SciELO - Scientific Electronic Library Online

 
vol.18 issue3Study on the relationship between intracellular metabolites and astaxanthin accumulation during Phaffia rhodozyma fermentation author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand

Journal

Article

Indicators

Related links

  • On index processCited by Google
  • Have no similar articlesSimilars in SciELO
  • On index processSimilars in Google

Share


Electronic Journal of Biotechnology

On-line version ISSN 0717-3458

Electron. J. Biotechnol. vol.18 no.3 Valparaíso May 2015

http://dx.doi.org/10.1016/j.ejbt.2015.02.001 

RESEARCH ARTICLE

Statistical optimization for tannase production by Aspergillus tubingensis in solid-state fermentation using tea stalks

 

Anfeng Xiaoa, b, c, *, Yufeng Huanga, Hui Nia, b, c, Huinong Caia, b, c, Qiuming Yanga, b, c

a College of Food and Biological Engineering, Jimei University, Xiamen 361021, China
b Fujian Provincial Key Laboratory of Food Microbiology and Enzyme Engineering, Xiamen, Fujian Province 361021, China
c The Research Center of Food Biotechnology, Xiamen 361021, China


ABSTRACT

Background

A sequential statistical strategy was used to optimize tannase production from Aspergillus tubingensis using tea stalks by solid-state fermentation.

Results

First, using a Plackett-Burman design, inoculum size and incubation time (among seven tested variables) were identified as the most significant factors for tannase yield. The effects of significant variables were further evaluated through a single steepest ascent experiment and central composite design with response surface analysis. Under optimal conditions, the experimental value of 84.24 units per gram of dry substrate (U/gds) closely matched the predicted value of 87.26 U/gds.

Conclusions

The result of the statistical approach was 2.09 times higher than the basal medium (40.22 U/gds). The results were fitted onto a second-order polynomial model with a correlation coefficient (R2) of 0.9340, which implied an adequate credibility of the model.

Keywords: Aspergillus tubingensis; Process optimization; Response surface methodology; Tea stalks; Tannase fermentation


 

1. Introduction

Tannin acyl hydrolase, normally known as tannase (EC 3.1.1.20) [1], catalyzes the hydrolysis of tannic acid by specifically breaking its ester and depside bonds to release glucose and gallic acid [2]. Tannase is an extracellular inducible enzyme produced in the presence of tannic acid by fungi, bacteria, and yeasts [3] and [4]. Recently, interest in tannase production from agricultural residues has remarkably increased because of its low cost and wide application. Tannase has been proven to have important applications in various industries [5], particularly in the food and pharmaceutical sectors, including the elaboration of instantaneous tea [6], production of wine [7], coffee-flavored soft drinks, clarification of beer and fruit juices [8] and [9], as well as gallic acid [10]. Moreover, tannase can also be used in feed preparation to increase the bioavailability of nutrients by hydrolyzing phenolic anti-nutritional factors [11].

Given the immense potential of tannase, optimizing its production for maximum yield is a worthwhile endeavor. The optimization of fermentation parameters is an important issue in the development of economically feasible bioprocesses [12]. To the best of our knowledge, only a few reports focused on the optimization of tannase production using statistical methods. The use of the statistical approach has recently gained considerable attention for achieving medium optimization and for understanding the interactions among various physicochemical parameters using a minimum number of experiments [13]. Compared with conventional methods, the statistical approach saves more time and cost, thus making it highly practical. Thus, the use of appropriate statistical experiment design tools for tannase optimization is essential. The Plackett-Burman design (PBD) is widely used to determine the independent variables with significant effects on the response [14], whereas response surface methodology (RSM) has long been used to solve process optimization problems in the fields of chemical engineering and agro-biotechnology [15]. These methods can help to evaluate effective factors and simultaneously solve multivariate equations to study the interactions between variables, as well as to determine the optimum combination of variables for a desirable response.

In this work, statistical tools were used to optimize tannase production from Aspergillus tubingensis CICC 2651 using tea stalks as an economical solid substrate. The effects of medium components and culture conditions were investigated by combining different statistical experimental designs.

2. Materials and methods

2.1.Microorganism

A. tubingensis CICC 2651, obtained from the China Center of Industrial Collection, was used to produce tannase. The culture was maintained as spores on potato dextrose agar (PDA) slants at 4°C and subcultured every 30 d.

2.2. Inoculum preparation

Before inoculation, A. tubingensis CICC 2651 was cultivated on PDA slants at 30°C for 4 d. Then, the spores were scraped from the surface of slants with sterile distilled water to prepare a spore suspension (1 × 108 spores/mL). Thereafter, the spore solution was cultivated synchronously at 4°C for 1 h.

2.3. Production of tannase under solid-state fermentation

Tea stalk powder (5 g) was placed in 250-mL Erlenmeyer flasks and moistened with a mineral salt solution (0.1% MgSO4·7H2O, 0.1% NaCl, and 0.1% KH2PO4) at an initial pH of 6.0, and then autoclaved at 121°C for 20 min. After cooling, the samples were inoculated with some spore inoculum (1 × 108 spores/mL). The contents were mixed thoroughly and incubated under different temperatures for a period of time. Different nutritional and growth conditions were adjusted according to the design suggested by the statistical model.

2.4. Sampling and extraction

The fermented mass was mixed with citrate buffer (0.05 N, pH 5.0) to a volume of 100 mL for each flask and agitated at 180 rpm for 1 h. The slurry was squeezed through a cheesecloth, followed by centrifugation at 10,000 × g for 10 min at 4°C. The supernatant was collected in vials and stored at 4°C for further analysis. The substrate was dried at 80°C to measure biomass.

2.5. Enzyme assay

Tannase activity was determined using the colorimetric method described by Sharma et al. [16]. One unit of enzyme activity was defined as the amount of enzyme that catalyzed the release of 1 µmol of gallic acid per minute under specific conditions, expressed as U/gds (units per gram of dry substrate).

2.6. Determination of fungal biomass

The biomass in the fermented substrate was determined by estimating the glucosamine present in the cell wall of the culture, and this was expressed in milligrams per gram of dry substrate (mg/gds). A known quantity of fermented matter, ground to powder, was placed in a test tube and hydrolyzed with concentrated sulfuric acid (60%, w/v) for 24 h. The mixture was then diluted and autoclaved at 121°C for 1 h. After cooling, the mixture was percolated, and the filtrate was neutralized to a pH of 7 by adding 1 N NaOH. The glucosamine content was estimated using the method of Sakurai et al. [17].

2.7. Screening of significant factor using PBD

PBD [14] was employed to evaluate the relative importance of various factors relative to tannase yield. In this work, seven physical and nutritional components (independent variables) were studied, along with four “dummy variables” generated by the software. These seven factors and the selection of the range were set according to the result of our pilot study [10]. Each variable was tested at two levels, high (+ 1) and low (-1). In PBD, a total of 12 experiments were generated and enzyme activities were measured, which are shown in Table 1. From the regression analysis, variables having a P-value of < 0.05 were considered to have a significant effect on tannase production.

 

Table 1. Experimental design and corresponding results of PBD.

Run Variables Tannase yield (U/gds) Biomass (mg/gds)
A B C D E F G
1
2
5
50
8
2.5
48
30
29.55 ± 1.36
24.22 ± 1.82
2
5
2
75
8
2.5
48
20
34.72 ± 2.27
20.12 ± 2.61
3
2
5
75
5
2.5
120
30
68.35 ± 2.56
26.43 ± 3.21
4
5
5
75
5
1
48
30
26.45 ± 2.89
35.36 ± 4.05
5
2
2
50
5
1
48
20
2.00 ± 0.24
7.14 ± 1.09
6
5
5
50
8
2.5
120
20
55.69 ± 0.44
24.63 ± 1.36
7
2
2
50
8
1
120
30
52.22 ± 4.20
23.21 ± 3.82
8
5
2
50
5
2.5
48
30
31.32 ± 2.95
22.38 ± 1.59
9
2
2
75
5
2.5
120
20
73.49 ± 2.10
33.24 ± 1.97
10
2
5
75
8
1
48
20
10.59 ± 2.39
15.56 ± 1.26
11
5
5
50
5
1
120
20
56.20 ± 3.84
23.57 ± 5.64
12
5
2
75
8
1
120
30
69.63 ± 0.42
40.75 ± 2.96

A: NH4Cl, % (w/w); B: α-lactose, % (w/w); C: humidity, % (v/w); D: initial pH; E: inoculum size, mL; F: incubation time, h; G: temperature, °C.

 

2.8. Steepest ascent method

Generally, the initial estimate of the optimal operating conditions for the system will be far from the actual optimum. In such circumstances, the levels of significant variables are optimized with respect to the responses of a single steepest ascent experiment.

2.9. Response surface methodology

The variables with significant effects on tannase production, according to the results of PBD and steepest ascent method (SAM), were further evaluated using a central composite design (CCD). The variables were studied at five different levels, namely, -1.414 (lowest), -1 (low), 0 (middle), + 1 (high), and + 1.414 (highest). The full experimental plan is given in Table 3. The quadratic model for predicting the optimal point is expressed according to the following equation:

[Equation 1]

where Y is the predicted response, xi is the coded independent variable, β0 is the intercept, βi is the linear coefficient, βii is the quadratic coefficient, and βij is the interactive coefficient.

2.10. Data analysis

All experiments were conducted in duplicate and all values were averaged. Data were analyzed using the Microsoft Excel program and Design-Expert 8.04 statistical software package to obtain the standard deviations.

3. Results and discussion

During the course of our earlier studies, the maximum production of tannase by A. tubingensis CICC 265 was observed for 120 h, and the value only reached 40.22 U/gds using an initial TS medium. Thus, statistical optimization of the fermentation parameters for maximal enzyme production was conducted in three stages by employing PBD and SAM, followed by RSM with CCD.

3.1. Evaluation of significant variables by PBD

A 12-run Plackett-Burman experiment (Table 1) was conducted to determine the significant factors affecting enzyme yield. As shown in Table 1, wide variation was observed in tannase activities from 2.0 to 73.49 U/gds, which demonstrated the importance of the optimization of fermentation parameters to achieve higher production. Furthermore, the results were fitted onto the following equation:

[Equation 2]

where Y is the response of tannase activity and A, B, C, D, E, F, and G represent NH4Cl, lactose, humidity, pH, inoculum size, incubation time, and temperature, respectively (Table 1).

The adequacy of the model was tested and the variables that were determined to have a statistically significant effect were screened via Fisher's test for analysis of variance (ANOVA). ANOVA showed that the model is significant given its F-value of 17.29 and P-value of 0.0076. We only found a 0.76% chance that a “Model F-value” could occur because of noise. Generally, factors evidencing P-values < 0.05 were desirable. The lowest P-value indicates the most significant factor on enzyme production. Furthermore, P-values of E (0.0347) and F (0.0006), as well as their regression coefficients (6.34 and 20.08, respectively), implied that inoculum size and incubation time had a significant positive effect on tannase production. Meanwhile, such factors as NH4Cl, lactose, moisture level, pH, and temperature, which had P-values > 0.05, had no significant effect on tannase activities. Therefore, all other insignificant variables were neglected and only the two significant variables were selected for further optimization using SAM and RSM.

3.2. Determination of proper regions by SAM

The second optimization step was a single steepest ascent experiment to identify the proper regions of the variables by increasing inoculum size and extending culture time regularly. Table 2 lists the directions of changes for the two variables, while the other insignificant variables remained in the 0 level, that is, NH4Cl was 2.5%, α-lactose was 2.5%, humidity was 62.5%, initial pH was 6.5, and temperature was 25°C. The tannase activities were observed to reach the peak on the third group. Therefore, this condition was selected for further research.

 

Table 2. Experimental design and corresponding results of SAM.

Run Inoculum size (mL) Incubation time (h) Tannase yield (U/gds) Biomass (mg/gds)
1
1.75
84
59.31 ± 3.21
26.09 ± 3.54
2
2.55
96
70.01 ± 2.82
31.49 ± 2.08
3
3.35
108
88.29 ± 4.14
34.58 ± 2.59
4
4.15
120
80.19 ± 3.78
28.79 ± 0.46
5
4.95
132
70.93 ± 2.38
27.72 ± 1.17
6
5.75
144
62.64 ± 2.71
27.63 ± 2.34

 

3.3. Optimization of significant factors by RSM

Based on the results of PBD and SAM, crucial factors and their interaction were analyzed and optimized by RSM using a CCD. A total of 13 experiments were performed in duplicate; their corresponding results are shown in Table 3. In all 13 runs, tannase activity varied from 69.83 to 89.02 U/gds. Further, the second-order polynomial [Equation 3] was used to explain the tannase production and interaction among the variables (in coded form) as follows:

[Equation 3]

where Y presents the tannase activity and X1 and X2 are the inoculum size and incubation time, respectively.

 

Table 3. Experimental design and corresponding responses of CCD.

Run Inoculum size (X1) Incubation time (X2) Tannase yield (U/gds) Biomass (mg/gds)
Actual Predicted
1
-1 (2.35 mL)
-1 (96 h)
69.83 ± 2.90
70.94
31.63 ± 2.11
2
1 (4.35 mL)
-1
75.60 ± 2.20
74.79
30.02 ± 2.51
3
-1
1 (120 h)
73.80 ± 3.09
73.87
32.23 ± 1.82
4
1
1
89.02 ± 2.57
87.16
44.11 ± 2.48
5
-1.414 (1.94 mL)
0 (108 h)
72.27 ± 2.16
71.27
33.66 ± 3.80
6
1.414 (4.76 mL)
0
81.67 ± 3.86
83.39
40.56 ± 1.82
7
0 (3.35 mL)
-1.414 (91 h)
71.02 ± 2.74
70.64
39.60 ± 3.82
8
0
1.414 (125 h)
80.35 ± 3.39
81.46
41.99 ± 4.50
9
0
0
83.88 ± 3.02
83.42
34.58 ± 2.92
10
0
0
80.95 ± 4.97
83.42
35.64 ± 2.82
11
0
0
84.05 ± 2.56
83.42
42.82 ± 4.15
12
0
0
81.69 ± 3.86
83.42
39.27 ± 1.99
13
0
0
86.55 ± 5.11
83.42
38.49 ± 3.39
ANOVA of CCD
Source Model X1X1 X2 X1X2 X12 X22 Residual Lack of fit Pure error Cor total
Sum of squares
427.24
146.90
116.99
22.27
64.53
94.53
30.18
10.67
19.50
457.41
DF
5
1
1
1
1
1
7
3
4
12
Mean square
85.45
146.90
116.99
22.27
64.53
94.53
4.31
3.56
4.88
-
F-value
19.82
34.08
27.14
5.17
14.97
21.93
-
0.73
-
-
P-value
0.0005
0.0006
0.0012
0.0573
0.0061
0.0023
-
0.5857
-
-

 

The ANOVA of the response surface model demonstrated that [Equation 3] was highly significant, as evident from the model F-value of 19.82 and P-value of 0.0005 (P < 0.01). We only observed a 0.05% chance that a “Model F-Value” could occur because of noise. Values of P (Prob > F) < 0.0500 indicate that model terms are significant. In this case, A (0.0006), B (0.0012), A2 (0.0061), and B2 (0.0023) were significant model terms.

The R2 value for the model was 0.9340, which indicated that 93.4% of the observed variations in the tannase yield could be explained by the model. This result also suggested that the prediction of experimental data was satisfactory. The lack-of-fit F-value of 0.73 implied that the lack of fit was not significant relative to the pure error. The “Pred R2” value of 0.7674 was in reasonable agreement with the “Adj R2” value of 0.8869. “Adeq Precision” measures the signal to noise ratio; a value > 4 is generally desirable. In the present study, the ratio of 11.710 indicated that an adequate signal was detected.

3.4. Effect of significant factors on tannase production

To investigate the interaction and to visualize the effects of the two factors on tannase activity better, a graphical representation of RSM (a) and contour plots (b) are presented in Fig. 1. The circular shape of the curve indicates that no interaction occurred, whereas the elliptical shape indicated good variation of the two variables [18]. The response surface obtained in this study was convex in nature, which suggested that the optimum conditions were well defined. Meanwhile, the shapes of the contour plots indicated the nature and extent of the interactions. Fig. 1a indicated that the interaction between the inoculum size and incubation time was not significant. Furthermore, increasing the inoculum size and prolonging the incubation time have a positive effect on the response until a certain optimum value, whereas extremely high values of these variables will inhibit tannase production (Fig. 1b). Similar results were reported by Sabu et al. [19], when Aspergillus niger ATCC 16620 was used to ferment palm kernel cake (Fig. 1).

Fig. 1. (a) Response surface curves and (b) contour plots of interaction effect between inoculum size and incubation time on tannase activity.

 

3.5. Experimental validation of the model

The response surface model predicted a maximum tannase content of 87.26 U/gds when the regression equation was solved by the numerical optimization function in the Design-Expert software. The optimum levels of significant variables were an inoculum size of 4.35 mL after 118 h of incubation. To confirm the predicted model, four replicate experiments were performed and the tannase activities of each were determined. Finally, the average maximum yield reached 84.24 U/gds, which was comparable to the predicted value (87.26 U/gds) and was 1.09-fold higher than that of the basal medium (40.22 U/gds). Consequently, the proposed model was considered to be accurate and reliable for predicting the production of tannase from A. tubingensis.

4. Conclusions

Tannase production from A. tubingensis with tea stalks as substrates was optimized through a series of statistical methodologies. Among the variables studied, the inoculum size and incubation time were found to affect tannase activity significantly. The optimal conditions of variables for maximum tannase production were an inoculum size of 4.35 mL and 118 h of incubation. The actual tannase activity of 84.24 U/gds was consistent with the predicted value of 87.26 U/gds. The optimization process doubled the tannase yield (from 40.22 U/gds to 84.24 U/gds) as compared with the basal medium.

Conflict of interest

None.

Financial support

This study was supported by Chinese Scientific and Technical Services for Enterprise Projects (2009GJC40050) and Key Projects of Science and Technology Program of Fujian Province (2011N1008).

References

1. Kumar R, Sharma J, Singh R. Production of tannase from Aspergillus ruber under solid-state fermentation using jamun (Syzygium cumini) leaves. Microbiol Res 2007;162:384-90. http://dx.doi.org/10.1016/j.micres.2006.06.012.         [ Links ]

2. Beena PS, Basheer SM, Bhat SG, Bahkali AH, Chandrasekaran M. Propyl gallate synthesis using acidophilic tannase and simultaneous production of tannase and gallic acid by marine Aspergillus awamori BTMFW032. Appl Biochem Biotechnol 2011;164:612-28. http://dx.doi.org/10.1007/s12010-011-9162-x.         [ Links ]

3. Sharma KP, John PJ. Purification and characterization of tannase and tannase gene from Enterobacter sp. Process Biochem 2011;46:240-4. . http://dx.doi.org/10.1016/j.procbio.2010.08.016.         [ Links ]

4. Böer E, Bode R, Mock HP, Piontek M, Kunze G. Atan1p-An extracellular tannase from the dimorphic yeast Arxula adeninivorans: Molecular cloning of the ATAN1 gene and characterization of the recombinant enzyme. Yeast 2009;26:323-37. http://dx.doi.org/10.1002/yea.1669.         [ Links ]

5. Chávez-González M, Rodríguez-Durán LV, Balagurusamy N, Prado-Barragán A, Rodríguez R, Contreras JC, et al. Biotechnological advances and challenges of tannase: An overview. Food Bioprocess Technol 2012;5:445-59. http://dx.doi.org/10.1007/s11947-011-0608-5.         [ Links ]

6. Aguilar CN, Rodriguez R, Gutierrez-Sanchez G, Augur C, Favela-Torres E, Prado-Barragan LA, et al. Microbial tannases: Advances and perspectives. Appl Microbiol Biotechnol 2007;76:47-59. http://dx.doi.org/10.1007/s00253-007-1000-2.         [ Links ]

7. Selwal MK, Yadav A, Selwal KK, Aggarwal NK, Gupta R, Gautam SK. Tannase production by Penicillium atramentosum KM under SSF and its applications in wine clarification and tea cream solubilization. Braz J Microbiol 2011;42:374-87. http://dx.doi.org/10.1590/S1517-83822011000100047.         [ Links ]

8. Mohapatra PKD, Mondal KC, Pati BR. Production of tannase by the immobilized cells of Bacillus licheniformis KBR6 in Ca-alginate beads. J Appl Microbiol 2007;102: 1462-7. http://dx.doi.org/10.1111/j.1365-2672.2006.03207.x.         [ Links ]

9. Lekha PK, Lonsane BK. Production and application of tannin acyl hydrolase: State of the art. Adv Appl Microbiol 1997;44:215-60. . http://dx.doi.org/10.1016/S0065-2164(08)70463-5.         [ Links ]

10. Wang F, Ni H, Cai H, Xiao A. Tea stalks-A novel agro-residue for the production of tannase under solid state fermentation by Aspergillus niger JMU-TS528. Ann Microbiol 2013;63:897-904. http://dx.doi.org/10.1007/s13213-012-0541-5.         [ Links ]

11. Graminha EBN, Goncalves AZL, Pirota RDPB, BalsalobreMAA, Da Silva R, Gomes E. Enzyme production by solid-state fermentation: Application to animal nutrition. Anim Feed Sci Technol 2008;144:1-22. http://dx.doi.org/10.1016/j.anifeedsci.2007.09.029.         [ Links ]

12. Kar S, Swain MR, Ray RC. Statistical optimization of alpha-amylase production with immobilized cells of Streptomyces erumpens MTCC 7317 in Luffa cylindrica L. sponge discs. Appl Biochem Biotechnol 2009;152:177-88. http://dx.doi.org/10.1007/s12010-008-8248-6.         [ Links ]

13. Hajji M, Rebai A, Gharsallah N, Nasri M. Optimization of alkaline protease production by Aspergillus clavatus ES1 in Mirabilis jalapa tuber powder using statistical experimental design. Appl Microbiol Biotechnol 2008;79:915-23. http://dx.doi.org/10.1007/s00253-008-1508-0.         [ Links ]

14. Plackett RL, Burman JP. The design of optimum multifactorial experiments. Biometrika 1946;33:305-25. http://dx.doi.org/10.1093/biomet/33.4.305.         [ Links ]

15. Tu YY, Xu XQ, Xia HL,Watanabe N. Optimization of the aflavin biosynthesis from tea polyphenols using an immobilized enzyme system and response surfacemethodology. Biotechnol Lett 2005;27:269-74. http://dx.doi.org/10.1007/s10529-004-8292-4.         [ Links ]

16. Sharma S, Bhat TK,Dawra RK. A spectrophotometric method for assay of tannase using rhodanine. Anal Biochem 2000;279:85-9. http://dx.doi.org/10.1006/abio.1999.4405.         [ Links ]

17. Sakurai Y, Lee TH, ShiotaH. On the convenientmethod for the glucosamine estimation in koji. Agric Biol Chem 1977;41:619-24. http://dx.doi.org/10.1271/bbb1961.41.619.         [ Links ]

18. Garg G, Mahajan R, Kaur A, Sharma J. Xylanase production using agro-residue in solidstate fermentation fromBacillus pumilus ASHfor biodelignification ofwheat strawpulp. Biodegradation 2011;22:1143-54. http://dx.doi.org/10.1007/s10532-011-9470-4.         [ Links ]

19. Sabu A, Augur C, Swati C, Pandey A. Tannase production by Lactobacillus sp. ASR-S1 under solid-state fermentation. Process Biochem 2006;41:575-80. http://dx.doi.org/10.1016/j.procbio.2005.05.011.         [ Links ]


*Corresponding author: E-mail address: xxaaffeng@jmu.edu.cn (A. Xiao).

Received 1 November 2014, Accepted 14 January 2015, Available online 18 March 2015

Copyright © 2015 Pontificia Universidad Católica de Valparaíso. Production and hosting by Elsevier B.V. All rights reserved. Peer review under responsibility of Pontificia Universidad Católica de Valparaíso.

 

Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License