- Email: [email protected]
Optimization of protease production by Microbacterium sp. in feather meal using response surface methodology
Process Biochemistry 41 (2006) 67–73 www.elsevier.com/locate/procbio
Optimization of protease production by Microbacterium sp. in feather meal using response surface methodology Roberta C.S. Thys, Samanta O. Guzzon, Florencia Cladera-Olivera, Adriano Brandelli * Laborato´rio de Bioquı´mica e Microbiologia Aplicada, Departamento de Cieˆncia de Alimentos, ICTA, Universidade Federal do Rio Grande do Sul, Av. Bento Gonc¸alves 9500, 91501-970 Porto Alegre, Brazil Received 29 March 2004; received in revised form 28 July 2004; accepted 15 March 2005
Abstract A 23 factorial design was performed with the aim of optimizing protease production by a strain of Microbacterium sp. isolated from feathers in decomposition by response surface methodology. Protease production was first tested on different nitrogen source (casein, peptone, yeast extract, gelatin, soybean protein, feather meal and cheese whey). Feather meal was the selected substrate to test the effect of three variables on protease production (temperature, initial pH and feather meal concentration) by RSM. The point was chosen with these conditions: temperature 37 8C, initial pH 7.0 and feather meal concentration 12.5 g l1. Statistical analysis of results showed that, in the range studied, only pH did not have a significant effect on protease production whereas interaction between pH and feather meal concentration was significant. The optimum conditions were 25 8C, initial pH 7.0 and 12.5 g l1 of feather meal. Under these conditions, the model predicted a protease activity of 202.7 U ml1. # 2005 Elsevier Ltd. All rights reserved. Keywords: Actinomycetes; Microbacterium; Protease; Experimental design; Optimization
1. Introduction Proteases constitute at least 65% of the total industrial enzyme market [1]. They are used for various industrial applications, such as laundry detergents, leather preparation, protein recovery or solubilization and organic synthesis [2]. In food industry, proteases have been routinely used for various purposes such as cheesemaking, baking, preparation of soy hydrolysates and meat tenderization [1]. Bacterial neutral proteases are active within a narrow pH range and have relatively low thermal tolerance. Due to their intermediate rate of reaction, neutral proteases generate less bitterness in hydrolyzed food proteins than do the animal proteinases and hence are valuable for use in the food industry [3]. Their low thermal tolerance is advantageous for controlling their reactivity during the production of food hydrolysates with a low degree of hydrolysis [1]. * Corresponding author. Tel.: +55 51 3316 6249; fax: +55 51 3316 7048. E-mail address: [email protected] (A. Brandelli). 1359-5113/$ – see front matter # 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.procbio.2005.03.070
Optimization of medium by the classical method involves changing one independent variable (i.e., nutrient, pH, temperature) while unchanging all others at a fixed level. This is extremely time-consuming and expensive for a large number of variables [4] and also may result in wrong conclusions [5]. Response surface methodology (RSM) is a collection of statistical techniques for designing experiments, building models, evaluating the effects of factors and searching optimum conditions of factors for desirable responses [6]. This method has been successfully applied in many areas of biotechnology such as bioconversion of cheese whey to mycelia of Ganoderma lucidum [7], optimization of neomycin production [4], enzyme production [8], enzyme kinetics [9] and bacteriocin production [10]. With respect to protease production, it was utilized for example for Bacillus species [11,12]. Extracellular protease production by microorganisms is greatly influenced by physical factors such as pH, temperature and incubation time and by others factors such as media composition and presence of metal ions [13–16].
68
R.C.S. Thys et al. / Process Biochemistry 41 (2006) 67–73
The feather-degrading Microbacterium sp. strain kr10 produces an extracellular keratinase belonging to the metalloprotease family. This enzyme was also proved to present desirable properties for de-hairing bovine pelts, which would be very beneficial to the leather industry [17]. The purpose of this study was to evaluate the effect of three variables (temperature, pH and feather meal concentration) in the production of this protease. Previously, enzyme production with different by-products as growth substrate was tested.
2. Materials and methods 2.1. Reagents and media Azocasein was from Sigma Chemical Co. (St. Louis, USA). Feather meal and soybean protein were from Bunge Alimentos S.A. (Brazil). Dried cheese whey powder was from Parmalat (Porto Alegre, Brazil). Nutrient agar and peptone were from Oxoid (Basingstone, UK) and yeast extract was from Biobras (Montes Claros, Brazil). Casein and gelatin were from Synth (Diadema, Brazil). 2.2. Bacterial strain and inoculum preparation Microbacterium sp. strain kr10 was used as the producer microorganism. BHI medium containing 20% (v/v) glycerol was used for maintenance of the strain at 20 8C. The cells were first propagated in milk agar plates (5 g l1 peptone, 3 g l1 yeast extract, 12.0 g l1 agar and 100 ml l1 UHT skimmed milk). The inoculum was prepared in feather meal previously hydrolyzed with 0.4N NaOH for 2 h, neutralized with HCl, filtered and than autoclaved (121 8C for 40 min). The inoculum absorbance at 600 nm was adjusted before use with a Hitachi U-1100 spectrophotometer (Hitachi, Japan) to obtain about 1 106 CFU ml1. 2.3. Selection of nitrogen source For selection of the best nitrogen source for protease production, various inorganic and complex nitrogen sources (casein, peptone, yeast extract, gelatin, soybean protein, feather meal and cheese whey) were tested individually (10 g l1) in the minimal mineral medium (0.5 g l1 NaCl, 0.3 g l1 K2HPO4, 0.4 g l1 KH2PO4, 0.015 g l1 CaCl22H2O). Erlenmeyer flasks of 200 ml containing 50 ml of medium were inoculated with 1% of inoculum (A600 nm at 0.350) and incubated at 30 8C under shaking (125 rpm). Protease yield was determined after 48 h. 2.4. Experimental design and protease production After selection of the best medium, the next stage was determination of the optimal levels of three variables, feather meal concentration, temperature and initial pH on
Table 1 Values of independent variables at different levels of the 23 factorial design Independent variables
Symbol
Initial pH Temperature (8C) Feather meal (g l1)
x1 x2 x3
Levels 1.68
1
0
+1
+1.68
5.0 25 0.0
5.8 30 5.0
7.0 37 12.5
8.29 45 20.0
9.0 50 25.0
protease production. For this purpose, the response surface approach by using a set of experimental design (central composite design with five coded levels) was performed. For the three factors, this design was made up of a full 23 factorial design with its eight points augmented with three replications of the centre points (all factors at level 0) and the six star points, that is, points having for one factor an axial distance to the centre of a, whereas the other two factors are at level 0. The axial distance a was chosen to be 1.68 to make this design orthogonal. A set of 17 experiments was carried out. The range and levels of experimental variables investigated are presented in Table 1. The central values (zero level) chosen for experimental design were: feather meal concentration 12.5 g l1, temperature 37 8C and initial pH 7.0. The enzyme activity values were tested after 2, 4 and 6 days of incubation but the regression equation was made with the values obtained after 4 days (maximal values). In developing the regression equation, the test factors were coded according to the following equation: xi ¼
X i X0 DXi
(1)
where xi is the coded value of the ith independent variable, Xi the natural value of the ith independent variable, X0 the natural value of the ith independent variable at the center point and DXi the step change value (DXi is 1.2 for initial pH, 8 for temperature and 7.5 for feather meal concentration). For a three factors system, the model equation is: Y ¼ b0 þ b1 x1 þ b2 x2 þ b3 x3 þ b11 x21 þ b22 x22 þ b33 x23 þ b12 x1 x2 þ b13 x1 x3 þ b23 x2 x3
(2)
where Y, predicted response, b0, intercept; b1, b2, b3, linear coefficients; b11, b22, b33, squared coefficients; b12, b13, b23, interaction coefficients. Results were analyzed by the Experimental Design Module of the Statistic 5.0 software (Statsoft, USA). The model permitted evaluation of the effects of linear, quadratic and interactive terms of the independent variables on the dependent variable. The statistical significance of the regression coefficients was determined by Student’s t-test and the second order model equation was determined by Fisher’s test. Three-dimensional surface plots were drawn to illustrate the main and interactive effects of the independent variables on protease production. The optimum values of the selected variables were obtained by solving the regression
R.C.S. Thys et al. / Process Biochemistry 41 (2006) 67–73
equation and also by analyzing the response surface contour plots [6]. The production medium (50 ml in 200 ml Erlenmeyer flask) was inoculated with 1% of inoculum (A600 nm at 0.350) and incubated under different temperatures for each set of experiment in a rotary shaker at 125 rpm. After incubation the culture broth was centrifuged at 10,000 g for 10 min and total protease yield was determined in the supernatant. 2.5. Enzyme assay Protease activity was assayed with azocasein as substrate by the following method. The reaction mixture contained 120 ml of cell-free supernatant and 480 ml of 10 g l1 azocasein in 100 mmol l1 Tris buffer pH 7.0 was incubated for 30 min at 45 8C (for the control, the reaction was stopped after the addition of the supernatant). The reaction was then stopped by the addition of tricoroacetic acid to a final concentration of 100 g l1 and incubated 30 min at 4 8C. After centrifugation at 10,000 g for 10 min, 200 ml of 1.8N NaOH were added at 800 ml of the supernatant and the absorbance was determined at 420 nm. One unit of enzyme activity was the amount of enzyme that caused a change of absorbance of 0.01 at 420 nm for 40 min at 45 8C. 2.6. Growth determination and protease production under optimal conditions Bacterial growth and protease activity were determinate under optimal conditions (25 8C, 12.5 g l1 of feather meal concentration and initial pH 7.0) on a rotary shaker at
69
Table 2 Effect of various nitrogen sources on protease production by Microbacterium arborescens after 48 h at 30 8C Nitrogen source (10 g l1)
Protease yield (U ml1)
Feather meal Soybean protein Gelatin Casein Yeast extract Cheese whey Peptone
96.5 3.8 73.8 4.7 45.8 7.7 44.0 9.7 24.4 15 21.5 3.8 18.1 6.8
125 cycles min1. At 24 h intervals, the bacterial suspension was diluted to 108 in 8.75 g l1 NaCl, samples were homogenized and then loaded (20 ml) in triplicate onto nutrient agar plates. Plates were incubated for 24 h at 37 8C and counts performed on plates having between 20 and 200 colonies. In parallel, protease production was determinate as described previously.
3. Results 3.1. Selection of growth substrate Among the various nitrogen sources studied, the Microbacterium sp. kr10 produced maximum protease in feather meal (96.5 U ml1), followed by soybean protein (73.8 U ml1) and gelatin (45.8 U ml1) as described in Table 2. Feather meal was selected for design of response surface methodology.
Table 3 Experimental design and results of the 23 factorial design Run number
x1 1 +1 1 +1 1 +1 1 +1 0 0 0 1.68 0 0 +1.68 0 0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 a b c *
Protease activity (U ml1)
Coded levels x2 1 1 +1 +1 1 1 +1 +1 0 0 0 0 1.68 0 0 +1.68 0
x3 1 1 1 1 +1 +1 +1 +1 0 0 0 0 0 1.68 0 0 +1.68
Observed
Predicted
Day 2
Day 4
Day 6
Day 2a
Day 4b
Day 6 c
74.2 51.7 12.5 22.5 186.7 55.8 36.7 25.0 20.0 30.8 16.7 11.7 179.2 47.5 11.7 7.5 13.3
49.2 70.8 10.8 27.5 106.7 49.2 66.7 23.3 13.3 4.2 9.2 15.8 270.8 62.5 11.7 11.7 10.0
45.0 93.3 23.3 103.3 152.5 111.7 10.0 10.0 0.0 4.2 1.7 5.8 90.0 25.0 10.8 0.0 10.0
91.1 63.0 0.0 * 41.3 158.7 65.5 16.2 0.0 * 21.7 21.7 21.7 37.3 168.8 26.3 0.0 * 30.9 47.5
88.2 112.8 2.8 27.4 123.0 78.0 37.5 0.0 * 7.5 7.5 7.5 16.0 202.7 26.0 0.0 * 59.2 26.0
33.4 71.6 13.7 88.2 132.9 86.6 0.0* 0.0* 0.0* 0.0* 0.0* 21.1 119.7 42.8 44.7 19.5 41.3
Model for day 2, with R2 = 0.8668 and Fc(9,7) = 5.06 > Ft(9,7) = 3.68. Model for day 4, with R2 = 0.7748 and Fc(5,11) = 7.57 > Ft(5,11) = 3.20. Model for day 6, with R2 = 0.8273 and Fc(9,7) = 3.72 > Ft(9,7) = 3.68. For these points, the models predicted negative values.
70
R.C.S. Thys et al. / Process Biochemistry 41 (2006) 67–73
Fig. 1. Bar graph of standardized estimated effects of the different variables tested in the prospective experiment on protease production by Microbacterium sp. The variables tested were temperature (T), initial pH (pH) and feather meal concentration (FM). The point at which the effect estimates were statistically significant (at P = 0.05) is indicated by the vertical line.
3.2. Optimization by response surface methodology
demonstrates significance for the regression model. The regression equation obtained indicated the R2 value of 0.7748 (a value of R2 > 0.75 indicates the aptness of the model). This value ensured a satisfactory adjustment of the quadratic model to the experimental data and indicated that 77.48% of the variability in the response could be explained by the model. The following regression equation was obtained (coefficients with less than 90% of significance were not included).
The enzyme activity values were tested after 2, 4 and 6 days of incubation. Results are shown in Table 3. The best values were obtained after 4 days, consequently the regression equation was made with these values. Treatment 13 showed the highest levels of protease activity (270.8 U ml1). Statistical analysis of results showed that, in the range studied, temperature has a strong effect on protease production. The interaction between pH and feather meal concentration have a significant effect and the quadratic effect of feather meal concentration and the linear effect of initial pH were in the limit of significance for 95% of significance (Fig. 1). The results of the second-order response surface model in the form of analysis of variance (ANOVA) are given in Table 4. Fischer F-test (F (5,11) = 7.57 > Ft(5,11) = 3,2)
Y ¼ 7:451181 5:09532x1 42:7196x2 þ 43:76328x22 þ 6:560895x23 17:3958x1 x3
(3)
with Y, protease production (response, in U ml1), x1, initial pH; x2, temperature and x3, feather meal concentration (coded values). The significance of each coefficient was determined by Student’s t-test and P-values, which are listed in Table 5. The larger the magnitude of the t-value and
Table 4 Analysis of variance (ANOVA) for the quadratic model Source of variations
Sum of squares
Degrees of freedom
Mean square
F-value
Regression Residual
51399.59 14939.38
5 11
10279.92 1358.126
7.57
Total
66338.97
16
Table 5 Model coefficients estimated by multiple linear regression (significance of regression coefficients) Factor
Coefficient
Computed t-value
P-value
Interception pH (linear) Temperature (linear) Temperature (quad) Feather meal (quad) pH (linear) by feather meal (linear)
7.451181 5.09532 42.7196 43.76328 6.560895 17.3958
3.7512 4.0994 34.3820 33.4533 5.0152 10.7204
0.064288 0.054671 0.000845 0.000892 0.037533 0.008589
R.C.S. Thys et al. / Process Biochemistry 41 (2006) 67–73
71
Fig. 2. Protease production (U ml1) observed as a response to the interaction of temperature (8C) and initial pH as variables and feather meal concentration at central point.
smaller the P-value, the more significant is the corresponding coefficent [6]. The response surface curves were then plotted (Figs. 2– 4). When protease production was observed as a response to the interaction of temperature and pH as variables and feather meal concentration at central point, it was observed that there was an enhancement in protease production at lower temperatures (Fig. 2). Maximum protease production was obtained at lower temperatures (25 8C) whereas the optimum growth temperature for this bacterium is around 30 8C (not shown). When production was tested at 18 8C (pH 7.0 and feather meal 12.5 g l1) protease production was very low (1.5 U ml1). The contour plot of interaction between temperature and
feather meal concentration (Fig. 3) confirmed that temperature has major influence on protease production. The contour plot showing the interaction between pH and feather meal concentration showed the activity changes are clearly visible upon changing the parameters toward their maximal (Fig. 4). 3.3. Growth and protease production at optimal conditions Microbacterium sp. kr10 was aerobically incubated at 25 8C in a rotary shaker. Cell growth reached the stationary phase after 3 days of cultivation and maximum protease activity was 270 U ml1, observed at day 4 (Fig. 5).
Fig. 3. Protease production (U ml1) observed as a response to the interaction of temperature (8C) and feather meal concentration (g l1) as variables and initial pH at central point.
72
R.C.S. Thys et al. / Process Biochemistry 41 (2006) 67–73
Fig. 4. Protease production (U ml1) observed as a response to the interaction of initial pH and feather meal concentration (g l1) as variables and temperature at central point.
4. Discussion The improving of microbial protease production is the aim of several investigations, being the production capacity of the organism depending on the successful selection of growth conditions and substrate [18,19]. The strain kr10 showed higher protease production on feather meal, which is an inexpensive and readily available substrate. Thus, the utilization of such substrate may result in a cost-effective process. The absence of carbohydrates in feather meal could be also beneficial by avoiding catabolite repression, which is often observed in extracellular protease production [18]. Optimization by a conventional ‘one-at-a-time-approach’ leads to a substantial increase in enzyme yields, however, this approach is not only awkward and time-consuming, but also has the limitations of ignoring the importance of
Fig. 5. Production of metalloprotease during growth of Microbacterium sp. in feather meal medium. pH 7.0 and 25 8C. CFU (*) and protease activity (~) were monitored. Each point represents the mean of two independent experiments.
interaction of various physicochemical parameters [6,12]. Response surface methodology used for optimization of metalloprotease production by Microbacterium sp. kr10 indicated a significant interaction between initial pH and feather meal concentration. When protease production was observed as a response to the interaction of these factors (temperature at central point), it was noted that at lower pH values the best values of feather meal concentration are the higher and, at high pH the best values of feather meal concentration are the lower. This suggests that production of protease by Microbacterium sp. kr10 is under complex control, although it is mainly influenced by temperature. Maximum protease production was obtained at lower temperatures than the optimum for growth, but production at 18 8C was very low. This suggests that temperatures around 25 8C are the best values and also the inferior limit for enzyme production. We have previously reported that for Chryseobacterium sp. strain kr6 maximum protease activity was at 25 8C whereas maximum cell growth was at 30 8C [20]. For Bacillus licheniformis PWD-1, Wang and Shih [21] reported that maximum keratinase activity was at 37 8C, a temperature much lower than the optimal for cell growth (50 8C). These authors also describe optimal temperature for protease production by the recombinant Bacillus subtilis FDB-29 is at lower temperature (37 8C) than the best for bacterial growth (42 8C) [21]. Our model indicate that temperature has a major effect on protease production by the strain kr10. Other RSM studies also showed that temperature is a very relevant factor for microbial protease production [22,23]. Response surface methodology has been used to optimization of protease production, particularly by Bacillus spp. [11,12]. This methodology was found to be very efficient to determine the optimal conditions for protease production by Bacillus sp. PE-11 [22,24]. For Microbacterium sp. kr10, an overall 3.6-fold increase in protease
R.C.S. Thys et al. / Process Biochemistry 41 (2006) 67–73
production was obtained compared to the observed before optimization (30 8C, pH 7.0 and 10 g l1 feather meal). This value is comparable with an increase of 2.6-fold in protease production obtained with Bacillus sp. RGR-14 [11], and a 4.2-fold increase observed for Bacillus mojavensis [12]. Response surface methodology proved to be a powerful tool in optimizing metalloprotease production by Microbacterium sp. Temperature had a strong influence in protease production and pH and feather meal concentration showed less significance. Maximum protease production was achieved at temperature 25 8C, pH 7.0 and feather meal concentration at 12.5 g l1.
[9]
[10]
[11]
[12]
[13]
Acknowledgements [14]
We thank Dr. S.H. Floˆres from Universidade Federal do Rio Grande do Sul for helpful in RSM analysis. R.C.S. Thys received a M.Sc. fellowship from CAPES. This work was supported by FAPERGS and CNPq.
[15]
[16]
References [17] [1] Rao MB, Tanksale AM, Ghatge MS, Deshpande VV. Molecular and biotechnological aspects of microbial proteases. Microbiol Mol Biol Rev 1998;62:597–635. [2] Anwar A, Saleemuddin M. Alkaline proteases: a review. Biores Technol 1998;64:175–83. [3] Lahl WJ, Braun SD. Enzymatic production of protein hydrolysates for food use. Food Technol 1994;48:68–71. [4] Adinarayana K, Ellaiah P, Srinivasulu B, Devi RB, Adinarayana G. Response surface methodological approach to optimize the nutritional parameters for neomycin production by Streptomyces marinensis under solid-state fermentation. Process Biochem 2003;38:1565–72. [5] Oh S, Rheem S, Sim J, Kim S, Baek Y. Optimizating conditions for the growth of Lactobacillus caseis YIT 9018 in tryptone–glucose medium by using response surface methodology. Appl Environ Microbiol 1995;61:3809–14. [6] Myers RH, Montgomery DC. Response surface methodology: process and product optimization using designed experiments. New York: Wiley, 2002. [7] Lee H, Song M, Hwang S. Optimizing bioconversion of deproteinated cheese whey to mycelia of Ganoderma lucidum. Process Biochem 2003;38:1685–93. [8] Bocchini DA, Alves-Prado HF, Baida LC, Roberto IC, Gomes E, Da Silva R. Optimization of xylanase production by Bacillus circulans D1
[18]
[19]
[20]
[21]
[22]
[23] [24]
73
in submerged fermentation using response surface methodology. Process Biochem 2002;38:727–31. Beg QK, Saxena RK, Gupta R. Kinetic constants determination for an alkaline protease from Bacillus mojavensis using response surface methodology. Biotechnol Bioeng 2002;78:289–95. Li C, Bai J, Cai Z, Ouyang F. Optimization of a cultural medium for bacteriocin production by Lactococcus lactis using response surface methodology. J Biotechnol 2001;93:27–34. Puri S, Beg QK, Gupta R. Optimization of alkaline protease production from Bacillus sp. by response surface methodology. Curr Microbiol 2002;44:286–90. Beg QK, Sahai V, Gupta R. Statistical media optimization and alkaline protease production from Bacillus mojavensis in bioreactor. Process Biochem 2003;39:203–9. Nehete PN, Shah VD, Kothari RM. Profiles of alkaline protease production as a function of composition of the slant, age, transfer and isolate number and physiological state of culture. Biotechnol Lett 1985;7:413–8. Kole MM, Draper I, Gerson DF. Production of protease by Bacillus subtilis using simultaneous control of glucose and ammonium concentrations. J Chem Technol Biotechnol 1988;41:196–206. Varela H, Ferrari MD, Belobradjic L, Weyrauch R, Loperena ML. Effect of medium composition on the production by a new Bacillus subtilis isolate of protease with promising unhearing activity. World J Microbiol Biotechnol 1996;12:643–5. Johnvesly B, Naik GR. Studies on production of thermostable alkaline protease from thermophilic and alkaliphilic Bacillus sp. JB-99 in a chemically defined medium. Process Biochem 2001;37:139–44. Riffel A, Ortolan S, Brandelli A. De-hairing activity of extracellular proteases produced by keratinolytic bacteria. J Chem Technol Biotechnol 2003;78:855–69. ¨ zdamar TH. Carbon sources affect metabolic capacities of C ¸ alik P, O Bacillus species for the production of industrial enzymes: theoretical analysis for serine and neutral proteases and a-amylase. Biochem Eng J 2001;8:61–81. Joo HS, Kumar CG, Park GC, Kim KT, Paik SR, Chang CS. Optimization of the production of an extracellular alkaline protease from Bacillus hirikoshii. Process Biochem 2002;38:155–9. Riffel A, Lucas FS, Heeb P, Brandelli A. Characterization of a new keratinolytic bacterium that completely degrades native feather keratin. Arch Microbiol 2003;179:258–65. Wang JJ, Shih JCH. Fermentation production of keratinase from Bacillus licheniformis PWD-1 and a recombinant B. subtilis FDB29. J Ind Microbiol Biotechnol 1999;22:608–16. Ellaiah P, Adinarayana K. Response surface optimization of the cultural conditions for the production of alkaline protease. J Sci Ind Res 2001;60:868–75. Wu LC, Hang YD. Acid protease production from Neosartorya fischeri. Lebenm Wiss Technol 2000;33:44–7. Adinarayana K, Ellaiah P. Response surface optimization of the critical medium components for the production of alkaline protease by a newly isolated Bacillus sp. J Pharm Pharm Sci 2002;5:281–7.
Optimization of protease production by Microbacterium sp. in feather meal using response surface methodology Roberta C.S. Thys, Samanta O. Guzzon, Florencia Cladera-Olivera, Adriano Brandelli * Laborato´rio de Bioquı´mica e Microbiologia Aplicada, Departamento de Cieˆncia de Alimentos, ICTA, Universidade Federal do Rio Grande do Sul, Av. Bento Gonc¸alves 9500, 91501-970 Porto Alegre, Brazil Received 29 March 2004; received in revised form 28 July 2004; accepted 15 March 2005
Abstract A 23 factorial design was performed with the aim of optimizing protease production by a strain of Microbacterium sp. isolated from feathers in decomposition by response surface methodology. Protease production was first tested on different nitrogen source (casein, peptone, yeast extract, gelatin, soybean protein, feather meal and cheese whey). Feather meal was the selected substrate to test the effect of three variables on protease production (temperature, initial pH and feather meal concentration) by RSM. The point was chosen with these conditions: temperature 37 8C, initial pH 7.0 and feather meal concentration 12.5 g l1. Statistical analysis of results showed that, in the range studied, only pH did not have a significant effect on protease production whereas interaction between pH and feather meal concentration was significant. The optimum conditions were 25 8C, initial pH 7.0 and 12.5 g l1 of feather meal. Under these conditions, the model predicted a protease activity of 202.7 U ml1. # 2005 Elsevier Ltd. All rights reserved. Keywords: Actinomycetes; Microbacterium; Protease; Experimental design; Optimization
1. Introduction Proteases constitute at least 65% of the total industrial enzyme market [1]. They are used for various industrial applications, such as laundry detergents, leather preparation, protein recovery or solubilization and organic synthesis [2]. In food industry, proteases have been routinely used for various purposes such as cheesemaking, baking, preparation of soy hydrolysates and meat tenderization [1]. Bacterial neutral proteases are active within a narrow pH range and have relatively low thermal tolerance. Due to their intermediate rate of reaction, neutral proteases generate less bitterness in hydrolyzed food proteins than do the animal proteinases and hence are valuable for use in the food industry [3]. Their low thermal tolerance is advantageous for controlling their reactivity during the production of food hydrolysates with a low degree of hydrolysis [1]. * Corresponding author. Tel.: +55 51 3316 6249; fax: +55 51 3316 7048. E-mail address: [email protected] (A. Brandelli). 1359-5113/$ – see front matter # 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.procbio.2005.03.070
Optimization of medium by the classical method involves changing one independent variable (i.e., nutrient, pH, temperature) while unchanging all others at a fixed level. This is extremely time-consuming and expensive for a large number of variables [4] and also may result in wrong conclusions [5]. Response surface methodology (RSM) is a collection of statistical techniques for designing experiments, building models, evaluating the effects of factors and searching optimum conditions of factors for desirable responses [6]. This method has been successfully applied in many areas of biotechnology such as bioconversion of cheese whey to mycelia of Ganoderma lucidum [7], optimization of neomycin production [4], enzyme production [8], enzyme kinetics [9] and bacteriocin production [10]. With respect to protease production, it was utilized for example for Bacillus species [11,12]. Extracellular protease production by microorganisms is greatly influenced by physical factors such as pH, temperature and incubation time and by others factors such as media composition and presence of metal ions [13–16].
68
R.C.S. Thys et al. / Process Biochemistry 41 (2006) 67–73
The feather-degrading Microbacterium sp. strain kr10 produces an extracellular keratinase belonging to the metalloprotease family. This enzyme was also proved to present desirable properties for de-hairing bovine pelts, which would be very beneficial to the leather industry [17]. The purpose of this study was to evaluate the effect of three variables (temperature, pH and feather meal concentration) in the production of this protease. Previously, enzyme production with different by-products as growth substrate was tested.
2. Materials and methods 2.1. Reagents and media Azocasein was from Sigma Chemical Co. (St. Louis, USA). Feather meal and soybean protein were from Bunge Alimentos S.A. (Brazil). Dried cheese whey powder was from Parmalat (Porto Alegre, Brazil). Nutrient agar and peptone were from Oxoid (Basingstone, UK) and yeast extract was from Biobras (Montes Claros, Brazil). Casein and gelatin were from Synth (Diadema, Brazil). 2.2. Bacterial strain and inoculum preparation Microbacterium sp. strain kr10 was used as the producer microorganism. BHI medium containing 20% (v/v) glycerol was used for maintenance of the strain at 20 8C. The cells were first propagated in milk agar plates (5 g l1 peptone, 3 g l1 yeast extract, 12.0 g l1 agar and 100 ml l1 UHT skimmed milk). The inoculum was prepared in feather meal previously hydrolyzed with 0.4N NaOH for 2 h, neutralized with HCl, filtered and than autoclaved (121 8C for 40 min). The inoculum absorbance at 600 nm was adjusted before use with a Hitachi U-1100 spectrophotometer (Hitachi, Japan) to obtain about 1 106 CFU ml1. 2.3. Selection of nitrogen source For selection of the best nitrogen source for protease production, various inorganic and complex nitrogen sources (casein, peptone, yeast extract, gelatin, soybean protein, feather meal and cheese whey) were tested individually (10 g l1) in the minimal mineral medium (0.5 g l1 NaCl, 0.3 g l1 K2HPO4, 0.4 g l1 KH2PO4, 0.015 g l1 CaCl22H2O). Erlenmeyer flasks of 200 ml containing 50 ml of medium were inoculated with 1% of inoculum (A600 nm at 0.350) and incubated at 30 8C under shaking (125 rpm). Protease yield was determined after 48 h. 2.4. Experimental design and protease production After selection of the best medium, the next stage was determination of the optimal levels of three variables, feather meal concentration, temperature and initial pH on
Table 1 Values of independent variables at different levels of the 23 factorial design Independent variables
Symbol
Initial pH Temperature (8C) Feather meal (g l1)
x1 x2 x3
Levels 1.68
1
0
+1
+1.68
5.0 25 0.0
5.8 30 5.0
7.0 37 12.5
8.29 45 20.0
9.0 50 25.0
protease production. For this purpose, the response surface approach by using a set of experimental design (central composite design with five coded levels) was performed. For the three factors, this design was made up of a full 23 factorial design with its eight points augmented with three replications of the centre points (all factors at level 0) and the six star points, that is, points having for one factor an axial distance to the centre of a, whereas the other two factors are at level 0. The axial distance a was chosen to be 1.68 to make this design orthogonal. A set of 17 experiments was carried out. The range and levels of experimental variables investigated are presented in Table 1. The central values (zero level) chosen for experimental design were: feather meal concentration 12.5 g l1, temperature 37 8C and initial pH 7.0. The enzyme activity values were tested after 2, 4 and 6 days of incubation but the regression equation was made with the values obtained after 4 days (maximal values). In developing the regression equation, the test factors were coded according to the following equation: xi ¼
X i X0 DXi
(1)
where xi is the coded value of the ith independent variable, Xi the natural value of the ith independent variable, X0 the natural value of the ith independent variable at the center point and DXi the step change value (DXi is 1.2 for initial pH, 8 for temperature and 7.5 for feather meal concentration). For a three factors system, the model equation is: Y ¼ b0 þ b1 x1 þ b2 x2 þ b3 x3 þ b11 x21 þ b22 x22 þ b33 x23 þ b12 x1 x2 þ b13 x1 x3 þ b23 x2 x3
(2)
where Y, predicted response, b0, intercept; b1, b2, b3, linear coefficients; b11, b22, b33, squared coefficients; b12, b13, b23, interaction coefficients. Results were analyzed by the Experimental Design Module of the Statistic 5.0 software (Statsoft, USA). The model permitted evaluation of the effects of linear, quadratic and interactive terms of the independent variables on the dependent variable. The statistical significance of the regression coefficients was determined by Student’s t-test and the second order model equation was determined by Fisher’s test. Three-dimensional surface plots were drawn to illustrate the main and interactive effects of the independent variables on protease production. The optimum values of the selected variables were obtained by solving the regression
R.C.S. Thys et al. / Process Biochemistry 41 (2006) 67–73
equation and also by analyzing the response surface contour plots [6]. The production medium (50 ml in 200 ml Erlenmeyer flask) was inoculated with 1% of inoculum (A600 nm at 0.350) and incubated under different temperatures for each set of experiment in a rotary shaker at 125 rpm. After incubation the culture broth was centrifuged at 10,000 g for 10 min and total protease yield was determined in the supernatant. 2.5. Enzyme assay Protease activity was assayed with azocasein as substrate by the following method. The reaction mixture contained 120 ml of cell-free supernatant and 480 ml of 10 g l1 azocasein in 100 mmol l1 Tris buffer pH 7.0 was incubated for 30 min at 45 8C (for the control, the reaction was stopped after the addition of the supernatant). The reaction was then stopped by the addition of tricoroacetic acid to a final concentration of 100 g l1 and incubated 30 min at 4 8C. After centrifugation at 10,000 g for 10 min, 200 ml of 1.8N NaOH were added at 800 ml of the supernatant and the absorbance was determined at 420 nm. One unit of enzyme activity was the amount of enzyme that caused a change of absorbance of 0.01 at 420 nm for 40 min at 45 8C. 2.6. Growth determination and protease production under optimal conditions Bacterial growth and protease activity were determinate under optimal conditions (25 8C, 12.5 g l1 of feather meal concentration and initial pH 7.0) on a rotary shaker at
69
Table 2 Effect of various nitrogen sources on protease production by Microbacterium arborescens after 48 h at 30 8C Nitrogen source (10 g l1)
Protease yield (U ml1)
Feather meal Soybean protein Gelatin Casein Yeast extract Cheese whey Peptone
96.5 3.8 73.8 4.7 45.8 7.7 44.0 9.7 24.4 15 21.5 3.8 18.1 6.8
125 cycles min1. At 24 h intervals, the bacterial suspension was diluted to 108 in 8.75 g l1 NaCl, samples were homogenized and then loaded (20 ml) in triplicate onto nutrient agar plates. Plates were incubated for 24 h at 37 8C and counts performed on plates having between 20 and 200 colonies. In parallel, protease production was determinate as described previously.
3. Results 3.1. Selection of growth substrate Among the various nitrogen sources studied, the Microbacterium sp. kr10 produced maximum protease in feather meal (96.5 U ml1), followed by soybean protein (73.8 U ml1) and gelatin (45.8 U ml1) as described in Table 2. Feather meal was selected for design of response surface methodology.
Table 3 Experimental design and results of the 23 factorial design Run number
x1 1 +1 1 +1 1 +1 1 +1 0 0 0 1.68 0 0 +1.68 0 0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 a b c *
Protease activity (U ml1)
Coded levels x2 1 1 +1 +1 1 1 +1 +1 0 0 0 0 1.68 0 0 +1.68 0
x3 1 1 1 1 +1 +1 +1 +1 0 0 0 0 0 1.68 0 0 +1.68
Observed
Predicted
Day 2
Day 4
Day 6
Day 2a
Day 4b
Day 6 c
74.2 51.7 12.5 22.5 186.7 55.8 36.7 25.0 20.0 30.8 16.7 11.7 179.2 47.5 11.7 7.5 13.3
49.2 70.8 10.8 27.5 106.7 49.2 66.7 23.3 13.3 4.2 9.2 15.8 270.8 62.5 11.7 11.7 10.0
45.0 93.3 23.3 103.3 152.5 111.7 10.0 10.0 0.0 4.2 1.7 5.8 90.0 25.0 10.8 0.0 10.0
91.1 63.0 0.0 * 41.3 158.7 65.5 16.2 0.0 * 21.7 21.7 21.7 37.3 168.8 26.3 0.0 * 30.9 47.5
88.2 112.8 2.8 27.4 123.0 78.0 37.5 0.0 * 7.5 7.5 7.5 16.0 202.7 26.0 0.0 * 59.2 26.0
33.4 71.6 13.7 88.2 132.9 86.6 0.0* 0.0* 0.0* 0.0* 0.0* 21.1 119.7 42.8 44.7 19.5 41.3
Model for day 2, with R2 = 0.8668 and Fc(9,7) = 5.06 > Ft(9,7) = 3.68. Model for day 4, with R2 = 0.7748 and Fc(5,11) = 7.57 > Ft(5,11) = 3.20. Model for day 6, with R2 = 0.8273 and Fc(9,7) = 3.72 > Ft(9,7) = 3.68. For these points, the models predicted negative values.
70
R.C.S. Thys et al. / Process Biochemistry 41 (2006) 67–73
Fig. 1. Bar graph of standardized estimated effects of the different variables tested in the prospective experiment on protease production by Microbacterium sp. The variables tested were temperature (T), initial pH (pH) and feather meal concentration (FM). The point at which the effect estimates were statistically significant (at P = 0.05) is indicated by the vertical line.
3.2. Optimization by response surface methodology
demonstrates significance for the regression model. The regression equation obtained indicated the R2 value of 0.7748 (a value of R2 > 0.75 indicates the aptness of the model). This value ensured a satisfactory adjustment of the quadratic model to the experimental data and indicated that 77.48% of the variability in the response could be explained by the model. The following regression equation was obtained (coefficients with less than 90% of significance were not included).
The enzyme activity values were tested after 2, 4 and 6 days of incubation. Results are shown in Table 3. The best values were obtained after 4 days, consequently the regression equation was made with these values. Treatment 13 showed the highest levels of protease activity (270.8 U ml1). Statistical analysis of results showed that, in the range studied, temperature has a strong effect on protease production. The interaction between pH and feather meal concentration have a significant effect and the quadratic effect of feather meal concentration and the linear effect of initial pH were in the limit of significance for 95% of significance (Fig. 1). The results of the second-order response surface model in the form of analysis of variance (ANOVA) are given in Table 4. Fischer F-test (F (5,11) = 7.57 > Ft(5,11) = 3,2)
Y ¼ 7:451181 5:09532x1 42:7196x2 þ 43:76328x22 þ 6:560895x23 17:3958x1 x3
(3)
with Y, protease production (response, in U ml1), x1, initial pH; x2, temperature and x3, feather meal concentration (coded values). The significance of each coefficient was determined by Student’s t-test and P-values, which are listed in Table 5. The larger the magnitude of the t-value and
Table 4 Analysis of variance (ANOVA) for the quadratic model Source of variations
Sum of squares
Degrees of freedom
Mean square
F-value
Regression Residual
51399.59 14939.38
5 11
10279.92 1358.126
7.57
Total
66338.97
16
Table 5 Model coefficients estimated by multiple linear regression (significance of regression coefficients) Factor
Coefficient
Computed t-value
P-value
Interception pH (linear) Temperature (linear) Temperature (quad) Feather meal (quad) pH (linear) by feather meal (linear)
7.451181 5.09532 42.7196 43.76328 6.560895 17.3958
3.7512 4.0994 34.3820 33.4533 5.0152 10.7204
0.064288 0.054671 0.000845 0.000892 0.037533 0.008589
R.C.S. Thys et al. / Process Biochemistry 41 (2006) 67–73
71
Fig. 2. Protease production (U ml1) observed as a response to the interaction of temperature (8C) and initial pH as variables and feather meal concentration at central point.
smaller the P-value, the more significant is the corresponding coefficent [6]. The response surface curves were then plotted (Figs. 2– 4). When protease production was observed as a response to the interaction of temperature and pH as variables and feather meal concentration at central point, it was observed that there was an enhancement in protease production at lower temperatures (Fig. 2). Maximum protease production was obtained at lower temperatures (25 8C) whereas the optimum growth temperature for this bacterium is around 30 8C (not shown). When production was tested at 18 8C (pH 7.0 and feather meal 12.5 g l1) protease production was very low (1.5 U ml1). The contour plot of interaction between temperature and
feather meal concentration (Fig. 3) confirmed that temperature has major influence on protease production. The contour plot showing the interaction between pH and feather meal concentration showed the activity changes are clearly visible upon changing the parameters toward their maximal (Fig. 4). 3.3. Growth and protease production at optimal conditions Microbacterium sp. kr10 was aerobically incubated at 25 8C in a rotary shaker. Cell growth reached the stationary phase after 3 days of cultivation and maximum protease activity was 270 U ml1, observed at day 4 (Fig. 5).
Fig. 3. Protease production (U ml1) observed as a response to the interaction of temperature (8C) and feather meal concentration (g l1) as variables and initial pH at central point.
72
R.C.S. Thys et al. / Process Biochemistry 41 (2006) 67–73
Fig. 4. Protease production (U ml1) observed as a response to the interaction of initial pH and feather meal concentration (g l1) as variables and temperature at central point.
4. Discussion The improving of microbial protease production is the aim of several investigations, being the production capacity of the organism depending on the successful selection of growth conditions and substrate [18,19]. The strain kr10 showed higher protease production on feather meal, which is an inexpensive and readily available substrate. Thus, the utilization of such substrate may result in a cost-effective process. The absence of carbohydrates in feather meal could be also beneficial by avoiding catabolite repression, which is often observed in extracellular protease production [18]. Optimization by a conventional ‘one-at-a-time-approach’ leads to a substantial increase in enzyme yields, however, this approach is not only awkward and time-consuming, but also has the limitations of ignoring the importance of
Fig. 5. Production of metalloprotease during growth of Microbacterium sp. in feather meal medium. pH 7.0 and 25 8C. CFU (*) and protease activity (~) were monitored. Each point represents the mean of two independent experiments.
interaction of various physicochemical parameters [6,12]. Response surface methodology used for optimization of metalloprotease production by Microbacterium sp. kr10 indicated a significant interaction between initial pH and feather meal concentration. When protease production was observed as a response to the interaction of these factors (temperature at central point), it was noted that at lower pH values the best values of feather meal concentration are the higher and, at high pH the best values of feather meal concentration are the lower. This suggests that production of protease by Microbacterium sp. kr10 is under complex control, although it is mainly influenced by temperature. Maximum protease production was obtained at lower temperatures than the optimum for growth, but production at 18 8C was very low. This suggests that temperatures around 25 8C are the best values and also the inferior limit for enzyme production. We have previously reported that for Chryseobacterium sp. strain kr6 maximum protease activity was at 25 8C whereas maximum cell growth was at 30 8C [20]. For Bacillus licheniformis PWD-1, Wang and Shih [21] reported that maximum keratinase activity was at 37 8C, a temperature much lower than the optimal for cell growth (50 8C). These authors also describe optimal temperature for protease production by the recombinant Bacillus subtilis FDB-29 is at lower temperature (37 8C) than the best for bacterial growth (42 8C) [21]. Our model indicate that temperature has a major effect on protease production by the strain kr10. Other RSM studies also showed that temperature is a very relevant factor for microbial protease production [22,23]. Response surface methodology has been used to optimization of protease production, particularly by Bacillus spp. [11,12]. This methodology was found to be very efficient to determine the optimal conditions for protease production by Bacillus sp. PE-11 [22,24]. For Microbacterium sp. kr10, an overall 3.6-fold increase in protease
R.C.S. Thys et al. / Process Biochemistry 41 (2006) 67–73
production was obtained compared to the observed before optimization (30 8C, pH 7.0 and 10 g l1 feather meal). This value is comparable with an increase of 2.6-fold in protease production obtained with Bacillus sp. RGR-14 [11], and a 4.2-fold increase observed for Bacillus mojavensis [12]. Response surface methodology proved to be a powerful tool in optimizing metalloprotease production by Microbacterium sp. Temperature had a strong influence in protease production and pH and feather meal concentration showed less significance. Maximum protease production was achieved at temperature 25 8C, pH 7.0 and feather meal concentration at 12.5 g l1.
[9]
[10]
[11]
[12]
[13]
Acknowledgements [14]
We thank Dr. S.H. Floˆres from Universidade Federal do Rio Grande do Sul for helpful in RSM analysis. R.C.S. Thys received a M.Sc. fellowship from CAPES. This work was supported by FAPERGS and CNPq.
[15]
[16]
References [17] [1] Rao MB, Tanksale AM, Ghatge MS, Deshpande VV. Molecular and biotechnological aspects of microbial proteases. Microbiol Mol Biol Rev 1998;62:597–635. [2] Anwar A, Saleemuddin M. Alkaline proteases: a review. Biores Technol 1998;64:175–83. [3] Lahl WJ, Braun SD. Enzymatic production of protein hydrolysates for food use. Food Technol 1994;48:68–71. [4] Adinarayana K, Ellaiah P, Srinivasulu B, Devi RB, Adinarayana G. Response surface methodological approach to optimize the nutritional parameters for neomycin production by Streptomyces marinensis under solid-state fermentation. Process Biochem 2003;38:1565–72. [5] Oh S, Rheem S, Sim J, Kim S, Baek Y. Optimizating conditions for the growth of Lactobacillus caseis YIT 9018 in tryptone–glucose medium by using response surface methodology. Appl Environ Microbiol 1995;61:3809–14. [6] Myers RH, Montgomery DC. Response surface methodology: process and product optimization using designed experiments. New York: Wiley, 2002. [7] Lee H, Song M, Hwang S. Optimizing bioconversion of deproteinated cheese whey to mycelia of Ganoderma lucidum. Process Biochem 2003;38:1685–93. [8] Bocchini DA, Alves-Prado HF, Baida LC, Roberto IC, Gomes E, Da Silva R. Optimization of xylanase production by Bacillus circulans D1
[18]
[19]
[20]
[21]
[22]
[23] [24]
73
in submerged fermentation using response surface methodology. Process Biochem 2002;38:727–31. Beg QK, Saxena RK, Gupta R. Kinetic constants determination for an alkaline protease from Bacillus mojavensis using response surface methodology. Biotechnol Bioeng 2002;78:289–95. Li C, Bai J, Cai Z, Ouyang F. Optimization of a cultural medium for bacteriocin production by Lactococcus lactis using response surface methodology. J Biotechnol 2001;93:27–34. Puri S, Beg QK, Gupta R. Optimization of alkaline protease production from Bacillus sp. by response surface methodology. Curr Microbiol 2002;44:286–90. Beg QK, Sahai V, Gupta R. Statistical media optimization and alkaline protease production from Bacillus mojavensis in bioreactor. Process Biochem 2003;39:203–9. Nehete PN, Shah VD, Kothari RM. Profiles of alkaline protease production as a function of composition of the slant, age, transfer and isolate number and physiological state of culture. Biotechnol Lett 1985;7:413–8. Kole MM, Draper I, Gerson DF. Production of protease by Bacillus subtilis using simultaneous control of glucose and ammonium concentrations. J Chem Technol Biotechnol 1988;41:196–206. Varela H, Ferrari MD, Belobradjic L, Weyrauch R, Loperena ML. Effect of medium composition on the production by a new Bacillus subtilis isolate of protease with promising unhearing activity. World J Microbiol Biotechnol 1996;12:643–5. Johnvesly B, Naik GR. Studies on production of thermostable alkaline protease from thermophilic and alkaliphilic Bacillus sp. JB-99 in a chemically defined medium. Process Biochem 2001;37:139–44. Riffel A, Ortolan S, Brandelli A. De-hairing activity of extracellular proteases produced by keratinolytic bacteria. J Chem Technol Biotechnol 2003;78:855–69. ¨ zdamar TH. Carbon sources affect metabolic capacities of C ¸ alik P, O Bacillus species for the production of industrial enzymes: theoretical analysis for serine and neutral proteases and a-amylase. Biochem Eng J 2001;8:61–81. Joo HS, Kumar CG, Park GC, Kim KT, Paik SR, Chang CS. Optimization of the production of an extracellular alkaline protease from Bacillus hirikoshii. Process Biochem 2002;38:155–9. Riffel A, Lucas FS, Heeb P, Brandelli A. Characterization of a new keratinolytic bacterium that completely degrades native feather keratin. Arch Microbiol 2003;179:258–65. Wang JJ, Shih JCH. Fermentation production of keratinase from Bacillus licheniformis PWD-1 and a recombinant B. subtilis FDB29. J Ind Microbiol Biotechnol 1999;22:608–16. Ellaiah P, Adinarayana K. Response surface optimization of the cultural conditions for the production of alkaline protease. J Sci Ind Res 2001;60:868–75. Wu LC, Hang YD. Acid protease production from Neosartorya fischeri. Lebenm Wiss Technol 2000;33:44–7. Adinarayana K, Ellaiah P. Response surface optimization of the critical medium components for the production of alkaline protease by a newly isolated Bacillus sp. J Pharm Pharm Sci 2002;5:281–7.