Construction and screening of CtLac mutants with enhanced laccase activity from random mutagenesis
A random mutagenesis library was constructed by error-prone PCR using GeneMorph Ⅱ EZClone Domain Mutagenesis Kit. Reaction conditions were set to introduce 6-8 mutations per 1000 bps of CtLac coding gene. A total of 1,300 transformants was randomly picked, and their laccase activities were examined using a resting cell assay in 96-well plates. Among the tested transformants, one transformant named RM484 showed a 30% increase in laccase activity than wild-type CtLac, and was selected for further experiments. The sequence of RM484 has two mutations at V243D and I468V among 516 amino acids of wild-type CtLac (Fig. S1). Considering isoleucine and valine are categorized into the same amino acid group due to their similar structures with a hydrophobic side chain, I468V was not considered a significant mutation point affecting the enzyme activity. Therefore, valine at the 243 position was assumed to be an essential factor influencing the enhanced laccase activity, and finally, it was selected as the target amino acid for site-directed mutagenesis.
Comparison of laccase activity among CtLac V243 mutants from site-directed mutagenesis based on the spectrophotometric assay and the real-time oxygen measurement at 25℃
pECtLac was used as a template to get substitutions at the position of V243; consequently, nineteen mutants were obtained from the site-directed mutagenesis. The laccase activity of the mutants was compared based on the spectrophotometric assay and the real-time oxygen measurement during the enzyme reaction at 25℃ and pH 8.0. The overall trends of increase or decrease in the product concentration and the oxygen consumption by CtLac V243 mutants were similar in both measurements (Fig. 1). Concentration of the oxidized product from 2,6-DMP by wild-type CtLac was 4.8 μM at 25℃ and pH 8.0. (Fig. 1a). The substitution of V243 by aspartic acid (D) showed 35% enhanced enzyme activity than wild-type CtLac under the spectrophotometric assay. The other substitutions showed varied enzyme activities. Specifically, six substitutions, including serine (S), threonine (T), asparagine (N), cysteine (C), glycine (G), proline (P), showed about 10% decreased enzyme activity than wild-type CtLac under the spectrophotometric assay (Fig. 1a). In addition, ten substituents including arginine (R), histidine (H), glutamic acid (E), glutamine (Q), alanine (A), isoleucine (I), leucine (L), phenylalanine (F), tyrosine (Y), tryptophan (W) showed about 20-40% decreased enzyme activity than wild-type CtLac. Remarkably, substitutions of V243 by lysine (K) or methionine (M) showed highly decreased laccase activity to 18.3% and 17.0%, respectively, compared to wild-type CtLac.
Oxygen consumption was evaluated by the real-time measurement of the dissolved oxygen in the reaction mixture during the enzyme reaction. The initial dissolved oxygen value of the reaction mixture was stabilized for 3 min and was around 295-310 μM (Fig. S2). After stabilizing the dissolved oxygen value, each enzyme was added to the reaction mixture. A bumping line was observed while the needle-type optical oxygen sensor was dipping into the reaction mixture through Teflon-lined rubber septa. This might be from the insertion of oxygen in the enzyme solution into the reaction mixture. It was commonly observed in all reactions within 5 μM; therefore, it was considered the initial dissolved oxygen concentration. Only V243D showed 25% higher enzyme activity among the mutants than wild-type CtLac (Fig. 1b).
It was found that there was considerable overestimation in the laccase activity from the result of the spectrophotometric assay. For example, six substitutions, including S, T, N, C, G, and P, exhibited only 10% decreased enzyme activity than wild-type CtLac from the spectrophotometric assay. However, they showed 18~38% decreased enzyme activity from the real-time oxygen measurement. This indicates one of the limitations of the spectrophotometric assay in evaluating oxidase enzyme activity, even though the assay method is simple and easy to perform. During the spectrophotometric assay, the result values could have interfered from various factors such as a limit of detection of substrates or products, a need for absorbance at a particular wavelength, and unintended side products by enzyme reactions [23].
During the laccase catalysis of 2,6-DMP, it undergoes single-electron oxidation and transforms to 2,6-dimethoxy-phenoxy radical species. Once the unpaired electron resonates with p-radical species in the benzene ring, the consecutive dimeric or polymeric reaction will be undergone non-specifically. There have been reported several dimeric compounds resulted from 2,6-dimethoxy-phenoxy radicals such as 3,3′,5,5′-tetramethoxydiphenoquinone, 3,3′,5,5′-tetramethoxybiphenyl-4,4′-diol, and 4-(2,6-dimethoxy-phenoxy)-2,6-dimethoxyphenol from 2,6-DMP by laccase-catalyzed reaction [24-26]. In addition, polymeric oxidative products such as C1-C1 biphenyl polymer or C1-O-C1 biphenyl ether polymers could be produced from consecutive radical-driven reactions [25]. However, the spectrophotometric assay generally targets to detect only 3,3′,5,5′-tetramethoxydiphenoquinone because of the absorbance of quinone-based structure at 468 nm. In other words, the spectrophotometric assay absorbs all types of quinone-based products from the non-specific reaction of radicals during the enzyme reaction, which could make an inconclusive result in the evaluation and comparison of the enzyme activity.
Furthermore, the spectrophotometric assay to measure the oxidized products from laccase activity could be influenced by assay conditions like pH and the reaction time. For example, phenolic compounds such as 2,6-DMP and caffeic acid show an increase in auto-oxidation above pH 8.5, while non-phenolic compounds such as 2,2′-azino-bis (3-ethylbenzothiazoline-6-sulfonate) (ABTS) and syringaldazine (SGZ) exhibit no auto-oxidation at a pH range of 2.6-8.0 [27]. Therefore, selecting optimal substrates is very important to determine laccase activity using the spectrophotometric assay.
Meanwhile, the measurement of oxygen consumption during the laccase reaction circumvents these drawbacks from the spectrophotometric assay. Laccase catalyzes single-electron oxidation of substrate and transfers four electrons to one molecule of oxygen which is reduced to two molecules of water. Therefore, comparing oxygen consumption might be an obvious criterion for evaluating enzyme activity. For example, it could be applied to develop a laccase-mediator-system (LMS) to improve the low redox potential of laccase itself [28-30]. Euring et al. evaluated the performance of different natural mediators in enzymatic oxidation of lignin by comparing oxygen consumption rates in different LMS [29]. It found that the mediator, 2,6-DMP, significantly enhanced the oxidation of lignosulfonate and Indulin AT with a more decrease in oxygen saturation. On the other hand, oxygen consumption of LMS with the help of ABTS was faster than that of 1-hydroxy-benzotriazole (HBT) supported-LMS in the oxidation of guaiacylglycerol-β-guaiacyl ether (GGGE) [30]. More interestingly, these LMS resulted in different oxidation pathways: HBT-LMS led to the extensive polymerization of GGGE; however, ABTS-LMS directly oxidized GGGE by both laccase and ABTS radical cations, making a stable product instead of further polymerization. These implied that the detection of the specific oxidized product by laccase reaction would not be a direct way to evaluate the enzyme activity. Furthermore, it would be very hard to predict various polymerized products in cases of complex substrates like lignin.
The measurement of oxygen consumption during the laccase reaction could be a more accurate method to evaluate and compare the enzyme activity among mutants than the spectrophotometric assay, especially in mutagenesis research. Moreover, the oxygen consumption measurement could be applied to other oxidases and oxygenase enzymes.
Comparison of biochemical characterization of purified CtLac V243 mutants with wild-type CtLac based on the spectrophotometric assay
Among nineteen V243 mutants, V243D was selected as the best mutant due to its highest laccase activity based on both the spectrophotometric assay and real-time oxygen measurement (Fig. 1). Therefore, its biochemical characterization was compared with wild-type CtLac and a mutant V243M, which showed the least laccase activity. The biochemical characterization of the selected V243D and V243M was carried out only based on the spectrophotometric assay because the needle-type optical fiber oxygen sensor has a limit to temperature ranges up to 50℃. V243D showed 16-42% enhanced laccase activity than wild-type CtLac in the tested ranges of temperatures from 20℃ to 90℃ (Fig. 2a). However, the laccase activity of V243D was not affected significantly by the change in pH except for pH 8.0 (Fig. 2b). Especially, V243D showed the most enhanced laccase activity at 70℃ and pH 8.0 by producing 25.3 μM of the oxidized product of 2,6-DMP, while wild-type CtLac produced 19.1 μM. Thus, we assumed that the condition of 70℃ and pH 8.0 is the optimal reaction condition for the enzymes V243D and wild-type CtLac. On the other hand, V243M showed about 80-90% decreased enzyme activity compared to wild-type CtLac, even at the optimal enzyme conditions, 70℃ and pH 8.0.
From the screening of mutants with the enhanced laccase activity based on the spectrophotometric assay and the real-time oxygen consumption measurement, two mutants V243K and V243M showed significantly decreased laccase activity compared to wild-type CtLac (Fig. 1). In the case of V243K, the substitution of Val to Lys might reduce the solubility of the protein because inclusion body was observed in the cell lysis step of the crude laccase enzyme preparation (data not shown). It might infer that V243K could not induce the appropriate conformational change of the enzyme, resulting in the significant loss of enzyme activity.
Meanwhile, the decrease in the laccase activity of V243M might be from the copper binding issue to the T1 copper binding site of laccase or the vicinity of the catalytic coppers. This could be deduced from the color difference of E. coli (pECtLac-V243M) cell pellet. To be specific, the color of the cell pellets of E. coli (pECtLac) and E. coli (pECtLac-V243D) turned greenish during the protein overexpression in the presence of CuCl2; however, the color of the cell pellet of E. coli (pECtLac-V243M) remained yellowish (Fig. S3). The green color might be from the coordination of Cu2+ in the Type 1 site of laccase, which is the characteristic adsorption of the oxidized Cu2+ state at 600 nm [31]. Therefore, there are possibilities that the substitution of Val to Met might affect Cu binding activity of wild-type CtLac. In other words, V243D might induce favorable conformational changes in the enzyme’s active site than wild-type CtLac because, generally, the catalytic efficiency of laccase is proportional to the redox potential of type 1 copper. Although this deduction might be imperfect without the exact protein structure of wild-type CtLac, it could consider Val at 243 position of CtLac to be one of the key amino acids contributing to the enzyme’s catalytic efficiency due to a noticeable difference among the mutants. Although it is necessary to predict critical residues of CtLac by using protein structure data of other laccases available from protein data bank (PDB), we found the low amino acid sequence (< 30%) with that of previously known structure for bacterial laccase [32]. Therefore, further study is needed based on the protein structural data of wild-type CtLac and the mutants to prove the abovementioned assumption.
Comparison of kinetic parameters of CtLac V243 mutants with wild-type CtLac based on the spectrophotometric assay and the real-time oxygen consumption measurement at 25℃ and pH 8.0
Kinetic parameters of V243D and V243M were investigated based on the spectrophotometric assay and the real-time oxygen consumption measurement with the different concentrations of 2,6-DMP. The data were analyzed using Michaelis-Menten equation by non-linear regression and compared with wild-type CtLac. From the spectrophotometric assay, V243D exhibited Km and kcat of 137.2 μM and 2.7 s-1 at 25℃ and pH 8.0, respectively, which was 12.7% lower and 145.5% higher than those of wild-type CtLac (Table 1a). The kcat/Km value of V243D was determined to be 0.02 s-1μM-1, which was about 2.9-fold larger in catalytic efficiency than wild-type CtLac. Especially, the specific activity of 10 μg purified V243D in the presence of 500 μM 2,6-DMP at 25℃ and pH 8.0 was 347.7±16.2 μmol/min/mg, which was 2.3-fold higher than that of wild-type CtLac. However, V243M showed Km and kcat of 674.2 μM and 0.1 s-1 at 25℃ and pH 8.0, respectively, which was 329.0% higher and 91.0% decreased values of wild-type CtLac.
Kinetic parameters of the mutants and wild-type CtLac were also obtained from the oxygen consumption measurement (Table 1b). The initial dissolved oxygen concentration in the reaction mixture containing different concentrations of 2,6-DMP was equilibrated as described above (Fig. S4). V243D exhibited Km(O2) and kcat(O2) values of 332.3 μM and 4.2 s-1, respectively, which were 33.1% lower and 20.0% higher than those of wild-type CtLac at 25℃ and pH 8.0 (Table 1b). However, V243M showed Km(O2) and kcat(O2) of 87.7 μM and 0.7 s-1 at 25℃ and pH 8.0, respectively, which was 82.4% lower and 80.0% decreased values of wild-type CtLac. There was a distinct difference between the oxygen consumption rate (μM O2/min) among wild-type CtLac, V243D, and V243M in the presence of 500 μM 2,6-DMP at 25℃ and pH 8.0, following the order V243D > wild-type CtLac > V243M with the values of 4.2, 2.4, and 0.6 (Fig. 3).
Comparison of kinetic parameters of CtLac V243 mutants with wild-type CtLac based on the spectrophotometric assay under the optimal enzyme conditions at 70℃ and pH 8.0
As investigated in our previous study, CtLac showed the highest laccase activity at a high temperature of 70℃ and pH 8.0 because it originated from thermoalkaliphilic bacterium, Caldalkalibacillus thermarum strain TA2.A1 [21]. In other words, the absolute concentrations of the oxidized product from wild-type CtLac and V243 mutants were substantially higher at 70℃ than those at 25℃ (Fig. S5 and Fig. 1b). Therefore, the investigation of kinetic parameters of V243D and V243M was additionally carried out at 70℃ and pH 8.0 to compare with those of wild-type CtLac. Notably, V243D showed 127.3 μM and 15.3 s-1 for Km and kcat values, respectively, which was 13.6% lower and 19.5% higher than wild-type CtLac (Table 2). Moreover, the kcat/Km value of V243D was 0.12 s-1μM-1 which was 1.4-fold larger in catalysis efficiency compared to wild-type CtLac. Moreover, the specific activity of 10 μg purified enzyme in the presence of 500 uM 2,6-DMP at 70℃ and pH 8.0 was 2141.8 μmol/min/mg, which was 1.1 fold higher than that of wild-type CtLac. However, V243M showed Km and kcat of 1127 μM and 1.5 s-1 at 70℃ and pH 8.0, respectively, which was 665.1% higher and 88.3% decreased values of wild-type CtLac.
The mutation at the V243 position of wild-type CtLac affected laccase activity by resulting in changes in the kinetics parameters. Especially, V243D showed improved enzyme activity with the decreased Km and increased kcat. Thus, our results could provide a path to improve catalytic activity of CtLac with further investigation of molecular structure. Indeed, most investigations on mutagenesis to enhance enzyme activity have been performed between well-known laccases with high amino acid identities, such as Bacillus licheniformis and Bacillus subtilis [33-35].
In the present study, we performed random and site-directed mutagenesis and compared enzyme activity using the spectrophotometric assay and the real-time oxygen measurement to select the best mutant with enhanced enzyme activity, although there were no references in the protein structures. Therefore, this atypical trial could be applied as a more powerful tool to evaluate and broaden the range of novel oxidase enzymes.
Enhancement of oxidative degradation of lignin model compound and rice straw using V243D at 70℃ and pH 8.0
Enhancement of lignin degradation by V243D was determined using GGGE and rice straw. Purified V243D and V243M were treated on each substrate at 70℃ and pH 8.0, and the enzymatic activity on the oxidative lignin degradation was compared with that of wild-type CtLac. After 6 h of the enzymatic treatment, 10% higher GGGE oxidation was observed in V243D than wild-type CtLac, while V243M slightly oxidized GGGE (Fig. 4a). The m/z value of the product was 637.2 of deprotonated ion in the negative ionization mode [M-H]-, which was expected as a dimer form of GGGE (Fig. 4b). In addition, the collision-induced dissociation (CID) fragmentation patterns of the product from GGGE by V243D treatment was exactly matched with the results of wild-type CtLac in our previous report [21]. Briefly, there were typical neutral losses of 15 Da (CH3•, methyl radical), 18 Da (H2O, water), 30 Da (CH2O, formaldehyde), 48 Da (H2O+CH2O), and 124 Da (C7H8O2, guaiacol) among fragment ions, which are well known as the characteristic fragmentations from β-aryl ether linkage, which account for more than 50% of inter-unit linkage found in all types of lignin [36, 37]. The m/z values of major fragment ions were 589.1, 541.1, 483.1, 435.1, 329.1, which indicated [M-H-H2O-CH2O]-, [M-H-2H2O-2CH2O]-, [M-H-guaiacol-CH2O]-, [M-H-2guaiacol-CH2O]- (Fig. 4b). These m/z fragment patterns were exactly matched with our previous results of CtLac treatment on GGGE [21], in other words, GGGE might be transformed to C5-C5 biphenyl tetramer by the initial enzymatic oxidation and radical-radical coupling (Fig. 4c). Considering V243D has lower Km and higher kcat value than wild-type CtLac, it might enhance oxidation activity on GGGE than wild-type CtLac.
Furthermore, V243D and V243M were treated on rice straw, representative lignocellulosic biomass, and the production of lignin-derived monomers was compared with wild-type CtLac. The six lignin-derived monomers were detected at m/z values of (a) 167.03, (b) 195.06, (c) 121.02, (d) 163.03, (e) 151.03, (f) 181.05 and identified as vanillic acid, homosyringaldehyde, p-hydroxybenzaldehyde, p-coumaric acid, vanillin, and syringaldehyde, respectively (Fig. 5). The lignin monomers were quantified using the peak area of extracted ion chromatogram (EIC) of compounds. The peak area was used to calculate the concentration of the compounds as compared to the EIC of authentic vanillin (0-200 μM). The above six lignin-derived compounds commonly have a phenolic group that is deprotonated in negative ionization mode. Therefore, the compounds’ concentrations could be calculated from this calibration curve, even in a lack of authentic chemicals such as homosyringaldehyde.
Specifically, the concentration of homosyringaldehyde, p-hydroxybenzaldehyde, and vanillin was 17.8 μM, 91.7 μM, and 98.0 μM, respectively, after 3 h of V243D treatment (Table 3). On the other hand, wild-type CtLac produced less concentration of compounds as 13.8 μM, 74.2 μM, and 90.2 μM than V243D. Significantly, vanillin was still increased at 6 h of enzyme treatment, even though homosyringaldehyde and p-hydroxybenzaldehyde were decreased due to possible employment in the laccase activity as mediators. On the other hand, V243M produced similar concentrations of these benzaldehyde compounds to the control sample containing only rice straw without enzymes, which meant no significant effect of V243M on the release of lignin monomers due to its low enzyme activity. The total concentration of lignin-derived monomer from rice straw was around 265 μM after V243D treatment, which was higher than wild-type CtLac. Consequently, V243D produced much more high-value benzaldehyde compounds from rice-straw than wild-type CtLac.
It can be assumed that the action of CtLac on the rice straw would be different from that on simple phenolic compounds such as 2,6-DMP and GGGE. CtLac could oxidize 2,6-DMP by directly abstracting single-electron from this phenolic substrate, resulting in phenoxyl radicals and sequential radical coupling. However, CtLac could not directly abstract electrons from the non-phenolic group of lignin due to its intrinsic low redox potential of laccase than the non-phenolics [38]. As proposed in our previous study, it can be overcome by the action of lignin-derived monomers from rice straw as natural mediators in CtLac-driven biodegradation of terminal β-O-4 type bond in rice straw [22]. These mediators oxidized by CtLac might abstract electrons from the C1 carbon of β-O-4 type lignin, and the C1 radical-driven cleavage of Cα-Cβ bond produced benzaldehyde such as vanillin. Based on this proposed mechanism, V243D could oxidize a larger amount of lignin-derived monomers than wild-type CtLac, resulting in much-oxidized mediators with faster oxygen consumption during the enzyme treatment. Consequently, V243D could foster the production of much high-value benzaldehyde chemicals such as p-hydroxybenzaldehyde and vanillin than wild-type CtLac.