Catalytic Efﬁciency of Basidiomycete Laccases: Redox Potential versus Substrate-Binding Pocket Structure

: Laccases are copper-containing oxidases that catalyze a one-electron abstraction from various phenolic and non-phenolic compounds with concomitant reduction of molecular oxygen to water. It is well-known that laccases from various sources have different substrate speciﬁcities, but it is not completely clear what exactly provides these differences. The purpose of this work was to study the features of the substrate speciﬁcity of four laccases from basidiomycete fungi Trametes hirsuta , Coriolopsis caperata , Antrodiella faginea and Steccherinum murashkinskyi , which have different redox potentials of the T1 copper center and a different structure of substrate-binding pockets. Enzyme activity toward 20 monophenolic substances and 4 phenolic dyes was measured spectrophotometrically. The kinetic parameters of oxidation of four lignans and lignan-like substrates were determined by monitoring of the oxygen consumption. For the oxidation of the high redox potential (>700 mV) monophenolic substrates and almost all large substrates, such as phenolic dyes and lignans, the redox potential difference between the enzyme and the substrate ( ∆ E ) played the deﬁning role. For the low redox potential monophenolic substrates, ∆ E did not directly inﬂuence the laccase activity. Also, in the special cases, the structure of the large substrates, such as dyes and lignans, as well as some structural features of the laccases (ﬂexibility of the substrate-binding pocket loops and some amino acid residues in the key positions) affected the resulting catalytic efﬁciency.


Introduction
Laccases (benzene diol:dioxygen oxidoreductases; EC 1.10.3.2) are copper-containing oxidases that catalyze a one-electron abstraction from various phenolic and non-phenolic compounds with concomitant reduction of molecular oxygen to water. Laccases can be used in wood, pulp and paper, textile, and food industries [1] as well as soil bioremediation and sewage treatment [2,3].
One of the main distinguishing features of laccases is their ability to oxidize a very wide range of compounds. Typical substrates of laccases are polyphenols, substituted phenols, and aromatic diamines. It is well-known that laccases from various sources have different substrate specificities, but it is not completely clear what exactly provides these differences [4,5]. For a long time, it was believed that the catalytic efficiency of substrate oxidation by laccases correlates with a redox potential difference between the substrate and the enzyme T1 copper center [6]. However, at the present moment Four different laccases from basidiomycetes, T. hirsuta (ThL), C. caperata (CcL), S. murashkinskyi (SmL), and A. faginea (AfL), were screened for their ability to oxidize 20 simple monophenolic compounds and 4 phenolic dyes. All chosen compounds form a representative set concerning different structural patterns observed in typical laccase substrates ( Figure 1). All four studied laccases can be divided into two groups according to their T1 center redox potentials: ThL (E = 780 mV) and CcL (E = 780 mV) are high redox potential laccases and SmL (E = 650 mV) and AfL (E = 620 mV) are middle redox potential laccases.
The data regarding activities of laccases toward different substrates are presented in Figure 1. To highlight overall patterns in the oxidation of substrates by all studied laccases, the log-transformed activity values are presented (Figure 1 upper panel). To facilitate the comparison among different laccases, the standardization (Z-scores) of activities was performed (Figure 1 lower panel). Additional information regarding the laccase activity is shown in Table S1.
In general, the benzoic acid derivatives were oxidized at a lower rate than the cinnamic acid derivatives with the same substituents. It is interesting to note that for the benzoic acid derivatives the methoxy substituent showed less influence on the oxidation rates than the hydroxy substituent, whereas for the cinnamic acid derivatives the opposite trend was observed.
All the laccases oxidized the aromatic compounds with the amino group (2,5-xylidine and 2-amino-4-methoxybenzoic acid) at a much lower rate compared to most compounds with the hydroxy group. The same tendency as for the simple monophenolic compounds was observed for the phenolic dyes. All laccases oxidize congo red and indigo carmine containing an amino group and a nitrogen in the indole ring slower than bromocresol green and phenol red containing hydroxy groups in the benzene rings. group. The same tendency as for the simple monophenolic compounds was observed for the phenolic dyes. All laccases oxidize congo red and indigo carmine containing an amino group and a nitrogen in the indole ring slower than bromocresol green and phenol red containing hydroxy groups in the benzene rings. It is worth noting that although all laccases considerably more slowly oxidized compounds with redox potentials above 700 mV (e.g., coumaric acids, vanillic acid, vanillin, 2,5-xylidine [15][16][17]), the relationship between the rate of the oxidation and the redox potential of the substrates was not so apparent for compounds with lower redox potentials. For example, caffeic acid, which has a lower potential than sinapic acid and ferulic acid [15], was oxidized less efficiently by all laccases. Moreover, in the case of the phenolic dyes, indigo carmine, which has a lower potential than congo red [18,19], was the worst oxidizable substrate for all laccases.
In addition to the fact that the higher rate of substrate oxidation cannot be fully attributed to the lower substrate redox potential, the higher redox potential of laccase does not always guarantee more efficient substrate oxidation. For such substrates as syringaldazine, sinapic acid, and guaiacol, the middle redox potential SmL and AfL showed a significantly higher oxidation rate compared to the high redox potential ThL and CcL. Nevertheless, in the case of the substrates with a redox potential higher than 700 mV, the difference between the high redox potential and the middle redox potential laccases became more obvious. The high redox potential laccases (ThL and CcL) oxidized such high redox potential substrates as coumaric acid isomers, vanillic acid, vanillin, and 2,5-xylidine with a higher rate.
It is interesting that in the case of the phenolic dyes, clearer separation in the oxidation rates among studied laccases was observed. ThL and CcL oxidized all four dyes much more efficiently than AfL and SmL. Apparently, in this case the significantly higher redox potentials of ThL and CcL played a crucial role. Similar to our results, Maijala and colleagues have also shown that the high redox potential laccase from T. hirsuta oxidizes large phenolic substrates, such as mataresinol and It is worth noting that although all laccases considerably more slowly oxidized compounds with redox potentials above 700 mV (e.g., coumaric acids, vanillic acid, vanillin, 2,5-xylidine [15][16][17]), the relationship between the rate of the oxidation and the redox potential of the substrates was not so apparent for compounds with lower redox potentials. For example, caffeic acid, which has a lower potential than sinapic acid and ferulic acid [15], was oxidized less efficiently by all laccases. Moreover, in the case of the phenolic dyes, indigo carmine, which has a lower potential than congo red [18,19], was the worst oxidizable substrate for all laccases.
In addition to the fact that the higher rate of substrate oxidation cannot be fully attributed to the lower substrate redox potential, the higher redox potential of laccase does not always guarantee more efficient substrate oxidation. For such substrates as syringaldazine, sinapic acid, and guaiacol, the middle redox potential SmL and AfL showed a significantly higher oxidation rate compared to the high redox potential ThL and CcL. Nevertheless, in the case of the substrates with a redox potential higher than 700 mV, the difference between the high redox potential and the middle redox potential laccases became more obvious. The high redox potential laccases (ThL and CcL) oxidized such high redox potential substrates as coumaric acid isomers, vanillic acid, vanillin, and 2,5-xylidine with a higher rate.
It is interesting that in the case of the phenolic dyes, clearer separation in the oxidation rates among studied laccases was observed. ThL and CcL oxidized all four dyes much more efficiently than AfL and SmL. Apparently, in this case the significantly higher redox potentials of ThL and CcL played a crucial role. Similar to our results, Maijala and colleagues have also shown that the high redox potential laccase from T. hirsuta oxidizes large phenolic substrates, such as mataresinol and hydroxymataresinol, more efficiently than the middle redox potential laccases from Melanocarpus albomyces and Thielavia arenaria [20]. However, some patterns cannot be explained based on the redox potential differences alone. Despite the lower potential than SmL, AfL exhibited slightly higher activity toward all phenolic dyes except congo red.

Kinetics of Oxidation of Lignans by Laccases
To investigate the oxidation of large phenolic substrates by the studied laccases, four different lignans and lignan-like compounds, including secoisolariciresinol (SECO), mangiferin (MANG), secoisolariciresinol diglucoside (SDG), and etoposide (ETO), were chosen ( Figure 2). The chosen lignans have a different number of glycosidic moieties, which make the substrate bulkier, and possess a different flexibility of their carbon backbone due to the presence of different condensed substructures. hydroxymataresinol, more efficiently than the middle redox potential laccases from Melanocarpus albomyces and Thielavia arenaria [20]. However, some patterns cannot be explained based on the redox potential differences alone. Despite the lower potential than SmL, AfL exhibited slightly higher activity toward all phenolic dyes except congo red.

Kinetics of Oxidation of Lignans by Laccases
To investigate the oxidation of large phenolic substrates by the studied laccases, four different lignans and lignan-like compounds, including secoisolariciresinol (SECO), mangiferin (MANG), secoisolariciresinol diglucoside (SDG), and etoposide (ETO), were chosen ( Figure 2). The chosen lignans have a different number of glycosidic moieties, which make the substrate bulkier, and possess a different flexibility of their carbon backbone due to the presence of different condensed substructures. The kinetic constants of oxidation of the lignans by all four laccases are presented in Table 1. In terms of the catalytic efficiency (Vmax/KM), the obtained data can be seen as an interplay between two general tendencies. In terms of laccases, the catalytic efficiencies of oxidation were decreased in the series ThL > CcL > AfL > SmL for all substrates. In terms of substrates, the catalytic efficiencies were decreased in the series SECO > ETO > SDG > MANG for all laccases. The exception in both series was the high catalytic efficiency of MANG oxidation by AfL. Glycoside moieties of ETO, SDG, and MANG negatively affected the oxidation efficiency of all laccases except AfL in the case of ETO oxidation. The negative effect of the presence of glucose residues in the substrate molecules on the laccase catalyzed reaction rate for the T. hirsuta laccase was described previously [21]. However, in our data, CcL oxidized SECO and its glycosylated form SDG with approximately the same Vmax.  The kinetic constants of oxidation of the lignans by all four laccases are presented in Table 1. In terms of the catalytic efficiency (V max /K M ), the obtained data can be seen as an interplay between two general tendencies. In terms of laccases, the catalytic efficiencies of oxidation were decreased in the series ThL > CcL > AfL > SmL for all substrates. In terms of substrates, the catalytic efficiencies were decreased in the series SECO > ETO > SDG > MANG for all laccases. The exception in both series was the high catalytic efficiency of MANG oxidation by AfL. Glycoside moieties of ETO, SDG, and MANG negatively affected the oxidation efficiency of all laccases except AfL in the case of ETO oxidation. The negative effect of the presence of glucose residues in the substrate molecules on the laccase catalyzed reaction rate for the T. hirsuta laccase was described previously [21]. However, in our data, CcL oxidized SECO and its glycosylated form SDG with approximately the same V max . Similar to the experiments with phenolic dyes, the oxidation efficiencies of lignans oxidation can be roughly attributed to the T1 center redox potential of the laccases. Although the differences in the oxidation efficiencies of lignans between the groups of the high and middle redox potential laccases are well-explained by the redox potentials of the laccases, for the explanation of the differences within these groups some other factors must be considered. In our study, ThL, while having the same redox potential as CcL, always demonstrated a higher efficiency of lignan oxidation. Moreover, AfL, while having a slightly lower redox potential than SmL, always oxidized lignans more efficiently. Apparently, the structure of the substrate-binding pockets of laccases must play a substantial role in the described situations.
It is obvious that the influence of a substrate's structure on its oxidation efficiency must be closely linked with the architecture of the laccase's substrate-binding pockets. Among our laccases, three main differences in the substrate-binding pockets can be observed: the size of the pocket, the flexibility of the pocket-forming loops, and the presence of specific residues that can promote better substrate binding ( Figure 3 and Table S2). Similar to the experiments with phenolic dyes, the oxidation efficiencies of lignans oxidation can be roughly attributed to the T1 center redox potential of the laccases. Although the differences in the oxidation efficiencies of lignans between the groups of the high and middle redox potential laccases are well-explained by the redox potentials of the laccases, for the explanation of the differences within these groups some other factors must be considered. In our study, ThL, while having the same redox potential as CcL, always demonstrated a higher efficiency of lignan oxidation. Moreover, AfL, while having a slightly lower redox potential than SmL, always oxidized lignans more efficiently. Apparently, the structure of the substrate-binding pockets of laccases must play a substantial role in the described situations.
It is obvious that the influence of a substrate's structure on its oxidation efficiency must be closely linked with the architecture of the laccase's substrate-binding pockets. Among our laccases, three main differences in the substrate-binding pockets can be observed: the size of the pocket, the flexibility of the pocket-forming loops, and the presence of specific residues that can promote better substrate binding (Figure 3 and Table S2). While the substrate-binding pocket of ThL is the hydrophobic cavity formed by the loops near the T1 center, the substrate-binding pockets of CcL, AfL, and SmL lack one of the pocket walls ( Figure  3). In ThL, this side wall is formed by the side chains of the residues Phe162 and Arg161. In addition, the Arg161 side chain forms a hydrogen bond with the main chain of the Ser335 residue, which is a part of the loop S4 of the substrate-binding pocket on the opposite side. In CcL, AfL, and SmL, Arg161 is substituted with Ala or Gly residues (Figure 3), which cannot form the hydrogen bonds via their side chains. The length of the substrate-binding loops also differs between the laccases (Table S2, Figure 3). In AfL and SmL, compared to ThL and CcL, the loop S1 is longer and the loop S5 is shorter, which result in more spacious substrate-binding pockets. Nevertheless, it seems that a more spacious substrate-binding pocket alone does not positively affect the catalytic efficiency or the affinity of AfL and SmL to lignans.
An analysis of the temperature factors of the residues in the loops of the substrate-binding While the substrate-binding pocket of ThL is the hydrophobic cavity formed by the loops near the T1 center, the substrate-binding pockets of CcL, AfL, and SmL lack one of the pocket walls ( Figure 3). In ThL, this side wall is formed by the side chains of the residues Phe162 and Arg161. In addition, the Arg161 side chain forms a hydrogen bond with the main chain of the Ser335 residue, which is a part of the loop S4 of the substrate-binding pocket on the opposite side. In CcL, AfL, and SmL, Arg161 is substituted with Ala or Gly residues (Figure 3), which cannot form the hydrogen bonds via their side chains. The length of the substrate-binding loops also differs between the laccases (Table S2, Figure 3). In AfL and SmL, compared to ThL and CcL, the loop S1 is longer and the loop S5 is shorter, which result in more spacious substrate-binding pockets. Nevertheless, it seems that a more spacious substrate-binding pocket alone does not positively affect the catalytic efficiency or the affinity of AfL and SmL to lignans.
An analysis of the temperature factors of the residues in the loops of the substrate-binding pockets of laccases showed that the flexibility of the loops increases in the series ThL < CcL < SmL < AfL.
Earlier, Galli and colleagues showed a detrimental effect of some point mutations in the enzyme substrate-binding pocket on the efficiency of oxidation of large substrates by the Trametes versicolor laccase [9]. The substitution of F162A and F265A by Ala residues almost in all cases reduced the oxidation efficiency, presumably due to the decrease of the substrate-binding pocket's hydrophobicity and the loss of the stacking interactions with the substrate. Among the all laccases studied, only ThL contains the phenylalanine residues in these positions (Table S2).
We suppose that the main difference within the groups of the high and middle redox potential laccases is not driven only by the redox potential difference but by the structure of the substrate-binding pockets of laccases. In the case of the high redox potential laccases, the main determinant of the catalytic efficiency seems to be the presence of Phe162 and Phe265 residues in ThL that can promote tighter bonding of the substrate compared to CcL. Regarding the middle redox potential laccases, the flexibility of the substrate-binding pocket loops plays a defining role. More flexible loops in AfL can result in a more advantageous substrate binding compared to SmL.
The exceptionally high catalytic efficiency of MANG oxidation by AfL also can be attributed to the substrate-binding pocket architecture. Unlike SECO, SDG, and ETO, in MANG the hydroxy group oxidized by laccase is attached to a system of three condensed benzene rings that make this part of the substrate extremely stiff and flat. While having a wider substrate-binding pocket than ThL, AfL also has the most flexible loops among all of the studied laccases. Another feature of the AfL substrate-binding pocket is the presence of Glu207 residue instead of Asp residue as in other laccases. The residue in this position is believed to participate in substrate binding [22,23]. The longer side chain of the Glu residue is more accessible to the substrate. These features allow positioning MANG more properly than in other laccases and apparently a more proper positioning of the substrate supersedes the redox potential factor in this particular case.
In summary, a high redox potential and a presence of the hydrophobic phenylalanine residues in the key positions made ThL almost always the most effective laccase for the oxidation of the large substrates. CcL has the same redox potential as ThL, but lacked these phenylalanine residues and so was less effective. In the case of the middle redox potential laccases AfL and SmL, the flexibility of the substrate-binding loops seems to be more important than the slight difference in the T1 copper center redox potentials. Moreover, for such a bulky and rigid substrate as MANG, AfL, despite having the lowest redox potential among all of the studied laccases, was the most efficient. The fungal strains were maintained at 4 • C on slants with a wort-agar medium, prepared by diluting the brewer wort (Northern Brewery, Saint-Petersburg, Russia) with water in a 1:4 (v/v) ratio and adding 2% agar (Difco Laboratories, Detroit, MI, USA).

Enzyme Structure Analysis
The flexibility of the loops forming the substrate-binding pockets was estimated in terms of B-factors using the method described in [25]. The B-factors of the Ca atoms were normalized as Z-scores and the mean B-factor for each loop was calculated. The structural pictures were constructed with CCP4MG [26].

Conclusions
The high redox potential laccases showed higher activity toward the high redox potential (>700 mV) monophenolic substrates. However, for the low redox potential monophenolic substrates, the redox potential difference between the enzyme and the substrate did not directly influence the laccase activity.
For the oxidation of the large substrates, such as phenolic dyes and lignans, the redox potential difference between the enzyme and substrate played the defining role. High redox potential laccases almost always oxidized large substrates more efficiently that the middle potential laccases. Nevertheless, the redox potential alone cannot explain the different oxidation efficiency of laccases with similar redox potential. Hence, we propose that the structure of the substrate as well as some structural features of the laccase's substrate-binding pocket affect the resulting catalytic efficiency.
The most effective was ThL, which has a high redox potential and two Phe residues at the key positions in the substrate-binding pocket. At the same time, AfL showed higher efficiency compared to SmL, which has a greater redox potential but more rigid substrate-binding loops. Moreover, in the case of such a bulky and rigid substrate as MANG, AfL, which has the most flexible substrate-binding loops, was the most efficient laccase.