Introduction

The use of agroindustrial residues for the generation of value-added products is one of the key challenges facing the transition to a sustainable bioeconomy. The fruit juice industry alone generates 5.5 million metric tons (MMT) of underutilized bagasse and pulp every year [1]. These residues are rich in polysaccharides and represent a potentially valuable resource, which could be enzymatically converted into sugars and oligosaccharides to generate a multitude of products within the biorefinery concept. Pectinolytic enzymes, or pectinases, designate a heterogeneous group involved in the degradation of pectin, a complex polysaccharide that is present in the cell wall and middle lamella of all land plants [2, 3]. Pectinases include pectate lyases (PL), polygalacturonases (PG), and pectinesterases (PE), and are used in a variety of applications, such as in the wine, food, and paper industries, and also in animal feed production, extraction of vegetable oil, textile, and for the depolymerization of lignocellulosic biomass for biofuel production [4, 5]. Due to their widespread application, pectinases occupy a leading position among commercially produced industrial enzymes with a 25% share of the global food and beverage enzyme market [4, 5].

Bacteria, fungi, plants, and yeast are the natural sources of pectinases and more than 30 distinct genera have been described as producers of these enzymes [4, 5]. However, the industrial application of pectinases is limited by low productivity, enzyme instability, high cost, and difficulty with enzyme recovery [6, 7]. Many of these limitations can be overcome by the use of immobilized enzymes. Enzyme immobilization is a powerful strategy in biocatalyst design [8] that allows the re-use of a relatively expensive catalyst and improved reaction control since the catalytic process can be interrupted by separating the immobilized enzyme from the reaction medium [9, 10]. Additionally, immobilization can improve the catalytic performance of an enzyme by altering properties such as specific activity, stability, selectivity, and decreased inhibition by reaction products [9]. There are several immobilization techniques, such as simple adsorption on porous substrates, weak chemical interactions (ion exchange and hydrophobic interaction, for example), and immobilization by one-point or multi-point covalent bonding [11]. Recently, the immobilization of pectinolytic enzymes using different methods including adsorption, entrapment, and covalent binding has been reported [6, 7, 12,13,14]

Covalent immobilization is among the most widely used strategies, with a frequent outcome of reduced enzyme losses and, therefore, long-term process stability. The use of magnetic nanoparticles has attracted attention in recent years due to the ease of recovery of the immobilized enzyme using an externally applied magnetic field and their high surface area that enables a high loading of enzymes per unit mass of support [6, 12, 15, 16]. We have previously demonstrated that glutaraldehyde-activated chitosan ferromagnetic nanoparticles are a promising low-cost alternative for enzyme immobilization. These nanoparticles were coated with chitosan and treated with the cross-linking agent glutaraldehyde that functionalizes the chitosan, stabilizing the nanoparticle coating and providing a chemically reactive surface for enzyme immobilization [17].

In this work, we have immobilized an endopolygalacturonase from Chondrostereum purpureum (EndoPGCp) expressed in Pichia pastoris [2] on glutaraldehyde-activated chitosan ferromagnetic nanoparticles, and have compared the catalytic properties of the free enzyme with the immobilized enzyme. In addition, we report the capacity of the immobilized EndoPGCp to generate sugars from agricultural residues derived from the fruit processing industry.

Material and Methods

EndoPGCp Production and Purification

The recombinant pectinase from C. purpureum was overexpressed in Pichia pastoris GS115 as previously described [2]. Briefly, a single colony of P. pastoris GS115 transformed with the pPIC9Kf1CT_endopg plasmid was inoculated into 100 mL of MG medium (3.4 g/L yeast nitrogen base without amino acids and ammonium sulfate, 10 g/L ammonium sulfate, 0.0004 g/L biotin, 1% (v/v) glycerol, 100 mmol/L potassium phosphate buffer, pH 6.0). The culture was incubated at 30 °C for 16 h in an orbital shaker at 250 rpm. Cells were harvested by centrifugation at 5000 g for 5 min, and the cell pellet was resuspended to an OD600 nm of 1 in 100 mL MM medium (MG medium in which glycerol was substituted with 1% (v/v) methanol) and the culture was grown for 3 days at 30 °C. Every 24 h, methanol was added to a final concentration of 1% and the pH was corrected by the addition of 1 mL of 1 mol/L potassium phosphate buffer, pH 8.0. After 3 days, the cells were centrifuged at 5000 g for 20 min and the recombinant EndoPGCp was purified from the culture supernatant by nickel affinity chromatography.

Estimation of Protein Concentration

The protein concentrations were determined according to Read and Northcote [18] using bovine serum albumin as standard.

Preparation of Ferromagnetic Nanoparticles (MNP)

The glutaraldehyde-treated magnetic nanoparticles were prepared using a protocol adapted from that previously described by Carneiro and Ward [17]. Briefly, the Fe3O4 ferromagnetic nanoparticles were prepared after controlled co-precipitation of Fe2+ and Fe3+ under alkaline conditions in the presence of Pluronic F127, a non-ionic surfactant that prevents spontaneous aggregation of the newly formed particles. The particles were then coated with chitosan 0.1% dissolved in acetic acid and the mixture was stirred for 30 min at room temperature. The MNPs coated with chitosan were treated with the cross-linking agent glutaraldehyde. The initial reaction was performed for 1 h at room temperature and the particle suspension was subsequently incubated at 4 °C for 16 h. The nanoparticles were separated with a neodymium magnet and repeatedly washed with water to remove excess reagent to remove excess glutaraldehyde. The glutaraldehyde-activated chitosan MNPs were stored in water at 4 °C until further use.

Enzyme Immobilization

EndoPGCp was immobilized on MNP by covalent bonding to produce EndoPGCp-MNP. Concentrations (Ci) of 50, 100, and 200 μg of pure EndoPGCp were prepared in 100 mmol/L sodium phosphate buffer; pH 8.0 in a final volume of 100 μL was incubated with 100 μL of a MNP suspension (10% w/v) at 4 °C for 16 h. The MNP were washed 4 times with 1.0 mL of wash buffer (50 mmol/L sodium phosphate, pH 8.0) to remove non-adsorbed proteins using a neodymium magnet (11-mm diameter, 3-mm thickness). The supernatants from the initial immobilization (S) and the subsequent washes (W1, W2, W3, and W4) were collected for protein quantifications. The amount of protein retained on the support was estimated using Eq. (1):

$$Immobilized protein content (\mu g) = {C}_{i} - (S + {W}_{1} + {W}_{2} + {W}_{3}+ {W}_{4})$$
(1)

where Ci, S, and W1-4 were the amount of protein measured in the initial solution and in the supernatants from the immobilization and washing steps, respectively.

The resultant EndoPGCp-MNP were resuspended in 200 μL of the wash buffer and stored at 10 °C. The immobilization yields and efficiencies of the immobilized enzymes were calculated according to Eqs. (2) and (3), respectively [15]. All immobilization tests were performed in triplicate.

$$\frac{Amount of protein retained on the support (mg protein)}{\mathrm{Amount of offered protein to immobilization }(\mathrm{mg protein})}\times 100$$
(2)
$$\frac{Specific of immobilized enzyme (U/mg)}{\mathrm{Specific activity of free enzyme }(\mathrm{U}/\mathrm{mg})}\times 100$$
(2)

Scanning Electron Microscopy (SEM)

Analysis of the surface topology characteristics of the magnetic nanoparticles both in the absence and presence of immobilized EndoPGCp was performed by scanning electron microscopy (SEM). A 5-μL drop of the particle suspension at 0.5% (w/v) was applied to a glass microscope slide and dried under vacuum prior analysis, and images were obtained using a JEOL model JSM 6610LV instrument (JEOL, Akishima, Tokyo, Japan) operated at 30 kV and at 1 × 104 and 2 × 104 magnification.

Determination of the Free and Immobilized Endopolygalacturonase Activity

Unless otherwise stated, the activities of the free and immobilized EndoPGCp were determined at 60 °C in McIlvaine buffer [19], pH 4.5, containing 0.5% (w/v) citrus pectin (Sigma-Aldrich Chem. Co.) as substrate in a final volume of 60 μL, as previously described [2]. The total released reducing sugars were quantified using the dinitrosalicylic acid (DNS) method [20]. The experimental conditions (reaction times, enzymatic units) of all enzymatic assays were adjusted to guarantee the reliable estimation of initial velocities. One unit of enzyme activity (U) was defined as the amount of enzyme that catalyzes the formation of 1 μmol of galacturonic acid per min under optimal conditions and converted to a specific activity (U/mg protein). Controls with heat-inactivated enzyme were included in all enzymatic assays, and all the experiments were performed in triplicate.

Effect of temperature and pH on immobilized EndoPGCp

The effect of temperature on the activity of the free and immobilized EndoPGCp was determined in the range from 45 to 85 °C in McIlvaine buffer [19], pH 4.5. The activation energy (Ea) for the hydrolysis of pectin by free and immobilized EndoPGCp was determined using the Arrhenius equation following Segel [21]. The thermal stability of the enzymes was evaluated by pre-incubating the free and immobilized EndoPGCp diluted in Millipore MilliQ ultrapure water for 30 min at different temperatures (30–80 °C). After cooling in an ice bath for 1 min, the residual activities were estimated at 60 °C as described in the previous section. The effects of pH on the activity of EndoPGCp and EndoPGCp-MNP were determined at 60 °C in McIlvaine buffer [19] over the pH range 3.0–8.0.

Estimation of Kinetic Parameters and Data Fitting

Maximum velocity (Vmax) and Michaelis–Menten constant (KM) for the hydrolysis of citrus pectin (concentrations ranging from 0.1 to 10 mg/mL) by EndoPGCp-MNP were estimated by non-linear regression using the SigrafW software [22]. The protocols of expression, purification, and immobilization of EndoPGCp were repeated three times, resulting in three independent preparations of EndoPGCp-MNPs. The experimental kinetic curves were repeated three times using independent enzyme preparations, with duplicate enzyme assays for each preparation. The kinetic parameters are presented as the mean ± SD for the three different enzyme preparations (n = 3).

Evaluation of Reusability of the EndoPGCp-MNps

The reusability of the immobilized EndoPGCp was determined over 10 reaction/washing cycles using the activity assay as described in the section describing the enzyme activity assay. At the end of each cycle, the immobilized enzyme was magnetically separated from the reaction media, washed with 1.0 mL of wash buffer, and then added to a fresh reaction mixture to start a new reaction cycle. The reducing sugar concentration in the supernatant was assayed using the DNS method [20].

Supplementation of a Commercial Enzyme Cocktail by Free and Immobilized EndoPGCp for Hydrolysis of Processing Fruits

The passion fruit and orange residues were prepared from the fruit peels after the removal of the pulps. The peels were washed with distilled water and dried for 72 h at 50 °C and subsequently milled in a Wiley-type mill (SL 31 – SOLAB, Piracicaba – BRA) using a 30-mesh sieve. The powders were washed 2 times with ethanol 100% to remove soluble sugars, dried to remove the ethanol, and used in the hydrolysis experiments.

The hydrolysis of lignocellulosic residues was performed at 45 °C and 200 rpm in hermetically sealed conical plastic tubes containing 50 mmol/L sodium acetate buffer, pH 5.5, 10 mmol/L sodium azide, and 10 mg of the passion fruit or orange residue in a final volume of 1 mL. The reaction was initiated by the addition of an aliquot of either free or immobilized EndoPGCp, or a cellulase-rich commercial enzyme preparation (Celluclast 1.5 L), or of a mixture of both. At defined time intervals, aliquots of the reaction medium were centrifuged at 10,000 g for 5 min, and total reducing sugars were determined by the DNS method [20]. In all experiments, the enzyme loading was constant at 10 FPU/g substrates with respect to Celluclast 1.5 L and 50 U/g substrates with respect to EndoPGCp (free or immobilized). Each hydrolysis experiment was repeated three times using three independent EndoPGCp preparations.

Statistical Analysis of the Enzyme Activity Data

All data were expressed as the mean and standard deviation of at least three independent experiments.

Results and Discussion

MNP Synthesis and Enzyme Immobilization Efficiency and Yields

Optimizing the initial EndoPGCp concentration for immobilization suggested that the ideal enzyme loading was 100 μg of protein for 100 μL of a MNP suspension (10% w/v) (data not shown). The immobilization yield of the EndoPGCp on the glutaraldehyde-treated magnetic nanoparticle support was approximately 20% and the specific activity of the immobilized enzyme was similar to that observed for a free enzyme, giving an efficiency yield of approximately 100%. The yield of covalent immobilization using glutaraldehyde as a spacer arm for most pectinases is in the range of 65–98% [13, 23,24,25]. However, similar to that observed for EndoPGCp, the polygalacturonase from Aspergillus niger [26] and Aspergillus ustus [27] also showed a low immobilization yield, of 37 and 20%, respectively. Low immobilization yields can be associated with different factors such as initial protein concentration, type of support, and the glutaraldehyde concentration used for enzyme cross-linking to the support [28,29,30].

The immobilization efficiency is a measure of the change in specific activity of the enzyme after immobilization, and few pectinases showed efficiency above 80% [13, 27, 31]. Alagöz and collaborators [25] observed a 90% reduction in the specific activity of an Aspergillus aculeatus pectinase after covalent immobilization on glutaraldehyde-activated magnesium silicate. However, the same enzyme presented ~ 75% of the specific activity as compared to the free enzyme after covalent immobilization on nanosilica activated with glutaraldehyde, demonstrating the influence of the support on the enzymatic activity after immobilization. The maintenance of the catalytic activity and the stability of the enzyme after the immobilization process are crucial for the viability of the operation. In general, this is governed by the orientation of the enzyme on the support, which largely depends on the distribution of the amino acid residues that participate in the covalent bond between enzyme and support. Techniques using glutaraldehyde as a coupling agent are probably the most used to enzyme immobilization. They can react mainly with the protein primary amino groups and eventually thiols, phenols, and imidazoles, and this could allow the EndoPGCp-MNP to be fully functional after the process [32,33,34]. Mohamad et al. [35] suggested that the enzyme activity is preserved when residues at the active site are not involved in binding to the support, and this effect may explain the high efficiency of immobilization in the present study.

Morphological Characterization of EndoPGCp-MNP by SEM

Scanning electron microscopy showed the formation of aggregates in the MNPs both in the presence and absence of the immobilized enzyme (Supplementary Information (SI)). These results could be related to the tendency of the ferromagnetic nanoparticles to agglomerate spontaneously [17, 36]. The nanoparticle aggregation effect was also observed in SEM by Carneiro and Ward [17]. The resulting particle sizes are in the range of 2–10 μM and present a highly irregular and porous surface. Immobilization of the EndoPGCp has no discernable effect on the morphology of the aggregates. Andrade et al. [37] observed the formation of aggregated structures in magnetic nanoparticles and immobilized β-galactosidase, relating this with magnetism. Podrepšek et al. [38] reported an increase in particle size caused by chitosan coating compared to the bare nanoparticle (22.7 nm). They found coated particles with an average size of 68.5 μm using a microemulsion process, 44.2 μm in the cross-linking process, and 58.8 nm by the covalent binding method. Gennari et al. [39] showed that magnetic nanoparticles coated with nanocellulose presented an average diameter of 74 nm that is about 2 × larger than the non-coated nanoparticles.

Influence of pH and Temperature on the Activity and Stability of Free and Immobilized EndoPG

The effect of temperature on the free and immobilized EndoPGCp is presented in Fig. 1a. The EndoPGCp-MNP presented a similar profile to that previously described for free EndoPGCp [1], and in both cases, the catalytic activity gradually increased from 45 to 65 °C, reaching the maximum at 70 °C and decreasing abruptly at higher temperatures. Previous reports demonstrate that the enzyme immobilization process may have positive [7] or negative [12] effects on the optimal reaction temperature. However, maintained optimal reaction temperature is a common feature for pectinases that are covalently immobilized with glutaraldehyde [13, 23,24,25]. Indeed, a commercial pectinase was immobilized on magnetic nanoparticles coated with silica and functionalized with glutaraldehyde [40] also showed an optimum temperature identical to the free enzyme. However, for this same enzyme, the immobilization process favored activity at higher temperatures and the enzyme showed a 30% increase in activity at temperatures above the optimum temperature for the free enzyme [40]. These effects are influenced by the immobilization method, the nature of the enzyme, and the support and cross-linker used for immobilization [9, 28, 30, 41]. Analysis of the Arrhenius plots (Fig. 1b) of the hydrolysis of pectin by both free and immobilized EndoPGCp yield values for Ea of 17.3 ± 1.3 kJ/mol (free EndoPGCp) and 17.1 ± 1.1 kJ/mol (EndoPGCp-MNPs), demonstrating that the activation energy (Ea) was unchanged, confirming that the immobilization process had no effect on the temperature dependence of the catalytic reaction rate.

Fig. 1
figure 1

Catalytic properties of immobilized (blue) and free (red) EndoPGCp. a Influence of temperature on the enzymatic activity. b Arrhenius plots for the effect of temperature on the enzymatic activity. c Influence of temperature on the stability of the enzymes. d Influence of pH on the enzymatic activity. One hundred percent specific activity corresponded to 4890.4 ± 368.3 U/mg for free EndoPGCp and 5085.9 ± 217.1 U/mg for EndoPGCp-MNPs

Full size image

The EndoPGCp-MNP retained about 50% of its initial activity after incubation at 70 °C for 30 min and is more stable than the free enzyme, which presented about 25% of the initial activity under the same conditions (Fig. 1c). The covalent bonds between EndoPGCp and the MNP are likely to decrease the thermal motions of the polypeptide chain leaving the enzyme less susceptible to thermal denaturation. This effect is widely observed in enzymes that are covalently immobilized to the support [6, 13, 14, 42].

The activity of the free and immobilized pectinase in various pH at a constant temperature (60 °C) is presented in Fig. 1d. The maximum activity of the free pectinase was obtained at a pH of 4.5 [2], and after immobilization, the optimum pH was shifted to 4.0. Furthermore, at pH 5.5, approximately 75% of the maximum activity is maintained for the immobilized enzyme, whereas the free enzyme maintains only 50% activity. Equally subtle effects of immobilization on the optimal pH of catalysis have been observed with other pectinases [31, 43,44,45] and may be associated with changes in enzyme conformation or pH effects on the concentration of substrate, product, charged species, hydrogen, and hydroxyl ions both in the immediate microenvironment of the immobilized enzyme [9, 28, 30, 41].

Kinetic Parameters

The EndoPGCp-MNP hydrolyzed citrus pectin with a Vmax = 5250.60 ± 315.59 U/mg and KM = 2.93 ± 0.30 mg/mL, giving in a catalytic efficiency (kcat/KM) of 2178.84 ± 105.20 mL/mg s (Table 1). These values are similar to those previously observed for the free EndoPGCp [2], indicating that the conformational flexibility of pectinase was maintained after immobilization. Although immobilization may limit the diffusion of the substrate to the active site of the enzyme or alter the dynamics of the enzyme molecule resulting in a compromised catalytic cycle, such effects would alter the kinetic parameters, as has been observed in the increased KM values and decreased in Vmax of immobilized enzymes previously reported for other pectinases immobilized by different methods [6, 16, 46, 47].

Table 1 Kinetic parameters for pectin hydrolysis by free and immobilized EndoPG
Full size table

Application of EndoPGCp-MNP as Supplementation of a Commercial Enzyme Cocktail for Hydrolysis of Citrus Fruit Residues

Brazil is the largest producer of sweet orange and passion fruit in the world and is responsible for over three-quarters of global orange juice exports [48, 49]. Processing these fruits generates peel, seed, and bagasse by-products, and the re-use of these residues is a challenge for the food industry. These by-products are lignocellulosic materials rich in polysaccharides, including pectin, which could potentially also be utilized for the production of value-added bioproducts [50]. The application of immobilized EndoPGCp-MNP as a supplement to commercial enzymatic cocktails for the hydrolysis of by-products of the fruit processing industry was evaluated and compared with the results obtained using the free enzyme (Fig. 2). The total reducing sugar released by a commercial enzyme cocktail supplemented with EndoPGCp-MNP using orange residues (Fig. 2b) gradually increased over time and reached about 1.8-fold higher as compared to the hydrolysis by the unsupplemented cocktail. The EndoPGCp alone (Fig. 2b) released approximately 45% of the total when the enzyme was added to the commercial cocktail, and the cocktail alone released 15% of the total. The sum of the separate activities of the EndoPGCp and the commercial cocktail was therefore only 60% of the total activity of the mixture, indicating a synergistic effect of the combined enzymes.

Fig. 2
figure 2

Time course analysis of the hydrolysis of orange (a and b) and passion fruit (c and d) residues by EndoPGCp-MNPs (a and c) or free EndoPGCp (b and d). Conditions: Celluclast + EndoPGCp (blue), Celluclast only (red), EndoPGCp-MNPs, or free EndoPGCp only (black)

Full size image

In experiments using passion fruit waste, no reducing sugar release was detected on treatment with either the immobilized (Fig. 2c) or free pectinase (Fig. 2d). Although the addition of the commercial cocktail resulted in a gradual increase in reducing sugar release over time to reach final levels that were 2 times higher than with the orange residues, the addition of the pectinase did not significantly alter the total reducing sugar release as compared to commercial enzyme cocktail. The increased reducing sugar release from passion fruit residues may reflect the higher pectin content in passion fruit residues than in orange residues [51]. Furthermore, the degree of esterification is higher in the pectin from passion fruit [51], and the differences in reducing sugar release by the EndoPGCp may indicate that the enzyme preferentially hydrolyses non-esterified substrates.

The synergistic effect observed between EndoPGCp and the enzymes that make up the commercial cocktail may be related to the structure of the orange plant cell wall, which presents the type I plant cell wall architecture [52]. In this type of cell wall, the content of the three main polysaccharides is 15–30% cellulose, 20–25% hemicelluloses, and 20–35% pectin. In addition, although the lignin content is extremely low, the recalcitrance of the material is related to the high concentration of pectin [52, 53]. The results obtained in this work for the orange pomace hydrolysis indicate that the EndoPGCp improves the saccharification efficiency by reduction of the recalcitrance of the cell wall, due to the hydrolysis of the pectin and increasing the access of other hydrolytic enzymes.

Reusability of EndoPGCp-MNP

The possibility of reusing immobilized enzymes is one of the main advantages that directly affects the cost reduction associated with their industrial application [6, 11]. The reusability of the immobilized pectinase is presented in Fig. 3. EndoPGCp-MNP retained more than 65% of the initial activity even after seven re-use cycles, and after 10 cycles, the enzyme still retained about 30% of its initial activity. Similar results have been reported for commercial pectinase immobilized onto magnetic nanoparticles via glutaraldehyde activation [40, 54]. The possibility of easy separation of the enzyme from the reaction medium using a magnetic field is an added advantage compared to enzymes immobilized on other types of non-magnetic supports [17] where the biocatalyst is normally separated by filtration or centrifugation [29, 30, 55]. However, in order to recover the immobilized enzyme by filtration, larger particles are required which tends to result in substantial loss of enzyme activity due to the increased diffusion limitation through larger particles [29, 30, 55].

Fig. 3
figure 3

Re-use of EndoPG-MNP determined by repeated use-wash cycles of the hydrolysis assay

Full size image

Conclusions

This study has demonstrated that the use of immobilized pectinase on chitosan-coated ferromagnetic particles (EndoPGCp-MNPs) can be advantageous for the hydrolysis of pectinolytic-rich residues as compared to the free enzyme. The immobilization process does not alter enzyme activity since the EndoPGCp-MNP retained essentially the same biochemical properties of the free enzyme such as optima temperature, activation energy, and kinetic parameters, and even demonstrated a slight increase in the thermostability. The use of the pectinase preparations to supplement the commercial cocktail Celluclast for the hydrolysis of orange residues increased the release of reducing sugars by about 1.8-fold and in the case of the EndoPGCp showed a synergistic effect compared with controls using Celluclast and pectinase alone. The immobilization allowed repeated use of the EndoPGCp-MNPs, which retained about 30% of the initial activity after 10 re-use cycles. The use of ferromagnetic particles allowed the easy recovery of the EndoPGCp-MNP from reaction media using a magnetic field. Taken together, our results showed that the pectinase immobilized on ferromagnetic particles has advantages that are compatible with industrial applications.