Immobilization of the Peroxygenase from Agrocybe aegerita . The Effect of the Immobilization pH on the Features of an Ionically Exchanged Dimeric Peroxygenase

: This paper outlines the immobilization of the recombinant dimeric unspeciﬁc peroxygenase from Agrocybe aegerita (r Aae UPO). The enzyme was quite stable (remaining unaltered its activity after 35 h at 47 ◦ C and pH 7.0). Phosphate destabilized the enzyme, while glycerol stabilized it. The enzyme was not immobilized on glyoxyl-agarose supports, while it was immobilized albeit in inactive form on vinyl-sulfone-activated supports. r Aae UPO immobilization on glutaraldehyde pre-activated supports gave almost quantitative immobilization yield and retained some activity, but the biocatalyst was very unstable. Its immobilization via anion exchange on PEI supports also produced good immobilization yields, but the r Aae UPO stability dropped. However, using aminated agarose, the enzyme retained stability and activity. The stability of the immobilized enzyme strongly depended on the immobilization pH, being much less stable when r Aae UPO was adsorbed at pH 9.0 than when it was immobilized at pH 7.0 or pH 5.0 (residual activity was almost 0 for the former and 80% for the other preparations), presenting stability very similar to that of the free enzyme. This is a very clear example of how the immobilization pH greatly affects the ﬁnal biocatalyst performance.


Introduction
The modern use of enzymes as catalysts in the organic chemistry industry is receiving growing interest due to their high activity under mild conditions (that makes them suitable from an environmental point of view) along with their high specificity and selectivity (that can save purification and protection/deprotection steps) [1][2][3][4]. Nevertheless, enzymes have been optimized by nature to fulfill physiological functions and are not necessarily well suited for the needs of industrial biotransformation [5]. To be efficient catalysts for chemical transformations, enzymes need to be active and stable under the reaction conditions (which may differ significantly from the physiological ones) while at the same time converting non-physiological substrates [6][7][8]. Recovery and reuse of enzymes after the reaction may be also a problem, as they are water-soluble catalysts [9][10][11].
These factors have considerably delayed the industrial implementation of enzymes; fortunately, there are many tools to improve the enzyme features that have experienced significant advancements in recent years, offering new solutions to solve natural enzyme limitations [12]. Advances in enzyme modelling along with site-directed mutagenesis can Scheme 1. Comparison of P450 monooxygenases (a) and peroxygenases (b) as catalysts for selective oxyfunctionalization reactions. Both enzyme classes utilize compound I (Cpd I, (c)) as an oxygen transfer agent.
Thus, in this contribution, the range of conditions under which the enzyme may b handled was studied as well as the effect of some additives on its stability. rAaeUPO im mobilization on different supports via covalent immobilization (using glyoxyl, vinyl su fone or glutaraldehyde-activated agarose) or physical exchange (on PEI or monoam noethyl-N-aminoethyl (MANAE)-activated agarose beads) was assayed. The expresse activity and stability of the biocatalysts were evaluated.

Handling of rAaeUPO
Before starting the immobilization trials, the range of conditions where the enz was stable and can be handled must be determined. The enzyme activity was determ under the conditions reported in literature (pH 4.4) [63,64]. At pH 7.0, the observed a ity versus ABTS was only 10% of this maximum activity. The enzyme was very stab Thus, in this contribution, the range of conditions under which the enzyme may be handled was studied as well as the effect of some additives on its stability. rAaeUPO immobilization on different supports via covalent immobilization (using glyoxyl, vinyl sulfone or glutaraldehyde-activated agarose) or physical exchange (on PEI or monoaminoethyl-Naminoethyl (MANAE)-activated agarose beads) was assayed. The expressed activity and stability of the biocatalysts were evaluated.

Handling of rAaeUPO
Before starting the immobilization trials, the range of conditions where the enzyme was stable and can be handled must be determined. The enzyme activity was determined under the conditions reported in literature (pH 4.4) [63,64]. At pH 7.0, the observed activity versus ABTS was only 10% of this maximum activity. The enzyme was very stable at pH 7.0, with not decrease in its activity even at 47 • C for 35 h. At pH 7.0, a strong dependence of the enzyme stability on the buffer was found. At 50 • C, the enzyme lost 50% of the activity after 3 h when 50 mM sodium phosphate was employed as buffer (Figure 2A), while if Tris or HEPES buffers were employed, the enzyme retained 100% of the initial activity. At 57 • C, the enzyme was inactivated at a similar rate using HEPES or Tris ( Figure 2B). This suggested that phosphate presented a negative effect for the enzyme stability. This negative effect of phosphate anions has been described for other enzymes even in an immobilized form [99,100]. This enzyme is dimeric; thus, we investigated whether the enzyme stability depends on the enzyme concentrations in both phosphate and Tris buffers ( Figure 3). In both buffers, the enzyme concentration did not present a relevant impact on the enzyme stability. This suggested that reversible dissociation was not the first step of the enzyme inactivation at pH 7.0 and that the negative effect of phosphate anions on enzyme stability does not favor this dissociation [32]. Considering that the Heme B group is deeply buried in the enzyme structure, it was most unlikely that its release could be the first step of the enzyme inactivation. Therefore, enzyme inactivation seems to be initialized by a distortion of the tridimensional structure or an induction of an erroneous enzyme assembly. This enzyme is dimeric; thus, we investigated whether the enzyme stability depends on the enzyme concentrations in both phosphate and Tris buffers ( Figure 3). In both buffers, the enzyme concentration did not present a relevant impact on the enzyme stability. This suggested that reversible dissociation was not the first step of the enzyme inactivation at pH 7.0 and that the negative effect of phosphate anions on enzyme stability does not favor this dissociation [32]. Considering that the Heme B group is deeply buried in the enzyme structure, it was most unlikely that its release could be the first step of the enzyme inactivation. Therefore, enzyme inactivation seems to be initialized by a distortion of the tridimensional structure or an induction of an erroneous enzyme assembly. ers, the enzyme concentration did not present a relevant impact on the enzyme stability. This suggested that reversible dissociation was not the first step of the enzyme inactivation at pH 7.0 and that the negative effect of phosphate anions on enzyme stability does not favor this dissociation [32]. Considering that the Heme B group is deeply buried in the enzyme structure, it was most unlikely that its release could be the first step of the enzyme inactivation. Therefore, enzyme inactivation seems to be initialized by a distortion of the tridimensional structure or an induction of an erroneous enzyme assembly. At pH 10.0 (pH necessary to immobilize the enzyme on glyoxyl-agarose beads) [40] and 25 °C, the enzyme was rapidly inactivated (residual activity was only 5% after 4 h), preventing immobilization of the enzyme under these conditions ( Figure 4). The addition of glycerol to the medium significantly stabilized the enzyme, and this was more significant when the glycerol concentration increased (the enzyme retained over 60% of the initial activity after 4 h in 40% glycerol) ( Figure 4). This enzyme stabilizing effect of glycerol has been recently examined, and although it is not universal, it is a quite general phenomenon [44]. At 4 °C and in 40% glycerol, the enzyme maintained around 75% of its activity after 4 h ( Figure 4). At pH 10.0 (pH necessary to immobilize the enzyme on glyoxyl-agarose beads) [40] and 25 • C, the enzyme was rapidly inactivated (residual activity was only 5% after 4 h), preventing immobilization of the enzyme under these conditions ( Figure 4). The addition of glycerol to the medium significantly stabilized the enzyme, and this was more significant when the glycerol concentration increased (the enzyme retained over 60% of the initial activity after 4 h in 40% glycerol) ( Figure 4). This enzyme stabilizing effect of glycerol has been recently examined, and although it is not universal, it is a quite general phenomenon [44]. At 4 • C and in 40% glycerol, the enzyme maintained around 75% of its activity after 4 h ( Figure 4).  We therefore found valid conditions to handle the enzyme at n values, both necessary for the immobilization strategies that we in enzyme.  We therefore found valid conditions to handle the enzyme at neutral and alkaline pH values, both necessary for the immobilization strategies that we intended to apply to this enzyme.

Covalent Immobilization of rAaeUPO
rAaeUPO was not immobilized in glyoxyl-agarose at pH 10.05 ( Figure S3) (immobilization yield was almost 0). Looking to the structure of the enzyme (Figure 1), it can be seen that although the enzyme is rich in Lys residues areas, the polysaccharide chains may hinder the simultaneous interaction of several primary amino groups of the enzyme with the support, thus avoiding the enzyme immobilization in glyoxyl-agarose. The considerable dilution of the enzyme solution (almost 1000 folds) discards the possibility of some aminated compounds on the enzyme solution that could prevent the enzyme immobilization [101].
The enzyme was not significantly immobilized on supports activated with DVS at pH 5.0 and 7.0 (immobilization yield under 20%) ( Figure 5A,B). Moreover, the small immobilized fraction decreased its expressed activity after immobilization by more than 50%, giving very poor final activity. At pH 9.0 ( Figure 5C), a higher immobilization yield (over 90%) was observed, but the enzyme became almost fully inactivated after immobilization (expressed activity was under 10%). This enzyme inactivation upon immobilization was only partially reduced by adding 40% glycerol (expressed activity was 40%) ( Figure 5D), but this additive also reduced the immobilization yield (to 50%). After 24 h, the expressed activity decreased to less than 5%, and if the biocatalysts were subjected to a blocking step (with ethylenediamine or aspartic acid), a necessary step to obtain a chemically inert support surface [38], the enzyme expressed activity fell below the detection limit. Thus, both covalent immobilization techniques were discarded for this enzyme. Next, the enzyme was immobilized in pre-activated amino-glutaraldehyde supports at pH 5.0 and 7.0 ( Figure 6). As the first step of the immobilization of enzymes in this support is the physical ion exchange of the enzyme on the support [48,102], this process was quite rapid, and the immobilization yield reached 100%. However, a significant percentage of activity was lost at both pH values, with expressed activities under 25% after 24 h of incubation. Immobilization of the enzyme at pH 9.0 on this support resulted in an almost instantaneous enzyme inactivation. This decrease in enzyme activity could be explained by a certain enzyme-support multipoint covalent attachment. This should result in more rigid enzyme structure, and the decrease in activity could be compensated if this Thus, both covalent immobilization techniques were discarded for this enzyme. Next, the enzyme was immobilized in pre-activated amino-glutaraldehyde supports at pH 5.0 and 7.0 ( Figure 6). As the first step of the immobilization of enzymes in this support is the physical ion exchange of the enzyme on the support [48,102], this process was quite rapid, and the immobilization yield reached 100%. However, a significant percentage of activity was lost at both pH values, with expressed activities under 25% after 24 h of incubation. Immobilization of the enzyme at pH 9.0 on this support resulted in an almost instantaneous enzyme inactivation. This decrease in enzyme activity could be explained by a certain enzyme-support multipoint covalent attachment. This should result in more rigid enzyme structure, and the decrease in activity could be compensated if this is correlated with an increase in enzyme stability [30,102,103]. However, as shown in Figure 7, the enzyme immobilized on this support was even less stable than the free enzyme. This decrease in the enzyme stability after immobilization on the support could be explained by the presence of hydrophobic and ionic groups near the enzyme surface that could stabilize some incorrect enzyme structures [35]. Another possibility is the production of some distortion of the assembly between both enzyme subunits caused by the multipoint covalent attachment, which could drive the higher tension of the multimeric structure and lower enzyme stability [33]. In any case, this immobilization strategy does not appear to be adequate to prepare an immobilized biocatalyst of this enzyme either.

Immobilization of rAaeUPO Via Ion Exchange
After the failure of the covalent immobilization protocols assayed above to produce a very active and stable immobilized biocatalyst using some covalent immobilization protocols, we assayed the rAaeUPO immobilization on MANAE-, PEI-10,000-and PEI-25,000activated agarose at pH 7.0. The enzyme exposes various anionic groups in its structure (see Figure 1) to permit the enzyme anion exchange on these cationic supports. Figure 8 shows the immobilization course in these supports. In all cases, immobilization is very rapid (immobilization yield was 100% after only 15 min). However, while in MANAEactivated agarose the activity is fairly preserved after immobilization (around 80%), and in PEI coated supports, a significant decrease of the enzyme activity is observed. After 48 h, these preparations did not maintain any activity even if maintained in the refrigerator

Immobilization of rAaeUPO Via Ion Exchange
After the failure of the covalent immobilization protocols assayed above to pr a very active and stable immobilized biocatalyst using some covalent immobilizatio tocols, we assayed the rAaeUPO immobilization on MANAE-, PEI-10,000-and PEI-2 activated agarose at pH 7.0. The enzyme exposes various anionic groups in its str (see Figure 1) to permit the enzyme anion exchange on these cationic supports. Fi

Immobilization of rAaeUPO Via Ion Exchange
After the failure of the covalent immobilization protocols assayed above to produce a very active and stable immobilized biocatalyst using some covalent immobilization protocols, we assayed the rAaeUPO immobilization on MANAE-, PEI-10,000-and PEI-25,000-activated agarose at pH 7.0. The enzyme exposes various anionic groups in its Catalysts 2021, 11, 560 9 of 21 structure (see Figure 1) to permit the enzyme anion exchange on these cationic supports. Figure 8 shows the immobilization course in these supports. In all cases, immobilization is very rapid (immobilization yield was 100% after only 15 min). However, while in MANAE-activated agarose the activity is fairly preserved after immobilization (around 80%), and in PEI coated supports, a significant decrease of the enzyme activity is observed. After 48 h, these preparations did not maintain any activity even if maintained in the refrigerator at 6 • C just after immobilization. However, after that time, the MANAE biocatalyst maintained its activity intact even at 25 • C. This negative result observed when immobilizing the enzyme in PEI coated agarose could be explained if the flexible polymer was able to accede to internal pockets of the protein, which can be relevant for the enzyme activity or stability or even accelerate erroneous subunit assembly [56]. The amino groups attached to the surface of MANAE agarose should be unable to penetrate these internal pockets and would not produce this negative effect on enzyme stability, and the enzyme activity would be maintained. As shown in Figure 9, this biocatalyst, when inactivated at 57 • C and pH 7.0, has a similar stability to that of the free enzyme, or even marginally higher. Considering the good properties of this biocatalyst compared to the covalent preparations, we tried to improve its performance using two different strategies.
First, the immobilization on MANAE agarose was also performed at pH 5.0 and 9.0 ( Figure 10) and compared to that at pH 7.0, as it has been shown that the immobilization pH may greatly alter the final properties of the enzymes immobilized via anion exchange [48,61,102]. Compared to the immobilization of the enzyme at pH 7.0, the main difference is a higher expressed activity when immobilizing the enzyme at pH 5.0 (over 95%), while at pH 9.0 the expressed activity is around 80%, similar to the results obtained at pH 7 ( Figure 8). However, immobilization yield was 90% for the immobilization performed at pH 5.0 and almost 100% if the immobilization was performed at pH 9. At pH 5.0, the cationic nature of the support is reinforced, while the anionic nature of the enzyme is de-creased (the theoretic isoelectric point of the enzyme from the enzyme primary sequence is 5.325 (http://isoelectric.org/ (accessed on 04/04/2021))), and perhaps a small proportion of the enzyme molecules, which are highly glycosylated, cannot immobilize on MANAE supports under this less favorable pH value.  Considering the good properties of this biocatalyst compared to the covalent preparations, we tried to improve its performance using two different strategies.
First, the immobilization on MANAE agarose was also performed at pH 5.0 and 9.0 ( Figure 10) and compared to that at pH 7.0, as it has been shown that the immobilization pH may greatly alter the final properties of the enzymes immobilized via anion exchange [48,61,102]. Compared to the immobilization of the enzyme at pH 7.0, the main difference is a higher expressed activity when immobilizing the enzyme at pH 5.0 (over 95%), while at pH 9.0 the expressed activity is around 80%, similar to the results obtained at pH 7 ( Figure 8). However, immobilization yield was 90% for the immobilization performed at pH 5.0 and almost 100% if the immobilization was performed at pH 9. At pH 5.0, the cationic nature of the support is reinforced, while the anionic nature of the enzyme is decreased (the theoretic isoelectric point of the enzyme from the enzyme primary sequence  The thermal inactivation courses of the 3 biocatalysts of the enzyme immobilized in MANAE agarose beads are shown in Figure 11. The enzyme immobilized at pH 5.0 and 7.0 gave very similar inactivation profiles to that of the free enzyme, while the enzyme immobilized at pH 9.0 was much less stable. This drastic difference of enzyme stability of enzymes immobilized in the same ion exchanger, but at different pH values, could be related to two factors: a different orientation of the enzyme in the support (involving different areas of the protein in the absorption when the immobilization pH changes and the ionization of the enzyme groups is altered) or the immobilization of different enzyme forms induced by the pH value [48,61,102]. The immobilization of lipases via interfacial activation on the same support (also a reversible immobilization protocol), but under different conditions, has proved to give different enzyme conformations, which are maintained after the immobilization [104][105][106]. The thermal inactivation courses of the 3 biocatalysts of the enzyme immobilized in MANAE agarose beads are shown in Figure 11. The enzyme immobilized at pH 5.0 and 7.0 gave very similar inactivation profiles to that of the free enzyme, while the enzyme immobilized at pH 9.0 was much less stable. This drastic difference of enzyme stability of enzymes immobilized in the same ion exchanger, but at different pH values, could be related to two factors: a different orientation of the enzyme in the support (involving different areas of the protein in the absorption when the immobilization pH changes and the ionization of the enzyme groups is altered) or the immobilization of different enzyme forms induced by the pH value [48,61,102]. The immobilization of lipases via interfacial activation on the same support (also a reversible immobilization protocol), but under different conditions, has proved to give different enzyme conformations, which are maintained after the immobilization [104][105][106]. ferent areas of the protein in the absorption when the immobilization pH changes and the ionization of the enzyme groups is altered) or the immobilization of different enzyme forms induced by the pH value [48,61,102]. The immobilization of lipases via interfacial activation on the same support (also a reversible immobilization protocol), but under different conditions, has proved to give different enzyme conformations, which are maintained after the immobilization [104][105][106].  In any case, this effect of the immobilization pH, even when the enzyme is reversibly immobilized, shows the significance of this parameter when preparing immobilized biocatalysts with this immobilization strategy.
Finally, we tried to improve the immobilized enzyme stability by treating the biocatalysts with glutaraldehyde, with the objective to establish enzyme-support crosslinking and, perhaps, some enzyme intermolecular crosslinking [39,48,50,102].
First, we treated the free enzyme with this reagent at pH 7.0 and 25 • C, ( Figure 12A). The effect of this treatment on enzyme activity was not very significant using 0.1% or 1% glutaraldehyde; enzyme activity was maintained similarly to the activity of the enzyme incubated in absence of glutaraldehyde after 1 h (sufficient time to modify the primary amino groups with one glutaraldehyde molecule). When this treatment was performed on the enzymes adsorbed on MANAE support, the enzyme immobilized at pH 7.0 was fully inactivated after only 40 min, while the biocatalysts prepared by immobilizing the enzyme at pH 5.0 and 9.0 maintained more than 40% of the initial activity ( Figure 12B). This also suggested the very different features of rAaeUPO ionically exchanged at different pH values. However, 24 h after washing the glutaraldehyde-modified biocatalysts, these biocatalysts were also fully inactive. These much more drastic effects of the glutaraldehyde on the activity of the enzyme adsorbed on aminated supports than on the free enzyme under similar treatment were fairly similar to those found using a similar strategy to immobilize ficin [49].
Thus, from the different protocols assayed in this paper, the optimal protocol for the immobilization of this enzyme is the immobilization of rAaeUPO on MANAE support at pH 5.0. enzyme at pH 5.0 and 9.0 maintained more than 40% of the initial activity ( Figure 12B). This also suggested the very different features of rAaeUPO ionically exchanged at different pH values. However, 24 h after washing the glutaraldehyde-modified biocatalysts, these biocatalysts were also fully inactive. These much more drastic effects of the glutaraldehyde on the activity of the enzyme adsorbed on aminated supports than on the free enzyme under similar treatment were fairly similar to those found using a similar strategy to immobilize ficin [49]. Thus, from the different protocols assayed in this paper, the optimal protocol for the immobilization of this enzyme is the immobilization of rAaeUPO on MANAE support at pH 5.0.

Release of rAaeUPO from the MANAE Support and Reuse of the Support
One of the advantages of the immobilization of enzymes via ion exchange is the possibility of reuse of the support by releasing the enzyme after its inactivation. Figure 13 shows how the enzyme could be easily released from the support at both pH 4.4 and 7.0. Full enzyme release was achieved using 200 mM ammonium sulfate at pH 4.4 or 500 mM at pH 7.0. The support could be reused to immobilize fresh enzyme samples by 10 successive cycles, without any alteration of the enzyme performance.

Release of rAaeUPO from the MANAE Support and Reuse of the Support
One of the advantages of the immobilization of enzymes via ion exchange is the possibility of reuse of the support by releasing the enzyme after its inactivation. Figure 13 shows how the enzyme could be easily released from the support at both pH 4.4 and 7.0. Full enzyme release was achieved using 200 mM ammonium sulfate at pH 4.4 or 500 mM at pH 7.0. The support could be reused to immobilize fresh enzyme samples by 10 successive cycles, without any alteration of the enzyme performance.

Operational Stability of MANAE-rAaeUPO
rAaeUPO was immobilized on the MANAE support at pH 5.0, but in this instance using a loading of 3 mg/g. The immobilization yield was higher than 80% after 4 h. Using standard ABTS assay conditions, the enzyme was almost fully released from the support in the first cycle. Fifty mM citrate was enough to release most of the enzyme from the support (see Figure 13). Then, the experiment was repeated using 5 mM of sodium acetate at pH 5.0. Under these conditions, however, more than one third of the oxidized ABTS become adsorbed on the support and was released in the subsequent reuse, making reliable quantification of the new reaction course impossible. Washing steps using high ionic strength buffers to remove the adsorbed ABTS oxidation product released the adsorbed oxidized ABTS, but it also resulted in enzyme leaching. A practical solution to study the operational stability of the biocatalyst was to reduce the concentration of ABTS to 0.03 mM. A reference of the previously used biocatalyst was incubated in the reaction medium

Operational Stability of MANAE-rAaeUPO
rAaeUPO was immobilized on the MANAE support at pH 5.0, but in this instance using a loading of 3 mg/g. The immobilization yield was higher than 80% after 4 h. Using standard ABTS assay conditions, the enzyme was almost fully released from the support in the first cycle. Fifty mM citrate was enough to release most of the enzyme from the support (see Figure 13). Then, the experiment was repeated using 5 mM of sodium acetate at pH 5.0. Under these conditions, however, more than one third of the oxidized ABTS become adsorbed on the support and was released in the subsequent reuse, making reliable quantification of the new reaction course impossible. Washing steps using high ionic strength buffers to remove the adsorbed ABTS oxidation product released the adsorbed oxidized ABTS, but it also resulted in enzyme leaching. A practical solution to study the operational stability of the biocatalyst was to reduce the concentration of ABTS to 0.03 mM. A reference of the previously used biocatalyst was incubated in the reaction medium (in the absence of fresh substrates) to determine the absorbance produced by the adsorbed during the previous cycles and now released oxidized ABTS. Then, this supernatant absorbance was subtracted from the absorbance of the new reaction cycle. We were therefore able to confirm that the immobilized enzyme could be reused in at least 10 reaction cycles without any significant reduction to activity (Figure 14). The activity in the 10th cycle was 75% of that of the initial one, very likely due to some enzyme release from the support even under these mild conditions.
The recombinant and evolved unspecific peroxygenase from Agrocybe aegerita (rAae-UPO) was produced as described by Molina-Espeja et al. [64]. Figure S2 shows an SDS-PAGE of the relatively pure enzyme solution utilized in this research.

Enzyme Preparation
The enzyme was produced via expression in Pichia pastoris as described by Molina-Espeja et al. [64]. The "unfiltered" solution contained Pichia pastoris cells and was centrifuged at 10,000 rpm at 4 °C for 20 min. The clarified supernatant was aliquoted and stored at −20 °C. The protein concentration (8.85 ± 0.15 mg/mL) was determined through Brad- The experiment was performed to demonstrate that the reuse of the biocatalyst is possible, but it is clear that the oxidation of ABTS is not the target reaction for the utilization of this biocatalyst. It is evident that the biocatalyst will be of special interest for the application of the enzyme in anhydrous media, and there are many applications of this enzyme in these media, as discussed in the introduction. Its application in aqueous medium requires the use of very low concentrations of buffers and substrates.
The recombinant and evolved unspecific peroxygenase from Agrocybe aegerita (rAaeUPO) was produced as described by Molina-Espeja et al. [64]. Figure S2 shows an SDS-PAGE of the relatively pure enzyme solution utilized in this research.

Enzyme Preparation
The enzyme was produced via expression in Pichia pastoris as described by Molina-Espeja et al. [64]. The "unfiltered" solution contained Pichia pastoris cells and was centrifuged at 10,000 rpm at 4 • C for 20 min. The clarified supernatant was aliquoted and stored at −20 • C. The protein concentration (8.85 ± 0.15 mg/mL) was determined through Bradford's method using bovine serum albumin as standard [111]. The semi-purified sample was utilized in all experiments. The enzyme features are similar to those described by Molina-Espeja et al. [64]; we did not re-characterize the enzyme sample.

Determination of rAaeUPO Activity
The activity of rAaeUPO was determined via ABTS assay using a spectrophotometer thermoregulated at 25 • C with magnetic stirring. The ABTS assay was performed using as reaction medium 2.5 mL of 50 mM sodium citrate at pH 4.4 containing 2 mM H 2 O 2 and 0.5 mM ABTS, adding the desired amount of rAaeUPO solution. The oxidation of ABTS was monitored by the change in absorbance at 418 nm (ε 418 = 36,000 M −1 cm −1 under these conditions [81]). In this procedure, 25 µL of a rAaeUPO sample diluted in 50 mM Tris buffer at pH 7.0 was added to a cuvette containing 2 mL of the reaction solution, obtaining an activity of 1111 U per mL in the crude extract. One unit (U) of activity was defined as the amount of enzyme that oxidizes 1 µmol of substrate per minute under the specified conditions.

SDS-PAGE Analysis
The SDS-PAGE analysis was carried out using 12% polyacrylamide gel with 5% polyacrylamide as concentrating gel [112]. The rAaeUPO samples were suspended in the rupture buffer (4% (w/v) SDS and 5% (v/v) mercaptoethanol in 200 mM Tris buffer at pH 7.0) to obtain final protein concentrations of 0.1 and 0.2 mg/mL. Subsequently, 15 µL of the samples were loaded on the gels, using 8 µL of low molecular weight marker protein as standard. After the separation, the gels were stained with Coomassie brilliant blue R-250, 3 mM in 40% (v/v) ethanol, and 10% (v/v) acetic acid for 1 h.

Handling of the Enzyme
The enzyme solution was diluted in different buffers, at pH 7.0 (50 mM Tris, 50 mM HEPES or 50 mM sodium phosphate) and 10.0 (50 mM sodium carbonate) to a final concentration of 0.5 U/mL, using different temperatures. Periodically, samples were withdrawn and their activity assayed using ABTS as described above. In some instances, 20 or 40% (v/v) glycerol was added to the enzyme solutions.

Immobilization of rAaeUPO
All experiments were performed at least in triplicate, and the results are presented as the mean values. Immobilization was characterized by immobilization course, immobilization yield and expressed activity characterized as recommended by Boudrant et al. [113]. Immobilization yield is defined as the percentage of the offered enzyme that is immobilized on the support (determined by the enzyme remaining in the supernatant and the activity maintained by the reference). Expressed activity is the percentage of observed activity of the immobilized enzyme considering the expected activity of the biocatalyst from the immobilization yield as 100%.

Immobilization of rAaeUPO on Glyoxyl-agarose Beads
Immobilization of rAaeUPO on this support was performed diluting rAaeUPO solution in 50 mM of sodium carbonate at pH 10.05 and 25 • C [40] with 20% or 40% glycerol to prevent the enzyme inactivation. The enzymatic activity of the suspension and supernatant was followed during the whole process [113].
Immobilization of rAaeUPO Via Anionic Exchange.
rAaeUPO solution was diluted in 10 mM sodium acetate at pH 5.0 (activity was 0.5 U/mL), 10 mM Tris at pH 7.0 or 10 mM sodium carbonate at pH 9.0 at 25 • C. Then, 1 g of MANAE-or PEI-activated agarose beads per 10 mL of enzyme solutions was added. It was kept in orbital stirring for 2 h. The enzymatic activity of the suspension and supernatant was followed during the whole process [113]. After the immobilization, the derivative was washed several times with distilled water and stored at 4-6 • C.
Immobilization of rAaeUPO on Glutaraldehyde-amino-agarose Support Immobilization on glutaraldehyde-amino-agarose was performed using 1 g of support per 10 mL of enzyme solution (0.5 U/mL) prepared in 10 mM sodium acetate at pH 5.0, 10 mM sodium phosphate at pH 7.0 or 10 mM of sodium carbonate at pH 9.0 at 25 • C. The enzymatic activity of the suspension and supernatant was followed during the whole process using the ABTS assay described above [113]. Finally, the biocatalyst was washed with distilled water, vacuum dried and stored at 4-6 • C.

Immobilization on Vinyl Sulfone-agarose (VS-agarose) Support
The 4% BCL agarose beads were activated with DVS according to the methodology previously described by Santos et al. [114] with some modifications. Ten grams of agarose was added to 200 mL of a 333 mM sodium carbonate solution at pH 11.5 containing 350 mM of DVS. The resulting solution was kept under constant stirring for 2 h at 25 • C and finally was washed with excess distilled water and stored at 4-6 • C. rAaeUPO was immobilized using 50 mM sodium acetate at pH 5.0, 50 mM sodium phosphate at pH 7.0 or 50 mM sodium carbonate at pH 10.0 and 25 • C. During the immobilization, the activity of the supernatant, suspension and a reference under identical conditions was periodically determined [113].

Thermal Inactivation of the Different Biocatalysts
The free or immobilized enzyme samples were diluted/suspended in different buffers, pH and temperatures. Samples were periodically withdrawn, and the residual activity was quantified using the ABTS assay described above. The temperatures were selected to ensure reliable inactivation courses, and the error in the established temperatures was ±1 • C for different experiments using different water baths.
3.2.7. Desorption Assay of rAaeUPO from MANAE Support rAaeUPO immobilized on MANAE agarose support as described above was suspended in 10 mM sodium phosphate at pH 7.0 or 10 mM sodium acetate pH 4.4, Then, solid ammonium sulphate was used to reach different concentrations. After 1 h of incubation, the activity of the suspension and the supernatant was measured to determine the percentage of released enzyme. A reference with free enzyme was subjected to the same treatment to determine any effect of the ammonium sulfate on enzyme activity (we found none).

Operational Stability of MANAE-rAaeUPO Biocatalyst
Lastly, 1 g of biocatalyst was suspended in 100 mL of 5 mM sodium acetate at pH 5.0 containing 0.125 mM of H 2 O 2 and 0.03 mM of ABTS. To determine the absorbance, 2 mL of the reaction suspension was taken, and the biocatalyst was discarded by centrifugation at Catalysts 2021, 11, 560 16 of 21 6000 rpm for 2 min using an Eppendorf centrifuge. When the absorbance did not increase after 5 min, the biocatalysts were recovered by filtration and utilized in a new reaction cycle.

Conclusions
The immobilization of rAaeUPO is a quite complex task. This dimeric and highly glycosylated enzyme is destabilized by phosphate anions and stabilized by glycerol, and may be considered a quite stable enzyme (e.g., activity is fully maintained at pH 7.0 and 47 • C for 2 h). As it is a highly glycosylated enzyme, multipoint-covalent immobilization is a challenge. Production of the enzyme in prokaryotic expression systems resulting in non-glycosylated variants may offer the opportunity to further explore multi-point immobilization. This can over-compensate activity and stability losses stemming from the missing glycosylation pattern. The covalent immobilization strategies that we assayed failed (no immobilization or enzyme inactivation was observed) or produced very unstable biocatalysts. The ionic exchange of rAaeUPO in a polymeric ionic bed also produced very unstable immobilized enzyme biocatalysts. However, using MANAE agarose, immobilization yield was very high, and the enzyme maintained a high expressed activity. The immobilization pH slightly alters these parameters, but they are particularly relevant to the enzyme stability: The enzyme immobilized at pH 9.0 was less stable than the free enzyme. Thus, this immobilization condition, which was previously studied for its effect on enzyme activity or immobilization yield, should be considered a key condition in the determination of the immobilized enzyme features. However, even the best biocatalysts prepared in this research could only be used to maintain the activity and stability of the free enzyme, having as their main advantage the possibility of enzyme reuse. Other immobilization strategies, such as encapsulation (using sol-gel or MOFs) [115][116][117][118], crosslinked enzyme aggregates [119][120][121] or nanoflowers [122][123][124] could be explored with the objective of improving the current results.