 (2004) 361-368.

[5] R. Gupta, N. Gupta, P. Rathi, Appl. Microbiol. Biotechnol. 64 (2004) 763-781.

[6] F. Hasan, A.A. Shah, A. Hameed, Enzyme Microb. Technol. 39 (2006) 235-251.

[7] K.E. Jaeger, T. Eggert, Curr. Opin. Biotechnol. 13 (2002) 390397.

[8] K.E. Jeager, B.W. Dijkstra, M.T. Reetz, Annu. Rev. Microbiol. 53 (1999) 315-351.

[9] L. Jing, C. Shijie, L. Lin, Y. Yingpeng, F. Zhongan, Dairong, Afr. J. Microbiol. Res. 6 (2012) 637-642.

[10] K.K. Jung, O.H. Baek Rock, J.S. Hyun, E. Chi-Yong, Si Wouk Kim, Process Biochem. 43 (2008) 1308-1312.

[11] S.J. Kalil, F. Maugeri, M.I. Rodrigues, Process Biochem. 35 (2000) 539-550.

[12] K.N. Kilcawley, M.G. Wilkinson, P.F. Fox, Enzyme Microb. Technol. 31 (2002) 310-320.

[13] Y. Kosugi, N. Azuma, J. Am. Oil Chem. Soc. 71 (1994) 13971403.

[14] A. Pandey, S. Benjamin, C.R. Soccol, P. Nigam, N. Krieger, V.T. Soccol, Biotechnol. Appl. Biochem. 29 (1999) 199-231.

[15] P. Rathi, V.K. Goswami, V. Sahai, R. Gupta, J. Appl. Microbiol. 93 (2002) 930-936.

[16] R.K. Saxena, P.K. Ghosh, R. Gupta, W. Sheba Dvidson, S. Bradoo, R. Gulati, Curr. Sci. 77 (1999) 101-115.

[17] G. Sharmila, B. Nidhi, C. Muthukumaran, Ind. Crops Prod. 44 (2013) 158-164.

[18] M. Singh, K. Saurav, N. Srivastava, K. Kannabiran, Curr. Res. J. Biol. Sci. 2 (2010) 241-245.

[19] K. Tamilarasan, M. Dharmendirakumar, Biocatal. Agric. Biotechnol. 1 (2012) 309-313.

[20] K. Tamilarasan, N. Peatciyammal, L. Rajamaniccam, M. DharmendiraKumar, Res. J. Eng. Technol. 2 (2011) 31-35.

[21] H. Treichel, D. Oliveira, M.A. Mazutti, M.D. Luccio, J.V. Oliveira, Food Bioprocess Technol. 3 (2010) 182-196.