- Email: [email protected]
Crystal structure of the blue multicopper oxidase from the white-rot fungus Trametes trogii complexed with p-toluate
Inorganica Chimica Acta 361 (2008) 4129–4137
Contents lists available at ScienceDirect
Inorganica Chimica Acta journal homepage: www.elsevier.com/locate/ica
Crystal structure of the blue multicopper oxidase from the white-rot fungus Trametes trogii complexed with p-toluate Irene Matera, Antonella Gullotto, Silvia Tilli, Marta Ferraroni, Andrea Scozzafava *, Fabrizio Briganti Bioinorganic Chemistry Laboratory, Department of Chemistry, University of Florence, Via della Lastruccia 3, 50019 Sesto Fiorentino, Florence, Italy
a r t i c l e
i n f o
Article history: Received 24 January 2008 Received in revised form 17 March 2008 Accepted 17 March 2008 Available online 23 March 2008 Dedicated to Professor Dante Gatteschi. Keywords: Trametes trogii Funalia trogii laccase X-ray structure
a b s t r a c t A multicopper oxidase, the fungal laccase glycoenzyme from the white-rot basidiomycete fungus Trametes (Funalia) trogii, was crystallized and its crystal structure was solved at 1.58 Å using molecular replacement techniques. Model refinement resulted in R-factor and R-free values of 17.4% and 19.0%, respectively. The T. trogii laccase structural model reveals the presence of a ligand bound to the T1 active site which resembles a ptoluate molecule, such bound compound is most probably a fungal metabolite. The p-toluate is bound into the T1 active site of the laccase forming, with one of the carboxylate oxygens, a H-bond with His455, one of the T1 copper ion ligands, whereas the methyl group presents hydrophobic interactions within a pocket composed by Phe331, Phe336, Pro390 and Val162. The coordination geometries, the bond distances and the oxidation states of the T1 and T2/T3 copper active sites are analyzed and discussed in terms of the enzymatic mechanism and catalytic functionality. Ó 2008 Elsevier B.V. All rights reserved.
‘‘My first scientific publication in 1972 was the result of joined work with Dante Gatteschi and Ivano Bertini, and this crucial collaboration resulted in a significant number of studies at the beginning of my career, mainly on single crystal electronic and esr spectra of coordination compounds. Then, our interests set apart, Dante became more and more involved into magnetic characterization of paramagnetic metalloclustered coordination compounds and myself oriented towards the study of metalloenzymes. It is now a pleasure for my co-workers and myself to present this contribution dedicated to Dante, where, in some way, the metalloenzymes investigated encounter some of the issues pioneered in his studies. Bioinorganic chemists evidenced that the clustering of paramagnetic metal ions in metalloproteins, beside affecting their magnetic properties, has a real functional role, exploiting a different chemistry compared to that of the single metal ions. So we can pragmatically say that the study of the magnetic properties of a metal cluster (a topic to which Dante brought many new theoretical contributions) give us essential information on the relative distances between the paramagnetic atoms and the strength of their spin-spin interactions but the original chemical reactivity exhibited by the clustered moieties has to be de novo characterized case by case”. Andrea Scozzafava * Corresponding author. Tel.: +39 055 4573273; fax: +39 055 4573385. E-mail address: andrea.scozzafava@unifi.it (A. Scozzafava). 0020-1693/$ - see front matter Ó 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.ica.2008.03.091
1. Introduction Laccases (EC 1.10.3.2, p-diphenol:dioxygen oxidoreductase) are blue-copper proteins distributed in plants, fungi and bacteria, that belong to the so-called multicopper protein family including ascorbate oxidase, ceruloplasmin and bilirubin oxidase that use molecular oxygen to oxidize various aromatic and non-aromatic compounds by a radical-catalyzed reaction mechanism. Laccases catalyze the reduction of dioxygen to water accompanied by the one electron oxidation of a substrate, typically a p-hydroxy phenol or other phenolic compounds, their reducing substrates depending mainly from the redox potential of the laccase which varies largely across the group. O2 þ 4e þ 4Hþ ! 4H2 O For the catalytic activity four copper atoms per protein unit are needed. The four copper ions are traditionally distinguished as follows [1] on the basis of their spectroscopic and/or magnetic properties: Type 1: paramagnetic ‘blue’ copper, absorbance at 610 nm (oxidized). Type 2: paramagnetic ‘‘normal” copper. Type 3: diamagnetic spin-coupled copper–copper pair, absorbance at 330 nm (oxidized). Several crystal structures of laccases of fungal origin from Lentinus tigrinus, Coprinus cinereus, Trametes versicolor, Melanocarpus
4130
I. Matera et al. / Inorganica Chimica Acta 361 (2008) 4129–4137
albomyces, Rigidoporus lignosus [2–7], and from bacteria: the spore coat protein cotA from Bacillus subtilis have been reported [8]. From these structures it results that Type 1 copper has a trigonal coordination, with two histines and a cysteine as conserved equatorial ligands. A further axial ligand at a longer distance is also observed in several multicopper oxidases. It has been widely argued that this axial position ligand strongly influences the oxidation potential of the enzyme, possibly providing the mechanism for regulating its activity. A mutation from phenylalanine to methionine significantly lowered the oxidation potential of a fungal laccase from Trametes villosa [9]. Type 1 copper confers the typical blue colour to multicopper proteins, which results from the intense electronic absorption caused by the covalent copper–cysteine bond. Due to its high redox potential of ca. 790 mV, Type 1 copper is the site where substrate oxidation takes place. Type 2 or normal Cu(II) site is characterized by the lack of strong absorption features in the visible region and reveals usual EPR spectra. It is strategically positioned close to the Type 3 copper, a binuclear center spectroscopically characterized by an electron adsorption at 330 nm (oxidized form) and by the absence of an EPR signal as the result of the anti-ferromagnetic coupling of the copper pair. The Type 3 copper center is also the common feature of another protein superfamily including the tyrosinases and haemocyanins [10]. Type 2 and Type 3 copper form a trinuclear cluster, where reduction of molecular oxygen and release of water takes place (Scheme 1). Type 2 copper is coordinated by two histidines and Type 3 copper atoms by six histidines. The strong anti-ferromagnetical coupling between the two Type 3 copper atoms, is related to the presence of a hydroxyl bridge. The chemistry, function and biotechnological use of laccases have recently been reviewed. The basic aspects of laccase structure and function were reviewed by Thurston [11], Leonowicz et al. [12] focused on the functional properties of fungal laccases and their involvement in lignin transformation and Mayer and Staples [13] dealt with the latest results about the roles of laccases in vivo and its biotechnological applications. The physico-chemical properties of multicopper oxidases have been comprehensively reviewed by several authors [14–18]. An overview of technological applications of oxidases including laccase was published by Duran et al. [19,20]. Recently our group reported the X-ray structure of some oxygen intermediates in the active site of a laccase from L. tigrinus [4].
HO
O 4x
4x -
R
+
4e +4H
R
His T1
From a mechanistic point of view, the simplest reactions catalyzed by laccases are those in which the substrate molecules are oxidized to the corresponding radicals by direct interaction with the copper cluster. Frequently, however, the substrates are not oxidized directly by laccases being to large to penetrate into the enzyme active site. Nature overcomes this limitation with the utilization of mediators, which are suitable compounds that act as intermediate substrates for laccases, whose oxidized radical forms are able to interact with the bulky substrate or enzyme targets. Indeed, laccase is considered to play a major role in the degradation of lignin by white-rot fungi. Degradation of the high M.W. lignin polymer is expected to occur through the involvement of natural mediators, although many aspects of the process still remain unclear. Indeed, owing to their high non-specific oxidation capacity, and the use of readily available molecular oxygen as an electron acceptor, laccases are useful biocatalysts for diverse biotechnological applications. Redox biocatalysts are highly desirable because of the selectivity, controllability and economy of their reactions, in comparison with conventional chemical reactions. Moreover, biocatalyst-based processes require less energy and minimize the amount of waste produced, whilst at the same time being able to improve the quality and functional specifications of products. In this paper we report the X-ray crystal structure of the fungal laccase isolated from Trametes trogii 201, a white-rot basidiomycete involved in wood decay worldwide, good producer of laccases and other ligninolytic enzymes [21–23]. This fungus was shown to be efficient in the degradation of several organic pollutants including PCB and PAH mixtures as well as textile dyes [24–30]. The X-ray structure of this metalloprotein, beside the well known structural organization of the four copper atoms, shows the unusual presence of a small aromatic molecule bound to the T1 copper. This finding will be discussed in the light of reaching a deeper understanding of the interaction between laccases and small organic molecules to be used as mediators. 2. Materials and methods 2.1. Organism and culture conditions The white-rot fungus T. Trogii 201 (DSM 11919) was maintained on BRM agar plates at 4 °C and periodically transferred onto fresh BRM agar plates and grown at 28 °C [31]. After 4–6 days 500 ml shaken flask cultures containing 150 ml liquid BRM were prepared by inoculating with 10 plugs of fungal mycelia (5 mm) and grown in the dark at 28°C with continuous agitation (130 rpm). After 4 days the grown mycelia were transferred (10%, v/v) in baffled 2000-ml Erlenmeyer flasks containing 1000 ml of fresh BRM liquid medium and grown under the same conditions. The laccase expression was further induced by addition of 150 lM CuSO4. When the extracellular laccase activity reached a maximum about on day 7, the culture supernatant was collected by filtration through Whatman No. 1 paper and concentrated using an ultrafiltration device with a 30,000 Da cut-off membrane.
Cys His
2.2. Enzyme assay
His T3
T3 T2
O2
2 H 2O
Scheme 1. Catalytic cycle of laccases.
Laccase activity was determined spectrophotometrically based on the capacity of this enzyme to oxidize the non-phenolic compound 2,20 -azino-bis(3-ethylbenzthiazoline-6-sulfonic acid) (ABTS) [32]. Oxidation of ABTS by laccase results in the production of a green–blue coloured radical cation (ABTS+) measurable at 420 nm (e420 = 36 000 M1 cm1). The assay mixture consists of 2 mM ABTS (final concentration) and 0.1 M Na-citrate buffer, pH 3, and the fungal extracellular medium containing the laccase activity to be measured at 25 °C.
4131
I. Matera et al. / Inorganica Chimica Acta 361 (2008) 4129–4137
1 Unit of laccase activity was defined as the amount of enzyme oxidizing 1 lmol of substrate per minute. 2.3. Enzyme purification Solid (NH4)2SO4 was slowly added up to a final concentration of 80% w/v to the extracellular fungal extract. After 1 h stirring at 4 °C the protein pellet was harvested by centrifugation at 8000g. The crude protein pellet was dissolved in a minimal amount of 10 mM sodium phosphate buffer (pH 6.0) and dialyzed against 200 volumes of the same buffer for at least 24 h then concentrated by ultrafiltration (membrane cut-off 30 000). All the following chromatographic steps were performed utilizing a HPLC Waters system composed by a 600 Solvent Module and a 996 diode array UV–Visible Detector interfaced to a personal computer running a Millennium chromatographic system. The dialyzed sample was loaded onto a DEAE macro prep ion exchange column (Bio-Rad, 2.6 14 cm) previously equilibrated with imidazole 20 mM pH 6.0 buffer. After the column was washed with the same buffer for 20 min at a flow rate of 4 ml/min, the enzyme fraction was eluted with a linear concentration gradient of 120 min, 0–0.4 M Na2SO4 in the same buffer and flow rate. The fractions showing laccase activity were collected and concentrated by ultrafiltration and the buffer exchanged to imidazole 20 mM pH 6.0 plus Na2SO4 0.7 M. The sample was then applied to a Phenyl-Sepharose HP column (1.6 10 cm), previously equilibrated with imidazole 20 mM pH 6.0, Na2SO4 0.7 M, the laccase containing sample was washed for 200 at a flow rate of 2 ml/min then eluted by applying a linear gradient of 1200 , 0.35–0 M Na2SO4 in imidazole 20 mM pH 6.0 at the same flow rate. The fractions containing laccase were concentrated and washed with imidazole 20 mM pH 6.0 and further applied to a Q10 ion exchange column (Bio-Rad, 1.2 8.8 cm). The column was subsequently washed for 200 with at a flow rate of 1 ml/min, 0.2 M Na2SO4, then a linear gradient of 1200 , 0.2–0.5 M Na2SO4 was applied to elute the enzyme at the same flow rate. The fractions showing the laccase activity and a visible spectrum showing the typical band at 600 nm of blue-copper proteins were collected and concentrated by ultrafiltration and the buffer exchanged to imidazole 20 mM pH 6.0 containing 0.1 M Na2SO4 then applied to a Superdex 75 column (1.6 62.5 cm) equilibrated and run with the same buffer at 0.5 ml/min. The fractions containing laccase with a A280/A600 ratio lower than 13 were pooled, concentrated and further used for purity tests and crystallization.
tor (10 kDa molecular weight cutoff, Amicon). Crystallization experiments were performed utilizing the sitting-drop vapor diffusion methods in 96-wells plates (Crystal Clear Strips, Douglas Instruments Limited, UK). Preliminary crystallization trials were performed utilizing the Crystal Screen kit I (Hampton Research) and the JBScreens 4 and 5 (Jena Bioscience) at 296 K and the most promising conditions were optimized. One microliter of protein solution were mixed with 1 ll of reservoir solution and equilibrated against 50 ll of precipitant solution. The best results were obtained with conditions B1 and C1 of JBScreen-5 (B1: 18% PEG 8000, Hepes 0.1 M pH 7.5, Ca acetate 0.2 M; C1: 22% PEG 8000, Mes 0.1 M pH 6.5, ammonium sulfate 0.2 M) and these were used for the optimization. Concentrations of the protein and of the salts and pH were systematically varied in order to grow larger and regularly shaped crystals. Moreover the effect of additives on crystal shape and regularity was tested, using the Additive Screen kit I (Hampton Research). A complete data set was collected at the BW7B beamline at the DORIS storage ring, Hamburg, Germany, using a MAR image plate detector and a wavelength of 0.8435 Å. Data were collected at 100 K adding 25% glycerol to the mother liquor as cryoprotectant to a maximum resolution of 1.58 Å. The data were processed and integrated with Mosflm and scaled by Scala, from the CCP4 package. Data processing gave 106 982 unique reflections, an overall completeness of 99.6% and an Rsym of 0.059 [33]. 2.8. Structure determination and refinement The structure was solved with molecular replacement using the structure of a laccase from L. tigrinus (PDB code: 2qt6) as a startings model and the program MolRep from the CCP4 package [33]. The alignment of the two sequences performed with program Align (http://www.ebi.ac.uk/emboss/align/) shows 70.2% of identity and 80.0% of similarity. The model was refined using the program Refmac 5.1.24 from the CCP4 program suite. Manual rebuilding of the model was performed using the program XTALVIEW [34]. Solvent molecules were introduced automatically using ARP [35]. Refinement results in R-factor and R-free values of 17.4% and 19.2%, respectively. Data processing and refinement statistics are summarized in Table 1. The stereochemistry of the final model was analyzed with PROCHECK [36]. The Ramachandran plot is of good quality; there are 415 non-glycine and non-proline residues; among these, 365 (88.0%) are in the most favored regions, 47
2.4. Electrophoresis Polyacrylamide (12%) gel slab electrophoresis in 0.1% SDS was performed according to a modification of the Laemmli method, molecular mass markers ranging from 14 000 to 140 000 Da were used throughout (Bio-Rad Laboratories, USA). 2.5. Protein determination Protein concentration was determined using the BioRad Protein Assay (BioRad), using bovine serum albumin as standard. 2.6. Spectroscopic and analytical methods UV–Visible absorbance spectra were recorded utilizing a Perkin Elmer spectrophotometer Lambda EZ 201 interfaced to a personal computer. 2.7. Crystallization and data collection
Table 1 Summary of data collection and atomic model refinement statistics Data collectiona Wavelength (Å) Limiting resolution (Å) Unique reflections Rsym (%)b Multiplicity Completeness overall (%) hI/r(I)i
0.8435 85.13–1.58 (1.67–1.58) 106 932 5.9 (36.2) 4.1 (3.3) 99.6 (98.6) 15.1 (3.2)
Refinement Resolution range (Å) Unique reflections, working/free R-factor (%)c R-free (%) Non-hydrogen atoms Water molecules r.m.s.d. bonds (Å) r.m.s.d. angles (°)
22.33–1.58 101 597/5068 17.4 19.2 4476 548 0.008 1.382
a
The fungal laccase was washed with 20 mM imidazole pH 6.0 and concentrated to 20 mg/ml utilizing a Centricon ultraconcentra-
b c
Values in parentheses are for the highest resolution shell. P P Rsym = jI hIij/ I. P P Rfactor = jFobs Fcalcj/ Fobs.
4132
I. Matera et al. / Inorganica Chimica Acta 361 (2008) 4129–4137
(11.3%) are in the additional allowed regions and 3 (0.7%) in the generously allowed regions. The overall mean B factor of the structure after refinement is 17.5 Å2 for the polypeptide chain, 33.4 Å2 for water molecules and 19.8 Å2 for all atoms. Protein coordinates have been deposited with the Protein Data Bank (Protein Data Bank accession number 2HRG).
3. Results and discussion The T. trogii strain 201 subject of this research secretes a major laccase and no peroxidases under the growth conditions utilized in this study [37–39]. This main laccase was purified to homogeneity and crystallization trials were performed as described in Section 2.
Fig. 1. (A) Schematic representation of the overall structure of TtL; the copper ions are depicted as magenta spheres. (B) Stereoview of the schematic representation of the four copper sites in LtL.
I. Matera et al. / Inorganica Chimica Acta 361 (2008) 4129–4137
The optimized crystallization conditions: PEG 8000 21%, calcium acetate 0.2 M, PEG 400 3%, Hepes 0.1 M pH 7.5, allowed to obtain diffraction quality crystals which grow after one week at 296 K using the sitting drop vapor diffusion method. Crystals belong to the primitive orthorhombic space group P212121 with unit cell dimension a = 84.4, b = 85.1, c = 108.6 Å. Assuming one molecule per asymmetric unit the solvent content is 64% of the unit cell (Vm = 3.5 Å3/Da).
4133
3.1. The overall structure The tridimensional structure was solved at 1.58 Å resolution by using the molecular replacement technique and employing, as a starting model, the coordinates of the L. tigrinus laccase (pdb code: 2qt6) [4,40]. The final model is composed by a single chain of 496 residues, 595 water molecules, 4 Cu ions, 6 saccharide moieties (4 molecules
Fig. 2. (A) Representation of the 2Fo Fc electron density map for the p-toluate moiety bound to the T1 active site of TtL. The electron density is contoured at 2r. The molecular surface of the site is also shown. (B) T2/T3 trinuclear cluster active site of TtL. The molecular surfaces of the oxygen channel (on top) reaching the trinuclear cluster as well as that of the water channel (on the bottom) are shown.
4134
I. Matera et al. / Inorganica Chimica Acta 361 (2008) 4129–4137
of N-acetyl-glucosamide and 2 of mannose), 3 Ca(II) ions, 3 glycerol molecules and a molecule of p-toluate. The schematic structure and folding topology of the monomeric globular glycoprotein TtL, with overall dimensions 70 60 50 Å, is shown in Fig. 1A. The overall structure is organized in three domains; each of them has a greek key ß-barrel topology, strictly related to the small copper protein like azurin and plastocyanin and common to all the members of the blue multicopper oxidase family, like the ascorbate oxidase and the mammalian ceruloplasmin: domain I is composed by the first 139 residues (1–139), domain II starts at residue 140 and ends at residue 319 and domain III is composed by the last 176 residues (320–496). The structure is stabilized by two disulfide bridges: Cys85–Cys485, between domains I and III, and Cys117–Cys204, between domains I and II. TtL is a monomer. In particular it shows a relevant structural similarity with the laccase from T. versicolor (TvL) (75.8% of sequence identity and 83.8% of sequence similarity). Superimposition of the two structures gives an r.m.s.d. of 3.97 Å. The density maps of TtL indicate the presence of glycosylation at Asn54 and Asn433 residues. The corresponding electronic densities
were modelled to one di(n-acetyl-D-glucosamine), one b-D-mannose, and one a-D-mannose moieties bound to Asn54, and one di(n-acetyl-D-glucosamine) moiety bound to Asn433. Three Ca ions were also found on the enzyme surface: one bound to Thr188 and 3 water molecules, a second bound to Ser224, Ala103 and 4 water molecules, the third bound to Asp183 and 2 water molecules. 3.2. The p-toluate ligand In the current structure a further electron density was found on the protein surface near to the T1 copper site; a molecule of p-toluate (p-methylbenzoate) or p-hydroxybenzoate could be modelled into such density (see Figs. 1B and 2A). Neither one of these compounds were added during the crystallization process or during the enzyme purification procedures, therefore we suppose that the bound molecule is a fungal metabolite. This moiety is located in the putative cavity for the reducing substrate: the carboxylate group is hydrogen bonded to His455 (at 2.7 Å) whereas the paragroup is hosted in a hydrophobic region composed by Phe331, Phe336, Pro390 and Val162 (see Fig. 2A). As p-hydroxybenzoate
Fig. 3. Schematic representations of the: T1 (A) and T2–T3 (B) coppers sites coordinations of TtL. The copper–ligand distances are reported in yellow.
I. Matera et al. / Inorganica Chimica Acta 361 (2008) 4129–4137
4135
is known to be a substrate for laccases its binding to the T1 copper center is likely to occur through an hydroxy–His455 hydrogen bond interaction. For this reason the electronic density hosted in the hydrophobic pocket was modeled as a methyl group (p-toluate) rather than as an hydroxyl group (p-hydroxybenzoate). A weak substrate of laccases, 2,5-xylidine (2,5-dimethylaniline), was observed to bind to the enzyme from T. versicolor in a similar mode [41]. The ring of the His residue coordinated to the T1 copper is close to the nitrogen of 2,5-xylidine at a distance consistent with a hydrogen bond (3.2 Å). The amino group of 2,5-xylidine was also hydrogen bonded to a terminal oxygen of a vicinal Asp side chain. We tested the properties of p-toluate in order to establish if this compound could have some effect on the enzyme activity and resulted that such molecule does not acts as a substrate and it is only a very weak inhibitor towards ABTS (2,20 -azino-bis(3-ethylbenzothiazoline-6-sulfonic acid)) oxidation enzymatic activity. Anyway the observed adduct may well suggest the mode of binding for a variety of compounds interacting with laccases since to our knowledge the T1 site does not exhibit strong interactions neither with substrates nor with inhibitors.
but also to the nature of the second sphere residues present in the structure which influences solvent accessibility, hydrogen bonding, and dielectric anisotropy around the site. In particular in the present structure the occurrence of two hydrophobic residues Phe460 and Ile452 in the near surroundings of the T1 copper ion (3.79 and 3.58 Å from the copper ion, respectively) would indeed contribute to the high redox potential observed (E0 = 760 mV). Furthermore, residue Phe460 is additionally surrounded by a large number of hydrophobic residues: Phe336, Phe329, Phe396, Ala389 and Ile338 that would also contribute to rise the redox potential of the copper center. In CcL (E0 = 550 mV) Phe460, Phe329, and Ala389 are substituted by Leu462, Leu333, and Val391 whereas in the low redox potential bacterial Bacillus subtilis laccase (E0 = 455 mV) Phe460, Phe336, Phe329, Ile338, Ala389 and Phe396 are substituted by Met502 (fourth ligand to the T1 copper), Leu386, Leu374, Leu388, Thr415 and Ile421, respectively. These observations support the fact that several mutations are needed to change the redox potential of the metal site in a 300 mV range.
3.3. The T1 active site
The electrons taken from the substrates are transferred, through two intramolecular electron transfer pathways [48], shown in Fig. 1B in magenta and formed by Cys450, His 449 and His451, from the T1 copper ion to the second active site (Fig. 2B), where the oxygen molecules bind and are reduced to water molecules. It is composed by a trinuclear copper cluster, with a T2 Cu ion and two T3 Cu ions arranged in a triangular way, embedded between domains I and III and bound to eight histidine residues and two water molecules. T2 Cu is tricoordinated and bound to His64, His397 and one water molecule (see Fig. 3B). The two T3 coppers are tetracoordinated: T3(a) Cu binds to residues His111, His399 and His449, T3(b) Cu binds to residues His66, His109 and His451; moreover there is a water molecule asymmetrically bound
Laccases have two distinct metal active sites: the T1 single copper site and The T2/T3 trinuclear copper site. The first active site is near to the external surface where the ptoluate molecule is found and where organic substrates (mainly phenols or arylamines) are oxidized releasing electrons to the T1 copper ion through a bound His residue (His455) (see Figs. 1B and 2A). The coordination of the T1 copper ion, which is not solvent exposed, is completed by His394 and Cys450 (Fig. 3A), all belonging to domain III. In laccases the T1 copper exhibits a planar triangular coordination with the above mentioned two histidines and one cysteine residue; contrarily to many other multicopper oxidases which show a tetrahedrical coordination of the T1 copper ion with an axial fourth ligand, usually a methionine, to complete the first coordination sphere [42]. A role of redox potential modulation was ascribed to this ligand nature in azurin and recently in a bacterial laccase: when the axial methionine ligand at the T1 copper was substituted by a leucine the redox potential increased of about 100 mV [43]. Additional experiments of site specific mutagenesis, showed that Leu-Phe substitution at such site resulted in no significant effect [44]. Further structural features, which could be utilized by blue copper oxidases to tune the potential over a larger range, have not been identified to date with the exception of the hypothesis from Piontek et al. that supposed to have identified a correlation between the redox potential E0 and the coordination distances around the copper T1 ion, at longer distances corresponding higher redox potentials [7]. They suggested a mechanism by which laccases can raise their redox potential by more than 200 mV comparing the coordination distances of the T1 copper in T. versicolor laccase (E0 = 800 mV) (TvL hereafter) and C. cinereus laccase (E0 = 550 mV) (CcL hereafter), the second having up to 0.17 Å shorter bonds. In their opinion a distortion of the coordination environment would be the cause of different redox potentials. Unfortunately EPR spectra, which are very sensitive to the coordination structure of the copper(II) centers, registered for laccases differing in potential more than 300 mV resulted to be very similar thus excluding substantial structural differences in the close surroundings [45–47]. For these reasons the variations in redox potential of the T1 site observed among laccases as well as in other metalloenzymes cannot be ascribed to a single structural feature but to a sum of factors such as the nature of the direct ligands and to their coordination geometries which seem to be somewhat invariant in many laccases
3.4. The T2/T3 trinuclear active site
Table 2 Coordination distances (Å) in the copper centers and copper–copper distances for Trametes trogii laccase Mononuclear copper center: T1 copper His394 2.06 Cys450 2.15 His455 2.03 Phe460 CD2 3.79 Ile452 CD1 3.58 Trinuclear copper center T2/T3 T3(a) copper His111 2.07 His449 2.01 His399 1.97 OH 2.23 O2 2.91 T3(b) copper His451 1.94 His66 2.03 His109 2.19 OH 2.98 T2 copper His397 1.87 His64 1.80 HOH 2.57 Copper–Copper T1–T3(a) 12.12 T1–T3(b) 13.10 T1–T2 14.80 T3(a)–T3(b) 5.19 T2–T3(a) 4.55 T2–T3(b) 4.09
4136
I. Matera et al. / Inorganica Chimica Acta 361 (2008) 4129–4137
between the two T3 copper ions (with a distance of 2.98 Å to T3(a) and 2.23 Å to T3(b)). The surface of the two solvent channels which provide access to the trinuclear copper cluster, located in the interior of the protein structure, are also shown in Fig. 2B. The first channel points towards the two T3 copper ions on one side of the T2/T3 cluster allowing to the molecular oxygen to enter and to bind to it. The second channel pointing towards the T2 copper ion on the other side of the cluster permits to the water molecules produced in the O2 reduction to move to the bulk solvent. 3.5. Redox states in copper sites The bond distances between copper ions and the corresponding ligands are reported in Table 2. It should be noted that differences observed among the structural data of multicopper oxidases published to date, and acquired using synchrotron radiation, are hard to interpret. In fact, it has been observed that metalloproteins are subjected to progressive reduction by exposure of crystals to high-intensity X-ray synchrotron beams. Synchrotron radiation can yield the partial or total reduction of the metal ions depending on the particular metalloprotein studied, on the data collection method and accordingly on the acquired radiation dose [49,50]. This strategy, newly applied by our group to multicopper oxidases, allowed to observe the progressive reduction of the copper ions and under particular circumstances of the molecular oxygen and monitor the nature of the intermediate states [4]. In the present structure the distances reported in Table 2 would suggest that the electron density picture observed could be the outcome of an average between the two oxidation states of the copper ions. The T3(a) and T3(b) copper ions are asymmetrically coordinated to an hydroxide/water moiety. Also in the R. lignosus laccase (RlL) structure [5] it has been observed the asymmetric binding of the hydroxide ion bridging the T3 copper ions (The T3(a)–OH bond being shorter than the T3(b)–OH bond); this was explained by assuming a reduced oxidation state for T2 and T3(b) copper ions. This asymmetry could have a crucial role in the progression of the dioxygen splitting. It has to be noted that in the present structure the T3(a) copper results to be about 0.75 Å closer to the T1 site than the T3(b) copper. Such asymmetry is generally observed in all multicopper oxidases. According to the theoretical calculations performed by Kyritsis et al. [48] for ascorbate oxidase, the particular conformation of the interposed aminoacidic residues, also observed in TtL, should render the T1–T3(a) pathway up to three times more efficient than the T1–T3(b), thus a differential reduction progress in the T2/T3 cluster would result in the observed asymmetries. As mentioned above for the T1 site, subtle differences observed among the structural data of multicopper oxidases published to date, and acquired using synchrotron radiation, are problematical to interpret since they represent averages of the starting oxidized state and of the partially/totally reduced coppers states, the weight of which rely on the X-ray doses acquired and on the other experimental conditions. Also in the laccase structure from M. albomyces a change of the redox states and asymmetrical assembly of the trinuclear structure upon data collected with diverse X-ray radiation doses has been observed [51]. At low X-ray doses a dioxygen/peroxide moiety was identified but at high doses an enzyme in the resting state was observed. A recent study from our laboratory further allowed to discern two central intermediates in the process of molecular oxygen reduction and splitting: the peroxide and the native intermediates which result from the two- and four-electrons molecular oxygen reduction, respectively. This was obtained by starting from aerobic crystals in which the complete dioxygen turnover was inhibited by the high pH due to the formation of a strong T2 Cu–OH complex
which hinders the subsequent transformation of reduced oxygen to water molecules (O2 + 2H+ ? H2O) [4]. The structural characterization of these new intermediates in the reduction of O2 to H2O in the multicopper oxidases allowed to propose a more detailed unified molecular mechanism [4]. Comparing the geometry of the metal sites of the different available laccase structures, we can notice substantial differences in Cu–ligand distances as well as Cu–Cu distances (see Supplementary material). As an example the distances between Cu ions in the T2/T3 cluster of the present enzyme are significantly higher than those expected for a completely oxidized cluster (3.6–3.8 Å). In TtL the T3(a) and T3(b) copper ions distance reaches 5.19 Å which is comparable to that observed in totally reduced laccases or other multicopper oxidases [52,53]. This lengthening of the copper–copper distances results in slight reductions of the bond distances between copper ions and their coordinated residues. We should mention that the presence of the OH/H2O moiety asymmetrically bound between the T3 copper ions has never been observed in completely reduced anaerobic clusters but usually in the resting oxidized state [7,41]. For all these reasons the picture coming out from the current data could suggest that, since during data collection the X-ray synchrotron radiation reduces the copper centers, the aerobic enzyme undergoes redox cycling. The crystals are exposed to air, then the molecular oxygen is taken and reduced into water molecules and most likely the catalytic cycle is repeated additional times resulting in collection of data generating electron density maps which represent an average of the several catalytic steps, the most significant being the completely reduced and the resting state with the OH/H2O moiety asymmetrically bound between the T3 copper ions. Acknowledgements We thank the Assessorato all’Istruzione, Formazione e Lavoro, Regione Toscana and we gratefully acknowledge the support of Progetto MECHOS, POR Ob. 3 2000/2006 Toscana, Progetti integrati di ricerca Mis. D4 (Decreto Regionale 03/04/2007 no. 1785). Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.ica.2008.03.091. References [1] B.G. Malmstrom, J. Leckner, Curr. Opin. Chem. Biol. 2 (1998) 286. [2] V. Ducros, A.M. Brzozowski, K.S. Wilson, S.H. Brown, P. Ostergaard, P. Schneider, D.S. Yaver, A.H. Pedersen, G.J. Davies, Nat. Struct. Biol. 5 (1998) 310. [3] V. Ducros, A.M. Brzozowski, K.S. Wilson, P. Ostergaard, P. Schneider, A. Svendson, G.J. Davies, Acta Crystallogr., Sect. D 57 (2001) 333. [4] M. Ferraroni, N.M. Myasoedova, V. Schmatchenko, A.A. Leontievsky, L.A. Golovleva, A. Scozzafava, F. Briganti, BMC. Struct. Biol. 7 (2007) 60. [5] S. Garavaglia, M.T. Cambria, M. Miglio, S. Ragusa, V. Iacobazzi, F. Palmieri, C. D’Ambrosio, A. Scaloni, M. Rizzi, J. Mol. Biol. 342 (2004) 1519. [6] N. Hakulinen, L.L. Kiiskinen, K. Kruus, M. Saloheimo, A. Paananen, A. Koivula, J. Rouvinen, Nat. Struct. Biol. 9 (2002) 601. [7] K. Piontek, M. Antorini, T. Choinowski, J. Biol. Chem. 277 (2002) 37663. [8] F.J. Enguita, L.O. Martins, A.O. Henriques, M.A. Carrondo, J. Biol. Chem. 278 (2003) 19416. [9] S.V.S. Kumar, P.S. Phale, S. Durani, P.P. Wangikar, Biotechnol. Bioeng. 83 (2003) 386. [10] H. Decker, T. Schweikardt, D. Nillius, U. Salzbrunn, E. Jaenicke, F. Tuczek, Gene 398 (2007) 183. [11] C.F. Thurston, Microbiol.-UK 140 (1994) 19. [12] A. Leonowicz, N.S. Cho, J. Luterek, A. Wilkolazka, M. Wojtas-Wasilewska, A. Matuszewska, M. Hofrichter, D. Wesenberg, J. Rogalski, J. Basic Microbiol. 41 (2001) 185. [13] A.M. Mayer, R.C. Staples, Phytochemistry 60 (2002) 551. [14] E.I. Solomon, R. Sarangi, J.S. Woertink, A.J. Augustine, J. Yoon, S. Ghosh, Acc. Chem. Res. 40 (2007) 581. [15] E.I. Solomon, Inorg. Chem. 45 (2006) 8012.
I. Matera et al. / Inorganica Chimica Acta 361 (2008) 4129–4137 [16] A. Decker, E.I. Solomon, Curr. Opin. Chem. Biol. 9 (2005) 152. [17] E.I. Solomon, R.K. Szilagyi, G.S. Debeer, L. Basumallick, Chem. Rev. 104 (2004) 419. [18] E.I. Solomon, P. Chen, M. Metz, S.K. Lee, A.E. Palmer, Angew. Chem., Int. Ed. Engl. 40 (2001) 4570. [19] N. Duran, M.A. Rosa, A. D’Annibale, L. Gianfreda, Enzyme Microb. Technol. 31 (2002) 907. [20] N. Duran, E. Esposito, Appl. Catal. B-Environ. 28 (2000) 83. [21] T. Deveci, A. Unyayar, M.A. Mazmanci, J. Mol. Catal. B-Enzym. 30 (2004) 25. [22] S.S. Kahraman, I.H. Gurdal, Bioresour. Technol. 82 (2002) 215. [23] L. Levin, F. Forchiassin, A. Viale, Process Biochem. 40 (2005) 1381. [24] L. Levin, A. Viale, A. Forchiassin, Int. Biodeterior. Biodegrad. 52 (2003) 1. [25] C. Haglund, L. Levin, F. Forchiassin, M. Lopez, A. Viale, Rev. Argent Microbiol. 34 (2002) 157. [26] O. Yesilada, S. Cing, D. Asma, Bioresour. Technol. 81 (2002) 155. [27] M.C. Colao, S. Lupino, A.M. Garzillo, V. Buonocore, M. Ruzzi, Microb. Cell Fact. 5 (2006) 31. [28] H. Zouari-Mechichi, T. Mechichi, A. Dhouib, S. Sayadi, A.T. Martinez, M.J. Martinez, Enzyme Microb. Technol. 39 (2006) 141. [29] L. Levin, A. Jordan, F. Forchiassin, A. Viale, Rev. Argent Microbiol. 33 (2001) 223. [30] I. Ciullini, S. Tilli, A. Scozzafava, F. Briganti, Bioresour. Technol. 2008 Feb 15 [Epub ahead of print]. [31] L. Bezalel, Y. Hadar, C.E. Cerniglia, Appl. Environ. Microbiol. 63 (1997) 2495. [32] B.S. Wolfenden, R.L. Wilson, J. Chem. Soc., Perkin Trans. 2 (1982) 805. [33] P.N. Collaborative Computational, Acta Crystallogr., Sect. D 50 (1994) 760. [34] D.E. McRee, J. Struct. Biol. 125 (1999) 156. [35] V.S. Lamzin, Acta Crystallogr., Sect. D 49 (1993) 129. [36] R.A. Laskowski, M.W. MacArthur, D.S. Moss, J.M. Thornton, J. Appl. Crystallogr. 26 (1993) 283. [37] A.M. Garzillo, M.C. Colao, C. Caruso, C. Caporale, D. Celletti, V. Buonocore, Appl. Microbiol. Biotechnol. 49 (1998) 545.
4137
[38] A.M. Garzillo, M.C. Colao, V. Buonocore, R. Oliva, L. Falcigno, M. Saviano, A.M. Santoro, R. Zappala, R.P. Bonomo, C. Bianco, P. Giardina, G. Palmieri, G. Sannia, J. Protein Chem. 20 (2001) 191. [39] M.C. Colao, A.M. Garzillo, V. Buonocore, A. Schiesser, M. Ruzzi, Appl. Microbiol. Biotechnol. 63 (2003) 153. [40] M. Ferraroni, I. Duchi, N.M. Myasoedova, A.A. Leontievsky, L.A. Golovleva, A. Scozzafava, F. Briganti, Acta Crystallogr., Sect. F 61 (2005) 205. [41] T. Bertrand, C. Jolivalt, P. Briozzo, E. Caminade, N. Joly, C. Madzak, C. Mougin, Biochemistry 41 (2002) 7325. [42] A. Messerschmidt, Multi-Copper Oxidases, World Scientific Publishing, Singapore, 1997. [43] P. Durao, I. Bento, A.T. Fernandes, E.P. Melo, P.F. Lindley, L.O. Martins, J. Biol. Inorg. Chem. 11 (2006) 514. [44] F. Xu, R.M. Berka, J.A. Wahleithner, B.A. Nelson, J.R. Shuster, S.H. Brown, A.E. Palmer, E.I. Solomon, Biochem. J. 334 (Pt 1) (1998) 63. [45] F. Xu, W. Shin, S.H. Brown, J.A. Wahleithner, U.M. Sundaram, E.I. Solomon, Biochim. Biophys. Acta 1292 (1996) 303. [46] A.E. Palmer, D.W. Randall, F. Xu, E.I. Solomon, J. Am. Chem. Soc. 121 (1999) 7138. [47] F. Xu, A.E. Palmer, D.S. Yaver, R.M. Berka, G.A. Gambetta, S.H. Brown, E.I. Solomon, J. Biol. Chem. 274 (1999) 12372. [48] P. Kyritsis, A. Messerschmidt, R. Huber, G.A. Salmon, A.G. Sykes, J. Chem. Soc., Dalton Trans. (1993) 731. [49] O. Carugo, K.D. Carugo, Trends Biochem. Sci. 30 (2005) 213. [50] R.B. Ravelli, E.F. Garman, Curr. Opin. Struct. Biol. 16 (2006) 624. [51] N. Hakulinen, K. Kruus, A. Koivula, J. Rouvinen, Biochem. Biophys. Res. Commun. 350 (2006) 929. [52] A.B. Taylor, C.S. Stoj, L. Ziegler, D.J. Kosman, P.J. Hart, Proc. Natl. Acad. Sci. USA 102 (2005) 15459. [53] I. Bento, L.O. Martins, L.G. Gato, C.M. Armenia, P.F. Lindley, Dalton Trans. (2005) 3507.
Contents lists available at ScienceDirect
Inorganica Chimica Acta journal homepage: www.elsevier.com/locate/ica
Crystal structure of the blue multicopper oxidase from the white-rot fungus Trametes trogii complexed with p-toluate Irene Matera, Antonella Gullotto, Silvia Tilli, Marta Ferraroni, Andrea Scozzafava *, Fabrizio Briganti Bioinorganic Chemistry Laboratory, Department of Chemistry, University of Florence, Via della Lastruccia 3, 50019 Sesto Fiorentino, Florence, Italy
a r t i c l e
i n f o
Article history: Received 24 January 2008 Received in revised form 17 March 2008 Accepted 17 March 2008 Available online 23 March 2008 Dedicated to Professor Dante Gatteschi. Keywords: Trametes trogii Funalia trogii laccase X-ray structure
a b s t r a c t A multicopper oxidase, the fungal laccase glycoenzyme from the white-rot basidiomycete fungus Trametes (Funalia) trogii, was crystallized and its crystal structure was solved at 1.58 Å using molecular replacement techniques. Model refinement resulted in R-factor and R-free values of 17.4% and 19.0%, respectively. The T. trogii laccase structural model reveals the presence of a ligand bound to the T1 active site which resembles a ptoluate molecule, such bound compound is most probably a fungal metabolite. The p-toluate is bound into the T1 active site of the laccase forming, with one of the carboxylate oxygens, a H-bond with His455, one of the T1 copper ion ligands, whereas the methyl group presents hydrophobic interactions within a pocket composed by Phe331, Phe336, Pro390 and Val162. The coordination geometries, the bond distances and the oxidation states of the T1 and T2/T3 copper active sites are analyzed and discussed in terms of the enzymatic mechanism and catalytic functionality. Ó 2008 Elsevier B.V. All rights reserved.
‘‘My first scientific publication in 1972 was the result of joined work with Dante Gatteschi and Ivano Bertini, and this crucial collaboration resulted in a significant number of studies at the beginning of my career, mainly on single crystal electronic and esr spectra of coordination compounds. Then, our interests set apart, Dante became more and more involved into magnetic characterization of paramagnetic metalloclustered coordination compounds and myself oriented towards the study of metalloenzymes. It is now a pleasure for my co-workers and myself to present this contribution dedicated to Dante, where, in some way, the metalloenzymes investigated encounter some of the issues pioneered in his studies. Bioinorganic chemists evidenced that the clustering of paramagnetic metal ions in metalloproteins, beside affecting their magnetic properties, has a real functional role, exploiting a different chemistry compared to that of the single metal ions. So we can pragmatically say that the study of the magnetic properties of a metal cluster (a topic to which Dante brought many new theoretical contributions) give us essential information on the relative distances between the paramagnetic atoms and the strength of their spin-spin interactions but the original chemical reactivity exhibited by the clustered moieties has to be de novo characterized case by case”. Andrea Scozzafava * Corresponding author. Tel.: +39 055 4573273; fax: +39 055 4573385. E-mail address: andrea.scozzafava@unifi.it (A. Scozzafava). 0020-1693/$ - see front matter Ó 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.ica.2008.03.091
1. Introduction Laccases (EC 1.10.3.2, p-diphenol:dioxygen oxidoreductase) are blue-copper proteins distributed in plants, fungi and bacteria, that belong to the so-called multicopper protein family including ascorbate oxidase, ceruloplasmin and bilirubin oxidase that use molecular oxygen to oxidize various aromatic and non-aromatic compounds by a radical-catalyzed reaction mechanism. Laccases catalyze the reduction of dioxygen to water accompanied by the one electron oxidation of a substrate, typically a p-hydroxy phenol or other phenolic compounds, their reducing substrates depending mainly from the redox potential of the laccase which varies largely across the group. O2 þ 4e þ 4Hþ ! 4H2 O For the catalytic activity four copper atoms per protein unit are needed. The four copper ions are traditionally distinguished as follows [1] on the basis of their spectroscopic and/or magnetic properties: Type 1: paramagnetic ‘blue’ copper, absorbance at 610 nm (oxidized). Type 2: paramagnetic ‘‘normal” copper. Type 3: diamagnetic spin-coupled copper–copper pair, absorbance at 330 nm (oxidized). Several crystal structures of laccases of fungal origin from Lentinus tigrinus, Coprinus cinereus, Trametes versicolor, Melanocarpus
4130
I. Matera et al. / Inorganica Chimica Acta 361 (2008) 4129–4137
albomyces, Rigidoporus lignosus [2–7], and from bacteria: the spore coat protein cotA from Bacillus subtilis have been reported [8]. From these structures it results that Type 1 copper has a trigonal coordination, with two histines and a cysteine as conserved equatorial ligands. A further axial ligand at a longer distance is also observed in several multicopper oxidases. It has been widely argued that this axial position ligand strongly influences the oxidation potential of the enzyme, possibly providing the mechanism for regulating its activity. A mutation from phenylalanine to methionine significantly lowered the oxidation potential of a fungal laccase from Trametes villosa [9]. Type 1 copper confers the typical blue colour to multicopper proteins, which results from the intense electronic absorption caused by the covalent copper–cysteine bond. Due to its high redox potential of ca. 790 mV, Type 1 copper is the site where substrate oxidation takes place. Type 2 or normal Cu(II) site is characterized by the lack of strong absorption features in the visible region and reveals usual EPR spectra. It is strategically positioned close to the Type 3 copper, a binuclear center spectroscopically characterized by an electron adsorption at 330 nm (oxidized form) and by the absence of an EPR signal as the result of the anti-ferromagnetic coupling of the copper pair. The Type 3 copper center is also the common feature of another protein superfamily including the tyrosinases and haemocyanins [10]. Type 2 and Type 3 copper form a trinuclear cluster, where reduction of molecular oxygen and release of water takes place (Scheme 1). Type 2 copper is coordinated by two histidines and Type 3 copper atoms by six histidines. The strong anti-ferromagnetical coupling between the two Type 3 copper atoms, is related to the presence of a hydroxyl bridge. The chemistry, function and biotechnological use of laccases have recently been reviewed. The basic aspects of laccase structure and function were reviewed by Thurston [11], Leonowicz et al. [12] focused on the functional properties of fungal laccases and their involvement in lignin transformation and Mayer and Staples [13] dealt with the latest results about the roles of laccases in vivo and its biotechnological applications. The physico-chemical properties of multicopper oxidases have been comprehensively reviewed by several authors [14–18]. An overview of technological applications of oxidases including laccase was published by Duran et al. [19,20]. Recently our group reported the X-ray structure of some oxygen intermediates in the active site of a laccase from L. tigrinus [4].
HO
O 4x
4x -
R
+
4e +4H
R
His T1
From a mechanistic point of view, the simplest reactions catalyzed by laccases are those in which the substrate molecules are oxidized to the corresponding radicals by direct interaction with the copper cluster. Frequently, however, the substrates are not oxidized directly by laccases being to large to penetrate into the enzyme active site. Nature overcomes this limitation with the utilization of mediators, which are suitable compounds that act as intermediate substrates for laccases, whose oxidized radical forms are able to interact with the bulky substrate or enzyme targets. Indeed, laccase is considered to play a major role in the degradation of lignin by white-rot fungi. Degradation of the high M.W. lignin polymer is expected to occur through the involvement of natural mediators, although many aspects of the process still remain unclear. Indeed, owing to their high non-specific oxidation capacity, and the use of readily available molecular oxygen as an electron acceptor, laccases are useful biocatalysts for diverse biotechnological applications. Redox biocatalysts are highly desirable because of the selectivity, controllability and economy of their reactions, in comparison with conventional chemical reactions. Moreover, biocatalyst-based processes require less energy and minimize the amount of waste produced, whilst at the same time being able to improve the quality and functional specifications of products. In this paper we report the X-ray crystal structure of the fungal laccase isolated from Trametes trogii 201, a white-rot basidiomycete involved in wood decay worldwide, good producer of laccases and other ligninolytic enzymes [21–23]. This fungus was shown to be efficient in the degradation of several organic pollutants including PCB and PAH mixtures as well as textile dyes [24–30]. The X-ray structure of this metalloprotein, beside the well known structural organization of the four copper atoms, shows the unusual presence of a small aromatic molecule bound to the T1 copper. This finding will be discussed in the light of reaching a deeper understanding of the interaction between laccases and small organic molecules to be used as mediators. 2. Materials and methods 2.1. Organism and culture conditions The white-rot fungus T. Trogii 201 (DSM 11919) was maintained on BRM agar plates at 4 °C and periodically transferred onto fresh BRM agar plates and grown at 28 °C [31]. After 4–6 days 500 ml shaken flask cultures containing 150 ml liquid BRM were prepared by inoculating with 10 plugs of fungal mycelia (5 mm) and grown in the dark at 28°C with continuous agitation (130 rpm). After 4 days the grown mycelia were transferred (10%, v/v) in baffled 2000-ml Erlenmeyer flasks containing 1000 ml of fresh BRM liquid medium and grown under the same conditions. The laccase expression was further induced by addition of 150 lM CuSO4. When the extracellular laccase activity reached a maximum about on day 7, the culture supernatant was collected by filtration through Whatman No. 1 paper and concentrated using an ultrafiltration device with a 30,000 Da cut-off membrane.
Cys His
2.2. Enzyme assay
His T3
T3 T2
O2
2 H 2O
Scheme 1. Catalytic cycle of laccases.
Laccase activity was determined spectrophotometrically based on the capacity of this enzyme to oxidize the non-phenolic compound 2,20 -azino-bis(3-ethylbenzthiazoline-6-sulfonic acid) (ABTS) [32]. Oxidation of ABTS by laccase results in the production of a green–blue coloured radical cation (ABTS+) measurable at 420 nm (e420 = 36 000 M1 cm1). The assay mixture consists of 2 mM ABTS (final concentration) and 0.1 M Na-citrate buffer, pH 3, and the fungal extracellular medium containing the laccase activity to be measured at 25 °C.
4131
I. Matera et al. / Inorganica Chimica Acta 361 (2008) 4129–4137
1 Unit of laccase activity was defined as the amount of enzyme oxidizing 1 lmol of substrate per minute. 2.3. Enzyme purification Solid (NH4)2SO4 was slowly added up to a final concentration of 80% w/v to the extracellular fungal extract. After 1 h stirring at 4 °C the protein pellet was harvested by centrifugation at 8000g. The crude protein pellet was dissolved in a minimal amount of 10 mM sodium phosphate buffer (pH 6.0) and dialyzed against 200 volumes of the same buffer for at least 24 h then concentrated by ultrafiltration (membrane cut-off 30 000). All the following chromatographic steps were performed utilizing a HPLC Waters system composed by a 600 Solvent Module and a 996 diode array UV–Visible Detector interfaced to a personal computer running a Millennium chromatographic system. The dialyzed sample was loaded onto a DEAE macro prep ion exchange column (Bio-Rad, 2.6 14 cm) previously equilibrated with imidazole 20 mM pH 6.0 buffer. After the column was washed with the same buffer for 20 min at a flow rate of 4 ml/min, the enzyme fraction was eluted with a linear concentration gradient of 120 min, 0–0.4 M Na2SO4 in the same buffer and flow rate. The fractions showing laccase activity were collected and concentrated by ultrafiltration and the buffer exchanged to imidazole 20 mM pH 6.0 plus Na2SO4 0.7 M. The sample was then applied to a Phenyl-Sepharose HP column (1.6 10 cm), previously equilibrated with imidazole 20 mM pH 6.0, Na2SO4 0.7 M, the laccase containing sample was washed for 200 at a flow rate of 2 ml/min then eluted by applying a linear gradient of 1200 , 0.35–0 M Na2SO4 in imidazole 20 mM pH 6.0 at the same flow rate. The fractions containing laccase were concentrated and washed with imidazole 20 mM pH 6.0 and further applied to a Q10 ion exchange column (Bio-Rad, 1.2 8.8 cm). The column was subsequently washed for 200 with at a flow rate of 1 ml/min, 0.2 M Na2SO4, then a linear gradient of 1200 , 0.2–0.5 M Na2SO4 was applied to elute the enzyme at the same flow rate. The fractions showing the laccase activity and a visible spectrum showing the typical band at 600 nm of blue-copper proteins were collected and concentrated by ultrafiltration and the buffer exchanged to imidazole 20 mM pH 6.0 containing 0.1 M Na2SO4 then applied to a Superdex 75 column (1.6 62.5 cm) equilibrated and run with the same buffer at 0.5 ml/min. The fractions containing laccase with a A280/A600 ratio lower than 13 were pooled, concentrated and further used for purity tests and crystallization.
tor (10 kDa molecular weight cutoff, Amicon). Crystallization experiments were performed utilizing the sitting-drop vapor diffusion methods in 96-wells plates (Crystal Clear Strips, Douglas Instruments Limited, UK). Preliminary crystallization trials were performed utilizing the Crystal Screen kit I (Hampton Research) and the JBScreens 4 and 5 (Jena Bioscience) at 296 K and the most promising conditions were optimized. One microliter of protein solution were mixed with 1 ll of reservoir solution and equilibrated against 50 ll of precipitant solution. The best results were obtained with conditions B1 and C1 of JBScreen-5 (B1: 18% PEG 8000, Hepes 0.1 M pH 7.5, Ca acetate 0.2 M; C1: 22% PEG 8000, Mes 0.1 M pH 6.5, ammonium sulfate 0.2 M) and these were used for the optimization. Concentrations of the protein and of the salts and pH were systematically varied in order to grow larger and regularly shaped crystals. Moreover the effect of additives on crystal shape and regularity was tested, using the Additive Screen kit I (Hampton Research). A complete data set was collected at the BW7B beamline at the DORIS storage ring, Hamburg, Germany, using a MAR image plate detector and a wavelength of 0.8435 Å. Data were collected at 100 K adding 25% glycerol to the mother liquor as cryoprotectant to a maximum resolution of 1.58 Å. The data were processed and integrated with Mosflm and scaled by Scala, from the CCP4 package. Data processing gave 106 982 unique reflections, an overall completeness of 99.6% and an Rsym of 0.059 [33]. 2.8. Structure determination and refinement The structure was solved with molecular replacement using the structure of a laccase from L. tigrinus (PDB code: 2qt6) as a startings model and the program MolRep from the CCP4 package [33]. The alignment of the two sequences performed with program Align (http://www.ebi.ac.uk/emboss/align/) shows 70.2% of identity and 80.0% of similarity. The model was refined using the program Refmac 5.1.24 from the CCP4 program suite. Manual rebuilding of the model was performed using the program XTALVIEW [34]. Solvent molecules were introduced automatically using ARP [35]. Refinement results in R-factor and R-free values of 17.4% and 19.2%, respectively. Data processing and refinement statistics are summarized in Table 1. The stereochemistry of the final model was analyzed with PROCHECK [36]. The Ramachandran plot is of good quality; there are 415 non-glycine and non-proline residues; among these, 365 (88.0%) are in the most favored regions, 47
2.4. Electrophoresis Polyacrylamide (12%) gel slab electrophoresis in 0.1% SDS was performed according to a modification of the Laemmli method, molecular mass markers ranging from 14 000 to 140 000 Da were used throughout (Bio-Rad Laboratories, USA). 2.5. Protein determination Protein concentration was determined using the BioRad Protein Assay (BioRad), using bovine serum albumin as standard. 2.6. Spectroscopic and analytical methods UV–Visible absorbance spectra were recorded utilizing a Perkin Elmer spectrophotometer Lambda EZ 201 interfaced to a personal computer. 2.7. Crystallization and data collection
Table 1 Summary of data collection and atomic model refinement statistics Data collectiona Wavelength (Å) Limiting resolution (Å) Unique reflections Rsym (%)b Multiplicity Completeness overall (%) hI/r(I)i
0.8435 85.13–1.58 (1.67–1.58) 106 932 5.9 (36.2) 4.1 (3.3) 99.6 (98.6) 15.1 (3.2)
Refinement Resolution range (Å) Unique reflections, working/free R-factor (%)c R-free (%) Non-hydrogen atoms Water molecules r.m.s.d. bonds (Å) r.m.s.d. angles (°)
22.33–1.58 101 597/5068 17.4 19.2 4476 548 0.008 1.382
a
The fungal laccase was washed with 20 mM imidazole pH 6.0 and concentrated to 20 mg/ml utilizing a Centricon ultraconcentra-
b c
Values in parentheses are for the highest resolution shell. P P Rsym = jI hIij/ I. P P Rfactor = jFobs Fcalcj/ Fobs.
4132
I. Matera et al. / Inorganica Chimica Acta 361 (2008) 4129–4137
(11.3%) are in the additional allowed regions and 3 (0.7%) in the generously allowed regions. The overall mean B factor of the structure after refinement is 17.5 Å2 for the polypeptide chain, 33.4 Å2 for water molecules and 19.8 Å2 for all atoms. Protein coordinates have been deposited with the Protein Data Bank (Protein Data Bank accession number 2HRG).
3. Results and discussion The T. trogii strain 201 subject of this research secretes a major laccase and no peroxidases under the growth conditions utilized in this study [37–39]. This main laccase was purified to homogeneity and crystallization trials were performed as described in Section 2.
Fig. 1. (A) Schematic representation of the overall structure of TtL; the copper ions are depicted as magenta spheres. (B) Stereoview of the schematic representation of the four copper sites in LtL.
I. Matera et al. / Inorganica Chimica Acta 361 (2008) 4129–4137
The optimized crystallization conditions: PEG 8000 21%, calcium acetate 0.2 M, PEG 400 3%, Hepes 0.1 M pH 7.5, allowed to obtain diffraction quality crystals which grow after one week at 296 K using the sitting drop vapor diffusion method. Crystals belong to the primitive orthorhombic space group P212121 with unit cell dimension a = 84.4, b = 85.1, c = 108.6 Å. Assuming one molecule per asymmetric unit the solvent content is 64% of the unit cell (Vm = 3.5 Å3/Da).
4133
3.1. The overall structure The tridimensional structure was solved at 1.58 Å resolution by using the molecular replacement technique and employing, as a starting model, the coordinates of the L. tigrinus laccase (pdb code: 2qt6) [4,40]. The final model is composed by a single chain of 496 residues, 595 water molecules, 4 Cu ions, 6 saccharide moieties (4 molecules
Fig. 2. (A) Representation of the 2Fo Fc electron density map for the p-toluate moiety bound to the T1 active site of TtL. The electron density is contoured at 2r. The molecular surface of the site is also shown. (B) T2/T3 trinuclear cluster active site of TtL. The molecular surfaces of the oxygen channel (on top) reaching the trinuclear cluster as well as that of the water channel (on the bottom) are shown.
4134
I. Matera et al. / Inorganica Chimica Acta 361 (2008) 4129–4137
of N-acetyl-glucosamide and 2 of mannose), 3 Ca(II) ions, 3 glycerol molecules and a molecule of p-toluate. The schematic structure and folding topology of the monomeric globular glycoprotein TtL, with overall dimensions 70 60 50 Å, is shown in Fig. 1A. The overall structure is organized in three domains; each of them has a greek key ß-barrel topology, strictly related to the small copper protein like azurin and plastocyanin and common to all the members of the blue multicopper oxidase family, like the ascorbate oxidase and the mammalian ceruloplasmin: domain I is composed by the first 139 residues (1–139), domain II starts at residue 140 and ends at residue 319 and domain III is composed by the last 176 residues (320–496). The structure is stabilized by two disulfide bridges: Cys85–Cys485, between domains I and III, and Cys117–Cys204, between domains I and II. TtL is a monomer. In particular it shows a relevant structural similarity with the laccase from T. versicolor (TvL) (75.8% of sequence identity and 83.8% of sequence similarity). Superimposition of the two structures gives an r.m.s.d. of 3.97 Å. The density maps of TtL indicate the presence of glycosylation at Asn54 and Asn433 residues. The corresponding electronic densities
were modelled to one di(n-acetyl-D-glucosamine), one b-D-mannose, and one a-D-mannose moieties bound to Asn54, and one di(n-acetyl-D-glucosamine) moiety bound to Asn433. Three Ca ions were also found on the enzyme surface: one bound to Thr188 and 3 water molecules, a second bound to Ser224, Ala103 and 4 water molecules, the third bound to Asp183 and 2 water molecules. 3.2. The p-toluate ligand In the current structure a further electron density was found on the protein surface near to the T1 copper site; a molecule of p-toluate (p-methylbenzoate) or p-hydroxybenzoate could be modelled into such density (see Figs. 1B and 2A). Neither one of these compounds were added during the crystallization process or during the enzyme purification procedures, therefore we suppose that the bound molecule is a fungal metabolite. This moiety is located in the putative cavity for the reducing substrate: the carboxylate group is hydrogen bonded to His455 (at 2.7 Å) whereas the paragroup is hosted in a hydrophobic region composed by Phe331, Phe336, Pro390 and Val162 (see Fig. 2A). As p-hydroxybenzoate
Fig. 3. Schematic representations of the: T1 (A) and T2–T3 (B) coppers sites coordinations of TtL. The copper–ligand distances are reported in yellow.
I. Matera et al. / Inorganica Chimica Acta 361 (2008) 4129–4137
4135
is known to be a substrate for laccases its binding to the T1 copper center is likely to occur through an hydroxy–His455 hydrogen bond interaction. For this reason the electronic density hosted in the hydrophobic pocket was modeled as a methyl group (p-toluate) rather than as an hydroxyl group (p-hydroxybenzoate). A weak substrate of laccases, 2,5-xylidine (2,5-dimethylaniline), was observed to bind to the enzyme from T. versicolor in a similar mode [41]. The ring of the His residue coordinated to the T1 copper is close to the nitrogen of 2,5-xylidine at a distance consistent with a hydrogen bond (3.2 Å). The amino group of 2,5-xylidine was also hydrogen bonded to a terminal oxygen of a vicinal Asp side chain. We tested the properties of p-toluate in order to establish if this compound could have some effect on the enzyme activity and resulted that such molecule does not acts as a substrate and it is only a very weak inhibitor towards ABTS (2,20 -azino-bis(3-ethylbenzothiazoline-6-sulfonic acid)) oxidation enzymatic activity. Anyway the observed adduct may well suggest the mode of binding for a variety of compounds interacting with laccases since to our knowledge the T1 site does not exhibit strong interactions neither with substrates nor with inhibitors.
but also to the nature of the second sphere residues present in the structure which influences solvent accessibility, hydrogen bonding, and dielectric anisotropy around the site. In particular in the present structure the occurrence of two hydrophobic residues Phe460 and Ile452 in the near surroundings of the T1 copper ion (3.79 and 3.58 Å from the copper ion, respectively) would indeed contribute to the high redox potential observed (E0 = 760 mV). Furthermore, residue Phe460 is additionally surrounded by a large number of hydrophobic residues: Phe336, Phe329, Phe396, Ala389 and Ile338 that would also contribute to rise the redox potential of the copper center. In CcL (E0 = 550 mV) Phe460, Phe329, and Ala389 are substituted by Leu462, Leu333, and Val391 whereas in the low redox potential bacterial Bacillus subtilis laccase (E0 = 455 mV) Phe460, Phe336, Phe329, Ile338, Ala389 and Phe396 are substituted by Met502 (fourth ligand to the T1 copper), Leu386, Leu374, Leu388, Thr415 and Ile421, respectively. These observations support the fact that several mutations are needed to change the redox potential of the metal site in a 300 mV range.
3.3. The T1 active site
The electrons taken from the substrates are transferred, through two intramolecular electron transfer pathways [48], shown in Fig. 1B in magenta and formed by Cys450, His 449 and His451, from the T1 copper ion to the second active site (Fig. 2B), where the oxygen molecules bind and are reduced to water molecules. It is composed by a trinuclear copper cluster, with a T2 Cu ion and two T3 Cu ions arranged in a triangular way, embedded between domains I and III and bound to eight histidine residues and two water molecules. T2 Cu is tricoordinated and bound to His64, His397 and one water molecule (see Fig. 3B). The two T3 coppers are tetracoordinated: T3(a) Cu binds to residues His111, His399 and His449, T3(b) Cu binds to residues His66, His109 and His451; moreover there is a water molecule asymmetrically bound
Laccases have two distinct metal active sites: the T1 single copper site and The T2/T3 trinuclear copper site. The first active site is near to the external surface where the ptoluate molecule is found and where organic substrates (mainly phenols or arylamines) are oxidized releasing electrons to the T1 copper ion through a bound His residue (His455) (see Figs. 1B and 2A). The coordination of the T1 copper ion, which is not solvent exposed, is completed by His394 and Cys450 (Fig. 3A), all belonging to domain III. In laccases the T1 copper exhibits a planar triangular coordination with the above mentioned two histidines and one cysteine residue; contrarily to many other multicopper oxidases which show a tetrahedrical coordination of the T1 copper ion with an axial fourth ligand, usually a methionine, to complete the first coordination sphere [42]. A role of redox potential modulation was ascribed to this ligand nature in azurin and recently in a bacterial laccase: when the axial methionine ligand at the T1 copper was substituted by a leucine the redox potential increased of about 100 mV [43]. Additional experiments of site specific mutagenesis, showed that Leu-Phe substitution at such site resulted in no significant effect [44]. Further structural features, which could be utilized by blue copper oxidases to tune the potential over a larger range, have not been identified to date with the exception of the hypothesis from Piontek et al. that supposed to have identified a correlation between the redox potential E0 and the coordination distances around the copper T1 ion, at longer distances corresponding higher redox potentials [7]. They suggested a mechanism by which laccases can raise their redox potential by more than 200 mV comparing the coordination distances of the T1 copper in T. versicolor laccase (E0 = 800 mV) (TvL hereafter) and C. cinereus laccase (E0 = 550 mV) (CcL hereafter), the second having up to 0.17 Å shorter bonds. In their opinion a distortion of the coordination environment would be the cause of different redox potentials. Unfortunately EPR spectra, which are very sensitive to the coordination structure of the copper(II) centers, registered for laccases differing in potential more than 300 mV resulted to be very similar thus excluding substantial structural differences in the close surroundings [45–47]. For these reasons the variations in redox potential of the T1 site observed among laccases as well as in other metalloenzymes cannot be ascribed to a single structural feature but to a sum of factors such as the nature of the direct ligands and to their coordination geometries which seem to be somewhat invariant in many laccases
3.4. The T2/T3 trinuclear active site
Table 2 Coordination distances (Å) in the copper centers and copper–copper distances for Trametes trogii laccase Mononuclear copper center: T1 copper His394 2.06 Cys450 2.15 His455 2.03 Phe460 CD2 3.79 Ile452 CD1 3.58 Trinuclear copper center T2/T3 T3(a) copper His111 2.07 His449 2.01 His399 1.97 OH 2.23 O2 2.91 T3(b) copper His451 1.94 His66 2.03 His109 2.19 OH 2.98 T2 copper His397 1.87 His64 1.80 HOH 2.57 Copper–Copper T1–T3(a) 12.12 T1–T3(b) 13.10 T1–T2 14.80 T3(a)–T3(b) 5.19 T2–T3(a) 4.55 T2–T3(b) 4.09
4136
I. Matera et al. / Inorganica Chimica Acta 361 (2008) 4129–4137
between the two T3 copper ions (with a distance of 2.98 Å to T3(a) and 2.23 Å to T3(b)). The surface of the two solvent channels which provide access to the trinuclear copper cluster, located in the interior of the protein structure, are also shown in Fig. 2B. The first channel points towards the two T3 copper ions on one side of the T2/T3 cluster allowing to the molecular oxygen to enter and to bind to it. The second channel pointing towards the T2 copper ion on the other side of the cluster permits to the water molecules produced in the O2 reduction to move to the bulk solvent. 3.5. Redox states in copper sites The bond distances between copper ions and the corresponding ligands are reported in Table 2. It should be noted that differences observed among the structural data of multicopper oxidases published to date, and acquired using synchrotron radiation, are hard to interpret. In fact, it has been observed that metalloproteins are subjected to progressive reduction by exposure of crystals to high-intensity X-ray synchrotron beams. Synchrotron radiation can yield the partial or total reduction of the metal ions depending on the particular metalloprotein studied, on the data collection method and accordingly on the acquired radiation dose [49,50]. This strategy, newly applied by our group to multicopper oxidases, allowed to observe the progressive reduction of the copper ions and under particular circumstances of the molecular oxygen and monitor the nature of the intermediate states [4]. In the present structure the distances reported in Table 2 would suggest that the electron density picture observed could be the outcome of an average between the two oxidation states of the copper ions. The T3(a) and T3(b) copper ions are asymmetrically coordinated to an hydroxide/water moiety. Also in the R. lignosus laccase (RlL) structure [5] it has been observed the asymmetric binding of the hydroxide ion bridging the T3 copper ions (The T3(a)–OH bond being shorter than the T3(b)–OH bond); this was explained by assuming a reduced oxidation state for T2 and T3(b) copper ions. This asymmetry could have a crucial role in the progression of the dioxygen splitting. It has to be noted that in the present structure the T3(a) copper results to be about 0.75 Å closer to the T1 site than the T3(b) copper. Such asymmetry is generally observed in all multicopper oxidases. According to the theoretical calculations performed by Kyritsis et al. [48] for ascorbate oxidase, the particular conformation of the interposed aminoacidic residues, also observed in TtL, should render the T1–T3(a) pathway up to three times more efficient than the T1–T3(b), thus a differential reduction progress in the T2/T3 cluster would result in the observed asymmetries. As mentioned above for the T1 site, subtle differences observed among the structural data of multicopper oxidases published to date, and acquired using synchrotron radiation, are problematical to interpret since they represent averages of the starting oxidized state and of the partially/totally reduced coppers states, the weight of which rely on the X-ray doses acquired and on the other experimental conditions. Also in the laccase structure from M. albomyces a change of the redox states and asymmetrical assembly of the trinuclear structure upon data collected with diverse X-ray radiation doses has been observed [51]. At low X-ray doses a dioxygen/peroxide moiety was identified but at high doses an enzyme in the resting state was observed. A recent study from our laboratory further allowed to discern two central intermediates in the process of molecular oxygen reduction and splitting: the peroxide and the native intermediates which result from the two- and four-electrons molecular oxygen reduction, respectively. This was obtained by starting from aerobic crystals in which the complete dioxygen turnover was inhibited by the high pH due to the formation of a strong T2 Cu–OH complex
which hinders the subsequent transformation of reduced oxygen to water molecules (O2 + 2H+ ? H2O) [4]. The structural characterization of these new intermediates in the reduction of O2 to H2O in the multicopper oxidases allowed to propose a more detailed unified molecular mechanism [4]. Comparing the geometry of the metal sites of the different available laccase structures, we can notice substantial differences in Cu–ligand distances as well as Cu–Cu distances (see Supplementary material). As an example the distances between Cu ions in the T2/T3 cluster of the present enzyme are significantly higher than those expected for a completely oxidized cluster (3.6–3.8 Å). In TtL the T3(a) and T3(b) copper ions distance reaches 5.19 Å which is comparable to that observed in totally reduced laccases or other multicopper oxidases [52,53]. This lengthening of the copper–copper distances results in slight reductions of the bond distances between copper ions and their coordinated residues. We should mention that the presence of the OH/H2O moiety asymmetrically bound between the T3 copper ions has never been observed in completely reduced anaerobic clusters but usually in the resting oxidized state [7,41]. For all these reasons the picture coming out from the current data could suggest that, since during data collection the X-ray synchrotron radiation reduces the copper centers, the aerobic enzyme undergoes redox cycling. The crystals are exposed to air, then the molecular oxygen is taken and reduced into water molecules and most likely the catalytic cycle is repeated additional times resulting in collection of data generating electron density maps which represent an average of the several catalytic steps, the most significant being the completely reduced and the resting state with the OH/H2O moiety asymmetrically bound between the T3 copper ions. Acknowledgements We thank the Assessorato all’Istruzione, Formazione e Lavoro, Regione Toscana and we gratefully acknowledge the support of Progetto MECHOS, POR Ob. 3 2000/2006 Toscana, Progetti integrati di ricerca Mis. D4 (Decreto Regionale 03/04/2007 no. 1785). Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.ica.2008.03.091. References [1] B.G. Malmstrom, J. Leckner, Curr. Opin. Chem. Biol. 2 (1998) 286. [2] V. Ducros, A.M. Brzozowski, K.S. Wilson, S.H. Brown, P. Ostergaard, P. Schneider, D.S. Yaver, A.H. Pedersen, G.J. Davies, Nat. Struct. Biol. 5 (1998) 310. [3] V. Ducros, A.M. Brzozowski, K.S. Wilson, P. Ostergaard, P. Schneider, A. Svendson, G.J. Davies, Acta Crystallogr., Sect. D 57 (2001) 333. [4] M. Ferraroni, N.M. Myasoedova, V. Schmatchenko, A.A. Leontievsky, L.A. Golovleva, A. Scozzafava, F. Briganti, BMC. Struct. Biol. 7 (2007) 60. [5] S. Garavaglia, M.T. Cambria, M. Miglio, S. Ragusa, V. Iacobazzi, F. Palmieri, C. D’Ambrosio, A. Scaloni, M. Rizzi, J. Mol. Biol. 342 (2004) 1519. [6] N. Hakulinen, L.L. Kiiskinen, K. Kruus, M. Saloheimo, A. Paananen, A. Koivula, J. Rouvinen, Nat. Struct. Biol. 9 (2002) 601. [7] K. Piontek, M. Antorini, T. Choinowski, J. Biol. Chem. 277 (2002) 37663. [8] F.J. Enguita, L.O. Martins, A.O. Henriques, M.A. Carrondo, J. Biol. Chem. 278 (2003) 19416. [9] S.V.S. Kumar, P.S. Phale, S. Durani, P.P. Wangikar, Biotechnol. Bioeng. 83 (2003) 386. [10] H. Decker, T. Schweikardt, D. Nillius, U. Salzbrunn, E. Jaenicke, F. Tuczek, Gene 398 (2007) 183. [11] C.F. Thurston, Microbiol.-UK 140 (1994) 19. [12] A. Leonowicz, N.S. Cho, J. Luterek, A. Wilkolazka, M. Wojtas-Wasilewska, A. Matuszewska, M. Hofrichter, D. Wesenberg, J. Rogalski, J. Basic Microbiol. 41 (2001) 185. [13] A.M. Mayer, R.C. Staples, Phytochemistry 60 (2002) 551. [14] E.I. Solomon, R. Sarangi, J.S. Woertink, A.J. Augustine, J. Yoon, S. Ghosh, Acc. Chem. Res. 40 (2007) 581. [15] E.I. Solomon, Inorg. Chem. 45 (2006) 8012.
I. Matera et al. / Inorganica Chimica Acta 361 (2008) 4129–4137 [16] A. Decker, E.I. Solomon, Curr. Opin. Chem. Biol. 9 (2005) 152. [17] E.I. Solomon, R.K. Szilagyi, G.S. Debeer, L. Basumallick, Chem. Rev. 104 (2004) 419. [18] E.I. Solomon, P. Chen, M. Metz, S.K. Lee, A.E. Palmer, Angew. Chem., Int. Ed. Engl. 40 (2001) 4570. [19] N. Duran, M.A. Rosa, A. D’Annibale, L. Gianfreda, Enzyme Microb. Technol. 31 (2002) 907. [20] N. Duran, E. Esposito, Appl. Catal. B-Environ. 28 (2000) 83. [21] T. Deveci, A. Unyayar, M.A. Mazmanci, J. Mol. Catal. B-Enzym. 30 (2004) 25. [22] S.S. Kahraman, I.H. Gurdal, Bioresour. Technol. 82 (2002) 215. [23] L. Levin, F. Forchiassin, A. Viale, Process Biochem. 40 (2005) 1381. [24] L. Levin, A. Viale, A. Forchiassin, Int. Biodeterior. Biodegrad. 52 (2003) 1. [25] C. Haglund, L. Levin, F. Forchiassin, M. Lopez, A. Viale, Rev. Argent Microbiol. 34 (2002) 157. [26] O. Yesilada, S. Cing, D. Asma, Bioresour. Technol. 81 (2002) 155. [27] M.C. Colao, S. Lupino, A.M. Garzillo, V. Buonocore, M. Ruzzi, Microb. Cell Fact. 5 (2006) 31. [28] H. Zouari-Mechichi, T. Mechichi, A. Dhouib, S. Sayadi, A.T. Martinez, M.J. Martinez, Enzyme Microb. Technol. 39 (2006) 141. [29] L. Levin, A. Jordan, F. Forchiassin, A. Viale, Rev. Argent Microbiol. 33 (2001) 223. [30] I. Ciullini, S. Tilli, A. Scozzafava, F. Briganti, Bioresour. Technol. 2008 Feb 15 [Epub ahead of print]. [31] L. Bezalel, Y. Hadar, C.E. Cerniglia, Appl. Environ. Microbiol. 63 (1997) 2495. [32] B.S. Wolfenden, R.L. Wilson, J. Chem. Soc., Perkin Trans. 2 (1982) 805. [33] P.N. Collaborative Computational, Acta Crystallogr., Sect. D 50 (1994) 760. [34] D.E. McRee, J. Struct. Biol. 125 (1999) 156. [35] V.S. Lamzin, Acta Crystallogr., Sect. D 49 (1993) 129. [36] R.A. Laskowski, M.W. MacArthur, D.S. Moss, J.M. Thornton, J. Appl. Crystallogr. 26 (1993) 283. [37] A.M. Garzillo, M.C. Colao, C. Caruso, C. Caporale, D. Celletti, V. Buonocore, Appl. Microbiol. Biotechnol. 49 (1998) 545.
4137
[38] A.M. Garzillo, M.C. Colao, V. Buonocore, R. Oliva, L. Falcigno, M. Saviano, A.M. Santoro, R. Zappala, R.P. Bonomo, C. Bianco, P. Giardina, G. Palmieri, G. Sannia, J. Protein Chem. 20 (2001) 191. [39] M.C. Colao, A.M. Garzillo, V. Buonocore, A. Schiesser, M. Ruzzi, Appl. Microbiol. Biotechnol. 63 (2003) 153. [40] M. Ferraroni, I. Duchi, N.M. Myasoedova, A.A. Leontievsky, L.A. Golovleva, A. Scozzafava, F. Briganti, Acta Crystallogr., Sect. F 61 (2005) 205. [41] T. Bertrand, C. Jolivalt, P. Briozzo, E. Caminade, N. Joly, C. Madzak, C. Mougin, Biochemistry 41 (2002) 7325. [42] A. Messerschmidt, Multi-Copper Oxidases, World Scientific Publishing, Singapore, 1997. [43] P. Durao, I. Bento, A.T. Fernandes, E.P. Melo, P.F. Lindley, L.O. Martins, J. Biol. Inorg. Chem. 11 (2006) 514. [44] F. Xu, R.M. Berka, J.A. Wahleithner, B.A. Nelson, J.R. Shuster, S.H. Brown, A.E. Palmer, E.I. Solomon, Biochem. J. 334 (Pt 1) (1998) 63. [45] F. Xu, W. Shin, S.H. Brown, J.A. Wahleithner, U.M. Sundaram, E.I. Solomon, Biochim. Biophys. Acta 1292 (1996) 303. [46] A.E. Palmer, D.W. Randall, F. Xu, E.I. Solomon, J. Am. Chem. Soc. 121 (1999) 7138. [47] F. Xu, A.E. Palmer, D.S. Yaver, R.M. Berka, G.A. Gambetta, S.H. Brown, E.I. Solomon, J. Biol. Chem. 274 (1999) 12372. [48] P. Kyritsis, A. Messerschmidt, R. Huber, G.A. Salmon, A.G. Sykes, J. Chem. Soc., Dalton Trans. (1993) 731. [49] O. Carugo, K.D. Carugo, Trends Biochem. Sci. 30 (2005) 213. [50] R.B. Ravelli, E.F. Garman, Curr. Opin. Struct. Biol. 16 (2006) 624. [51] N. Hakulinen, K. Kruus, A. Koivula, J. Rouvinen, Biochem. Biophys. Res. Commun. 350 (2006) 929. [52] A.B. Taylor, C.S. Stoj, L. Ziegler, D.J. Kosman, P.J. Hart, Proc. Natl. Acad. Sci. USA 102 (2005) 15459. [53] I. Bento, L.O. Martins, L.G. Gato, C.M. Armenia, P.F. Lindley, Dalton Trans. (2005) 3507.