Biotechnological

Communication
Biosci. Biotech. Res. Comm. 11(4): 691-698 (2018)

Statistical optimization of extracellular tannase

production by Streptomyces sp. AT 13 using response
surface methodology and Plackett-Burmen design
Archana D. Tripathi* and Lakshmi B.
Department of Biotechnology, Kadi Sarva Vishwavidyalaya, Gandhinagar, Pin-382028, Gujarat, India

ABSTRACT
Tannase has many important applications in animal feed, chemical, food, and pharmaceutical industries. In the present study,
optimization of tannase production by Streptomyces sp. AT13 was carried out using statistical experimental designs. Initially, a
Plackett-Burmen design (PBD) was employed to screen the preferable nutriments (carbon and nitrogen sources of the medium) to
produce tannase. The result showed that only tannic acid was found to be significant for the production of tannase by Strepto-
myces sp. AT 13. The significant factor was further optimized by using Box-Behnken design under response surface methodology
(RSM). From among 6 fermentative variables that were studied, 5 significant variables were picked up by PBD. Among 5 variables
from PBD, 3 were further optimized by Box-Behnken design. The parameters studied through RSM were 1% Tannic Acid, 0.5 %
KCl and 0.1 % KH2PO4. Under optimized conditions tannase activity was 18.12 U/ml/min. This activity was almost three times
higher as compared to the amount obtained by ‘one-at-a-time’ approach. (5.19 U/ml/min)

INTRODUCTION Based on their structure and properties, tannins are dis-

tributed into two major groups – hydrolysable and con-
Tannins are polyphenolic compounds with molecular densed tannins. Hydrolysable tannins are composed of
weights ranging from 500 to 3000 daltons which occur esters of gallic acid (gallotannins) or ellagic acid (ella-
widely ina variety of plants including monocots, dicots gitannins) with a sugar core, which is usually glucose,
and ferns (Bate-Smith and Swain 1962; Mcleod 1974; and are readily hydrolysed by acids and enzymes into
Haslam 1989). They are the fourth most abundant plant monomeric products. Hydrolysable tannins are notably
constituents after cellulose, hemicellulose and lignin. absent in monocots. Commercially, hydrolysable tannins

ARTICLE INFORMATION:
Corresponding Authors: archana315@[Link],
lakjayan@[Link]
Received 19th Sep, 2018
Accepted after revision 18st Dec, 2018
BBRC Print ISSN: 0974-6455
Online ISSN: 2321-4007 CODEN: USA BBRCBA
Thomson Reuters ISI ESC / Clarivate Analytics USA
Mono of Clarivate Analytics and Crossref Indexed
Journal Mono of CR
NAAS Journal Score 2018: 4.31 SJIF 2017: 4.196
© A Society of Science and Nature Publication, Bhopal India

691
2018. All rights reserved.
Online Contents Available at: http//[Link]/
DOI: 10.21786/bbrc/11.4/21
Archana and Lakshmi

are extracted from Chinese gall (Rhus semialata), Sumac grade and purchased from Hi-Media, India. Sterptomy-
(R. coriara), Turkish gall (Quercus infectoria), Tara (Cae- ces sp. AT 13 used for the present study was screened
salpina spinosa), Myrobalan nuts (Terminalia chebula) and isolated from Tannery Effluent at G.I.D.C Naroda,
and chestnuts (Castania sativa) (Bhat et al., 1998). Ahmedabad and the organism was maintained on tannic
Tannase (tannin acyl hydrolase) catalyses the hydroly- acid agar (TAA) medium. Tannase production by Sterp-
sis of ester and depside bonds of hydrolysable tannins to tomyces sp. AT 13 was carried out using tannase pro-
produce glucose and gallic acid (Barthomeuf et al.,1994). duction medium containing NaNO3 0.6 % w/v, KCl 0.5
Hydrolysis of tannic acid by tannase results in the liber- % w/v, MgSO4 0.05 % w/v, K2HPO4 0.05 % w/v, KH2PO4
ation of glucose, gallic acid and various galloyl esters of 0.05 % w/v and 1 % (w/v) filtered sterilized tannic acid
glucose (Van de Lagemaat and Pyle, 2006). This enzyme (designated as TA broth). The broth was assayed for tan-
finds widespread applications in food processing, brew- nase activity.
ing, pharmaceuticals, medicine, textiles, detergents and
tea industry (Lekha and Lonsane, 1997). A combination TANNASE ASSAY
of trimethoprim and sulphonamide is effective against
many antibiotic resistant species of bacteria. More than The tannase activity was estimated by modified spectro-
8000 tonnes per annum of gallic acid is manufactured photometric method of Sharma et al., 2000. Tannic Acid
out of which, almost 70% is used in production of tri- was used as substrate. The basic principle of this assay
methoprim (Beena, 2010). is based on the formation of chromogen between gallic
The use of a sequential experimental design strategy acid (released by the action of tannase on tannic acid)
is a useful tool for process optimization. Response sur- and rhodanine (2-thio-4-ketothiazolidine). A standard
face methodology (RSM) provides important information curve was prepared using gallic [Link] enzyme was
regarding the optimum level of each variable along with used for the assay. All the tests were performed in trip-
its interactions with other variables and their effects on licates. One unit of tannase activity is defined as the
product yield. It reduces the number of experiments with- amount of enzyme required to liberate 1μM of gallic
out neglecting the interactions among the parameters. acid /min under defined conditions. Enzyme activity
This multivariate approach also improves statistical inter- was expressed as U/ml/min.
pretation possibilities and evaluates the relative signifi- Plackett–Burman design
cance of several contributing factors even in the presence
of complex interactions (Dilipkumar et al., 2011). RSM is Plackett–Burman design, an efficient technique for
widely used for multivariable optimization studies in sev- medium component selection (Plackett-Burman, 1946)
eral biotechnological processes such as the optimization of was employed to establish the factors that significantly
media, process conditions etc. (Mannan et al., 2007; Pan enhance the tannase production. Five independent vari-
et al., 2008). Statistical optimization allows the interaction ables (Table 1) were analysed in 12 experimental runs
amid possible influencing parameters to be evaluated with (Table 2) for the production of tannase. Triplicates were
a limited number of experiments (Rodriguez et al., 2008). maintained for each experimental set up. The response
Statistical designs such as Plackett-Burman and Response of these factors for the production of tannase was meas-
surface methodology are common in practice for testing ured by Spectrophotometer method as proposed by
multiple factors and interactions that can be quantified in Sharma et al., (2000). The low level (-1) and the high
an error-free and robust manner (Mohan, 2014). level (+1) of each factors are demonstrated in Table 2.
In order to fully exploit the prospective of this The statistical software design expert 6.0 was used for
enzyme for various applications, it is imperative to analysing the experiment. Total 5 independent variables
investigate the possibility of enhancing its production
by using more efficient production strategies (Rana and Table 1. Screening of variables using a Plackett–
Bhat, 2005). Hence in this study the process parameters Burman design
were optimized for maximum enzyme production adopt- Variable Variables Low High
ing Plackett-Burman (PB) and Response Surface Meth- codes level (-1) level (+1)
odology (RSM) with Box Behnken design. The statisti- A Tannic Acid -1 +1
cal software package DesignExpert® 7.0 (Stat Ease Inc.,
Minneapolis, and USA) was used. B KCl -1 +1

C NaNO3 -1 +1
MATERIALS AND METHODS D KH2PO4 -1 +1
Tannic Acid was purchased from Sigma Aldrich, USA. E K2HPO4 -1 +1
All other chemicals and reagents were used of analytical

692 STATISTICAL OPTIMIZATION OF EXTRACELLULAR TANNASE PRODUCTION BIOSCIENCE BIOTECHNOLOGY RESEARCH COMMUNICATIONS
 Archana and Lakshmi

Table 2. Plackett Burman design matrix for the screening of variables

F1 F2 F3 F4 F5 F6 F7 F8 F9 F 10 F11
A:Tannic
Run B:KCl C:NaNO3 D:KH2PO4 E:K2HPO4 F:F G:G H:H J:J K:K L:L R1
Acid
1 -1.00 1.00 1.00 1.00 -1.00 -1.00 -1.00 1.00 -1.00 1.00 1.00 1.62
2 1.00 1.00 1.00 -1.00 -1.00 -1.00 1.00 -1.00 1.00 1.00 -1.00 12.99
3 1.00 -1.00 1.00 1.00 -1.00 1.00 1.00 1.00 -1.00 -1.00 -1.00 7.17
4 -1.00 -1.00 -1.00 -1.00 -1.00 -1.00 -1.00 -1.00 -1.00 -1.00 -1.00 10.12
5 1.00 1.00 -1.00 1.00 1.00 1.00 -1.00 -1.00 -1.00 1.00 -1.00 16.53
6 -1.00 1.00 -1.00 1.00 1.00 -1.00 1.00 1.00 1.00 -1.00 -1.00 5.61
7 -1.00 -1.00 -1.00 1.00 -1.00 1.00 1.00 -1.00 1.00 1.00 1.00 5.91
8 1.00 -1.00 1.00 1.00 1.00 -1.00 -1.00 -1.00 1.00 -1.00 1.00 16.2
9 -1.00 -1.00 1.00 -1.00 1.00 1.00 -1.00 1.00 1.00 1.00 -1.00 2.58
10 -1.00 1.00 1.00 -1.00 1.00 1.00 1.00 -1.00 -1.00 -1.00 1.00 0.3
11 1.00 1.00 -1.00 -1.00 -1.00 1.00 -1.00 1.00 1.00 -1.00 1.00 15.75
12 1.00 -1.00 -1.00 -1.00 1.00 -1.00 1.00 1.00 -1.00 1.00 1.00 0.36
+ “sign indicates high level of the variable a nd” – “sign indicates low level”

along with 5 dummy variables generated by the soft- test (F test). P (probability) values and determination
ware. All the experiments were done in triplicate and the of coefficient obtained determines regression model’s
average of lipase production was taken as response. The goodness of fit.
obtained results are matched with the polynomial equa-
Box-Behenken Design
tion 1 in significant order 1:
After optimizing the various nutritional variables by
Y = 0 +  iXi (i=1, 2, 3……………k) (1)
PBD, the 3 most significant variables (Tannic Acid, KCl
Where Y is the response (Enzyme activity), 0 is model and KH2PO4) were further chosen for response surface
intercept, i is linear coefficient, Xi is level of the inde- methodology using Box-Behnken design. Design-Expert®
pendent variables. The model was statistically analyzed statistical software was used to analyze the experimental
and the overall significance of the model was evaluated design. A Box-Behnken design with a set of 12 experi-
by ANOVA (Analysis of variance) involving Fischer’s ments was generated (Table 4). Each variable was stud-

Table 3. ANOVA for the experiment with PBD for Tannase Production by
AT 13
Analysis of variance table [Partial sum of squares - Type III]
Sum of Mean F p-value
Source Squares df Square Value Prob > F
Model 427.76 10 42.78 285.87 0.0460 significant
A-TANNIC ACID 153.08 1 153.08 1023.04 0.0199
B-KCl 9.12 1 9.12 60.93 0.0811
C-NaNO3 15.01 1 15.01 100.30 0.0634
D-KH2PO4 9.97 1 9.97 66.65 0.0776
E-K2HPO4 11.96 1 11.96 79.93 0.0709
G-G 77.32 1 77.32 516.71 0.0280
H-H 69.89 1 69.89 467.08 0.0294
J-J 43.85 1 43.85 293.07 0.0371
K-K 19.15 1 19.15 127.99 0.0561
L-L 18.40 1 18.40 122.98 0.0573
Residual 0.15 1 0.15
Cor Total 427.91 11

BIOSCIENCE BIOTECHNOLOGY RESEARCH COMMUNICATIONS STATISTICAL OPTIMIZATION OF EXTRACELLULAR TANNASE PRODUCTION 693
Archana and Lakshmi

FIGURE 1. Pareto chart showing the effect of media components on tannase

activity using AT 13

ied at three levels, +1, 0, -1 where “0” is the central coefficients, ii the squared coefficients and ij the
coded value, “+1” high value and “-1” low value. The interaction coefficients. The model was statistically
fermentation was carried out in 250 ml flasks containing analyzed and the overall significance of the model was
100 ml of the production medium. All experiments were evaluated by ANOVA (Analysis of variance) involving
done in triplicate and tannase production obtained was Fischer’s test (F test). P (probability) values and coef-
taken as the response. ficient of determination obtained determines regression
The obtained results are matched with the polynomial model’s goodness of fit. The optimum values of each of
equation 2 in significant order the 3 significant variables were determined by solving
regression equation.
Y = 0 +  iXi +  iiXi 2 +  ijXiXj (2)
The interactive effects of the variables on tannase
where Y is the predicted response, 0 is the intercept production were studied by analyzing the 3D and coun-
term, Xi and Xj are the input variables, i the linear ter plots which depicted the interactions graphically.
Validation of the model: The statistical model was again
Table 4. Box-Behenken (BB) design matrix of tested by carrying out the fermentation under the opti-
optimizing factors influence for tannase activity mum conditions that were obtained through the Box-
Tannic Acid KCl KH2PO4 E.A Behnken design.
Run A:A B:B C:C R1
1 -1.00 1.00 0.00 12.99 RESULTS AND DISCUSSION
2 -1.00 0.00 -1.00 14.61
3 0.00 1.00 1.00 17.34 SCREENING OF VARIABLES BY PBD
4 0.00 -1.00 1.00 12.33
Iimportant fermentative parameters can be effectively
5 0.00 1.00 -1.00 11.52 and quickly picked up by applying PBD. The use of sta-
6 1.00 0.00 1.00 9.27 tistical tools not only saves time by simultaneously opti-
7 1.00 0.00 -1.00 10.71 mizing several process variables with few experimental
8 1.00 1.00 0.00 2.61 runs but also reduces the cost of fermentation. A total
9 -1.00 -1.00 0.00 9.66 of 5 important variables Viz., tannic acid, KCl,NaNO3,
KH2PO4 and K2HPO4 which were selected for statistical
10 1.00 -1.00 0.00 6.57
optimization using PBD (Table 1).
11 0.00 -1.00 -1.00 17.79
Experiments were carried out based on Plackett–Bur-
12 -1.00 0.00 1.00 18.12 man design and the results obtained are given in Table 2.

694 STATISTICAL OPTIMIZATION OF EXTRACELLULAR TANNASE PRODUCTION BIOSCIENCE BIOTECHNOLOGY RESEARCH COMMUNICATIONS
 Archana and Lakshmi

FIGURE 2a. Response surface plot of tannase production by AT 13 showing interaction between
tannic acid and KCl

FIGURE 2b. Response surface plot of tannase production by AT 13 showing interaction

between KCl and KH2PO4

BIOSCIENCE BIOTECHNOLOGY RESEARCH COMMUNICATIONS STATISTICAL OPTIMIZATION OF EXTRACELLULAR TANNASE PRODUCTION 695
Archana and Lakshmi

FIGURE 2c. Response surface plot of tannase production by AT 13 showing interaction between KH2PO4 and
tannic acid

From the table, it was observed that the variation in tan- ing important factors that are mainly responsible for
nase activity was 0.3–16.53 U/ml/min. From the Pareto enzyme production. From the chart it was evident that
chart (Fig. 1), only tannic acid was found to be signifi- the most important contributing factor for tannase pro-
cant for the production of tannase by AT 13. duction was tannic acid only.
Figure 1 show the Pareto chart of effects of vari- The ANOVA results are given in Table 3. The p-value
ables on tannase production which helps in. identify- of ANOVA table serves as a tool for checking the sig-

Table 5. ANOVA for Response Surface Quadratic Model for the production of
tannase by AT 13
Analysis of variance table [Partial sum of squares - Type III]
Sum of Mean F p-value
Source Squares df Square Value Prob > F
Model 239.38 8 29.92 187.04 0.0006 significant
A-A 85.94 1 85.94 537.18 0.0002
B-B 0.45 1 0.45 2.79 0.1934
C-C 0.74 1 0.74 4.61 0.1209
AB 13.29 1 13.29 83.05 0.0028
AC 6.13 1 6.13 38.29 0.0085
BC 31.81 1 31.81 198.84 0.0008
A2 92.14 1 92.14 575.97 0.0002
B2 54.50 1 54.50 340.66 0.0003
C2 0.000 0
Residual 0.48 3 0.16 187.04
Cor Total 239.86 11

696 STATISTICAL OPTIMIZATION OF EXTRACELLULAR TANNASE PRODUCTION BIOSCIENCE BIOTECHNOLOGY RESEARCH COMMUNICATIONS
 Archana and Lakshmi

nificance of each of the coefficients and is indicative centration of tannic acid (A) reaching to its maximum
of interaction strength of each independent variable. at 18.12 U/ml/min, where as the enzyme activity gradu-
P-value less than 0.05 indicate that the model terms are ally increased with increase KCl (B) concentration and
significant. In this experiment, the Model F-value of then declined (Figure 2a).Significant interactions were
285.87 implied that the model is significant in enhanc- also present between KCl (B) and KH2PO4(C)with a low
ing enzyme production In this case A (Tannic acid ), G p-value (0.0008) and between KH2PO4(C) and tannic
(Dummy), H (Dummy) and J (Dummy) gave significant acid(A) with a low p-value (0.0085) (Figure 2b and 2c).
terms and B, C, D, K and E gave negative significant
Validation of the Model: Validation of the response sur-
terms.
face model was confirmed by using optimum conditions
Optimization of significant variables by Response obtained by the model. The experimental values were
surface methodology (RSM) very close to the predicted values and hence it was con-
cluded that the model was successfully validated.
Tannic acid was found to be significant by PBD was
optimized by RSM using Box-Behnken design for tan-
nase production. The response values in terms of tan-
CONCLUSION
nase activity and the matrix design were represented in
Table 4. The response surface methodology was applicable for
The results obtained by Box-Behnken were analyzed the production of tannase from Streptomyces sp. AT13.
by Analysis of variance (ANOVA) with a model F-value In the present study,optimum conditions for tannase
of 187.04 indicating that the model was significant production by Streptomyces sp. AT13 was done using
(Table 5). Values of Prob > F less than 0.05 indicate that Plackett-Burmen design followed by Response surface
the model terms are significant. In the present model methodology using Box-Behnken design. From among
A, AB, BC, AC, A2, B2 are significant model terms. The 6 fermentative variables that were studied, 5 significant
Model F-value of 187.04 implies the model is signifi- variables were picked up by PBD. Among 5 variables
cant. There is only a 0.06% chance that an F-value this from PBD, 3 were further optimized by Box-Behnken
large could occur due to noise. The Values of “Prob > design. Analysis of variance (ANOVA) showed the sig-
F” less than 0.0500 indicate model terms are signifi- nificance of the model and the validity of the model was
[Link] fit of the model was checked with the coef- confirmed by the verification experiments. P-value less
ficient of determination R2, which was calculated to be than 0.05 indicate that the model terms are significant.
0.9680. The determination coefficient (R2) of the model The optimum values of parameters obtained through
was 0.9980 indicating that 99.80% of variability in the RSM were 1 % tannic acid, 0.5 % KCl and 0.1% KH2PO4
response could be accounted by the model. to achieve maximum tannase production.
The ANOVA gave the following regression equation:
R1 = +19.97 -3.28* A -0.24* B+0.30* C-1.82*
AB-1.24* AC+2.82* BC-6.79* A2-5.22*+0.000* C2 ACKNOWLEDGMENTS
Where R1 is the tannase activity; A is tannic acid; B is The Authors are thankful to the head of Department of
KCl; C is KH2PO4 and the interactions AB, BC, AC, A2, B2 Biotechnology and Microbiology, KSV University for
are significant. providing all the laboratory facilities. The authors are
also thankful to Biogene department of Gujarat State
3D Surface Plots Biotechnology Mission (GSBTM), Gandhinagar, Gujarat,
The three-dimensional response surface curves were India for bacterial strain identification.
plotted to study the interaction among different phys-
icochemical factors used and to find out the optimum COMPLIANCE WITH ETHICAL STANDARDS
concentration of each factor for maximum tannase pro-
Conflict of Interests: There are no conflicts of interest.
duction.
For the construction of 3D plots, effect of two vari-
ables was considered while the other variable was held REFERENCES
at zero. These plots were helpful in understanding both Barthomeuf, C., Regerat, F., Pourrat, H. (1994). Production,
the individual and the interaction effects of the two fac- purification and characterization of a tannase from Aspergil-
tors. For the construction of 3Dplots, effect of two vari- lus niger LCF 8. J. Ferment. Bioeng. 3: 320-323.
ables was considered while the other variable was held Bate-Smith, E.C. and Swain, T. (1962). Flavonoid compounds.,
at zero. From the 3D plots, it was clear that the tannase In: Comparative biochemistry. Academic Press, Pp 755-809
activity significantly increased with increase in the con- (Eds) Mason and A.M. Florkin. New York.

BIOSCIENCE BIOTECHNOLOGY RESEARCH COMMUNICATIONS STATISTICAL OPTIMIZATION OF EXTRACELLULAR TANNASE PRODUCTION 697
Archana and Lakshmi

Beena P.S., Soorej, M.B., Elyas, K.K., Sarita, G.B., Chandrasekaran, Mohan, S. K. (2014). Optimization of tannase production from
M. (2010). Acidophilic tannase from marine Aspergillus awamori agro industrial wastes- kinetics and modelling. From: Ph. D.
BTMFW032. J Microbiol Biotechnol., 20:1403–1414. Thesis. Department of Chemical Engineering, Annamalai Uni-
Bhat, T.K., Singh, B. and Sharma, O.P. (1998). Microbial degrada- versity, Annamalainagar, Tamilnadu, India.
tion of tannins–a current perspective. Biodegradation, 9:343-357. Pan, C.M., Fan, Y.T., Xing, Y., Hou, H.W., Zhang, M.L. (2008).
Dilipkumar, M., Rajasimman, M., Rajamohan, N. (2011). Appli- Statistical optimization of process parameters on biohydrogen
cation of statistical design for the production of inulinase production from glucose by Clostridium sp. Fanp2. Bioresour
by  Streptomyces sp. using press mud.  Front Chem Sci Eng.; Technol.; 99(8):3146-54.
5(4):463–470. doi: 10.1007/s11705-011-1112-1. Rana, N. K. and Bhat, T. K. (2005). Effect of Fermentation
Haslam, E. (1989). In Plant Polyphenols Plant Polyphenols: Vege- System on the Production And Properties of Tannase from
table Tannins Revisited. Cambridge: Cambridge University Press. Aspergillus niger van Tighem MTCC 2425. Journal of Applied
Cambridge, Pp 230. Cambridge. Microbiology. 51: 203-212.

Lekha, P.K., Lonsane, B.K. (1997). Production and application Rodríguez, H., De las Rivas, B., Gómez-Cordovés, C., Munoz, R.
of tannin acyl hydrolase: state of the art. Adv Appl Microbiol (2008). Degradation of tannic acid by cell-free extracts of Lac-
44:215–260. tobacillus plantarum. Food Chem.107:664–670.

Mannan S., Fakhruĺ-Razi, A., Alam, M.Z. (2007). Optimiza- Sharma, S., Bhat, T.K., Dawra, R.K. (2000). A Spectrophoto-
tion of process parameters for the bioconversion of activated metric Method for Assay of Tannase Using Rhodonine. Anal.
sludge by Penicillium corylophilum, using response surface Biochem. 278: 85-89.
methodology. J Environ Sci. (China):19(1):23-8. Van de Lagemaat, J. and Pyle, D.L. (2006). Tannase. In: Enzyme
Mcleod, M.N. (1974). Plant tannins – their role in forage qual- Technology. Pp 380-397 (Eds) A. Pandey, A.C. Webb, C.R. Soc-
ity. Nutrition Abstracts and Reviews 44:803–815. col and C. Larroche,1st Edn., Springer, New York.

698 STATISTICAL OPTIMIZATION OF EXTRACELLULAR TANNASE PRODUCTION BIOSCIENCE BIOTECHNOLOGY RESEARCH COMMUNICATIONS