Cytosolic pyruvate kinase (ATP:pyruvate 2-O-phosphotransferase, EC 22.214.171.124) is an important glycolytic enzyme, but the post-translational regulation of this enzyme is poorly understood. Sequence analysis of the soybean seed enzyme suggested the potential for two phosphorylation sites: site-1 (FVRKGS220DLVN) and site-2 (VLTRGGS407TAKL). Sequence- and phosphorylation state-specific antipeptide antibodies established that cytosolic pyruvate kinase (PyrKinc) is phosphorylated at both sites in vivo. However, by SDS–PAGE, the phosphorylated polypeptides were found to be smaller (20–51 kDa) than the full length (55 kDa). Biochemical separations of seed proteins by size exclusion chromatography and sucrose-density gradient centrifugation revealed that the phosphorylated polypeptides were associated with 26S proteasomes. The 26S proteasome particle in developing seeds was determined to be of approximately 1900 kDa. In vitro, the 26S proteasome degraded associated PyrKinc polypeptides, and this was blocked by proteasome-specific inhibitors such as MG132 and NLVS. By immunoprecipitation, we found that some part of the phosphorylated PyrKinc was conjugated to ubiquitin and shifted to high molecular mass forms in vivo. Moreover, recombinant wild-type PyrKinc was ubiquitinated in vitro to a much greater extent than the S220A and S407A mutant proteins, suggesting a link between phosphorylation and ubiquitination. In addition, during seed development, a progressive accumulation of a C-terminally truncated polypeptide of approximately 51 kDa was observed that was in parallel with a loss of the full-length 55 kDa polypeptide. Interestingly, the C-terminal 51 kDa truncation showed not only pyruvate kinase activity but also activation by aspartate. Collectively, the results suggest that there are two pathways for PyrKinc modification at the post-translational level. One involves partial C-terminal truncation to generate a 51 kDa pyruvate kinase subunit which might have altered regulatory properties and the other involves phosphorylation and ubiquitin conjugation that targets the protein to the 26S proteasome for complete degradation.
Pyruvate kinase (ATP:pyruvate 2-O-phosphotransferase, EC 126.96.36.199) is an enzyme of glycolysis that catalyzes the irreversible conversion of phospho(enol)pyruvate (PEP) and ADP to pyruvate and ATP. Because pyruvate is a key metabolic intermediate of many pathways, such as energy production and the biosynthesis of amino acids, organic acids, and fatty acids, there is much interest in elucidating the regulatory and kinetic properties of pyruvate kinase. Allosteric regulation of eukaryotic pyruvate kinase is common, but the effectors of pyruvate kinase in plants appear to be different from those in animals (Plaxton, 1996). In mammalian liver, PyrKinc is activated by fructose-1,6-P2, the product of ATP-linked phosphofructokinase (PFK), whereas plant PyrKinc is not. Rather, PEP – the substrate of PyrKinc– is an inhibitor of plant PFK (Plaxton, 1996). In addition, amino acids are important allosteric effectors of plant PyrKinc. A recent study involving rapeseed identified aspartate as an activator that could reverse the inhibition by glutamate (Smith et al., 2000) which provided a link between C and N metabolism.
We were interested in the role of PyrKinc in developing soybean seeds. Soybean is one of the most important economic crop plants in the world, and the regulation of carbon partitioning between storage protein and oil production is a central issue for crop improvement. It is currently thought that sucrose, taken by developing seed, is converted to PEP via glycolysis in the cytoplasm. The PEP can either be transported into plastids to support fatty acid biosynthesis (Ruuska et al., 2002) or converted to pyruvate in the cytoplasm by PyrKinc for energy production and organic acids required for amino acid interconversion for storage protein biosynthesis. Therefore, PyrKinc is a branch-point enzyme and its regulation may be one of the factors controlling storage product accumulation. Earlier studies suggested that PyrKinc activity is correlated positively with both storage protein and oil biosynthesis in developing soybean seeds (Platt and Bassham, 1978; Smith et al., 1989). While these experimental findings are not consistent with the proposed role of PyrKinc as a branch-point enzyme, they do support a role for this enzyme in storage product biosynthesis and thus it is important to identify the biological mechanisms controlling activity. While allosteric control of PyrKinc is known (Smith et al., 2000), it remains unclear whether control by reversible phosphorylation occurs in vivo as it occurs in mammals.
In the present study, we investigated the phosphorylation of PyrKinc and its regulation in developing soybean seeds. By sequence analysis, we found that the enzyme has two potential phosphorylation sites: FVRKGS220DLVN and VLTRGGS407TAKL. Both sites match the minimal consensus motif (φ-X-Basic-X-X-S/T, where φ is a hydrophobic residue) that is recognized by calcium-dependent protein kinases (CDPKs) and SnRK1 (SNF1 related) kinases in plants (Halford and Hardie, 1998; Huang and Huber, 2001). Sequence- and phosphorylation-specific antibodies were produced for each potential site in PyrKinc in order to monitor phopshorylation in vitro and in vivo. Similar affinity-purified antibodies produced against synthetic phosphopeptides have been used to monitor the phosphorylation of specific sites in other plant enzymes, including PEP carboxylase (Ueno et al., 2000) and sucrose synthase (Hardin et al., 2002). The results obtained in the present study suggest that there are two mechanisms for the post-translational modification of PyrKinc. The first is a C-terminal truncation of the 55 kDa full-length PyrKinc to form an approximately 51 kDa polypeptide that increased during seed development and might alter regulatory properties. The second mechanism is phosphorylation that could target the enzyme to the ubiquitin/26S proteasome pathway for degradation.
Pyruvate kinase consists of 51 and 55 kDa subunits in soybean seed
An immunological approach was taken to monitor PyrKinc during seed development. An N-terminal antibody (Ab-NT) was produced to detect the PyrKinc polypeptide, and sequence- and phosphorylation-specific antibodies were produced that specifically recognized the phosphoserine −220 (Ab-pSer220) or phosphoserine −407 (Ab-pSer407) sites (Figure 1a). Initially, we probed PyrKinc in protein extracts from seeds (Glycine max (L.) cv. Ransom) at different stages of development by immunoblot analysis with Ab-NT. To minimize the possible proteolysis and de-phosphorylation during protein extraction, the frozen powdered seeds were homogenized directly in SDS sample buffer supplemented with a cocktail of protease and phosphatase inhibitors. We found two major polypeptides detected by Ab-NT: one was a 55 kDa full-length PyrKinc and the other was approximately of 51 kDa . During seed development, the 55 kDa full-length PyrKinc decreased, while the 51 kDa polypeptide progressively increased (as shown in Figure 1b). Similar developmental profiles were observed in three other cultivars (data not shown).
Previously, only one cDNA encoding PyrKinc from soybean somatic embryos was available in the public database. The full-length deduced sequence (GB# Q42806) encoded a protein of 511 amino acid. However, no complete open reading frame cDNA sequences for PyrKinc were reported from developing soybean seeds. By RT-PCR, we cloned two cDNAs for PyrKinc from developing seeds of Glycine max (L.) cv. Ransom, herein named PyrKinc1 and PyrKinc2. PyrKinc1 (GB# AY128951) was 511 amino acid in length and PyrKinc2 (GB# AY128952) was 510 amino acid in length. The identity of each to the original sequence for soybean PyrKinc (GB# Q42806) is 99.4 and 98%, respectively, and both share very high identity with cDNAs for PyrKinc in other plant species (Figure 1d). The theoretical molecular weights of the two PyrKinc isoforms are approximately 55 kDa. The transcriptional profiles of the two PyrKinc genes were examined by RT-PCR using isoform-specific primers under high-stringency PCR conditions. A dominant expression pattern of PyrKinc1 was observed (Figure 1e), and both declined in abundance as seed development proceeded.
In order to determine if the 51 kDa polypeptide resulted from an independent transcription event or alternative splicing, we performed a nested RNA gel blot analysis with two different 5′-region cDNAs (378 and 719 bp) and the full-length sequence as templates for probe labeling. The subsequent hybridization was carried out under low-stringency conditions. Because the 51 kDa polypeptide was immunologically related to the full-length PyrKinc (Figures 1b and 3e), it was reasonable to anticipate that RNAs for the 55 kDa full-length and the 51 kDa truncated polypeptide might be detected in the RNA gel blots, especially in RNA samples from the 9- and 10-week-old seeds that had the highest level of the 51 kDa polypeptide (Figure 1b). However, we could only detect a signal band corresponding to the predicted 55 kDa PyrKinc subunit (Figure 1c). No signal was observed in RNA gel blots of samples from the 9- and 10-week-old seeds, consistent with the results obtained by RT-PCR (Figure 1e). In addition, because the 51 kDa polypeptide continued to increase from 8 to 10 weeks while the 55 kDa subunit decreased and transcripts were essentially undetectable, these experiments excluded the possibility that the 51 kDa polypeptide was produced from either an independent transcriptional event or an alternative splicing. Rather, it seemed that the 51 kDa polypeptide might be derived from the 55 kDa form at the post-translational level by C-terminal cleavage.
The 51 kDa polypeptide might act as a functional subunit of PyrKinc in the soybean seed
For investigation of potential functions of the novel 51 kDa polypeptide, we partially purified PyrKinc by Resource Q (RQ) anion-exchange Fast Protein Liquid Chromatography (FPLC) from extracts of 5-week-old seeds. Enzymatic assays were conducted at pH 6.9 at which the cytoplasmic form is active while the plastid form is inactive (Smith et al., 2000). Under these conditions, there were two major peaks of PyrKinc activity (fraction nos. 18–26 and 27–35, designated as peak I and II, respectively; Figure 2a). To further separate possible PyrKinc oligomeric forms, we separated proteins by non-denaturing PAGE followed by in-gel activity staining as described by Rivoal et al. (2002) (Figure 2b). After native electrophoresis, each active PyrKinc band (white boxes in Figure 2b) was excised and analyzed by SDS–PAGE and immunobloting. As shown in Figure 2(c), active bands eluted from the first activity peak consisted mainly of the 51 kDa subunit. Upon SDS–PAGE, the most active bands resolved by native PAGE from RQ peak I (bands 3 and 4) contained only small amounts of 55 kDa polypeptide. In contrast, all active bands from RQ peak II had the 55 kDa polypeptide as a major component and a lesser amount of the 51 kDa polypeptide. All the active bands resolved by native PAGE migrated slower than the 200 kDa standard, suggesting that the soybean PyrKinc can exist as a tetramer or higher order oligomer. Indeed, hexameric (Hu and Plaxton, 1996) and decameric (Wu and Turpin, 1992) forms of PyrKinc have been observed. Even though not all forms of PyrKinc might have retained activity during native PAGE, the results suggest that the 51 and 55 kDa polypeptides were active subunits of PyrKinc oligomers.
In order to investigate metabolite regulation of soybean seed PyrKinc, active forms resolved by native PAGE were assayed in the presence and absence of amino acids. For example, fraction no. 18 from RQ peak I contained a major and minor band of activity after native PAGE (band 3 and 2, respectively). Both bands were unaffected by Gln, Asn, and Glu, but were activated by Asp (Figure 2d). In contrast, of the three activity bands resolved by native PAGE from RQ peak II, the most pronounced effect was inhibition by Glu. Thus, the observed regulatory properties appeared to vary depending on the subunit composition and/or oligomeric structure of the enzyme. Of particular significance is the association of Asp activation with forms containing the 51 kDa polypeptide.
To further test if the 51 kDa polypeptide had altered kinetic activity and/or regulatory properties, we compared the total PyrKinc activity in preparations from 2-week-old versus 7-week-old seeds. Immunoblot analysis of aqueous extracts of the seeds indicated that the 55 kDa polypeptide predominated in 2-week-old seeds, whereas the 51 kDa polypeptide predominated in 7-week-old seeds (Figure 2e). Quantification of immunoreactive bands by densitometry indicated that the PyrKinc protein (sum of the 51 and 55 kDa polypeptides) in 7-week-old seeds was reduced to 70 ± 2% of the amount present in 2-week-old seeds (Figure 2f). Maximum PyrKinc activity in 7-week-old seeds was reduced to roughly the same extent (62% compared to 2-week-old seeds; Figure 2g). Therefore, these results suggest that the 51 kDa polypeptide must be an active subunit with equivalent catalytic activity compared with the full-length PyrKinc subunit. Glutamate has been reported to inhibit PyrKinc (Smith et al., 2000). Consistent with this, 30 mm Glu inhibited PyrKinc activity by about 30% in extracts of 2- and 7-week-old seeds (data not shown). However, aspartate at 10 mm activated total PyrKinc activity in 7-week-old seeds (16.6 ± 2% stimulation) but slightly inhibited activity in extracts from 2-week-old seeds (10 ± 1% inhibition; Figure 2g). These results suggest that the 51 kDa polypeptide might have somewhat altered the regulatory properties compared to the 55 kDa polypeptide.
The investigation of phosphorylation of pyruvate kinase in soybean seeds
The phosphorylation of PyrKinc in plants has not been reported so far (Podesta and Plaxton, 1991). However, in silico analysis suggested that plant PyrKinc contains two sequences that match the phosphorylation motif targeted by plant CDPKs and SnRK1s (Huang and Huber, 2001; see Figure 3a). Interestingly, the mammalian L- and R-type PyrKinc also contain these two potential phosphorylation sites, but phosphorylation at these two sites has not been reported (Lone et al., 1986). Rather, phosphorylation of the mammalian L- and R-type pyruvate kinase occurs at the serine residue in the sequence LRRAS, located close to the N-terminus of the protein.
We first determined if synthetic peptides based on the two predicted sites could be phosphorylated by endogenous seed protein kinases. Initially, we fractionated seed proteins by RQ-FPLC. Two major peptide kinase activity peaks were resolved (Figure 3b). The first peak had a twofold higher activity with the Ser220 peptide compared to the Ser407peptide, was Ca2+-dependent, and had a native Mr of approximately 60 kDa by size exclusion chromatography (Figure 3c), and therefore appeared to be a member of the CDPK family of protein kinases (Harmon et al., 2000). The second peak from RQ-FPLC showed a preference for the Ser407 peptide, compared to the Ser220 peptide and was not Ca2+-dependent. The native Mr was approximately 150 kDa, and immunoblot analysis using antimaize SNF1 catalytic subunit antibodies confirmed the existence of SnRK1 in those fractions (data not shown).
In order to determine if the full-length PyrKinc could be phosphorylated by CDPK or SnRK1, we performed in vitro phosphorylation using recombinant PyrKinc as substrate and RQ-CDPK or RQ-SnRK1 as the protein kinases. Phosphorylation at the Ser220 site by RQ-CDPK, as monitored with sequence- and phosphorylation-specific antibodies (Ab-pSer220), proceeded in a time-dependent manner, and was completely blocked by the addition of the calcium chelator EGTA (Figure 3d). We could not demonstrate the phosphorylation of recombinant PyrKinc at the Ser407 site by protein kinase and no phosphorylation of either site was catalyzed in vitro by RQ-SnRK1 (data not shown).
We tried to detect the in vivo phosphorylation of PyrKinc using the phospho-specific antibodies. There was no significant signal band corresponding to the phosphorylated full-length PyrKinc, but rather abundant phosphorylated and truncated polypeptides in the range between approximately 20 and 51 kDa were observed (Figure 3e). This result implied that phosphorylation and proteolytic cleavage could be closely coupled.
Association of phosphorylated and truncated pyruvate kinase polypeptides with the 26S proteasome in soybean seeds
Because phosphorylated PyrKinc polypeptides only appeared in truncated polypeptides in vivo, the proteolytic machinery might be involved. There are two general pathways for protein turnover in eukaryotic cells: one is the ubiquitin/26S proteasome pathway and the other involves soluble proteases present in the vacuole or lysosome. Selective protein degradation usually involves the 26S proteasome, which is a huge complex with a molecular weight of approximately 2000 kDa in Arabidopsis (Peng et al., 2001). Therefore, we reasoned that size exclusion chromatography of seed extracts could distinguish which pathway was active in degradation of PyrKinc. Cytoplasmic pyruvate kinase activity eluted from the size-exclusion chromatography column in a broad peak consistent with active tetrameric and dimeric forms (Figure 4a). Enzymatic activity was coincident with the 55 kDa polypeptide, detected with Ab-NT (Figure 4b). As shown in the immunoblots probed with Ab-pSer220 or Ab-pSer407 (Figure 4b), the heavily phosphorylated and truncated PyrKinc polypeptides partitioned in the flow-through fractions which were devoid of PyrKinc activity.
Because the phosphorylated and truncated PyrKinc polypeptides eluted near the column void volume, we speculated that they could be associated with proteasomes. In a subsequent size exclusion chromatography experiment, we detected a chymotrypsin-like activity of a huge proteolytic complex using Suc-LLVY-AMC as the substrate (Figure 5a). The maximum catalytic activity of the proteolytic complex was observed in the presence of MgATP. No significant 20S-like activity was observed. Moreover, a 26S S6′ ATPase-like component was detected in this proteolytic complex by immunoblot analysis using anti-Hela cell 26S S6′ ATPase (Figure 5b). Therefore, we believe that the high molecular complex is the 26S proteasome. The size of the soybean 26S proteasome particle, determined by size exclusion chromatography, was approximately 1900 kDa, which is similar to that reported in Arabidopsis (Peng et al., 2001). Indeed, supplementing the extraction buffer with the proteasome-specific inhibitors, MG132 and NLVS, increased the recovery of full-length (55 kDa) and 51 kDa polypeptides (cf. Figure 5c,d) and phosphorylated polypeptides (cf. Figure 5e–h) co-eluting with the 26S proteasomes.
In parallel, we also measured trypsin activity as an internal control. The elution volume for detectible typsin-like activity using the substrate Z-ARR-AMC was located at approximately 24 kDa, similar to the Mr of the trypsin-like protease (Shrestha et al., 2002). No trypsin activity was observed in the flow-through fractions (Figure 5a). Although there was a clear 36 kDa polypeptide recognized by Ab-NT that co-eluted with trypsin-like protease, there were no obvious phosphorylated and truncated PyrKinc polypeptides that co-eluted with the trypsin-like activity (Figure 5e–h). These results suggest that the phosphorylated and truncated PyrKinc polypeptides were specifically associated with the 26S proteasome.
Because of their large size, sucrose-gradient centrifugation is a routine technique to isolate proteasomes (Fujinami et al., 1994; Hu et al., 1999). By this approach, we examined 26S proteasomes in developing soybean seeds. After 19 h of centrifugation at 100 000 g at 4°C in a 10–40% sucrose gradient, all of the 26S proteasome activity was at the bottom of the gradient when MgATP was present throughout (Figure 6a) and there was no detectable 20S-like proteasome activity (data not shown). When MgATP was omitted, 26S-like proteasome activity was undetectable (data not shown) and only a low level of 20S-like proteasome activity was observed (Figure 6b). The results suggest that the maintenance of the 26S proteasome particle is ATP-dependent. Consistently, as an internal control as in size exclusion chromatography experiments, we found typsin-like activity exclusively in the low-molecular weight fraction that remained at the top of the gradient (Figure 6c).
In the sucrose-gradient centrifugation experiments, incidences of proteasomes and phosphorylated PyrKinc polypeptides were examined by immunoblot analysis. Including proteasome inhibitors increased recovery of PyrKinc polypeptides reactive with Ab-NT (cf. Figure 6d,e), and Ab-pSer407 (cf. Figure 6f,g) in the proteasome-containing fractions. These results are consistent with those obtained from size-exclusion chromatography experiments, and suggest that the phosphorylated and truncated PyrKinc polypeptides are degradative intermediates associated with 26S proteasomes. Notably, the 55 kDa full-length PyrKinc was found associated with the 26S proteasome when MG132 and NLVS were included in the extraction buffer, suggesting that not all subunits of the PyrKinc oligomer are post-translationally modified prior to degradation by the proteasome.
Ubiquitination of pyruvate kinase in vivo and in vitro
Ubiquitination is an essential step for the targeting of most proteins for degradation through the 26S proteasome pathway. In plants, ubiquitin-conjugation and proteasome-mediated degradations have been confirmed for a number of important regulatory proteins such as PhyA and AUX/IAA proteins (Cough and Vierstra, 1997; Gray et al., 2001). Accordingly, we tested if PyrKinc can be ubiquitinated either in vivo or in vitro. We carried out in vitro ubiquitination using two different approaches. First, we demonstrated in vitro ubiquitination using recombinant PyrKinc and a soluble protein fraction prepared from 2-week-old seeds that was depleted of proteasomes by ultracentrifugation and supplemented with exogenous ubiquitin. A similar approach was used by O'Hagan et al. (2000) with S100 cell extracts. The ubiquitination of recombinant His6-PyrKinc was detected by the appearance of high-molecular weight PyrKinc conjugates, immunoprecipitated with anti-His6 antibodies. The ubiquitin conjugation state was verified by immunoblot analysis using antiubiquitin antibodies (data not shown). The second approach was attempted, because the components of the ubiquitin–proteasome pathway have been shown to be pre-assembled at the cytoplasmic surface of the ER in mammals and yeast (Biederer et al., 1997; de Virgilio et al., 1998). We fractionated the membrane preparation from 5-week-old seeds by ultracentrifugation at 150 000 g for 6.5 h, and after re-suspension tested in vitro ubiquitination of recombinant PyrKinc. Time-dependent ubiquitination of recombinant His6-PyrKinc, immunoprecipitated with anti-His6 antibodies, was observed. Only the wild-type (WT) recombinant PyrKinc was ubiquitinated in a time-dependent manner (Figure 7a). Among the phosphorylation site mutants, S220A showed less ubiquitination signal and S407A did not show any ubiquitination (Figure 7a). These results suggest that the phosphorylation of serine residues in PyrKinc is essential for ubiquitination.
To examine the existence of ubiquitinated conjugates of PyrKincin vivo, we immunoprecipitated PyrKinc from seeds at different stages of development using Ab-NT, followed by immunoblot analysis with anti-Ubi (Figure 7b-1), Ab-NT (Figure 7b-2), Ab-pSer407 (Figure 7b-3), or Ab-pSer220 (Figure 7b-4). High Mr ubiquitin conjugates were observed after probing with anti-Ubi antibody (Figure 7b-1) or with Ab-NT (Figure 7b-2). These results confirm the existence of endogenous ubiquitin conjugates of PyrKinc in young developing seeds. The abundance of those conjugates was higher in the first 5 weeks of seed development, then decreased gradually until at maturity there were no detectable ubiquitin conjugates. This pattern is in agreement with the finding of total protein-conjugates in Lupin (Lupinus albus L.) seeds during formation (Ferreira et al., 1995).
In mammals, phosphorylation and ubiquitination are closely coupled processes that target many proteins for complete degradation (Attaix et al., 2001), in particular, those recognized by the SCF E3 ligase (Hershko and Ciechanover, 1998). To determine if the ubiquitin conjugates of PyrKinc are also phosphorylated, we probed the blots of Ab-NT-immunoprecipitates with Ab-pSer220 or Ab-pSer407. As shown in Figure 7(b-3, b-4), both serine residues were phosphorylated in the ubiquitin-conjugates in a developmental profile. Thus, in agreement with our observation of in vitro ubiquitination (Figure 7b), the ubiquitin conjugates of phosphorylated PyrKinc polypeptides have been confirmed and the results suggest that phosphorylation of PyrKinc occurs in concert with ubiquitination.
The degradation of pyruvate kinase polypeptides associated with 26S proteasomes in vitro
We next examined the degradation of PyrKinc polypeptides associated with proteasomes contained in the flow-through fraction from size exclusion chromatography (as in the experiment shown in Figure 5a). As shown in Figure 8(a), immunoblot analysis using Ab-NT indicated that the degradation was time-dependent and was prevented by MG132. The full-length PyrKinc and its truncations were only degraded during incubation at 30°C in the presence of ATP but in the absence of MG132.
Similarly, we tested the proteasome-containing fraction from sucrose-gradient centrifugation following the removal of MG132 by dialysis. The time-dependent degradation of proteasome-associated phosphorylated polypeptides at 30°C was observed on immunoblots probed with Ab-pSer407 (Figure 8b) or Ab-pSer220 (data not shown). These results provide additional evidence for PyrKinc degradation through the 26S proteasome pathway.
In this study, we have cloned two nearly identical cDNAs encoding paralogs of PyrKinc; both encode polypeptides of approximately 55 kDa and are expressed in a similar developmental manner in developing soybean seeds. However, during development, a 51 kDa polypeptide accumulates, apparently produced by C-terminal cleavage of the 55 kDa polypeptide. Active PyrKinc appears to be a heterooligomeric protein composed of 55 and 51 kDa subunits depending on the developmental stage. We also found that developing seeds contain two kinds of protein kinases that can potentially phosphorylate PyrKinc at two sites (Ser220 and Ser407). Biochemically, one shows CDPK activity and the other appears to be a SnRK1. Furthermore, developing soybean seeds are shown to contain typical 26S proteasome particles with a molecular weight of approximately 1900 kDa determined by size exclusion chromatography. Collectively, we propose that there are two pathways for PyrKinc modification at the post-translational level. One involves conversion of the 55 kDa to the 51 kDa polypeptide, which may produce a novel pyruvate kinase activity. The second involves phosphorylation (at the Ser220 and Ser407 sites) and ubiquitination and might be involved in the proteasome-mediated degradation of the enzyme.
The 51 kDa subunit is produced from the 55 kDa pyruvate kinase subunit by a C-terminal truncation
Immunoblot analysis using Ab-NT revealed that developing seeds accumulate a C-terminally truncated 51 kDa PyrKinc polypeptide that appears to retain PyrKinc activity. As a part of this study, we attempted to elucidate the basis for production of the 51 kDa polypeptide. The 51 kDa polypeptide could be produced: (i) from an independent transcription event, (ii) from an alternative splicing event of the full length transcript, or (iii) from proteolytic processing of the 55 kDa polypeptide. We were unable to detect a transcript encoding a 51 kDa polypeptide, which rules the first two possibilities as unlikely. Furthermore, while alternative splicing to generate a shortened mRNA is common in mammals (35% of genes), it occurs less often in plants (Miller et al., 2001). In rat muscle, alternative splicing produces the M1- and M2-type PyrKinc and the alternative splicing occurs at the 9th and 10th exons (Noguchi et al., 1986). The two M-types of PyrKinc have different enzymatic properties and tissue specificity (Imamura and Tanaka, 1982). To understand the possible gene structure of soybean PyrKinc, we examined the Arabidopsis homolog (GB# NP_196474, TIGR accession no. at5g08570, http://www.tigr.org), which has three exons and two introns. Therefore, the gene structure encoding plant PyrKinc is totally different from that of the mammalian genes. No alternative splicing for the plant gene was suggested by computational analysis (http://www.tigr.org; search locus At5g08570). This computational prediction is consistent with the nested RNA gel blot results; no detectible shorter transcripts were observed.
Therefore, proteolytic processing might be responsible. This is supported by the observation that the 51 kDa polypeptide continued to increase from the 9th to 10th week, even though the RNA gel blots and RT-PCR profiles showed almost no transcripts for PyrKinc at that time. Several critical questions remain. First, which proteolytic pathway or enzyme (such as proteasome or an unknown protease) is engaged in this C-terminal truncation and how is it regulated? Secondly, what is the physiological significance of the 51 kDa polypeptide? It is possible that C-terminal cleavage of the 55 kDa polypeptide of PyrKinc may provide a novel mechanism to complement transcriptional regulation in developing soybean seed. We speculated that the 51 kDa polypeptide may have increased stability, as it showed a steady accumulation during development while the full-length PyrKinc decreased. Removal of approximately 4 kDa (approximately 35 amino acid residues) from the C-terminus could have a significant impact on both stability and enzymatic properties.
Phosphorylation and ubiquitination of pyruvate kinase are linked to its degradation via 26S proteasomes
Embryogenesis is an active phase of metabolism and requires highly accurate control. The regulation of metabolic enzymes by turnover contributes to the control extended by transcriptional and translational control. Ferreira et al. (1995) reported that there were abundant and dramatic changes of ubiquitin and ubiquitin–protein conjugates during lupin seed development. Thus, ubiquitination and protein turnover is significant in developing seeds. Similarly, Doelling et al. (2001) demonstrated the importance of the ubiquitin/26S proteasome pathway to embryo development in Arabidopsis. They showed arrest of embryo development at the globular stage in knockouts of ubiquitin-specific protease (UBP14). In agreement with these findings, our data suggest that metabolic enzymes such as PyrKinc are degraded via the ubiquitin/26S proteasome pathway.
Phosphorylation of PyrKinc at the Ser220 and Ser407 occurred in vivo primarily on N-terminally truncated polypeptides that co-fractionated with 26S proteasomes during size exclusion chromatography (Figures 4 and 5) or sucrose-gradient centrifugation (Figure 6). We presume that the phosphorylation of both sites was reversible by the action of protein phosphatases. This was suggested by the observation that phosphatase inhibitors increased recovery of phosphorylated PyrKinc polypeptides (data not shown). However, the extent of reversibility was not investigated and is beyond the scope of the present study. More importantly, the recovery of the phosphorylated polypeptides with proteasomes was increased by addition of proteasome-specific inhibitors, MG132 and NLVS, and when proteasomes were incubated in vitro (in the absence of inhibitors), PyrKinc polypeptides were degraded (Figure 8). High Mr Ubi–PyrKinc conjugates could be identified in vivo. Ubiquitination of recombinant WT PyrKinc was demonstrated in vitro, but ubiquitination of phosphorylation site mutants (S220A or S407A) was significantly reduced (Figure 7a). These results suggest that phosphorylation is an essential step prior to ubiquitination. Collectively, these results suggest that phosphorylation and ubiquitination are essential for PyrKinc degradation via proteasome pathway.
In recent years, the ubiquitin/26S proteasome pathway has been shown to play pleiotrophic roles in plant biology such as hormone regulation (auxin, cytokinins, and abscisic acid signaling), photomorphogenesis, senescence, programmed cell death, and plant defense responses (reviewed by Callis and Vierstra, 2000; Estelle, 2001; Schwechheimer and Deng, 2001). Our results suggest that degradation of metabolic enzymes such as PyrKinc and sucrose synthase (Hardin et al., 2002) can be added to the list as well. For both enzymes, phosphorylation of minor or ‘cryptic’ sites has been implicated as a possible trigger for proteasome-mediated degradation. If this is correct, there may be recognition elements contained within the sequences surrounding Ser220 and Ser407 of PyrKinc and Ser170 of sucrose synthase (Hardin et al., 2002) that may be recognized by an E3 ubiquitin ligase such as SCF complex. Thus, certain phosphorylation sites in metabolic enzymes may serve as ‘phospho-degrons’ as described for the phosphorylation-driven and ubiquitination-mediated destruction of cyclin-dependent kinases (Harper, 2002).
All peptide-specific and phosphopeptide-specific antibodies were produced and affinity purified by Bethyl Laboratories, Inc. (Montgomery, TX). The peptide sequences were as follows: MANID IEGILKQQQP for Ab-NT; FVRKGpS220DLVN for Ab-pSer220; VLTRGGpS407 TAKL for Ab-pSer407. The ratio of specificity of the antiphosphopeptide antibodies for the phosphopeptide antigen compared to the unphosphorylated sequence was greater than 99 : 1 by ELISA (as determined by Bethyl Laboratories). [γ-32P] ATP (111 TBq mmol−1) and [α-32P] dCTP (111 TBq mmol−1) were purchased from Perkin Elmer (Boston, MA); microcystin-LR, proteasome inhibitors (MG132, NLVS) and the fluorogenic substrate (Suc-LLVY-AMC) for proteasome assay were obtained from CalBiochem (La Jolla, CA). Pfu DNA polymerase was from Stratagene (Carlsbad, CA) and Taq DNA polymerase was from Roche (Switzerland). All other chemicals were from Sigma–Aldrich (St. Louis, MO), unless stated otherwise.
Soybean [Glycine max (L.) cv. Ransom] plants were grown in controlled growth rooms with 9 h photoperiod and 26°C day/22°C night temperature regime in the South-eastern Plant Environment Laboratory. A 3 h light interruption was included during the dark period for the first 5 weeks. Pods were harvested at the size depending on the designed experiments and seeds were frozen directly in the liquid N2.
Sample preparation for SDS–PAGE and immunoblot analysis
Frozen seeds were powdered in liquid N2 and mixed with four volumes of 1.5× stock of SDS sample buffer (no dye), which was supplemented with a cocktail of protease inhibitors (1 mm 4-(2-aminoethyl)-benzene sulfonyl fluoride hydrochloride (AEBSF-HCl), 1 mm phenylmethylsulfonyl fluoride (PMSF), 1 mm benzamidine HCl, 5 µmtrans-epoxysuccinyl-l-leucylamido-(4-guani dinobutane (E64), 50 µm leupeptin, 20 µg ml−1 soybean trypsin inhibitor and 5 mm ethylenediamine tetraacetic acid Na2-salt (EDTA)), and phosphatase inhibitors (50 mm sodium fluoride and 1 µm microcystin-LR). Proteins were denatured by boiling at 100°C for 4 min, and 20 µg total protein was loaded into each lane and separated by 12.5% SDS–PAGE, which was followed by a standard immunoblot analysis (ECL; Amersham Biosciences, UK).
Partial purification of protein kinases from soybean seeds
Partial purification of protein kinases was performed as detailed by Bachmann et al., (1996). Briefly, frozen seeds (6 g) were homogenized in 50 ml of extraction buffer containing 50 mm MOPS-NaOH, pH 7.5, 10 mm MgCl2, 5 mm DTT, 1 mm EDTA, 0.5 mm phenylmethylsulfonyl fluoride, and 0.1% (v/v) Triton X-100 in a cold mortar. The homogenate was filtered through two layers of Miracloth (CalBiochem, LaJolla, CA). Soluble proteins that precipitated from 5 to 15% (w/v) polyethylene glycol 8000 (PEG 8000) were collected by centrifugation at 34 540 g for 20 min at 4°C. The protein pellet was re-suspended in 10 ml buffer A (50 mm MOPS-NaOH, pH 7.5, 10 mm MgCl2, 2.5 mm DTT). After clarification by centrifugation at 34 540 g, 20 min, 4°C, the sample was applied to a 10 ml Resource Q anion exchange column (Amersham Biosciences, UK). After washing with buffer A, the bound protein was eluted with an 80 ml linear gradient from 0 to 500 mm NaCl in buffer A at a flow rate of 1 ml min−1. Active fractions were pooled separately and dialyzed against buffer A without NaCl.
Peptide kinase assay
The peptide kinase assay was carried out in a 40 µl volume of 50 mm MOPS-NaOH, pH 7.5, 10 mm MgCl2, 0.1 mm ATP ([γ-32P] ATP 200 CPM pmol−1), 0.1 mg ml−1 peptide, 0.1 mm CaCl2 or 1 mm EGTA (as specified in the text) and containing 4 µl RQ column fraction. Reactions were initiated by addition of ATP and incubated for 10 min at room temperature. After incubation, 30 µl of the reaction mixture was spotted onto P81 phosphocellulose paper squares (4 cm2). After three 10 min washes with 75 mmO-phosphoric acid to remove unincorporated γ-32P ATP, the radioactivity on the paper square was detected by liquid scintillation spectrometry.
Pyruvate kinase assay
The PyrKinc assay followed the procedure of Turner and Plaxton (2000). Briefly, the assay buffer system consisted of 50 mm Hepes/KOH (pH 6.9), 25 mm KCl, 10 mm MgCl2, 2 mm PEP, 1 mm ADP, 1 mm DTT, 5% (w/v) PEG 8000, 0.15 mm NADH, and 4 units ml−1 lactate dehydrogenase in a total volume of 1 ml. Assays were initiated by the addition of ADP and PEP and incubated for 5 min at 25°C. The decrease in absorbance at 340 nm as a result of the oxidation of NADH was followed with time.
Cloning of cDNAs encoding pyruvate kinase from developing seeds
Total RNA was isolated from 0.3 g tissue of 4–5-week-old seeds following the routine RNA isolation protocol (Kiefer et al., 2000). Total RNA (1 µg) was used for the reverse transcription reaction following instruction from the manufacturer (InVitrogen, Carlsbad, CA) except that incubation was at 45°C for 1 h. The forward primer, GAGAACATATGGCGAACATAGACATCG, and the reverse primer, GAGCCCGGGTTATTTGACTATGCAGATC, were chosen from the cDNA sequence in the NCBI database (GB#: L08632). The PCR was catalyzed by Pfu DNA polymerase (Stratagene, La Jolla, CA) under the conditions of 94°C for 2 min followed by 36 cycles of 94°C, 50 sec denaturation; 47°C, 45 sec annealing; 72°C, 2 min extension. The PCR product was checked by electrophoresis in 0.8% agarose in 1× TAE buffer. For TA cloning, a dA tail was added to the PCR product by Taq DNA polymerase, and the desired band was excised and then eluted by the Qiaquick gel extraction Kit (Qiagen, Germany). The eluted cDNA was ligated into pGEM-T-easy vector (Promega, Madison, WI). Positive clones were identified by sequencing, and sequence analysis was carried out by using the online software of sequence utilities at BCM Search Launcher (http://searchlauncher.bcm.tmc.edu) and phylogeny phylip program (http://bioweb.pasteur.fr/seganalphylogeny/phylip-uk.html). The two cDNAs cloned for soybean PyrKinc (PyrKinc1 and PyrKinc2) were submitted to the NCBI database and their accession numbers are AY128951 and AY128952.
RT-PCR analysis and RNA gel blot analysis
The expression of PyrKinc genes was analyzed by RT-PCR. The isoform-specific primers for each PyrKinc were as follows: the forward primer for PyrKinc1 was TTGAAGCAGCAGCAGCCTT and the reverse primer was TGTTGGGTACACCCCATCC; the forward primer for PyrKinc 2 was ATCTTGAAGCAGCAGCTG and the reverse primer was TGTTGGGTACACCCCATTG. The PCR was driven by Taq DNA polymerase (Roche, Switzerland) under the conditions of 94°C for 2 min followed by 28 cycles of 94°C, 50 sec denaturation; 58°C, 45 sec annealing; 72°C, 1 min extension.
For nested RNA gel blots, the DNA for probe 1 was 378 bp in length covering 5′ region, which was released by PCR. The forward primer was GAGAACATATGGCGAACATAGACATCG and reverse primer was CTCCTGATTCTCAACC. The DNA for probe 2 was 719 bp in length covering 5′ region and produced by PCR using the forward primer GAGAACATATGGCGAACATAGACATCG and the reverse primer GGACAGCAGATGATGC. The probe labelings were using the Random Labeling Kit following the instruction from the manufacturer (Roche, Switzerland).
Expression of recombinant pyruvate kinase protein
The vector plasmid DNA pET15b (Novagen, Germany) was sequentially digested by BamH1, filled with Klenow, then digested with NdeI, and fractionated by 0.8% agarose gel electrophoresis. The insert DNA for PyrKinc was prepared by sequential double digestion by Smal and NdeI, ligated with the linearized vector DNA, and transformed into E. coli DH5α. Positive transformants were identified and the plasmid DNA was used to transform E. coli BL21. The 37°C overnight culture of transformed BL21 cell was diluted into 100 volumes in LB medium. After growing at room temperature for 12 h, the induction of recombinant PyrKinc was initiated by addition of isopropyl β-d-thiogalacto-pyranoside (IPTG) to a final concentration of 0.5 mm and continued for 10 h at room temperature. Soluble (His)6-PyrKinc and its mutants (S220A and S407A) were affinity-purified following the manufacturer's instruction (Qiagen, Germany).
In vitro phosphorylation of pyruvate kinase
The procedure for in vitro phosphorylation was the same as for peptide kinase assays, except that only non-radioactive ATP was applied. Specific phosphorylation of the Ser220 and Ser407 sites was determined by immunoblot analysis using sequence- and phosphorylation-specific antibodies for both sites.
Size exclusion chromatography separation of 26S proteasome particle
Six grams of soybean seeds were homogenized in 50 ml of extraction buffer containing 50 mm MOPS-NaOH, pH 7.5, 100 mm NaCl, 10 mm MgCl2, 1.5 mm ATP, 2 mm DTT, 10% glycerol, and 0.5 mm phenylmethylsulfonyl fluoride (PMSF) in a cold mortar. The homogenate was filtered through two layers of Miracloth (CalBiochem, La Jolla, CA) and fractionated with PEG 8000. Soluble proteins that precipitated between 5 and 15% (w/v) PEG 8000 were collected by centrifugation at 34 540 g for 20 min at 4°C. The protein pellet was re-suspended in 10 ml of extraction buffer. After an additional centrifugation at 100 000 g for 1 h at 4°C, 6 ml of the supernatant was applied to a Fractogel TSK HW-55 column (bed volume 126 ml) (MCB manufacturing chemists. Inc., Darmstadt, Germany) linked to FPLC (Amersham Biosciences, UK). The equilibration buffer and elution buffer were 50 mm MOPS-NaOH, pH 7.5, containing 10% glycerol, 100 mm NaCl, 3 mm MgCl2, 1.6 mm ATP, and 2 mm DTT. The flow speed was at 30 ml h−1 and the separation was performed at 4°C.
Partial purification of proteasome by sucrose-gradient centrifugation
Powdered soybean seeds (2.5 g) were homogenized in 10 ml of extraction buffer containing 25 mm Tris–HCl, pH 7.5, 10 mm MgCl2, 4 mm ATP, and 1 mm DTT. The homogenate was filtered through Miracloth (CalBiochem, La Jolla, CA) and centrifuged for 15 min at 34 540 g and 4°C. Supernatants were collected and 1 ml was applied to 4 ml 10–40% sucrose gradients in extraction buffer. The gradient was then centrifuged at 100 000 g and 4°C for 19 h in a Beckman L8-80 Ultracentrifuge. Fractions (200 µl) were collected manually after ultracentrifugation.
Proteasome activity assays
Proteasome assays were performed as described by Kisselev et al. (1999) with minor modifications. For 26S proteasome activity, 10 µl enzyme samples were added to 0.99 ml of 20 µm Suc-LLVY-AMC in 20 mm Tris–HCl, pH 7.5, 1 mm DTT, 5 mm MgCl2, and 1 mm ATP. After a 30 min incubation at 30°C, the released AMC was detected by fluorescence with 380 nm excitation 460 nm−1 emission. For 20S proteasome activity, 0.02% SDS was substituted for 1 mm ATP in the same buffer and incubation conditions.
Immunoprecipitation and protein immunoblot analysis
Immunoprecipitation was performed in 100 µl volume consisting of 50 mm Tris–HCl, pH 7.5, 150 mm NaCl, 0.1% (v/v) Triton X-100, 5 mm EDTA, 1 mm PMSF, and 10 µm MG132. The antibody was at a concentration of 10 µg ml−1. After binding for 2 h at 4°C with gentle shaking, 10 µl washed protein-G beads (Sigma–Aldrich, St. Louis, MO) were added and binding was continued for another 1 h at 4°C. After extensive washing with the buffer above, the protein bound to the beads was eluted by addition of SDS sample buffer and boiled at 100°C for 10 min. After electrophoresis, the protein was transferred to nitrocellulose membrane (Hybond ECL) and probing followed the protocol detailed by the manufacturer (Amersham Bioscience, UK).
In vitro ubiquitination
Soybean seed extracts were prepared following the procedure described by O'Hagan et al. (2000) with some modifications. Briefly, 5 g frozen seeds were homogenized in 10 ml extraction buffer containing 40 mm Hepes-KOH, pH 7.2, 5 mm MgCl2, 8 mm KCl, 0.5 mm PMSF, and 10 µg ml−1 leupeptin. The homogenate was clarified by centrifugation at 34 540 g for 10 min at 4°C. The resulting supernatant was then centrifuged at 100 000 g for 1 h at 4°C in a Beckman L8-80 Ultracentrifuge. For the membrane preparation, after 100 000 g centrifugation for 1 h at 4°C, the pellet was discarded and the supernatant was centrifuged again (150 000 g, 4°C, 6.5 h). The in vitro ubiquitination assay was performed according to the protocol detailed by Podust et al. (2000) with some modifications. Briefly, the 40 µl reaction mixture contained 10 mm Hepes-KOH, pH 7.2, 6 mm MgCl2, 2 mm ATP, 1 µm okadaic acid, 30 µm ubiquitin, 10 nm MG132, 3 µg recombinant protein, and 10 µl S100-like supernatant or membrane preparation, and incubated at 30°C for 0–60 min. The conjugates were immunoprecipitated by anti-His6 and probed with antiubiquitin antibodies (Sigma–Aldrich, St. Louis, MO) on immunoblots.
Pyruvate kinase native gel activity assay
The 8% native gel system was the same as described by Doucet et al. (1990) except that the SDS was omitted. Activity staining of pyruvate kinase was performed by the protocol detailed by Rivoal et al. (2002) with some modifications. Briefly, the electrophoresis was run for 2 h at 120 V and 4°C. After that, the gel was immediately equilibrated in 25 ml of 50 mm Hepes-KOH, pH 6.9, 25 mm KCl, 10 mm MgCl2, 1 mm DTT, and 5% (w/v) PEG 8000 for 15 min at room temperature. The gel was developed in the same buffer containing 2 mm PEP, 1 mm ADP, 0.2 mm NADH, and 4 units ml−1 lactate dehydrogenase. After incubation for 10 min at room temperature, the stained gel was photographed on the UV transilluminator. The active PyrKinc appeared in dark bands against the fluorescent background.
This research represents cooperative investigations of the US Department of Agriculture (USDA), Agricultural Research Service, and the North Carolina Agricultural Research Services. Mention of a trademark or proprietary product does not constitute a guarantee or warranty of the product by the USDA or the North Carolina Agricultural Services and does not imply its approval to the exclusion of other products that might also be suitable. The research was supported in part by fund from the United Soybean Board (project number 1240) and the US Department of Energy (Grant DE-AI05–91ER20031 to SCH).