O R I G I N A L A R T I C L E
Guang-Huar Young Æ Han-Min Chen Æ Chi-Tsai Lin Kuang-Ching Tseng Æ Jiann-Shing Wu
Rong-Huay Juang
Site-specific phosphorylation of L-form starch phosphorylase
by the protein kinase activity from sweet potato roots
Received: 5 May 2005 / Accepted: 29 June 2005 / Published online: 3 September 2005 Springer-Verlag 2005
Abstract A 78-amino acid insertion (L78) is found in the low-affinity type form) of starch phosphorylase (L-SP, EC 2.4.1.1). This insertion blocks the starch-binding site on the L-SP molecule, and it decreases the binding affinity of L-SP toward starch. The computational analysis of the amino acid sequence on L78 predicts several phosphorylation sites at its Ser residues. Indeed, from the immunoblotting results using antibodies against phosphoamino acids, we observed that the purified L-SP from mature sweet potato (Ipomoea ba-tatas) roots is phosphorylated. This observation led us to the detection of a protein kinase activity in the protein fraction of the crude extract from the sweet potato roots. The kinase was partially purified by liquid chromatog-raphy, and its native molecular mass was estimated as 338 kDa. An expressed peptide (L78P) containing the essential part of L78 was intensively phosphorylated by the kinase. However, H-SP (the high-affinity isomer of SP lacking the L78 insertion) and the proteolytic mod-ified L-SP, which lost its L78 fragment, could not be phosphorylated. Furthermore, using L78P mutants by site-directed mutagenesis at Ser residues on L78, we demonstrate that only one Ser residue on L78 is phos-phorylated by the kinase. These results imply that this kinase is specific to L-SP, or more precisely, to the L78 insertion. The in vitro phosphorylated L-SP shows
higher sensitivity to proteolytic modification, but has no change in its kinetic parameters.
Keywords Ipomoea Æ Protein kinase Æ Protein phosphorylation Æ Sweet potato roots Æ Starch phosphorylase
Abbreviations CDPK: Calcium-dependent protein kinase Æ Glc-1-P: Glucose 1-phosphate Æ GP: Glycogen phosphorylase Æ H-SP: High-affinity type SP (Pho2) Æ L-SP: Low-affinity type SP (Pho1) Æ MBP: myelin basic protein Æ PVDF: Polyvinylidene fluoride Æ SP: Starch phosphorylase
Introduction
Starch phosphorylase (SP, EC 2.4.1.1) catalyzes the reversible phosphorolysis of starch and produces glucose 1-phosphate (Glc-1-P) as one of its products (Hanes 1940a,1940b). Plants express two isoforms of SP, which are classified as the low-affinity (L-form SP, L-SP or Pho1) and the high-affinity (H-form SP, H-SP or Pho2) types according to their binding affinities to starch (Nakano and Fukui 1986; Nakano et al. 1989; Mori et al.1991; Albrecht et al.1998). When compared to the amino acid sequence of glycogen phosphorylase (GP), H-SP shows higher homology to GP. In contrast, a peptide containing 78 amino acids (L78) is inserted in the center of L-SP, where the starch-binding site is lo-cated (Mori et al.1993). The L78 might be derived from an intron (Camirand et al. 1990), and its amino acid sequence contains several interesting features, as pdicted by the computer programs, including PEST re-gions, phosphorylation sites, and a polyproline II helix (Chen et al. 2002). These structural features are sum-marized in Fig.1. During the cell division cycle, proteins containing a PEST region are degraded rapidly (Rogers et al.1986). Indeed, most L-SP molecules isolated from mature potato tubers (Brisson et al. 1989) or sweet po-tato roots (Chiang et al. 1991) are proteolytically
mod-H.M. Chen and C.T. Lin contributed equally to this work. G.-H. Young Æ K.-C. Tseng Æ J.-S. Wu Æ R.-H. Juang (&) Department of Biochemical Science and Technology, Institute of Microbiology and Biochemistry,
National Taiwan University, AC 2-520, Taipei 106, Taiwan E-mail: juang@ntu.edu.tw
Tel.: +886-2-33664448 Fax: +886-2-23631704 H.-M. Chen
Department of Life Science, Catholic Fu-Jen University, Taipei 242, Taiwan
C.-T. Lin
Institute of Bioscience and Biotechnology,
National Taiwan Ocean University, Keelung 202, Taiwan DOI 10.1007/s00425-005-0103-1
ified. Nevertheless, L-SP retains its quaternary structure as well as full catalytic activity. On the SDS-PAGE gel, the purified L-SP reveals an intact 110 kDa band (SP110) and a group of bands (F50s) having a molecular mass of approximately 50 kDa, which are produced by the proteolytic modification of L-SP (Chen et al.2002). Glycogen phosphorylase from rabbit muscle has complex regulatory mechanisms including the well-known phosphorylation at the near N-terminal Ser res-idue by its specific protein kinase (Sprang et al. 1988). On the other hand, SP had never been reported as having any regulation mechanism during the past dec-ades, and its physiological role remains unclear (Nelson and Pan 1995), although pea SP has been reported as inhibited by the physiological concentrations of ADP-glucose (Kruger and Ap Rees 1983). However, obser-vations have shown that the accumulation of starch is connected with the increase of SP activity in potato tu-bers (Mingo-Castel et al. 1976; Albrecht et al. 2001), maize endosperm (Liu and Shannon 1981), rice (Baun et al.1970), wheat (Schupp and Ziegler2004), and sweet potato roots (Chang et al. 2000). Starch content is also in proportion to the expression of SP in potato (Brisson et al. 1989; St-Pierre and Brisson 1995; Duwenig et al. 1997; Albrecht et al. 2001), spinach (Duwenig et al. 1997), and pea (Van Berkel et al. 1991). We have re-ported that the proteolytic modification of L-SP on L78 has a regulatory role for its catalytic behavior (Chen et al.2002). But how the proteolysis would be conducted and what the upstream regulatory elements would be are still unknown.
Recently, Tetlow et al. (2004) have reported that wheat L-SP is phosphorylated and could form multi-protein complexes with the phosphorylated starch branching enzymes (SBEI and SBEIIb). The computa-tional analysis of the amino acid sequence reveals that several Ser residues on the L78 fragment might be phosphorylated (Chen et al. 2002). In this study, by using the commercial antibodies against phosphoamino
acids, we found that L-SP isolated from sweet potato roots is phosphorylated. This led us to search for a protein kinase that is responsible for the phosphoryla-tion of L-SP. Consequently, a protein kinase activity was detected and partially purified that specifically phosphorylated L-SP at one Ser residue on L78. The results in this study imply that the phosphorylation of L78 might play a key role in the regulation of the cat-alytic features of L-SP in sweet potato roots.
Materials and methods
Materials and reagents
Mature roots of sweet potato [Ipomoea batatas (L.) Lam. cv Tainong 57] and potato tubers (Solanum tu-berosum L.) were collected from the local farm near Taipei; [c 32P]ATP was obtained from NEN Life Sci-ences, Perkin-Elmer. Mouse monoclonal antibodies against phosphoamino acids (Ser, phospho-Thr, or phospho-Tyr), cellulose phosphate, and protein kinase inhibitor (staurosporine) were from Sigma. The antibody titre has not been determined, and the dilution used for working solutions in immunoblotting followed the directions from Sigma. The antibody against the His tag on the expressed peptides was from Amersham Biosciences (Anti-His antibody). The Polyvinylidene fluoride (PVDF) membrane (Immobilon-P or Immobi-lon-PSQ) was purchased from Millipore. Calf intestine alkaline phosphatase was from Boehringer Mannheim. Starch (1.2% solution, w/v) was prepared by dissolving 1.2 g soluble starch powder (Nacalai Tesque, 321-22, Kyoto, Japan; or Sigma S-9765) in 100 ml warm water (80C). Glucose 1-phosphate (Glc-1-P) was purchased from Sigma (G-7000). All other reagents or equipment are specified in the text. Starch phosphorylases (L-SP and H-SP), monoclonal antibody H7c, and monospecific antibody Anti-L2 were prepared as described previously
Fig. 1 a–c The structural features of glycogen phosphorylase from rabbit muscle (a) and L-form starch phosphorylase from sweet potato (b) are compared. The L78 insertion on L-SP is emphasized and expanded to show its amino acid sequence (Lin et al.1991) and significant features (c) including the PEST region and the phosphorylation target for GP kinase. The range of an expressed peptide (L78P) containing the L78 fragment is marked with red triangle flags. The two Ser residues, which were replaced by Ala in site-directed mutants are indicated with asterisks. The amino acid sequence of the L78 insertion from potato (Nakano and Fukui1986) is aligned, showing 63% homology with the sweet potato L78 (identical amino acid pairs are marked with a vertical dash)
(Chen et al. 2002). Phosphorylated L-SP was prepared in vitro by incubating 200 lg purified L-SP at 37C in a mixture containing 50 mM HEPES (pH 7.0), 5 mM MgCl2, 1 mM ATP, protease inhibitor cocktail, and
50 lg partially purified kinase from this study. After overnight incubation, the mixture was subjected to Su-perose 6 HR10/300 (Pharmacia) to separate the modi-fied L-SP from LSK and ATP. The fractions containing L-SP were collected and concentrated by ultrafiltration (Microcon YM30, Millipore, Bedford, MA, USA). Protein content was determined by the dye-binding method (Bradford 1976) using the microassay system from Bio-Rad (Protein Assay Kit). Bovine serum albu-min was used as the standard.
Polyacrylamide gel electrophoresis, protein transfer, and immunoblotting
The SDS-PAGE was performed according to the method described by Laemmli (1970) with a slight modification. After electrophoresis, the proteins on the gel were transferred to the PVDF membrane that was then stained by antibodies to reveal the target proteins. The detailed protocols are described in our previous report (Chen et al. 2002).
Assays for enzyme activity
The assays for SP in the synthetic and the phosphoro-lytic directions were carried out according to the pro-cedures described by Chen et al. (2002) and Mori et al. (1993), respectively.
The kinase activity was assayed by measuring the incorporation of 32P from [c 32P]ATP to L-SP. The reaction mixture contained 50 mM HEPES (pH 7.0), 5 mM MgCl2, 3.7·104 Bq [c 32P]ATP, and 4 lg L-SP
as the substrate. An appropriate amount of the enzyme solution was added to the reaction mixture for a final volume of 10 ll. This was incubated at 37C for 10 min, terminated by adding to the SDS sample buffer (Lae-mmli 1970), and then denatured at 95C for 10 min. After gel electrophoresis (12.5% SDS-PAGE) at 150 V, the gel was stained with Coomassie Brilliant Blue R-250, and was dried by the cellophane sandwich method (Ju-ang et al.1984). The radioactive bands were exposed on an imaging plate and visualized by an image analyzer (Fujix BAS 1000). Alternatively, the gel was exposed to Amersham Hyperfilm MP directly for a conventional autoradiogram. For quantification of the radioactivity of the phosphorylated proteins, the bands were cut out directly from the gel after the staining and put into the counting vials. After adding the scintillation fluid, the radioactivity was measured in a liquid scintillation counter (Beckman LS5000). The enzyme unit is defined as the femtomole of [c 32P]ATP incorporated into L-SP per minute.
Isolation of L-SP kinase from sweet potato roots
All purification procedures were performed at 4C. Fresh sweet potato roots (150 g) were homogenized in a blender in 150 ml of buffer A (50 mM Tris–HCl, pH 7.4, 150 mM NaCl) containing 1% polyvinylpoly-pyrrolidone (PVPP) and was centrifuged for 20 min at 21,600 g (Beckman Avanti J25, rotor JLA 16.25). The supernatant was filtered through glass wool and the filtrate collected was then fractionated by ammonium sulfate precipitation at a saturation ranging between 20 and 40%. The precipitated protein was collected by centrifugation as above and was resuspended in a minimal volume of buffer A. The crude protein mixture was dialyzed overnight against buffer A. The precipi-tate formed was removed by centrifugation (29,400 g, 40 min) before applying it to a DEAE-Sephacel col-umn (i.d. 2.6·20 cm, pre-equilibrated in buffer A). The column was washed with buffer A, and the kinase activity was eluted directly. The active fractions were pooled and then loaded into a phosphocellulose col-umn (total volume 40 ml, pre-equilibrated in buffer A). This was eluted by a linear Na3PO4 gradient (0–
200 mM in buffer A, total 150 ml); the kinase activity appeared at a concentration of approximately 100 mM Na3PO4. The active fractions were pooled and further
purified by gel filtration using Sephacryl S-300 (i.d. 1.6·100 cm, equilibrated in buffer A). The fractions containing kinase were collected and concentrated by ultrafiltration (Amicon YM-30). This purification pro-tocol allowed us to obtain a partially purified kinase, which was used in all the experiments throughout this study.
In-gel kinase assay for L-SP kinase using non-denaturing gel
Isolated kinase was separated by non-denaturing gel electrophoresis in 10% native polyacrylamide gel con-taining L-SP (0.3 mg/ml). After electrophoresis, the gel was equilibrated for 30 min in a reaction buffer (50 mM HEPES, pH 7.0, 5 mM MgCl2). Then it was incubated
for 2 h at 37C in 10 ml of the reaction buffer containing 7.4·105
Bq of [c 32P]ATP. The reaction was termi-nated by transferring the gel into a bath of 5% trichlo-roacetic acid and 1% Na3PO4. The unincorporated ATP
was removed by washing the gel in the bath solution several times. The radioactive band on the gel was visualized by autoradiography, as described in the assay method for the L-SP kinase activity.
Recombinant L78 DNA preparation and site-directed mutagenesis
The SPc39 (GeneBank accession no. M64362) is a cDNA clone of 3,292 nucleotide residues containing an
open reading frame encoding for the 955 amino acids of L-SP from sweet potato roots (Lin et al. 1991). According to the SPc39 sequence, a forward primer (5¢ GAATTCGGTATCAGAATATGGCACGTC 3¢) and a reverse primer (5¢ CTGACGTGCCATATTCTGATA CCGAATTC 3¢) were synthesized. A 0.2-kb DNA fragment encoding L78 was amplified using SPc39 as a template. The PCR product was ligated with pCR2.1 and then transformed into Escherichia coli TOPO10 (Invitrogen). The 0.2-kb insert DNA was then subcloned into pET-20b(+) (Novagen) and transformed into E. coliBL21 (DE3), which was used for the preparation of the recombinant L78. The crude protein was purified by a Ni+-nitrilotriacetic acid Sepharose superflow column (Qiagen). The expressed polypeptide (L78P) covers the essential part of L78 and its C-terminal flanking se-quence (total 89 amino acids), which is shown in Fig. 1c. L78P was used to immunize BALB/c mice and produce hybridoma secreting monoclonal antibody (mAb anti-L78P) following the protocol described previously (Chen et al.2002).
The recombinant L78 DNA in pET-20b(+) prepared above was used as the wild-type template for site-di-rected mutagenesis using the QuikChange Site-Disite-di-rected Mutagenesis Kit (Stratagene). Three mutants at residues Ser42 or/and Ser71 were prepared, and their sequences were verified by nucleotide sequencing. The expressed proteins [L78P(S42A), L78P(S71A), and L78P(S42A/ S71A)] were isolated and purified by affinity chroma-tography as described above.
Analysis for phosphoamino acids on the phosphorylated L-SP
Free phosphoamino acids can be separated and analyzed by a two-dimensional thin-layer electrophoresis system (Boyle et al. 1991). The radioactive labeled proteins, as described in the assay method for the L-SP kinase, were separated by 12.5% SDS-PAGE and then transferred to the PVDF membranes. The radioactive bands were ex-cised, hydrolyzed in 200 ll of 5.7 N HCl for 1 h at 110C, and then dried to a powder form in a vacuum microcentrifuge. Five microliters of the first-dimension buffer (0.58 M formic acid, 1.36 M glacial acetic acid, pH 1.9) containing the mixture of phosphoamino acid standards (Sigma, P-Ser, P0878; P-Thr, P1053; P-Tyr, P9405; 1.7 lg/ll, respectively) was added to dissolve the powder, which was then applied as a single spot to the DC-Platten cellulose thin-layer plate (Merck 1.05716, 20·20 cm). After wetting the plate with the first-dimension buffer, electrophoresis was performed for 20 min at 1,500 V. The plate was dried completely, and then the second dimension was performed for 16 min at 1,300 V (second-dimension buffer: 0.5% pyridine, 0.87 M glacial acetic acid, pH 3.5). Phosphoamino acid standards were visualized by spraying the plate with 0.25% (w/v) ninhydrin in methanol and heating it at
65C for 30 min. For the radioactive labeled amino acids, the X-ray film was exposed to the plate for 1 d at
80C.
Fig. 2 a Purified L-SP (5 lg per lane) was analyzed by monoclonal antibodies against phospho-Ser (P-Ser), phospho-Thr (P-Thr), or phospho-Tyr (P-Tyr). Molecular weight marker (M) from Phar-macia (LMW electrophoresis calibration kit) was used as the negative control for immunoblotting. One of the markers (phos-phorylase b) was stained slightly. b and c L-SP was phosphorylated in vitro by protein fractions from sweet potato roots using radioactive ATP. Ammonium sulfate fractions of the crude extract at 0–20% (P1), 20–40% (P2), 40–60% (P3), and 60–80% (P4) saturation were used as the source of the protein kinase. After the enzyme reaction, the L-SP was pulled down by a mixture of Protein A-Sepharose and monoclonal antibody H7c (specific to L-SP). The precipitate was analyzed by SDS-PAGE and then transfered to a membrane. Antibody H7c was used again to reveal L-SP bands on the blot (b), which was then exposed on a phospho-image plate, and visualized on the autoradiogram (c). The kinase isolated from sweet potato roots also phosphorylated L-SP from potato tubers (d, 6 lg protein per lane)
Results
The L-SP is found phosphorylated in sweet potato roots or phosphorylated in vitro by a protein fraction from the root extract
The L-SP isolated from the sweet potato roots was separated on SDS-PAGE, transferred to the PVDF membrane and then probed with commercial antibodies against phosphoamino acids. The immunoblotting re-sults (Fig.2a) showed that L-SP in the sweet potato roots is phosphorylated intensively on its Ser or Thr residues, but only faintly on Tyr. Both the intact L-SP (SP110) and the protease-cleaved fragments of L-SP (F50s) are phosphorylated, but the SP110 bands show a higher intensity of staining, indicating that the intact L-SP might be phosphorylated more extensively than the protease-cleaved form in the sweet potato, or the 110 kDa protein is more abundant in the roots. To further verify the phosphorylation of L-SP, we per-formed the cell-free reaction by mixing fractions of ammonium sulfate precipitation from the crude extract
of sweet potato roots with the purified L-SP as the substrate. The fraction of 20 to 40% ammonium sulfate saturation showed the highest phosphorylation activity for L-SP (Fig.2c). In contrast to the phosphorylation detected by monoclonal antibodies (Fig.2a), the prote-olytic modified L-SP (F50s) is phosphorylated more intensively in the in vitro reaction. In addition to the L-SP from sweet potato, the kinase also phosphorylates L-SP from potato (Fig.2d). Notably, the sweet potato L-SP was phosphorylated more intensively on its F50s fragments, whereas the intact SP110 was the major phosphorylation target in the potato L-SP (Fig.2d). In Fig3, the L-SP was phosphorylated by the kinase, stopped by staurosporine, and then the phosphorylation was reversed by calf intestine alkaline phosphatase, which removed most of the labeled radioactive phos-phate from L-SP. The above results indicate the exis-tence of a protein kinase in sweet potato roots, which might phosphorylate L-SP on its Ser or Thr residues.
A protein kinase is partially purified from sweet potato roots
The kinase activity in the sweet potato roots was traced by using L-SP as the substrate in a purification protocol. A putative L-SP kinase was partially purified by con-ventional liquid chromatography (Fig.4) starting from the fraction of 20 to 40% saturation of ammonium sulfate. The major kinase activity flowed through the DEAE ion exchange column directly and split into two peaks on the chromatogram (Fig.4a). We repeated this operation several times and confirmed that these two peaks are actually from one larger peak, whose activity is affected by an unknown factor eluted coincidently. The final purification table (Table 1) shows an increasing specific activity through the whole purification protocol except in the cellulose phosphate step. This is due to the inhibitory effect of the phosphate used in the gradient elution for the adsorbed proteins. Nevertheless, the activity could be fully recovered by removing the phos-phate from the sample in the next gel filtration step. Totally, the enzyme was purified by 500-fold and 36% activity was recovered. Its native molecular mass was determined as 338 kDa by gel filtration (Fig.4d). This partially purified kinase was visualized by the in-gel ki-nase assay using non-denaturing PAGE that contained L-SP as the substrate in the gel (Fig.4d, inset).
L-SP is phosphorylated specifically on a Ser residue of L78
To examine the substrate specificity of the kinase, par-tially purified H-SP, the extensively proteolytic modified L-SP (L-SP*), and an expressed peptide (L78P) con-taining L78 were used as the alternative substrates for comparison. H-SP is the high-affinity isoform of SP, which shows higher homology to GP. It contains no L78
Fig. 3 The phosphorylation of L-SP by the protein kinase (closed circles) and the subsequent dephosphorylation by calf intestine alkaline phosphatase (open circles) were monitored during the reaction. Partially purified kinase (2.8 lg) was added to 50 lg of L-SP in a total volume of 100 ll. Phosphorylation was terminated after 20 min by adding the kinase inhibitor staurosporine (1 lg/ll, indicated by an arrow), and then the calf intestine alkaline phosphatase (five units) was added to remove the phosphate group on the protein. The autoradiogram shows the accumulation and disappearance of the phosphorylated bands at selected time points (10 ll aliquot was sampled for each lane). One aliquot of the original phosphorylation mixture was kept to complete the whole reaction after the addition of staurosporine (closed circle at 35 min)
in its molecule and has a smaller molecular mass of 97 kDa (Mori et al. 1993). L-SP* was prepared by removing the L78 fragment from L-SP by controlled
proteolysis (Chen et al. 2002). Figure 5 shows that no phosphorylation was found on H-SP and L-SP*. On the other hand, the kinase intensively phosphorylated the expressed peptide L78P that contains the L78 fragment. To determine the amino acid that was phosphory-lated by this kinase, protein substrates including L-SP, L78P and myelin basic protein (MBP) were phosphor-ylated by using radioactive ATP. The phosphorphosphor-ylated proteins were separated by SDS-PAGE and transferred to the PVDF membrane, which were then cut out for analysis. SP110 and F50s were collected separately for L-SP. They were then subjected to acid hydrolysis and separated by two-dimensional electrophoresis as de-scribed in the experimental procedures. Figure6reveals that phospho-Ser is the major phosphoamino acid in all the cases. A trace amount of phospho-Thr was observed in SP110, F50s, and MBP. Notably, L78P shows only phospho-Ser. The above observations imply that L-SP is the specific target for the kinase that could phosphory-late Ser residue(s) on L78.
The exact phosphorylation site on L-SP was deter-mined by using mutants of L78P by site-directed muta-genesis, in which two key Ser residues were replaced by Ala: L78P(S42A), L78P(S71A), and L78P(S42A/S71A). The two residues were selected because S42 is flanked by a predicted sequence of GP kinase target (KETSI) (Chan et al. 1982), and Ser71 was predicted as the strongest phosphorylated site on L78 (Chen et al.2002). The amino acid sequence of L78 is aligned in Fig.1c, which is numbered from 1 to 78 using underlined numbers (in the total L-SP molecule, L78 starts from Val 414 as the numbering system of rabbit GP is em-ployed). Figure7 shows that wild-type L78P and L78P(S42A) are phosphorylated intensively, but muta-tion at Ser71 totally suppresses the phosphorylamuta-tion. This result suggests that Ser71 is the only residue being phosphorylated on L78.
The phosphorylated L-SP is sensitive to proteolytic modification, but has no change in its kinetic parameters
The crude extract from the sweet potato roots was mixed with the in vitro phosphorylated L-SP and incubated to
Fig. 4 The kinase was isolated by ammonium sulfate fractionation first and then subjected to liquid chromatography using DEAE Sephacel (a), cellulose phosphate (b), and Sephacryl S-300 (c). Kinase activity was determined by measuring the radioactivity of the L-SP bands on SDS-PAGE gel. Since L-SP was resolved into the intact subunit (SP110) and a group of fragments with molecular mass around 50 kDa (F50s), their radioactivity was measured separately (SP110, close circles; F50s, open squares). The molecular mass of the putative L-SP kinase was estimated as 338 kDa by gel filtration using Sephacryl S-300 (d). The molecular weight markers (Bio-Rad, 151-1901) are thyroglobulin (670 kDa), bovine gamma globulin (158 kDa), chicken ovalbumin (44 kDa), equine myoglo-bin (17 kDa), and vitamin B12 (1,350 Da). The kinase eluted from the gel filtration was collected and concentrated, and then a fraction of 2 lg protein was subjected to the in-gel kinase assay (inset in d). Representative data shown (n=6)
observe the effect of phosphorylation on L-SP. It was found that the phosphorylated enzyme was degraded after 12 h of incubation (Fig. 8). The intact SP110 dis-appeared completely, and the F50s showed an increase of its electrophoretic mobility on SDS-PAGE gel as re-vealed by using mAb J3b, which recognizes an epitope on the N-terminal side of the protein (Fig. 8b). On the other hand, mAb anti-L78P was prepared by using L78P as the antigen, and it might recognize a specific epitope on the L78 of L-SP. If anti-L78P was used instead of mAb J3b, a distinct contrast was observed (Fig.8c). Anti-L78P cannot stain any band on the blot after 15 h, indicating that the L78 insertion might be removed completely during the proteolytic modification induced by the phosphorylation of L-SP; otherwise the unmod-ified L-SP is unaffected by the crude extract throughout the incubation (left panels in Fig.8). Apparently, the crude extract might contain an unidentified proteolytic
system, which could specifically recognize the phos-phorylated form of L-SP and cleave the enzyme on its L78 insertion. Further study is needed to elucidate this effect.
The phosphorylation did not significantly affect the kinetic parameters (Km or Kcat) of L-SP, either in the
direction of phosphorolysis or amylose synthesis (Ta-ble2). The effect of phosphorylation on L-SP is there-fore very different from the phosphorylation on the Ser14 of GP, which triggers the change of protein con-formation and thereby activates the enzyme (Sprang et al. 1988). Interestingly, the ratio of the Kcat in the
synthetic direction over the phosphorolytic direction is about five. This observation correlates with our obser-vation that when Glc-1-P is used as the exclusive sub-strate and catalyzed by L-SP for the synthesis of amylose, the polymerization reaction achieved an equi-librium state where the ratio of the product (Pi) to the
Fig. 5 The catalytic specificity of the kinase was examined on SDS-PAGE gel (12.5%) by using L-SP and H-SP as the substrates. Modified protein (L-SP*) or expressed peptide derived from L-SP (L78P) were used for comparison. Neither H-SP nor L-SP* could be phosphorylated. L-SP* contained two large peptides after the controlled proteolytic modification on L-SP (denoted with N and C), which have been identified by amino acid sequencing. The N-terminal sequence of the C-peptide was determined as KDEDPVPAPI that is indicated in Fig.1c. The H-SP used in this experiment was not homogeneous, but the 97-kDa band on the gel has been identified as H-SP by tryptic digestion and mass spectroscopic analysis
Fig. 6 Several substrate proteins (L-SP, L78P, and MBP, 5 lg each) were phosphorylated using [c 32P]ATP, and then the phosphoamino acids were analyzed by two-dimensional thin-layer electrophoresis. Results show that Ser is the primary phosphorylation target on L-SP. S phospho-Ser, T phospho-Thr, Piinorganic
phosphate, ‘‘ x’’ marks the origin spot of the sample, and the asterisks denote
phosphopeptides by partial hydrolysis
remaining substrate (Glc-1-P) was approximately five (data not shown).
Discussion
In this study, we show that L-SP isolated from sweet potato roots is phosphorylated, as observed by using antibodies against phosphoamino acids (Fig.2a). The enzymatic phosphorylation activity was further recon-stituted in vitro by mixing the protein fractions extracted from sweet potato roots with the purified L-SP (Fig.2c). In a time-course study, the kinase activity was termi-nated by the protein kinase inhibitor staurosporine, and the phosphorylated L-SP was then dephosphorylated by the alkaline phosphatase from the calf intestine (Fig.3). Therefore, we conclude that a portion of L-SP has been phosphorylated in the cell before it is isolated from sweet potato roots, and a protein kinase activity is in existence in the root extract. On the other hand, some isolated L-SP might remain unphosphorylated and could be labeled with radioactive phosphate in the test tube. This is re-flected in the distinct patterns of intensity between the immunoblot (Fig.2a) and the autoradiogram (Fig.2c). However, unidentified protein phosphatase might be present in the sweet potato roots, and the phosphate group of the phosphorylated L-SP could be removed during the extraction procedure for L-SP. We then iso-lated this kinase (Fig. 4 and Table1), which could be visualized by the in-gel kinase assay using a
non-dena-turing gel, and showed its low mobility on the gel (inset in Fig.4d), which might be caused by the high affinity between the kinase and L-SP. However, we were not able to visualize a clear protein band of the kinase on SDS-PAGE gel owing to the low abundance of the en-zyme in the cell. Besides, the in-gel assay for the kinase activity in SDS-PAGE was not successful. This may be due to the nature of the L-SP structure, whose confor-mation might be sensitive to SDS.
If the L78 was removed from L-SP by a controlled proteolytic modification, the remaining peptide chains still keep the intact quaternary structure and full cata-lytic activity (Chen et al.2002). However, this modified L-SP could not be phosphorylated by the kinase (L-SP* in Fig.5). Similarly, the H-SP isoform contained no L78 when it was biosynthesized and was not phosphorylated. Therefore, we might regard this kinase as specific to the L78 insertion on the L-SP molecule, although it also phosphorylated MBP, which is a broad-spectrum sub-strate for most protein kinases (Fig.6). There are ten Ser and five Thr residues on L78. Analysis for the hydroly-sate of the phosphorylated proteins showed that L-SP was intensively phosphorylated on Ser and slightly on Thr residue(s) (Fig.6). This result supports the obser-vation of using antibodies against phospho-amino acids (Fig.2a) and is consistent with the observation for the phosphorylation of L-SP in wheat (Tetlow et al.2004). The exact phosphorylation site was further determined by single amino acid substitutions at two Ser residues on L78. To our surprise, only Ser71 was phosphorylated,
Fig. 7 a–d Mutants of L78P by site-directed mutagenesis replacing Ser residue(s) with Ala were constructed. The expressed peptides (3 lg per lane) were phosphorylated by the kinase using [c 32P]ATP and then separated on SDS-PAGE gel, which was stained with Coomassie
Brilliant Blue (a), or transferred to the PVDF membrane and then detected by Anti-His antibody (c) or monospecific antibody Anti-L2 (d). Anti-L2 was prepared by immunizing mice with a synthetic peptide L2, which contained 23 amino acids starting from Lys66 on L78. The Coomassie Brilliant Blue stained gel was dried and then exposed on an imaging plate to visualize the phosphorylated peptides (b). No phosphorylation was observed on mutants whose Ser71 residues were replaced by Ala. The two blots (c and d) look like they have been mistakenly placed upside down; after a careful examination and duplicate experiment, the bands were proved to be stained like that. This might be caused by the small molecular mass of L78P, which contains a high percentage of charged amino acids, and a relatively high concentration of the SDS gel (17.5%)
and Ser42 was not a target (Fig.7). Notably, Ser42 is covered in the PEST sequence, as well as a consensus sequence (KXXSI, Fig. 1c), which was predicted as a specific target for GP kinase in mammals (Chan et al. 1982). The GP kinase from rabbit muscle is classified as a member of the calcium-dependent protein kinase (CDPK) subfamily (Hardie 1999). Several CDPK type kinases have been found in plants. For example, sucrose synthase from maize leaf is phosphorylated by a CDPK, and the Km values for its substrates are reduced by
fivefold (Huber et al. 1996). The family class of the ki-nase in this study is unknown at present, which could be predicted from the amino acid sequence deciphered from its cDNA sequence.
The L-SP (Pho1a) from potato tubers also contains the L78 insertion, and it could be phosphorylated by the kinase isolated from sweet potato roots (Fig. 2d). These two plant enzymes share high homology in amino acid
sequences (88% similarity and 84% identity). However, sequences of their L78 insertions show only 63% iden-tity as compared in Fig.1c. In addition, the corre-sponding Ser71 position on the potato L78 is replaced by an Asp. The Ser42 residue is conserved on potato L78, but the KXXSI sequence flanking Ser42 has been changed (-XXSV) in which the Lys residue is deleted, and the Ile is replaced by Val. This suggests that potato L-SP might be phosphorylated by an alternative mech-anism. The exact site for phosphorylation on the potato L-SP remains to be clarified.
The results allow us to conclude that L-form SP is site-specifically phosphorylated by a protein kinase in plants; what physiological role might it play? First, one possible effect is the modification of its catalytic behavior due to the conformational change induced by phosphorylation. Our observations on the kinetic parameters for the phosphorylated L-SP decline this
Fig. 8 a–c The phosphorylated L-SP was sensitive to
proteolytic modification as compared with the unmodified L-SP in the reconstitute reaction. Fresh crude extract from the sweet potato roots was mixed with the L-SP
(unmodified or
phosphorylated), and their final protein concentrations were adjusted to 1.5 lg/ll. The reactions were incubated at 37C for 24 h. Samples of 15 ll aliquot from the reaction mixture were collected every 3 h and then subjected to 12.5% SDS-PAGE. The gels were either stained with Coomassie Brilliant Blue (a) or transferred to PVDF membranes that were then stained with monoclonal antibodies J3b (b) or anti-L78P (c)
Table 1 The purification of the L-SP kinase from sweet potato roots Purification steps Total activity
(units) Total protein (mg) Specific activity (units mg 1) Fold purification Activity recovered Crude extraction 342 1,845 0.19 1.0 100
Ammonium sulfate fractionation (20–40% saturation)
418 835 0.50 2.6 122
DEAE Sephacel 236 105 2.25 12.0 69
Sephacryl S-300 123 1.3 95 500 36
Purification was started from 150 g of fresh sweet potato roots (about 5 cm in diameter) One enzyme unit is defined as one fmole of [c 32P]ATP incorporated in L-SP per minute
possibility (Table 2). Second, site-specific phosphoryla-tion was reported to target the inhibitor of a transcrip-tion factor (IjBa) to the ubiquitin-proteasome pathway (Chen et al. 1996), which leads to the proteolytic mod-ification of the protein. Our previous work has reported that the proteolytic modification of the L78 on L-SP might regulate the catalytic direction of the enzyme (Chen et al. 2002). Besides, two destruction boxes (RXXLXXXXN) near the N-terminus were identified on the L-SP from sweet potato as shown in Fig. 1b. Protein with destruction box has been reported to be degraded through the ubiquitin-dependent proteolysis (Glotzer et al. 1991). Indeed, we have detected protea-some in the crude extract of mature sweet potato roots, which might be associated with L-SP (Chang et al. 2000). Furthermore, the reconstitute experiment that mixed the in vitro phosphorylated L-SP with the crude extract from sweet potato roots shows that phosphory-lation might direct the enzyme to the proteolytic modi-fication, which removes the L78 from the L-SP completely (Fig. 8). Third, the recent report by Zeeman et al. (2004) suggests that plastidial a-glucan phosphor-ylase is not required for starch degradation but plays a role in the tolerance of abiotic stress in Arabidopsis leaves. Tetlow et al. (2004) showed the phosphorylation of SP in wheat and the formation of multiprotein com-plexes with other enzymes related to starch biosynthesis. These observations suggest a wider role for the phos-phorylation of L-SP in the control of starch metabolism. To our knowledge, L-SP is the only enzyme catalyzing the synthesis of a glucan from monosaccharide (Glc-1-P) in the absence of a pre-existing primer (data not shown). Moreover, Albrecht et al. (2001) have reported that an antibody specific to the L78 insertion on the potato functional dimer of L-SP (Pho1a)2 almost completely
impedes the capacity for the unprimed glucan synthesis of L-SP. Their finding implies that L78 might involve
itself in the biosynthesis of starch. The above observa-tions might suggest a key role for the phosphorylation of L78 in the regulation of L-SP catalytic activity through the proteolytic pathway, which might subsequently in-volve in controlling the regulatory network of starch metabolism in plants. Observations in this study propose a possible role for the L-SP and its phosphorylation, which does not infringe the current view of starch bio-synthesis in plants (Ball et al.1996).
Acknowledgments We thank Drs. D.S. Bendall and B.G. Schlarb-Ridley (University of Cambridge); Dr. S.C. Chang (Harvard Medical School); Drs. Y.Y. Charng and C.S. Chen (Academia Si-nica, ROC); Dr. C.M. Chiang (Case Western Reserve University); Drs. P.L. Huang, A.Y. Wang, and C.C. Yang (National Taiwan University); and Dr. A.B. Tobin (University of Leicester) for their helpful suggestions and discussions. We also thank Dr. J. Hook for reading the article. Part of this work was supported by the National Science Council, ROC.
References
Albrecht T, Greve B, Pusch K, Kossmann J, Buchner P, Wobus U, Steup M (1998) Homodimers and heterodimers of Pho1-type phosphorylase isoforms in Solanum tuberosum L as revealed by sequence-specific antibodies. Eur J Biochem 251:343–352 Albrecht T, Koch A, Lode A, Greve B, Schneider-Mergener J,
Steup M (2001) Plastidic (Pho1-type) phosphorylase isoforms in potato (Solanum tuberosum L.) plants: expression analysis and immunochemical characterization. Planta 213:602–613 Ball S, Guan HP, James M, Myers A, Keeling P, Mouille G,
Bu-leon A, Colonna P, Preiss J (1996) From glycogen to amylo-pectin—a model for the biogenesis of the plant starch granule. Cell 86:349–352
Baun CC, Palmiano EP, Perez CM, Juliano BO (1970) Enzymes of starch metabolism in the developing rice grain. Plant Physiol 46:429–434
Boyle WJ, Van der Geer P, Hunter T (1991) Phosphopeptide mapping and phosphoamino acid analysis by two-dimensional separation on thin-layer cellulose plates. Methods Enzymol 201:110–149
Table 2 Enzyme kinetic analysis of the L-SP and its phosphorylated form L-SP Synthetic direction (Chen et al.2002)
Km Kcat(l/s)
Soluble starch (%, w/v) Glc-1-P (mM) Fixed [Glc-1-P]a Fixed [soluble starch]b
Unmodified 0.077 1.052 100.1 99.2
Phosphorylated 0.070 1.090 98.6 97.1
Phosphorolytic direction (Mori et al.1993)
Km Kcat
Soluble starch (%, w/v) Pi(mM) Fixed [Pi]c Fixed [soluble starch]d
Unmodified 0.115 1.498 22.39 18.46
Phosphorylated 0.108 1.443 21.92 18.06
Data were taken from the average of three experiments
a [Glc-1-P]=4 mM b[soluble starch]=0.3% c[P i]=5 mM d [soluble starch]=0.2%
Bradford MM (1976) A rapid and sensitive method for the quan-titation of microgram quantities of protein utilizing the prin-ciple of protein-dye binding. Anal Biochem 72:248–254 Brisson N, Giroux H, Zollinger M, Camirand A, Simard C (1989)
Maturation and subcellular compartmentation of potato starch phosphorylase. Plant Cell 1:559–566
Camirand A, St-Pierre B, Marineau C, Brisson N (1990) Occur-rence of a copia-like transposable element in one of the introns of the potato starch phosphorylase gene. Mol Gen Gene 224:33–39
Chan KF, Hurst MO, Graves DJ (1982) Phosphorylase kinase specificity. A comparative study with cAMP-dependent protein kinase on synthetic peptides and peptide analogs of glycogen synthase and phosphorylase. J Biol Chem 257:3655–3659 Chang SC, Lin PC, Chen HM, Wu JS, Juang RH (2000) The
iso-lation and characterization of chaperonin 60 from sweet potato roots—involvement of the chaperonins in starch biosynthesis. Bot Bull Academia Sin 41:105–111
Chen ZJ, Parent L, Maniatis T (1996) Site-specific phosphorylation of IjBa by a novel ubiquitin-dependent protein kinase activity. Cell 84:853–862
Chen HM, Chang SC, Wu CC, Cuo TS, Wu JS, Juang RH (2002) Regulation of the catalytic behaviour of L-form starch phos-phorylase from sweet potato roots by proteolysis. Physiol Plant 114:506–515
Chiang C-L, Lu Y-L, Juang RH, Lee P-D, Su JC (1991) Native and degraded forms of sweet potato starch phosphorylase. Agric Biol Chem 55:641–646
Duwenig E, Steup M, Kossmann J (1997) Induction of genes encoding plastidic phosphorylase from spinach (Spinacia oler-acea L.) and potato (Solanum tuberosum L) by exogenously supplied carbohydrates in excised leaf discs. Planta 203:111–120 Glotzer M, Murray AW, Kirschner MW (1991) Cyclin is degraded
by the ubiquitin pathway. Nature 349:132–138
Hanes CS (1940a) The breakdown and synthesis of starch by an enzyme system from pea seeds. Proc R Soc (Lond) B 128:421–450 Hanes CS (1940b) The reversible formation of starch from glucose-1-phosphate catalyzed by potato phosphorylase. Proc R Soc (Lond) B 129:174–208
Hardie DG (1999) Plant protein serine/threonine kinases: classifi-cation and functions. Annu Rev Plant Physiol Plant Mol Biol 50:97–131
Huber SC, Huber JL, Liao PC, Gage DA, McMichael RW Jr, Chourey PS, Hannah LC, Koch K (1996) Phosphorylation of serine-15 of maize leaf sucrose synthase. Occurrence in vivo and possible regulatory significance. Plant Physiol 112:793–802 Juang RH, Chang YD, Sung HY, Su JC (1984) Oven-drying
method for polyacrylamide gel slab packed in cellophane sandwich. Anal Biochem 141:348–350
Kruger NJ, Ap Rees T (1983) Properties of a-glucan phosphorylase from pea-chloroplasts. Phytochemistry 22:1891–1898
Laemmli UK (1970) Cleavage of structural proteins during the assembly of the head bacteriophage T4. Nature 227:680–685 Lin CT, Yeh KW, Lee PD, Su JC (1991) Primary structure of sweet
potato starch phosphorylase deduced from its cDNA sequence. Plant Physiol 95:1250–1253
Liu TTY, Shannon JC (1981) A nonaqueous procedure for iso-lating starch granules with associated metabolites from maize endosperm. Plant Physiol 67:518–524
Mingo-Castel AM, Young RE, Smith OE (1976) Kinetin-induced tuberization of potato in vitro: on the mode of action of kinetin. Plant Cell Physiol 17:557–570
Mori H, Tanizawa K, Fukui T (1991) Potato tuber type H phos-phorylase isoenzyme. Molecular cloning, nucleotide sequence, and expression of a full-length cDNA in Escherichia coli. J Biol Chem 266:18446–18453
Mori H, Tanizawa K, Fukui T (1993) A chimeric a-glucan phos-phorylase of plant type L and H isozymes: functional role of 78-residue insertion in type L isozyme. J Biol Chem 268:5574–5581 Nakano K, Fukui T (1986) The complete amino acid sequence of
potato a-glucan phosphorylase. J Biol Chem 261:8230–8236 Nakano K, Mori H, Fukui T (1989) Molecular cloning of cDNA
encoding potato amyloplast a-glucan phosphorylase and the structure of its transit peptide. J Biochem (Tokyo) 106:691–695 Nelson O, Pan D (1995) Starch synthesis in maize endosperms.
Annu Rev Plant Physiol Plant Mol Biol 46:475–496
Rogers S, Wells R, Rechsteiner M (1986) Amino acid sequences common to rapidly degraded proteins: the PEST hypothesis. Science 234:364–368
Schupp N, Ziegler P (2004) The relation of starch phosphorylases to starch metabolism in wheat. Plant Cell Physiol 45:1471–1484 Sprang SR, Acharya KR, Goldsmith EJ, Stuart DI, Varvill K, Fletterick RJ, Madsen NB, Johnson LN (1988) Structural changes in glycogen phosphorylase induced by phosphoryla-tion. Nature 336:215–221
St-Pierre B, Brisson N (1995) Induction of the plastidic starch-phosphorylase gene in potato storage sink tissue. Planta 195:339–344
Tetlow IJ, Wait R, Lu Z, Akkasaeng R, Bowsher CG, Esposito S, Kosar-Hashemi B, Morell MK, Emes MJ (2004) Protein phosphorylation in amyloplasts regulates starch branching en-zyme activity and protein-protein interactions. Plant Cell 16:694–708
Van Berkel J, Conrads Strauch J, Steup M (1991) Glucan-phos-phorylase forms in cotyledons of Pisum sativm L.: localization, developmental change, in vitro translation, and processing. Planta 185:432–439
Zeeman SC, Thorneycroft D, Schupp N, Chapple A, Weck M, Dunstan H, Haldimann P, Bechtold N, Smith AM, Smith SM (2004) Plastidial a-glucan phosphorylase is not required for starch degradation in Arabidopsis leaves but has a role in the tolerance of abiotic stress. Plant Physiol 135:849–858
![Fig. 6 Several substrate proteins (L-SP, L78P, and MBP, 5 lg each) were phosphorylated using [c 32 P]ATP, and then the phosphoamino acids were analyzed by two-dimensional thin-layer electrophoresis.](https://thumb-ap.123doks.com/thumbv2/9libinfo/8854208.243290/7.892.266.823.767.1066/substrate-proteins-phosphorylated-using-phosphoamino-analyzed-dimensional-electrophoresis.webp)
