Expression and characterization of a recombinant unique acid phosphatase from kidney bean hypocotyl exhibiting chloroperoxidase activity in the yeast pichia pastoris

Protein Expression and Purification 53 (2007) 31–39 www.elsevier.com/locate/yprep

Expression and characterization of a recombinant unique acid phosphatase from kidney bean hypocotyl exhibiting chloroperoxidase activity in the yeast pichia pastoris q Tohru Yoneyama a, Maya Taira a, Tomonori Suzuki a, Masao Nakamura b, Koichi Niwa a, Toshihiro Watanabe a, Tohru Ohyama a,* a

Department of Food Science and Technology, Faculty of Bioindustry, Tokyo University of Agriculture, 196 Yasaka, Abashiri 099-2493, Japan b Department of Chemistry, Asahikawa Medical College, Midorigaoka-Higashi 2-1-1-1, Asahikawa 078-8510, Japan Received 22 September 2006, and in revised form 15 November 2006 Available online 20 December 2006

Abstract We previously purified and characterized a novel acid phosphatase (KhACP) from kidney bean hypocotyls that exhibited vanadatedependent chloroperoxidase (V-CPO) activity. In the present study, a functional recombinant KhACP (rKhACP) was successfully produced at a high expression level by the methylotrophic yeast Pichia pastoris. The KhACP cDNA excising signal peptide sequence was subcloned into the pPICZaA vector and then integrated into the genome of P. pastoris strain X-33 under control of the alcohol oxidase 1 promoter. The rKhACP protein, with a molecular mass of 60 kDa on SDS–PAGE, was secreted into the culture medium as a C-terminal His-tagged fusion protein. Purification was facile using only nickel affinity chromatography. The apparent molecular mass of the purified rKhACP was estimated to be around 110 kDa by analytical gel filtration. PAGE analysis showed that rKhACP was a glycosylated dimeric enzyme, consisting of two 60-kDa subunits linked non-covalently, which was similar to the dominant form of the natural enzyme isolated from plant material. Furthermore, the rKhACP exhibited V-CPO activity when ortho-vanadate (VO4 3 ) was added to the apo enzyme, and it showed broad substrate specificity and kinetic parameters comparable to the natural enzyme. This expression system produces sufficient protein to allow us to attempt to determine the three-dimensional crystal structure, which will shed light on its unique mechanism of converting KhACP to vanadate-dependent chloroperoxidase.  2007 Elsevier Inc. All rights reserved. Keywords: Acid phosphatase; Chloroperoxidase; Vanadate; Kidney bean; Methylotrophic yeast Pichia pastoris

Acid phosphatase (ACP)1 (EC 3.1.3.2) enzymes catalyze hydrolysis of phosphate monoesters, releasing inorganic phosphorus from phosphorylated substrates in vitro in the pH range 4–7. ACPs are ubiquitous in a wide variety of animals, plants and microorganisms, and they exhibit q The nucleotide sequence reported in this paper has been submitted to the GenBank database with Accession No. AB116719 for cDNA. * Corresponding author. Fax: +81 152 48 2940. E-mail address: [email protected] (T. Ohyama). 1 Abbreviations used: ACP, acid phosphatase; CPO, chloroperoxidase; SDS–PAGE, sodium dodecyl sulfate–polyacrylamide gel electrophoresis; PAS, periodic acid-Schiff; p-NPP, p-nitrophenyl phosphate; 2-ME, 2mercaptoethanol; MCD, monochlorodimedon.

1046-5928/$ - see front matter  2007 Elsevier Inc. All rights reserved. doi:10.1016/j.pep.2006.12.010

low substrate specificity [1,2]. Plant ACPs are present in various organs of germinating seeds and also in different cell compartments, suggesting that these enzymes are involved at various cellular metabolic levels. Plant ACPs are also induced under various environmental and developmental conditions, including salt or drought stress, seed germination, flowering and pathogen infection, making their cellular roles difficult to define [3–5]. However, the stimulation of phosphatase activities in response to phosphate starvation is well documented, and ACPs play a role in the utilization of phosphate compounds [1]. Purple ACPs are commonly found in a wide range of eukaryotic organisms and in some bacterial species. They

32

T. Yoneyama et al. / Protein Expression and Purification 53 (2007) 31–39

belong to a family of non-specific ACPs containing a bimetal nucleus in their active center [6–8], that also includes phosphoprotein phosphatase and other types of phosphomonoesterase [9,10]. The characteristic purple color of the purified enzymes is the result of a tyrosine to Fe3+ charge transfer transition [11]. The purple colored ACP from the kidney bean is the best characterized plant enzyme, due to the determination of its three-dimensional structure. It is a homodimeric glycoprotein with a molecular mass of 110 kDa, containing a Fe3+–Zn2+ metal site in each of two subunits that are connected by a disulfide bond [9,12]. Recently, we purified and characterized a novel colorless acid phosphatase (KhACP) from kidney bean (Phaseolus vulgaris cv. Ohfuku) hypocotyls, which was distinct from the kidney bean purple acid phosphatase [13]. When ortho-vanadate (VO4 3 ) was added to the apo-form of the enzyme, KhACP uniquely exhibited chloroperoxidase activity with concomitant loss of phosphatase activity. This is the first demonstration that KhACP is a vanadate-dependent chloroperoxidase (CPO) in plants, and it suggests that KhACP may play a role in modifying a wide variety of chlorinated compounds that are present in higher plants. KhACP is also a dimeric glycoprotein with a molecular mass of 96 kDa consisting of dominant 56 kDa and minor 45 kDa subunits that are connected by a non-covalent bond. Unlike most purple ACPs that contain a binuclear Fe3+–Zn2+ (Fe) center at their active sites [5], KhACP develops no purple color and shows CPO activity after the addition of vanadate. Additionally, Northern blot analysis demonstrated that expression of the KhACP gene was observed specifically in hypocotyls of the kidney bean, and that it coincided with elongation of the hypocotyl during germination, strongly suggesting that KhACP is clearly distinct from other ACPs that have been reported in plants. However, we have not obtained enough data to elucidate the physiological functions of KhACP, due to the limited yield of protein purified from the bean hypocotyl. Further characterization of the enzyme, such as determining its precise structure and physiological substrates, will require a large amount of KhACP. We also isolated a gene encoding KhACP (GenBank Accession No. AB116719) from a kidney bean hypocotyls cDNA library [13]. The full-length KhACP gene is closely related to the soybean purple ACP-like gene GmPAP3 [14] with 92% amino acid sequence identity, and to lupine ACP [15] with 87% sequence identity, but it has only 54% identity with the red kidney bean purple ACP gene [16]. Computer-assisted motif analysis and phylogenetic analysis suggest that KhACP belongs to a large purple ACP subfamily consisting of proteins with no disulfide bridge between their subunits. Nevertheless, KhACP is regarded as a novel ACP that differs from previously identified purple ACPs in plants. The purpose of the present study is to establish an efficient system for expression of this unique KhACP,

allowing us to produce milligram amounts of this protein. The methylotrophic yeast P. pastoris, which is one of the dominant expression systems in molecular biology due to its stable and high-level expression of heterologous proteins [17,18], can be easily grown to high cell density in defined minimal media. The yeast is also able to introduce eukaryotic post-translational modifications such as glycosylation. In the present work, we report our strategy for overexpression of this enzyme by P. pastoris. The cDNA encoding KhACP was successfully expressed in yeast cells as a His-tag fusion protein. The optimized purification procedure for obtaining milligram amounts of homogenous active recombinant enzyme is presented. We also confirm here that the expressed recombinant KhACP (rKhACP) is functionally equivalent to the natural enzyme. Materials and methods Yeast culture media Pichia pastoris cells were cultured in YPD medium (1% yeast extract, 2% peptone, and 2% D-glucose) or BMGY medium (1% yeast extract, 2% peptone, 1.34% yeast nitrogen base, 4 · 105% biotin, 1% glycerol, 1% casamino acids, and 100 mM potassium phosphate, pH 6.0). YPDS plates (YPD medium plus 1 M sorbitol and 2% agar (w/v)) containing 100 lg/ml zeocin (Invitrogen, San Diego, CA) were used for selection of Pichia transformants. BMMY medium (1% yeast extract, 2% peptone, 1.34% yeast nitrogen base, 4 · 105% biotin, 0.5% methanol, 1% casamino acids, and 100 mM potassium phosphate, pH 6.0) was used for protein induction. Strains and vectors Escherichia coli TOP10 cells and the plasmid vector pCR4-TOPO were used for cloning. For yeast transformation, the P. pastoris transfer vector pPICZaA containing the 5 0 alcohol oxidase 1 (AOX1) promoter and the 3 0 AOX1 transcription termination sequences was used. pPICZaA also contains the dominant selectable marker zeocin, which is bifunctional in both Pichia and E. coli. Pichia pastoris host strain X-33 was used for protein expression experiments. These products were purchased from Invitrogen (Carlsbad, CA). Construction of expression vector and P. pastoris transformation The KhACP gene (AB116719) without a native signal peptide sequence (1275 bp) was amplified by PCR using kidney bean hypocotyls cDNA library as a template [13] with sense (5 0 -GAA TTC GGG ATC ACT AGC TCC TTC-3 0 ) and antisense (5 0 -CTA GAT ACC AAT ACT GGT TAT GCA ATA C-3 0 ) primers introducing an EcoRI site (sense) and XbaI site (antisense) (underlined, respectively) into the region immediately upstream of the first

T. Yoneyama et al. / Protein Expression and Purification 53 (2007) 31–39

codon GGG and downstream of the final codon ACC, respectively. The PCR was carried out using a Gene-Amp automated thermal cycler (Model 9700, Applied Biosystems, Foster City, CA) and using LA Taq polymerase (Takara bio, Tokyo, Japan) with the following conditions: an initial denaturation for 5 min was followed by 30 cycles of denaturation (94 C, 1 min), annealing (55 C, 1 min) and extension (68 C, 3 min), and the final extension (68 C, 5 min). The PCR product encoding KhACP was cloned into the pCR4-TOPO vector. The pCR4-TOPOKhACP plasmid was transformed into E. coli TOP10 cells. Plasmid DNAs were isolated from E. coli cell cultures using a GFX Micro Plasmid Prep Kit (Amersham Biosciences, Piscatway, NJ) according to the manufacturer’s instruction. After EcoRI–XbaI digestion of isolated plasmid DNA, the KhACP coding sequence was cloned into pPICZaA in frame between the EcoRI (5 0 end) and XbaI (3 0 end) restriction sites to generate the plasmid pPICZaA-KhACP for expression of protein containing the 6· His-tag at its Cterminus. Restriction enzyme digestion and ligation steps with T4 DNA ligase were performed as recommended by the enzyme suppliers. The pPICZaA-KhACP construct was transformed into the E. coli TOP10 strain. Proper insert orientation was tested by restriction digestion and DNA sequencing. Sequencing reactions were performed with 5 0 AOX1 and 3 0 AOX1 vector-specific primers, and dye-labeled extension products were analyzed with a 3130 DNA sequencer using ABI PRISM Big Dye Terminator chemistry (Applied Biosystems). Electrocompetent P. pastoris X-33 cells were prepared using standard methods [19] and their transformation was performed by electroporation. The pPICZaA-KhACP construct (50 lg) was transformed into P. pastoris wild type strain X-33. The recombinant yeast clones were selected on YPDS plates containing 100 lg/ml zeocin. The colonies were subsequently screened by performing direct PCR on yeast colonies to confirm integration of the KhACP gene into the yeast genome. Expression and purification of rKhACP in P. pastoris For overexpression, a single P. pastoris X-33-KhACP colony was inoculated into 25 ml BMGY medium and grown at 28 C in a shaker incubator until the culture reached an OD600 of 4.0. Cells were harvested by centrifugation for 5 min at 3250 rpm. The cell pellet was resuspended to an OD600 of 1.0 in 100 ml BMMY medium and the culture was grown for 48 h at 28 C with shaking when the culture reached OD600 of 20–30. To induce rKhACP, methanol was added every 24 h to a final concentration of 0.5%. Finally, the culture supernatant (100 ml) was collected by centrifugation (10 min, 15,000 rpm, 4 C). Purification of the rKhACP protein was carried out in only one chromatographic step. Proteins in the culture supernatant were precipitated with 80% ammonium sulfate fractionation. The precipitate was then suspended in a small volume of 50 mM Tris–HCl (pH 7.4) containing

33

0.5 M NaCl and dialyzed against the same buffer. Protein that precipitated during dialysis was removed by centrifugation. The enzyme solution was applied to a ProBond resin nickel affinity column (2.6 · 20 cm, Invitrogen) equilibrated with 50 mM Tris–HCl (pH 7.4) containing 0.5 M NaCl. Non-specific adsorbed materials were washed off with an equilibration buffer containing 20 mM imidazole. The adsorbed materials were eluted with 50 mM Tris– HCl (pH 7.4) containing 300 mM imidazole. The fractions having enzyme activity were collected and concentrated to minimal volume using the membrane filter YM10 (Amicon, Beverly, MA). The proteins were quantified by Lowry method using bovine serum albumin as a standard. Natural KhACP was purified from hypocotyls of kidney bean (P. vulgaris cv. Ohfuku) as described previously [13]. PAGE and N-terminal amino acid sequence analysis SDS–PAGE was performed using a 10% polyacrylamide gel in the presence and absence of 2-mercaptoethanol. The molecular size markers were phosphorylase b (106.9 kDa), BSA (93.6 kDa), egg ovalbumin (52.2 kDa), carbonic anhydrase (37.2 kDa), soybean trypsin inhibitor (28.2 kDa) and lysozyme (18.8 kDa) (Bio-Rad laboratories, Hercules, CA). The separated protein bands were stained with Coomassie Brilliant Blue R-250 (CBB). Glycoprotein staining of the gel was performed according to a periodic acid-Schiff (PAS) staining protocol [20]. For the analysis of N-terminal amino acid sequence, the proteins separated by SDS–PAGE were transferred onto a polyvinyldine difluoride membrane according to the method as previously described [21]. The amino acid sequence of each component was determined using an automated protein sequence analyzer (Model 492HT, Applied Biosystems). Western blot analysis for detection of the 6 · His epitope tag fusion protein The proteins separated by SDS–PAGE were electroblotted onto a nitrocellulose membrane (Amersham Bioscience) and probed with an anti 6 · His epitope tag IgG (Affinity Reagents, Golden, CO) (1:1000 dilution in Trisbuffered solution (TBS) containing 3% skim milk and 0.1% Tween 20) by incubating the membrane at room temperature for 1 h. After rinsing with TBS, the membranes were incubated with an anti-rabbit IgG-HRP conjugate (1:1000 dilution in TBS containing 3% skim milk and 0.1% Tween 20) at room temperature for 1 h. After washing the membrane with the same solution, the bound antibody was visualized with chemiluminescence detection (Amersham Bioscience). Analytical gel filtration A Superdex 200 HR 10/30 gel filtration column (1.0 · 30 cm) (Amersham Bioscience) was equilibrated with

34

T. Yoneyama et al. / Protein Expression and Purification 53 (2007) 31–39

20 mM Tris–HCl (pH 7.5) containing 0.5 M NaCl. The both enzymes were detected absorbance at 280 nm and ACP activity. The molecular masses of the natural and recombinant enzymes were estimated by calculation using the standard proteins thyroglobulin (669 kDa), ferritin (440 kDa), aldolase (158 kDa), BSA (67 kDa), egg ovalbumin (43 kDa) and ribonuclease A (13.7 kDa). The molecular mass was estimated from the mean result of three experiments. Deglycosylation analysis Deglycosylation of the natural KhACP and rKhACP was performed using an N-glycosidase F Deglycosylation Kit (Roche, Penzberg, Germany) according to the manufacturer’s instruction. The purified enzyme (50 lg) was incubated in 10 ll of denaturation buffer (pH 8.6) containing 1% SDS at 95 C for 3 min, and was transferred into reaction buffer (pH 7.2) containing 0.5% Nonidet P40. The reaction mixture was incubated with 12 U of N-glycosidase F (Roche) for 12 h at 37 C. Samples were loaded on SDS–PAGE, and the gels were stained by CBB and PAS reagent. Assay for ACP activity The ACP activity of the enzyme was routinely determined by following the hydrolysis of p-nitrophenyl phosphate (p-NPP) at 410 nm. The optimal pH of the enzyme was around pH 6.0, and it showed activity in the range of pH 5–7 (data not shown). The routine assay was performed at 30 C by adding 50 ll of enzyme solution to 0.5 ml of reaction mixture consisting of 50 mM p-NPP and 200 mM Tris–maleate (pH 6.0). One enzyme unit was defined as 1 lmole of p-NPP hydrolyzed per 1 min. Substrate specificities of the natural KhACP and rKhACP were estimated by incubating the purified enzyme with various substrates (17 kinds). The assay was carried out at 30 C by adding 50 ll of the enzyme solution to a reaction mixture containing 10 mM target substrate and 50 mM Tris–maleate (pH 6.0), and phosphorus released during the 2 min incubation was measured according to a method described previously [13]. In all assays one unit of enzyme activity was defined as 1 lmole of Pi released per 1 min. Assays to examine the effects of metal ions and chemical agents (nine kinds) on ACP activity were carried out as above, by incubating p-NPP at 30 C for 30 min with natural enzyme or recombinant enzyme that was dialyzed against buffer containing EDTA. All measurements were repeated five times and the results represent the arithmetic means. Assay for CPO activity The vanadate-substituted enzyme (V-KhACP and V-rKhACP) was prepared by incubating the apoenzyme

with various concentration of ortho-vanadate in 100 mM Tris–HCl (pH 7.5) at 30 C for 30 min. The purified enzyme was then dialyzed against 100 mM Tris–HCl (pH 7.5) containing 20 mM EDTA. The CPO activity was measured by following the decrease in the absorbance at 290 nm due to chlorination of monochlorodimedon (MCD) in the presence of hydrogen peroxide [13]. The standard assay was carried out at 30 C in a reaction mixture containing 2 mM H2O2, 5 mM KCl and 100 mM citrate buffer (pH 4.7). One unit of CPO was defined as 1 lmole of MCD chlorinated per 1 min. Results and discussion Expression and purification of rKhACP in P. pastoris Purification of KhACP from kidney bean hypocotyls is a relatively laborious process yielding only a small amount of protein [13]. The P. pastoris gene expression system is an attractive method with which to produce a variety of intercellular and extracellular proteins [17]. Since, we previously isolated and identified a cDNA encoding this protein, we decided to clone the DNA sequence encoding KhACP without its native signal peptide in the P. pastoris pPICZaA plasmid, in frame with the yeast a-factor signal sequence, for secretion of protein containing a 6· HisTag at its C-terminus. The pPICZaA-KhACP plasmid was then transformed and targeted to the P. pastoris genome by means of homologous recombination. The presence of the KhACP coding sequence in the genomic DNA isolated from Pichia transformants was confirmed by PCR. In the P. pastoris expression system, recombinant protein expression is strictly controlled by the AOX1 promoter. Expression was induced by addition of methanol to a final methanol concentration of 0.5%. The time course of the secretion of recombinant protein by a P. pastoris transformant is shown in Fig. 1. A 60-kDa protein that was not present before methanol induction was detected by SDS–PAGE and subsequent Western blot analysis with an anti-His-tag probe. The 60-kDa protein gradually accumulated in the medium during the cultivation after the methanol induction (Fig. 1a and b), and the ACP activity also increased consistently during this period (Fig. 1c). The 60-kDa recombinant protein in the 48-h culture supernatant was purified to near homogeneity though the single step of nickel chelating resin chromatography. As shown in Fig. 2, the purified protein was revealed as a single band on SDS–PAGE and Western blot analyses; its Nterminal amino acid sequence was determined to be EFGITSSFIRSEWPAVDIPL. Since two residues (EF) at the N-terminus were derived from the restriction site sequence, the N-terminal sequence of rKhACP begins with the glycine at position 3. Our previous data demonstrated that the N-terminal amino acid sequence of the natural mature KhACP was SEWPAVDIPL (Fig. 2, underlined), in which eight residues were removed from the N-terminus due to processing by a trypsin-like protease that was in the

T. Yoneyama et al. / Protein Expression and Purification 53 (2007) 31–39

M

kDa

Culture time (h) 0 12 24 36

48

M

kDa

106.9 93.6

Culture time (h) 0 12 24 36

48

90 80 70

60 kDa

60 50

52.3

40 37.2 30

28.2

20

18.8 SDS−PAGE

Western blot

ACP activity (U / ml)

c 0.04

signal peptide was correctly excised without additional proteolytic processing. Table 1 shows a summary of the purification procedure parameters. The rKhACP was purified 25-fold to a final pNPP specific activity of 32.2 · 102 U/mg protein with a recovery of 87%. The nickel chelating resin affinity chromatography step effectively served to remove contaminating proteins, and a single chromatographic peak of ACP activity was observed. Approximately 6 mg pure rKhACP was obtained from 100 ml of BMMY medium (0.06 mg/ml). Since the yield of natural enzyme is only 0.96 mg from 600 g of kidney bean hypocotyls, the much larger amount now available will allow more extensive characterization of rKhACP properties. Molecular characteristics of rKhACP

0.03

0.02

0.01

0 0

12

24

36

48

Culture time (h)

Fig. 1. Time course for the secretion of recombinant protein by methanolinduced P. pastoris X-33-rKhACP cells. (a) Culture supernatant aliquots (30 ll) were subjected to SDS–PAGE. Samples were taken at the indicated culture time (h) after methanol induction. The gel was stained by CBB. (b) Culture supernatant aliquots (10 ll) were subjected to SDS–PAGE followed by Western blot analysis. Samples were taken at the indicated culture time (h) after methanol induction. For detection of recombinant protein, anti 6 · His epitope tag IgG was utilized (1:1000 dilution). (c) ACP activities in culture supernatants at each culture time (h) after methanol induction.

M

35

1

M

2

kDa 106.9 93.6

1

2

kDa 80 70 60 50 40

52.3

60 kDa ( E F GIT S S F IRS E WP AVD I PL. . . )

37.2 30 28.2 20 18.8

SDS−PAGE

Western blot

Fig. 2. SDS–PAGE and Western blot analysis of purified rKhACP. Left panel shows the CBB stained SDS–PAGE gel. Right panel shows the protein detected using anti 6 · His epitope tag IgG (1:1000 dilution). Lane 1, 80% ammonium sulfate fraction. Lane 2, purified rKhACP with Nterminal amino acid sequence after ProBond nickel affinity chromatography.

hypocotyls tissue, or that became active during purification [13]. Accordingly, the active, tagged 60-kDa protein was confirmed to be rKhACP. The vector-derived a-factor

The purified 60-kDa rKhACP band on the SDS–PAGE gel was stained with a PAS reagent, strongly suggesting that the 60-kDa protein is glycosylated. Deglycosylation of the rKhACP with N-glycosidase F under the denaturation condition was successful. After 12 h incubation, the 60-kDa protein band migrated as a 52-kDa band on SDS–PAGE, and this band no longer stained with the PAS reagent (Fig. 3b). The estimated molecular mass of 52 kDa for the deglycosylated rKhACP agreed very closely with that predicted from the vector sequence that contained two extra residues at the N-terminus, 17 residues of the c-myc epitope and six residues in the His-tag at the C-terminus (452 amino acids, 51611 Da). These results confirm that rKhACP is N-glycosylated, as assumed from the four potential N-glycosylation sites in the rKhACP sequence. On the other hand, deglycosylation of natural KhACP with N-glycosidase F resulted in a decrease of 1.4–2 kDa molecular mass in each respective subunit, corresponding to at least 7–12 sugar residues (Fig. 3d). We can speculate that the partial fucosylation of some oligosaccharide core residues may make them inaccessible to N-glycosidase F [13]. Generally, P. pastoris can add both O- and N-linked carbohydrate moieties to secretory proteins [22]. It is likely that the glycosylation pattern of rKhACP differs from that of natural KhACP due to species differences between higher plants and yeast. The apparent molecular mass of the purified rKhACP was estimated to be around 110 kDa by analytical gel filtration on a Superdex 200 HR 10/30 column (Fig. 4). Taking into account-specific interactions between glycosylated residues and the gel filtration resin, the native rKhACP appears to be a homodimer consisting of identical 60kDa subunits. There are no intersubunit disulfide linkages, as no dimer band was seen on SDS–PAGE in the absence of reducing agent (Fig. 3a). On the other hand, the molecular mass of natural KhACP was estimated to be 96 kDa by gel filtration (Fig. 4), as previously reported [13]. This corresponds to a dimer consisting of dominant glycosylated 56- and minor 45-kDa subunits linked to each other non-covalently (Fig. 3c). Since both subunits have an

36

T. Yoneyama et al. / Protein Expression and Purification 53 (2007) 31–39

Table 1 Summary of the rKhACP purification from the 48 h culture supernatant Activity (U/ml)

Total activity (U)

Total protein (mg)

Specific activity (·102 U/mg)

Purification-fold

Yield (%)

(NH4)2SO4 fractionation Nickel affinity chromatography

0.18 0.15

2.29 1.98

177 6.15

1.30 32.2

1.0 25.0

100 87

d

Purification stage

(+)

M

Thyroglobulin Ferritin

106.9 93.6 60 kDa 52 kDa

52.3

37.2

Molecular mass (kDa)

(−)

hA

M

C P D e rK gl hA yc C osy P la

kDa

M (+)

rK

b

2-mercaptoethanol

a

a 1000

te

Recombinant KhACP

Recombinant KhACP (110 kDa) Natural KhACP (96 kDa) Aldolase 100 BSA Egg ovalbumin

N-glycosidase F

28.2

Ribonuclease A CBB

CBB

PAS

10

PAS

8

9

10

11

12

Natural KhACP

16

17

18

19

20

te la

b

D Kh egl AC yco P sy

AC

1.0

13.94

Recombinant KhACP

3

(+)

M

2 0.5 1

A280

106.9 93.6

52.3

56 kDa

54 kDa

45 kDa

43.6 kDa

37.2

0

0

1.0

3

Natural KhACP

14.20

2 0.5 1

ACP activity

A280

N-glycosidase F 0

28.2

ACP activity (U)

(−)

Kh

M

15

P

2-mercaptoethanol (+)

14

d

d

c kDa M

13

Elution volume (ml)

0

5

10

15

0 20

Elution volume (ml) CBB

PAS

CBB

PAS

Fig. 3. SDS–PAGE analyses in the presence and absence of 2-mercaptoethanol of natural KhACP and rKhACP. (a and c) SDS–PAGE of the purified rKhACP (a) and natural KhACP (c) in the presence of 2mercaptoethanol (lane 1, (+)) and absence of 2-mercaptoethanol (lane 2, ()) using a 10% separating gel. Protein bands in both lanes were stained by CBB. Lane 3, glycoprotein staining of purified rKhACP (a) and KhACP (c) separated on SDS–PAGE. Protein bands were stained by PAS. (b and d) Deglycosylation of the purified rKhACP (b) and natural KhACP (d) with N-glycosidase F separated from SDS–PAGE. Protein bands were stained by CBB and PAS.

identical N-terminal amino acid sequence, the minor 45kDa subunit was likely derived from deletion of about 100 amino acids from the C-terminal end of the dominant 56-kDa subunit, or the difference may have been due to a change in the glycosylation pattern of each subunit. Although natural KhACP seemingly could exist as a mixture of three dimeric forms, a homodimer of the 56-kDa subunit or the 45-kDa subunit and a heterodimer contain-

Fig. 4. Molecular size estimation of KhACP and rKhACP by analytical gel filtration. (a) Purified natural KhACP or rKhACP was applied to a Superdex 200 HR 10/30 gel filtration column equilibrated with 20 mM Tris–HCl (pH 7.5) containing 0.5 M NaCl. Measurements were repeated three times, and the results represent the arithmetic means. (b) Superdex 200 HR 10/30 gel filtration profiles of Purified natural KhACP and rKhACP. Numerical values with arrowheads show the elution volume of the both enzyme.

ing each subunit, these data suggest that the rKhACP is structurally similar to a homodimer of a 56-kDa subunit of natural KhACP. Thus, native rKhACP that has not undergone proteolytic processing offers great advantages to examine molecular properties, comparing with mixtures of hetero-dimeric natural KhACP. Enzymatic properties of rKhACP To elucidate the enzymatic properties of rKhACP, kinetic parameters were determined and compared with

T. Yoneyama et al. / Protein Expression and Purification 53 (2007) 31–39 Table 2 Kinetic constants of purified natural KhACP and rKhACP

Natural KhACP Recombinant KhACP

Table 4 Effect of metal ions and chemical agents on KhACP activity

Vmax (mM/min)

Km (mM)

17.4 ± 3.3 17.3 ± 2.6

0.56 ± 0.17 0.61 ± 0.15

Vmax and Km values determined using p-NPP as a substrate. Data represent the arithmetic means of five experiments.

those of natural KhACP. As shown in Table 2, both Vmax and Km values of purified rKhACP using p-NPP as substrate were similar to those of the natural enzyme. The pH dependence of rKhACP yielded a curve with a pH optimum at pH 6.0 (data not shown) that was almost identical to that of natural KhACP. The purified rKhACP and natural KhACP were tested for phosphorylated substrate specificity (Table 3). Like natural KhACP, rKhACP also showed a broad range of substrate specificity, and it catalyzed dephosphorylation of phosphorylated-Ser, -Thr and -Tyr. However, the relative activities of rKhACP against various phosphorylated compounds appeared to be somewhat higher than for the natural enzyme. Effects of various chemicals on rKhACP and natural KhACP activities As shown in Table 4, the effects of various chemicals on rKhACP were very similar to effects on natural KhACP. However, only Fe3+ at 100 lM enhanced by 1.88-fold the activity of rKhACP, but not that of the natural enzyme. Normally, plant purple ACPs are known to contain a dinuclear Fe3+–Me center (where Me can be Zn2+ or Mn2+) in their active site, such as Fe3+–Zn2+ in red kidney bean [12] and soybean [5], or Fe3+–Mn2+ in sweet potato [23]. The purple color is due to a charge transfer complex Table 3 Substrate specificity of purified natural KhACP and rKhACP Substrate

Relative activity (%)a Natural KhACP

Recombinant KhACP

p-NPP ADP 5 0 -AMP ATP a-Naphtylphosphate b-Naphtylphosphate bis-p-NPP Glucose-1-phosphate Glucose-6-phosphate Pyrphosphate 2-Phosphoglycerate 3-Phosphoglycerate Phosphoenolpyruvate Phytate Phosphorylated-Thr Phosphorylated-Ser Phosphorylated-Tyr

100 54 13 40 32 59 NDb 58 46 65 32 34 67 16 42 28 63

100 79 34 72 50 107 7 27 66 109 60 69 86 35 55 42 106

a b

The activity against p-NPP was taken to be 100%. ND, not detected.

37

Addition

2+

Zn Fe3+ Mn2+ VO4 3 Mo6+ Tartrate Citrate NaF EDTA a

Relative activity (% of control)a Natural KhACP

Recombinant KhACP

10 (lM)

100 (lM)

10 (lM)

100 (lM)

101 108 106 5 38 98 99 114 97

70 102 94 10 13 110 99 107 104

103 115 105 60 30 99 102 102 103

74 188 113 17 11 105 105 101 101

Data represent the arithmetic means of five experiments.

between Fe3+ and a nearby Tyr residue. However, addition of Fe3+ and Zn2+ to the rKhACP protein after dialysis against buffer containing EDTA did not lead to development of a purple color. We have as yet no explanation for the effect of Fe3+ on ACP activity, and it is therefore a high priority for us to obtain a crystal structure of rKhACP, as this could explain interaction between the active site and Fe3+.

Vanadate-substituted CPO activity Fig. 5a and c shows that both vanadate-substituted natural KhACP and rKhACP uniquely exhibited CPO activity with a loss of ACP activity. However, the CPO activity of rKhACP was about one fifth that of the natural enzyme. Modified Hill plots with vanadate for both enzymes are shown in Fig. 5b and d. The interaction between both enzymes and vanadate showed a linear relationship, and their Hill constants were calculated from the slope to be 1.0, respectively. This suggests that one molecule of vanadate binds to either subunit, or that two molecules bind to both subunits with equivalent affinity. However, the apparent dissociation constant for vanadate bound to rKhACP, Kd, was determined to be 21 lM, which was higher than that for the natural enzyme (6.6 lM vanadate). The discrepancies between CPO activity and Kd values between rKhACP and the natural enzyme may reflect differences in their structures due to altered glycosylation patterns, or may be due to the presence of the minor subunit found in the natural enzyme. Further detailed structural analysis may be required to better clarify these issues. To our knowledge, recombinant plant purple ACPs are available only by heterologous expression from baculovirus-infected insect cells [24–26], but expression of purple ACP from kidney bean was tested in yeast without success [24]. This is the first report describing expression and purification of functional recombinant KhACP in the P. pastoris expression system. We have shown that the methylotrophic yeast expression system used is a convenient method to prepare large amounts of stable and functional rKhACP. The recombinant protein is structurally close to

38

T. Yoneyama et al. / Protein Expression and Purification 53 (2007) 31–39

Recombinant KhACP 6

b

120

2

5

100

4

80

3

60

2

40

ACP activity

Log(V0/V0−V)−1

CPO activity ACP activity (%)

CPO activity (µM/min)

a

1 Slope = 0.92 0 −6

−1

Kd = 21.0 × 10

(M)

20

1

−2 −3

0

0 0

50

100

150

−2

Vanadate [µM]

−1

0

Log [V]

Natural KhACP 25

d

120

2

CPO activity

80 15 60 10 40 ACP activity

5

0 50

100

1 Slope = 0.95 0

−1

−6

Kd = 6.64 × 10

(M)

20

0 0

Log(V0/V0−V)−1

100

20

ACP activity (%)

CPO activity (µM/min)

c

150

Vanadate [µM]

−2 −6

−5

−4

−3

Log [V]

Fig. 5. ACP and CPO activities of the vanadate-substituted natural KhACP and rKhACP, and a modified Hill plot. (a and c) After the enzyme was dialyzed against buffer containing EDTA, the enzyme was incubated with various concentration of vanadate for 30 min at 30 C. The activities for ACP (D) using p-NPP and for V-CPO (s) using chlorination of MCD in the presence of H2O2 as substrates were taken to be 100%, respectively. (b and d) Modified Hill plots of CPO activity versus vanadate concentration. The Hill constant was calculated from the linear slope, and the apparent dissociation constant Kd was obtained from the intercept on the horizontal axis at half activity, as indicated by the arrow. V0 and V represent maximum velocity and velocities of the CPO reactions at each concentration of vanadate, respectively.

the 56-kDa homodimer of natural KhACP, and it exhibits a unique vanadate-substituted CPO activity. The data suggest that the enzymatic properties of rKhACP were comparable to those of the natural enzyme isolated from kidney bean hypocotyls, except for a few properties. Our future studies will involve functional analysis of this unique KhACP, utilizing recombinant proteins constructed using sitedirected mutagenesis. We believe that this present expression system will help us to obtain the three-dimensional crystal structure and conversion mechanism of this unique KhACP to vanadate-substituted CPO. Acknowledgments This work was supported by the Research Fellowships of the Japan Society for Promotion of Science (JSPS) for Young Scientist (No. 18Æ6215). The authors thank Masae Shiozawa, Shuhei Okada, Masatomo Etoh, and Wataru Hibino for technical assistance. References [1] S.M.G. Duff, G. Sarath, W.C. Plaxton, The role of acid phosphatase in plant phosphorus metabolism, Physiol. Plant 90 (1994) 791–800.

[2] J.B. Vincent, M.W. Crowder, B.A. Averill, Hydrolysis of phosphate monoesters; a biological problem with multiple chemical solutions, Trends Biochem. Sci. 17 (1992) 105–110. [3] A.R. Penheiter, S.M.G. Duff, G. Sarath, Soybean root nodule acid phosphatase, Plant Physiol. 114 (1997) 597–604. [4] G.G. Bozzo, K.G. Raghothama, W.C. Plaxton, Purification and characterization of two secreted purple acid phosphatase isozymes from phosphate-starved tomato (Lycopersicon esculentum) cell cultures, Eur. J. Biochem. 269 (2002) 6278–6286. [5] G. Schenk, Y. Ge, L.E. Carrington, C.J. Wynne, I.R. Searle, B.J. Carroll, S. Hamilton, J. de Jersey, Binuclear metal centers in plant purple acid phosphatases: Fe–Mn in sweet potato and Fe-Zn in soybean, Arch. Biochem. Biophys. 370 (1999) 183–189. [6] J.B. Vincent, G.L. Oliver-Lilly, B.A. Averill, Proteins containing oxobridged dinuclear iron centers: a bioinorganic perspective, Chem. Rev. 90 (1990) 1447–1467. [7] D. Barford, A.K. Das, M.P. Egloff, The structure and mechanism of protein phosphatase: insights into catalysis and regulation, Annu. Rev. Biomol. Struct. 27 (1998) 133–164. [8] G. Schenk, M.L.J. Korsinczky, D.A. Hume, S. Hamilton, J. de Jersey, Purple acid phosphatases from bacteria: similarities to mammalian and plant enzymes, Gene 255 (2000) 419–424. [9] T. Klabunde, N. Strater, R. Frohlich, H. Witzel, B. Krebs, Mechanism of Fe(III)-Zn(II) purple acid phosphatase based on crystal structures, J. Mol. Biol. 259 (1996) 737–748. [10] J. Goldberg, H.B. Huang, Y.G. Kwon, P. Greengard, A.C. Nairn, J. Kuriyan, Three-dimensional structure of the catalytic subunit of protein serine/threonine phosphatase-1, Nature 376 (1995) 745–753.

T. Yoneyama et al. / Protein Expression and Purification 53 (2007) 31–39 [11] B.P. Gaber, J.P. Sheridan, F.W. Bazer, R.M. Roberts, Resonance Raman scattering from uteroferrin, the purple glycoprotein of the porcine uterus, J. Biol. Chem. 254 (1979) 8340–8342. [12] J.L. Beck, L.A. MaConachie, A.C. Summors, W.N. Arnold, J. de Jersey, B. Zerner, Properties of a phosphatase from red kidney bean: a zinc–iron metalloenzyme, Biochim. Biophys. Acta 869 (1986) 61–68. [13] T. Yoneyama, M. Shiozawa, M. Nakamura, T. Suzuki, Y. Sagane, Y. Katoh, T. Watanabe, T. Ohyama, Characterization of a novel acid phosphatase from embryonic axes of kidney bean exhibiting vanadate-dependent chloroperoxidase activity, J. Biol. Chem. 279 (2004) 37477–37484. [14] H. Liao, F.L. Wong, T.H. Phang, M.Y. Cheung, W.Y. Li, G. Shao, X. Yan, H.M. Lam, GmPAP3, a novel purple acid phosphatase-like gene in soybean induced by NaCl stress but not phosphorus deficiency, Gene 318 (2003) 103–111. [15] M. Olczak, W. Watorek, B. Morawiecka, Purification and characterization of acid phosphatase from yellow lupin (Lupinus luteus) seeds, Biochim. Biophys. Acta 1341 (1997) 14–25. [16] T. Klabunde, B. Stahl, H. Suerbaum, S. Hahner, M. Karas, F. Hillenkamp, B. Krebs, H. Witzel, The amino acid sequence of the red kidney bean Fe(III)–Zn(II) purple acid phosphatase, Eur. J. Biochem. 226 (1994) 369–375. [17] J.M. Cregg, T.S. Vedvick, W.C. Raschke, Recent advances in the expression of foreign genes in Pichia pastoris, Biotechnology (NY) 11 (1993) 905–910. [18] M. Romanos, C. Scorner, K. Sreekrshna, J. Clare, The generation of microscopy recombinant strains, Methods Mol. Biol. 103 (1998) 55–72.

39

[19] J.M. Cregg, K.A. Russell, Transformation, Methods Mol. Biol. 103 (1998) 27–39. [20] P.L. Devine, J.A. Warren, Glycoprotein detection on immobilon PVDF transfer membrane using the periodic acid/Schiff reagent, Biotechniques 8 (1990) 492–495. [21] H. Hirano, T. Watanabe, Microsequencing of proteins electrotransferred onto immobilizing matrices from polyacrylamide gel electrophoresis: application to an insoluble protein, Electrophoresis 11 (1990) 573–580. [22] J.L. Cereghino, J.M. Cregg, Heterologous protein expression in the methylotorophic yeast Pichia pastoris, FEMS Microbiol. Rev. 24 (2000) 45–66. [23] A. Durmus, C. Eichen, B.H. Sift, A. Kratel, R. Kappl, J. Huttermann, B. Krebs, The active site of purple acid phosphatase from sweet potatoes (Ipomoea batatas) metal content and spectroscopic characterization, Eur. J. Biochem. 260 (1999) 709–716. [24] A. Vogel, T. Borchers, K. Marcus, H.E. Meyer, B. Krebs, F. Spener, Heterologous expression and characterization of recombinant purple acid phosphatase from red kidney bean, Arch. Biochem. Biophys. 401 (2002) 164–172. [25] T. Waratrujiwong, B. Krebs, F. Spener, P. Visoottiviseth, Recombinant purple acid phosphatase isoform 3 from sweet potato is an enzyme with a diiron metal center, FEBS J. 273 (2006) 1648–1659. [26] M. Olczak, T. Olczak, Expression and purification of active plant diphosphonucleotide phosphatase/phosphodiesterase from baculovirus-infected insect cells, Protein Expr. Purif. 39 (2005) 116–123.