Optimization of emulsifying activity and bioemulsifier production
To evaluate the effect of three independent variables on the emulsifying activity and production of the bioemulsifier by T. mycotoxinivorans CLA2, a CCRD methodology was employed to evaluate the coefficients in a quadratic mathematical model. RSM was used to calculate the maximum production based on a few sets of experiments in which all the factors were varied within chosen ranges.
Under the conditions used in the CCRD, the bioemulsifier production ranged from 0
g/L to 7.077
mg/ml, and the emulsifying activity ranged from 0 to 82%. The central points for both responses revealed only small variations, indicating good reproducibility of the process. Table
presents the CCRD matrix with the levels of variables chosen for trials in the central composite design. Twenty experimental runs with different combinations of three factors were performed for each of the three repetitions. Table
also shows the observed and expected values of bioemulsifier production and emulsifying activity. The expected values were obtained using the polynomial equations model, based on regression coefficient findings.
Experimental design and results of the CCRD matrix with observed and expected values of the bioemulsifier production and emulsifying activity
Through multiple regression analysis of the experimental data, a second-order polynomial equation was obtained for bioemulsifier production (Equation
1) and emulsifying activity (Equation
The statistical significance of Equations
1 and 2 was verified by the F test and by analysis of variance of the quadratic model of the RSM and is shown in Table
. At a 95% confidence level, the models for bioemulsifier production and emulsifying activity present a calculated F-value higher than the table value
], showing that the regression models were significant.
Analysis of variance (ANOVA) for the response of bioemulsifier production (Y1)aand emulsifying activity ( Y 2)bof T. mycotoxinivorans cultivated in medium containing biodiesel refinery residue
The significance of the coefficients of the full second-order polynomial model for bioemulsifier production and emulsifying activity were evaluated by Student’st
-test and the p
). The p
-values lower than 0.05 indicate that there is a significant correlation between the coefficients. The statistical significance of Eq. (1) and Eq. (2) was checked by the F-test, and the analysis of variance (ANOVA) for the response surface quadratic model is shown in Table
For both analyses, the X1 factor (yeast extract) was not significant. On the other hand, the X2 (biodiesel residue) and X3 ((NH4)2SO4) factors were highly significant. For bioemulsifier production, the X2 * X3 interaction was significant, exhibiting a synergistic effect on bioemulsifier production. The coefficient of the linear term X2 was negative, which means that if there were no effects of the X2 * X3 interaction, increases in the levels of X2 would result in decreased bioemulsifier production.
For the emulsifying activity, it was found that the second order parameters of the model represented by X1, X2 and X3, the coefficients of the linear terms, were all positive, meaning that a higher emulsifying activity should be obtained by increasing the levels of each of the factors. However, because the coefficients of the quadratic terms are all negative, increases in the levels of the factors will also tend to decrease the response. This indicates that this model reached the optimum region of emulsifying activity.
The coefficient of variation (CV) was reasonable for both models (31.45 for bioemulsifier production; 23.87 for emulsifying activity), indicating good precision and reliability of the experiments. The precision of the models was determined by the determination coefficient (R2). The R2 value implies that the sample variation of 83.18% for bioemulsifier production and 68% for emulsifying activity were attributed to the independent variables, and only about 16.82% and 32%, respectively, of the total variation cannot be explained by the models.
The response surfaces plots described by the regression models are presented in Figures
a. The figures provide a surface three-dimensional visualization of the estimated trend for the variation of bioemulsifier production and emulsifying activity, respectively, by T. mycotoxinivorans
CLA2, with different concentrations of biodiesel residue and ammonium sulfate. Although Figure
a shows that the optimal region of the response has not yet been obtained, it indicates the parameters necessary for optimization. Figure
b facilitates the identification of the maximum bioemulsifier production point within the range studied, located at a level delimited by 60.5 and 75
g/L biodiesel residue and 4.4 and 5
g/L ammonium sulfate. The estimated maximum bioemulsifier production occurred at 75
g/L biodiesel residue and 5
g/L ammonium sulfate. Under these conditions, the estimated bioemulsifier production was 9.525
a-b. Response surface plot and contour plot of the combined effects of residue biodiesel and ammonium sulfate on the bioemulsifier production by T. mycotoxinivorans CLA2 in a constant cultivation time (24 hours).
a-b. Response surface plot and contour plot of the combined effects of residue biodiesel and ammonium sulfate on the emulsifying activity by T. mycotoxinivorans CLA2 in a constant cultivation time (24 hours).
a provides evidence that the model encompassed the optimum region for emulsifying activity, located at the surface peak. Figure
b shows that the range of the estimated highest emulsifying activity was located at a level delimited by 40 and 75
g/L biodiesel residue and 3.4 and 5
g/L ammonium sulfate. The estimated maximum emulsifying activity occurred at 58
g/L biodiesel residue and 4.6
g/L ammonium sulfate. Under these conditions, the estimated emulsifying activity was 85% (E24
In the contour plots (Figures
b), it is possible to visualize whether the mutual interactions between the independent variables are significant or not. Clearly, the fitted model for bioemulsifier production presents interactions between independent variables, due an elliptical nature of the contour plot.
RSM is the most accepted statistical analysis for bioprocess optimization, allowing the relationship between a set of experimental factors and observed results to be examined
]. With regard to the production of microbial SACs as biosurfactants, the RSM has been successfully employed to optimize the production of these compounds from waste agricultural industries and to reduce the amount of organic material to be released in effluents
]. Furthermore, other products, such as cheese, whey and molasses, are added to microbial growth media as an alternative to reduce the cost of producing microbial SACs using probiotic bacteria
The residue derived from the production processes and purification of biodiesel consisting of organic compounds impregnated in diatomaceous earth was successfully employed to produce bioemulsifiers by T. mycotoxinivorans
CLA2. The residue used has an organic matter concentration of 24.78%, composed mainly of fatty acids methyl esters and glycerol byproducts, which is capable of supporting the growth of microbial cells, without adding other carbon sources. In this case, the reduction of bioemulsifier production costs by employing biodiesel processing waste would be an advantage for the production of these compounds by yeasts. Due to concerns about the disposition of the diatomaceous earth adsorbents filters and because some landfills are refusing this waste due to the concern of spontaneous combustion
], the microbial degradation of the compounds adsorbed in filters is advantageous and of great interest. It can eliminate the dangers of explosion and environmental contamination.
The described metabolic potential of the Trichosporon
species has made the application of these yeasts for environmental purposes very promising
]. Recently, a strain of T. mycotoxinivorans
was described as promising for the detoxification of effluents containing mycotoxins, especially the cleavage of zearalenone
]. Additionally, strains of the genus Trichosporon
are capable of producing bioemulsifiers from the metabolism of sunflower oil, which is added as the sole carbon source for cellular growth
]. To our knowledge, this is the first report on bioemulsifier production by T. mycotoxinivorans
from growth medium supplemented with residue from the processing of biodiesel.
Emulsifying activity of the bioemulsifier using hydrophobic substrates
In addition to the measure of surface tension, the stabilization of an oil/water emulsion is commonly used as surface activity indicator. The emulsifier specificity (5
mg/mL) produced by T. mycotoxinivorans
CLA2 was assayed with various hydrophobic substrates (Figure
). A high emulsifying activity value, approximately around 75%, was obtained using xylene. The emulsifying activities for toluene, hexadecane, and kerosene were approximately 71.7%. Statistical analysis of the test data by the Duncan test with 5% probability showed no significant differences between their values. Nevertheless, significant differences were observed between the values of emulsifying activity obtained for hexane, octane, cyclohexane, and gasoline.
Emulsifying activity (E24) of the crude bioemulsifier produced by T. mycotoxinivorans CLA2 using different hydrophobic substrates.
Several studies have shown that biosurfactants and bioemulsifiers may vary in their ability to emulsify different hydrophobic compounds. It was suggested that the emulsifying activity depends on the bioemulsifier’s affinity for hydrocarbons substrates, which involves a direct interaction with the hydrocarbon itself, rather than an effect on the surface tension of the medium
]. Poor emulsification of some hydrocarbons might be due to the inability of the bioemulsifier to stabilize the microscopic droplets. The broad-spectrum emulsifying activity is essential for the use of a bioemulsifier in industrial processes, such as the treatment of industrial effluents, washing of oily deposits and pumping of heavy oils, considering that the processes have different mixtures of hydrophobic compounds.
The bioemulsifier produced by T. mycotoxinivorans
CLA2 from biodiesel residue appears to be a very effective and efficient emulsifier for aromatic and aliphatic hydrocarbons. Emulsan, one of the most effective emulsifiers, stabilizes hydrocarbon emulsions in water emulsions (the percentage of hydrocarbons vary from 0.01-0.10%) at low concentrations (0.02 to 0.2
mg/mL), and it exhibits considerable substrate specificity but does not emulsify pure aliphatic, aromatic or cyclic hydrocarbons. Nevertheless, all mixtures that contain an appropriate mixture of aliphatic and aromatic (or cyclic alkane) residues are efficiently emulsified
Purification and characterization of the bioemulsifier
Purification of the bioemulsifier produced by T. mycotoxinivorans CLA2 was performed by gel filtration chromatography. Fractionation of the bioemulsifier resulted in only one peak, with the resulting fractions being positive for total carbohydrate and emulsifying activity. The final purified yield of the bioemulsifier, based on the recovered mass fractionation, was 90% (data not shown).
The analysis of the fatty acid methyl esters of the lipid portion of the bioemulsifier indicated the presence of octadecanoic acid, hexadecanoic acid and 9-octadecenoic acid (Z) at concentrations of 48.01%, 43.16% and 8.83%, respectively. The content of fatty acids of this bioemulsifier resembles that of a high-molecular-weight bioemulsifier produced by bacteria and yeast, in which the highest percentage of fatty acids contain chains with 12, 16 and 18 carbons.
] studying the bioemulsifier Yansan, observed that the fatty acid content corresponds mainly to fatty acids of 16 and 18 carbons, such as hexadecanoic acid (35. 8%) and octadecanoic acid (21.4%). However, the number of carbons in fatty acid chains of the bioemulsifier Emulsan produced by Acinetobacter calcoaceticus
RAG-1 seems to be dependent on the type and combination of hydroxy and hydroxylated fatty acids used for microbial growth. Therefore, different structures of the bioemulsifier can be obtained by varying the type of the precursor
Monosaccharides in the bioemulsifier were identified by GC–MS analysis of their alditol acetate derivatives. The analysis showed the presence of xylose (49.27%), mannose (39.91%) and glucose (10.81%). The composition of monosaccharides in the bioemulsifier from CLA2 is similar to that described for the bioemulsifier produced by T. loubieri
CLV20, which contains mannose and glucose at concentrations of 41% and 8%, respectively
]. Additionally, this composition is similar to that described for a biosurfactant produced by Y. lipolytica
IMUFRJ 50682, which showed a predominance of mannose
The high-molecular-weight SACs are not effective in reducing surface and interfacial tension of liquids but are very effective in stabilizing oil in water emulsions
]. Generally, this hydrophobic portion is required for emulsification, such as in the emulsifiers produced by Acinetobacter calcoaceticus
BD4 and A. calcoaceticus
]. Similarly, the SACs produced by the yeast T. mycotoxinivorans
CLA2 exhibited the ability to stabilize emulsions with kerosene as the organic phase. However, the bioemulsifier is not capable in significantly reducing the surface tension of water (data not shown).
Characterization of the bioemulsifier by nuclear magnetic resonance (1
The results obtained with a 400
H NMR spectrum of the bioemulsifier molecule in D2
O are shown in Figure
. The high overlap of the signals only allows characteristic regions of chemical shifts of the hydrogen signals in saccharide groups to be identified. The 1
H NMR signals related to the anomeric hydrogens of the saccharide residues present in the polymer molecule appear at δ 4.6 and δ 5.5
ppm, with higher intensities at δ 5.05
ppm, indicating a higher proportion of this sugar in the polymer chain. According to the literature, anomeric hydrogens of saccharides present chemical shifts between δ 4.5 and δ 6.0
ppm (xylose - δ 4.4
ppm; glucose - δ 4.8
ppm (β) - δ 5.23
ppm (α) and mannose - δ 4.89
ppm (β)). In the positions between δ 3.0 and δ 4.5
ppm, overlapping signals from the other hydrogens of the sugars present in the polymer chain are observed; a low resolution spectrum is not observed at the coupling constants between the hydrogens of this part of the molecule. The singlet at δ 2.2 corresponds to the resonance signal of long-chain CH2
groups. The integration of these signals indicates that equivalent hydrogens are present in a relative proportion of 4:1, with respect to the anomeric hydrogen located at δ 5.23, assigned to the glucose residue. Signals are also observed due to the resonances of protons belonging to methyl groups of the polymer chain, confirming the results of chromatography. The presence of a spot indicative of correlation of scalar coupling in the 2D-COSY spectrum confirms the link from the polymer sample with its glycosidic portion (Figure
). This result is consistent with the literature data that suggests that glycolipids or hetero-polysaccharide bioemulsifiers contain covalently linked hydrophobic side chains
]. The low resolution and the great superposition of the signals observed in 1
H NMR spectra and confirmed by the COSY spectra suggest a high-molecular-weight bioemulsifier molecule with a polymeric pattern containing sugars and long chain aliphatic groups
Figure 4 1H nuclear magnetic resonance (NMR) spectrum of the bioemulsifier produced by T. mycotoxinivorans CLA2 at 27°C (presaturation, 400MHz, D2O).
Figure 5 COSY contour map of the bioemulsifier produced by T. mycotoxinivorans CLA2 at 27°C (400MHz, D2O). Scalar correlations between the polysaccharides group of the molecule and the hydrophobic side chains are indicated.