New Heterofunctional Supports Based on Glutaraldehyde-Activation: A Tool for Enzyme Immobilization at Neutral pH

Immobilization is an exciting alternative to improve the stability of enzymatic processes. However, part of the applied covalent strategies for immobilization uses specific conditions, generally alkaline pH, where some enzymes are not stable. Here, a new generation of heterofunctional supports with application at neutral pH conditions was proposed. New supports were developed with different bifunctional groups (i.e., hydrophobic or carboxylic/metal) capable of adsorbing biocatalysts at different regions (hydrophobic or histidine richest place), together with a glutaraldehyde group that promotes an irreversible immobilization at neutral conditions. To verify these supports, a multi-protein model system (E. coli extract) and four enzymes (Candida rugosa lipase, metagenomic lipase, β-galactosidase and β-glucosidase) were used. The immobilization mechanism was tested and indicated that moderate ionic strength should be applied to avoid possible unspecific adsorption. The use of different supports allowed the immobilization of most of the proteins contained in a crude protein extract. In addition, different supports yielded catalysts of the tested enzymes with different catalytic properties. At neutral pH, the new supports were able to adsorb and covalently immobilize the four enzymes tested with different recovered activity values. Notably, the use of these supports proved to be an efficient alternative tool for enzyme immobilization at neutral pH.


Introduction
A simple, cheap and efficient immobilization strategy represents one of the main bottlenecks in the implementation of many industrial-scale enzymatic processes. Immobilization technology, as a very powerful tool, has been intensively applied to prepare various high-performance and economically-feasible biocatalysts improving activity, stability and selectivity [1][2][3][4]. By careful adjusting of the immobilization strategies, heterogeneous biocatalysts can work in broader pH and temperature ranges, and show greater thermal stability than their native free form [5,6]. Currently, a variety of methodologies for the immobilization of enzymes have been described. Among them, covalent immobilization on solid supports has been shown to be able to stabilize a variety of enzymes. Thus, supports activated with different functional groups (epoxy, aldehyde, glutaraldehyde, and others) have been developed and applied. However, the use of these supports only permit the random immobilization or the immobilization through only one region (e.g., richest in lysine region for aldehyde supports). This makes it so that only one kind of catalyst can be obtained using a functional covalent group; the product of the rigidification through one region or the average of the properties of the different enzyme molecules by random immobilization [7][8][9][10]. Because of these properties, several studies have shown the efficiency of immobilizing different enzymes in heterofunctionalized supports using various adsorbing groups (octyl, primary or secondary amino groups, ionic exchangers, metallic chelates and others) and reactive moieties (glyoxyl, epoxy, aldehyde, divinyl sulfone, amino, or glutamic groups), which are capable of adsorbing the protein physically and subsequently establishing covalent or ionic linkages with nucleophilic groups on the adsorbed protein [6,[11][12][13][14][15][16]. One of the main advantages of use this type of supports is that they allows the orientation of the enzyme immobilization in different zones in its surface, being an alternative to the monofunctional supports. Heterofunctional supports are currently being studied as a useful tool in the immobilization of lipases (lipases from Pseudomonas stutzeri, Rhizomucor miehei, Thermomyces lanuginosus, and Candida antarctica), for instance, immobilizing silica or agarose supports containing distinct hydrophobic adsorbing groups (different alkyl groups) and reactive moieties [12,13,[15][16][17]. In addition to the lipases, β-galactosidases (e.g., Aspergillus oryzae and Thermus sp.) and other enzymes have been immobilized on heterofunctional supports, obtaining good results with respect to their use [18][19][20][21]. For most strategies involving the application of heterofunctional supports, the immobilization process needs to be carried out in at least two steps: first the enzyme is adsorbed into the different adsorbent groups under mild conditions (i.e., low ionic strength and neutral pH), followed by elevation to alkaline conditions (around pH 10) for the occurrence of multiple interactions between nucleophilic groups on the enzyme (typically amino, hydroxyl, and/or thiol moieties) and reactive groups on the surface of the support. In some cases an additional step is carried out involving the reduction process necessary to establish multiple covalent linkages between the support and the enzyme [9]. However, the reduction step can cause some deleterious effects on the protein structure, cleaving disulfide and peptide linkages or reducing some groups essential for catalysis [22,23].
Although these immobilization methods are very useful, their uses have several limitations. One disadvantage of its use is the need to apply alkaline pH conditions (around pH 10) to enhance the reactivity of the nucleophilic residues found on the protein surface. Thus enzymes that are intrinsically unstable at alkaline pH values (e.g., the lipase from Candida rugosa) can be inactivated by the immobilization process [9,24,25]. An alternative process is the use of the reagent glutaraldehyde. The chemical covalent attachment process using glutaraldehyde is particularly attractive, since it provides a carefully regulated connection with specific groups found on the proteins under mild pH conditions [26]. Furthermore, the chemical reactivity shown by glutaraldehyde means that the reduction process with sodium borohydride is not strictly necessary [9].
Standard immobilization techniques using glutaraldehyde as the coupling agent are quite simple, efficient and probably the most used technique to carry out enzyme immobilization. Glutaraldehyde can react with different portions of the enzyme, mainly involving the primary amino groups of proteins, although it may eventually react with other groups (thiols, phenols and imidazoles) [9,26]. Supports functionalized with glutaraldehyde groups are described as being readily constructed from different supports containing primary amino groups (i.e., ethylenediamine-activated supports). However, the chemistry of glutaraldehyde in aqueous solution as well as the structures related to protein crosslinking or to enzyme immobilization are not yet fully understood. Some structures have been proposed for aqueous solutions of glutaraldehyde, which reported that in the pH from 3.0 to 8.0 and under dilute conditions, the glutaraldehyde reagent may exist as monomers (free aldehyde form or pre-dominantly cyclic hemiacetal), whereas in concentrations above 25% or under acidic conditions, oligomeric hemiacetals are formed. The equilibrium between the linear and cyclic monomeric forms is described as being linearly oriented as the temperature increases. Under basic conditions, oligomeric aldehydes may be formed by way of intermolecular aldol condensations [9,26,27].
Apart from the structural complexity presented by glutaraldehyde, its configuration already characterizes it as having distinct heterofunctionalities. Depending on the conditions applied during the process of immobilization with glutaraldehyde-activated supports, the proteins can be immobilized by three different mechanisms. On using very high ionic strength, the protein may first be immobilized by way of hydrophobic adsorption before forming the covalent bonds, whereas when using low ionic strength, primary immobilization will be by way of anion exchange. If the ionic strength applied were moderate, protein immobilization is described as being mainly due to covalent bonds [9,26,[28][29][30].
In this work, a new generation of heterofunctional supports was proposed using different groups (hydrophobic or metal-chelate groups) capable of adsorbing proteins via different regions, followed by activation with the glutaraldehyde group, which can react covalently with the adsorbed proteins under neutral pH conditions (Scheme 1). Therefore, this work proposed the use of glutaraldehyde for the activation and formation of covalent linkages as a viable alternative to the alkaline mechanics currently applied. After the construction, a study was carried out with a multi-protein model system (E. coli BL21 crude extract) to define the conditions under which a pure immobilization mechanism occurred by way of adsorbed groups, and to define the conditions under which a connection can occur via only covalent bonds using the reagent glutaraldehyde. Structural analyses of the supports were carried out using Fourier Transformed Infrared (FTIR) spectroscopy with the aim of demonstrating the connection between the amine groups present in the support and the glutaraldehyde groups. Moreover studies using four different enzymes (Candida rugosa lipase-CRL, metagenomic lipase-LipC12, β-galactosidase-KlBgal and β-glucosidase-EaBglA) were evaluated to verify the ability of the new supports to adsorb and form covalent linkages under neutral pH conditions. The properties of the immobilized enzymes, such as their activity and thermal stability, were also studied.
characterizes it as having distinct heterofunctionalities. Depending on the conditions applied during the process of immobilization with glutaraldehyde-activated supports, the proteins can be immobilized by three different mechanisms. On using very high ionic strength, the protein may first be immobilized by way of hydrophobic adsorption before forming the covalent bonds, whereas when using low ionic strength, primary immobilization will be by way of anion exchange. If the ionic strength applied were moderate, protein immobilization is described as being mainly due to covalent bonds [9,26,[28][29][30]. Scheme 1. Schematic representation of the chemical modification in the surface of the agarose. The structures of the supports were idealized based on previous reviews concerning glutaraldehyde chemistry [26,27]. IDA (iminodiacetic acid); EDA (ethylenediamine); NaBH4 (sodium borohydride); NaIO4 (sodium periodate); GLU (glutaraldehyde); and HS-C8 (1-octanothiol).
In this work, a new generation of heterofunctional supports was proposed using different groups (hydrophobic or metal-chelate groups) capable of adsorbing proteins via different regions, followed by activation with the glutaraldehyde group, which can react covalently with the adsorbed proteins under neutral pH conditions (Scheme 1). Therefore, this work proposed the use of glutaraldehyde for the activation and formation of covalent linkages as a viable alternative to the alkaline mechanics currently applied. After the construction, a study was carried out with a multi-protein model system (E. coli BL21 crude extract) to define the conditions under which a pure immobilization mechanism occurred by way of adsorbed groups, and to define the conditions under which a Scheme 1. Schematic representation of the chemical modification in the surface of the agarose. The structures of the supports were idealized based on previous reviews concerning glutaraldehyde chemistry [26,27]. IDA (iminodiacetic acid); EDA (ethylenediamine); NaBH 4 (sodium borohydride); NaIO 4 (sodium periodate); GLU (glutaraldehyde); and HS-C8 (1-octanothiol).

Construction and Analysis of New Heterofunctional Supports
Agarose-based beads were chosen as a base matrix for the construction of different heterofunctional supports. Agarose is a strongly hydrophilic, lyophilic and inert colloid that can reversibly form stable and firm gels. Its suitability as an excellent support is confirmed by its high area rich in hydroxyl groups [31,32]. Briefly, agarose was activated with epichlorohydrin, a bifunctional reagent used to obtain epoxy groups [33]. The reaction was performed in alkaline conditions where most of the primary hydroxyl groups of the support were deprotonated. The primary reaction formed epoxy groups but as the reaction occurred in an alkaline environment, some of the epoxy groups created were hydrolyzed and yielded diol groups (Scheme 1). The total amount of activated primary hydroxyl groups obtained was around 65 ± 0.3 µmoL g −1 , with epoxy groups accounting for 23 ± 0.4 µmoL g −1 and diol groups accounting for 42 ± 0.4 µmoL g −1 . By obtaining a basic structure containing epoxy/diol groups, the new heterofunctional supports were prepared using the following strategy: (i) the epoxy groups formed were functionalized with different bifunctional reagents (i.e., hydrophobic or carboxylic/metal groups); and (ii) the diol groups were firstly oxidized with sodium periodate (NaIO 4 ), and then activated with amino groups using ethylenediamine (EDA) (Scheme 1). The amino groups were applied to future activation with glutaraldehyde molecules. The new supports were constructed using very stable bifunctional groups and inert amino groups that allow storage for long periods prior to activation with the glutaraldehyde molecules.
The new heterofunctional supports are expected to perform an initial protein adsorption only by the inserted heterofunctional groups (i.e., hydrophobic groups or metal chelate linked to the carboxylic groups), and only later to establish a covalent interaction with the glutaraldehyde group. However, reactive groups used in their construction, such as amino (from ethylenediamine, EDA), may interfere in a pure immobilization process. A conjugation chemistry that introduces a charged functional group into the support can cause nonspecific binding by promoting ion-exchange effects. However, in some cases, as in the traditional glutaraldehyde activated supports, a pure immobilization process is expected. Therefore, a linkage that alters the flow and binding characteristics of the support, e.g., EDA, is not of interest for the work. Thus, a possible unspecific adsorption via ion-exchange by the amino groups was studied. A multi-protein model system (E. coli BL21 crude extract, 5 mg g −1 of support) was offered to a pure amino support (activated only with ethylenediamine/Aga-EDA), and then the adsorption process at different ionic strengths was analyzed (Figure 1a). The E. coli extract was applied because it contains proteins with different properties and characteristics, therefore providing different affinities for the adsorption [30]. At pH 7 and low ionic strength (25 mM sodium phosphate buffer), 50% of proteins were quickly adsorbed in 30 min (Figure 1a). An increase in the ionic strength promoted a decrease in the adsorption by amino groups until a negligible adsorption of proteins was found in 500 mM sodium phosphate buffer (even at 200 mM sodium phosphate buffer, the adsorption was almost negligible during the first hour). The results show that the new supports could not be used under low ionic strength due to protein interaction being conducted via a double mechanism, which would be promoted by amino groups (from ethylenediamine), and by the heterofunctional groups inserted in the new supports. Aga-C8-GLU; Aga-IDA-Ni 2+ -EDA-GLU; Aga-EDA-GLU; Aga-EDA (Control); and Aga-EDA-GLU-Control (Control). The results are expressed as the average of triplicate assays ± the standard error of the mean.
Similarly, a possible adsorption by the glutaraldehyde structure was studied. Depending on the conditions applied, the immobilization process with glutaraldehyde groups can occur by a double step: in a first step, there is the adsorption of proteins on the support and then, the covalent immobilization [26,30]. Thus, a study using a support activated only by glutaraldehyde groups but reduced with sodium borohydride (Aga-EDA-GLU-Control) was done. The reduction step with sodium borohydride was used to transform the Schiff's bases (-N=C-double bond) into stable primary amino bonds [34]. A negligible adsorption was observed using either 200 mM or 500 mM sodium phosphate buffer (Figure 1b). On the other hand, when low ionic strengths were applied, a significant protein adsorption was found (21% in 100 mM). Therefore, the results indicate that the process of the immobilization on the new heterofunctional supports has to be carried out at a moderate ionic strength (200-500 mM) to prevent unspecific interaction, such as ion-exchange. Aga-C8-GLU; Aga-IDA-Ni 2+ -EDA-GLU; Aga-EDA-GLU; Aga-EDA (Control); and Aga-EDA-GLU-Control (Control). The results are expressed as the average of triplicate assays ± the standard error of the mean.
Similarly, a possible adsorption by the glutaraldehyde structure was studied. Depending on the conditions applied, the immobilization process with glutaraldehyde groups can occur by a double step: in a first step, there is the adsorption of proteins on the support and then, the covalent immobilization [26,30]. Thus, a study using a support activated only by glutaraldehyde groups but reduced with sodium borohydride (Aga-EDA-GLU-Control) was done. The reduction step with sodium borohydride was used to transform the Schiff's bases (-N=C-double bond) into stable primary amino bonds [34]. A negligible adsorption was observed using either 200 mM or 500 mM sodium phosphate buffer (Figure 1b). On the other hand, when low ionic strengths were applied, a significant protein adsorption was found (21% in 100 mM). Therefore, the results indicate that the process of the immobilization on the new heterofunctional supports has to be carried out at a Aga-C8-GLU; Aga-IDA-Ni 2+ -EDA-GLU; Aga-EDA-GLU; Aga-EDA (Control); and Aga-EDA-GLU-Control (Control). The results are expressed as the average of triplicate assays ± the standard error of the mean.
Similarly, a possible adsorption by the glutaraldehyde structure was studied. Depending on the conditions applied, the immobilization process with glutaraldehyde groups can occur by a double step: in a first step, there is the adsorption of proteins on the support and then, the covalent immobilization [26,30]. Thus, a study using a support activated only by glutaraldehyde groups but reduced with sodium borohydride (Aga-EDA-GLU-Control) was done. The reduction step with sodium borohydride was used to transform the Schiff's bases (-N=C-double bond) into stable primary amino bonds [34]. A negligible adsorption was observed using either 200 mM or 500 mM sodium phosphate buffer (Figure 1b). On the other hand, when low ionic strengths were applied, a significant protein adsorption was found (21% in 100 mM). Therefore, the results indicate that the process of the immobilization on the new heterofunctional supports has to be carried out at a Similarly, a possible adsorption by the glutaraldehyde structure was studied. Depending on the conditions applied, the immobilization process with glutaraldehyde groups can occur by a double step: in a first step, there is the adsorption of proteins on the support and then, the covalent immobilization [26,30]. Thus, a study using a support activated only by glutaraldehyde groups but reduced with sodium borohydride (Aga-EDA-GLU-Control) was done. The reduction step with sodium borohydride was used to transform the Schiff's bases (-N=C-double bond) into stable primary amino bonds [34]. A negligible adsorption was observed using either 200 mM or 500 mM sodium phosphate buffer (Figure 1b). On the other hand, when low ionic strengths were applied, a significant protein adsorption was found (21% in 100 mM). Therefore, the results indicate that the process of the immobilization on the new heterofunctional supports has to be carried out at a Similarly, a possible adsorption by the glutaraldehyde structure was studied. Depending on the conditions applied, the immobilization process with glutaraldehyde groups can occur by a double step: in a first step, there is the adsorption of proteins on the support and then, the covalent immobilization [26,30]. Thus, a study using a support activated only by glutaraldehyde groups but reduced with sodium borohydride (Aga-EDA-GLU-Control) was done. The reduction step with sodium borohydride was used to transform the Schiff's bases (-N=C-double bond) into stable primary amino bonds [34]. A negligible adsorption was observed using either 200 mM or 500 mM sodium phosphate buffer (Figure 1b). On the other hand, when low ionic strengths were applied, a significant protein adsorption was found (21% in 100 mM). Therefore, the results indicate that the process of the immobilization on the new heterofunctional supports has to be carried out at a Similarly, a possible adsorption by the glutaraldehyde structure was studied. Depending on the conditions applied, the immobilization process with glutaraldehyde groups can occur by a double step: in a first step, there is the adsorption of proteins on the support and then, the covalent immobilization [26,30]. Thus, a study using a support activated only by glutaraldehyde groups but reduced with sodium borohydride (Aga-EDA-GLU-Control) was done. The reduction step with sodium borohydride was used to transform the Schiff's bases (-N=C-double bond) into stable primary amino bonds [34]. A negligible adsorption was observed using either 200 mM or 500 mM sodium phosphate buffer (Figure 1b). On the other hand, when low ionic strengths were applied, a significant protein adsorption was found (21% in 100 mM). Therefore, the results indicate that the process of the immobilization on the new heterofunctional supports has to be carried out at a  Similarly, a possible adsorption by the glutaraldehyde structure was studied. Depending on the conditions applied, the immobilization process with glutaraldehyde groups can occur by a double step: in a first step, there is the adsorption of proteins on the support and then, the covalent immobilization [26,30]. Thus, a study using a support activated only by glutaraldehyde groups but reduced with sodium borohydride (Aga-EDA-GLU-Control) was done. The reduction step with sodium borohydride was used to transform the Schiff's bases (-N=C-double bond) into stable primary amino bonds [34]. A negligible adsorption was observed using either 200 mM or 500 mM sodium phosphate buffer (Figure 1b). On the other hand, when low ionic strengths were applied, a significant protein adsorption was found (21% in 100 mM). Therefore, the results indicate that the process of the immobilization on the new heterofunctional supports has to be carried out at a Similarly, a possible adsorption by the glutaraldehyde structure was studied. Depending on the conditions applied, the immobilization process with glutaraldehyde groups can occur by a double step: in a first step, there is the adsorption of proteins on the support and then, the covalent immobilization [26,30]. Thus, a study using a support activated only by glutaraldehyde groups but reduced with sodium borohydride (Aga-EDA-GLU-Control) was done. The reduction step with sodium borohydride was used to transform the Schiff's bases (-N=C-double bond) into stable primary amino bonds [34]. A negligible adsorption was observed using either 200 mM or 500 mM sodium phosphate buffer (Figure 1b). On the other hand, when low ionic strengths were applied, a significant protein adsorption was found (21% in 100 mM). Therefore, the results indicate that the process of the immobilization on the new heterofunctional supports has to be carried out at a Similarly, a possible adsorption by the glutaraldehyde structure was studied. Depending on the conditions applied, the immobilization process with glutaraldehyde groups can occur by a double step: in a first step, there is the adsorption of proteins on the support and then, the covalent immobilization [26,30]. Thus, a study using a support activated only by glutaraldehyde groups but reduced with sodium borohydride (Aga-EDA-GLU-Control) was done. The reduction step with sodium borohydride was used to transform the Schiff's bases (-N=C-double bond) into stable primary amino bonds [34]. A negligible adsorption was observed using either 200 mM or 500 mM sodium phosphate buffer (Figure 1b). On the other hand, when low ionic strengths were applied, a significant protein adsorption was found (21% in 100 mM). Therefore, the results indicate that the process of the immobilization on the new heterofunctional supports has to be carried out at a Aga Aga-EDA (Control); and Aga average of triplicate assays ± the stand Similarly, a possible adsorption by conditions applied, the immobilization step: in a first step, there is the ads immobilization [26,30]. Thus, a study reduced with sodium borohydride (A sodium borohydride was used to tra primary amino bonds [34]. A negligib sodium phosphate buffer (Figure 1b). significant protein adsorption was fou process of the immobilization on the moderate ionic strength (200-500 mM Similarly, a possible adsorption by the glutaraldehyde structure was studied. Depending on the conditions applied, the immobilization process with glutaraldehyde groups can occur by a double step: in a first step, there is the adsorption of proteins on the support and then, the covalent immobilization [26,30]. Thus, a study using a support activated only by glutaraldehyde groups but reduced with sodium borohydride (Aga-EDA-GLU-Control) was done. The reduction step with sodium borohydride was used to transform the Schiff's bases (-N=C-double bond) into stable primary amino bonds [34]. A negligible adsorption was observed using either 200 mM or 500 mM sodium phosphate buffer (Figure 1b). On the other hand, when low ionic strengths were applied, a significant protein adsorption was found (21% in 100 mM). Therefore, the results indicate that the process of the immobilization on the new heterofunctional supports has to be carried out at a Similarly, a possible adsorption by the glutaraldehyde structure was studied. Depending on the conditions applied, the immobilization process with glutaraldehyde groups can occur by a double step: in a first step, there is the adsorption of proteins on the support and then, the covalent immobilization [26,30]. Thus, a study using a support activated only by glutaraldehyde groups but reduced with sodium borohydride (Aga-EDA-GLU-Control) was done. The reduction step with sodium borohydride was used to transform the Schiff's bases (-N=C-double bond) into stable primary amino bonds [34]. A negligible adsorption was observed using either 200 mM or 500 mM sodium phosphate buffer (Figure 1b). On the other hand, when low ionic strengths were applied, a significant protein adsorption was found (21% in 100 mM). Therefore, the results indicate that the process of the immobilization on the new heterofunctional supports has to be carried out at a moderate ionic strength (200-500 mM) to prevent unspecific interaction, such as ion-exchange. Furthermore, it has been described that protein immobilization using glutaraldehyde groups with moderate ionic strength occur by direct covalent attachment [9,26,35], which is of interest in the present study.
Considering these data, the E. coli BL21 crude extract was offered to new supports using moderate ionic strength (200 mM sodium phosphate buffer at pH 7) to prevent unspecific absorptions, and to promote a direct covalent and irreversible immobilization by glutaraldehyde groups. Using the new supports (Aga-C8-GLU and Aga-IDA-Metal-EDA-GLU), more than 95% of all proteins from E. coli BL21 crude extract were immobilized after 5 h (Figure 1c). To compare the immobilization efficiency of the new supports, the crude extract was offered to the traditional support activated using glutaraldehyde (Aga-EDA-GLU), and a similar result was obtained after 5 h (Figure 1c). The controls (Aga-EDA and Aga-EDA-GLU-Control) were repeated and no protein adsorption was found ( Figure 1c). The findings suggest that the proteins can be immobilized on new glutaraldehyde-activated supports at moderate ionic strength, where unspecific interactions are not found.
FTIR spectra analysis was used to characterize the activation with the glutaraldehyde group, reflecting the effectiveness of the developed procedure in producing different glutaraldehyde-activated supports. The FTIR spectra were taken from untreated and treated with the glutaraldehyde group, using EDA-functionalized supports to determine any changes in chemical bonding states ( Figure 2). The new band around 1660 cm −1 (as shown in Figure 2a(2), b(2)) was ascribed to the presence of imine (-N=C-) in the support as the result of the reaction between glutaraldehyde and the amino groups of the EDA found on the supports. Previous studies indicate that the reaction of the agent glutaraldehyde with primary amine groups in the treated support is well accepted [27,36]. FTIR spectra analysis was used to characterize the activation with the glutaraldehyde group, reflecting the effectiveness of the developed procedure in producing different glutaraldehydeactivated supports. The FTIR spectra were taken from untreated and treated with the glutaraldehyde group, using EDA-functionalized supports to determine any changes in chemical bonding states ( Figure 2). The new band around 1660 cm −1 (as shown in Figure 2a(2) and 2b (2)) was ascribed to the presence of imine (-N=C-) in the support as the result of the reaction between glutaraldehyde and the amino groups of the EDA found on the supports. Previous studies indicate that the reaction of the agent glutaraldehyde with primary amine groups in the treated support is well accepted [27,36].

Immobilization of Different Biocatalysts on the New Heterofunctional Supports
The new supports were used for the covalent immobilization of different enzymes in mild conditions (neutral pH). For the immobilization assays on these supports, four enzymes of biotechnological interest were used, namely: Candida rugosa lipase (CRL), metagenomic lipase

Immobilization of Different Biocatalysts on the New Heterofunctional Supports
The new supports were used for the covalent immobilization of different enzymes in mild conditions (neutral pH). For the immobilization assays on these supports, four enzymes of biotechnological interest were used, namely: Candida rugosa lipase (CRL), metagenomic lipase (LipC12), β-galactosidase (KlBgal) and β-glucosidase (EaBglA).
To prevent non-specific adsorptions, the immobilization tests were carried out using a moderate ionic strength (200 mM at pH 7). In all cases, the non-specific adsorption was negligible. Table 1 shows the tests with the different glutaraldehyde-activated supports using the lipase from Candida rugosa (CRL). Lipase CRL is an enzyme used in several biocatalytic processes; however, it shows great structural instability especially at alkaline pH [25,37]. At pH 7, immobilization tests with lipase CRL on the new glutaraldehyde-activated supports were done, and the data showed that the final recovered activity was strongly dependent on the support used. The recovered activity (R) values were from 210% (new hydrophobic group-glutaraldehyde support, Aga-C8-GLU) to 54% (new metal group-glutaraldehyde support, Aga-IDA-Metal-EDA-GLU). Traditional glutaraldehyde support (Aga-EDA-GLU) showed almost unaltered recovered activity (R = 95%) compared to the initial activity of the free enzyme. For further studies, lipase CRL was also immobilized on a commercial support activated only with hydrophobic groups (Octyl-Sepharose). The commercial hydrophobic support showed a recovered activity of R = 225%, similar to that found by the new hydrophobic group-glutaraldehyde support (Aga-C8-GLU, Table 1). Both hydrophobic supports have shown a hyperactivation described as being related to the interfacial activation of the enzyme, whereby the catalyst may be adsorbed in its open form, compared to the soluble form in the standard conformational equilibrium [38]. Thus, the presence of the agent glutaraldehyde on Aga-C8-GLU did not significantly alter lipase adsorption by the hydrophobic group. The results for lipase CRL immobilized on the Aga-C8-GLU were in general agreement with a previous study with octyl activated with divinyl sulfone (OCDVS, more than 3-fold recovered activity) [14]. Aga-IDA-Ni 2+ -EDA-GLU 100 100 Aga-EDA-GLU 100 42 C8 (1-octanothiol); Ni 2+ (nickel); EDA (ethylenediamine); IDA (iminodiacetic acid); Aga (agarose beads); and GLU (glutaraldehyde). a Calculated as the difference between the initial and final activities in the supernatant after 1 h of immobilization. b Recovered activity (%), measured as the ratio between the real activity (U g −1 of support) of immobilized enzymes (CRL or LipC12) and theoretical activity of the immobilized enzymes (CRL or LipC12) (U g −1 of support).
To analyze if CRL was covalently attached to the new supports and not only by adsorption, the different CRL preparations were boiled in Laemmli's disruption buffer (which contains mercaptoethanol and SDS) and analyzed by SDS-polyacrylamide gel electrophoresis (SDS-PAGE, Figure 3). CRL was used as a model because it is a monomeric enzyme. The glutaraldehyde is known to establish covalent bonds at pH 7 [9,39]. According to SDS-PAGE results, the new glutaraldehyde-activated supports did not allow the release of the immobilized CRL, proving that the new supports can promote the adsorption and subsequent covalent immobilization by the glutaraldehyde group at pH 7. CRL immobilized on Octyl-Sepharose produced a band in SDS-PAGE (Figure 3), which indicates only the adsorption to the hydrophobic groups. used as a model because it is a monomeric enzyme. The glutaraldehyde is known to establish covalent bonds at pH 7 [9,39]. According to SDS-PAGE results, the new glutaraldehyde-activated supports did not allow the release of the immobilized CRL, proving that the new supports can promote the adsorption and subsequent covalent immobilization by the glutaraldehyde group at pH 7. CRL immobilized on Octyl-Sepharose produced a band in SDS-PAGE (Figure 3), which indicates only the adsorption to the hydrophobic groups. Further experiments of lipase CRL desorption from the different supports were done by the addition of Triton ×100, NaCl and imidazol ( Table 2). The tests were used to prove that lipase CRL Further experiments of lipase CRL desorption from the different supports were done by the addition of Triton ×100, NaCl and imidazol ( Table 2). The tests were used to prove that lipase CRL was not hydrophobic or ion-bound to the new supports. In the analysis, the activities of supernatants and suspensions were measured to verify the desorption of the immobilized CRL. The use of the non-ionic detergent Triton ×100 produced a significant desorption of the immobilized CRL on Octyl-Sepharose (81% initial activity), while a negligible activity was obtained in the supernatant for the new hydrophobic group-glutaraldehyde support/Aga-C8-GLU (Table 2). Higher concentrations of Triton ×100 were not able to remove CRL from the new glutaraldehyde-activated support (data not shown). Analysis using NaCl and imidazol showed a negligible activity found in the supernatant for the immobilized CRL on Aga-EDA-GLU and Aga-IDA-Metal-EDA-GLU supports. These results confirm the covalent immobilization of lipase CRL by the glutaraldehyde group at pH 7.  Another monomeric lipase obtained from metagenomics (lipase LipC12, [40]) was also studied ( Table 1). All the assayed supports allowed the complete immobilization of lipase LipC12. The recovered activities were different depending on the support used in the immobilization (Table 1). Upon immobilization on all supports, a high-recovered activity (R = 269%) was obtained with the new hydrophobic group-glutaraldehyde support (Aga-C8-GLU) showing a hyperactivation of lipase LipC12 (Table 1). Moreover, LipC12 immobilized on Aga-IDA-Metal-EDA-GLU and Aga-EDA-GLU (traditional support) showed recovered activity of 100% and 42%, respectively. The traditional glutaraldehyde-activated support used as control system showed a low recovered activity of LipC12 in relation to the new supports. A similar study of covalent attachment by the glutaraldehyde group was accomplished, and no desorption was found for LipC12 immobilized on the new supports (data not shown).
Tests analyzing the possibility of using the new glutaraldehyde-activated supports on the immobilization of two complex multimeric enzymes were studied. β-galactosidase (KlBgal) and β-glucosidase (EaBglA) are described as tetrameric enzymes [41,42]. The structural complexity of these catalysts represents an especially complex problem for immobilization because small disturbances in their quaternary-structures may alter the precise distribution of the native state ensemble [39,43], with dissociation of subunits producing enzymatic inactivation. Therefore, these enzymes are fragile even in mild conditions. Table 3 shows the results for KlBgal and EaBglA with a final recovered activity ranging from 81 to 70% for the β-galactosidase and from 64 to 0% for the β-glucosidase. KlBgal immobilized on the new supports showed recovered activity values of 81% (Aga-C8-GLU) and 78% (Aga-IDA-Co 2+ -EDA-GLU). The traditional GLU supports (Aga-EDA-GLU) had an R 70%. The traditional glutaraldehyde-activated support (control system) showed a low recovered activity of KlBgal in relation to the new supports. A similar result has been reported for the covalent immobilization of Kluyveromyces lactis β-galactosidase in glyoxyl-Sepharose support [20], retaining approximately 82% of its initial activity after immobilization. In the case of EaBglA, Aga-IDA-Co 2+ -EDA-GLU showed the highest R (64%), following the traditional Aga-EDA-GLU with R 50%. The new hydrophobic-glutaraldehyde support yielded an inactive derivative after the immobilization process. Zanphorlin et al. [42] describe that the selective pressure at low temperatures favored mutations that redesigned the EaBglA protein surface, reducing the number of salt bridges, and exposing more the hydrophobic regions to the solvent. Its can thus be concluded that one reason for the inactivation of EaBglA when immobilized in the hydrophobic support may be due to the interaction between a high number of hydrophobic groups in the protein surface and the hydrophobic group of the support. As in the other cases, no activity was detected after desorption of the different preparations after incubation in Triton ×100, NaCl and imidazol.

Thermal Stability
Enzymatic stabilization has great importance due to the increasing number of applications of enzymes in almost all areas and it is one of the main aims when the immobilization of enzymes is performed [6]. Therefore, the thermal stabilities of different immobilized preparations and free forms were studied in 100 mM phosphate buffers, pH 7 using distinct temperatures. Figure 4a shows the tests using lipase CRL preparations. The stability of the adsorbed preparation (Octyl-Sepharose) was considerably similar to the free form with around 3.6-fold stabilization. The maximal stability was obtained with the traditional glutaraldehyde support (Aga-EDA-GLU) with a half-life of around 1560 min after incubation at 50 • C ( Table 4). Considering that the half-life of the free CRL was 21 min, the stabilization factor obtained was around 74.3-fold. CRL immobilized on Aga-C8-GLU and Aga-IDA-Ni 2+ -EDA-GLU was 5.7 and 2.8-fold more stable than the free form.
The thermal stability of lipase LipC12 preparations was assessed by incubation at 70 • C (Figure 4b, Table 4). At this temperature, the native free LipC12 had a half-life time of 13 min. The most stable preparation was found using the new hydrophobic group-glutaraldehyde support (Aga-C8-GLU) with a stabilization factor around 32.3-fold compared to the free form. The stability of LipC12 immobilized on the traditional amino (Ag-EDA-GLU) and new metal group-glutaraldehyde support (Aga-IDA-Ni 2+ -EDA-GLU) was, respectively, 5.8 and 1.8-fold higher than the free LipC12. The high improvement of lipase LipC12 after immobilization (Aga-C8-GLU/LipC12) is important because it permits the transformation of a mesophilic enzyme into an enzyme with properties that are similar to those of thermophilic organisms, such as Bacillus thermocatenolatus lipase (BTL) and Thermus thermophilus lipase (TTL) [44,45]. Figure 5 shows the thermal stability of the immobilized KlBgal and EaBglA at 40 • C. For KlBgal preparations, only Aga-IDA-Co 2+ -EDA-GLU promoted a significant stabilization (4.8-fold factor) compared to the free form (half-life of 33 min at 40 • C, Figure 5a and Table 4). All other preparations showed a lower stabilization than the free form (Aga-C8-GLU 0.52 and Aga-EDA-GLU 0.36-fold factor, Table 4). For studies with the immobilized EaBglA, it was observed that the Aga-IDA-Co 2+ -EDA-GLU (half-life time of 230 min) showed the best result, which was more stable (5.9-fold factor) than the traditional GLU preparation (Aga-EDA-GLU), and 10.9-times more stable than free enzyme (half-life time of 21 min at 40 • C, Figure 5b and Table 4). The Aga-EDA-GLU preparation also presents a half-life of 39 min (1.8-fold factor more stable than free form). In all cases, the stability of different preparations was strongly dependent on the support used in the immobilization process.   enzymes in almost all areas and it is one of the main aims when the immobilization of enzymes is performed [6]. Therefore, the thermal stabilities of different immobilized preparations and free forms were studied in 100 mM phosphate buffers, pH 7 using distinct temperatures. Figure 4a shows the tests using lipase CRL preparations. The stability of the adsorbed preparation (Octyl-Sepharose) was considerably similar to the free form with around 3.6-fold stabilization. The maximal stability was obtained with the traditional glutaraldehyde support (Aga-EDA-GLU) with a half-life of around 1560 min after incubation at 50 °C ( Table 4). Considering that the half-life of the free CRL was 21 min, the stabilization factor obtained was around 74.3-fold. CRL immobilized on Aga-C8-GLU and Aga-IDA-Ni 2+ -EDA-GLU was 5.7 and 2.8-fold more stable than the free form.
The thermal stability of lipase LipC12 preparations was assessed by incubation at 70 °C ( Figure  4b, Table 4). At this temperature, the native free LipC12 had a half-life time of 13 min. The most stable preparation was found using the new hydrophobic group-glutaraldehyde support (Aga-C8-GLU) with a stabilization factor around 32.3-fold compared to the free form. The stability of LipC12 immobilized on the traditional amino (Ag-EDA-GLU) and new metal group-glutaraldehyde support (Aga-IDA-Ni 2+ -EDA-GLU) was, respectively, 5.8 and 1.8-fold higher than the free LipC12. The high improvement of lipase LipC12 after immobilization (Aga-C8-GLU/LipC12) is important because it permits the transformation of a mesophilic enzyme into an enzyme with properties that are similar to those of thermophilic organisms, such as Bacillus thermocatenolatus lipase (BTL) and Thermus thermophilus lipase (TTL) [44,45]. Figure 5 shows the thermal stability of the immobilized KlBgal and EaBglA at 40 °C. For KlBgal preparations, only Aga-IDA-Co 2+ -EDA-GLU promoted a significant stabilization (4.8-fold factor) compared to the free form (half-life of 33 min at 40 °C, Figure 5a and Table 4). All other preparations showed a lower stabilization than the free form (Aga-C8-GLU 0.52 and Aga-EDA-GLU 0.36-fold factor, Table 4). For studies with the immobilized EaBglA, it was observed that the Aga-IDA-Co 2+ -EDA-GLU (half-life time of 230 min) showed the best result, which was more stable (5.9-fold factor) than the traditional GLU preparation (Aga-EDA-GLU), and 10.9-times more stable than free enzyme (half-life time of 21 min at 40 °C, Figure 5b and Table 4). The Aga-EDA-GLU preparation also presents a half-life of 39 min (1.8-fold factor more stable than free form). In all cases, the stability of different preparations was strongly dependent on the support used in the immobilization process.  ) Aga-IDA-Metal-EDA-GLU; ( enzymes in almost all areas and it is one of the main aims when the immobilization of enzymes is performed [6]. Therefore, the thermal stabilities of different immobilized preparations and free forms were studied in 100 mM phosphate buffers, pH 7 using distinct temperatures. Figure 4a shows the tests using lipase CRL preparations. The stability of the adsorbed preparation (Octyl-Sepharose) was considerably similar to the free form with around 3.6-fold stabilization. The maximal stability was obtained with the traditional glutaraldehyde support (Aga-EDA-GLU) with a half-life of around 1560 min after incubation at 50 °C ( Table 4). Considering that the half-life of the free CRL was 21 min, the stabilization factor obtained was around 74.3-fold. CRL immobilized on Aga-C8-GLU and Aga-IDA-Ni 2+ -EDA-GLU was 5.7 and 2.8-fold more stable than the free form. The thermal stability of lipase LipC12 preparations was assessed by incubation at 70 °C ( Figure  4b, Table 4). At this temperature, the native free LipC12 had a half-life time of 13 min. The most stable preparation was found using the new hydrophobic group-glutaraldehyde support (Aga-C8-GLU) with a stabilization factor around 32.3-fold compared to the free form. The stability of LipC12 immobilized on the traditional amino (Ag-EDA-GLU) and new metal group-glutaraldehyde support (Aga-IDA-Ni 2+ -EDA-GLU) was, respectively, 5.8 and 1.8-fold higher than the free LipC12. The high improvement of lipase LipC12 after immobilization (Aga-C8-GLU/LipC12) is important because it permits the transformation of a mesophilic enzyme into an enzyme with properties that are similar to those of thermophilic organisms, such as Bacillus thermocatenolatus lipase (BTL) and Thermus thermophilus lipase (TTL) [44,45]. Figure 5 shows the thermal stability of the immobilized KlBgal and EaBglA at 40 °C. For KlBgal preparations, only Aga-IDA-Co 2+ -EDA-GLU promoted a significant stabilization (4.8-fold factor) compared to the free form (half-life of 33 min at 40 °C, Figure 5a and Table 4). All other preparations showed a lower stabilization than the free form (Aga-C8-GLU 0.52 and Aga-EDA-GLU 0.36-fold factor, Table 4). For studies with the immobilized EaBglA, it was observed that the Aga-IDA-Co 2+ -EDA-GLU (half-life time of 230 min) showed the best result, which was more stable (5.9-fold factor) than the traditional GLU preparation (Aga-EDA-GLU), and 10.9-times more stable than free enzyme (half-life time of 21 min at 40 °C, Figure 5b and Table 4). The Aga-EDA-GLU preparation also presents a half-life of 39 min (1.8-fold factor more stable than free form). In all cases, the stability of different preparations was strongly dependent on the support used in the immobilization process.  ) Aga-C8-GLU; ( enzymes in almost all areas and it is one of the main aims when the immobilization of enzymes is performed [6]. Therefore, the thermal stabilities of different immobilized preparations and free forms were studied in 100 mM phosphate buffers, pH 7 using distinct temperatures. Figure 4a shows the tests using lipase CRL preparations. The stability of the adsorbed preparation (Octyl-Sepharose) was considerably similar to the free form with around 3.6-fold stabilization. The maximal stability was obtained with the traditional glutaraldehyde support (Aga-EDA-GLU) with a half-life of around 1560 min after incubation at 50 °C ( Table 4). Considering that the half-life of the free CRL was 21 min, the stabilization factor obtained was around 74.3-fold. CRL immobilized on Aga-C8-GLU and Aga-IDA-Ni 2+ -EDA-GLU was 5.7 and 2.8-fold more stable than the free form.
The thermal stability of lipase LipC12 preparations was assessed by incubation at 70 °C ( Figure  4b, Table 4). At this temperature, the native free LipC12 had a half-life time of 13 min. The most stable preparation was found using the new hydrophobic group-glutaraldehyde support (Aga-C8-GLU) with a stabilization factor around 32.3-fold compared to the free form. The stability of LipC12 immobilized on the traditional amino (Ag-EDA-GLU) and new metal group-glutaraldehyde support (Aga-IDA-Ni 2+ -EDA-GLU) was, respectively, 5.8 and 1.8-fold higher than the free LipC12. The high improvement of lipase LipC12 after immobilization (Aga-C8-GLU/LipC12) is important because it permits the transformation of a mesophilic enzyme into an enzyme with properties that are similar to those of thermophilic organisms, such as Bacillus thermocatenolatus lipase (BTL) and Thermus thermophilus lipase (TTL) [44,45]. Figure 5 shows the thermal stability of the immobilized KlBgal and EaBglA at 40 °C. For KlBgal preparations, only Aga-IDA-Co 2+ -EDA-GLU promoted a significant stabilization (4.8-fold factor) compared to the free form (half-life of 33 min at 40 °C, Figure 5a and Table 4). All other preparations showed a lower stabilization than the free form (Aga-C8-GLU 0.52 and Aga-EDA-GLU 0.36-fold factor, Table 4). For studies with the immobilized EaBglA, it was observed that the Aga-IDA-Co 2+ -EDA-GLU (half-life time of 230 min) showed the best result, which was more stable (5.9-fold factor) than the traditional GLU preparation (Aga-EDA-GLU), and 10.9-times more stable than free enzyme (half-life time of 21 min at 40 °C, Figure 5b and Table 4). The Aga-EDA-GLU preparation also presents a half-life of 39 min (1.8-fold factor more stable than free form). In all cases, the stability of different preparations was strongly dependent on the support used in the immobilization process.  ) Aga-EDA-GLU; ( enzymes in almost all areas and it is one of the main aims when the immobilization of enzymes is performed [6]. Therefore, the thermal stabilities of different immobilized preparations and free forms were studied in 100 mM phosphate buffers, pH 7 using distinct temperatures. Figure 4a shows the tests using lipase CRL preparations. The stability of the adsorbed preparation (Octyl-Sepharose) was considerably similar to the free form with around 3.6-fold stabilization. The maximal stability was obtained with the traditional glutaraldehyde support (Aga-EDA-GLU) with a half-life of around 1560 min after incubation at 50 °C ( Table 4). Considering that the half-life of the free CRL was 21 min, the stabilization factor obtained was around 74.3-fold. CRL immobilized on Aga-C8-GLU and Aga-IDA-Ni 2+ -EDA-GLU was 5.7 and 2.8-fold more stable than the free form. The thermal stability of lipase LipC12 preparations was assessed by incubation at 70 °C ( Figure  4b, Table 4). At this temperature, the native free LipC12 had a half-life time of 13 min. The most stable preparation was found using the new hydrophobic group-glutaraldehyde support (Aga-C8-GLU) with a stabilization factor around 32.3-fold compared to the free form. The stability of LipC12 immobilized on the traditional amino (Ag-EDA-GLU) and new metal group-glutaraldehyde support (Aga-IDA-Ni 2+ -EDA-GLU) was, respectively, 5.8 and 1.8-fold higher than the free LipC12. The high improvement of lipase LipC12 after immobilization (Aga-C8-GLU/LipC12) is important because it permits the transformation of a mesophilic enzyme into an enzyme with properties that are similar to those of thermophilic organisms, such as Bacillus thermocatenolatus lipase (BTL) and Thermus thermophilus lipase (TTL) [44,45]. Figure 5 shows the thermal stability of the immobilized KlBgal and EaBglA at 40 °C. For KlBgal preparations, only Aga-IDA-Co 2+ -EDA-GLU promoted a significant stabilization (4.8-fold factor) compared to the free form (half-life of 33 min at 40 °C, Figure 5a and Table 4). All other preparations showed a lower stabilization than the free form (Aga-C8-GLU 0.52 and Aga-EDA-GLU 0.36-fold factor, Table 4). For studies with the immobilized EaBglA, it was observed that the Aga-IDA-Co 2+ -EDA-GLU (half-life time of 230 min) showed the best result, which was more stable (5.9-fold factor) than the traditional GLU preparation (Aga-EDA-GLU), and 10.9-times more stable than free enzyme (half-life time of 21 min at 40 °C, Figure 5b and Table 4). The Aga-EDA-GLU preparation also presents a half-life of 39 min (1.8-fold factor more stable than free form). In all cases, the stability of different preparations was strongly dependent on the support used in the immobilization process.  ) Octyl-Sepharose; and ( enzymes in almost all areas and it is one of the main aims when the immobilization of enzymes is performed [6]. Therefore, the thermal stabilities of different immobilized preparations and free forms were studied in 100 mM phosphate buffers, pH 7 using distinct temperatures. Figure 4a shows the tests using lipase CRL preparations. The stability of the adsorbed preparation (Octyl-Sepharose) was considerably similar to the free form with around 3.6-fold stabilization. The maximal stability was obtained with the traditional glutaraldehyde support (Aga-EDA-GLU) with a half-life of around 1560 min after incubation at 50 °C ( Table 4). Considering that the half-life of the free CRL was 21 min, the stabilization factor obtained was around 74.3-fold. CRL immobilized on Aga-C8-GLU and Aga-IDA-Ni 2+ -EDA-GLU was 5.7 and 2.8-fold more stable than the free form. The thermal stability of lipase LipC12 preparations was assessed by incubation at 70 °C ( Figure  4b, Table 4). At this temperature, the native free LipC12 had a half-life time of 13 min. The most stable preparation was found using the new hydrophobic group-glutaraldehyde support (Aga-C8-GLU) with a stabilization factor around 32.3-fold compared to the free form. The stability of LipC12 immobilized on the traditional amino (Ag-EDA-GLU) and new metal group-glutaraldehyde support (Aga-IDA-Ni 2+ -EDA-GLU) was, respectively, 5.8 and 1.8-fold higher than the free LipC12. The high improvement of lipase LipC12 after immobilization (Aga-C8-GLU/LipC12) is important because it permits the transformation of a mesophilic enzyme into an enzyme with properties that are similar to those of thermophilic organisms, such as Bacillus thermocatenolatus lipase (BTL) and Thermus thermophilus lipase (TTL) [44,45]. Figure 5 shows the thermal stability of the immobilized KlBgal and EaBglA at 40 °C. For KlBgal preparations, only Aga-IDA-Co 2+ -EDA-GLU promoted a significant stabilization (4.8-fold factor) compared to the free form (half-life of 33 min at 40 °C, Figure 5a and Table 4). All other preparations showed a lower stabilization than the free form (Aga-C8-GLU 0.52 and Aga-EDA-GLU 0.36-fold factor, Table 4). For studies with the immobilized EaBglA, it was observed that the Aga-IDA-Co 2+ -EDA-GLU (half-life time of 230 min) showed the best result, which was more stable (5.9-fold factor) than the traditional GLU preparation (Aga-EDA-GLU), and 10.9-times more stable than free enzyme (half-life time of 21 min at 40 °C, Figure 5b and Table 4). The Aga-EDA-GLU preparation also presents a half-life of 39 min (1.8-fold factor more stable than free form). In all cases, the stability of different preparations was strongly dependent on the support used in the immobilization process.  enzymes in almost all areas and it is one of the main aims when the immobilization of enzymes is performed [6]. Therefore, the thermal stabilities of different immobilized preparations and free forms were studied in 100 mM phosphate buffers, pH 7 using distinct temperatures. Figure 4a shows the tests using lipase CRL preparations. The stability of the adsorbed preparation (Octyl-Sepharose) was considerably similar to the free form with around 3.6-fold stabilization. The maximal stability was obtained with the traditional glutaraldehyde support (Aga-EDA-GLU) with a half-life of around 1560 min after incubation at 50 °C ( Table 4). Considering that the half-life of the free CRL was 21 min, the stabilization factor obtained was around 74.3-fold. CRL immobilized on Aga-C8-GLU and Aga-IDA-Ni 2+ -EDA-GLU was 5.7 and 2.8-fold more stable than the free form.
The thermal stability of lipase LipC12 preparations was assessed by incubation at 70 °C ( Figure  4b, Table 4). At this temperature, the native free LipC12 had a half-life time of 13 min. The most stable preparation was found using the new hydrophobic group-glutaraldehyde support (Aga-C8-GLU) with a stabilization factor around 32.3-fold compared to the free form. The stability of LipC12 immobilized on the traditional amino (Ag-EDA-GLU) and new metal group-glutaraldehyde support (Aga-IDA-Ni 2+ -EDA-GLU) was, respectively, 5.8 and 1.8-fold higher than the free LipC12. The high improvement of lipase LipC12 after immobilization (Aga-C8-GLU/LipC12) is important because it permits the transformation of a mesophilic enzyme into an enzyme with properties that are similar to those of thermophilic organisms, such as Bacillus thermocatenolatus lipase (BTL) and Thermus thermophilus lipase (TTL) [44,45]. Figure 5 shows the thermal stability of the immobilized KlBgal and EaBglA at 40 °C. For KlBgal preparations, only Aga-IDA-Co 2+ -EDA-GLU promoted a significant stabilization (4.8-fold factor) compared to the free form (half-life of 33 min at 40 °C, Figure 5a and Table 4). All other preparations showed a lower stabilization than the free form (Aga-C8-GLU 0.52 and Aga-EDA-GLU 0.36-fold factor, Table 4). For studies with the immobilized EaBglA, it was observed that the Aga-IDA-Co 2+ -EDA-GLU (half-life time of 230 min) showed the best result, which was more stable (5.9-fold factor) than the traditional GLU preparation (Aga-EDA-GLU), and 10.9-times more stable than free enzyme (half-life time of 21 min at 40 °C, Figure 5b and Table 4). The Aga-EDA-GLU preparation also presents a half-life of 39 min (1.8-fold factor more stable than free form). In all cases, the stability of different preparations was strongly dependent on the support used in the immobilization process.  zymes in almost all areas and it is one of the main aims when the immobilization of enzymes is erformed [6]. Therefore, the thermal stabilities of different immobilized preparations and free forms ere studied in 100 mM phosphate buffers, pH 7 using distinct temperatures. Figure 4a shows the sts using lipase CRL preparations. The stability of the adsorbed preparation (Octyl-Sepharose) was nsiderably similar to the free form with around 3.6-fold stabilization. The maximal stability was tained with the traditional glutaraldehyde support (Aga-EDA-GLU) with a half-life of around 1560 in after incubation at 50 °C ( Table 4). Considering that the half-life of the free CRL was 21 min, the abilization factor obtained was around 74.3-fold. CRL immobilized on Aga-C8-GLU and Aga-IDAi 2+ -EDA-GLU was 5.7 and 2.8-fold more stable than the free form. The thermal stability of lipase LipC12 preparations was assessed by incubation at 70 °C ( Figure  , Table 4). At this temperature, the native free LipC12 had a half-life time of 13 min. The most stable reparation was found using the new hydrophobic group-glutaraldehyde support (Aga-C8-GLU) ith a stabilization factor around 32.3-fold compared to the free form. The stability of LipC12 mobilized on the traditional amino (Ag-EDA-GLU) and new metal group-glutaraldehyde support ga-IDA-Ni 2+ -EDA-GLU) was, respectively, 5.8 and 1.8-fold higher than the free LipC12. The high provement of lipase LipC12 after immobilization (Aga-C8-GLU/LipC12) is important because it ermits the transformation of a mesophilic enzyme into an enzyme with properties that are similar those of thermophilic organisms, such as Bacillus thermocatenolatus lipase (BTL) and Thermus ermophilus lipase (TTL) [44,45]. Figure 5 shows the thermal stability of the immobilized KlBgal and EaBglA at 40 °C. For KlBgal reparations, only Aga-IDA-Co 2+ -EDA-GLU promoted a significant stabilization (4.8-fold factor) mpared to the free form (half-life of 33 min at 40 °C, Figure 5a and Table 4). All other preparations owed a lower stabilization than the free form (Aga-C8-GLU 0.52 and Aga-EDA-GLU 0.36-fold ctor, Table 4). For studies with the immobilized EaBglA, it was observed that the Aga-IDA-Co 2+ -DA-GLU (half-life time of 230 min) showed the best result, which was more stable (5.9-fold factor) an the traditional GLU preparation (Aga-EDA-GLU), and 10.9-times more stable than free enzyme alf-life time of 21 min at 40 °C, Figure 5b and Table 4). The Aga-EDA-GLU preparation also presents half-life of 39 min (1.8-fold factor more stable than free form). In all cases, the stability of different reparations was strongly dependent on the support used in the immobilization process.  ) Aga-C8-GLU; ( mes in almost all areas and it is one of the main aims when the immobilization of enzymes is ormed [6]. Therefore, the thermal stabilities of different immobilized preparations and free forms studied in 100 mM phosphate buffers, pH 7 using distinct temperatures. Figure 4a shows the using lipase CRL preparations. The stability of the adsorbed preparation (Octyl-Sepharose) was iderably similar to the free form with around 3.6-fold stabilization. The maximal stability was ined with the traditional glutaraldehyde support (Aga-EDA-GLU) with a half-life of around 1560 after incubation at 50 °C (Table 4). Considering that the half-life of the free CRL was 21 min, the ilization factor obtained was around 74.3-fold. CRL immobilized on Aga-C8-GLU and Aga-IDA-EDA-GLU was 5.7 and 2.8-fold more stable than the free form. The thermal stability of lipase LipC12 preparations was assessed by incubation at 70 °C ( Figure  able 4). At this temperature, the native free LipC12 had a half-life time of 13 min. The most stable aration was found using the new hydrophobic group-glutaraldehyde support (Aga-C8-GLU) a stabilization factor around 32.3-fold compared to the free form. The stability of LipC12 obilized on the traditional amino (Ag-EDA-GLU) and new metal group-glutaraldehyde support -IDA-Ni 2+ -EDA-GLU) was, respectively, 5.8 and 1.8-fold higher than the free LipC12. The high rovement of lipase LipC12 after immobilization (Aga-C8-GLU/LipC12) is important because it its the transformation of a mesophilic enzyme into an enzyme with properties that are similar ose of thermophilic organisms, such as Bacillus thermocatenolatus lipase (BTL) and Thermus ophilus lipase (TTL) [44,45]. Figure 5 shows the thermal stability of the immobilized KlBgal and EaBglA at 40 °C. For KlBgal arations, only Aga-IDA-Co 2+ -EDA-GLU promoted a significant stabilization (4.8-fold factor) pared to the free form (half-life of 33 min at 40 °C, Figure 5a and Table 4). All other preparations ed a lower stabilization than the free form (Aga-C8-GLU 0.52 and Aga-EDA-GLU 0.36-fold r, Table 4). For studies with the immobilized EaBglA, it was observed that the Aga-IDA-Co 2+ --GLU (half-life time of 230 min) showed the best result, which was more stable (5.9-fold factor) the traditional GLU preparation (Aga-EDA-GLU), and 10.9-times more stable than free enzyme -life time of 21 min at 40 °C, Figure 5b and Table 4). The Aga-EDA-GLU preparation also presents lf-life of 39 min (1.8-fold factor more stable than free form). In all cases, the stability of different arations was strongly dependent on the support used in the immobilization process.  ) Aga-EDA-GLU; and ( performed [6]. Therefore, the thermal stabilities of d were studied in 100 mM phosphate buffers, pH 7 u tests using lipase CRL preparations. The stability of considerably similar to the free form with around obtained with the traditional glutaraldehyde suppo min after incubation at 50 °C (Table 4). Considering stabilization factor obtained was around 74.3-fold. C Ni 2+ -EDA-GLU was 5.7 and 2.8-fold more stable tha The thermal stability of lipase LipC12 prepara 4b, Table 4). At this temperature, the native free LipC preparation was found using the new hydrophob with a stabilization factor around 32.3-fold comp immobilized on the traditional amino (Ag-EDA-GL (Aga-IDA-Ni 2+ -EDA-GLU) was, respectively, 5.8 an improvement of lipase LipC12 after immobilizatio permits the transformation of a mesophilic enzyme to those of thermophilic organisms, such as Baci thermophilus lipase (TTL) [44,45]. Figure 5 shows the thermal stability of the imm preparations, only Aga-IDA-Co 2+ -EDA-GLU prom compared to the free form (half-life of 33 min at 40 ° showed a lower stabilization than the free form ( factor, Table 4). For studies with the immobilized EDA-GLU (half-life time of 230 min) showed the be than the traditional GLU preparation (Aga-EDA-GL (half-life time of 21 min at 40 °C, Figure 5b and Table  a half-life of 39 min (1.8-fold factor more stable tha preparations was strongly dependent on the suppo (a) ) free enzyme. The results are expressed as the average of triplicate assays ± the standard error of the mean.      [46]. The stability factors were done in relation to the free form.   [46]. The stability factors were done in relation to the free form.   [46]. The stability factors were done in relation to the free form.   [46]. The stability factors were done in relation to the free form.       ) free enzyme. The results are expressed as the average of triplicate assays ± the standard error of the mean. C8 (1-octanothiol); Ni 2+ (nickel); Co 2+ (cobalt); EDA (ethylenediamine); IDA (iminodiacetic acid); Aga (Agarose beads); and GLU (glutaraldehyde). a Thermal stability of lipase CRL preparations was checked by incubation at 50 • C. b Thermal stability of lipase LipC12 preparations was checked by incubation at 70 • C. c Thermal stability of Klgal and EaglA preparations was checked by incubation at 40 • C. Aliquots were periodically withdrawn for quantification of the residual enzymatic activity to estimate the half-life according to Henley and Sadana [46]. The stability factors were done in relation to the free form.

Standard Determination of Enzymatic Activities
β-Glucosidase assays. The activity was determined using 4-nitrophenyl β-D-glucopyranoside (p-NPG) as substrate. The free or immobilized enzyme was added to a mixture solution (5 mM p-NPG in 50 mM sodium phosphate buffer at pH 7) and the increase in absorbance was monitored at 410 nm (pH 7.0, ε 410 nm = 7320 M −10 cm −1 ) [47]. One unit (U) of enzyme was defined as the amount of enzyme required to release1 µmoL of p-nitrophenol per min and the specific activity was defined as the number of units per mg of protein.
Lipase assays. The activities were determined using p-nitrophenyl propionate (p-NPP) as substrate. The free or immobilized enzymes were added to a mixture solution (0.4 mM p-NPP in 50 mM sodium phosphate buffer at pH 7) and the increase in absorbance was monitored at 348 nm (pH 7.0, ε 348 nm = 5150 M −1 cm −1 ) [29]. One unit (U) of enzyme was defined as the amount of enzyme required to release 1 µmoL of p-nitrophenol per min and the specific activity was defined as the number of units per mg of protein.
β-Galactosidase assays. The activity was determined using O-nitro-phenyl-β-D-galactopyranoside (O-NPG) as substrate. The free or immobilized enzyme was added to a mixture solution (5 mM O-NPG in 50 mM potassium phosphate buffer at pH 7 with the addition of 2 mM MgCl 2 ) and the increase in absorbance was monitored at 410 nm (pH 7.0, ε 410 nm = 3500 M −1 cm −1 ) [48]. One unit (U) of enzyme was defined as the amount of enzyme required to release 1 µmoL of O-NP per min and the specific activity was defined as the number of units per mg of protein.

Activation of Agarose-Based Beads
Epoxy-activated agarose was prepared by the reaction of the hydroxyl groups on the agarose beads with epichlorohydrin, as described by Mateo et al. [24]. Under gentle agitation and in an ice bath, 50 g of agarose 4 BCL, previously washed with distilled water and vacuum dried, was mixed with 220 mL of distilled water, 16.4 g of NaOH, 1.0 g of sodium borohydride (NaBH 4 ), 80 mL of acetone, and 55 mL of epichlorohydrin. The suspension was incubated for 16 h at 25 • C, and then washed with the excess of distilled water, vacuum dried, and stored at 4 • C.
The number of epoxy/ligand groups was calculated from the difference in periodate consumption between the hydrolyzed support and the initial epoxy support. Periodate consumption was quantified using potassium iodide, as previously described [49].

Modification of the Epoxy-Activated Agarose with Different Reactive Groups Cationic Supports
Epoxy-activated agarose (10 g) was modified using 1 M triethylamine solution (acetone/water, 50:50 v/v) for 18 h at 25 • C [24]. After this, the support was oxidized with 100 mL of 100 mM sodium periodate (NaIO 4 ) for 2 h at 25 • C, washed with distilled water, vacuum dried and stored at 4 • C.

Support Activated with the Glutaraldehyde Group
The support activated using the glutaraldehyde group was prepared as previously described by Betancor et al. [35]. Briefly, the epoxy-activated agarose (10 g) was hydrolyzed with 100 mL of 1 M

Immobilization of Enzymes
For all immobilization experiments, β-glucosidase (EaBglA) and lipases (CRL and LipC12) were solubilized in 200 mM sodium phosphate buffer at pH 7, and β-galactosidase (KlBgal) was solubilized in 200 mM potassium phosphate buffer at pH 7 with the addition of 2 mM MgCl 2 . Then, different immobilization supports were suspended in an enzyme solution (1 g support: 10 mL of enzyme solution in the immobilization buffer, 1 mg g −1 of support). The immobilization suspensions were gently stirred at 25 • C. Periodically, samples of supernatants were withdrawn, and their enzymatic activities were analyzed. The immobilization was considered complete when there was no activity in the supernatant.
After immobilization, the enzymes were washed with phosphate buffers. The immobilization efficiency (IE, %) was calculated as Equation (1): where Ai is the amount activity (U) of the enzyme solution before immobilization and A f is the amount activity (U) remaining in the supernatant at the end of the immobilization procedure.
The recovered activity (R, %) was calculated as Equation (2): where A o is the as the ratio between the real specific activity (U g −1 of support) of immobilized preparation and A T is the theoretical specific activity of the immobilized preparation (U g −1 of support). Escherichia coli BL21 crude extract was used as a multi-protein model system during the immobilization tests with the different supports. One gram of activated supports was suspended in 10 mL of E. coli BL21 crude cell extract (5 mg g −1 of support). The adsorption of proteins was monitored using the Bradford method.

Fourier Transform Infrared (FTIR)
Identification of the chemical groups on the untreated and treated support was performed using the FTIR spectroscopy Vertex 70 Bruker. The spectra were obtained with a wave number range from 4000 to 400 cm −1 at a resolution of 4 cm −1 over 32 cumulative scans. FTIR spectra were not used as a quantitative relationship, but only as a qualitative reference.

Desorption Process Analysis
Samples of 1 g of immobilized enzymes were suspended in 3 mL of phosphate buffer (5 mM, at pH 7). Then, Triton ×100, NaCl and imidazol were added to a final concentration of 0.5% (v/v), 1.0 M and 0.5 M, respectively. The different preparations were gently stirred at 25 • C and, periodically, samples from supernatants were withdrawn and their enzymatic activities were analyzed.

Thermal Stability
Thermal stabilities were assessed by incubation in sodium phosphate buffer (100 mM, pH 7 at 40 • C) for β-glucosidase, potassium phosphate buffer containing 2 mM MgCl 2 (100 mM, pH 7 at 40 • C) for β-galactosidase, and sodium phosphate buffer (100 mM, pH 7 at 50 • C or 70 • C) for lipases (CRL and LipC12). Aliquots were periodically withdrawn for the quantification of the residual enzymatic activity to estimate the half-life time according to Henley and Sadana [46]. The relative activities of free and immobilized forms without incubation were defined as control and attributed a relative activity of 100%. Inactivation parameters were determined from the best-fit model of the experimental data which was the one based on a two-stage series inactivation mechanism with residual activity. Half-lives (time in which the residual enzyme activity is half of its initial value) were used to compare the stability of the different preparations, being determined by the interpolation from the respective models described in [52].

Conclusions
The new generation of supports proposed in this study overcomes some of problems associated with the immobilization of enzymes intrinsically unstable at alkaline pH. The new supports were capable of immobilizing distinct enzymes with significant recovered activity values. Thus, this strategy allowed obtaining heterogeneous catalysts with different catalytic properties (i.e., activity and stability) under mild conditions (neutral pH).
The strategy of this study can be applied to other supports activated in similar ways, such as cellulose, permitting the improvement of the materials or the use of supports with distinct morphologies or properties.