Targeted Mutations in a Trametes villosa Laccase

Trametes villosa laccase was mutated on a tetrapeptide segment near the type 1 site. The mutations F463M and F463L were at the position corresponding to the type 1 copper axial methionine (M517) ligand in Zucchini ascorbate oxidase. The mutations E460S and A461E were near the T1 copper site. The mutatedTrametes laccases were expressed in an Aspergillus oryzae host and characterized. The E460S mutation failed to produce a transformant with meaningful expression. The F463L and A461E mutations did not significantly alter the molecular and enzymological properties of the laccase. In contrast, the F463M mutation resulted in a type 1 copper site with an EPR signal intermediate between that of the wild type laccase and plastocyanin, an altered UV-visible spectrum, and a decreased redox potential (by 0.1 V). In oxidizing phenolic substrate, the mutation led to a more basic optimal pH as well as an increase in k cat and K m . These effects are attributed to a significant perturbation of the T1 copper center caused by the coordination of the axial methionine (M463) ligand.

Trametes villosa laccase was mutated on a tetrapeptide segment near the type 1 site. The mutations F463M and F463L were at the position corresponding to the type 1 copper axial methionine (M517) ligand in Zucchini ascorbate oxidase. The mutations E460S and A461E were near the T1 copper site. The mutated Trametes laccases were expressed in an Aspergillus oryzae host and characterized. The E460S mutation failed to produce a transformant with meaningful expression. The F463L and A461E mutations did not significantly alter the molecular and enzymological properties of the laccase. In contrast, the F463M mutation resulted in a type 1 copper site with an EPR signal intermediate between that of the wild type laccase and plastocyanin, an altered UV-visible spectrum, and a decreased redox potential (by 0.1 V). In oxidizing phenolic substrate, the mutation led to a more basic optimal pH as well as an increase in k cat and K m . These effects are attributed to a significant perturbation of the T1 copper center caused by the coordination of the axial methionine (M463) ligand.
Laccase (p-diphenol:dioxygen oxidoreductase, EC 1.10.3.2) is a copper-containing oxidase that couples the oxidation of substrate (usually diphenols, aryl diamines, or amino phenols) with the reduction of dioxygen to water (1)(2)(3)(4). Sequence comparisons, crystal structure analyses, and spectroscopy indicate that all multicopper oxidases contain at least one type 1 (T1) 1 copper center, one type 2 (T2) copper center, and one type 3 (T3) copper center (3)(4)(5)(6)(7). These copper sites are defined by their spectroscopic properties. The T1, or "blue," copper is characterized by an intense (⑀ ϳ 5000 M Ϫ1 cm Ϫ1 ) absorption band around 600 nm and an unusually small (Ͻ100 ϫ 10 Ϫ4 cm Ϫ1 ) parallel hyperfine coupling in EPR. The T2, or normal, copper site does not exhibit a strong feature in the visible absorption spectrum, but has a parallel hyperfine coupling Ͼ 160 ϫ 10 Ϫ4 cm Ϫ1 . The T3, or coupled binuclear, copper center consists of two copper atoms connected by a hydroxide bridge. This bridge provides a strong superexchange pathway and therefore mediates antiferromagnetic coupling. The T3 center is EPR-silent, but is characterized by an absorption band around 330 nm (⑀ ϳ 5000 M Ϫ1 cm Ϫ1 ). The T2 and T3 sites form a trinuclear copper cluster that is the site for O 2 reduction (5,8,9). Laccase is the simplest of the multicopper oxidases, containing one of each type of copper site for a total of 4 copper atoms.
Based on a wide range of comparative studies, including sequence homology and crystal structure analysis, the copper site coordination is very similar among the multicopper oxidases. The main difference is in the coordination sphere of the T1 copper site. The typical T1 site, such as that found in the blue copper protein plastocyanin (Pc) and the multi-copper protein Zucchini ascorbate oxidase (zAO), contains two histidines (His), a cysteine (Cys) that forms a short S-Cu bond, and a methionine (Met) that forms a long S-Cu bond. These four ligands bind T1 copper in a distorted tetrahedral coordination geometry (Fig. 1, A and B). In contrast, most fungal laccases with known primary sequence have either a leucine (Leu) or phenylalanine (Phe) at the position corresponding to the axial Met ligand (3,5). Neither Leu nor Phe would be expected to coordinate to the copper that would render a tri-coordinate T1 site. The recent crystal structure of Coprinus cinereus laccase (CcL) confirms this and shows that the T1 site has only three ligands (two His and one Cys) in a trigonal planar geometry ( Fig. 1C) (7). An interesting question is how this difference in geometry is manifested in terms of differences in the electronic structure and the electron transfer function of the T1 site. The redox potential (E°) of fungal laccases ranges from 0.48 V to 0.78 V, whereas Pc and zAO have E°in the range of 0.3 to 0.4 V (3). It has been proposed that the lack of an axial ligand might be responsible for the high E°observed in some fungal laccases (10 -12), as it is expected that the elimination of the axial Met donor interaction would preferentially stabilize the reduced Cu(I) state. A number of mutagenesis studies (13)(14)(15) have been performed on the T1 copper site in azurin, and it was found that the nature of the axial ligand can influence E°to some extent. Mutation of Met-121 was found to tune the E°o ver a range of Ϫ0.105 V to 0.138 V with respect to the E°of the wild-type (wt) azurin at pH 7.0.
The present study utilized site-directed mutagenesis to examine how the composition of the amino acid residues in the vicinity of the T1 copper affect the structure and E°of this copper in a Trametes villosa laccase (TvL) (16). We were also interested in determining how the mutations would impact the enzymatic properties, k cat , K m , and the pH dependence of en-* This work was partially supported by National Institutes of Health Grant DK 31450 (to E. I. S.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

EXPERIMENTAL PROCEDURES
Materials, Methods, and Enzyme Assays-Chemicals used as buffers and substrates were commercial products of at least reagent grade. Britton and Robinson buffer (pH 2.7-11, made by mixing 0.1 M boric acid, 0.1 M acetic acid, and 0.1 M phosphoric acid with 0.5 M NaOH to desired pH) was used for the pH profile. Recombinant TvL (isoform-1) was purified as reported previously (16). The protocols for molecular biology experiments (including restriction digests, DNA ligations, gel electrophoresis, and DNA preparations) were adapted from either the instruction of the manufacturer or standard procedures. Oligonucleotides were synthesized by an Applied Biosystems model 294 DNA/RNA synthesizer. Nucleotide sequences were determined by an Applied Biosystems automatic DNA sequencer, model 373A-1.2.0. EPR spectra were obtained using a Bruker ER 220-D-SRC spectrometer. All samples were run at 77 K in a liquid nitrogen finger Dewar flask. Spectrometer settings were 10 milliwatts (or 13 decibels) power, 500 ms time constant, 20 Gauss (G) modulation, 1100 G sweep width, and 3050 G center field. The ground state spin Hamiltonian parameters g Ќ , g ʈ , and A ʈ were obtained from simulations using the Quantum Chemistry Program Exchange (QCPE) computer program "sim 15." Atomic absorption spectroscopy was performed on a Perkin-Elmer 2380 instrument. The UVvisible absorption spectra were recorded on a Shimadzu UV160U spectrophotometer with 1-cm quartz cuvette. The extinction coefficient at 280 nm was determined by amino acid analysis. The E°of T1 copper in laccase was measured as reported previously (17) using the redox titrant couple Fe(2,2Ј-dipyridyl) 2 Cl 3 /Fe(2,2Ј-dipyridyl) 2 Cl 2 . Laccase-catalyzed oxidation of syringaldazine (SGZ) or 2,2Ј-azinobis-(3-ethylbenzothiazoline-6-sulfonic acid) (ABTS) and kinetic analysis were performed as reported previously (18). These two substrates were chosen (to assay the activity of the mutants) because of the two classes of the pH activity profiles that they possess (19).
Preparation of TvL Mutants-First, a 1.5-kilobase BamHI-HindIII fragment encoding the T1 copper-ligating peptide segment of TvL was subcloned from pDSY10 (16) into pUC118 (20). Second, a singlestranded DNA was prepared from this intermediate and used as a template for site-directed mutagenesis (21) with the primers 5Ј-dCTC-GAGGCCGGCCTCGCCGTCGTG-3Ј (for F463L mutant), 5Ј-dGACTTC-CACCTCTCGGCCGGCTTC-3Ј (for E460S mutant), 5Ј-dCACCTCGAG-GAGGGCCTCGCCGTCGTG-3Ј (for A461E/F463L mutant), and 5Ј-dCTCGAGGCCGGCATGGCCGTCGTG-3Ј (for F463M mutant). The mutants were identified by hybridization with the corresponding radiolabeled oligonucleotide primer and verified by DNA sequence analysis. The DNA segments encoding TvL mutants were reinserted into pDSY10, then used to transform an Aspergillus oryzae host. The transformation was carried out by a procedure previously described (18). Cultures producing high levels of extracellular laccase activity were fermented in a laboratory-scale fermentor. The purification of the recombinant laccases was similar to that described previously (16). Recovery yield of 15, 8, and 24%, and 300-, 540-, and 40-fold purification were obtained for the F463L, A461E/F463L, and F463M mutants, re-spectively. Overall, the transformation and expression of the F463M, F463L, and A461E/F463L mutants were comparable to the wt TvL, and the expressed three mutants exhibited similar chromatographic elution patterns to the wt TvL. Because of the lack of detectable laccase activity in shaker flask, the E460S mutant was not purified and characterized.

RESULTS AND DISCUSSION
F463L Mutant and A461E/F463L Double Mutant-The purified mutants each had an UV-visible spectrum quite similar to the wt TvL (Fig. 2, Table II). The mutant enzymes had ϳ 4 copper/protein and ϳ 50% paramagnetism. The EPR spectrum of the F463L mutant was obtained and showed a slight increase in g ʈ and decrease in A ʈ at the T1 site and a small change in g ʈ of the T2 copper site (Fig. 3, Table II). However, these changes were minor, indicating that the F463L mutation did not significantly alter the electronic structure of the T1 site. In 10 mM MES, pH 5.5, F463L exhibited an E°(T1 copper) that was only 0.05 V lower than that of the wt TvL (Table III). This suggests that changing the non-coordinating, bulky, electronrich Phe to Leu did not have a dramatic impact on the T1 copper site. This result is consistent with a previous study in which mutation of Leu to Phe in Rhizoctonia solani and Myceliophthora thermophila laccases did not result in alteration of the EPR parameters and E°for their T1 copper (18). Table III and Fig. 4 summarize the SGZ and ABTS oxidase activities of both the mutants. K m , k cat , and optimal pH (pH opt ) were not significantly altered in either the single or the double mutant. The lack of a significant effect of these mutations on the electronic properties of the T1 copper is consistent with the lack of an impact on the kinetic parameters of the enzyme. Previous studies have shown that, under steady state conditions, the rate-limiting step in the catalytic cycle is most likely the oxidation of the substrate by transfer of an electron from the   (22). If the electronic structure of the T1 copper were significantly altered by a mutation-induced perturbation of the site, it would seem likely that the rate of substrate oxidation would also be affected. F463M Mutant-The transformation and expression of the F463M mutant was comparable to that of the wt TvL. The expressed mutant exhibited a chromatographic elution pattern similar to that of the wt TvL. Copper content measurements revealed a stoichiometry number close to 4 copper/protein for the wt and mutant. The purified protein had an UV-visible absorption spectrum characteristic of laccase (Fig. 2, Table II). The absorption band at 600 nm had a maximal absorption wavelength ( max ) shifted 13 nm lower in energy, and the shoulder around 750 nm is more pronounced, indicating that the mutation perturbed the T1 copper. To further assess the extent of structural and electronic changes of the T1 and T2 copper centers, we obtained the EPR spectra for the mutant and wt enzymes (Fig. 3) and simulated the spectra to extract the ground state spin Hamiltonian parameters (Table II). The T1 copper site of the F463M mutant exhibited an increased g ʈ (2.214 as compared with 2.194 in the wt enzyme) and a decreased A ʈ (78 ϫ 10 Ϫ4 cm Ϫ1 as compared with 90 ϫ 10 Ϫ4 cm Ϫ1 for the wt enzyme). The T2 copper remained unaltered. These changes are consistent with the T1 copper site becoming more like the site in Pc and Rhus vernicifera laccase (RvL). The EPR spectrum of Pc is characterized by a g ʈ of 2.226 and an A ʈ of 63 ϫ 10 Ϫ4 cm Ϫ1 , and RvL has a T1 copper with g ʈ of 2.30 and A ʈ of 43 ϫ 10 Ϫ4 cm Ϫ1 . While there is no sequence available for RvL, the enzyme is believed to have a ligating Met in the axial position of the T1 copper site, a structure analogous to that of Pc and zAO (Fig. 1). Both the absorption and the EPR results indicate that the mutation led to a significant change in the electronic structure of the T1 copper in TvL.
In 10 mM MES, pH 5.5, the mutant exhibited an E°0.1 V lower than that of the wt TvL (Table III). It has been proposed that the additional Met ligand would stabilize the oxidized Cu(II) over the reduced Cu(I) site, thus leading to a lower E° (3,10,11). Our EPR and absorption data strongly suggested that the introduction of Met significantly perturbed the T1 site and that Met probably coordinated to the T1 copper based on the spectral similarities to Pc. The decrease in the E°was also consistent with this interpretation. The extent of the E°change (0.1 V) agrees with previous studies on the T1 copper site in azurin in which the reverse mutation was carried out; replacement of Met by a non-ligating Leu led to a 0.1-V increase in E° ( 13,14). These observations support the hypothesis that ligation of the axial amino acid can tune the E°to some extent. However it is important to note that the E°of 0.68 V observed for the F463M mutant is still 0.1-0.4 V higher than that of zAO, human serum ceruloplasmin, and Myrothecium verrucaria bilirubin oxidase, whose T1 copper coordination also includes the axial Met (3,4,17). Thus, other factors such as hydrogen bonding, solvent accessibility, and orientation of local dipoles clearly play an important role in tuning the E° (4,23).
The pH profiles of the E°for the wt and F463M mutant are depicted in Fig. 5. The profile of both enzymes changed very little between pH 4 and 8 but showed an increase at pH Ͼ 8. This increase was more pronounced in the F463M mutant. It has been reported that the E°of Pc and azurin increases as the pH decreases, due to the titration of a T1 copper-ligating H in the reduced site (24). The E°(T1) of RvL also increases when pH decreases (25). Thus for TvL, the pH effect on E°was in the opposite direction. Fig. 5 also shows the pH profile of max . No change was observed in either the wt or F463M mutant over pH 2 through 8 but at pH Ͼ 8, a significant decrease in max was observed. This change in max indicates a pH-dependent structural change of the T1 copper site. While further spectroscopic studies are needed to elucidate the origin of this change, it may be noteworthy that the decrease in max coincides with the increase in E°.
The F463M mutation led to a 5-fold increase in K m for the SGZ substrate and a 38-fold increase in K m for the ABTS substrate (at pH 5.5). The mutation also resulted in a 2-fold increase in k cat (Table III). For an electron-transfer reaction (such as the one between the substrate and the T1 copper), its rate (k ET ) is determined by thermodynamic driving force (E°d ifference between donor and acceptor), electronic coupling (orbital overlap between donor and acceptor), and Frank-Condon barrier (reorganization energy associated with electron transfer) (26). The parameter k cat is reflective of the ratelimiting step that was shown to be the substrate oxidation under steady state conditions (22). While a number of steps may contribute to k cat , the fact that k cat was strongly correlated with the difference in E°between the substrate and the T1 copper for a series of substrates suggests that k ET is a major component of k cat . Our data illustrate an increase in k cat of substrate oxidation that can be explained by an increase in k ET . The decreased E°of the T1 copper in the F463M mutant would lead to a decreased thermodynamic driving force and thus a lower k ET . Thus, in order to account for the increase in k cat the mutation must have affected the electronic coupling and/or the reorganization energy. The EPR spectrum of F463M provides initial evidence that the mutation caused a change in the ground state of the T1 copper. This could result in a change in the orbital overlap between the donor (substrate) and acceptor (the half occupied highest molecular orbital of the T1 copper). There is some evidence from ENDOR that the spin density on the imidazole nitrogen ligands of different T1 copper sites can vary, depending on the nature of the ground state (27). This  difference could influence the orbital overlap between the donor and acceptor. It is also possible that the mutation altered the Frank-Condon barrier to electron transfer. Evaluation of electronic structure differences between wt and F463M, which would relate to this difference in reactivity, is presently under way. The F463M mutation led to a shift in pH opt for the oxidation of SGZ but not ABTS (Fig. 4). It has been shown that oxidation of a phenolic substrate depends on whether it is protonated; the deprotonated phenol has a lower redox potential and therefore is more easily oxidized (19). The fact that the pH dependence of activity was only altered when SGZ, a phenolic substrate, was used but not when ABTS, a non-phenolic substrate, was used indicates that the mutation must have caused a slight change in the substrate-binding pocket. This is also consistent with the change in K m and could perturb the protonation equilibrium of the phenolic substrate.
Concluding Remarks-Understanding how the E°of copper sites in proteins is regulated and how E°and geometric and electronic structure perturbations influence the electron transfer function of a protein is one of the major challenges in the field of metallo-biochemistry. Various theories have been proposed to explain how two seemingly similar T1 copper sites could have different E°and different reactivity (1)(2)(3)(4)17). For these comparisons, it is important to obtain a detailed description of the electronic structure of the copper in a similar envi-ronment. Our results provide initial evidence that a more "classic" blue copper site can be created in fungal laccase through site-directed mutagenesis and that the resulting electronic structure appears to be intermediate between the classic Pc T1 copper site and that of the wt fungal laccase, which has no axial Met. This leads to a lowering of E°in the same protein environment consistent with the axial Met stabilizing the oxidized, Cu(II) state more than the reduced, Cu(I) state. Detailed spectral studies are currently being pursued to understand the effects of the associated electronic structure changes on the reactivity of these enzymes.