Poly (Ethylene-Alt-Maleic Anhydride) Ionic Modification of Lipase B from Candida antarctica Immobilized on Octyl Agarose Beads Alters Its Catalytic Properties

TL;DR

This study shows how modifying immobilized Candida antarctica lipase B with a polymer can significantly change its activity and stability, depending on various factors.

Contribution

The study introduces a novel approach to altering lipase B properties through ionic modification with poly(ethylene-alt-maleic anhydride).

Findings

Polymer modification can increase enzyme activity by more than double or reduce it by 5–6 times depending on conditions.

Stability improvements were observed, such as maintaining 80% activity compared to 5% in unmodified biocatalysts under certain conditions.

The modification prevents enzyme release during inactivation, indicating a structural stabilization effect.

Abstract

The lipase B from Candida antarctica was immobilized on octyl-agarose using low and high (one that saturated the support surface with enzyme) loadings. Then, both biocatalysts were aminated, and the aminated and non-aminated biocatalysts were used in further experiments. The enzyme activity was determined using substrates with different structures. The modification of the four biocatalysts with poly (ethylene-alt-maleic anhydride) revealed that only a marginal covalent reaction occurs. That way, the ion exchange of the polymer on the immobilized enzyme surface should be responsible for the enzyme functional changes. The modification of the biocatalysts with this polymer produced mixed results for enzyme activity (depending on the enzyme loading, use of aminated or non-aminated enzyme, polymer concentration and used substrate), in some instances more than doubling the activity, in others reducing it by 5–6 times the activity when compared to the unmodified biocatalyst. The effects on biocatalyst stability were also mixed, depending on the same factors; in some instances, great stabilization could be found (e.g., in inactivation of the highly loaded aminated biocatalyst at pH 7.0, the unmodified biocatalyst kept 5% of the initial activity, while the biocatalyst modified with 1% of the polymer maintained 80%), but in other instances, enzyme stability was reduced after modification. It was shown that one of the effects of the polymer modification was the prevention of the enzyme release during inactivation.

Figures

Full text

1. Introduction

Lipases are among the most utilized enzymes in both academia and industry, due to their wide specificity (enabling the use of a wide range of substrates), high stability (enabling their use on a diversity of reaction media), absence of cofactors, etc. [1,2]. While their natural function is the hydrolysis of glycerides to fatty acids and glycerin [3], in vitro, they can catalyze many other reactions, such as esterifications [4,5], transesterifications [6,7], acidolysis [8,9], interesterifications [10], amidations [11,12] and even promiscuous reactions [13,14]. That way, lipases have found a wide diversity of applications in very different areas, such as food modification [5,15,16], biofuel [17] and biolubricant [18,19] production, fine chemistry [20], polymer production or degradation [21,22] (including plastic materials [23,24]) and bioremediation [25]. This diversity of applications of lipases is coupled to a characteristic catalytic mechanism, the interfacial activation, where a polypeptide chain secludes the enzyme activity center from the media, but it moves and facilitates adsorption of the lipases on hydrophobic drops of substrate, exposing the active centers and making them “interfacial enzymes” [26].

Enzyme immobilization started as a strategy to recover and reuse the enzymes, enabling the utilization of a wide diversity of reactors [27,28]. Soon, researchers noted that enzyme immobilization, if using an adequate protocol, enables the improvement of some enzyme features, such as enzyme stability [29,30], purity [31], activity, selectivity, specificity [32,33] and inhibition [34], and all these factors can widen the biocatalyst operation window [34,35]. Against these advantages, the main drawback of enzyme immobilization is that it increases the biocatalyst cost, the impact of which may be reduced if it enables enzyme reuse by multiple reaction cycles [35]. These immobilization costs involve the price of the support, the immobilization process itself, the work force, the transportation and storage of the support and the discarding of the inactivated biocatalysts, together with likely losses of enzyme activity caused by the immobilization [35].

In this context, there is a lipase immobilization protocol that can produce most of the benefits of enzyme immobilization while reducing its costs. This is the interfacial activation of lipases on hydrophobic supports at low ionic strength [36]. Hydrophobic supports are very stable, their storage is simple, and the immobilization process is very fast and permits the one-step immobilization-purification-stabilization-hyperactivation of most lipases [36]. Moreover, after enzyme inactivation during operation, the enzyme may be released from the support, enabling the support to be reused in new lipase immobilization cycles [36,37]. This means that biocatalyst costs are reduced because the support can be reused, while the immobilization usually produces an increase in enzyme activity (becoming a positive factor in final price calculations). Furthermore, the process is simple and fast, also saving costs in the process itself [36]. Moreover, most of the expected advantages of enzyme immobilization may be achieved [35]. However, the reversibility of the enzyme immobilization following this protocol poses a drawback: during operation, the enzyme may be released from the support to the medium under certain conditions (high temperatures, presence of hydrophobic organic co-solvents, reactants with detergent features, etc.) [36,37,38]. Some solutions have been proposed to solve the undesired enzyme release during operation when using this immobilization strategy, including the use of heterofunctional acyl supports. These supports can simultaneously contain octyl and anionic [39] or cationic moieties [40], enabling mixed physical adsorptions that make enzyme release more difficult, while maintaining the reversibility of the process. Another alternative is the use of chemically reactive groups, using glyoxyl [41], glutaraldehyde [42] or vinyl sulfone [43] combined with the acyl chains to covalently attach the enzymes to the supports after benefitting from the immobilization on hydrophobic supports. Another alternative to prevent enzyme release is the intermolecular crosslinking of the immobilized enzyme molecules, via chemical [44,45] or physical [46] crosslinking. This intermolecular crosslinking is favored when using high support loadings (saturating the support surface with enzyme molecules) where the enzyme molecules are near each other [47]. The modification of the immobilized lipase with polymers can alter other enzyme features, such as activity or specificity, and in many instances, it alters enzyme stability. The covalent massive intermolecular crosslinking using chemically reactive polymers permits one to nullify the possibility of enzyme release, but this means that the support cannot be reused after enzyme inactivation [44,45]. The physical intermolecular crosslinking using ionic polymers has, as its main advantage, the reversibility of the enzyme immobilization, enabling enzyme release from the support after its inactivation (accompanied by the ionic polymer) [46]. Among the cationic polymers, polyethyleneimine has been used in many instances for this purpose, in certain cases with positive results in terms of improvements in biocatalyst activity and stability [46]. Among the anionic polymers, sulfate-dextran has been utilized in some instances, but despite reducing the release of the immobilized lipase molecules [46], it often has a negative effect on immobilized enzyme activity and/or stability [48], making the search for alternative anionic polymers mandatory. The use of ionic polymers to obtain this massive intermolecular physical crosslinking may have additional advantages, as they can generate a hydrophilic environment around the enzyme molecules, producing the partition of some compounds with potential negative effects for the enzyme (e.g., oxygen or organic hydrophobic co-solvents) [49,50]. It should be stressed that the adsorption of enzyme molecules on ionic polymers or of ionic polymers on enzyme molecules is a multipoint ion exchange process that does not require that the polymer and enzyme present opposite global charges, but offers the possibility of establishing multiple enzyme–polymer positive interactions [51]. In fact, most proteins from E. coli can be immobilized at pH 7.0 on supports activated with either polyethyleneimine or sulfate dextran [52,53].

In this paper, we propose the use of polymers derived from maleic anhydride as alternative anionic polymers to modify immobilized enzymes. They may present some interest because of their comonomeric units [54]. It is simple to produce polymers with different comonomers, such as poly (ethylene-alt-maleic anhydride) [55] or poly (styrene-alt-maleic anhydride) [56], generating polymers with different properties. Sigma has commercialized poly (ethylene-alt-maleic anhydride), an alternative polymer formed by ethylene and maleic anhydride units. The anhydride moiety could permit the intermolecular covalent crosslinking of immobilized lipase molecules. If the anhydride is not stable enough to covalently react with the enzyme and it is hydrolyzed to yield maleic acid, it will become an anionic polymer, thus altering the enzyme environment and enabling physical intermolecular crosslinking.

As a model biocatalyst to test the potential of this polymer to improve immobilized lipase features, we selected the lipase B from Candida antarctica (CALB) immobilized on octyl-agarose. This enzyme is among the most utilized lipases, as it is very stable [57]. It has a small lid that is unable to isolate the active center from the reaction medium in a complete way [58], but it can still become adsorbed on hydrophobic surfaces. One feature of this immobilized biocatalyst is that in kinetic studies, the enzyme remains as 1st order at the limit of the substrate solubility in aqueous medium, making it impossible to calculate Km or Kcat and suggesting that Km is greatly increased while Kcat is also significantly increased [59]. The lipase has been immobilized at two different support loadings. One of these loadings will be well under the maximum loading capacity of the support, to analyze the effect of the polymer modification on isolated enzyme molecules. The other offered enzyme loadings will exceed the capacity of the support, to analyze the effect of the proximity of enzyme molecules to each other in the intermolecular crosslinking (the enzyme loading has been described to also affect immobilized CALB activity/stability) [60]. The enzyme was also aminated using ethylenediamine and carbodiimide to facilitate the covalent reaction/ionic interaction between the enzyme and the polymer [61]. SDS-PAGE of the immobilized and modified enzyme was used to identify the potential of the polymer as a covalent enzyme modifier reagent, and different substrates were utilized to quantify the effects of the modification on enzyme activity/specificity. The selected substrates presented different structures (Figure S1), while maintaining a high solubility in aqueous medium. Triacetin was the substrate most similar to the natural substrate of lipases, but with a very short-chain carboxylic acid; it is a multifunctional prochiral compound routinely used to determine lipase activity, but it may be an interesting material for more complex products [62]. Mandelic acid and esters have some interest in the pharmaceutical industry [63]; they are chiral compounds (enabling the analysis of changes in the enzyme enantiospecificity) and they present an aromatic carboxylic acid. p-Nitro phenyl butyrate (pNPB) is a standard substrate to determine lipase activity, and in this instance, the aromatic ring is in the nucleophile side.

2. Results and Discussion

2.1. Immobilization of CALB at Different Loads on Octyl-Agarose Beads

Figure 1 shows the immobilization courses of CALB on octyl-agarose beads at low and high loadings. In the first case, all enzyme activity was already immobilized in the first measure (15 min), and some increase in the enzyme activity was promoted by the immobilization (exhibiting 125% of the activity of the free enzyme). This can be explained by the stabilization of the open form of the lipase, although the lid of CALB is unable to fully seclude the active center to the medium [58], the enzyme conformational changes promoted by this adsorption may produce this increase in enzyme activity. During the preparation of the highly loaded biocatalyst, also most of the offered enzyme (80%) was immobilized before the first measure, but around 20% of the enzyme remained in the supernatant after 1 h, confirming that the support surface was fully coated by enzyme molecules. In this instance, the increase in enzyme activity was smaller, as the enzyme immobilization suspension increased the activity by only around 5%; this means that as only 80% of the enzyme was immobilized, the activity of the immobilized enzyme increased by more than 6%. This lower increase in enzyme activity can be caused by the existence of substrate diffusional limitations that prevent the substrate from reaching the enzyme molecules located on the core of the particles, as it is hydrolyzed by the enzyme molecules located nearer the external area of the pores of the biocatalyst particles’ medium [64]. That way, the lightly loaded biocatalyst presented an enzyme loading of 1 mg/g of biocatalyst, while the highly loaded biocatalysts exhibited an enzyme loading of 12 mg/g of biocatalyst. To confirm that this was the maximum support loading for CALB, maintaining the enzyme concentration (using larger volume), we tested 20 mg/g, and only 60% of the enzyme was immobilized, confirming that the maximum loading of the support was 12 mg/g.

The activities of both biocatalysts were determined versus the four substrates utilized in this research (Table S1). One first point to be noted is that while the specific activity versus pNPB of the enzyme in the lightly loaded biocatalyst was higher than that of the highly loaded biocatalyst (as should be expected by the increase in the substrate diffusional problems [64] and from Figure 1) when using triacetin, the specific activity of the highly loaded biocatalyst was higher than that of the lightly loaded biocatalyst. This cannot be explained by substrate diffusional factors (that will have negative effects on biocatalyst activity versus all substrates), and may be a consequence of conformational changes induced by enzyme molecule–enzyme molecule interactions, as reported in other instances [65]. The enzyme in the highly loaded biocatalyst presented a value of specific activity that was lower than 30% using pNPB, as can be deduced because it was only 3.3 folds more active than the lightly loaded biocatalyst per gram of biocatalyst, although the loading difference was a factor of 12. However, the specific enzyme activity versus triacetin in the highly loaded biocatalyst was over 10% higher than that of the enzyme in the lightly loaded biocatalyst: the highly loaded biocatalyst activity per gram of biocatalyst versus this substrate was about 13 fold higher than activity of the lightly loaded biocatalyst. Here, the use of 50 mM of triacetin can minimize the impact of substrate diffusional problems on the activity of the biocatalyst. Using both isomers of methyl mandelate, the specific activity of the enzyme in the highly loaded biocatalyst was in the range 45–50% of that of the lightly loaded biocatalyst for both substrates. That way, the highly loaded biocatalyst presented a mass activity almost 5.4 fold higher than that of the lightly loaded biocatalyst using the preferred isomer (the R-isomer), or 5.9 fold for the S-isomer. This implies a change in enzyme specificity for the different substrates depending on the enzyme loading. For example, the ratio between the specific activities with triacetin and with R-methyl mandelate was over 2.7 using the lightly loaded biocatalyst, but both specific activities became almost identical using the highly loaded biocatalyst (the value of the ratio was only 1.1). The enantiospecificity is not so different using R or S isomers of methyl mandelate (almost 11 using the lightly loaded biocatalyst and less than 10 using the highly loaded biocatalyst).

2.2. Amination of Both Octyl-CALB Biocatalysts

Table S1 shows the effect of amination on the activity of lightly and highly loaded biocatalysts versus different substrates (the activity is expressed as specific activity, that is, U/mg of enzyme). Results depend on the used substrate and biocatalyst. The enzyme in the lightly loaded biocatalyst decreased its activity after amination when using pNPB (to less than 60%), while the specific activity increased versus triacetin (by more than 40%), and slightly decreased using methyl mandelate (by less than 5% using the R-isomer and by almost 13% using the S-isomer). This means that the R/S methyl mandelate preference of the biocatalyst slightly increased from 10.9 to 12.0.

Using the highly loaded biocatalyst, the specific activity of CALB decreased using triacetin (to around 60%) and R-methyl mandelate (by around 20%) after amination, while it significantly increased over 4.1 fold using pNPB and around 2 fold using S-methyl mandelate. That means that the specific activity versus this substrate of CALB on the highly loaded immobilized biocatalyst further decreased compared to that of the lightly loaded biocatalyst, from 2.2 fold for the unmodified biocatalyst to 2.7 after amination.

These strong differences in the effects of the chemical amination on enzyme activity for biocatalysts with different enzyme loadings have been previously reported [47], and may be explained by the effects of the enzyme–enzyme interactions, existing in the highly loaded biocatalyst and not so intense for the lightly loaded biocatalysts, that will be quite different when all carboxylic acids have been transformed in cationic groups.

2.3. Modification of the Four CALB Biocatalysts with P: Study of the Covalent Modification

Highly and lightly Octyl-CALB and Octyl-CALB-A were incubated with P as described in Materials and Methods. Before studying the changes in the functional properties of the immobilized enzymes, it was assessed by SDS-PAGE if the polymer was able to produce covalent modifications of CALB molecules or intermolecular crosslinkings. Using alkaline pH values, the reactivity of the amino groups in the enzyme will be increased, but the stability of the anhydride groups will decrease. Due to the large size of the polymer, it may be expected that the modified enzymes can behave as much larger composites, while the crosslinked enzyme molecules should produce even larger adducts. Figure 2 shows the results using pH 9.0, where some bands in the upper half of the gel can be visualized, which can correspond to the formation of enzyme–polymer composites, but in any case, they account for a very minor amount of the original CALB since the band of this protein remains almost unaltered after modification. These minor bands were even more minor using pH 5.0 or 7.0 (Figure S2). Using the non-aminated enzyme, there were almost no additional bands compared to the initial biocatalyst; using the aminated and highly loaded biocatalyst, just very tiny bands can be observed at the top of the gel, while the aminated-CALB band is almost unaltered by the treatment.

Therefore, considering the SDS-PAGE results, it can be assumed that due to the low stability of the anhydride groups in aqueous buffer (visible by the pH decrease during the incubation by even 1 pH unit using the higher concentration of P) and the low reactivity of amino groups at low pH value, the covalent modification of the enzyme with P can be discarded as the main way to modify the enzyme under the utilized conditions. However, as in the case of dextran-sulfate, after opening of the anhydride groups, the resulting polymer is a poly-anion that could become ionically adsorbed on the enzyme surface, with adsorption reinforced when using aminated enzymes. This ionic enzyme modification will remain undetected by SDS-PAGE, as the formed composite should be broken after boiling in SDS. Hence, it was then analyzed if the enzyme incubation with P can be used to alter the enzyme functional features even without mainly producing covalent bonds.

2.4. Modification of the Four CALB Biocatalysts with P: Effect on Enzyme Activity

Table 1, Table 2, Table 3 and Table 4 show the effects of the incubation on the enzyme activity versus the four substrates used in this paper, which are structurally very different. The great effect of the incubation of the biocatalysts in P solutions confirms that some modification of the enzyme with P has been achieved.

Using the lightly loaded octyl-CALB (Table 1), the modification with P promoted a great decrease in the enzyme activity versus pNPB. After modification at pH 5.0, Octyl-CALB-0.5P was the least active biocatalyst (decreasing to only 8.7% of the initial activity); when the modification was performed at pH 7.0 or 9.0, the lowest activity was achieved using Octyl-CALB-0.05P (11% or 8.7%, respectively). The situation was very different using triacetin, where a general increase in enzyme activity could be detected after modification with P. The highest activity when the biocatalyst was modified at pH 5.0 was found using Octyl-CALB-0.05P (by 40%), although similar results could be found using the other P concentrations. When the modification was performed at pH 7.0, the activity of Octyl-CALB-0.05P was almost 12% higher than that of the unmodified biocatalyst, and then decreased when increasing the P concentration (becoming fairly similar to that of the unmodified biocatalyst using Octyl-CALB-1P). If the enzyme was modified at pH 9.0, again the activity progressively increased when increasing the concentration of P, with Octyl-CALB-0.05P presenting almost the same activity as the unmodified biocatalysts, while Octyl-CALB-1P was almost 31% more active.

Using R-methyl mandelate, the modification at pH 5.0 with the lowest concentration of P produced a decrease in enzyme activity (to around 65%), but then, the activity progressively increased when increasing the concentration of P, and when using 1% P, the activity became over 2.6 fold higher than that of the unmodified biocatalyst. If the modification was performed at pH 7.0, the activity slightly increased using 0.05% P, and then progressively decreased when increasing the concentration of P (to less than 50% for Octyl-CALB-1P). The modification at pH 9.0 of the immobilized enzyme with P produced an increase in the enzyme activity that increased when using higher [P], more than doubling the activity of Octyl-CALB-1P compared to the activity of Octyl-CALB.

Using S-methyl mandelate, the modification with P at pH 5.0 produced a progressive increase in enzyme activity (Octyl-CALB-1P exhibited almost double the activity of the unmodified biocatalyst). After modification at pH 7.0, the changes in activity were very small (it is remarkable that when using 0.1% P, the activity decreased to around 80%), and the modification at pH 9.0 seemed to have an almost null effect on the activity with this substrate.

Hence, the incubation of the lightly loaded biocatalyst with different P solutions had different impacts on the enzyme features, on the effect of the P concentration depending on the initial pH value used in the modification, and on the substrate. Again, the modification of the enzymes with P produced a drastic change in enzyme specificity, as the effects of the modification on the specific activity of immobilized CALB strongly depended on the substrate used. It should be noted that the final pH after enzyme modification was decreased (by over 1 unit in the extreme case), and the final value depended on the concentration of P.

From the scarce P-enzyme covalent reaction, and considering that the enzyme modification should be just via an ionic exchange between the surface of the enzyme and the polymer, it may be proposed that P could be previously incubated at alkaline pH to open the anhydrides, and then adjust the pH value. However, in this research, we decided to continue using the commercial material without any previous treatment, as this was a straightforward way to reproduce the studies.

Next, we will discuss the results obtained using the highly loaded Octyl-CALB (Table 2). Starting with the activity versus pNPB, in contrast with the drastic decrease in enzyme activity found using the lightly loaded biocatalyst, the enzyme activity was fairly maintained and even slightly increased, ranging from 66% of the activity of the unmodified biocatalyst in the case of the Octyl-CALB-0.1P modified at pH 5.0 to 106% for the Octyl-CALB-0.5P modified at pH 5.0 or Octyl-CALB-0.1P modified at pH 7.0.

Using triacetin, in contrast to the lightly loaded biocatalyst, the modification with P produced a general decrease in enzyme activity, but this depended on the pH of the modification and the concentration of P. The only exception was Octyl-CALB-0.5P prepared at pH 9.0, where the enzyme activity increased, albeit very slightly (by 4%). When modified at pH 5.0, the final activity increased with increasing P concentration, from 62% (for Octyl-CALB-0.05P) to 76% (for Octyl-CALB-1P), while at pH 7, the minimum activity was observed for the biocatalyst Octyl-CALB-0.5P (61%), and when the modification was performed at pH 9.0 using the biocatalyst Octyl-CALB-1P, the activity decreased more compared to the unmodified biocatalyst.

Using R-methyl mandelate, the modification at pH 5.0 produced biocatalysts that increased their activity when increasing the P concentration (going from 85% using Octyl-CALB-0.1-P to almost 110% using Octyl-CALB-1P). The modification at pH 7.0 had almost no effect on the enzyme activity, whereas if the modification was performed at pH 9.0, Octyl-CALB-0.1P reduced its activity to 75%, while Octyl-CALB-0.1P increased the activity to almost 110% for Octyl-CALB-0.5P.

The situation was again different using S-methyl mandelate. When modifying the enzyme at pH 5.0, the decrease in activity of Octyl-CALB-1P (81%) was the most relevant change. If the modification was performed at pH 7.0, the enzyme activity increased in all cases, and this improvement was progressive when increasing the P concentration, reaching 143% with Octyl-CALB-1P. At pH 9.0, a progressive improvement in enzyme activity could be observed when increasing P concentration, but in any case, improved the activity of the Octyl-CALB (with Octyl-CALB-1P and Octyl-CALB-0.05P presenting 94% and 60% of the unmodified biocatalyst activity, respectively).

Again, the modification effects on specific enzyme activity depended on the modification pH, on the P concentration, and on the used substrate. Very interestingly, the effects were very different from those observed for the lightly loaded biocatalysts, suggesting that the protein–protein interactions may be tuned after the modification with P, and this enzyme interaction tuning was more relevant for the highly loaded biocatalysts.

Next, we analyzed the effects of this modification using the aminated biocatalysts. Starting with the lightly loaded Octyl-CALB-A (Table 3), the activity versus pNPB decreased in all cases, as in the case with Octyl-CALB, ranging for 18% using Octyl-CALB-A-1P prepared at pH 9.0, to 44% for Octyl-CALB-A-0.05P prepared at pH 5.0.

Using triacetin, the activity decreased when incubating Octyl-CALB-A in P solutions at pH 5.0 (even to 46% using Octyl-CALB-A-1P) except using Octyl-CALB-A-0.5P showed a 150% increase in activity compared to that of Octyl-CALB-A. Octyl-CALB-A-0.05P and Octyl-CALB-A-0.5 prepared at pH 7.0 showed scarce changes in enzyme activity compared to Octyl-CALB-A. Octyl-CALB-A-1P decreased the activity to 73%, and Octyl-CALB-A-0.1P increased the activity to 125%.

When the activity was evaluated using R-methyl mandelate, the modification with P at pH 5 increased the activity in all cases, with a maximum using Octyl-CALB-A-0.05P (122%) and a minimum using Octyl-CALB-A-1P (115%). When incubating the biocatalyst with P at pH 7.0, the activity decreased in all cases in a quite similar way, with a minimum using Octyl-CALB-A-1P (around 60%), while modifying the biocatalyst at pH 9.0 resulted in around 75% of the activity of the unmodified biocatalyst, except for Octyl-CALB-A-0.5P (that retained 85% of its activity after the modification).

Using S-methyl mandelate, Octyl-CALB-A activity increased after modification with P at pH 5.0 (with a maximum of 1.9 fold using Octyl-CALB-A-0.5P), but the positive effect decreased when the P concentration increased and became null using Octyl-CALB-A-1P. When the modification was performed at pH 7.0, the increase in activity was very significant for Octyl-CALB-A-0.05P (by 3.5 fold) but progressively decreased with increasing the P concentration (to 88%). At pH 9.0, the situation was similar to that at pH 7, but with lower activities, with Octyl-CALB-A-0.05P and -0.1P presenting 113% of the activity of Octyl-CALB-A, while Octyl-CALB-A-0.5P and Octyl-CALB-A-1P presented activity in the range of 70–75%.

Again, the changes were different depending on the P concentration, pH of modification and substrate. There were no optimal conditions for all substrates, and results differed from those found using lightly loaded Octyl-CALB, showing the relevance of both the amination and the enzyme–enzyme interactions in the final effects of the modification with P.

Next, we analyzed the effects of the incubation in P of the highly loaded Octyl-CALB-A (Table 4). Starting with the activity versus pNPB, the activity decreased in all cases, with the lowest activity found for Octyl-CALB-A-0.05P and 0.1P prepared at pH 7.0 (just over 45%) and the highest activity obtained using Octyl-CALB-A-0.1P prepared at pH 5 (85%).

Using triacetin, the modification at pH 5.0 had mixed results, as the activity decreased when increasing P concentration, going from 113% for Octyl-CALB-A-0.05P to 44% for Octyl-CALB-A-1P. In the modification at pH 7.0, the biocatalyst activity decreased in a progressive way, reaching a value of 67% for Octyl-CALB-A-1P. The modification at pH 9.0 produced an increase in enzyme activity with this substrate, decreasing with the P concentration (from 145% for Octyl-CALB-A-0.05P to 121% for Octyl-CALB-A-1P).

The use of R-methyl mandelate produced again a completely different picture. The modification with P at pH 5.0 produced an increase in enzyme activity that was reduced when the concentration of P increased (from 126% using Octyl-CALB-A-0.05P to 104% using Octyl-CALB-A-1P). When the modification was performed at pH 7.0, the effect was always negative, increasing in a progressive way with the P concentration (with final activities from 85% for Octyl-CALB-A-0.05P to 71% for Octyl-CALB-A-1P). At pH 9.0, the modification with P produced a progressive increase in enzyme activity, except when using the lowest P concentration, Octyl-CALB-A-0.05P yielded 91% of the activity of Octyl-CALB-A, while Octyl-CALB-A-1P yielded 120%. Again the picture changed when using S-methyl mandelate as substrate. In all cases, the activity of immobilized CALB after modification with P decreased, being more intense when the modification was performed at pH 9.0 (where the activity was under 50%).

The results presented to this point show the strong co-interaction between variables such as enzyme loading, previous enzyme amination, modification of pH and concentration of P, and substrate used in determining the enzyme activity. Considering the very different effects on the activity with R- or S-methyl mandelate and the different studied variables, and as with using these biocatalysts, any partition differential effect can be discarded, the only explanation is that the enzyme properties are altered after the modifications. Accordingly, these strong changes in specificity of CALB when submitted to the different treatments suggest that the enzyme features are significantly altered after the modifications, very likely by a combination of the affects previously described. These effects are influenced by enzyme–enzyme interactions, as previously reported [66].

The results suggest that P modification may be utilized to improve the activity of CALB immobilized via interfacial activation, but the modification conditions must be carefully optimized for each specific substrate. Usually, the optimal activity was found not when using the highest concentration of P, but when using some intermedium concentration.

The modification of the enzyme with P may promote several simultaneous effects on the immobilized enzyme molecule’s features, even if it is just a physical modification. First, the enzyme interactions with the polymer can alter the enzyme structure, producing biocatalysts with higher or lower enzyme activities versus different substrates. This can depend on the modification conditions. Moreover, the enzyme will be in a hydrophilic environment (a polymeric bed surrounding the enzyme molecules); this should be a general effect, more intense when the polymeric bed better coats the enzyme molecules. This can alter enzyme activity and promote some partition effects. It has been reported that hydrophilic compounds may accumulate in the enzyme environment, while hydrophobic ones will be partitioned away [67]. This can produce a great effect on enzyme activity, as the biocatalyst is in 1st order, as discussed in the Introduction [59]. The large size of the pores of agarose 4BCL [68,69] and the random coil structure of P can eliminate the effects of increasing the substrate diffusional factors when penetrating inside the particles. However, mainly in the highly loaded biocatalysts, they may not be eliminated if there is some increase in the substrate diffusional factors from the light of the pore to the enzyme active center, looking to the support surface, promoted by the P coating of the enzyme molecules (the space between enzyme molecules should be the only way for the substrate to reach the active center of the immobilized enzyme molecules). Furthermore, the intermolecular physical crosslinking of the enzyme molecules can reinforce the immobilization strength, preventing their release from the support, with an impact on the stability of enzyme activity [44,46]. In fact, all these positive and negative phenomena will be simultaneously occurring, becoming the observed effects on the enzyme features and their combined effects.

2.5. Modification of the Four CALB Biocatalysts with P: Effect on Enzyme Stability

Figure S3 shows the comparison of the inactivation courses of the four different biocatalysts utilized in this paper. When inactivating the enzyme at pH 7.0, as previously reported, the lightly loaded biocatalyst is more stable than the highly loaded biocatalyst [60,66]. This occurred even considering that the substrate diffusional limitations produce a decrease in the detected activity for the highly loaded biocatalyst., Hence, if the stabilities of the enzyme in both preparations were identical, the apparent stability of the highly loaded biocatalyst could be found to be higher. As the opposite happened, it can be assumed than the enzyme–enzyme interaction on the highly loaded biocatalyst could be responsible for this lower stability of the highly loaded biocatalyst. The amination of the enzyme produced a certain enzyme destabilization in both cases, more noticeably for the highly loaded biocatalyst. At pH 5.0 and 9.0, the lightly loaded biocatalyst was still more stable than the highly loaded biocatalyst. Amination of the enzyme, however, presented a quite different effect on enzyme stability depending on the inactivation pH and loading of the support. While the lightly loaded biocatalyst stability increased at pH 5.0, a significant decrease in the stability of the highly loaded biocatalyst could be observed. At pH 9.0, amination was negative for the enzyme stability of both biocatalysts, but the decrease was more intense for the lightly loaded biocatalyst. Hence, the effect of the surface amination of CALB was negative in most cases for the enzyme stability, with the exception of the inactivation at pH 5.0 using the lightly loaded biocatalyst.

Next, we investigated the effects of modification with different concentrations of P on biocatalyst stability at different pH values. Starting with the lightly loaded Octyl-CALB (Figure 3), the modification with P seemed to have a negative effect on enzyme stability at pH 5.0 (except in the first moments of the inactivation of Octyl-CALB-0.1P) and at pH 9.0. In both instances, the biocatalyst Octyl-CALB-0.5P presented the lowest stability. However, the modification had a positive effect on the stability of the lightly loaded Octyl-CALB at pH 7.0, except for the biocatalyst Octyl-CALB-1P. The inactivation courses were similar for the other biocatalysts.

Figure 4 shows the inactivation courses of the highly activated biocatalyst after modification with P. The inactivation at pH 5.0, as in the case of the lightly loaded biocatalyst, shows a biocatalyst destabilization, again becoming an exception with the initial inactivation time of Octyl-CALB-0.1P. At pH 7.0 and 9.0, the effects on enzyme stability of the modification with P were not very relevant, generally promoting a marginal increase in enzyme stability. The situation was quite different from the results obtained using the lightly loaded biocatalyst. Using the highly loaded biocatalyst, we could expect to have the individual molecules coated by P, in a way similar to that of the results found using the lightly loaded biocatalyst, but it would be easier to have intermolecular physical crosslinking. This should mean an increase in enzyme stability, but the results did not point to this situation at pH 7.0, where the stabilization by the P modification was more significant using the lightly loaded biocatalyst, and the effect of the alteration of the enzyme molecule–enzyme molecule interactions may be more relevant. At pH 9.0, the destabilization caused by the P modification using the lightly loaded biocatalyst was avoided, and a slight positive effect was observed, perhaps as a consequence of a higher positive effect of the intermolecular crosslinking. The coating with P could also somehow reduce the negative effects of the enzyme–enzyme interactions on the enzyme stability.

The modification of lightly loaded Octyl-CALB-A with P produced a completely different effect than the modification of the lightly loaded non-aminated biocatalyst (Figure 5). In inactivation at pH 5.0, the effect of the modification with P was positive, except using Octyl-CALB-A-0.1P, with a lower stability compared to the unmodified non-aminated biocatalyst. It is curious that this biocatalyst is the only one that increased its stability after amination at this pH value (Figure S2), and also, the coating with an anionic polymer had a positive effect. The stability at pH 7.0 showed a significant increase, mainly using Octyl-CALB-A-0.05P, while Octyl-CALB-A-0.05P showed an inactivation course very similar to that of Octyl-CALB-A. At pH 9.0, the biocatalyst stability decreased after modification, with the exception of the biocatalyst Octyl-CALB-A-0.05P, which became much more stable than Octyl-CALB-A. The best stabilization was found in inactivation at pH 7.0, where Octyl-CALB-A preserved just over 25% of its initial activity after 4 h of incubation, while Octyl-CALB-A-0.05P maintained over 80% of its initial activity.

Using the highly loaded Octyl-CALB-A modification, the modification with P had a clearly positive effect at pH 5.0 and 7.0, with the highest stabilization obtained using the two highest P concentrations (Figure 6). At pH 9.0, these two biocatalysts remained more stable than the untreated Octyl-CALB-A, while the other two biocatalysts were less stable. Octyl-CALB-A retained less than 5% of its initial activity, whereas Octyl-CALB-A-1P retained almost 80% when the inactivation was performed at pH 7.0. These somehow better (or at least less negative) effects on the stability of the aminated enzyme with the P modification may be related to a decrease in the negative effects of amination on enzyme stability after the modification with this anionic polymer.

To analyze whether the effects of the modification with P on enzyme stability were due, at least partially, to preventing enzyme release during inactivation, SDS-PAGE studies of the biocatalysts treated or not with P were performed. Figure 7 shows a representative example, where the intermolecular physical crosslinking is favored, and therefore, it is more difficult to release the enzyme from the support during thermal inactivation: highly loaded Octyl-CALB-A-XP. It is clearly evident how the biocatalyst unmodified with P presented a lower CALB intensity band after the thermal inactivation, while the biocatalysts treated with P retained a higher amount of CALB molecules, increasing when moving from Octyl-CALB-A-0.05P to Octyl-CALB-A-0.1P, and later slightly decreasing using higher P concentrations.

This can help to understand some of the previous results. First, as explained in the Introduction, it should be borne in mind that the enzyme is highly stabilized after immobilization on octyl-agarose; consequently, the release of the enzyme from the support led to a lower stability of the biocatalysts. If the enzyme remained attached to the support, a higher stability of the biocatalysts should be expected. Using a low concentration of P, the polymer amount is not enough to fully modify all the enzyme molecules. Using very high concentrations of P, the modification of the individual enzyme molecules is so rapid that the “intermolecular crosslinking” becomes disfavored. Better “physical crosslinking” could be achieved using intermediate P concentrations, where the competition between new enzyme molecules modified with P and intermolecular crosslinking is not favored, but the amount of P is enough to have a significant number of enzyme molecules modified (P later could interact with enzyme molecules in the vicinity of those already interacting with molecules of P). A similar effect was described in the activation of support with polymers for enzyme immobilization via ion exchange [52].

3. Materials and Methods

3.1. Materials

Liquid CALB (lipase B from Candida antarctica, 24.77 mg of protein/mL) was kindly provided by Novozymes A/S (Madrid, Spain), and it was used without any further purification (the enzyme became purified after immobilization on octyl agarose). Octyl agarose beads were supplied by Agarose Bead Technologies (Madrid, Spain). Poly-(ethylene-alt-maleic-anhydride) (P) (with a molecular weight in the range of 100,000–500,000 Da following supplier information), ethylcarbodiimide hydrochloride (ECD) and ethylenediamine (EDA) were obtained from Sigma-Aldrich (St. Louis, MO, USA). 4-Nitrophenylbutyrate (pNPB) and triacetin were obtained from GE Healthcare (Madrid, Spain). R-Methyl mandelate and S-methyl mandelate were purchased from Alfa Aesar (Fisher Scientific, Madrid, Spain). Other used reagents were of analytical grade. The concentration of protein was quantified employing the method designated by Bradford [70].

3.2. Methods

All experiments were performed at least three times; the paper results are given as the means of these values, including their respective standard deviations.

3.2.1. Enzyme Activity Determination Using pNPB

pNPB was used as substrate in the standard activity quantification protocol to determine the enzyme activity during immobilization, stability tests, etc. A cuvette containing a volume of 2.5 mL of 25 mM sodium phosphate at pH 7.0 was prepared, adding 20 μL of a pNPB solution (50 mM in acetonitrile). The hydrolysis reaction was initiated by adding 50 μL of either free enzyme solution or enzyme immobilized suspension, under continuous magnetic stirring. A Jasco V-730 spectrophotometer (Jasco, Madrid, Spain) was employed to follow the increase in absorbance at 348 nm (corresponding to the isosbestic point of p-nitrophenol, ε = 5150 M^−1^ cm^−1^ under these conditions [71]) reflecting the formation of p-nitrophenol promoted by the hydrolysis of pNPB. The increase in absorbance was recorded over 90 s. Enzyme activity was expressed in activity units (U), defined as micromoles of released p-nitrophenol per minute.

3.2.2. Enzyme Immobilization

CALB was immobilized on octyl-agarose beads using 1 g of support per 10 mL of enzyme solution diluted in 50 mM sodium phosphate at pH 7.0 and 25 °C [60], using enzyme solutions at two different concentrations. One biocatalyst was prepared with a concentration of 0.1 mg of enzyme per mL of solution (i.e., an enzyme loading of 1 mg/g) and another one with a concentration of 1.5 mg of enzyme per mL of solution that guaranteed the complete coating of the support surface by lipase (using an enzyme excess of 15 mg/g) [60]. The use of 50 mM sodium phosphate was to reduce the possibility of CALB molecules being adsorbed on the fully enzyme-coated biocatalyst [72].

The immobilization suspension was maintained at room temperature under gentle agitation. At defined times, the catalytic activities of the supernatant and suspension, together with that of a reference solution (the enzyme under the same conditions but in the absence of support), were measured using pNPB as substrate. After 1 h, the immobilized biocatalysts were washed with distilled water, filtered under vacuum, and finally stored at 6 °C until further use. This biocatalyst was named Octyl-CALB.

3.2.3. Amination of the Immobilized CALB Surface

A mass of 1 g of the biocatalyst was added to 10 mL of 2M EDA at pH 4.75 and 25 °C. Subsequently, solid ECD was added until reaching a concentration of 10 mM. This suspension was stirred for 2 h at room temperature. This transformed the exposed carboxylic groups of the enzyme in primary amino groups [61]. Then, the biocatalyst was recovered by filtration under vacuum and washed with an excess of distilled water and stored at 6 °C until further use. The aminated immobilized lipase was named Octyl-CALB-A.

3.2.4. Immobilized Lipase Modification with Poly (Ethylene-Alt-Maleic Anhydride) (P)

A 2 g mass of Octyl-CALB or Octyl-CALB-A (highly or lightly loaded) biocatalyst was suspended in a 20 mL solution of 50 mM sodium acetate at pH 5.0, sodium phosphate at pH 7.0, or 50 mM sodium carbonate at pH 9.0. The suspension was stirred, and then P was added to reach the desired concentration to be incubated (0.05, 0.10, 0.50 or 1.0% (wt/v). Suspensions were gently shaken for 2 h at 25 °C without controlling the pH. After that, the suspensions were filtered and rinsed with the buffers where they were prepared and finally rinsed with an excess of distilled water. The biocatalysts were named by appending the % of P used in the modification, suffixed with the letter “P” (for example, Octyl-CALB modified using 0.05% P was named Octyl-CALB-0.05P).

3.2.5. Sodium Dodecyl Sulfate-Polyacrylamide Gel Electrophoresis (SDS-PAGE) Analysis of the Different Lipase Biocatalysts

The Laemmli protocol [73] was used to carry out the SDS-PAGE experiments. The different biocatalysts (free or immobilized enzymes) were suspended in a solution of 4% SDS and 10% mercaptoethanol, thus obtaining samples with 0.5 mg of protein per milliliter of solution. These suspensions were boiled for 8 min. After this time, the samples were centrifuged at 10,000 rpm for 2 min, and the support was discarded. Aliquots of 15 µL of each supernatant or 7 µL of marker molecular weight solutions (14.4–97 kDa) (GE Healthcare) were taken. Samples were injected on 12% polyacrylamide gels, applying a 100 V potential difference.

3.2.6. Enzyme Thermal Inactivation Experiments

A mass of 1 g of all the lipase biocatalysts was incubated in 5 mL of 50 mM sodium acetate pH 5.0, 50 mM Tris-HCl pH 7.0, and 50 mM sodium carbonate pH 9.0 at 74 °C. Samples of 50 µL of each biocatalyst inactivation suspension were periodically taken, and their activities were determined using pNPB as substrate. Activity was determined as described in the enzymatic assay section. The inactivation courses represented the biocatalysts’ residual activities along time, calculated as the percentage of their initial activities. The temperature was chosen to achieve a fast but reliable inactivation of the enzyme biocatalysts.

3.2.7. Activity of the Prepared Biocatalysts with Different Substrates

Hydrolysis of Triacetin

A solution of 2 mL containing 50 mM triacetin in 50 mM sodium phosphate at pH 7.0 and 25 °C was prepared. Subsequently, 0.02 (highly loaded) or 0.2 (lightly loaded) g of wet biocatalyst was added under continuous stirring. At pH 7.0, rapid acyl migration occurred in the enzymatic product, yielding a mixture of 1,2-diacetin and 1,3-diacetin [74]. Enzyme activity was evaluated by determining the percentage of hydrolysis at various time points of this reaction. For this purpose, an HPLC system equipped with a Kromasil C18 column, Göteborg, Sweden (15 cm × 0.46 cm) and a UV detector set at 230 nm was used. The mobile phase consisted of 15% (v/v) acetonitrile and 85% (v/v) Milli-Q water, with a constant flow rate of 1 mL/min. Biocatalyst activity was evaluated based on the formation of diacetins (1,2- and 1,3-diacetin), using conditions under which both isomers co-eluted. Only data corresponding to hydrolysis levels between 15% and 20% were considered for activity calculations, to reduce their transformation to monoacetins [37]. Retention times were 4 min for both diacetins and 18 min for triacetin. Activity is given as micromoles of produced diacetin per minute.

Hydrolysis of R- or S-Methyl Mandelate

Three-milliliter solutions containing 50 mM of R- or S-methyl mandelate in 50 mM sodium phosphate at pH 7.0 and 25 °C were prepared, and 0.02 (highly loaded) or 0.2 (lightly loaded) g of the wet biocatalysts was added to those solutions. The reaction mixture was stirred continuously. Both mandelic acid and its ester concentrations were analyzed using an HPLC system equipped with a UV/VIS detector set at 230 nm and a Kromasil C18 column (15 cm × 0.46 cm). The mobile phase consisted of 35% (v/v) acetonitrile and 65% (v/v) of an aqueous 10 mM ammonium acetate solution at pH 2.8, with a flow rate of 1 mL/min. Conversions in the range of 15–20% were used to calculate the initial reaction rates. Activity is given as micromoles of produced mandelic acid per minute.

4. Conclusions

The enzyme loading on octyl agarose produces biocatalyst with very different enzyme-specific activity and stability. The further amination produces again changes in the functional properties, changes that also depend on the loading of the support. The use of poly-(ethylene-alt-maleic-anhydride) to modify these CALBs immobilized in octyl agarose biocatalysts has been found to have great potential for tuning immobilized lipase properties, in some instances greatly improving enzyme activity or stability. This occurs even when the covalent reaction between the enzyme and the polymer is negligible and the modification of the enzyme features comes mostly from the ionic exchange of the polymer on the immobilized enzyme. Hence, opening the anhydride before performing the enzyme modification may be suggested to offer better pH control. In fact, in some instances, we have permitted the opening of the anhydride in the aqueous medium before adding the immobilized enzyme, and the results were almost identical to those obtained using the commercial polymer. The final effects depend on the concentration of P used for the enzyme modification, the loading of the biocatalyst, the previous amination of the immobilized enzyme and the conditions and substrate used for the evaluation of enzyme features. The use of aminated enzymes and the enzyme loading seem to be critical for the final properties of the biocatalyst. Therefore, the modification with P can be positive (even very positive) for the enzyme activity or stability in certain circumstances and negative (even very negative) in others. The concentration of P seems to determine the possibility of obtaining intermolecular physical crosslinking between different immobilized CALB molecules. However, the highest stability was not always found using the same P concentration; it depends on the inactivation pH, the previous amination, or not, of the enzyme, and the support loading with the enzyme. This can be explained by the competition between individual enzyme molecule modifications and enzyme molecule–enzyme molecule simultaneous interactions (that is, physical crosslinking).

This polymer may be a very good one for obtaining anionic polymeric supports, which can be used to immobilize enzymes via cation exchange, as an alternative to more expensive sulfate dextran. Activating supports, the polymer can be employed in anhydrous conditions where the stability of the anhydride group may be much higher. However, for enzyme immobilization, aqueous medium will be required, and that will promote the opening of the anhydride rings. Hence, it may be better to fully open the anhydride before use the support to obtain fine control of the immobilization conditions, as the prospects of using it as a support for covalent immobilization of enzymes will be as poor as its use for covalent modification of enzymes. In fact, preliminary experiments in our laboratory showed the immobilization of CALB on supports activated with this polymer (confirming that enzyme and polymer can ionically interact). Similarly, CALB could be immobilized and modified using dextran sulfate, as previously reported [53,75,76,77].

Overall, the results seem to show a complex interaction between many factors (co-interactions already visualized from the initial biocatalyst features), and the final selection of the optimal modification protocol should be defined using the target substrate and operational conditions; otherwise, there are many possibilities of using a non-optimized biocatalyst. The optimization should involve the initial enzyme biocatalyst (aminated or not aminated, the support loading with the enzyme), the concentration of P and the pH of the modification. In many instances, an increase in enzyme activity versus some substrates is accompanied by a decrease in enzyme stability under certain conditions, and vice versa. The potential of this polymer in biocatalysis must be explored for other applications (generation of enzyme hydrophilic nano-environments, use as crosslinking reagent in CLEA preparation, preparation of polymeric anionic supports, etc.), even if its use as a covalent reagent to interact with enzymes seems to be not very useful.