Structural Biology Communications Crystallization and Preliminary X-ray Crystallographic Analysis of Polyphenol Oxidase from Juglans Regia (jrppo1)

Tyrosinase is a type 3 copper enzyme that catalyzes the ortho-hydroxylation of monophenols to diphenols as well as their subsequent oxidation to quinones, which are precursors for the biosynthesis of melanins. The first plant tyrosinase from walnut leaves (Juglans regia) was purified to homogeneity and crystallized. During the purification, two forms of the enzyme differing only in their C-termini [jrPPO1(Asp 101 –Pro 444) and jrPPO1(Asp 101 –Arg 445)] were obtained. The most abundant form jrPPO1(Asp 101 –Arg 445), as described in Zekiri et al. , and diffracted to 2.39 A ˚ resolution. Crystals were only obtained from solutions containing at least 30% polyethylene glycol 5000 monomethyl ether in a close-to-neutral pH range.

To date, only crystal structures of bacterial and fungal tyrosinases have been reported in the literature. The first crystal structure of a tyrosinase was obtained from the bacterium Streptomyces castaneoglobisporus and was published in 2006 (Matoba et al., 2006), followed by the crystal structure of the bacterial tyrosinase from Bacillus megaterium (Sendovski et al., 2011). Two crystal structures of fungal tyrosinase from the mushroom Agaricus bisporus, abPPO3 (UniProt C7FF04) and abPPO4 (UniProt C7FF05) (Ismaya et al., 2011;Mauracher, Molitor, Al-Oweini et al., 2014), have recently been published. To date, only two crystal structures of catechol oxidases have been reported: the 39 kDa isoform of the Ipomoea batatas (sweet potato) catechol oxidase (UniProt Q9ZP19; Klabunde et al., 1998) and the catechol oxidase from the grape Vitis vinifera (UniProt P4331; Virador et al., 2010).
The active site of tyrosinase consists of two copper ions, each coordinated by three conserved histidine residues. It has been proposed that tyrosinases are involved in defence mechanisms against pathogens because of the bacteriostatic properties of oquinones and melanins. The study of tyrosinases in plants has focused on their role in the process of post-harvest browning, whereby cut or damaged plant tissues turn brown owing to the polymerization of tyrosinase-derived quinones (Queiroz et al., 2008). Walnut presents an interesting model to further explore the function of tyrosinase in plants based on the high content of various phenolic compounds in walnut tissues (Colaric et al., 2005;Solar et al., 2006;Araji et al., 2014).

Sample preparation
The isolation and purification process of the enzyme has been described in detail by Zekiri et al. (2014). Briefly, extraction of the enzyme from the natural source (walnut leaves) was performed using the method developed by Mauracher, Molitor, Michael et al. (2014) based on detergent and soluble polymer polyethylene glycol (PEG) phase separations. The active tyrosinase was purified to homogeneity by fast protein liquid chromatography (FPLC) using several ionexchange columns. During the purification, two forms of the enzyme differing only in their C-termini [jrPPO1(Asp 101 -Pro 444 ) and jrPPO1(Asp 101 -Arg 445 )] were obtained and were determined by peptide sequencing applying nanoUHPLC-ESI-MS/MS. The purity of jrPPO1 was monitored by SDS-PAGE. For crystallization experiments jrPPO1(Asp 101 -Arg 445 ), the most abundant form, was concentrated to 10 mg ml À1 in 20 mM HEPES pH 7.5.

Protein crystallization
Initial screening for crystallization was carried out by the sittingdrop vapour-diffusion method (96-well CrystalQuick plates, Greiner Bio-One) employing a nanodispenser robot (Gryphon, Art Robbins). Initial hits were obtained by screening over a broad variety of commercially available screening kits (JBScreen Classic 1-10 and JBScreen Membrane 1-3 from Jena Bioscience) mixing the protein solution with the reservoir solution in a 1:2 ratio. Further optimization of the crystallization conditions was performed manually in 15well EasyXtal plates (Qiagen) applying the hanging-drop vapourdiffusion method at 293 K. Single crystals suitable for diffraction measurements were obtained after 2-3 d by mixing 1 ml protein solution (10 mg ml À1 ) with 0.5 ml reservoir solution and equilibrating against 500 ml reservoir solution consisting of 30% PEG 5000 MME (MME, monomethyl ether), 200 mM ammonium sulfate, 100 mM MES pH 6.5 (Fig. 1). The crystals stopped growing after approximately 6 d.

Data collection and processing
Single crystals were harvested by transferring them with a cryoloop (10 mm, 0.1 Â 0.2 mm; Hampton Research) into a 0.5 ml drop of cryoprotectant solution (30% PEG 5000 MME, 20% glycerol, 200 mM ammonium sulfate, 100 mM MES pH 6.5) before flashcooling them in liquid nitrogen. X-ray diffraction measurements of about 20 crystals of suitable size were carried out at DESY (Hamburg, Germany) on the monochromatic (1.033 Å ) beamline P11. Diffraction data were collected with a PILATUS 6M detector at 100 K with an oscillation range of 0.2 and an exposure time of 0.190 s. The best crystal diffracted to 2.39 Å resolution using a crystalto-detector distance of 435 mm. Data processing was carried out using the XDS program package (Kabsch, 2010). The space group was determined using the program POINTLESS from the CCP4 suite (v.6.4.0; Winn et al., 2011). The data set presented here had strong anisotropy as indicated by phenix.xtriage from the PHENIX suite (v.1.8.4; Adams et al., 2010) and was therefore truncated (discarding reflections falling outside a specific 'ellipsoid' with dimensions 1/2.4, 1/2.4 and 1/2.4 Å along a*, b* and c*, respectively) and anisotropically scaled using the Diffraction Anisotropy Server at UCLA MBI (Strong et al., 2006).

Results and discussion
By applying the method described by Zekiri et al. (2014), the enzyme was purified to homogeneity. Initial attempts to crystallize the most abundant form jrPPO1(Asp 101 -Arg 445 ), as described in Zekiri et al. (2014), covering a wide range of crystallization conditions proved to be very successful. Crystals suitable for X-ray diffraction experiments were obtained using 30% PEG 5000 MME, 200 mM ammonium sulfate, 100 mM MES pH 6.5 (Fig. 1). The crystals were plate shaped and of reasonable size (100 Â 50 Â 10 mm) for X-ray diffraction analysis. Crystallization of jrPPO1 has been achieved by conditions that differ in almost all crystallization parameters (pH, precipitation agent, temperature or additives) from those reported previously for PPO crystallization. The utilization of PEG as a precipitation agent is the only constant. All of the published crystallization conditions of fungal (Ismaya et al., 2011;Mauracher, Molitor, Al-Oweini et al., 2014) and bacterial tyrosinases (Matoba et al., 2006;Sendovski et al., 2011) as well as plant catechol oxidases (Klabunde et al., 1998;Virador et al., 2010) contain PEG as precipitation agent with different molecular masses (PEG 4000-8000). Plate-shaped crystals obtained using 30% PEG 5000 MME, 200 mM ammonium sulfate, 100 mM MES pH 6.5.
Processing statistics for the X-ray diffraction measurements are presented in Table 1. The crystals belonged to space group C121, with unit-cell parameters a = 115. 56, b = 91.90, c = 86.87 Å , = 90, = 130.186, = 90 , and diffracted to a maximum resolution of 2.39 Å . The solvent content was determined with the Matthews formula using a molecular mass of 39.047 kDa (Zekiri et al., 2014). This gives a Matthews coefficient (Matthews, 1968) of 2.26 Å 3 Da À1 and a solvent content of 45.52% assuming the presence of two monomers per asymmetric unit.
We are currently attempting to solve the crystal structure by using molecular replacement (MR). Two models for MR are available: V. vinifera catechol oxidase (sequence identity 62%; UniProt P4331; Virador et al., 2010) and I. batatas catechol oxidase (sequence identity 56%; UniProt Q9ZP19; Klabunde et al., 1998). The data set was anisotropic and was therefore truncated and anisotropically scaled. However, both the truncated and original data sets will be used in refinement to determine the best produced electron-density map, because phenix.refine carries out anisotropic scaling by default; thus, the truncated and pre-scaled data could be redundant, leading to a featureless map. This work was supported by a grant from the 'Fonds zur Fö rderung der wissenschaftlichen Forschung' (FWF; grant No. P25217-N28). We thank Dr Saravanan Panneerselvam, beamline scientist at P11 (DESY, Hamburg, Germany), for his generous support during data collection (proposal No. I-20130227 EC). The research leading to these results has received funding from the European Community's Seventh Framework Programme (FP7/2012-2014) under BioStruct-X (grant agreement No. 283570).