Abstract: Cyclodextrin glycosyltransferase (CGTase) activity was produced by the Bacillus sp., subgroup alcalophilus in a culture medium containing cassava starch. A central composite design and response surface methodology were used to study the influence of carbon source (cassava starch), nitrogen sources (yeast extract and tryptone) and sodium carbonate in the production medium. Assays were performed in 300 mL Erlenmeyer flasks containing 100 mL of production medium maintained in a shaker at 150 rpm at 35±1°C for 72 h of fermentation. The independent variables [0.75% cassava starch, nitrogen sources (0.375% yeast extract and 0.375% tryptone) and 1% Na2CO3] produced an enzyme activity of 96.07 U mL-1.
INTRODUCTION
Cyclodextrins (CDS) are products of the enzyme, cyclodextrin glucanotransferase (CGTase) and are cyclic oligosaccharides formed by glycosidic units through α-1,4 bonds (Qi et al., 2004). The most industrially used CDS are α-CD, β-CD and λ-CD, made up of 6, 7 and 8 glucose units (Seon et al., 2009). Cyclodextrin have a truncated conical shape, the outer side of which is hydrophilic due to primary and secondary hydroxyls and the inner cavity is apolar due to the high electron density caused by pairs of free electrons from the oxygens involved in the glycosidic bonds being directed toward the interior of the cavity (Ribeiro and Sera, 2007). The inclusion complex may be formed by molecules of different shapes and sizes, capable of either completely or partially adapting within the apolar cavity, causing changes in undesirable properties of the host molecules (Gornas et al., 2009; Haiyee et al., 2009). The inclusion of CDS in the rigid matrix generally causes changes in the physiochemical properties of the inserted molecules, such as an increase in stability and biocompatibility (Moraes et al., 2007). These properties have made CDS increasingly more important for industrial applications and are employed in pharmaceutical, food and agro-industrial products (Aime et al., 2009). The considerable variety of inclusion complexes (Abdel-Shafi and Ai-Shihry, 2009) provides beneficial effects, such as protection from lipid oxidation (Haiyee et al., 2009), reduction or stability of aromas and flavors (Seon et al., 2009), color stability (Provenzi et al., 2006) vitamin stability (Cho et al., 2001) and the protection of active ingredients from oxidation (Szente and Szejtli, 2004) thereby reducing storage costs and increasing the shelf life of food products.
The enzyme CGTase (EC 2.4.1.19) catalyzes starch conversion reactions by means of cycling, coupling and disproportioning in different proportions of α-CD, β-CD and γ-CD formation, depending on culture conditions, environmental factors and the microorganism used in the production (Biwer et al., 2002). Thus, optimizing the production of the enzyme has been the aim of many studies using different starch sources, such as cassava (Higuti et al., 2003).
The present study used the response surface method through a Central Composite Design (CCD) to select carbon and nitrogen sources as well as sodium carbonate variables that may influence the production of CGTase by the microorganism, Bacillus sp., subgroup alcalophilus. This experiment design is an effective method that uses multifactor productivity measures, thereby economizing time and cost in the investigation of the effect of variables on the results of experiments (Seon et al., 2009).
MATERIALS AND METHODS
Microorganism
The Bacillus sp., subgroup alcalophilus strain was obtained
from the culture collection of the industrial microbiology laboratory of the
Biological Sciences Institute of Rio Claro (SP, Brazil) a culture isolated from
a cassava starch mill (Riberão Bonito, SP, Brazil). This microorganism
was characterized by the André Tosello Foundation (Campinas, SP, Brazil).
Identification of the Microorganism
The identification of the microorganism was achieved from the phylogenetic
analysis of partial 16S ribosomal DNA (16S rDNA) sequences, which were compared
to 16S rDNA sequences from organisms represented in the Ribosomal Database Project
(Wisconsin, USA) and Genbank databases. Distance matrices were calculated using
the Jukes and Cantor (1969) model and the design of the phylogenetic tree (Fig.
1) based on evolutional distances was performed using the Neighbor-Joining
method (Saitou and Nei, 1987).
Activation of Inoculum
The Bacillus sp., subgroup, alcalophilus was cultured in a
medium containing (w/v) 1% cassava starch, 0.5% yeast extract, 0.5% tryptone,
0.1% K2HPO4 and 0.02% MgSO4 7H2O,
completed with 10% Na2CO3 in Erlenmeyer flasks (250 mL)
at 150 rpm and 35±1°C for 56 h, through successive transferences
to fresh media at 8 h intervals. The production of the CGTase enzyme was observed
by the formation of a yellow halo around the bacterial colony. The viability
of microorganism and the possibility of enzyme production were determined in
a solid, Nakamura and Horikoshi (1976) medium and analyzed based on color intensity
and diameter of the halos formed in the medium.
| Fig. 1: | Phylogenetic tree based on analysis of partial 16S rDNA of sample 010200 and stains of Bacillus and Bacillus circulans (the latter of which was used as the outgroup). No. following species name correspond to the access acronym Genbank database |
Enzyme Production Through Shaker Fermentation
After selecting the parameters that influence CGTase production, such as
type and concentration of carbon source (ground cassava, nitrogen sources (yeast
extract and tryptone) and Na2CO3, the option was made
to use 1% cassava starch (w/v), 0.5% yeast extract (w/v), 0.5% tryptone (w/v)
and 1.5% Na2CO3 (w/v). These parameters were used as the
central point in the CCD. The Bacillus sp., Subgroup alcalophilus
was cultured in 300-mL Erlenmeyer flasks containing 100 mL of the culture
medium incubated at 150 rpm at 35±1°C for 18 h. From this culture,
10 mL (10% v/v) was inoculated in 90 mL of the Nakamura and Horikoshi (1976)
culture in 300 mL Erlenmeyer flasks. At the end of fermentation periods of 24,
48 and 72 h, samples were removed and centrifuged at 10,000 rpm for 10 min.
The cell-free supernatant was used for the determination of enzyme activity.
All assays were performed in duplicate.
Enzyme Activity
Enzyme Activity (EA) was measured based on the phenolphthalein method (Makela
et al., 1988). Five mililiter of the supernatant from the enzyme broth
and 5.0 mL of 1% starch solution [0.1 g of soluble starch, 1.0 mL of CaCl2
solution (0.05 M), 1.0 mL of Tris-HCl buffer solution (0.05 M; pH 8.0) and distilled
water for a final volume of 10 mL] were placed in a thermostatic reactor at
55±1°C. Samples were removed at 0, 3, 6, 9 and 12 min and inactivated
at 100°C for 5 min. Cyclodextrin was quantified as a function of time by
adding 2.5 mL of phenolphthalein solution (pH 10.5) to 0.5 mL of the inactivated
samples. Absorbance of the final solution was read from a spectrophotometer
at 550 nm. The EA was defined as the amount of cyclodextrin produced per minute.
Calculation of Enzyme Activity
Graphs were designed with the slope value of each assay, with a linear adjustment
performed on the OriginPro software program (version 7.5). The EA was determined
by the following equation:
| Table 1: | Levels of variables used in the central composite design |
| (1) |
where, A is EA (μmol/min/mL); α is the angle coefficient of the standard curve (μmol.mL-1/ABS); β is the angle coefficient of the experimentally obtained curve (ABS/min.); v is the enzyme volume from the reactor (Vsample (mL)/V(mL)) and Fdil is the dilution factor.
Central Composite Design
The experiments were carried out based on a CCD. To assess the tendency
of variables in the CGTase production process, a two-level factor design was
used with three variables, two replicates on the central point and six experiments
on the axial points (α), totaling 16 experiments. The variables were coded
based on the following equation:
| (2) |
where, xi is the coded value of the independent variable, Xi is the real value of the independent variable, Xcp is the real value of an independent variable at the central point and ΔXi is the step change value.
Table 1 shows the variations and levels of the independent variables analyzed. The Root Mean Square (RSM) was generated from an empirical equation produced by CCD.
Errors in the production measurements were estimated by means of duplicate assays. The results were analyzed with the aid of the Statistica 7 software program (Statistic 7 Stat Soft, USA).
RESULTS AND DISCUSSION
The microorganism was identified as a Bacillus sp., from the subgroup alcalophilus, still without species definition and therefore not described in the literature. The current classification of species within the genus Bacillus and related genera is well established and is based on a combination of numerous experimental approaches (Xu and Côté, 2003). A tree of Bacillus 16S rRNA sequences, including 8 type strains of Bacillus species, clustered all of the isolates belonging to the previously identified species to the corresponding specie together (Fig. 1). We found that the sample 010200 isolate was distributed in the subgroup alcalophilus of the tree (Fig. 1).
Optimization of Significant Medium Components Using Central Composite Design
The CCD in AE was employed to investigate the interactions among the significant
factors (carbon source, nitrogen sources and sodium carbonate) and determine
the optimal values of the enzyme activity. The design matrix and the corresponding
results of AE experiments to determine the effects of three independent variables
are shown in Table 2. The best concentrations were 0.75% carbon
and nitrogen and 1% for Na2CO3. CGTase production increased
from 10.7±0.43 to 96.0±0.02 U mL-1.
| Table 2: | Central composite design with real and predicted enzyme activity values |
| x1: Cassava starch; x2: Tryptone and yeast extract; x3: Sodium carbonate | |
| Table 3: | Regression results of CGTase production |
| R2: 0.901; R: 0.95; Adjusted R2: 0.80; fT: 10.16; fC (4.11): 3.36 | |
The results obtained during the peak EA were analyzed using multiple regression, with the isolated variables, quadratic variables and their interactions as factors. Table 3 shows the results of the significant effects of the variables, together with the coefficient of determination (R2) and F value from the Fisher distribution as the response to CGTase production.
The lowest and the highest values of CGTase activity were obtained when maximal and minimal levels of cassava starch (X1) and nitrogen sources (X2) were used, respectively. These results suggested that this variable significantly affect the CGTase production and hence, its activity. The low level of cassava starch (-1.0 in coded value or 0.75%, w/v) and nitrogen sources (-1.0 in coded value or 0.75%, w/v) allowed the strain to produce a higher CGTase activity than the higher level of carbon and nitrogen sources concentration as showed in Table 2. We can observe in Table 2 by indication of software that with a decrease in concentrations of variables X1 (experiments 9 and 10) and X2 (experiments 11 and 12) for the same values of X3, the EA increased.
As can be seen from Table 3, the Student’s t-values for factors cassava starch concentration (X1), nitrogen sources concentration (X2), interaction between cassava starch and yeast carbonate sodium (X1X3) and interaction between nitrogen sources and carbonate sodium (X2X3) were above the 1% level of significance. The result of the calculated F (Fc) was greater than the tabulated F (FT) at the 1% level of significance, thereby allowing the rejection of H0 at a 1% level of significance. Thus, there is 99% confidence that the model is significant. The coefficient of determination (R2) of 0.80 indicates an adequate adjustment of the experimental data for the EA response based on the proposed empirical equation, that is, 80% of the variability in the data was explained by the equation from the adjusted model.
The model can be applied in screening crucial and critical medium components. The Eq. 3 was obtained from analysis of multiple regression and the significant terms were included in the following equation:
| (3) |
where, EA is enzyme activity (U mL-1), X1 is cassava starch concentration (%), X2 is concentration of ½ tryptone and ½ yeast extract; and X3 is Na2CO3 concentration.
Based on the experimental results of CGTase production, the quadratic model of CGTase activity as a function of carbon, nitrogen and Na2CO3 concentrations generated two linear terms and two interactions, using the Statistic 7 software program.
The variable with positive fitted constant has an enhancer effects towards CGTase production than the one with negative coefficient, which had inhibitory effects. Therefore, the negative sign of the isolated variables (X1 and X2) means that a reduction in the concentration of cassava starch and nitrogen sources increases enzyme activity and positive sign of the interactions of variables (X1X3 and X2X3) means that a reduction in the concentration of cassava starch/carbonate sodium and nitrogen sources/carbonate sodium decrease enzyme activity.
Response surfaces were designed in order to facilitate the visualization of the effect of the independent variables (cassava starch, nitrogen sources and Na2CO3) on EA (Fig. 2-4).
Production of the CGTase enzyme was achieved by the substitution of the concentrations of nitrogen sources, carbon source and sodium carbonate. The greatest EA was achieved when the soluble starch was substituted with cassava starch and CCD variations were performed on the concentrations of carbon and nitrogen sources and sodium carbonate in the culture medium. Moreover, enzyme production by Bacillus sp., subgroup alcalophilus was best at the fermentation time of 72 h. Analyzing the response surfaces (Fig. 2-4), an optimal region for the EA of 96.07 U mL-1 was determined at the concentrations of cassava starch (0.75 %), nitrogen sources (0.75%) and sodium carbonate (1%).
| Fig. 2: | Response surface of CGTase production as a function of cassava starch concentration and concentration of nitrogen sources |
| Fig. 3: | Response surface of CGTase production as a function of cassava starch concentration and concentration of sodium carbonate |
| Fig. 4: | Response surface of CGTase production as a function of sodium carbonate concentration and concentration of nitrogen sources |
The CCD proved an effective method for determining the production of CGTase by the Bacillus sp., subgroup alcalophilus. The following is the optimized culture medium for intensifying CGTase production (%): cassava starch, 0.75; tryptone, 0.375; yeast extract, 0.375; K2HPO4, 0.1; MgSO4, 7H2O, 0.02; Na2CO3, 1 and initial pH of 9.2, which resulted in a 14.28% increase when compared to the original medium.
In a earlier study using the microorganism Bacillus sp., subgroup, alcalophilus (Freitas et al., 2004), the CGTase activity of 88.6 U mL-1 was achieved using 1.5% starch, 1.5% nitrogen and 1% sodium carbonate in 72 h of fermentation.
Adriana et al. (2002) achieved an EA of 5.8 U mL-1 with optimized concentrations of 1.5 g L-1 of cassava starch and 0.4 g L-1 ammonium sulfate using CCD. Zain et al. (2007) achieved an EA of 80.12 U mL-1 with optimized concentrations of 3.3 g L-1 of carbon and 0.13 g L-1 of nitrogen, analyzed using the response surface method. According to Khairizal et al. (2004) optimized values of 1.48 and 1.89% starch and yeast extract during the fermentation for CGTase production yielded an activity of 84 U mL-1, using a mathematical model provided by the response surface method, thereby demonstrating the importance of using an experimental design with variations in concentrations of carbon and nitrogen sources for CGTase production. This happen due to the presence of concentration of some essential nutrients to stimulate CGTase production. The CGTase production was influenced by the level of cassava starch, nitrogen sources and sodium carbonate (Table 3). As shown in response surfaces (Fig. 3-4), the effect of cassava starch, nitrogen sources and sodium carbonate concentration on CGTase production by Bacillus sp., subgroup alcalophilus was simulated according to the model.
Experimental Validation of the Optimized Conditions
To confirm the model adequacy and the results from an analysis of the response
surface, three additional experiments were repeated the optimal medium composition.
As a result, the mean value of AE was 93.1±0.06 U mL-1, which
was close to the predicted value (102.96 U mL-1). The correlation
between predicted and measured values verifies the model validation, indicating
that the model was adequate for obtaining the optimum region in the range of
studied parameters.
CONCLUSION
The value of 96.07 U mL-1 obtained in the present study had a 6.545% error in relation to the predicted value of 102.96 U mL-1, thereby demonstrating the considerable reliability of the results. The results indicate the potentiality of using milled cassava as a fermentation substrate for the production of CGTase using the microorganism, Bacillus sp., subgroup alcalophilus and the use of experimental design in the medium optimization to obtain high enzyme activity.
ACKNOWLEDGMENT
We thank Conselho Nacional de Desenvolvimento Científico e Tecnológico for fellowships and São Paulo State University-Unesp for financial support.