Molecular and Regulatory Properties of Leucoplast Pyruvate Kinase from Brassica napus (Rapeseed) Suspension Cells

Archives of Biochemistry and Biophysics Vol. 400, No. 1, April 1, pp. 54 – 62, 2002 doi:10.1006/abbi.2002.2782, available online at http://www.idealibrary.com on

Molecular and Regulatory Properties of Leucoplast Pyruvate Kinase from Brassica napus (Rapeseed) Suspension Cells 1 William C. Plaxton,* ,† ,2 Christopher R. Smith,† and Vicki L. Knowles* *Department of Biology and †Department of Biochemistry, Queen’s University, Kingston, Ontario, Canada K7L 3N6

Received November 22, 2001, and in revised form January 7, 2002

Plastidic pyruvate kinase (PK p) from Brassica napus suspension cells was purified 431-fold to a final specific activity of 28 ␮mol phosphoenolpyruvate (PEP) utilized/ min/mg protein. SDS–PAGE, immunoblot and gel filtration analyses indicated that this PK p exists as a 380-kDa heterohexamer composed of equal proportions of 64- (␣subunit) and 58-kDa (␤-subunit) polypeptides. The Nterminal sequence of the PK p ␣- and ␤-subunits exhibited maximal identity with the corresponding regions deduced from putative PK genes of Arabidopsis thaliana and Methylobacterium extorquens, respectively. B. napus PK p displayed a sharp pH optimum of pH 8.0, and hyperbolic saturation kinetics with PEP and ADP (K m ⴝ 0.052 and 0.14 mM, respectively). 6-Phosphogluconate functioned as an activator (K a ⴝ 0.12 mM) by increasing V max by approximately 35% while decreasing the K m(PEP) and K m(ADP) values by 40 and 50%, respectively. 2-Oxoglutarate and oxalate were the most effective inhibitors (I 50 ⴝ 8.3 and 0.23 mM, respectively). A model is presented which highlights the role of 6-phosphogluconate in coordinating stromal NADPH and ATP production for anabolic processes of B. napus leucoplasts. © 2002 Elsevier Science (USA)

Key Words: pyruvate kinase; plant glycolysis; leucoplast metabolism; Brassica napus (rapeseed).

Pyruvate kinase (PK 3; EC 2.7.1.40) is an important regulatory enzyme of glycolysis that catalyzes the irre1 This work was supported by Research and Equipment Grants from the Natural Sciences and Engineering Research Council of Canada (NSERC). 2 To whom correspondence and reprint requests should be addressed. Fax: (613) 545-6617. E-mail: [email protected]. http://biology.queensu.ca/faculty/plaxton.html. 3 Abbreviations used: BDA, blue dextran agarose; GS, glutamine synthetase; GOGAT, glutamine 2-oxoglutarate aminotransferase; 2-OG, 2-oxoglutarate; OPPP, oxidative pentose phosphate pathway; PEG, polyethylene glycol; PEP, phosphoenolpyru-

54

versible substrate level phosphorylation of ADP at the expense of phosphoenolpyruvate (PEP), yielding pyruvate, and ATP. In all eukaryotes PK is located in the cytosol, but vascular plant and green algal PKs are known to exist as cytosolic (PK c) and plastid (PK p) isozymes that differ markedly in their respective physical, immunological and kinetic/regulatory characteristics (1). Plant PK is of particular interest because considerable evidence indicates that it is a primary control site of glycolytic flux to pyruvate. A reduction in PEP levels, brought about by an enhancement of PK activity, will stimulate ATP-dependent phosphofructokinase because PEP is a potent inhibitor of plant phosphofructokinases (1). The Arabidopsis Genome Initiative (2) recently indicated that the PK biochemistry of vascular plants is relatively complex. They reported that Arabidopsis thaliana contains at least seven genes that encode different PK polypeptides, with an additional five for PK-like proteins. The molecular and kinetic properties of a variety of highly purified or homogeneous vascular plant PK cs have been studied in detail (3–11). Vascular plant PK c appears to exist as tissue-specific isozymes that demonstrate considerable differences in their respective physical and kinetic and regulatory properties (3–9). By contrast, far less is known about the relatively labile PK p, which has been purified to near or complete homogeneity only from the green alga Selenastrum minutum (12) and endosperm of developing castor (Ricinus communis) beans (13, 14). S. minutum PK p was shown to exist as a novel 235-kDa monomer (12), whereas the purified castor bean PK p was reported to consist of equal proportions of 63.5 (␣-subunit) and 54

vate; PK, pyruvate kinase; PK c and PK p , cytosolic and plastidic pyruvate kinase isozymes, respectively. 0003-9861/02 $35.00 © 2002 Elsevier Science (USA) All rights reserved.

Brassica napus LEUCOPLAST PYRUVATE KINASE

kDa (␤-subunit) polypeptides forming a unique 334kDa ␣ 3␤ 3 heterohexameric native enzyme (14, 15). That Brassica napus (commonly known as rapeseed or canola) and castor bean PK ps share a similar heteromeric subunit structure was suggested by Sangwan and coworkers (16) who observed equal intensity staining anti-(castor bean PK p) IgG immunoreactive polypeptides of approximately 64 and 58 kDa on immunoblots of clarified extracts from developing B. napus zygotic (i.e., developing seed) and microspore-derived embryos. During the enzyme’s extraction and purification, the N-termini of the ␣- and ␤-subunits of castor bean PK p are quite susceptible to partial in vitro proteolysis by an endogenous asparaginyl endopeptidase (14, 15, 17). A kinetic study of partially purified castor bean PK p (18) was probably performed with a substantially proteolyzed enzyme (15). By initiating purification with isolated leucoplasts, and including the thiol protease inhibitor 2,2⬘-dipyridyl disulfide in purification buffers, partial PK p degradation was minimized, but not completely prevented, during its isolation from developing castor beans (14). Very minor proteolytic clipping can exert a profound detrimental influence on the allosteric properties of a plant regulatory enzyme (19), and has been reported to significantly alter the kinetic behavior of human erythrocyte PK (20). Thus far, kinetic data for a nonproteolyzed vascular plant PK p have not been reported. However, relative to castor PK p, the PK p of B. napus embryos and cell cultures appears to be far less susceptible to partial in vitro degradation by endogenous proteases (16). Previous studies revealed that the developmental period during which PK p activity and concentration show maximal increases is coincident with the onset of the most active phase of storage-lipid accumulation by developing castor and tobacco seeds (15, 21), and microspore-derived B. napus embryos (16). These data indicate that a significant role for PK p in developing oilseeds may be to generate precursors (i.e., pyruvate and ATP) for long-chain fatty acid biosynthesis in leucoplasts. PK p could also be important to generate ATP for other plastid-localized anabolic processes such as NH 3⫹-assimilation by glutamine synthetase (GS) (22). There has been considerable attention in the use of nonzygotic embryos of oil-seed crops as a model system for studies of the biochemistry and gene regulation of oil-seed embryogenesis (23). Cell suspension cultures of embryos derived in vitro from pollen grains of B. napus have been reported to closely resemble their developing zygotic (seed embryo) counterpart with respect to fatty acid and storage lipid composition (23), and expression of PK c and PK p (16). We recently described the purification and characterization of PK c and PEP carboxylase from heterotrophic B. napus cell suspension cultures (10, 24). A model was presented

55

that highlights the pivotal role of Asp and Glu in the coordinate allosteric control of these key PEP-utilizing cytosolic enzymes during and following NH 4⫹-assimilation (10). The present paper concerns the purification to near homogeneity of non-proteolyzed B. napus PK p and reports molecular, immunological, and kinetic/regulatory properties of the purified enzyme. MATERIALS AND METHODS Antibodies and plant material. Monospecific rabbit polyclonal antibodies against B. napus PK c, and castor bean PK c and PK p were obtained as previously described (4, 10, 13). An embryogenic pollenderived heterotrophic cell suspension of winter oilseed rape (canola) (B. napus L. cv. Jet Neuf) was cultured and harvested as previously outlined (24). Enzyme and protein assay. The PK reaction was coupled to the lactate dehydrogenase reaction and assayed at 24°C by monitoring NADH oxidation at 340 nm, in a final volume of 1 ml. Unless otherwise stated PK p assay conditions were 50 mM Hepes-KOH (pH 8.0), 2 mM PEP, 1 mM ADP, 1 mM dithiothreitol, 5% (w/v) PEG (average M r 8 kDa), 50 mM KCl, 15 mM MgCl 2, 0.15 mM NADH, and 2.5 units/ml of desalted rabbit muscle lactate dehydrogenase. Assays were: (a) initiated by the addition of enzyme preparation, (b) corrected for contaminating PEP phosphatase activity by omitting ADP from the reaction mixture, and (c) linear with respect to time and PK concentration. One unit of PK activity is the amount of enzyme resulting in the utilization of 1 ␮mol PEP/min at 24°C. Protein concentrations were determined using a Coomassie blue G-250 dyebinding method (25) with bovine ␥-globulin as the protein standard. Kinetic studies. These were performed using a Dynatech MR5000 Microplate reader and a final volume of 0.2 ml for the PK reaction mixture. Apparent V max, K m or S 0.5 , and Hill coefficient (n H) values for substrates and cofactors were calculated from the Hill equation, fitted to a nonlinear least-squares regression computer kinetics program (26). I 50 and K a values ([inhibitor] and [activator] producing 50% inhibition and activation of PK activity, respectively) were likewise calculated. Stock solutions of metabolites were adjusted to pH 8.0, and the concentration of nucleotides was verified spectrophotometrically using published extinction coefficients. The free [Mg 2⫹] was calculated based upon its binding to organophosphates, nucleotides, organic acids, P i, and Cl ⫺ ions using a computer program that automatically corrects for temperature, pH, and ionic strength (27). PK p activity was independent of free [Mg 2⫹] in the range of 4 to 20 mM. For studies of the enzyme’s substrate saturation kinetics and response to metabolite effectors, the stock solutions of nucleotides, PEP, organic acids, and P i were made equimolar with MgCl 2, thus maintaining free [Mg 2⫹] in excess of 4 mM. Metabolite or substrate concentrations stated in the text refer to their total concentration in the assay medium. Enzyme purification. All steps were conducted at 4°C, unless otherwise noted. The protocol for PK p extraction and purification via PEG fractionation, and Butyl-Sepharose hydrophobic interaction and DEAE-Fractogel anion-exchange chromatography were as previously described (10), except that 1 mM 2,2⬘-dipyridyl disulfide was included in all purification buffers. Peak PK p activity fractions from the DEAE-Fractogel column were concentrated to 9 ml using an Amicon YM-30 ultrafilter. A solution of PEG (50% [w/v]) was added to the concentrate to bring the final PEG concentration to 25% (w/v). The solution was stirred for 30 min and centrifuged at 37,000g for 20 min. The resulting pellet was solubilized with 5 ml of blue dextran agarose (BDA) buffer (20 mM Tris–HCl, pH 8.0, containing 1 mM dithiothreitol, 1 mM 2,2⬘-dipyridyl disulfide, 5 mM MgCl 2, 1 mM EDTA, and 20% (v/v) glycerol), centrifuged as above, and loaded at 0.5 ml/min onto a column (0.5 ⫻ 9 cm) of BDA that had been

56

PLAXTON, SMITH, AND KNOWLES TABLE I

Purification of PK p from 385 g of B. napus Suspension Cells Step

Volume (ml)

Activity (units)

Protein (mg)

Specific activity (units/mg)

Purification (fold)

Yield (%)

Clarified extract PEG (4–20%) fractionation Butyl-Sepharose DEAE-fractogel BDA

1040 510 57 9 0.8

411 a 341 a 216 a 68 41

6344 4131 492 36 1.5

0.065 0.082 0.44 1.9 28

— 1.3 6.8 29 431

100 83 52 17 10

a

Calculated by assuming that PK p represents 35% of the total PK activity present in clarified extracts of B. napus suspension cells.

preequilibrated with the BDA buffer. The column, which was equilibrated and run at room temperature, was washed with 30 ml of BDA buffer and 20 ml of BDA buffer containing 2 mM ADP. PK p activity was eluted with BDA buffer containing 2 mM ADP and 2 mM PEP (fraction size ⫽ 3 ml). Peak activity fractions were pooled and concentrated to 1 ml with a YM-100 ultrafilter, diluted to 6 ml with BDA buffer and reconcentrated to 0.8 ml. The concentrated sample was adjusted to contain 2 mM dithiothreitol and 40% (v/v) glycerol, frozen in liquid N 2, and stored at ⫺80°C in 25-␮l aliquots. The activity of the purified PK p was stable for at least 4 weeks when stored frozen. Electrophoresis, immunoblotting, and N-terminal microsequencing. The procedures for SDS–PAGE, determination of subunit M r, and immunoblotting were as described previously (10). For N-terminal microsequencing, purified PK p (30 ␮g) was subjected to SDS– PAGE (using a 1.5-mm-thick slab gel, and a 9% (w/v) monomer concentration in the separating gel), electroblotted onto a Bio-Rad poly(vinylidene) difluoride membrane, stained with Ponceau S (Sigma), and destained with 1% (v/v) acetic acid. Polypeptides corresponding to PK p’s ␣- (64 kDa) and ␤- (58 kDa) subunits were identified by aligning the Ponceau S-stained lane with an adjacent lane containing 50 ng of the final preparation that had been immunoprobed with anti-(castor bean PK p) IgG. The Ponceau S-stained PK p subunits were sequenced by automated Edman degradation at the Harvard Microchemistry Facility (Cambridge, MA). Similarity searches were conducted with the BLAST program using the ‘short but nearly exact match’ option available on the National Center for Biotechnology Information World Wide Web site (28). Other procedures. Native M r estimation (by gel filtration fast protein liquid chromatography on a calibrated Superose 6 HR 10/30 column) and immunotitration of PK p activity were carried out as previously described (10). Extraction of B. napus suspension cells under denaturing conditions was performed according to Wu and Wang (29). This procedure involves homogenization of the tissue in 10% (w/v) trichloroacetic acid, followed by resolubilization of precipitated proteins in SDS–PAGE sample buffer.

Overall recovery of PK activity (representing PK c and PK p) following DEAE-Fractogel chromatography was 30%, and of this total about 35% represented PK p. As shown in Table I, B. napus PK p was purified approximately 430-fold to a final specific activity of 28 units/mg protein and an overall yield of 10%. PEP phosphatase activity was absent in the final PK p preparation. In contrast to B. napus PK c (10), but similar to castor oil plant and green algal PK ps (4, 12–14, 18), the purified B. napus PK p was relatively heat labile, retaining 83, 72, 50, and 0% of its activity following a 3-min incubation at 45, 50, 55, and 60°C, respectively. SDS–PAGE of the final preparation resolved four major Coomassie blue staining bands of about 66, 64, 58, and 56 kDa (Fig. 1A). Only the 64- and 58-kDa polypeptides cross-reacted with anti-(castor bean PK p)-IgG

RESULTS

Enzyme purification, physical, and immunological properties. The maximal PK activity of B. napus suspension cells was about 3 units/g fresh weight. As previously reported (10), DEAE-Fractogel anion-exchange chromatography of the pooled Butyl Sepharose fractions resolved major and minor peaks of PK activity eluting at approximately 100 and 200 mM KCl, respectively. Immunoblotting of the respective pooled fractions with anti-(castor bean PK c or PK p)-IgG established that the early and later eluting PK activity peaks corresponded to PK c and PK p, respectively (10).

FIG. 1. (A) SDS–PAGE (10% (w/v) separating gel) analysis of 4 ␮g of purified B. napus PK p. Protein staining was performed using Coomassie blue R-250. (B) Immunoblot analysis was performed using affinity-purified rabbit anti-(castor bean PK p)-IgG (13). Lane 1 contains 60 ng of the final B. napus PK p preparation; lane 2 contains 30 ␮g of protein from a clarified B. napus extract that was prepared under denaturing conditions in 10% (w/v) trichloroacetic acid according to Wu and Wang (29). Abbreviations used: O, origin; TD, tracking dye front.

Brassica napus LEUCOPLAST PYRUVATE KINASE

FIG. 2. Superose 6 gel filtration fast protein liquid chromatography of purified B. napus PK p. V o denotes the void volume of the column. (Inset) SDS–PAGE followed by immunoblotting using affinity-purified anti-(castor bean PK p)-IgG (13) of 10-␮l aliquots of the PK p peak activity fractions. Abbreviations used: O, origin; TD, tracking dye front.

(Fig. 1B, lane 1), suggesting that the 66- and 56-kDa protein staining bands are contaminating polypeptides. The 64- (␣-subunit) and 58-kDa (␤-subunit) PK p polypeptides copurified in an approximate 1:1 ratio during hydrophobic interaction, anion-exchange, and BDA affinity chromatography. Equal intensity staining 64- and 58-kDa anti-(castor bean PK p) IgG immunoreactive polypeptides were also observed on an immunoblot of a clarified B. napus extract that was prepared under denaturing conditions in the presence of 10% (w/v) trichloroacetic acid (Fig. 1B, lane 2). By contrast, no cross-reactivity was found when an immunoblot of 0.25 ␮g of purified B. napus PK p was probed with anti-(B. napus or castor bean PK c)-IgG.

57

The effect of anti-(B. napus PK c or castor bean PK p) immune serum on the activity of the final B. napus PK p preparation was examined. Complete removal of PK p activity occurred at about 200 ␮l of anti-(castor bean PK p) immune serum per unit of activity. By contrast, anti-(B. napus PK c) immune serum, which effectively immunoprecipitated B. napus PK c (10), exerted no influence on the activity of B. napus PK p. Native molecular mass determination. The native M r of B. napus PK p as estimated by gel filtration of the final preparation on a calibrated Superose 6 column was approximately 380 ⫾ 10 kDa (n ⫽ 3 ⫾ SE). Immunoblot analysis of fractions 62 to 74 from the Superose 6 column demonstrated that the single peak of PK p activity coeluted with 64- and 58-kDa polypeptides that cross-reacted with anti-(castor bean PK p) IgG with similar intensities (Fig. 2). N-Terminal microsequencing. Polypeptides corresponding to the ␣- and ␤-subunits of B. napus PK p were separated by SDS–PAGE, electroblotted onto a poly(vinylidene) difluoride membrane, and subjected to Nterminal microsequencing (Fig. 3). Similarity searches were conducted with the BLAST program using the “short but nearly exact match” option (28). A sequence of 12 amino acid residues of the N-terminus for the PK p ␣-subunit best aligned with a portion of a putative PK sequence deduced from the A. thaliana genome (Fig. 3A). N-terminal analysis of the PK p ␤-subunit yielded a sequence of 14 amino acid residues that exhibited maximal identity with the corresponding region deduced from the nucleotide sequence of a putative PK gene from the eubacterium Methylobacterium extorquens (Fig. 3B). Kinetic properties. The final PK p preparation exhibited a relatively sharp pH-activity profile with op-

FIG. 3. Comparison of the N-terminal amino acid sequence of polypeptides corresponding to the ␣- and ␤-subunits of B. napus PK p and other PKs. (A) Alignment of the N-terminal amino acid sequence of the 64-kDa ␣-subunit with a portion of the deduced amino acid sequences for a putative PK gene from A. thaliana. (B) Alignment of the N-terminal sequence obtained for the 58-kDa ␤-subunit of B. napus PK p with the N-terminal portion of the deduced amino acid sequences for a putative PK gene from the eubacterium M. extorquens. Sequences of A. thaliana and M. extorquens PKs were taken from the GenBank data base (Accession Nos. AY056196 and 005118, respectively). Hyphens denote amino acid residues that are identical to those of the respective B. napus PK p subunits. Underlined letters indicate positions with the conservative substitutions Glu, Asp, Asn, and Gln; Ile, Leu, Met, and Val; Phe, Tyr, and Trp; Ala, Gly and Pro; Ser and Thr, and Arg and Lys using the standard one-letter abbreviations. Numbers in parentheses indicate the position of right side amino acid residues from the N-terminus. M r values for the PK p subunits are based upon their relative mobility during SDS–PAGE (see Fig. 1), whereas those for A. thaliana and M. extorquens PK are predicted values.

58

PLAXTON, SMITH, AND KNOWLES TABLE III

Influence of Various Metabolites on the Activity of B. napus PK p a

FIG. 4. Activity of purified B. napus PK c and PK p as a function of assay pH. Assays were conducted using saturating concentrations of PEP and ADP (2 and 1 mM, respectively) and were buffered by a mixture of 25 mM Mes and 25 mM Bis-Tris-propane, titrated to the appropriate pH with either KOH or HCl. Data for PK c were taken from Smith et al. (10). All values represent the means (⫾SE) of n ⫽ 4 separate determinations.

timal activity occurring at about pH 8.0 (Fig. 4). All subsequent kinetic studies were conducted at pH 8.0. Cofactor requirements and substrate saturation kinetics. Table II summarizes the kinetic constants obtained for metal cation cofactors and substrates. B. napus PK p exhibited an absolute requirement for the simultaneous presence of both monovalent and divalent metal cations, with K ⫹ and Mg 2⫹ fulfilling this role. Mg 2⫹ exhibited cooperative binding, whereas K ⫹, PEP, and ADP followed hyperbolic saturation kinetics (Table II). The elimination of 5% (w/v) PEG from the

TABLE II

Kinetic Constants of B. napus PK p for Substrates and Metal Cation Cofactors a Parameter

Addition

V max (units/mg protein)

— 0.5 mM 6-P gluconate 8 mM 2-oxoglutarate — 0.5 mM 6-P gluconate 8 mM 2-oxoglutarate — 0.5 mM 6-P gluconate 8 mM 2-oxoglutarate — —

K m (PEP) (mM)

K m (ADP) (mM) K m (K ⫹) (mM) S 0.5 (Mg 2⫹) (mM)

29.5 39.8 20.8 0.052 0.029 0.24 0.14 0.073 0.15 2.6 0.67 (1.6)

a All assays were conducted at pH 8.0. Hill coefficients were equivalent to 1.0 except where the value is indicated in parentheses. All values are the means of at least four independent determinations and are reproducible to within ⫾10% (SE) of the mean value.

Metabolite

Concentration tested (mM)

Relative activity

Ka (mM)

I 50 (mM)

6-P Gluconate 2-Oxoglutarate Citrate Oxalate Isocitrate Malate ATP

1 10 10 0.2 10 10 2

205 38 67 48 70 72 84

0.12 — — — — — —

— 8.3 (2.0) 18.4 0.23 ND b ND ND

a Assays were conducted at pH 8.0 using subsaturating concentrations of PEP and ADP (0.06 and 0.15 mM, respectively). Enzymatic activity in the presence of effectors is expressed relative to the respective control set at 100%. Hill coefficients were equivalent to 1.0 except where the value is indicated in parentheses. All values are the means of at least four independent determinations and are reproducible to within ⫾10% (SE) of the mean value. b ND, not determined.

PK p assay mixture caused an approximate 25% reduction in the enzyme’s apparent V max without affecting the K m(PEP) value. This organic solute has been demonstrated to significantly enhance the activity of homogeneous castor bean PK c by stabilizing the active tetrameric structure of the native enzyme in dilute solutions (7). Although B. napus PK p can employ alternative nucleotide diphosphates as the phosphoryl acceptor, activities obtained with saturating (10 mM) GDP, CDP, UDP, and IDP were only 57, 25, 22, and 16%, respectively, of those obtained with saturating ADP. Thus, ADP is the preferred nucleotide substrate for the enzyme. No inhibition of B. napus PK p by NDP concentrations up to 10 mM was observed. Metabolite effects. A wide variety of compounds were tested as possible effectors of purified B. napus PK p with subsaturating concentrations of PEP and ADP (0.06 and 0.15 mM, respectively). The following compounds had no influence (⫾ 15% of the control rate) on PK p activity: KPi, Gln, Glu, Asp, and Asn (20 mM each); acetate and succinate (10 mM each); Ala, Gly, Ser, AMP, Lys, ADP-glucose, ribose 5-P, Fru 1,6-P 2, Fru 6-P, Glc 6-P, 3-P glycerate, 2-P glycerate, and glycerol 3-P (5 mM each); Arg, Cys, UDP-galactose, dihydroxyacetone-P, glycolate, NaNO 3, NH 4Cl, and Triton X-100 (2 mM each); PP i, NADH, NADPH, NAD ⫹, and NADP ⫹ (0.5 mM each); Phe, Trp, Tyr, CoA, malonyl-CoA, acetyl-CoA, shikimate, quercetin, and rutin (0.1 mM each); Fru 2,6-P 2, oleoyl-CoA, and oleate (50 ␮M each). Table III lists those compounds that significantly influenced the activity of the purified enzyme.

Brassica napus LEUCOPLAST PYRUVATE KINASE

Activators. The only activator identified was 6-P gluconate (Tables II and III). As an activator, 6-P gluconate functions by increasing apparent V max by approximately 35%, while lowering the K m(PEP) and K m(ADP) values by 40 and 50%, respectively (Table II). The presence of 2 mM 2-oxoglutarate (2-OG) did not significantly influence the enzyme’s K a value for, or fold-activation by, 6-P gluconate. Inhibitors. The most effective inhibitors were 2-OG and oxalate, with somewhat less pronounced inhibition caused by malate, citrate, isocitrate, and ATP (Table III). 2-OG showed cooperative saturation kinetics (n H ⫽ 2.0) and appeared to function as a mixed-type inhibitor with respect to PEP since the addition of 8 mM 2-OG caused a 30% reduction in V max and increased the K m(PEP) value by almost fivefold (Table II). In contrast, 8 mM 2-OG did not alter the K m(ADP) value. The presence of 1 mM 6-P gluconate did not influence the I 50 or n H values for 2-OG. DISCUSSION

Enzyme purification, physical, and immunological properties. The final specific activity of the purified B. napus PK p (28 units/mg; Table I) compares favorably with that of 41 units/mg reported for the near homogenous, but proteolyzed, castor bean PK p (13), and 51 units/mg for homogenous B. napus PK c (10). All attempts to further purify B. napus PK p resulted in inactivation or marked reductions in its specific activity. The native M r of the purified enzyme was determined to be approximately 380 kDa. This value is significantly greater than those of 200 to 250 kDa that have been reported for most other plant and nonplant PKs, but is similar to the value of 334 kDa reported for developing castor bean PK p (14). When the final preparation was subjected to SDS–PAGE and immunoblotting, equal intensity staining polypeptides of 64 and 58 kDa were observed that strongly cross-reacted with anti-(castor bean PK p)-IgG (Fig. 1). The detection of the 64- and 58-kDa immunoreactive polypeptides following trichloroacetic acid extraction of B. napus suspension cells (Fig. 1B) indicates that the 58-kDa ␤-subunit was not a proteolytic degradation product of the 64-kDa ␣-subunit. These results are analogous to those of Sangwan and coworkers (16), who also observed equally abundant immunoreactive polypeptides of about 64 and 58 kDa when immunoblots of clarified extracts from developing B. napus seed or microsporederived embryos were probed with the anti-(castor bean PK p)-IgG. Overall, the data indicate that the final preparation is a PK p, and that similar to the developing castor bean enzyme (14, 15) the native B. napus PK p exists as an unusual heterohexameric protein composed of equal proportions of ␣ (64 kDa) and ␤ (58 kDa) subunits.

59

N-terminal sequence analyses. The N-terminal sequence of the B. napus PK p ␣-subunit showed maximal alignment (66%) with a portion of a putative PK sequence deduced from the genome of A. thaliana (Fig. 3A). This sequence similarity rises to 92% if conservative amino acid substitutions are taken into account. In developing castor oil seeds, transit peptides of 44 and 60 amino acids are respectively cleaved from the PK p ␣- and ␤-subunit preproteins following their import into the leucoplast (14). Therefore, the 90 amino acids upstream of the B. napus N-terminus that are observed in the deduced A. thaliana PK sequence may represent a transit peptide. The N-terminal sequence of the B. napus PK p ␤-subunit was 57% identical (71% conservative) to the corresponding region deduced from the nucleotide sequence of a putative PK gene from the eubacterium M. extorquens (Fig. 3B). This supports the findings of Hattori and coworkers (30) who conducted phylogenetic analyses of the primary structures of plant and nonplant PKs to conclude that castor and tobacco seed PK p sequences are more closely related to eubacterial homologs, than they are to eukaryotic PK. These results are also consistent with the widely accepted tenet that plastids arose via the endosymbiosis of a cyanobacterium-like ancestor by a eukaryotic phagotroph. Kinetic properties. B. napus PK p exhibited a narrow pH-activity profile with optimal activity occurring at pH 8.0, which contrasts with the relatively broad pH optimum of pH 7.0 observed for purified B. napus PK c (Fig. 4). Several other leucoplast-localized enzymes, including developing castor bean PK p (18) and phosphofructokinase (31), and the pyruvate dehydrogenase complex and acetyl-CoA carboxylase from pea root plastids (32) also demonstrate optimal activity at pH 8.0. Light-dependent alkalinization of the chloroplast stroma is thought to contribute to the metabolic control of several Calvin cycle enzymes. However, we are unaware of any stromal pH estimates for leucoplasts from nongreen plant tissue. This, as well as the potential role of pH in the control of leucoplast metabolism awaits further research. In common with other vascular plant and green algal PKs, B. napus PK p demonstrated hyperbolic saturation kinetics for PEP, ADP, and K ⫹, whereas Mg 2⫹ binding showed positive cooperativity (Table II). The absolute requirement of B. napus PK p for the simultaneous presence of monovalent (fulfilled by K ⫹) and divalent (fulfilled by Mg 2⫹) metal cations is shared by most PKs studied to date. However, the apparent cyanobacterial origin of plant PK p (30) is difficult to reconcile with the strongly sigmoidal PEP saturation kinetics (n H ⫽ 2.6) and total absence of monovalent cation dependency that was recently observed with purified PK from the cyanobacterium Synechococcus PCC 6301 (33).

60

PLAXTON, SMITH, AND KNOWLES

FIG. 5. A model for the allosteric control of PK p in heterotrophic B. napus suspension cells. The activation of PK p by 6-P gluconate provides a mechanism for the coordinate stromal generation of NADPH and ATP for leucoplast NH 4⫹-assimilation and/or long chain fatty acid biosynthesis. See text for details. The abbreviations are: ACC, acetyl-CoA carboxylase; FAS, fatty acid synthetase complex; FNR, ferredoxin reductase; FD OX and FD RED, oxidized and reduced forms of ferredoxin, respectively; G6PDH, Glc 6-P dehydrogenase; GOGAT, glutamine 2-oxoglutarate transaminase; GS, glutamine synthetase; NIR, nitrite reductase; OPPP, oxidative pentose-P pathway; Ru5P, ribulose 5-P; 6-PG, 6-P gluconate; 6PGDH, 6-P gluconate dehydrogenase;1,3-diPGA, 1,3-P 2 glycerate; 3-PGA, 3-P glycerate; PGK, P-glycerate kinase.

Metabolite effectors. The relatively weak inhibition of B. napus PK p by ATP (Table III), coupled with the lack of effect of AMP, indicates that similar to B. napus PK c (10) energy charge does not play a major role in regulating the in vivo activity of the enzyme. Apart from their dissimilar pH-activity profiles (Fig. 4), significant kinetic differences between B. napus PK c and PK p include the observations that the flavonoids rutin and quercetin were potent inhibitors of B. napus PK c (I 50 values ⱕ 0.1 mM) (10), but exerted no influence on the activity of B. napus PK p. Although the physiological relevance of these results is unknown, the differential influence of quercetin and rutin on B. napus PK c and PK p could provide an effective means to discriminate between their respective activities in clarified B. napus extracts. B. napus PK p activity was also not influenced by Asp or Glu, metabolites demonstrated to be key allosteric effectors of B. napus PK c (Asp functions as an activator by reversing the inhibition of PK c by Glu) and PEP carboxylase (inhibited by both Asp

and Glu) (10, 24). Asp and Glu were hypothesized to have an important function in the coordinate control of the cytosolic PEP branchpoint of B. napus, while feedback regulating the glycolytic provision of C-skeletons and respiratory substrates needed to support NH 4⫹assimilation by GS/GOGAT (10). The activity of B. napus PK p was responsive to several metabolites involved in C- and N-metabolism (Tables II and III). As with most other PKs, the enzyme demonstrated potent inhibition by oxalate, which is believed to arise from the close structural similarity between oxalate and the enolate form of pyruvate. Measurement of the plastid oxalate pool is necessary to determine the physiological relevance of the inhibition of B. napus PK p by oxalate. Inhibition by 2-OG indicates a potential role for this key intermediate of plant C- and N-metabolism in the control of B. napus PK p in vivo. The central importance of 2-OG as the immediate C-skeleton precursor for plastidic NH 4⫹-assimilation by GS/GOGAT is well understood (22, 34). As 2-OG can

Brassica napus LEUCOPLAST PYRUVATE KINASE

reflect C/N status, it has been postulated to have a signaling role in coordinating plant C- and N-metabolism (34). This is consistent with the ability of 2-OG to inhibit several plant enzymes including PK c, PEP carboxylase, citrate synthase, and the alternative oxidase (34). The physiological relevance of the inhibition of B. napus PK p by 2-OG is equivocal since the enzyme’s I 50 for this compound (8.3 mM) is about two orders of magnitude greater than the 2-OG concentration of 0.07 mM estimated for the stroma of illuminated spinach leaves (35). Nevertheless, it will be of interest to assess the 2-OG concentration of the leucoplast stroma, and how this level may fluctuate during transient NH 4⫹assimilation. An interesting kinetic feature of B. napus PK p was its allosteric activation by 6-P gluconate (Tables II and III), which to our knowledge has not been reported for any PK. The enzyme’s K a value for 6-P gluconate (0.12 mM) is likely well within the physiologically relevant concentration range of this compound within the leucoplast stroma. 6-P Gluconate is the product Glc 6-P dehydrogenase, which catalyzes the first committed step of the oxidative pentose-P pathway (OPPP) (Fig. 5). There is substantial evidence that the OPPP of nongreen plastids, including B. napus leucoplasts, is essential for generating the reducing power (NADPH) required by various plastid-localized anabolic pathways, particularly long chain fatty acid synthesis and N-assimilation (22, 36). The activity of Glc 6-P dehydrogenase of nongreen plastids is potently controlled by the ratio of [NADPH]/[NADP ⫹] (22). A decreased redox ratio leads to Glc 6-P dehydrogenase activation and vice versa. In addition to NADPH, plastid anabolism depends upon the stromal provision of ATP, which must either be generated internally or imported from the cytosol via the adenylate translocator of the plastid envelope (22). Acetate-dependent fatty acid synthesis in developing castor bean leucoplasts was stimulated fourfold by the addition of PEP as compared to exogenously added ATP (37). Similarly, when exogenous ATP was replaced by PEP, a significant rate of fatty acid synthesis from 14C-pyruvate by isolated leucoplasts of developing B. napus embryos was observed (PEP supported 40% of the ATP-dependent rate) (Kubis and Rawsthorne, personal communication). Thus, PK p-generated ATP appears to be capable of either replacing or supplementing cytosolic ATP in supporting anabolic processes of nongreen plastids. The marked expression of a PEP-specific translocator in the inner envelope of nongreen, but not photosynthetic, plastids from C 3-plants (38) implies a fundamental metabolic importance for the PK p-mediated substrate level phosphorylation of ADP to ATP within the leucoplast stroma. In particular, ATP is needed as a co-substrate for GS and acetyl-CoA carboxylase that

61

respectively catalyze the first committed steps of NH 4⫹assimilation and fatty acid synthesis (22). CONCLUSIONS

The native PK p from leucoplasts of heterotrophic B. napus suspension cells appears to possess a complex ␣ 3␤ 3 heterohexameric subunit structure. Elucidation of the functional roles for the enzyme’s ␣- and ␤-subunits remains a challenging problem. Kinetic studies indicate an important physiological role for PK p in the complex carbon metabolism of B. napus leucoplasts. As outlined in the model presented in Fig. 5, feedforward activation of B. napus PK p by 6-P gluconate is hypothesized to help coordinate reductant generation (via OPPP) with stromal ATP production for anabolic processes within B. napus leucoplasts. Any demand for reductant will reduce the stromal ratio of [NADPH]: [NADP ⫹], leading to Glc 6-P dehydrogenase activation and enhanced OPPP flux. The consequent rise in [6-P gluconate] will feedforward activate PK p for ATP provision, whilst the hydrolysis of ATP by anabolic enzymes such as GS or acetyl-CoA carboxylase directly regenerates the PK p cosubstrate ADP. PK p activation by 6-P gluconate would also stimulate the production of pyruvate, which can be utilized as C-precursor for several anabolic pathways (i.e., fatty acid synthesis), or exported to the cytosol to serve as a respiratory substrate for the mitochondrion. The properties displayed in vitro by this novel regulatory enzyme appear to make it well suited to serve as an ATP and carbon substrate donor for anabolic pathways of nongreen plastids. ACKNOWLEDGMENT The authors thank Dr. Steve Rawsthorne for helpful discussions.

REFERENCES 1. Plaxton, W. C. (1996) Annu. Rev. Plant Physiol. Plant Mol. Biol. 47, 185–214. 2. The Arabidopsis Genome Initiative (2000) Nature 408, 796 – 813. 3. Plaxton, W. C. (1988) Plant Physiol. 86, 1064 –1069. 4. Plaxton, W. C. (1989) Eur. J. Biochem. 181, 443– 451. 5. Podesta´ , F. E., and Plaxton, W. C. (1991) Biochem J. 279, 495– 501. 6. Podesta´ , F. E., and Plaxton, W. C. (1992) Biochim. Biophys. Acta 1160, 213–220. 7. Podesta´ , F. E., and Plaxton, W. C. (1993) Plant Physiol. 103, 285–288. 8. Podesta´ , F. E., and Plaxton, W. C. (1994) Planta 194, 381–387. 9. Hu, Z-H., and Plaxton, W. C. (1996) Arch. Biochem. Biophys. 333, 298 –307. 10. Smith, C. R., Knowles, V. L., and Plaxton, W. C. (2000) Eur. J. Biochem. 267, 4477– 4485. 11. Turner, W. L., and Plaxton, W. C. (2000) Biochem. J. 252, 875– 882.

62

PLAXTON, SMITH, AND KNOWLES

12. Knowles, V. L., Dennis, D. T., and Plaxton, W. C. (1989) FEBS Letts. 259, 130 –132. 13. Plaxton, W. C., Dennis, D. T., and Knowles, V. L. (1990). Plant Physiol. 94, 1528 –1534. 14. Negm, F. B., Cornel, F. A., and Plaxton, W. C. (1995) Plant Physiol. 109, 1461–1469. 15. Plaxton, W. C. (1991) Plant Physiol. 97, 1334 –1338. 16. Sangwan, R. S., Gauthier, D., Turpin, D. H., Pomeroy, M. K., and Plaxton, W. C. (1992) Planta 187, 198 –202. 17. Cornel, F. A., and Plaxton, W. C. (1994) Physiol. Plant. 91, 599 – 604. 18. Ireland, R. J., DeLuca, V., and Dennis, D. T. (1980) Plant Physiol. 65, 1188 –1193. 19. Plaxton, W. C., and Preiss, J. (1987) Plant Physiol. 83, 105–112. 20. Marie, J., Garreau, H., and Kahn, A. (1977) FEBS Letts. 78, 9194. 21. Gottlob-McHugh, S., Knowles, V. L., Blakeley, S. D., Sangwan, R. S., Dennis, D. T., Plaxton, W. C., and Miki, B. L. (1995) Physiol. Plant. 95, 507–514. 22. Neuhaus, H. E., and Emes, M. J. (2000) Annu. Rev. Plant Physiol. Plant Mol. Biol. 51, 111–140. 23. Weselake, R. J., and Taylor, D. C. (1999) Prog. Lipid Res. 38, 401– 460. 24. Moraes, T. F., and Plaxton, W. C. (2000) Eur. J. Biochem. 267, 4465– 4476. 25. Bollag, D. M., Rozycki, M. D., and Edelstein, S. J. (1996) Protein Methods, pp. 62– 67, Wiley-Liss, New York.

26. Brooks, S. P. G. (1992) Biotechniques 13, 906 –911. 27. Brooks, S. P. G. , and Storey, K. B. (1992) Anal. Biochem. 201, 119 –126. 28. Altshul, S. F., Madden, T. L., Schaffer, A. A., Zhang, J., Zhang, Z., Miller, W., and Lipman, D. J. (1997) Nucleic Acids Res. 25, 3389 –3402. 29. Wu, F. S., and Wang, M. Y. (1984) Anal. Biochem. 139, 100 –103. 30. Hattori, J., Baum, B. R., McHugh, S. G., Blakeley, S. D., Dennis, D. T., and Miki, B. L. (1995) Biochem. Syst. Ecol. 23, 773–780. 31. Knowles, V. L., Greyson, M. F., and Dennis, D. T. (1990) Plant Physiol. 92, 155–159. 32. Qi, Q., Trimming, B. A., Kleppinger-Sparace, K. F., Emes, M. J., and Sparace, S. A. (1996) J. Exp. Bot. 47, 1889 –1896. 33. Knowles, V. L., Smith, C. S., Smith, C. R., and Plaxton, W. C. (2001) J. Biol. Chem. 276, 20966 –20972. 34. Lancien, M., Gadal, P., and Hodges, M. (2000) Plant Physiol. 123, 817– 824. 35. Winter, H., Robinson, D. G., and Heldt, H. W. (1994) Planta 193, 530 –535. 36. Rawsthorne, S., Kang, F., and Eastmond, P. J. (1999) in Regulation of Primary Metabolic Pathways in Plants (Kruger N. J., Hill, S. A., and Ratcliffe, R. G., Eds.), Proceedings of the Phytochemical Society of Europe, Vol. 42, pp. 137–157, Kluwer Academic, Dordrecht, The Netherlands. 37. Boyle, S. A., Hemmingsen, S. M., and Dennis, D. T. (1990) Plant Physiol. 92, 151–154. 38. Flu¨ gge, U-I. (1999) Annu. Rev. Plant. Physiol. Plant. Mol. Biol. 50, 27– 45.