Plasticity of laccase generated by homeologous recombination in yeast Angela M. Cusano* , , Yasmina Mekmouche*, Emese Megleczà and Thierry Tron Laboratoire Biosciences, Institut des Sciences Mole ´ culaires de Marseille, Universite ´ Aix-Marseille, ISM2 CNRS UMR 6263, Marseille Cedex 20, France Keywords cupredoxin domains; functional hybrids; heterologous expression; multicopper enzyme; recombination Correspondence T. Tron, Laboratoire Biosciences, Institut des Sciences Mole ´ culaires de Marseille, Universite ´ Aix-Marseille, ISM2 CNRS UMR 6263, Avenue Escadrille Normandie Niemen, case 342, F-13397 Marseille Cedex 20, France Fax: +33 491 288440 Tel: +33 491 289196 E-mail: thierry.tron@univ-cezanne.fr *These authors contributed equally to this work Present addresses INRA NANCY, UMR 1136 Interactions Arbres–Micro-organismes, Equipe de Pathologie Forestie ` re, Route d’amance, 54280 Champenoux, France àIMEP, Case 36, Universite ´ de Provence, 3 Place Victor Hugo, 13331 Marseille Cedex 3, France Database The sequences of the laccase hybrid cDNAs lac131, lac232 and lac 535 have been sub- mitted to the GenBank database under the accession numbers FJ817449, FJ817450 and FJ817451, respectively (Received 28 May 2009, revised 9 July 2009, accepted 23 July 2009) doi:10.1111/j.1742-4658.2009.07231.x Laccase-encoding sequences sharing 65–71% identity were shuffled in vivo by homeologous recombination. Yeast efficiently repaired linearized plas- mids containing clac1, clac2 or clac5 Trametes sp. C30 cDNAs using a clac3 PCR fragment. From transformants secreting active variants, three chimeric laccases (LAC131, LAC232 and LAC535), each resulting from double crossovers, were purified, and their apparent kinetic parameters were determined using 2,2¢-azino-bis(3-ethylbenzthiazoline-6-sulphonic acid) and syringaldazine (SGZ) as substrates. At acidic pH, the apparent kinetic parameters of the chimera were not distinguishable from each other or from those obtained for the LAC3 enzyme used as reference. On the other hand, the pH tolerance of the variants was visibly extended towards alka- line pH values. Compared to the parental LAC3, a 31-fold increase in apparent k cat was observed for LAC131 at pH 8. This factor is one of the highest ever observed for laccase in a single mutagenesis step. Abbreviations ABTS, 2,2’-azino-bis(3-ethylbenzthiazoline-6-sulphonic acid); BCBD, blue copper binding domain; bp, base pair; SGZ, syringaldazine (4-hydroxy-3,5-dimethoxybenzaldehyde azine). FEBS Journal 276 (2009) 5471–5480 ª 2009 The Authors Journal compilation ª 2009 FEBS 5471 Introduction Laccases (p-diphenol oxidase, EC 1.10.3.2) are polyphe- nol oxidases that catalyse the reduction of dioxygen to water, with a concomitant oxidation of phenolic com- pounds. The enzyme active site comprises four copper atoms classified into types T1, T2 or T3, according to their spectroscopic characteristics. Substrate oxidation occurs at the T1 copper site, while the T2–T3 tri-atomic cluster is responsible for O 2 reduction [1]. The overall outcome of the catalytic cycle is reduction of one mole- cule of dioxygen into two molecules of water, coupled with oxidation of four substrate molecules (phenols or anilines) into four radicals that can form dimers, oligo- mers and polymers. These enzymes are common in plants, fungi, insects and bacteria [2,3]. Laccases are intensely studied for their potential uses in industrial processes. They generally work under mild conditions: room temperature and atmospheric pressure, with water as solvent [4–7]. Over the past decade, a significant number of reports focusing on applications of this eco-friendly enzyme in technologi- cal and bioremediation processes, in addition to their use in organic synthesis, have been published [5]. For industrial use, the current challenge is to obtain both enhanced expression levels and improved laccases with desirable physicochemical characters such as a higher redox potential, optimal activity at neutral or alkaline pH, and thermostability [8]. Strategies to obtain such variants include natural biodiversity screening and optimization of nature-derived scaffolds. Mutagenesis (rational or random) is often used to generate laccase variants. In their pioneering work, Xu et al. [9] have reported significant changes in pH opti- mum, K M and k cat for triply mutated fungal laccases. Replacement of the aspartic acid D206 by alanine in a Trametes versicolor laccase resulted in a threefold increase in k cat [10]. A similar improvement factor was also reported for variants found in simple libraries of in vitro randomly generated mutants from Fomes ligno- sus [11] or Pleurotus ostreatus [12]. On the other hand, combination of in vitro mutation and in vivo recombi- nation strategies to evolve a Myceliophthora thermo- phila laccase led to a 170-fold increase in total laccase activity, corresponding to a 22-fold improvement in k cat [13]. In a recent report, a similar approach allowed authors to isolate a variant of a M. thermophila laccase capable of resisting a wide array of co-solvents at con- centrations as high as 50% v ⁄ v [14]. In all available examples of molecular evolution of laccase, variants with improved properties have been derived from lac- case sequences from a single origin at a time. Com- pared to the shuffling of randomly mutated sequences, recombination of distantly related sequences allows large distances in sequence space to be travelled with- out disturbing the function and ⁄ or structure, but this method has yet to be applied to laccase. In a model organism such as Saccharomyces cerevisiae, homolo- gous recombination properties have largely been used for gene targeting and allele cloning. Utilizing free DNA ends as efficient substrates for homologous recombination, the gap repair methodology allow effi- cient rescue of a replicative linearized plasmid by inter- molecular recombination within co-introduced sequence-related DNA. On the other hand, it has been shown that recombination involving similar but not identical DNA sequences (homeologous DNA) can occurs at rates proportional to the length of homology [15,16]. Thus, some groups have used in vivo homeolo- gous recombination to yield low-complexity chimeric enzymes [15–17]. Usually, a chimera generated in vivo results from the shuffling of large blocks of sequence corresponding to one or more structural domains. When high-complexity chimeric enzymes are desired, in vitro recombination methods, either random [18,19] or structure-oriented [20], are preferred. The scaffold of laccases and related copper-contain- ing proteins of various functions (e.g. bacterial nitrite reductase, plant ascorbate oxidase, the E. coli metallo- oxidase CueO, human ceruloplasmin etc.) consist of repeats of a homologous sequence domain (blue cop- per-binding domain) that shares distant homology to the single-domain cupredoxins [21,22]. The evolution- ary path from a single-domain cupredoxin to a three- domain laccase (D1, D2, D3) is thought to involve a duplication of genes and recruitment of a domain [22]. During evolution from an electron transfer protein to an oxidase, proto-laccase lost unnecessary blue copper- binding sites (in D1 and D2), acquired a T2–T3 cluster binding site (the dioxygen reduction site mapping at the boundary of D1 and D3) and substrate-binding sites (one for the electron donor and one for O 2 )in neo-formed clefts [21,22] (see Fig. 1). Taking inspiration from evolutionary pathways within the blue copper binding domain (BCBD) protein family, we aim to evolve laccases into artificial catalysts performing new activities. In a first approach, basic protein engineering techniques – such as fusion of laccase with an interacting domain [23] – were used to explore properties of simple artificial laccases expressed in heterologous hosts. Here, we report on the construction of laccase chimeras through yeast- mediated homeologous recombination of Trametes sp. strain C30 laccase cDNAs sharing 65–71% identity. Functional hybrids of laccases from Trametes C30 sp. A. M. Cusano et al. 5472 FEBS Journal 276 (2009) 5471–5480 ª 2009 The Authors Journal compilation ª 2009 FEBS Active variants of laccase were selected directly on transformation plates. Expression, purification and analysis of the pH activity profile allowed the charac- terization of a variant of laccase presenting unusual oxidation activity at pH 8 corresponding to a substan- tial increase in k cat . Results Homeologous recombination of laccase-encoding sequences Chimeric laccase-encoding sequences were obtained in three independent homeologous intermolecular recom- bination experiments. In each experiment, two parental laccase-encoding cDNAs were introduced in yeast by co-transformation of a linearized expression vector, containing either clac1, clac2 or clac5, in the presence of an overlapping double-stranded PCR fragment of clac3. Upon transformation, intermolecular recombina- tion within the homeologous sequences led to re-circu- larization of the replicative plasmid, and yeast transformants were selected on the plasmid-borne URA3 marker without selection for the point of recombination between the homeologous genes. The frequency of recombination ranged from 10 2 (cla535) to 10 4 (cla131 and cla232) transformants per lgof DNA. The frequency of recombination depends on the homology of the sequences that are being recombined, and therefore frequencies one to two orders of magni- tude lower than the frequencies reported for the recombination of homologous sequences [15] probably reflect the level of identity between the sequences we used (cla1 versus cla c3, 68.4%; cl a2 versus clac3,71%; cla5 versus clac3, 65.3%). Among the transformants, active laccase-secreting clones were detected as those able to oxidize the 2-meth- oxyphenol present in the selective medium. Plasmids recovered from these transformants were first analysed by restriction mapping in order to confirm their hybrid nature (results not shown) and then sequenced. Hybrid laccase-encoding genes lac131, lac232 and lac535 obtained by recombination were all found to contain a clac3 central sequence (700–800 bp) (Fig. 1). In recom- binant sequences, junctions were found to map within short stretches of identity varying from 5–45 bp. Similar lengths for 5¢ and 3¢ recombination zones were found in other randomly picked clones (data not shown). Deduced amino acid sequences of LAC131, LAC232 and LAC535 hybrids were found to resemble more clo- sely that of LAC3 (89, 94 and 85% identity, respec- tively) than that of either LAC1 (81%), LAC2 (83%) or LAC5 (83%). All together, recombination induced swapping of amino acids for 94 positions (19% of the residues) in the original LAC3 sequence. The C- and N-termini of the hybrids were 33–37% and 9–26% different, respectively, from that of LAC3. Expression, purification and characterization of the laccase hybrids LAC3 and the LAC131, LAC232 and LAC535 hybrids were heterologously expressed in Saccharomyces cerevi- siae W303-1A, and extracellular laccase production was analysed. The production levels in the hybrids were on average six times lower than observed for LAC3 (300 UÆL )1 versus 2000 U ÆL )1 using SGZ). This may be due either to differences in the activity of the enzymes or differences in expression conditions (differ- ent plasmid context, glycosylation level etc.; see below and Discussion). Recombinant laccases were purified from 10 L fermentor cultures in three steps according to our previous protocol [24]. For all these enzymes, we obtained a yield of 20% of pure enzyme, with a specific activity of 300 UÆmg )1 determined in acetate buffer (0.1 m, pH 5.5) using SGZ as the substrate. A B Fig. 1. Schematic representation of gene structures obtained in intermolecular recombination assays. (A) Parental (lac1, lac2, lac3 and lac5) and hybrid (lac131, lac232 and lac535) sequences are represented by rectangles of variable lengths. Recombinant junctions are indicated by vertical bars. (B) Representation of cupredoxin domain organization in the laccase structure. Black diamond, T1 copper atom; black circle, T2 copper atom; white circle, T3 copper atoms. A. M. Cusano et al. Functional hybrids of laccases from Trametes C30 sp. FEBS Journal 276 (2009) 5471–5480 ª 2009 The Authors Journal compilation ª 2009 FEBS 5473 The apparent molecular mass of LAC3 and hybrids estimated by SDS–PAGE was found to be substantially higher than that expected from the amino acid sequences (Fig. 2). Previous analysis on LAC3 has suggested that these differences are due to N-hyperglycosylation, a well-known process occurring during expression of for- eign proteins in S. cerevisiae [25]. Among the laccases studied, the apparent molecular mass of LAC535 (approximately 79 kDa) was slightly lower than that observed for the three other proteins (approximately 87 kDa), probably because of the replacement of N-gly- cosylation sites during the recombination process. This is supported by an in silico analysis of the hybrid amino acid sequences, which indicated that the LAC535 sequence contains two potential N-glycosylation sites fewer than the other sequences (data not shown). Kinetic results The apparent kinetic parameters measured for LAC3 and hybrids using 2,2¢-azino-bis(3-ethylbenzthiazoline- 6-sulphonic acid) (ABTS) and syringaldazine as substrates are reported in Table 1. At pH 5.7 (in Mes buffer), a pH at which oxidation of both phenolic (SGZ) and non-phenolic (ABTS) substrates was found to be maximal for the original LAC3 enzyme, the cata- lytic efficiency of hybrids was not distinguishable from that of LAC3. Apparent K M values determined for SGZ were in the micromolar range for all the enzymes, but were in the millimolar range for ABTS. We quickly checked the pH tolerance of the hybrids in Britton–Robinson buffer using ABTS as the colori- metric substrate, as ABTS is known to be stable at 4.0 £ pH £ 11.0 [26] (Fig. 3). Various patterns of activ- ity were observed from pH 6.0–8.0, with LAC131 and LAC232 having substantial activity at neutral to alka- line pH (Fig. 3). Similar behaviour was observed with SGZ as substrate, but precipitation and reversible transformation of SGZ at neutral pH led us to discon- tinue this experiment with this substrate. Based on these initial observations, we recorded the kinetics of ABTS oxidation for LAC3 and the three hybrids in Britton–Robinson buffer at various pH within a pH range of 4.5–8.0. As expected from previous reports on laccase kinetics, the catalytic efficiency of the tested enzymes towards ABTS decreased rapidly as pH increased, reaching values < 10% of the original (i.e. at pH 4.5) between pH 7.5 and 8.0. Variations in the apparent K M , k cat and k cat ⁄ K M values as function of pH were plotted. Below pH 6.0, all enzymes behaved almost identically. Above pH 6.0, the apparent K M value for LAC131 was almost stable (a threefold decrease was observed at pH 8.0), whereas the values for LAC3, LAC232 and LAC535 were 10–20 times lower than those observed at acidic pH. Apparent k cat values decreased rapidly, but the enzymes appeared to be differently affected: the apparent k cat values for LAC3 and LAC535 were three orders of magnitude lower than the corresponding values at pH 5.0, whereas LAC232 and LAC131 values were reduced by factors of 500 and 50, respectively (Table 2). Thus, between pH 4.5 and 8.0, the LAC3 and LAC535 enzymes appear undistinguishable from a kinetic point of view. On the other hand, the LAC131 and LAC232 enzymes appear to be more tolerant to alkaline pH, as their activity profiles were found to be shifted by at Fig. 2. Coomassie staining of 8% SDS–PAGE of purified laccase enzymes. Lane 1, LAC3; lane 2, LAC131; lane 3, LAC232; lane 4, LAC535; lane M, molecular mass standards (kDa). Each well contained 4 lg of protein. Table 1. Apparent kinetic parameter values for SGZ and ABTS in 50 mM MES buffer, pH 5.7, at 30 °C. Enzyme SGZ ABTS k cat (min )1 ) K M (lM) k cat ⁄ K M (min )1 ÆlM )1 ) k cat (min )1 ) K M (lM) k cat ⁄ K M (min )1 ÆlM )1 ) LAC3 38 784 23.75 1633 82 348 2754 29.9 LAC131 22 777 11.87 1919 50 713 1599 31.7 LAC232 31 140 17.72 1757 56 851 1978 28.9 LAC535 37 419 14.70 2546 64 746 1769 36.6 Functional hybrids of laccases from Trametes C30 sp. A. M. Cusano et al. 5474 FEBS Journal 276 (2009) 5471–5480 ª 2009 The Authors Journal compilation ª 2009 FEBS least one pH unit toward alkalinity. Relative to the LAC3 parental enzyme, if one considers a higher activity at pH 8.0 as an improvement, the apparent k cat values for the LAC232 and LAC131 enzymes improved 5- and 31-fold, respectively (Fig. 4). In terms of catalytic efficiency, this corresponds to improve- ments of 5- and 12-fold, respectively. Discussion Basidiomycete genomes contain multiple genes encod- ing laccase isoenzymes, with large variations in identity (for example ranging from 38–86% in Coprinopsis cine- rea [27]). This natural diversity within in a single organism can be used for protein engineering purpose. In Trametes sp. C30, we previously characterized five genes [24,28–31], four of which encoded expressed pro- teins and were used here for molecular breeding experi- ments. The lac131, lac232 and lac535 hybrid genes contain more than half of the lac3 gene sequence, flanked by the 5¢ and 3¢ lac1, lac2 and lac5 regions, respectively. In all three chimeric genes, the recombina- tion points between parental sequences more than 1500 nucleotides long involve less than 50 nucleotides. However, in the donor sequence (lac3), the 5¢ recombi- nation zone (279 nucleotides long) is about five times larger than the 3¢ one (58 nucleotides long). In the 5¢ recombination zone, blocks of identical nucleotides are short and spread over the entire segment (66% overall identity within the four sequences), whereas in a win- dow of comparable size (about 280 nucleotides) cen- tred on the 3¢ recombination zone, the highest identity is found in the central 58 nucleotide block (87% over- all identity within the four sequences). Studying ho- meologous recombination of P450 sequences, Me ´ zard et al. [16] concluded that the preferred points of recombination could be those corresponding to maxi- mal identity in the overall alignment of the parental sequences. However, the short window of recombina- tion found in the 3¢ zone suggests a bias in the selec- tion of recombination points. It has been suggested that optimal recombination points allow swapping of structural blocks [19], and combination of large pro- tein fragments in our chimera led to fully functional enzymes. Moreover, as recombination of nature- selected sequences is conservative [32], crossovers lead- ing to functional hybrids occur at positions that mini- mize disruption of interactions [18]. In the chimera, recombination preserved the integrity of domain D1 and the very end part of domain D3, two regions that interact precisely in the natural laccase fold (Fig. 5). Because of the bias introduced by linearization of the receptor fragment at restriction sites, it is difficult to interpret the position of the recombination points rela- tive to domain D2. However, it seems that D2 ⁄ D3 interactions are favoured in the chimera (all LAC3), whereas D1 ⁄ D2 interactions are favoured only in Fig. 3. Variations in ABTS oxidation rates as function of pH for LAC3 and the hybrids LAC131, LAC232 and LAC535. ABTS (5.5 m M final concentration) was added to the appropriate enzyme solution (0.6 nM) at the desired pH. Oxidation rates are proportional to variations in the absorbance at 410 nm per minute and are indi- cated as DA ⁄ min. Inset: microtitre plate with enzyme ⁄ substrate mixtures at various pH values; the photograph was taken after 3 min of incubation at 30 °C. Table 2. Apparent kinetic parameter values for ABTS at various pH in Britton–Robinson buffer adjusted to the relevant pH at 30 °C. Enzyme k cat (min )1 ) K M (lM) k cat ⁄ K M (min )1 ÆlM )1 ) pH 5.0 pH 8.0 pH 5.0 pH 8.0 pH 5.0 pH 8.0 LAC3 18 052 10 573 57 31.5 0.18 LAC131 16 082 316 410 143 39.2 2.2 LAC232 23 336 46 466 60 50 0.78 LAC535 22 728 12 473 109 48 0.11 Fig. 4. Relative increase in apparent k cat for laccase hybrids as a function of pH. A. M. Cusano et al. Functional hybrids of laccases from Trametes C30 sp. FEBS Journal 276 (2009) 5471–5480 ª 2009 The Authors Journal compilation ª 2009 FEBS 5475 LAC535 (all LAC5). Nevertheless, the 5¢ recombina- tion points apparently match structural block limits in chimeras as junctions were found: at the limit of domain D2 in LAC131, at the limit of domain D1 in LAC232, and at the position (or nearby) of the cyste- ine residue C228 (LAC3 numbering) that is involved in a disulfur bridge with C140 (D1) in LAC535. Based on the present observations, a better knowledge on toler- ance to block exchange in the laccase enzyme should be obtained by in vitro sequence permutation experi- ments and swapping of cupredoxin domains (D1, D2, D3). Such experiments are in progress. One of the beneficial effects of production of the present Trametes sp. C30 laccase chimeras was to create hybrid sequences that are better expressed in the host than the parental sequences clac1, clac2 and clac5. Among the parental sequences used for this recombi- nation study, only constructions bearing the sequence encoding the LAC3 isoenzyme have previously been found to lead to substantial production of recombinant enzyme in yeast [29,30]. Whereas LAC1 and LAC2 have been purified and fully characterized from Tra- metes sp. C30 [28,31], their recombinant counterparts produced in yeast are barely detectable on activity plates (T. Tron, unpublished results). These differences in expression of recombinant enzyme coding sequences are likely largely related to inappropriate codon usage by the heterologous host, as low-frequency codons can cause translation pauses depending on their position and abundance. Upon recombination with clac3 Fig. 5. Molecular models of LAC3 and hybrids. Models were constructed using the structure of the laccase 2HRG from T. trogii as template. A ribbon representation is used for the LAC3 model; copper atoms are represented as grey spheres; the p-methylbenzoate present in the structure 2HRG is used in the models to indicate a potential substrate-interacting zone. A surface representation is used for the hybrids; the surface of the parts of hybrids originating from LAC3 is coloured in light blue; the surface of the parts of hybrids originating from either LAC1, LAC2 or LAC5 but identical to LAC3 is coloured in dark blue; the surfaces of the parts of hybrids corresponding to LAC1, LAC2 or LAC5 substitutions are coloured in yellow, red and magenta, respectively. Functional hybrids of laccases from Trametes C30 sp. A. M. Cusano et al. 5476 FEBS Journal 276 (2009) 5471–5480 ª 2009 The Authors Journal compilation ª 2009 FEBS sequence, the clac1-, clac2- and clac5-based construc- tions led to functional expression of the hybrid sequences clac131, clac232 and clac535. A simple inspection of coding preference plots [33] for the parental and chimeric sequences confirmed that exchanging large sequence segments with the clac3 sequence substantially reduces the number of codons that potentially cause translation pauses. During func- tional expression of a M. thermoplila laccase gene in S. cerevisiae by directed evolution, synonymous muta- tions to more frequently used codons improved pro- duction of the recombinant laccase up to eightfold [13]. For our laccase chimera, it is difficult to calculate a fold improvement in production relative to LAC1, LAC2 or LAC5 because of the absence of a reference level for these parental enzymes. On the other hand, compared to LAC3, as the steady-state kinetic parame- ters for all the enzymes are of the same order of mag- nitude, the ratio of the total volumetric activities reflects a decrease in production by the hybrids of approximately fivefold. This suggests that the codon usage can probably be improved further, for example through design of synthetic sequences. LAC3 is representative of a class of laccases found in basidiomycetes: it is an acidic enzyme that works best in a pH window from 4.5–6.0. Under catalysis conditions previously established for LAC3 (buffer, pH, temperature), all three variants are as active as LAC3. This is remarkable as a 60–90% decrease in activity has been reported for P450 chimera (similar to our laccases in sequence size, identity, recombination area), although, in this case, chimera activities may account both for intrinsic kinetic differences in sub- strate oxidation and differences in interaction with a reductase [16]. In our case, as discussed above, recom- bination essentially preserved domain interactions as well as the architecture of coordination sites. More- over, the substrate-interacting zone, as defined by the location of substrate analogues in the crystal structures of Coriolaceae laccases 1KIA [34] and 2HRG (http:// www.rcsb.org), is identical to that of the LAC3 enzyme, either because it is entirely composed of LAC3 sequence (LAC131, LAC232) or because residue variations in that zone are conservative (LAC535) (Fig. 5). These may be major reasons why the kinetic behaviour of the hybrids is closer to that of LAC3 rather than that of the other parental enzymes [28,31]. Like other basidiomycetous laccases, LAC3 variant activities are progressively inhibited by an increasing concentration of OH ) [35], which binds the T2 copper, but significant differences distinguish them from each other. Thus, the LAC3 and LAC535 pH profiles are superimposable, suggesting strong conservation of the original LAC3 properties upon recombination, although the protein sequence of this hybrid is the least related to that of LAC3 (Fig. 5). On the other hand, LAC131 and LAC232 hybrids oxidize ABTS 31- and 5- fold faster, respectively, than the parental LAC3 enzyme at pH 8.0. As the kinetic behaviour of all of our enzymes is very similar below pH 6.0, these results probably reflect a significant improvement in the stabil- ity of LAC131 and LAC232 hybrids at alkaline pH. Further studies on this type of mutants should help to deepen our knowledge on protein regions modulating laccase activity in response to pH changes. In conclusion, recombination of large fragments of sequence coding for laccase isoenzymes leads to the exchange of structural blocks, allowing synthesis of hybrid enzymes with properties that distinguish them from the parental enzymes. Differences in laccase activity observed at pH 8.0 do not reflect an enhance- ment in k cat but rather reflect an enhancement of the enzyme stability at alkaline pH. Nevertheless, the cata- lytic efficiency of the best-performing hybrid (LAC131) is more than 12 times that of the parental enzyme (LAC3). Compared to studies involving mutagenesis, such a factor is one of the highest ever observed in a single step. Thus, hybrids obtained by homeologous recombination constitute a valuable tool set to study the plasticity of the enzyme. Experimental procedures Materials and reagents Chemicals were purchased from Sigma-Aldrich (St Louis, MO, USA) and were of the highest available grade. The Britton–Robinson buffer was produced by mixing 0.1 m boric acid, 0.1 m acetic acid and 0.1 m phosphoric acid with 45% NaOH to the desired pH. 2-(N-morpholino)ethane- sulfonic acid (Mes) buffer was adjusted to pH 5.7 with NaOH. Spectroscopic measurements were performed using either a CARRY 50 spectrophotometer (Varian, Palo Alto, CA, USA) or a KC4 microtitre plate reader (BioTek, Winooski, VT, USA). A DuoFlow FPLC apparatus (Bio- Rad, Hercules, CA, USA) was used for chromatographic separations. Strains and vectors used for cloning and expression S. cerevisiae W303-1A (MATD, ade2-1, his3-11, 15, leu2- 3/112, trp1-1, ura3-1, can1-100) was used for expression of laccase. Yeast expression vector pDP51 (2l, Amp r , URA3, GAL10 ⁄ CYC1) pBM258 (GAL1 ⁄ GAL10, CEN4 ⁄ ARS1, Amp r , URA3) and pSAL4 (2l, Amp r , URA3, CUP1) were A. M. Cusano et al. Functional hybrids of laccases from Trametes C30 sp. FEBS Journal 276 (2009) 5471–5480 ª 2009 The Authors Journal compilation ª 2009 FEBS 5477 respectively obtained from Dr D. Pompon (Centre de Ge ´ ne ´ tique Mole ´ culaire, Gif ⁄ Yvette, France), Dr D. Bot- stein (Princeton University, NJ, USA) and Dr R. Gaxiola (Departamento de Biologı ´ a Molecular de Plantas, UNAM, Cuernavaca, Morelos, Mexico). Standard techniques were used for cloning, transformation and analysis [36]. Construction of chimera and gap repair The four parental laccase-encoding sequences clac1, clac2, clac3 and clac5 have been previously isolated from Trametes sp. strain C30 and heterologously expressed in S. cerevisiae in our laboratory [24,29–31]. Expression vectors bearing clac1 (AKY160), clac2 (EMY162) or clac5 (EMY164) sequences were linearized at the SmaI, Kpn2I and ClaI restriction sites, respectively, located in the laccase-coding region. The clac3 sequence was amplified by PCR from the construct pAKY145 [29,30] using EM53 (5¢-TTCCTTTTG GCTGGTTTTGC-3¢) and EM54 (5¢-CAGTTATTACCC TATGCGGTGTGA-3¢), respectively, as forward and reverse primers. The resulting 2015 bp amplicon was gel-purified and further used in co-transformation assays (lg donor DNA/lg vector DNA=4) with various linearized laccase-encoding vectors. Transformants were plated on selective medium (per litre: yeast nitrogen base without amino acids and ammonium sulfate, 6.7 g; casaminoacids, 5 g; adenine sulfate, 30 mg; CuSO 4 100 lm; succinate buffer 50 mm, pH 5.3; 1.5% agar) containing 2% galactose as the carbon source and 0.05% v ⁄ v guaı ¨ acol as the laccase substrate. Laccase-active transformants were picked and further studied. Enzyme production Yeasts were cultivated at 28 °C. Pre-cultures were obtained in two stages from a single colony freshly grown on a selec- tive plate. Cells were first grown in 15 mL tubes containing 5 mL of selective medium for 24 h on a rotating wheel. A volume of suspension sufficient to reach a final attenuance at 600 nm of 0.1 was then used to inoculate 250 mL Erlen- meyer flasks containing 50 mL of selective medium, and cells were then grown for 24 h on a reciprocal shaker (150 rpm). Bio-reactor cultivations (batch) were performed in a 15 L fermentor vessel (B. Braun Biotech International GmbH, Melsungen, Germany) containing 10 L of selective medium. The inoculum was added to a final attenuance at 600 nm of 0.1, and yeasts were grown under stirring (220 rpm) and with an air flow of 16 LÆh )1 . Samples (1 mL) to be used for laccase activity and cell density determination were withdrawn and analysed regularly throughout cultivation. Purification Cells were sedimented by centrifugation at 1600 g and 4 °C for 10 min. Culture supernatant (10 L) was successively filtered through Whatman paper glass (porosity 1.5 lm), 0.45 and 0.22 lm poly(vinylidene difluoride) membranes. Filtrate was then concentrated 50-fold by ultrafiltration using a Prep ⁄ Scale cartridge (approximately 0.23 m 2 YM10 membrane, Amicon ⁄ Millipore, Bedford, MA, USA), and buffer-exchanged with 20 mm phosphate, pH 6.0 (buffer A). The sample was further concentrated to 50 mL on 76 mm diameter YM10 membrane and applied to an ion- exchange DEAE-Sepharose column (2.5 · 20 cm, Amer- sham Pharmacia Biotech Europe GmbH, Freiburg, Ger- many) pre-equilibrated with the same buffer. Proteins were eluted at a flow rate of 4 mLÆmin )1 with a step gradient of NaCl: 0.1, 0.15, 0.2, 0.25, 0.3 and 1 m. Fractions containing laccase activity were pooled and concentrated to a volume of 600 lL by ultrafiltration on a 25 mm diameter YM10 membrane, and loaded on a Superdex S200 column (Amer- sham Pharmacia) equilibrated with 20 mm phosphate, pH 6.0, 200 mm NaCl. Fractions containing laccase activity were pooled and concentrated. Exchange with buffer con- taining no salt, concentration and addition of 15% glycerol were undertaken for long-term storage of the protein ()20 °C). Enzyme purity in active fractions was then con- firmed by SDS–PAGE. Standard enzyme assay Protein concentration was determined by the Bradford method using BSA as standard, or by UV-vis spectroscopy (e 600 nm =5· 10 3 m )1 Æcm )1 for the T1 copper) [37]. Lac- case activity was routinely assayed at 30 °C using SGZ as the substrate. Oxidation of SGZ was detected by measuring the absorbance increase at 525 nm (e 525nm = 6.5 · 10 4 m )1 Æcm )1 ) after 2 min using a spectrometer (Carry 50 UV- vis spectrophotometer) [38]. The reaction mixture (1 mL) contained 10 lL of appropriately diluted enzyme sample and 980 lL of Mes buffer (50 mm, pH 5.7), and 10 lLof 0.8 mgÆmL )1 SGZ in MeOH was added to initiate the reac- tion. One unit (U) of laccase oxidizes one micromole of substrate per minute. Kinetic parameter determination and effect of pH Determination of kinetics parameters was undertaken using two substrates: SGZ and ABTS. For SGZ, the same condi- tions were used as those for the standard enzyme assay. ABTS oxidation was determined in both MES and Britton– Robinson buffers by monitoring the absorbance change at 414 nm with an extinction coefficient of 3.5 · 10 4 m )1 Æcm )1 [39]. Variation of the ABTS oxidation rate as function of pH was assayed in a 96-well plate at 30 °C for 2 min using Britton–Robinson buffer adjusted to a pH from 4.5–8.0. Apparent K M and k cat values were obtained from the initial rate (v), enzyme concentration (E) and substrate concentra- tion (S) according to the equation v = k cat ES⁄ (K M + S) (non-linear regression fitting using prizm program, Graph- Functional hybrids of laccases from Trametes C30 sp. A. M. Cusano et al. 5478 FEBS Journal 276 (2009) 5471–5480 ª 2009 The Authors Journal compilation ª 2009 FEBS pad, San Diego, CA). Because laccase catalysis involves two substrates and the [O 2 ] was invariant and assumed to be saturating in this study, the measured K M for the vari- ous substrates used should be considered apparent. Because of the assumption that 100% of the laccase participated in the catalysis as active enzyme, the measured k cat should also be considered apparent. Molecular models 3D models were obtained from the Swiss Model Server (swissmodel.expasy.org) using the crystallographic coordi- nates from Trametes trogii laccase 2HRG obtained from the Research Collaboratory for Structural Bioinformatics (RCSB) Protein Data Bank (http://www.rcsb.org). Acknowledgements This work was partly supported by the European Commission, Sixth Framework Program (NMP2- CT2004-505899, SOPHIED). Angela Cusano was the recipient of a Re ´ gion Provence Alpes Coˆ te d’Azur Postdoc fellowship. Emese Meglecz was the recipient of a Ministe ´ re de le Recherche Postdoc fellowship. We thank Marius Re ´ glier, Jalila Simaan, Erin Wallace- Bomati and Gilles Iacazio (Laboratoire Biosciences, Institut des Sciences Mole ´ culaires de Marseille, Univer- site ´ Aix-Marseille, France) for helpful discussions. References 1 Solomon EI, Sundaram UM & Machonkin TE (1996) Multicopper oxidases and oxygenases. Chem Rev 96, 2563–2605. 2 Baldrian P (2006) Fungal laccases – occurrence and properties. FEMS Microbiol Rev 30, 215–242. 3 Gianfreda L, Xu F & Bollag J-M (1999) Laccases: a useful group of oxidoreductive enzymes. Bioremediat J 3, 1–25. 4 Faber K (2004) Biotransformations in Organic Chemis- try. Springer, Berlin. 5 Riva S (2006) Laccases: blue enzymes for green chemis- try. Trends Biotechnol 24, 219–226. 6 Rodrı ´ guez Couto S & Tocca Herrera J (2006) Industrial and biotechnological applications of laccases: a review. Biotechnol Adv 24, 500–513. 7 Wells A, Teria M & Eve T (2006) Green oxidations with laccase-mediator systems. Biochem Soc Trans 34, 304–308. 8 Zuma ´ rraga M, Camarero S, Shleev S, Martı ´ nez-Arias A, Ballesteros A, Plou FJ & Alcalde M (2008) Altering the laccase functionality by in vivo assembly of mutant libraries with different mutational spectra. Proteins 71, 250–260. 9 Xu F, Berka RM, Wahleithner JA, Nelson BA, Shuster JR, Brown SH, Palmer AE & Solomon EI (1998) Site- directed mutations in fungal laccase: effect on redox potential, activity and pH profile. Biochem J 334, 63–70. 10 Madzak C, Mimmi MC, Caminade E, Brault A, Baumberger S, Briozzo P, Mougin C & Jolivalt C (2006) Shifting the optimal pH of activity for a laccase from the fungus Trametes versicolor by structure-based mutagenesis. Protein Eng Des Sel 19, 77–84. 11 Hu MR, Chao YP, Zhang GQ, Yang XQ, Xue ZQ & Qian SJ (2007) Molecular evolution of Fome lignosus laccase by ethyl methane sulfonate-based random muta- genesis in vitro. Biomol Eng 24, 619–624. 12 Festa G, Autore F, Fraternali F, Giardina P & Sannia G (2008) Development of new laccases by directed evo- lution: functional and computational analyses. Proteins 72, 25–34. 13 Bulter T, Alcalde M, Sieber V, Meinhold P, Schlacht- bauer C & Arnold FH (2003) Functional expression of a fungal laccase in Saccharomyces cerevisiae by directed evolution. Appl Environ Microbiol 69, 987–995. 14 Zumarraga M, Bulter T, Shleev S, Polaina J, Martinez- Arias A, Plou FJ, Ballesteros A & Alcalde M (2007) In vitro evolution of a fungal laccase in high concentra- tions of organic cosolvents. Chem Biol 14, 1052–1064. 15 Me ´ zard C & Nicolas A (1994) Homologous, homeolo- gous, and illegitimate repair of double-strand breaks during transformation of a wild-type strain and a rad52 mutant strain of Saccharomyces cerevisiae . Mol Cell Biol 14, 1278–1292. 16 Me ´ zard C, Pompon D & Nicolas A (1992) Recombina- tion between similar but not identical DNA sequences during yeast transformation occurs within short stretches of identity. Cell 70, 659–670. 17 Volkov AA, Shao Z & Arnold FH (1999) Recombina- tion and chimeragenesis by in vitro heteroduplex forma- tion and in vivo repair. Nucleic Acids Res 27, e18. 18 Crameri A, Raillard S, Bermudez E & Stemmer WP (1998) DNA shuffling of a family of genes from diverse species accelerates directed evolution. Nature 391, 288– 291. 19 Ostermeier M, Nixon AE & Benkovic SJ (1999) Incre- mental truncation as a strategy in the engineering of novel biocatalysts. Bioorg Med Chem 10, 2139–2144. 20 Voigt CA, Martinez C, Wang ZG, Mayo SL & Arnold FH (2002) Protein building blocks preserved by recom- bination. Nat Struct Biol 9, 553–558. 21 Nakamura K & Go N (2005) Function and molecular evolution of multicopper blue proteins. Cell Mol Life Sci 62, 2050–2066. 22 Nersissian AM & Shipp EL (2002) Blue copper-binding domains. Adv Protein Chem 60, 271–340. 23 Balland V, Hureau C, Cusano AM, Liu Y, Tron T & Limoges B (2008) Oriented immobilization of a fully A. M. Cusano et al. Functional hybrids of laccases from Trametes C30 sp. FEBS Journal 276 (2009) 5471–5480 ª 2009 The Authors Journal compilation ª 2009 FEBS 5479 active monolayer of histidine-tagged recombinant laccase on modified gold electrodes. Chemistry 14, 7186–7192. 24 Klonowska A, Le Petit J & Tron T (2001) Enhance- ment of minor laccases production in the basidiomycete Marsasmius quercophilus C30. FEMS Microbiol Lett 200, 25–30. 25 Romanos MA, Scorer CA & Clare JJ (1992) Foreign gene expression in yeast: a review. Yeast 8, 423–488. 26 Xu F, Shin WS, Brown SH, Wahleithner JA, Sundaram UM & Solomon EI (1996) A study of a series of recom- binant fungal laccases and bilirubin oxidase that exhibit significant differences in redox potential, substrate speci- ficity, and stability. Biochim Biophys Acta 1292, 303–311. 27 Kilaru S, Hoegger PJ & Ku ¨ es U (2006) The laccase multi-gene family in Coprinopsis cinerea has seventeen different members that divide into two distinct subfami- lies. Curr Genet 50, 45–60. 28 Dedeyan B, Klonowska A, Tagger S, Tron T, Iacazio G, Gil G & Le Petit J (2000) Biochemical and molecu- lar characterization of a laccase from Marasmi- us quercophilus. Appl Environ Microbiol 66, 925–929. 29 Klonowska A (2000) Expression he ´ te ´ rologue de laccases du champignon Marasmius quercophilus chez la levure Saccharomyces cerevisiae. PhD Thesis, Universite ´ Aix-Marseille III, Marseille, France. 30 Klonowska A, Gaudin C, Asso M, Fournel A, Reglier M & Tron T (2005) LAC3, a new low redox potential lac- case from Trametes sp. strain C30 obtained as a recombi- nant protein in yeast. Enzyme Microb Technol 36, 34–41. 31 Klonowska A, Gaudin C, Fournel A, Asso M, Le Petit J, Giorgi M & Tron T (2002) Characterization of a low redox potential laccase from the basidiomycete C30. Eur J Biochem 269, 6119–6125. 32 Ness JE, Del Cardayre ´ S, Minshull J & Stemmer WPC (2001) Molecular breeding: the natural approach to protein design. Adv Protein Chem 55, 261–292. 33 Gribskov M, Devereux J & Burgess RR (1984) The codon preference plot: graphic analysis of protein coding sequences and prediction of gene expression. Nucleic Acids Res 12, 539–549. 34 Bertrand T, Jolivalt C, Briozzo P, Caminade E, Joly N, Madzak C & Mougin C (2002) Crystal structure of a four-copper laccase complexed with an arylamine: insights into substrate recognition and correlation with kinetics. Biochemistry 41, 7325–7333. 35 Xu F (1997) Effects of redox potential and hydroxide inhibition on the pH activity profile of fungal laccases. J Biol Chem 272, 924–928. 36 Sambrook J, Fritsch EF & Maniatis T (1989) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York. 37 Bradford M (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein–dye binding. Anal Biochem 2, 248–254. 38 Bauer R & Rupe CO (1971) Use of syringaldazine in a photometric method for estimating ‘free’ chlorine in water. Anal Chem 43, 421–425. 39 Childs RE & Bardsley WG (1975) The steady-state kinetics of peroxidase with 2,2¢-azino-di-(3-ethyl-benz- thiazoline-6-sulphonic acid) as chromogen. Biochem J 45, 93–103. Functional hybrids of laccases from Trametes C30 sp. A. M. Cusano et al. 5480 FEBS Journal 276 (2009) 5471–5480 ª 2009 The Authors Journal compilation ª 2009 FEBS . Plasticity of laccase generated by homeologous recombination in yeast Angela M. Cusano* , , Yasmina Mekmouche*, Emese Megleczà. recombination of laccase- encoding sequences Chimeric laccase- encoding sequences were obtained in three independent homeologous intermolecular recom- bination