Introduction

Inulinase (β-2,1-d-fructan fructanohydrolase, EC 3.2.1.7) targets on the β-2,1 linkage of inulin (a polyfructan consisting of linear β-2,1 linked fructose) and hydrolyzes it into fructose, producing fructose syrup in a single step (Wei et al. 1998).

Fructose production by inulin hydrolysis is more advantageous than conventional process based on starch, which includes the action of α-amylase, amyloglucosidase, and glucose isomerase, yielding only 45% of fructose in the final product, due to the thermodynamical equilibrium of the reaction. Inulinase based hydrolysis of inulin can yield products with 95% of fructose (Mazutti et al. 2006).

However, a more potential application of such enzymes in food industry is the production of fructoligosaccharides (FOS), by the transfructosilation of sucrose catalyzed by inulinase or by β-fructosiltransferase. The FOS present many interesting functional properties, besides their ability to increase shelf-life and flavor of many products (Sangeetha et al. 2005). Nevertheless, the industrial application of inulinase will only be viable when this enzyme is available in large quantities at competitive market prices. Hence, it is critical to define the most adequate microorganism and cultivation medium, as well as to specify the stirring rate, pH, temperature, and aeration conditions for feasible scale-ups of fermentation process.

The use of industrial media is a good alternative to produce inulinase at low cost, if the enzyme activity is improved or if the activity is at least the same obtained in a synthetic medium. The production of inulinase using sugarcane molasses (SCM) and corn steep liquor (CSL) may be an interesting alternative to add value to such residues and lower enzyme production costs. Inulinase can be produced by solid state fermentation using untreated sugarcane bagasse, CSL and SCM as substrates, obtaining maximum inulinase activities of 390 and 444 U g−1, respectively (Mazutti et al. 2006; Bender et al. 2006). Similar results were obtained using wheat bran with different supplements (Xiong et al. 2007).

On the other hand, the composition of industrial media based in agro-industrial residues is very complex and the presence of heavy metals (iron, zinc, magnesium, calcium, copper, manganese) may inhibit inulinase production (Lazarridou et al. 2002; Roukas 1998). For instance, SCM contains a large amount of melanin, which is toxic to microorganisms due to its antioxidant properties (Lee et al. 2000). Thus, a study of the upstream processing for removal of contaminants of the fermentation media for enzyme production may contribute to increase enzyme yields and even simplification of the downstream steps, which usually represents 50 to 80% of product final costs.

The present work aimed to study the pretreatment of agro-industrial residues used in the production of inulinase. Corn steep liquor (CSL) and sugarcane molasses (SCM) were pretreated to improve enzyme yields. Experimental design and response surface methodology were used to optimize the conditions of acid treatment of substrates and the formulation of fermentation medium.

Materials and Methods

Pretreatment

Sugarcane molasses (SCM) was obtained from a sugar refinery (Ester Sugar Refinery, Campinas SP, Brazil) and corn steep liquor (CSL) was obtained from Corn Products Brazil.

Preliminary tests indicated that treatment with phosphoric acid and sulfuric acid were the most effective methods for clarification of CSL and SCM, respectively. The preliminary results helped to define experimental designs to study the clarification of substrates. A complete factorial design with triplicate at the central point was used to study SCM clarification using sulfuric acid. A Central Composite Rotatable Design (CCRD) was used for optimization of CSL clarification by phosphoric acid. Table 1 present the investigated levels in both designs. The studied factors included initial pH of treatment, resting time and final pH.

Table 1 Range of factors for molasses and CSL clarification
Full size table

At each experimental run, the pH of the substrate was adjusted with the respective acid and the solution was set to rest at 24°C for the time defined by the experimental design. The media was then centrifuged at 5,000×g for 15 min, and the final pH was adjusted with NaOH 10 M until the pH defined in the experimental design. The clarification of the medium was quantified by optical density measurements at 600 nm in a spectrophotometer (Agilent 8453). The analysis of data was performed using software Statistica® 5.1 (Statsoft).

Microorganism and Medium

The strain of Kluyveromyces marxianus NRRL Y-7571 was employed for inulinase production. The microorganism was grown on YM agar medium containing (g l−1): yeast extract 3.0, malt extract 3.0, peptone 5.0, glucose 10.0, agar 20.0, and was subcultured every 3 weeks.

Cell production for the preinoculum cultures was carried out in a medium containing 2% sucrose at pH 6.8 in 500 ml flasks with 100 ml of culture medium, at 30°C, 150 rpm for 24 h.

Submerged Fermentation

The inulinase production was carried out in 500 ml Erlenmeyer flasks with 100 ml pretreated culture medium. The fermentations started with 10% (v/v) inoculum, at initial pH 5.0, 36°C, and 150 rpm rotation in orbital shaker PSYCROTHERM (New Brunswick Scientific, NJ, USA). The activity was determined after 72 h of fermentation.

The composition of the medium was optimized for highest inulinase production. The optimization was performed using a Central Composite Rotatable Design (CCRD), with three replications on central point. Table 1 shows the range of factors investigated in this step. The analysis of data was performed using software Statistica® 5.1 (Statsoft).

Inulinase Assay

Activity was assayed as follows: 1 ml enzyme solution was mixed with 9 ml of 2% (m/m) sucrose or inulin on acetate buffer 0.1 M pH 4.6. The mixture was maintained at 50°C and the reducing sugars were determined by the 3,5-dinitrosalicylic acid method (DNS) (Miller 1959). One unit of inulinase activity is defined as the amount of enzyme necessary to hydrolyze 1 μmol of sucrose per minute under the conditions mentioned above (sucrose as substrate).

Microbial Growth

The microbial growth was determined by an indirect dry weight method, by comparing the cell suspensions optical density at 600 nm to standard curve, which was previously prepared with the same microorganism produced in the same culture medium (Silva-Santisteban and Maugeri 2005).

Results and Discussion

Pretreatment

Table 2 presents the results of the complete factorial design used for definition of the pretreatment of molasses with sulfuric acid. Optical density (OD) of molasses solution without the pretreatment was 0.749. The results show that sulfuric acid treatment is effective for color removal of molasses, since a decrease in the OD was detected in all runs. The condition in run 2 presented the highest removal of color (70%). The clarification action of sulfuric acid is possibly related to oxidation of suspended organic matter in the medium, since sulfuric acid is known as an oxidizing and dehydrating agent (Atkins and Jones 2005).

Table 2 Matrix of the experimental design (coded and real values) with responses in terms of optical density (OD) for clarification of molasses
Full size table

The results in Table 2 were statistically analyzed, yielding an empirical model as a function of the resting time and the final pH of the medium. The model was validated by analysis of variance (ANOVA) presented in Table 3. The factors that were not significant (p < 0.05) were added to the lack of fit for the analysis of variance (ANOVA). The ANOVA shows a high R (0.87, correlation coefficient) and a good performance of the F test for regression (calculated value about six times the listed one). Therefore, Eq. 1 is predictive of SCM clarification in the investigated range of factors, and consists in a first order function of resting time (t) and final pH (pHfinal).

$${\text{OD = 0}}{\text{.49 + 0}}{\text{.14 }} \times {\text{ }}t{\text{ + 0}}{\text{.09 }} \times {\text{ pH}}_{{\text{final}}} $$
(1)
Table 3 ANOVA for molasses clarification as response
Full size table

Equation 1 shows that the initial pH does not influence clarification of molasses in the investigated range (3.0–5.0). Clarification is then a function of resting time and final pH. The final pH of treatment and resting time presented positive main effects on optical density. This means that an increase on these factors would result in a lower clarification of the medium. Therefore, these factors were set to the minimum level in medium optimization. A Central Composite Rotatable Design (CCRD) was unnecessary in this case because a sufficient color removal was obtained in this design. (Rodrigues and Iemma 2005).

Table 4 presents the matrix of the results of pretreatment of CSL with phosphoric acid. Optical density of the medium without treatment was 0.835. For runs 4, 7, and 8, an increase in OD is observed, showing that these conditions were inadequate for clarification of CSL. Good color removal (70%) was obtained at runs 15–17, which correspond to the central point of the experimental design, i.e., initial pH 3.0, 24 h of resting time, final pH 5.5.

Table 4 Matrix of the experimental design (coded and real values) with responses in terms of optical density (OD) for clarification of CSL
Full size table

The results presented in Table 4 were submitted to statistical analysis, yielding a second order empirical model (Eq. 2). The model was validated by ANOVA, presented in Table 5. The factors that were not significant (p < 0.05) were added to the lack of fit for the analysis of variance (ANOVA). The ANOVA shows a high R (0.87, correlation coefficient) and a good performance of the F test for regression (calculated values about 1 times the listed one). Equation 2 is predictive for clarification of CSL in the range investigated and explains 87.0% of total variation of the experiments.

$${\text{OD}} = 0.23 + 0.07 \times {\text{pH}}_{{\text{initial}}} + 0.11 \times {\text{pH}}_{{\text{initial}}}^{\text{2}} 0.14 \times t + 0.13 \times t^2 + 0.03 \times {\text{pH}}_{{\text{final}}} + 0.15 \times {\text{pH}}_{{\text{final}}}^{\text{2}} + 0.03 \times t \times {\text{pH}}_{{\text{final}}} $$
(2)
Table 5 ANOVA for CSL clarification as response
Full size table

The fitted response surface for the clarification of CSL obtained by Eq. 2 is presented in Fig. 1. The results show that the clarification of CSL is optimized, since the minimum OD was obtained at the treatment conditions of the central point of the CCRD.

Fig. 1
figure 1

Response surface for clarification of CSL as a function of resting time and initial pH

Full size image

Many studies dealing with pretreatment of molasses are reported in literature. Some reports showed that only 80% of the glucose present in the molasses could be used by the microorganism in the medium without pretreatment, while 99% was consumed in the medium that was filtered with membrane (Kasseno and Kokugan 1997). The optimization of the production of xanthan gum using molasses as substrate, when the medium was formulated with molasses without any treatment xanthan resulted in a production of 46.9 g L−1 in 48 h of fermentation. For a medium containing molasses treated with sulfuric acid at pH 4.0, production was 36.9 g L−1. The pretreatment did not show a significant effect (Kalogiannis et al. 2003). Treatment with sulfuric acid and activated carbon for removal of heavy metals and colored compounds from molasses can be used for production of pullulan (Lazarridou et al. 2002).

Optimization of Inulinase Production

Table 6 presents the results of the CCRD used to study the influence of concentration of SCM, CSL and Prodex Lac on inulinase production. Experimental runs 10 and 13 showed the lowest enzyme productions, yielding activities 178 and 156.9 U ml−1, respectively. The highest inulinase yield was found at the central point of the CCRD (1,139 ± 111 U ml−1), showing that the experimental design was effective to optimize the process conditions. In the present work, a single CCRD was enough to optimize of inulinase production using the following concentration of pretreated substrates: 100 g l−1 CSL, 100 g l−1 SCM, and 6 g l−1 of Prodex Lac. In this condition, the activity of the enzyme on inulin was determined. The ratio between the activity on sucrose and the activity on inulin (S/I) was lower than 50, characterizing this enzyme as an inulinase rather than an invertase (Ettalibi and Baratti 1987).

Table 6 Matrix of the experimental design (coded and real values) for inulinase production (72 h of fermentation)
Full size table

The results show a large variability, suggesting that the concentration of such substrates present a great influence in the enzyme production, probably due to their complex composition of nutrients. A detailed composition of CSL and SCM before and after pretreatment with activated carbon is presented in literature (Valduga et al. 2007). Even after the treatment, residual metal cations were detected, like: zinc, manganese, copper, iron, magnesium, potassium, phosphorous, which can act as inhibitors above certain limit concentrations.

The inulinase activity obtained in the present work is the highest reported in literature so far. Studies about the effect of stirring rate, aeration, and the type of impeller in the production of inulinase found a maximum inulinase activity of 176 U ml−1 (Silva-Santisteban and Maugeri 2005). The use of synthetic media based in sucrose for inulinase production by K. marxianus can yield maximum inulinase production of 127 U ml−1 (Kalil et al. 2001). The use of inulin as carbon source and K. marxianus ATCC 52466 resulted in 470 U l−1 of inulinase (Selvakumar and Pandey 1999).

In the present work, production of 1,139 ± 111 U ml−1 shows the feasibility to produce inulinase using low cost substrates as SCM, CSL, and yeast hydrolysate.

The results presented in Table 6 were submitted to statistical analysis, yielding a second order empirical model of activity as function of concentration of molasses (SCM), corn steep liquor (CSL) and Prodex Lac (PL). The model was validated by ANOVA, presented in Table 7. The factors that were not significant (p < 0.05) were added to the lack of fit for the analysis of variance (ANOVA).The ANOVA shows a high R (0.93, correlation coefficient) and a good performance of the F test for regression (calculated values about 2 times the listed one). Therefore, Eq. 3 is predictive for inulinase production in the investigated range of factors.

$${\text{Activity}} = 1116.2 - 149.4 \times M - 175.4 \times M^2 - 185.5 \times CSL^2 - 229.3 \times PL^2 $$
(3)
Table 7 ANOVA for inulinase production as response
Full size table

Equation 3 shows that the three factors studied present second order significant (p < 0.05) parameters. This means that there is an optimum point for inulinase production as a function of concentration of sugarcane molasses, corn steeo liquor and Prodex Lac. Maximum inulinase production was obtained at 100 g l−1 SCM, 100 g l−1 CSL, and 6.0 g l−1 of Prodex Lac. Different media compositions yielded lower inulinase activities.

The relative difference between the predicted and the experimental responses was not higher than 10% in all experimental designs. This is a reasonable result for bioproductions using residues.

Figure 2 shows the kinetic behavior of inulinase production in the optimized fermentation condition. The pH of the medium was kept nearly constant during fermentation. Maximum growth occurred at 24 h, and the biomass decreased during the rest of fermentation run. A steep decrease in sugar concentration is observed in the first 12 h of fermentation. It is interesting to note that inulinase production is more evident after 12 h and keeps increasing until 72 h, when maximum production is observed. The results then suggest that inulinase production is partially associated to growth, since the highest rate of production occurred when microbial growth is negligible.

Fig. 2
figure 2

Kinetics of inulinase production at optimum conditions (100 g l−1 SCM, 100 g l−1 CSL, and 6.0 g l−1 of Prodex Lac). TRS Total reducing sugars

Full size image

Conclusions

The results presented in this work show the technical feasibility of inulinase production using low cost agro-industrial residues, after some pretreatment. Maximum inulinase activity obtained was 1139 U ml−1, which is the highest inulinase yield ever reported so far. The treatment with sulfuric acid was effective in the clarification of sugarcane molasses (SCM) and treatment with phosphoric acid was effective for clarification of corn steep liquor (CSL). The acid treatment was able to reduce up to 80% of color of both substrates. This is an important result, since this upstream step can simplify the steps of recovery and purification of the enzyme, which may contribute from 50 to 80% of total enzyme production costs.