International Conference on Challenges in Biotechnology and Food Technology

Process variables optimization by taguchi design and artificial

intelligence techniques for amylase production by Bacillus subtilis using
agro residual wastes
a
Mohandas Bhat S and bPrabhakar A
a
Department of Biotechnology, MSRIT, Bangalore 560054, India
Email: mohandasbhats@yahoo.com
b
Department of Chemical Engineering, SVU College of Engineering, Sri
Venkateswara University, Tirupati 517502, AP, India
Email: prabhakar52@gmail.com

ABSTRACT

Advances in industrial biotechnology offer potential opportunities for economic

utilization of agro-industrial residues such as banana peel and sweet corn stem stump.
Due to their rich organic nature, they can serve as an ideal substrate for microbial
processes for the production of value added products like alpha amylases. This paper
describes optimization of process variables for amylase production using banana peel and
sweet corn stem stump by Bacillus subtilis by orthogonal array design and artificial
intelligence techniques. The orthogonal array design was used to study the effects of
carbon and nitrogen source, pH, temperature, innoculum size on amylase production. Of
the five parameters examined, the effects of process variables for amylase production by
Bacillus subtilis were in the order of temperature>pH>carbon source>nitrogen
source>innoculum size respectively. Artificial Neural Networks (Bayesian training
function, trainbr) and genetic algorithms are used to model and further optimize the
significant parameters from orthogonal array design. The temperature, pH, carbon source
and nitrogen source served as inputs to the neural network models, and alpha amylase
activity served as a single output of each model. Genetic algorithms were used to
optimize the input space of the neural network models to find the optimum settings for
maximum enzyme production. Using this procedure, orthogonal array and artificial
intelligence techniques have been effectively integrated to create a powerful tool for
process modeling and optimization.

Corresponding Author:

Prof. Mohandas Bhat S

Department of Biotechnology,
M. S. Ramaiah Institute of Technology,
MSRIT Post
Bangalore 560054, India
Email: mohandasbhats@yahoo.com
Ph: 080- 23600822 (Extn: 167)
0 9341223016

Introduction

Banana (musa sapientum) is grown extensively in tropical and subtropical countries and
37% of the world’s production is shared by India. From the banana produce, in addition
to the fruit waste, the stem, leaves and pseudostem are also accumulated as waste in the
environment posing serious environmental problems. Sweet corn (Zea mays), also called
indian corn, sweetcorn, sugar corn, pole corn, or simply corn, is a variety of maize with a
high sugar content. Sweet corn is the result of a naturally-occurring recessive mutation in
the genes which control conversion of sugar to starch inside the endosperm of the corn
kernel. United States shares approximately 40% of the entire world’s production which
results in millions of tons of agricultural wastes. Several attempts have been made to
utilize these wastes through ensilaging and to eliminate or reduce the negative nutritional
effects.

In recent years the new potential of using microorganisms as biotechnological sources of

industrially relevant enzyme has stimulated renewed interest in the exploration of extra
cellular enzymatic activity in several microorganisms (Bilinski and Stewart, 1990;
Omemu et al. 1999; Buzzini and martini, 2002). Amylases are important enzymes
employed in starch processing industries for the hydrolysis of polysaccharides and also
find application in food, baking, brewing, detergent, textile, paper and distilling industry.
Alpha amylases (endo-1, 4-α -d-glucanglycohydrolase EC 3.2.1.1) constitute the family
of; endo amylases that randomly cleave the 1,4-α -d-glycosidic linkages between
adjacent glucose unit in the linear amylose chain with retention of alpha – anomeric
configuration in the products.

Production of alpha amylases has been investigated through submerged (SMF)

Fermentation (Carlsen et al. 1996; Francis et al. 2002; Pandey et al. 1999). In our
experiment the production of α -amylase was investigated through submerged
fermentation using banana peel and sweet corn stem stump agro residual wastes. The
optimization of the process parameters is done to enhance the yield of α -amylase using
orthogonal array technique and artificial neural networks. Process parameters like pH,
inoculum size, substrate carbon and nitrogen concentrations, temperature, incubation time
were considered for study. The results obtained indicated that all the parameters
considered had significant effect on the production of amylase and the optimum range of
these parameters were established experimentally.
Materials and Methods

Microorganism
Bacterial strain of Bacillus Subtilis MTCC 441 obtained from the Microbial Type Culture
Collection (MTCC, Chandigarh, India) was used for amylase production throughout this
study.

Preparation of substrate
The banana peel and sweet corn stem stump were sliced, spread on trays and oven dried
at 50°C for 48 hrs. The dried slices of banana peel and sweet corn stem were separated,
grounded and sieved through standard-mesh sieves to obtain particles of various sizes
ranging from 0.4 mm to 3.2 mm that were stored in polyethylene bags at room
temperature (30 ± 2°C) until use.

Submerged Fermentation (SMF) Medium:

The strain was cultivated on the slants of Nutrient Agar Media and incubated at 350C for
24 days. The medium was a chemically defined medium developed in our laboratory
consisting: Yeast extract 2 g/L, Beef extract 1 g/L, Peptone 5 g/L and Sodium Chloride 5
g/L. The initial pH was adjusted to 7. The temperature and agitation rate were 350C and
200 rpm during the fermentation processes, respectively.

Alpha-amylase production using banana agro-residual waste (banana peel) and sweet
corn stem stump
The banana agro-residual waste (banana peel of particle sizes ranging from 600-1200
µ m) or sweet corn stem stump (particle ranging from 500 – 1200µ m) was weighed 5 g
in 250 ml-Erlenmeyer flasks containing 100 ml of the above SMF medium, autoclaved at
121°C for 60 min, cooled to about 30°C and inoculum with 1ml of Bacillus Subtils MTCC
441 (106 cells per ml). Fermentation was carried out on a rotatory shaker at 200 rpm at
35°C. Samples were taken at 12-h intervals, filtered through Whatman filter paper No. 1
and kept at 4°C for further assays.

Enzyme assays
The reaction mixture consisting of 0.5 ml of soluble starch (1% (w/v)) solution made in
50 mM phosphate buffer (pH 6) and 50 µl of appropriately diluted cell-free culture-
supernatant was incubated at 50°C for 10 min. The reducing sugar formed in the reaction
was measured by 3,5-dinitrosalicylic acid (DNS) method using glucose as the standard
(Miller, 1959). One unit of a-amylase is defined as the amount of enzyme required to
produce 1 µmol of reducing equivalents per minute from soluble starch under the assay
conditions.

Orthogonal arrays for process variable optimization

To examine the interactions among process variables for amylase production, L18
orthogonal array was used to examine the effects of five factors (pH, Temperature,
Carbon Source (banana peel/corn stem stump), Peptone and innoculum size) with four
levels separately for banana peel and sweet corn stem stump respectively (Table 1). The
orthogonal arrays, data analysis and ANOVA were obtained using Design Expert Trail
Version (2007) (Stat-Ease, USA) software based on Taguchi method.

Table 1:

Factors Level 1 Level 2 Level 3 Level 4

Temperature (A) 25 30 35 40
pH (B) 5 6 7 8
Innoculum Size (C) (ml) 1 2 3 4
Bannana Peel (D) (g) 4 6 8 10
Sweet corn Stump (D) (g) 4 6 8 10
Peptone (E) (g) 0.1 0.2 0.3 0.4

Results and Discussion

Experimental design for the orthogonal array

Factor and level assignments including main effects of each factor on amylase activity are given
in Table 1. Each value for levels was based on the optimized medium determined by the one-
factor-at-a-time method (unpublished results). Based on L16 orthogonal array design, we carried
out only 16 experiments in triplicate. In full-factorial experimental designs, to reach the same
results as those of the orthogonal array method with this variety of factors, at least 625
experiments are necessary. The experimental conditions for each run are listed in Table 2,
including the enzyme activities in the last column. The ANOVA for the experimental results
obtained by Design Expert software and optimal levels of each factor for obtaining higher
amylase are given in the upper part of Table 3. The order of factor effects (contribution percent)
on amylase production was found to be Temperature>pH>carbon Source>Peptone>innoculum
size. The results show that the effects of Temperature and pH were more significant than those of
the other factors. To confirm the data, an experiment was carried out using optimal
concentrations of nutrients according to upper part of the last column of Table 3. The measured
amylase activity obtained was _______ units per ml.
Table 2:

Enzyme Activity Enzyme Activity

(U/ml) (U/ml)
S.No. A B C D E
(Sweet Corn Stem (Banana Peel as
Stump as substrate substrate
1 23.75
13.76
1 1 1 1 1
2 45.13
13.86
1 2 2 2 2
3 26.39
16.13
1 3 3 3 3
4 47.30
9.26
1 4 4 4 4
5 51.10
14.65
2 1 2 3 4
6 68.19
19.72
2 2 1 4 3
7 17.43
18.97
2 3 4 1 2
8 67.54
21.34
2 4 3 2 1
9 52.31
14.71
3 1 3 4 2
1 67.29
0 21.01
3 2 4 3 1
1 35.45
1 14.18
3 3 1 2 4
1 36.41
2 11.06
3 4 2 1 3
1 47.77
3 9.90
4 1 4 2 3
1 64.52
4 3.82
4 2 3 1 4
1 71.66
5 20.28
4 3 2 4 1
1 23.75
6 13.76
4 4 1 3 2