Nội dung

Production and characterization of a secreted, C-terminally processed tyrosinase from the filamentous fungus Trichoderma reesei Emilia Selinheimo1, Markku Saloheimo1, Elina Ahola2, Ann Westerholm-Parvinen1, Nisse Kalkkinen2, Johanna Buchert1 and Kristiina Kruus1 1 VTT Technical Research Centre of Finland, Espoo, Finland 2 Protein Chemistry Research Group and Core Facility, Institute of Biotechnology, University of Helsinki, Finland Keywords fungal; secreted; Trichoderma reesei; tyrosinase Correspondence E. Selinheimo, VTT Technical Research Centre of Finland, PO Box 1000, Espoo FIN-02044 VTT, Finland Fax: +358 20 722 7071 Tel: +358 20 722 7135 E-mail: Emilia.Selinheimo@vtt.fi (Received 16 May 2006, revised 7 July 2006, accepted 27 July 2006) doi:10.1111/j.1742-4658.2006.05429.x A homology search of the genome database of the filamentous fungus Trichoderma reesei identified a new T. reesei tyrosinase gene tyr2, encoding a protein with a putative signal sequence. The gene was overexpressed in the native host under the strong cbh1 promoter, and the tyrosinase enzyme was secreted into the culture supernatant. This is the first report on a secreted fungal tyrosinase. Expression of TYR2 in T. reesei resulted in good yields, corresponding to approximately 0.3 and 1 gÆL)1 tyrosinase in shake flask cultures and laboratory-scale batch fermentation, respectively. T. reesei TYR2 was purified with a three-step purification procedure, consisting of desalting by gel filtration, cation exchange chromatography and size exclusion chromatography. The purified TYR2 protein had a significantly lower molecular mass (43.2 kDa) than that calculated from the putative amino acid sequence (61.151 kDa). According to N-terminal and C-terminal structural analyses by fragmentation, chromatography, MS and peptide sequencing, the mature protein is processed from the C-terminus by a cleavage of a peptide fragment of about 20 kDa. The T. reesei TYR2 polypeptide chain was found to be glycosylated at its only potential N-glycosylation site, with a glycan consisting of two N-acetylglucosamines and five mannoses. Also, low amounts of shorter glycan forms were detected at this site. T. reesei TYR2 showed the highest activity and stability within a neutral and alkaline pH range, having an optimum at pH 9. T. reesei tyrosinase retained its activity well at 30 C, whereas at higher temperatures the enzyme started to lose its activity relatively quickly. T. reesei TYR2 was active on both l-tyrosine and l-dopa, and it showed broad substrate specificity. Tyrosinase (monophenol, o-diphenol:oxygen oxidoreductase, EC 1.14.18.1) is a copper-containing metalloprotein that is ubiquitously distributed in nature. Tyrosinases are found in prokaryotic as well as in eukaryotic microorganisms, and in mammals, invertebrates and plants. Tyrosinase is a mono-oxygenase and a bifunctional enzyme that catalyzes the o-hydroxyla- tion of monophenols and subsequent oxidation of o-diphenols to quinones [1,2]. The activities are also referred to as cresolase or monophenolase and catecholase or diphenolase activities, respectively. Tyrosinase thus accepts monophenols and diphenols as substrates, and the monophenolase activity is the initial rate-determining reaction [2,3]. Abbreviations TYR2, tyrosinase 2 from Trichoderma reesei; Q-TOF, quadrupole time-of-flight. 4322 FEBS Journal 273 (2006) 4322–4335 ª 2006 The Authors Journal compilation ª 2006 FEBS E. Selinheimo et al. In mammals, tyrosinase catalyzes reactions in the multistep biosynthesis of melanin pigments, being responsible, for instance, for skin and hair pigmentation [4]. Tyrosinases play an important role in regulation of the oxidation–reduction potential, and the wound-healing system in plants [5,6]. They are also related to browning reactions of fruit and vegetables [7]. Tyrosinase activity has an essential role in some plant-derived food products, e.g. tea, coffee, raisins and cocoa, where it produces distinct organoleptic properties [8]. Most commonly, tyrosinase-mediated reactions in plants, however, are related to the browning reactions that are considered harmful [9]. To date, the information on the physiologic role of tyrosinases in microorganisms is very limited. The most extensively investigated fungal tyrosinases, from both a structural and a functional point of view, are from Agaricus bisporus [10] and Neurospora crassa [1]. Studies with N. crassa have shown that the enzyme is completely absent in the vegetative stage. However, under stress conditions high levels of the enzyme can be induced [1]. This suggests that tyrosinases are not essential to the metabolism of the fungi, but improve the survival and competence of the fungi by producing melanins. Tyrosinases have been shown to share a similar active site with catechol oxidase and hemocyanin, a protein involved in oxygen transport in arthropods and molluscs [11]. These proteins are type 3 copper proteins with a diamagnetic spin-coupled copper pair in the active center. Each of the two copper atoms is coordinated by three conserved histidine residues [12]. Molecular oxygen is used as an electron acceptor and it is reduced to water in tyrosinase-catalyzed reactions. On the basis of thorough chemical and spectroscopic analyses of tyrosinases, the binuclear active site is known to exist in three states: oxy-tyrosinase, mettyrosinase and deoxy-tyrosinase [13–15]; a catalytic cycle, in which these states are alternated, has been proposed [16]. The met state is the resting state of the enzyme, and in the absence of substrate about 85–90% of the enzyme is in this state [16]. Both the met and oxy states of tyrosinases can catalyze the diphenoloxidase reaction, whereas the monohydroxylase reaction requires the oxy state. Just recently, the first tyrosinase structure from Streptomyces castaneoglobisporus [17] became available, and will enable more detailed analysis of the exact reaction mechanisms. The tyrosinase structure, wherein the active site was located at the bottom of a large vacant space and one of the six histidine ligands appeared to be highly flexible, was determined with a help of a caddie protein, ORF378, at 1.2–1.8 Å resolution [17]. Knowledge of fungal tyrosinases is still limited, and the work has been hampered Secreted tyrosinase from Trichoderma reesei by relatively low production yields of the enzymes. In this article, the cloning, production and characterization of a novel tyrosinase from the filamentous fungus T. reesei is reported. Results Isolation of a tyrosinase gene from Trichoderma reesei A homology search was performed against the genome sequence of T. reesei (http://gsphere.lanl.gov/trire1/ trire1.home.html). This revealed two uncharacterized genes showing clear similarity with known tyrosinase sequences. Analysis of the deduced protein sequence encoded by the tyr2 gene with the program signalp [24] indicated that the protein has a signal sequence, and should thus be a secreted enzyme. The T. reesei tyr2 gene and the corresponding cDNA were cloned by PCR and sequenced in order to verify the sequence at the genome website, to exclude PCR mutations and to localize the introns. The gene is interrupted by seven short introns. The encoded protein consists of 571 amino acids, including a predicted signal sequence of 18 amino acids, and three potential N-glycosylation sites. The closest homologs of the T. reesei TYR2 protein are putative tyrosinases from the fungi Gibberella zeae (46% amino acid identity), N. crassa (35% identity) and Magnaporthe grisea (34% identity). All these three proteins are predicted by the signalp program to have a signal sequence; however, none of them has been characterized at the protein level. The amino acid identity of TYR2 to the intracellular tyrosinase from Pycnoporus sanguineus is 34% [25], whereas the amino acid identity to other fungal tyrosinases is around 25–33%. Alignment of T. reesei TYR2 with the G. zeae tyrosinase places the suggested signal sequence cleavage sites of both enzymes precisely at the same location (Fig. 1). When these two sequences are aligned with the intracellular tyrosinase characterized from N. crassa [26], the suggested N-termini of the mature secreted enzymes coincide with the N-terminus of the intracellular enzyme (Fig. 1). The segments in the N-terminal portion around the copper ligand amino acids of the active site are well conserved between the proteins, whereas the C-terminal domains are less conserved. Overexpression of the tyrosinase gene in Trichoderma reesei Tyrosinase production by an untransformed T. reesei strain was tested, but no tyrosinase activity in culture FEBS Journal 273 (2006) 4322–4335 ª 2006 The Authors Journal compilation ª 2006 FEBS 4323 Secreted tyrosinase from Trichoderma reesei E. Selinheimo et al. Fig. 1. Alignment of T. reesei TYR2 (TrTYR2) amino acid sequence with the putatively secreted tyrosinase of Gibberella zeae (GzTYR) and the intracellular tyrosinase characterized from Neurospora crassa (NcTYR1). Identical amino acids are indicated by asterisks, conserved substitutions by colons, and similar amino acids by dots. The hisitidines acting as ligands for the Cu atoms A and B are shaded. The Cys–His thioester bond in the active site is indicated. Signal sequence cleavage sites of TrTYR2 and GzTYR are indicated by arrows. The last amino acids of processed TrTYR2 and NcTYR1 are marked by triangles. Putative N-glycosylation sites are in bold. supernatants or in cell lysates could be detected. Therefore, the tyrosinase was overexpressed in T. reesei. An expression construct in which the proteincoding region of the genomic tyr2 is between the cbh1 promoter and terminator was made by in vivo recombination with the Gateway recombination system. The cbh1 promoter is a strong inducible promoter and active throughout cultivation. The construct (pMS190) was transformed into T. reesei, and the transformants were tested with a plate activity assay with tyrosine as the indicator substrate. A number of transformants developing a stronger brown color around the streaks than the parental strain were found (data not shown). These uninucleate clones were isolated and tested for tyrosinase production in shake flask cultures. The best 4324 transformant produced 40.1 nkatÆmL)1 of tyrosinase activity. The first test cultures were made with 0.1 mm CuSO4 in the medium. The effect of copper concentration on the production level was studied by using 0–6 mm CuSO4 in the medium in cultures of the best transformant. The optimal copper concentration was 2 mm, but relatively good production was obtained at 1–4 mm. The highest tyrosinase production obtained in shake flask cultures was 96 nkatÆmL)1. The best tyr2-overexpression transformant was grown in a laboratory fermenter in a volume of 20 L. The enzyme production increased continuously during culture, and the activity level of 300 nkatÆmL)1 was reached after 6 days of cultivation. Although the activity was still increasing, the fermentation had to be stopped because FEBS Journal 273 (2006) 4322–4335 ª 2006 The Authors Journal compilation ª 2006 FEBS E. Selinheimo et al. Secreted tyrosinase from Trichoderma reesei Table 1. Purification of T. reesei TYR2. Purification step Culture filtrate Desalting Cation exchange chromatography Gel filtration Total activity (nkat) Total protein (mg) Specific activity (nkatÆmg)1) Activity yield (%) Purification factor 50 000 43 165 19 120 816 694 74 61.3 62.2 258.4 86 38 1.0 4.2 7270 24 303.0 15 4.9 of foaming problems. According to the specific activity of the purified tyrosinase (300 nkatÆmg)1; Table 1), the highest activity obtained in fermentation, 300 nkatÆmL)1, corresponds to about 1 g of the enzyme per liter of culture supernatant. Enzyme purification Enzyme purification was started with desalting by gel filtration (Sephadex G25). The following cation exchange chromatography was performed in 10 mm Tris ⁄ HCl, pH 7.3. Tyrosinase eluted at an NaCl concentration of 120 mm. Because of the high pI of T. reesei TYR2, most of the Trichoderma cellulases and hemicellulases could be separated from the tyrosinase-containing fractions. The final purification step was carried out with gel filtration (Sephacryl S-100). The overall recovery of activity in the three-step purification procedure was 15% (Table 1). ponding to a pI around 9.5. The purified T. reesei TYR2 appeared as a double protein band on SDS ⁄ PAGE gel (Fig. 2), with an apparent molecular mass of 43 kDa, which is far below the theoretical value of 61 151 Da calculated from the encoded amino acid sequence (including the signal sequence). The result suggested that T. reesei TYR2 is processed, as also described for several other fungal tyrosinases [25– 28]. The purified tyrosinase had an absorption maximum at around 350 nm, which is an indication of a T3-type copper pair in its oxidized form with a bridging hydroxyl moiety, assigned as an O22– fi Cu2+ charge transfer transition [29]. For molecular characterization, the purified enzyme was first subjected to reversed-phase chromatography, where it eluted as one symmetric peak (Fig. 3). Further analysis of the reversed-phase purified protein by SDS ⁄ PAGE still gave a double protein band corresponding to a molecular mass of about 43 kDa. AU 214 nm Biochemical characterization IEF of the purified T. reesei TYR2, and subsequent staining with l-dopa, showed a band in the gel corres- 2.5 kDa MW 1 97.0 66.0 45.0 2.0 30.0 20.1 MW 1 2 3 4 MW kDa 1.5 203.6 116.1 92.3 1.0 50.4 37.0 14.4 0.5 0.0 0 10 20 30 40 50 min 28.9 20.0 6.9 Fig. 2. Purification of T. reesei TYR2 as analyzed by SDS ⁄ PAGE (12% Tris ⁄ HCl gel). Gel lanes: MW, molecular mass markers; 1, culture filtrate; 2, desalted culture filtrate, enzyme preparation after cation exchange; 3, enzyme preparation after gel filtration. Fig. 3. Reversed-phase chromatographic analysis of T. reesei TYR2 from the last gel filtration purification step. Chromatography was performed on a 1 · 20 mm TSKgel TMS-250 column using a linear gradient of acetonitrile (3–100% in 60 min) in 0.1% trifluoroacetic acid and a flow rate of 40 lLÆmin)1. The eluted protein peak was collected in two fractions, which were analyzed by 12% SDS ⁄ PAGE (insert). Gel lanes: 1 and 2, equal samples from the first half and second half of the peak, respectively; MW, molecular mass markers. FEBS Journal 273 (2006) 4322–4335 ª 2006 The Authors Journal compilation ª 2006 FEBS 4325 Secreted tyrosinase from Trichoderma reesei E. Selinheimo et al. By MALDI-TOF MS, the reversed-phase purified protein gave a single peak corresponding to an average molecular mass of 43.3 kDa (not shown). Furthermore, in an electrospray MS analysis, using a quadrupole time-of-flight (Q-TOF) instrument, which has a considerably better resolution, the same protein preparation resulted in a set of masses among which 43 124.0 Da and 43 204.0 Da were dominant (not shown). The results thus indicate that the protein exists in different post-translationally modified forms. To further characterize the molecule, the reversed-phase purified protein was subjected to N-terminal sequencing by Edman degradation. No amino acid derivatives comparable to the amount of analyzed protein (200 pmol) could be obtained, suggesting that the protein has a blocked N-terminus. For further characterization, the protein was alkylated and fragmented by trypsin. The tryptic peptides obtained were first directly analyzed by MALDI-TOF peptide mass fingerprinting, where most of the obtained peptide masses could be correlated with theoretical tryptic peptide masses calculated from the deduced protein sequence. Notably, no tryptic peptide masses correlating with the C-terminal part after Lys394 (Fig. 1) of the encoded sequence could be found. The most N-terminal tryptic peptide found that correlated with the theoretical tryptic peptide map of the deduced protein was QNINDLAK (m ¼ 914.482 Da), indicating that the possible signal sequence cleavage site is located N-terminally to this peptide. Homology comparisons suggested that the signal sequence cleavage site could be at the A(18)–Q(19) bond in the deduced sequence (shown by an arrow in Fig. 1). Often, this kind of cleavage is followed by cyclization of the N-terminal glutamine to form pyroglutamic acid. The tryptic peptide mass fingerprint of TYR2 contained a peptide mass of 2136.108 Da, which was suggested to correspond to the N-terminal blocked tryptic peptide (< QGTTHIPVTGVPVSPGAAVPLR, m ¼ 2136.196 Da). The identity of this peptide was then confirmed by MALDI-TOF ⁄ TOF fragment ion analysis, where partial sequences of this peptide were obtained from the ladders of b-fragment and y-fragment ions. Subsequently, for specifying the C-terminus of the protein, the most C-terminal tryptic peptide, as compared with the theoretical tryptic peptide map of the deduced protein, was found to be SQAQIK (m ¼ 673.376 Da). The mass of the following theoretical tryptic peptide (SSVTTIINQLYGPNSGK, m ¼ 1777.927 Da) could not be found, indicating that the C-terminus of the protein is within this sequence. In the peptide mass fingerprint, a mass corresponding to the peptide SSVTTIINQLYGPNSG 4326 (m ¼ 1649.826 Da) was found. The identity of this C-terminal peptide was further confirmed by MALDI-TOF ⁄ TOF fragment ion analysis as well as by Edman degradation of the corresponding peptide after purification by reversed-phase chromatography. During the search for and confirmation of the N-terminus and C-terminus of the protein, many other peptides were also analyzed, and the results confirmed most of the remaining deduced amino acid sequence. Edman sequencing of purified tryptic peptides covered 39.1% of the sequence. The molecular masses of the tryptic peptides, either from the mass fingerprint or purified peptides, covered 60.3% of the sequence, and the masses of cyanogen bromide fragments, including the N-terminal and C-terminal ones, covered 94.9% of the sequence. Together with the results from the glycopeptide analysis (see below), the peptide analyses completely confirmed the deduced amino acid sequence of the secreted protein. The calculated average molecular mass of the polypeptide chain of TYR2 with the determined N-terminus and C-terminus is 41 862.7 Da, whereas the mass determined by MS is about 43 200 Da. Thus, there is a mass difference of about 1300 Da between the determined and calculated mass, due to post-translational modifications. The purified polypeptide chain contains one potential N-glycosylation site in the tryptic peptide SGPQWDLYVQA MYNMSK (m ¼ 2016.907 Da). In order to analyze the possible N-glycosylation, the protein was digested with trypsin and the potential glycopeptides were bound to a ConA column. MALDI-TOF MS analysis of the eluted material revealed a few peptides, of which the largest had a molecular mass of 3255.309 Da. This could correspond to the sodium adduct of the abovementioned tryptic peptide with a high-mannose-type glycan, (GlcNAc)2(Hex)5, attached to it. Further MALDI-TOF ⁄ TOF MS analysis of this peptide selected as the precursor ion (Fig. 4) revealed a ladder of b-ions corresponding to the suggested glycopeptide with a sequential loss of Na+, five hexoses, and two N-acetylglucosamines, respectively. The resulting protonated mass of 2017.0 Da fits well with the mass of the nonglycosylated peptide. Further downstream in the fragment ion spectrum, a b-ion ladder corresponding to the amino acid sequence LYVQAM was detected, which confirmed the identity of the peptide. Thus, the purified protein is N-glycosylated at the asparagine residue (N62, Fig. 1) having a high-mannose-type glycan consisting of two N-acetylglucosamines and five hexoses. From the ConA eluate, other masses corresponding to the same tryptic peptide but with a shorter glycan (e.g. peptide + 1 · GlcNAc, m ¼ 2220.12 Da) were also detected, which indicates that the presence of FEBS Journal 273 (2006) 4322–4335 ª 2006 The Authors Journal compilation ª 2006 FEBS E. Selinheimo et al. Secreted tyrosinase from Trichoderma reesei Y V Q A M Y- Δm GlcNAc = 203.20 Da Δm Hexose = 162.14 Da 8.58.02017.0 7.5- 3256.309 3235.0 (-Na+) -L GlcNAc Abs. Int. × 1000 7.06.56.0 5.55.04.5- 783.800 2.01.5- 670.931 1045.842 1244.723 945.792 1174.001 3257.0 2.5- Hex Hex Hex 3.0- Hex 3.5- Hex GlcNAc 4.0- 1539.592 1375.497 1.00.5600 800 1000 1200 1400 1600 1800 2000 2200 2400 2600 2800 3000 3200 m/z Fig. 4. The MALDI-TOF MS ⁄ MS spectrum of a ConA affinity-purified tryptic glycopeptide from T. reesei TYR2. The peptide with a determined monoisotopic protonated mass of 3256.309 Da (shown in the insert) was selected as the precursor ion and analyzed in the LID-LIFT mode without collision gas. The resulting fragment ions correspond to a sequential loss of one Na+, five hexoses and two N-acetylglucosamines, resulting in a molecule with a protonated mass of 2017.0 Da. The fragmentation ion ladder at the lower molecular mass range corresponds to a sequence LYVQAMY, which confirms the identity of the glycopeptide. other shorter glycan structures cannot be excluded. N-Glycans with five mannoses have been found as the predominant form previously in the Cel7A cellulase of T. reesei [30]. In the same study, N-glycosylation sites with a single GlcNAc were also found. In order to clarify the reason for the existence of TYR2 as a double band in SDS ⁄ PAGE, the two protein bands were individually cut out and ‘in-gel’ digested with trypsin. Mass fingerprint analysis of the tryptic fragments by MALDI-TOF MS did not show significant differences in the mass fingerprints, thus leaving the reason for the double band unclear. T. reesei TYR2 was shown to be almost fully active within a pH range of 6–9.5, with an optimum at pH 9. Considering the stability of T. reesei TYR2 within a pH range of 2–8, the enzyme showed good stability at neutral and alkaline pH. When the pH was under 7, the enzyme started to lose activity; after 1 h at pH 5, activity loss was 50%, and after 1 h at pH 4.0, the enzyme had totally lost its activity. Although l-tyrosine was chosen as the substrate to diminish the substrate auto-oxidation effect in the pH optimum and stability determination, the disturbance of auto-oxidation could not be totally eliminated, because l-tyrosine is first hydroxylated to diphenolic l-dopa and then further oxidized to quinones in tyrosinase-catalyzed reactions. An alkaline environment also changes the redox potential of the phenolic substrates, making them more easily oxidized. Therefore, the pH profile reflects not only the optimal behavior of the enzyme, but also changes in the substrate. With regard to temperature stability, T. reesei TYR2 was found to be stable up to 30 C. However, at higher temperatures it started to lose its activity relatively quickly; the enzyme showed half-lives of 18 h, 3 h 45 min and 15 min at 30 C, 40 C and 50 C, respectively. Among the tested substrates (Table 2), the highest affinity of T. reesei TYR2 was observed with p-tyrosol (Km ¼ 1.3 mm), followed by p-coumaric acid (Km ¼ 1.6 mm) and l-dopa (Km ¼ 3.0 mm). The highest turnover number, kcat, was observed with l-dopa, at 22 s)1. Substrate specificity determination for T. reesei TYR2 showed that the enzyme was able to oxidize various substituted monophenols, which had the OH group in the para position (Table 3). The activity of the enzyme on diphenols was substantially higher than on monophenols; for example, catechol was oxidized approximately 10 times faster than phenol. Interestingly, aniline, containing no hydroxyl groups in the aromatic ring, but an amino group, was also oxidized by the tyrosinase, although slowly. Any side chain ortho to the phenolic hydroxyl group prevented FEBS Journal 273 (2006) 4322–4335 ª 2006 The Authors Journal compilation ª 2006 FEBS 4327 Secreted tyrosinase from Trichoderma reesei E. Selinheimo et al. Table 2. Determination of Km and kcat values for T. reesei TYR2 on L-dopa, p-coumaric acid and p-tyrosol. Substrate Km (mM) kcat (s)1) kcat ⁄ Km (s)1ÆmM)1) L-Dopa 3.0 1.6 1.3 22 8 7 7 5 6 p-Coumaric acid p-Tyrosol Table 4. Stereospecificity of T. reesei TYR2. Substrate (2.5 mM) Activity (%) relative to L-dopa and L-tyrosine L-Dopa 100 46 18 100 40 7 DL-Dopa D-Dopa L-Tyrosine DL-Tyrosine D-Tyrosine Table 3. Substrate specificity of T. reesei TYR2 as determined relative to L-dopa. ND, not determined due to the low solubility. Relative activity (%) on monophenols and polyphenols from oxygen consumption (nmolÆL)1Æs)1) was calculated according to the stoichiometry that one monophenol molecule needs one oxygen molecule, and one polyphenol molecule needs half an oxygen molecule, in the reaction to form a quinone. c, substrate concentration. Activity (%) relative to L-dopa Substrate c ¼ 2.5 mM c ¼ 10 mM L-Dopa 100 11 8 0 12 1 0 3 23 25 0 0 1 96 142 87 66 100 ND 8 0 8 ND 0 3 16 16 0 0 0 89 73 72 52 L-Tyrosine Phenol 4-Mercaptophenol p-Cresol 4-Aminophenol 3-Hydroxyanthranilic acid Tyramine p-Tyrosol p-Coumaric acid o-Coumaric acid Ferulic acid Aniline (–)-Epicatechin (+)-Catechin hydrate Pyrocatechol Pyrogallol Table 5. Degree of inhibition of T. reesei TYR2 as determined in the presence of 15 mM L-dopa, as analyzed by oxygen consumption and spectrophotometric assay. oxidation of the substrate, presumably because of steric hindrance. The presence and the position of an amine group in the substrate structure appeared to be critical, considering the oxidation of the substrate by T. reesei TYR2. The closer to the hydroxyl group of phenol the amino group was, the slower was the oxidation of the substrate. For most of the substrates studied, increasing the substrate concentration from 2.5 mm to 10 mm (or to 20 mm, data not shown) did not substantially affect the activity as calculated relative (%) to l-dopa (Table 3). However, different stereo-forms of catechin behaved differently. As the concentration of (+)-catechin and (–)-catechin was increased from 2.5 to 10 mm, (–)-catechin was oxidized faster, suggesting that T. reesei TYR2 has a lower Km value for (–)-catechin than for (+)-catechin. Furthermore, T. reesei TYR2 was found to be stereospecific; it 4328 oxidized the l-forms of dopa and tyrosine noticeably better than the d-forms (Table 4). Various potential inhibitors of T. reesei TYR2 were tested (Table 5). Kojic acid and b-mercaptoethanol were the most effective inhibitors, even at low concentrations. Sodium chloride and EDTA did not inhibit the enzyme very efficiently. Glutathione caused only moderate inhibition, inhibiting the enzyme with 20% efficiency, as measured with the oxygen consumption assay. However, as measured with the spectrophotometric assay, inhibition efficiency was 100%, suggesting that glutathione does not inhibit the enzymatic reaction, but has more effect on the subsequent nonenzymatic reactions, as also reported in other studies [31]. T. reesei TYR2 was able to oxidize the tested model peptides glycine–tyrosine and glycine–glycine–tyrosine (Table 6). The oxidation rate was dependent on the Inhibitor Sodium azide Kojic acid b-Mercaptoethanol SDS Benzaldehyde Glutathione NaCl EDTA Inhibitor (mM) Degree (%) of inhibition as analyzed by oxygen consumption assay Degree (%) of inhibition as analyzed by spectrophotometric assay 10 1 10 1 10 1 10 1 10 1 10 1 100 10 10 1 50 39 99 95 100 100 68 39 64 30 18 17 15 1 39 11 91 75 100 98 100 100 73 44 42 13 100 100 49 0 13 7 FEBS Journal 273 (2006) 4322–4335 ª 2006 The Authors Journal compilation ª 2006 FEBS E. Selinheimo et al. Secreted tyrosinase from Trichoderma reesei Table 6. Activity of T. reesei TYR2 on 2.5 mM dipeptides and tripeptides in relation to L-tyrosine (%). Y, tyrosine; G, glycine. Substrate Activity (%) relative to Y Y GY GGY 100 292 335 length of the peptide, the tripeptide being more readily oxidized than the dipeptide. Discussion Tyrosinase enzymes and their genes have previously been characterized from bacteria, fungi, plants and mammals. The most extensively investigated fungal tyrosinases, from both a structural and a functional point of view, are from Agaricus bisporus [10] and N. crassa [1]. Also, a few bacterial tyrosinases have been reported, of which Streptomyces tyrosinases are the most thoroughly characterized [32,33]. In addition, tyrosinases have been reported, for example, from Pseudomonadacae [34], Bacillus, Myrothecium [35], Mucor [36], Miriococcum [37], Aspergillus, Chaetotomastia, Ascovaginospora [38], Trametes [39] and Pycnoporus [40]. Our aim was to discover novel fungal tyrosinases, and we used the genome sequence of a well-known industrial enzyme producer T. reesei for the search. A homology search in the genome database of this fungus revealed a new tyrosinase gene tyr2, which, according to sequence analysis, has a signal sequence. The gene was overexpressed in the native host; thus, the gene product was verified to be secreted. This is exceptional in this class of enzymes, as all the plant, animal and fungal tyrosinases studied thus far have been intracellular. The characterized Streptomyces tyrosinases are secreted but do not have signal sequences; their secretion is assisted by a second protein that has a signal sequence [41,42]. It appears that other ascomycetous fungi also have secreted tyrosinases, because the three closest homologs of T. reesei TYR2 from G. zeae, N. crassa and Magnaporthe grisea have putative signal sequences. In fact, N. crassa has both secreted and intracellular tyrosinases; the enzyme cloned and studied previously is intracellular [26]. From an industrial point of view, a naturally secreted tyrosinase can be considered beneficial, as such an enzyme is likely to be compatible with the secretory system of the host organism in attempts to produce substantial amounts of enzyme for applications. Although microbial tyrosinases have been produced heterologously, e.g. in Eschericia coli [32,43,44] and in Saccharomyces cerevisiae [45], the expression levels reported thus far have been relatively low. The availability of the enzyme has hampered its detailed characterization as well as testing it in various applications. The T. reesei TYR2 tyrosinase gene was expressed in T. reesei under the strong cbh1 promoter. In shake flasks, the highest production level was approximately 320 mgÆL)1, whereas production levels were over three times higher than this in fermenter conditions. The addition of copper to the T. reesei medium had a positive effect on tyrosinase production. Because the tyrosinase was expressed under the cbh1 promoter, which is not activated by copper, the improved production levels were presumably not caused by higher transcription rates. In addition, no effect of copper addition on fungal growth was observed, which implies that the higher enzyme yields may have been due to improved folding of the active enzyme in the presence of elevated copper concentrations. The importance of high copper concentrations has been reported in laccase production in S. cerevisiae, where the overexpression of two copper-trafficking enzymes from Trametes versicolor led to significantly improved recombinant laccase yields [46]. Added copper can improve correct folding of recombinant laccase, as previously detected in Aspergillus nidulans and Aspergillus niger expressing a laccase from Ceriporiopsis subvermispora [47], and in T. reesei producing the laccase of Melanocarpus albomyces [48]. C-terminal processing of fungal tyrosinases has been reported previously, and also the molecular mass of purified TYR2 tyrosinase, 43.2 kDa, suggested extensive processing of the protein. The intracellular tyrosinases from N. crassa [26–28] and Agaricus [28] and Pycnoporus species [25] have an additional C-terminal domain that is proteolytically released from the catalytic domain. It has been postulated that the function of the C-terminal domain is to keep the enzyme inactive until the activity is needed [26]. According to our results, the secreted T. reesei TYR2 is also C-terminally processed (after Gly410) (Fig. 1). However, in this case the peptidase performing the cleavage must reside in the secretory pathway or be extracellular. The precise processing site has previously been determined only for the N. crassa tyrosinase (after Phe408) (Fig. 1). According to the alignment of T. reesei TYR2 and the N. crassa tyrosinase, the positions of the processing sites in these two enzymes coincide exactly, even though the sequences are not conserved in that region. This is compatible with the idea that this site is at a domain border that would be susceptible to proteases. The processed N. crassa tyrosinase ends with a phenylalanine, and thus it was assumed that it is FEBS Journal 273 (2006) 4322–4335 ª 2006 The Authors Journal compilation ª 2006 FEBS 4329 Secreted tyrosinase from Trichoderma reesei E. Selinheimo et al. cleaved by a chymotrypsin-like enzyme [26]. The Agaricus bisporus tyrosinase can be processed in vitro by the serine proteases trypsin and subtilisin [28]. The processed T. reesei TYR2 ends with a glycine residue. Analysis of the whole sequence with the program peptidecutter [49], which searches for all known peptidase cleavage sites, did not indicate that the protein could be cleaved at that site. The C-terminal glycine of the mature TYR2 is followed in the sequence by the amino acids Lys–Lys–Arg. This contains a recognition sequence for the KEX2 ⁄ furin-type protease, which resides in the Golgi complex and processes a number of secreted enzymes and other proteins after dibasic recognition sites [50]. The putatively secreted tyrosinase of G. zeae has Lys–Arg at the same position (Fig. 1). For these reasons, it is possible that TYR2 is first cleaved by a T. reesei KEX2-type endopeptidase during secretion and is further processed by an exopeptidase. Further analyses are needed to elucidate the role of the C-terminal processing. As for the T. reesei TYR2, the pH optimum in the alkaline pH range has been reported for Thermomicrobium roseum (pH 9.5) [51] and pine needle tyrosinase (9–9.5) [52]. Many fungal tyrosinases have their pH optima at neutral and slightly acidic pH, e.g. N. crassa and Aspergillus flavipes at pH 6.0–7.0 [53,54] and Pycnoporus sanguineus at pH 6.5–7 [40]. T. reesei TYR2 was not able to retain substantial activity at temperatures above 30 C. Longer half-lives have been reported, e.g. 2 h at 50 C for P. sanguineus tyrosinase. However, at 60 C, P. sanguineus tyrosinase was also inactivated completely within 20 min [40]. In general, mammalian and plant-derived tyrosinases are not very thermostable; even a short incubation at 70–90 C inactivates the enzymes completely [52,55]. Also, inactivation of A. flavipes [54] and N. crassa [56,57] tyrosinases at relatively low temperatures has been reported. The enzyme showed relatively high Km values for all tested substrates, l-dopa, p-coumaric acid, and p-tyrosol. The values were in accordance with values reported in the literature. Km values for l-dopa were 3.0 mm for T. reesei TYR2, 0.74–1.09 mm for N. crassa [1,58], 5.0 mm for A. flavipes [54], 5.97 mm for Streptomyces glaucescens [59] and 8.7–10 mm for pine needle [52]. Trichoderma reesei TYR2 showed surprisingly broad substrate specificity and higher oxidation activity for diphenols than for monophenolic substrates. Ferulic acid, as well as other compounds with a side group ortho to the phenolic hydroxyl group, was not oxidized by the enzyme, presumably because of steric hindrance. The substituted phenols, such as 2-aminophenol and 4-nitrophenol, or benzene derivatives, such as benzoic and naphthoic acids, have been reported to be 4330 efficient tyrosinase inhibitors, and the inhibitory mechanism is suggested to be competitive docking, due to the similarity between the structures of these inhibitory compounds and those of phenol or tyrosine [60,61]. For instance, Piquemal et al. [61] showed in a theoretical study that 2-aminophenol forms a more stable and energetically favored complex with tyrosinase than phenol does. The amine group also seemed to act as a substrate analog for T. reesei TYR2. Also, a thiol group in the phenolic ring inhibited the enzyme. Toussaint and Lerch [62] and Ga˛sowska et al. [63] showed that N. crassa tyrosinase oxidizes aromatic amines and o-aminophenols, structural analogs of monophenols and ortho-diphenols. Similar catalytic reactions, ortho hydroxylation and oxidation, took place, although the reaction rates observed for aromatic amines were relatively slow as compared to those for monophenols. T. reesei TYR2 was also found to oxidize phenylalanine, although extremely slowly. The tyrosyl residue was oxidized by T. reesei TYR2 in the dipeptide glycine–tyrosine and the tripeptide glycine–glycine–tyrosine. The relative oxidation rate increased as the length of the peptide increased. Similarly, protein-bound tyrosyl was oxidized by the enzyme, and subsequent protein crosslinking was observed, as analyzed by SDS ⁄ PAGE (data not shown). Because of difficulties in the production and purification of microbial tyrosinases in sufficient amounts, knowledge of their structure–function relationships and exact reaction mechanisms is still limited. The availability of the enzyme has also hampered its testing and use in applications. We have reported here for the first time the production, purification and characterization of a novel tyrosinase from the well-known protein producer T. reesei. The high production levels of the tyrosinase also allow the testing of the enzyme for applications. Experimental procedures Isolation of the tyrosinase gene from Trichoderma reesei The tyr2 gene was amplified from genomic T. reesei DNA with the following primers: forward, GGG GAC AAG TTT GTA CAA AAA AGC AGG CTA TCA TGC TGT TGT CAG GTC CCT CTC G; and reverse, GGG GAC CAC TTT GTA CAA GAA AGC TGG GTC AGT GGT GGT GGT GGT GGT GCA GAG GAG GGA TAT GGG GAA CGG CAA A. The PCR reaction was done with the Dynazyme EXT thermostable polymerase (Finnzymes, Helsinki, Finland) in a reaction mixture recommended by the manufacturer. The PCR program comprised an initial denatura- FEBS Journal 273 (2006) 4322–4335 ª 2006 The Authors Journal compilation ª 2006 FEBS E. Selinheimo et al. tion step of 3 min at 94 C, followed by 25 cycles of 30 s at 94 C, 45 s at 52 C and 2.5 min at 72 C. This was followed by a final elongation step of 5 min at 72 C. The tyr1 gene fragment was cloned into the pCR2.1TOPO vector with the TOPO-TA Cloning Kit (Invitrogen, Carlsbad, CA, USA) and subsequently transferred into the pDONR221 vector (Invitrogen) with a BP recombination reaction carried out with the Gateway Recombination kit (Invitrogen). The tyr2 cDNA was isolated by RT-PCR from a cDNA expression library of T. reesei RutC-30 [18] with primers that were designed to create an N-terminal His6 tag and add EcoRI and KpnI restriction endonuclease sites to the 5¢ and 3¢ ends, respectively. The primers used were as follows: forward primer, GTT GGA ATT CCA TCA TCA TCA TCA TCA TCA GGG CAC GAC ACA CAT CCC C; and reverse primer, GAT CGG TAC CTC ATT ACA GAG GAG GGA TAT GGG GAA C. The PCR reaction was done as described above. The amplified PCR product was inserted into the EcoRI and KpnI sites of the vector pPICZáA (Invitrogen) and the sequence of the product was verified. Overexpression of the tyrosinase gene in Trichoderma reesei The genomic tyr2 gene fragment was transferred by an LR recombination reaction from the pDONR221 vector to the T. reesei expression vector pMS186, giving rise to the plasmid pMS190. The pMS186 contains the Gateway reading frame cassette C (RfC) inserted between the cbh1 (cellobiohydrolase 1) promoter and terminator, and a hygromycin resistance cassette. The LR recombination reaction was done with the Gateway Recombination kit (Invitrogen) according to the manufacturer’s instructions. The plasmid pMS190 was transformed into the T. reesei strain VTT-D-00775, essentially as described [19], and transformants were selected for hygromycin resistance on plates containing 125 lgÆmL)1 of hygromycin B. The transformants were streaked on the selective medium for three successive rounds and tested for tyrosinase activity with a plate assay. In the assay plates, Trichoderma minimal medium [19] with 2% lactose as a carbon source, 1% potassium phthalate as a buffering agent (pH 5.5), 0.1 mm CuSO4 and 1% tyrosine as an indicator substrate was used. The transformants were streaked on the plates and grown for 7 days, and tyrosinase activity was observed on the plates as a brown color appearing around the streaks. Positive transformants were isolated by single-spore cultures. In order to quantify tyrosinase production in liquid cultures, the transformants positive in the plate assay were grown in shake flasks for 8 days in 50 mL of Trichoderma minimal medium [19] supplemented with 4% lactose, 2% spent grain, 100 mm piperazine-N-N¢-bis(3-propanesulfonic acid) and 0.1–2 mm CuSO4, and tyrosinase activity was measured Secreted tyrosinase from Trichoderma reesei with 3,4-dihydroxy-l-phenylalanine (l-dopa) substrate as described below. Trichoderma reesei was cultivated in a Braun Biostat C-DCU 3 fermenter (B. Braun Biotech International, GmbH, Melsungen, Germany) in 20 L of a medium containing (gÆL)1): lactose (20), distiller’s spent grain (10), and KH2PO4 (15), and 2 mm CuSO4.5H2O. The medium pH was adjusted to 5.5–6 with NH4OH and H3PO4, and the cultivation temperature was 28 C. The dissolved oxygen level was kept above 30% with agitation at 450 r.p.m., aeration at 8 LÆmin)1 and 0–30% O2 enrichment of incoming air. Foaming was controlled by automatic addition of Struktol J633 polyoleate antifoam agent (Schill & Seilacher, Hamburg, Germany). After fermentation, cells were harvested by centrifugation and the culture supernatant was concentrated 2.5 times by ultrafiltration. Protein and enzyme activity assays Tyrosinase activity was measured according to Robb [2] with a few modifications, using 15 mm l-dopa and 2 mm l-tyrosine as substrates. Activity assays were carried out in 0.1 m sodium phosphate buffer (pH 7.0) at 25 C, monitoring dopachrome formation at 475 nm (edopachrome ¼ 3400 m)1Æcm)1). Tyrosinase activity was also determined by following the consumption of the cosubstrate oxygen with a single-channel oxygen meter (Precision Sensing GmbH, Regensburg, Germany). The activity was determined by measuring the oxygen consumption during the reaction in a sealed and a fully filled sample vial (1.8 mL) at 25 C. The reaction was initiated by addition of the enzyme to the substrate solution, and the oxidation rate (nmolÆL)1Æs)1) was calculated from the linear part of the oxygen consumption curve. The protein concentration was determined with the Bio-Rad DC protein assay kit (Bio-Rad, Richmond, CA, USA), with BSA as standard. During enzyme purification, to estimate protein contents for pooling fractions, protein content determinations were done by monitoring absorbance at 280 nm. Enzyme purification The concentrated culture supernatant was first desalted on a Sephadex G-25 Coarse column (2.6 · 27 cm; Pharmacia Biotech, Uppsala, Sweden) in 10 mm Tris ⁄ HCl buffer, pH 7.3. The subsequent purification steps were carried out with an ÄKTApurifier (Amersham Biosciences, Uppsala, Sweden). The sample was applied to a HiPreptm 16 ⁄ 10 CM Sepharose Fast Flow column, in 10 mm Tris ⁄ HCl buffer, pH 7.3. Bound proteins were eluted with a linear NaCl gradient (0–180 mm in six column volumes) in the equilibration buffer. Tyrosinase-positive fractions were pooled, concentrated with a Vivaspin concentrator (20 mL, 10 000 molecular weight cut-off; Vivascience, Hannover, Germany), and subjected to gel filtration in a Sephacryl S-100 FEBS Journal 273 (2006) 4322–4335 ª 2006 The Authors Journal compilation ª 2006 FEBS 4331