Recombinant expression, purification, and characterization of polyphenol oxidase 2 (VvPPO2) from “Shine Muscat” (Vitis labruscana Bailey × Vitis vinifera L.)

Abbreviations: PPO, polyphenol oxidase; 4-MC, 4-methylcatechol

Abstract

Polyphenol oxidases (PPOs) catalyze browning reactions in various plant organs, therefore controlling the reactions is important for the food industry. PPOs have been assumed to be involved in skin browning of white grape cultivars; however, the molecular mechanism underlying PPO-mediated browning process remains elusive. We have recently identified a new PPO gene named VvPPO2 from “Shine Muscat” (Vitis labruscana Bailey × V. vinifera L.), and have shown that the gene is transcribed at a higher level than the previously identified VvPPO1 in browning, physiologically disordered berry skins at the maturation stage. In this study, we expressed VvPPO2 in Escherichia coli and, using the purified preparation, revealed unique physicochemical characteristics of the enzyme. Our study opens up a way to not only understand the berry skin browning process but also to elucidate the enzymatic maturation process of grape PPOs.

Characterization of Polyphenol Oxidase 2 (VvPPO2) from ‘Shine Muscat’ (Vitis labruscana Bailey × Vitis vinifera L.)

Key words:

• polyphenol oxidase
• Vitis labruscana Bailey × Vitis vinifera L
• berry skin browning

Polyphenol oxidase (PPO: EC 1.10.3.1) is a dicopper enzyme that has tyrosinase/monophenolase and/or catecholase/diphenolase activities, and is widespread among plants, fungi, and bacteria.^1,2) Using molecular oxygen, the enzyme catalyzes hydroxylation of monophenol at the ortho-position and oxidization of ortho-diphenols to produce reactive ortho-diquinones.^3–6) PPO is considered to be involved in the plant defense system because the expression is up-regulated by herbivore-, pathogen-, and wound-induced cellular damage.^2,7) The reactive ortho-quinones produced by PPO are non-enzymatically polymerized to produce brown pigments that accumulate in intracellular space, thereby physically constraining further spread of the cellular damage.⁸⁾ PPO is localized mainly in chloroplasts and plastids of photosynthetic tissues,⁹⁾ while polyphenols are accumulated in the vacuoles. Consequently, the browning reaction triggered by PPO is spatially prevented in normal cells. Contact of PPOs with polyphenols can occur when physical damage or senescence occurs in cells.^9,10) Although the PPO-mediated reaction is considered to be important for the plant defense system, the browning significantly decreases the storage life and commercial value of plant products.¹⁰⁾

PPOs localized in chloroplasts are known to exist in a latent form.¹¹⁾ The latent enzyme consists of a catalytic domain followed by a C-terminal domain that shields the catalytic pocket of the enzyme. PPOs purified from plant sources generally possess the catalytic domain only, and therefore it is assumed that the C-terminal shielding domain is proteolytically removed upon introduction of cellular damages. This maturation process of PPOs has also been a target of study. However, because recombinant expression of plant PPOs was not successful until relatively recently, it was difficult to address the mechanism underlying its maturation process. Plant PPOs were known to easily form an insoluble aggregate when expressed in bacteria, and even if soluble enzyme was obtained, the preparation showed marginal activity.⁴⁾ In 2013, Dirks-Hofmeister et al. succeeded in expressing latent form of plant (dandelion) PPOs using Escherichia coli expression system by adding copper ion in the medium and culturing the recombinant strain at low temperature.⁴⁾ Subsequently, using the purified enzymes, they showed the recombinant PPOs can represent the characteristics of native enzymes purified from plant sources.

“Shine Muscat” is one of the yellow–green skin grape cultivars bred in Japan, and the cultivation area has drastically increased recently to reflect market demand. However, during cultivation, berry skin browning, which we call “Kasuri-shou” (Fig. 1), frequently arises in this cultivar. “Kasuri-shou” usually occurs at the berry maturation stage around 70‒80 days after full bloom. Because this physiological disorder considerably decreases its market value, technical developments to prevent the skin browning are strongly desired. In the previous study, we reported the identification of a new gene, Vitis vinifera PPO2, as a paralog of VvPPO1. Expression analysis revealed that VvPPO2, not VvPPO1, transcription is up-regulated in damaged fruits, which strongly suggests the involvement of VvPPO2 in the skin browning process.¹²⁾ VvPPO1 is identical to GPO1 from “Sultana” grape isolated by Dry and Robinson.¹³⁾ GPO1 has been purified from berries of several grape cultivars and characterized enzymatically and structurally^13–15); however, there have been no reports on the purification and characterization of VvPPO2. Here, we describe recombinant expression, purification, and characterization of VvPPO2. This is the first report describing the characterization of VvPPO2 and also the first to demonstrate active preparation of recombinant PPO from grape cultivars. These results contribute to deciphering the mechanism underlying the berry skin browning and also to elucidating the catalytic reaction and maturation process of grape PPOs.

Fig. 1. Appearance of “Shine Muscat” berries at harvesting time. (A) Normal matured berries without skin browning. (B) Matured berries with skin browning. Bar = 10 mm.

Material and methods

Phylogenetic analyses of PPOs

Amino acid sequences of characterized and predicted plant PPO genes (35 genes from 20 different plant species) were obtained from public databases including DNA Databank of Japan, European Molecular Biology Laboratory Bank, Genbank, and Grape Genome Browser (http://www.cns.fr/externe/GenomeBrowser/Vitis/). The sequences were aligned using the ClustalW program and the phylogenetic tree was created by the neighbor-joining method using MEGA software (ver. 7.0) with 1000 bootstrap replicates.¹⁶⁾

Heterologous expression of VvPPO2

VvPPO2 has a predicted chloroplast transit peptide at the N-terminal end (amino acid residues 1–98), which is deduced by comparing with VvPPO1 from “Shine Muscat”, GPO1 from “Sultana”, and PPO from “Grenache”.^12–14) The VvPPO2 cDNA fragment with and without the transit peptide region were amplified by high-fidelity PCR (Phusion, New England Biolabs, Ipswich, MA, USA) from the full-length cDNA clone isolated in our previous study.¹²⁾ The primers used for the PCR were specialized for the Gateway system (Invitrogen, Carlsbad, CA, USA) and their sequences are shown in Supplemental Table S1. The amplified fragment was cloned into pDONR221 entry vector of the Gateway system according to the manufacturer’s instruction, and then recombined into pDEST17 vector to construct expression plasmids pDEST17_VvPPO2-Full and pDEST17_VvPPO2-ΔN, respectively. Plasmid pDEST17_VvPPO2-ΔN was designed to express VvPPO2 without the transit peptide, but with hexahistidine (His)-tag at the N-terminus of the protein, which can simplify the protein purification. The DNA sequence of VvPPO2 cDNA fragment inserted into the vector was confirmed by sequencing. The expression plasmid was used to transform Escherichia coli Rosetta 2 (DE3) pLysS (Novagen Inc., Madison, WI, USA). The recombinant strains were cultivated in LB medium supplemented with 50 μg/mL ampicillin, 34 μg/mL chloramphenicol and 1% glucose at 37 °C with shaking at 200 rpm until the OD_600 nm reached a value between 0.5 and 1. The culture was then used to inoculate AIM medium (LB medium supplemented with 0.5% glycerol, 0.05% glucose, 0.2% lactose, 25 mM Na₂HPO₄, 25 mM KH₂PO₄, 50 mM NH₄Cl, 5 mM Na₂SO₄, 2 mM MgSO₄, and 0.2 mM CuCl₂)⁴⁾ to give an OD_600 nm value of 0.02. After cultivation at 26 °C for 48 h under shaking at 120 rpm, the cells were harvested and stored at −80 °C until use.

Purification of VvPPO2-ΔN

The cells were thawed on ice, suspended in 50 mM sodium phosphate buffer (pH 8.0) containing 300 mM NaCl and 10 mM imidazole, and disrupted by sonication (Vibra-Cell ultrasonic liquid processor VC-505, Sonics & Materials, Inc., CT, USA) . After removal of cell debris by centrifugation, His-tagged VvPPO2-ΔN was first purified by an affinity chromatography using nickel-nitrilotriacetic acid-agarose (Qiagen, Hilden, Germany). Purification was carried out according to the manufacturer’s instruction. The enzyme was further purified by MonoQ 5/50 GL (GE Healthcare) column chromatography, in which the protein was eluted by a linear gradient of NaCl (0‒1 M in 10 mM Tris–HCl, pH 8.0). Purity of the enzyme was assessed by SDS-polyacrylamide gel electrophoresis using a 12% separating gel. The purified protein was concentrated using an Amicon Ultra 30 K (Millipore, MA). Protein concentration was determined using a Bradford protein assay kit (Bio-Rad). Bovine serum albumin was used as a standard.

Assays

Standard reaction mixture consisted of 50 mM sodium citrate buffer (pH 5.0), 0.35 mM SDS, 1 mM 4-methylcatechol (4-MC), and enzyme in a total volume of 200 μL. Reaction was initiated by adding the enzyme, and the mixture was incubated at 30 °C. The reaction was monitored by measuring absorbance at 405 nm at 3 s intervals for 15 s (Multiskan GO, ThermoFisher Scientific, Waltham, MA), in which linearity of the reaction rate was observed. The absorption coefficient of oxidized 4-MC at 405 nm was 1090 (M⁻¹∙cm⁻¹).⁴⁾ The kinetic parameters were calculated by curve-fitting the experimental data to a Michaelis–Menten equation using Kaleida Graph version 4.0 (Synergy Software).

The optimal pH was examined using 100 mM buffers of sodium citrate (pH 4.0–6.0) and potassium phosphate (pH 6.0–8.0) in the presence of SDS. To determine the optimal concentration of SDS, the detergent was varied from 0 to 2 mM. Stability of VvPPO2-ΔN was examined by measuring the residual activity following incubation at 25 °C or 37 °C in 20 mM Tris–HCl (pH 8.5) or sodium citrate (pH 5.0) buffer in the presence and absence of 0.35 mM SDS.

Substrate specificity was examined using mono- and diphenolic natural compounds. The reaction mixture containing 50 mM sodium citrate (pH 5.0), 0.35 mM SDS, 1 mM substrate, and the enzyme (3 to 140 μg/mL) was incubated at 30 °C. The reaction was stopped by adding trichloroacetic acid at the final concentration of 20%. A high performance liquid chromatography (HPLC) system equipped with a C18 Inertsil ODS-2 analytical column (250 × 4.6 mm, GL Sciences, Japan) was used for separating the substrate from the products. Specific activity was calculated by measuring the amounts of the substrates consumed in the reactions. The elution was performed under constant flow (1 mL/min) with a linear gradient between solvent A (0.1% (v/v) phosphoric acid) and solvent B (0.1% (v/v) phosphoric acid in acetonitrile) (A: 90% to 60% for 40 min), and the absorbance at 280 nm was monitored.

Molecular weight estimation

The native molecular mass of VvPPO2-ΔN was estimated by size exclusion chromatography using Superdex 200 10/300 GL (GE healthcare). Elution was carried out at different pHs, i.e. 20 mM Tris–HCl buffer (pH 8.5) containing 150 mM NaCl and 20 mM sodium citrate buffer (pH 5.0) containing 150 mM NaCl. SDS was added at the final concentration of 0.35 mM. Ferritin (440 kDa), aldolase (158 kDa), ovotransferrin (conalbumin) (75 kDa), and ovalbumin (43 kDa) were used as standard molecular size markers. 1,2-α-L-Fucosidase and galacto-N-biose/lacto-N-biose I-binding protein from Bifidobacterium bifidum and Bifidobacterium longum, respectively, were recombinantly expressed in E. coli, and purified as described previously.^17–19) These proteins were used for gel filtration analysis under the same conditions described above.

Results

Phylogenetic analysis of plant PPOs

A phylogenic analysis of several plant PPOs based on amino acid sequences showed that the grape PPOs are clustered closely except for VviPPO4 (Fig. 2(A)). PPO1 isolated from “Shine Muscat” (VvPPO1) is 99.7% identical to GPO1 from “Sultana”¹³⁾ and 99.5% to PPO from “Grenache”¹⁴⁾ (Fig. 2(A)). Identity between VvPPO1 and VvPPO2 from “Shine Muscat” was found to be 63%. VviPPO3, which was identified in the genomic analysis of Vitis vinifera, shares 95% and 64% amino acid identity with VvPPO1 and VvPPO2, respectively, and was found to be located in the same cluster with VvPPO1. In contrast, VviPPO4 shows less similarity with the other grape PPOs and is 41, 43, and 42% identical with VvPPO1, VvPPO2, and VviPPO3, respectively. In the tree, VviPPO4 is clustered with PPO13 from Populus trichocarpa, which are known to be localized in the vacuole.²⁰⁾ Based on the analysis using subcellular localization prediction tools such as TargetP 1.1^21,22) and WoLF PSORT,²³⁾ VvPPO1, VvPPO2 and VviPPO3 were predicted to be located in chloroplast by the transit peptides (Fig. 2(B)), while VviPPO4 was predicted to be located elsewhere other than chloroplast.²⁾ Other plant PPOs diverged widely. AmAS1, an aureusidin synthase from Antirrhinum majus,²⁴⁾ an aurone synthase (AUS1) from Coreopsis grandiflora,²⁵⁾ and LtLH, (+)-larreatricin hydroxylase from Larrea tridentate,²⁶⁾ are distantly related to the grape PPOs.

Fig. 2. Phylogenetic analysis of plant PPOs. (A) Phylogenetic tree of 35 selected plant-derived PPO genes created based on the amino acid sequence. The scale bar indicates the estimated number of amino acid substitutions per site. The bootstrap values of >50% are shown. Asterisks indicate predicted genes in the grape genome database.²⁾ The accession numbers and references for PPOs are as follows: GPO1 “Sultana”(Vitis vinifera, P43311),¹³⁾ PPO “Granache” (Vitis vinifera, P93622_VITVI),¹⁴⁾ VvPPO1 and VvPPO2 “Shine Muscat” (Vitis labrusca × Vitis vinifera, BAO79386, BAO79387),¹²⁾ VviPPO1, VviPPO2, VviPPO3 (Vitis vinifera, GSVIVP00011782001, GSVIVP00011776001, GSVIVP00011780001),²⁾ VviPPO4 (GSVIVP00036366001, XP_010645091.1),²⁾ CsPPO (Camellia sinensis, EF635860),³⁵⁾ LcPPO (Litchi chinensis, AEQ30073),³⁶⁾ FaPPO1 (Fragaria ananassa, gene loci 30435),³⁷⁾ pAPO5 (Malus × domestica, AAA69902),³⁸⁾ PyPPO (Pyrus pyrifolia, BAB64530),³⁹⁾ AmAS1 (Antirrhinum majus, Q9FRX6),¹⁸⁾ ToPPO1, ToPPO2, ToPPO3, ToPPO4, ToPPO5, ToPPO6 (Taraxacum officinale, ABX09994, CAQ76694, CBZ41490, CBZ41491, CBZ41492, CCA94610),^34,40–42) JrPPO1 (Juglans regia ACN86310,⁴³⁾ PtrPPO11, PtrPPO13 (Populus trichocarpa, AEH41424, AEH41425),²⁰⁾ PtdPPO1, PtdPPO2, PtdPPO3 (Populus trichocarpa × P. deltoides, AAG21983, AAU12256, AAU12257),⁴⁴⁾ MsPPO (Medicago sativa, AAP33165),⁴⁵⁾ VfPPO (Vicia faba, CAA77764),⁴⁶⁾ LtLH (Larrea tridentata, AAQ67412),²⁶⁾ PPO1 IPOBA (Ipomoea batatas, Q9ZP19),⁴⁷⁾ LePPO (Solanum lycopersicum, CAA78300),⁴⁸⁾ AUS1 (Coreopsis grandiflora AHN09736),⁴⁹⁾ OsPPO (Oryza sativa, ABG23042),⁵⁰⁾ AcoPPO (Ananas comosus AAO16865),⁵¹⁾ AchPPO (Annona cherimola ABJ90144).⁵²⁾ (B) Alignment of amino acid sequences of grape PPOs. Broken line indicates N-terminal chloroplast transit peptides of VvPPO1, VvPPO2, and VviPPO3. Double line indicates C-terminal shielding domain. Arrowheads indicate the predicted cleavage sites to form mature proteins. CuA and CuB represent conserved copper-binding domains.

Expression and purification of VvPPO variants

VvPPO2 was recombinantly expressed with (VvPPO2-Full) and without (VvPPO2-ΔN) a chloroplast transit peptide in E. coli. Both variants were obtained in the soluble fraction as judged from western blot analysis using anti-His-tag antibodies (data not shown); however, PPO activity was detected only for the cells expressing VvPPO2-ΔN. This construct is a so-called latent form and comprises the catalytic domain (amino acid residues 99‒462) and the C-terminal shielding domain (aa 463–602) (Fig. 2(B)). The purified final preparation of VvPPO2-ΔN migrated as a single band in SDS-polyacrylamide gel electrophoresis (Fig. 3). The molecular size under the denaturing conditions was estimated to be 60 kDa, which agrees with the calculated mass of 59.6 kDa.

Fig. 3. SDS-PAGE of the purified recombinant VvPPO2-ΔN. The so-called latent form of VvPPO2 was expressed in E. coli and purified to homogeneity. The gel was stained with Coomassie Brilliant Blue R-250.

Characterization of VvPPO2-ΔN

The enzyme showed maximum activity at pH 5.0−5.5, which represents the pH values inside vacuoles (Supplemental Fig. S1(A)). But the activity drastically decreased at the slightly alkaline conditions (pH 8), which correspond to the pH value of chloroplastidic stroma. VvPPO2-ΔN required detergent SDS to exhibit its activity, and its optimal concentration was determined to be 0.35 mM (Supplemental Fig. S1(B)). Under the optimized conditions, K_m and k_cat values of the enzyme for 4-MC were calculated to be 1.4 mM and 190 s⁻¹, respectively, which resulted in high catalytic efficiency of this enzyme (140 s⁻¹∙mM⁻¹) (Supplemental Fig. S2). Monophenolase activity was considerably lower than diphenolase activity (Table 1). Specific activity for caffeic acid was 25-fold higher than that for p-coumaric acid (0.98 s⁻¹ vs. 0.039 s⁻¹), and the enzyme acted on 3,4-dihydroxybenzoic acid while it was inert on 4-hydroxybenzoic acid under the conditions examined. The enzyme preferred the substrates that do not have bulky substitution groups at the phenolic ring. For example, the activity differed by 95-fold between 4-MC and caffeic acid.

Table 1. Substrate specificity of recombinant VvPPO2-ΔN.

Substrate	Structure	Specific activity (s^−1)^a
4-Methylcatechol		93
Caffeic acid		0.98
3,4-Dihydroxybenzoic acid		0.059
p-Coumaric acid		0.039
Tyrosol		1.1
Tyrosine		0.12
4-Hydroxybenzoic acid		nd^b

^a Specific activity was determined as described in the Materials and Methods section.

^b not detectable.

Size exclusion chromatography suggested that VvPPO2-ΔN exists as a hexamer (294 kDa), a trimer (165 kDa), and a monomer (51.9 kDa) at the physiological pH of chloroplastidic stroma (pH 8.5) in the absence of SDS. The monomeric form might be the dominant species under the conditions (Fig. 4(A)). However, when SDS was included in the elution buffer, peaks possibly corresponding to hexamer and trimer disappeared and a monomer peak that was slightly shifted and broadened to a lower size (44.2 kDa) was detected. At pH 5.0, which represents the physiological pH inside vacuoles, the enzyme appeared to form a trimer (136 kDa) and a monomer (45.1 kDa) in the absence of SDS, with the monomer being the primary form (Fig. 4(B)). In the presence of SDS, the peak presumably corresponding to a trimer disappeared and the monomer peak was detected. Interestingly, however, the peak area corresponding to the monomer considerably reduced, although the same amount of the enzyme was injected. No other apparent peak was detectable in the chromatogram. Susceptibility of proteins to pH change and SDS was found to differ among different proteins (Supplemental Fig. S3). 1,2-α-L-Fucosidase and galacto-N-biose/lacto-N-biose I-binding protein from bacteria, both of which are extracellular enzymes without containing disulfide bond, appear to be resistant to acidic pH and SDS. Bovine serum albumin eluted slightly faster at pH 5 in the presence of SDS as compared with when eluted at pH8.5 in the absence of SDS, but essentially the same shape of the peak was detected under the two different conditions. In contrast, ovotransferrin was found to be susceptible to the acidic pH and the detergent, and the peak area was decreased at pH 5 in the presence of SDS as compared with that observed at pH 8.5.

Fig. 4. Gel filtration analysis of recombinant VvPPO2-ΔN. The purified enzyme was injected onto Superdex 200 10/300 GL column that was pre-equilibrated with 20 mM Tris–HCl (pH 8.5) + 150 mM NaCl (A) or 20 mM sodium citrate (pH 5.0) + 150 mM NaCl (B) (black circles). SDS was included at the final concentration of 0.35 mM (A and B) (gray circles). Molecular size was estimated based on the standard curve created using molecular standard markers.

Stability of VvPPO2-ΔN was examined by determining the residual activity after incubation of the enzyme at pH 5 or pH 8.5 and in the absence and presence of SDS. The incubation temperature was also altered to represent the day/night ambient air conditions, i.e. 37 °C/20 °C, of the “Shine Muscat” cultivation field in Shimane Prefecture. In the absence of the detergent, the remaining activity did not change upon the incubation of the enzyme up to 70 min regardless of pH value and temperature (Fig. 5(A)). However, in the presence of SDS, the residual activity of the enzyme was found to be decreased (Fig. 5(B)). Particularly, drastic decrease was observed when the enzyme was incubated at pH 5 and 37 °C. The activity diminished by 10-fold after 3 min incubation, and became 1/1000 in 30 min under the conditions. Decrease of the remaining activity was also observed when the enzyme was kept at 37 °C and pH 8.5 in the presence of the detergent; however, the enzyme retained 20% activity even after 30 min incubation. At 20 °C, the enzyme lost its activity by 90% and 50% at pH 5 and pH 8.5, respectively, within 30 min in the presence of SDS. None of the data fitted to the equation of first-order reaction.

Fig. 5. Stability of recombinant VvPPO2-ΔN. The purified enzyme was incubated at different pHs (5.0 and 8.5) and in the absence (A) and presence (B) of 0.35 mM SDS for indicated times. The residual activity was determined by the standard method described in the Materials and Methods section.

Discussion

Recent genome sequence analyses revealed ubiquitous occurrence of PPOs in Viridiplantae except for green algae and Arabidopsis,²⁾ indicating that the enzyme holds an important roles in plant physiology. In addition to generally accepted functions in defensive response, some PPOs have been shown to be involved in the synthesis of particular secondary metabolites. C. grandiflora and L. tridentate have evolved to exploit PPO as aurone synthase and (+)-larreatricin hydroxylase, respectively.^25,26) Using walnut, Araji et al. showed that silencing of the PPO gene can cause necrotic lesions on the leaves.²⁷⁾ These findings suggest that the functions of PPOs are diverse among plants and, considering that several plant genomes encode multiple PPO paralogs, could even be different in one species.²⁾

The V. vinifera genome is assumed to encode four PPO genes (Fig. 2). While the VvPPO1 (GPO1) and VvPPO2 genes have been cloned already, occurrence of the VviPPO3 and VviPPO4 genes are still at the in silico prediction level in the genome annotation database.²⁾ As mentioned above, the roles of the four grape PPOs could be diverged and VvPPO2 should have different functions from VvPPO1 (GPO1) and VviPPO3, even though they are clustered closely in the phylogenetic tree. PPO1 (GPO1) is the only characterized enzyme because it is highly expressed in and easily obtained from immature berry and leaves.^14,28) Accordingly, enzyme preparation from grape berries in previous studies inevitably corresponds to the product of the VvPPO1 gene. Consequently, the enzymatic characterization of the other grape PPO genes has not been elucidated. PPO prepared from natural sources is suitable for determining the characteristics of the mature and native forms of the enzyme; however, this approach intrinsically involves methodological difficulty in preparing mutant or variant enzymes due to the lack, or laborious nature, of genetic tools available in grape. Recombinant expression is thus the most effective way to enhance our understanding of the structure-function relationship of target enzymes, but to the best of our knowledge, there have been no reports describing successful recombinant expression of grape PPO.

In the present study, we, for the first time, succeeded in recombinantly expressing catalytically active grape PPO. The PPO2 gene was chosen in this study because it was found to be up-regulated in the skin browning berries during the maturation stage and thus could be involved in the incidence of “Kasuri-shou”.¹²⁾ We used a VvPPO2-ΔN construct for expression and purification because it represents the latent form of PPO and thus is thought to be appropriate for analyzing the activation process of the enzyme. The purified enzyme showed very low activity at pH 8.5 in the absence of detergent (0.041 s⁻¹ at 1 mM 4-MC); however, it showed increased activity either when the pH was shifted to acidic conditions (pH 5.0) (0.25 s⁻¹ at 1 mM 4-MC) or when detergent SDS was added to the reaction mixture (0.18 s⁻¹ at 1 mM 4-MC). The latent form of recombinant VvPPO2 was fully activated at pH 5.0 in the presence of SDS (93 s⁻¹ at 1 mM 4-MC). The specific activity of VvPPO2-ΔN under the conditions was comparable to that of the C-terminally truncated mature form of VvPPO1 purified from “Riesling” grape berries (44 s⁻¹ at 2 mM 4-MC).²⁹⁾ It should be mentioned that although the C-terminal shielding domain was not cleaved off from the recombinant VvPPO2-ΔN construct, PPO activation is thought to involve both proteolytic and non-proteolytic processes in planta.^30,31) Detergent-dependent activation of latent enzymes has been demonstrated for various PPOs.^30–33) VvPPO2-ΔN can thus partly recapitulate the activation process of PPOs in the cells, provided that non-proteolytic activation process naturally occurs in planta and that a natural detergent exists in cells. Taken together, VvPPO2-ΔN should serve as one of the powerful tools to elucidate the reaction mechanism and the maturation process of PPOs.

Occurrence of monophenolase activity in PPOs is still a controversial issue, but recently increasing evidence has shown that PPOs do possess this activity. We detected degradation of monophenol substrates by VvPPO2-ΔN in HPLC analysis (Table 1). The reaction mixture containing tyrosine as the substrate was indeed colored pink during incubation. These results demonstrate that VvPPO2 possesses monophenolase activity.

As mentioned above, dandelion PPOs is the first example of successful recombinant expression of latent, but acid-, and detergent-activatable PPOs of plant origin.^4,34) The dandelion PPO1 and PPO2 were activated at pH 5 and pH 6 in the presence of 1 and 0.75 mM SDS, respectively, and their kinetic parameters for 4-MC were determined to be 1 mM and 125‒160 s⁻¹. Thus, although amino acid sequence identity is not very high between VvPPO2 and dandelion PPOs (~65%) (Fig. 2(A)), the enzymatic characteristics were quite similar between them. The activation process from a latent form to a mature form may be generally conserved in PPOs from different plant origins, as previously suggested.²⁾ In this regard, it is interesting to note that domain-swapping analysis between dandelion PPO1 and PPO2 revealed that C-terminal shielding domain determines the pH optimum level for PPO activation while the linker between the catalytic domain and C-terminal domain determines the optimal SDS concentration required for full activation.³¹⁾ The linker region was also found to significantly influence enzyme stability during incubation in the presence of SDS. In some swapped dandelion PPO variants, the half-lives were decreased to ~5 min. VvPPO2-ΔN partly formed oligomer in solution at pH 5.0 and pH 8.5 in the absence of SDS and, under the conditions, no destabilization was observed up to 70 min incubation at 20 and 37 °C (Figs. 4 and 5). Role of the oligomer formation in the latent enzyme is unclear. Addition of the detergent to the solution however caused varied response for the enzyme depending on the pH value. Significant destabilization was observed at pH 5.0 and at 37 °C, under the conditions of which the half-life of the enzyme was less than 3 min. The peak area of the enzyme in the chromatogram of gel filtration also significantly decreased, strongly suggesting denaturation of the enzyme (Fig. 4). Indeed, precipitant appeared in the incubation mixture when we examined the residual activity of the enzyme (Fig. 5). Interestingly, under the most unstable conditions, VvPPO2-ΔN exhibited maximal activity. The catalytic domains of recombinant VvPPO2-ΔN and dandelion PPOs do not contain disulfide bonds that are present in all PPOs prepared from natural sources, and therefore the in vitro physicochemical behaviors observed for the recombinant enzymes could be different from those which would be obtained for native latent enzyme. Nonetheless, the observed stability of VvPPO2-ΔN under the plastid-mimicking conditions (pH 8.5) and instability of it under the damaged cell-mimicking conditions (pH 5.0 and detergent) is worth consideration. Quick degradation of PPOs might be beneficial for plant cells to transiently respond to the biotic and abiotic stresses and to prevent detrimental prolonged polymerization reactions caused by quinones produced by the enzymes. Given this, recombinant production of the C-terminal truncated, disulfide bond-containing mature form of VvPPO2 and comparison of its properties with that of VvPPO2-ΔN are necessary to decipher the molecular mechanism underlying PPO maturation and destabilization.

In conclusion, we succeeded in the preparation of recombinant VvPPO2. Latent VvPPO2 was fully activated but significantly destabilized under the acidic conditions in the presence of detergent. The enzyme showed higher diphenolase activities than monophenolase activities. Our study furthers understanding of not only the berry skin browning process but also the maturation processes of PPOs. Isolating as-yet uncharacterized VviPPO3 and VviPPO4 as well as preparing latent- and matured forms of the four grape PPOs may also help us explore currently ambiguous PPO-related browning reactions involving phenolic compounds, which may enable comprehensive understanding of physiological roles of PPOs in grape.

Authors’ contributions

A. K.-I., H. I., T. and T. E. conceived and designed the experiments. Y. S., A. K.-I., K. J., and T. K. performed the experiments. A. K.-I., Y. S., T. K. and T. E. wrote the paper. All authors discussed the results for the completion of the manuscript. A. K.-I., T. K., and T. E. edited the manuscript.

Disclosure statement

No potential conflict of interest was reported by the authors.

Funding

This work was supported in part by a Grant-in-Aid for Young Scientists (B) [No. 22780029], [No. 26850019] (to A. K.-I) from the Ministry of Education, Culture, Sports, Science and Technology, Japan, and by a Grant for the Promotion of Research from Actree Corporation.

Supplemental materials

Supplemental material for this article can be accessed at https://doi.org/10.1080/09168451.2017.1381017

Notes

Abbreviations: PPO, polyphenol oxidase; 4-MC, 4-methylcatechol