Purification and characterization of pyrophosphateand ATP-dependent phosphofructokinases from banana fruit

Transcription

1 Planta (2003) 217: DOI /s ORIGINAL ARTICLE William L. Turner Æ William C. Plaxton Purification and characterization of pyrophosphateand ATP-dependent phosphofructokinases from banana fruit Received: 19 August 2002 / Accepted: 20 November 2002 / Published online: 14 January 2003 Ó Springer-Verlag 2003 Abstract Pyrophosphate-dependent phosphofructokinase (PFP; EC ) and two isoforms of ATPdependent phosphofructokinase (PFK I and PFK II; EC ) from ripened banana (Musa cavendishii L. cv. Cavendish) fruits were resolved via hydrophobic interaction fast protein liquid chromatography (FPLC), and further purified using anion-exchange and gel filtration FPLC. PFP was purified 1,158-fold to a final specific activity of 13.9 lmol fructose 1,6-bisphosphate produced (mg protein) )1 min )1. Gel filtration FPLC and immunoblot analyses indicated that this PFP exists as a 490-kDa heterooctomer composed of equal amounts of 66- (a) and 60-kDa (b) subunits. PFP displayed hyperbolic saturation kinetics for fructose 6-phosphate (Fru 6-P), PPi, fructose 1,6-bisphosphate, and Pi (K m values = 32, 9.7, 25, and 410 lm, respectively) in the presence of saturating (5 lm) fructose 2,6-bisphosphate, which elicited a 24-fold enhancement of glycolytic PFP activity (K a =8 nm). PFK I and PFK II were each purified about 350-fold to final specific activities of lmol fructose 1,6-bisphosphate produced (mg protein) )1 min )1. Analytical gel filtration yielded respective native molecular masses of 210 and 160 kda for PFK I and PFK II. Several properties of PFK I and PFK II were consistent with their respective designation as plastid and cytosolic PFK isozymes. PFK I and PFK II exhibited: (i) ph optima of 8.0 and 7.3, respectively; (ii) hyperbolic saturation kinetics for ATP (K m =34 and 21 lm, respectively); and (iii) sigmoidal saturation kinetics for Fru 6-P (S 0.5 =540 and 90 lm, respectively). Allosteric effects of phosphoenolpyruvate (PEP) and Pi W.L. Turner Æ W.C. Plaxton (&) Department of Biology, Queen s University, Kingston, Ontario, K7L 3N6, Canada Fax: plaxton@biology.queensu.ca UR: W.C. Plaxton Department of Biochemistry, Queen s University, Kingston, Ontario, K7L 3N6, Canada on the activities of PFP, PFK I, and PFK II were characterized. Increasing concentrations of PEP or Pi progressively disrupted fructose 2,6-bisphosphate binding by PFP. PEP potently inhibited PFK I and to a lesser extent PFK II (I 50 =2.3 and 900 lm, respectively), while Pi activated PFK I by reducing its sensitivity to PEP inhibition. Our results are consistent with: (i) the respiratory climacteric being regulated by fine (allosteric) control of pre-existing enzymes; and (ii) primary and secondary glycolytic flux control being exerted at the levels of PEP and Fru 6-P metabolism, respectively. Keywords Banana (ripening) Æ Carbohydrate metabolism Æ Musa (fruit) Æ Plant glycolysis Æ Glycolytic control Abbreviations FPLC: fast protein liquid chromatography Æ Fru 6-P: fructose 6-phosphate Æ Fru 1,6-P 2 : fructose 1,6-bisphosphate Æ Fru 2,6-P 2 : fructose 2,6- bisphosphate Æ PEP: phosphoenolpyruvate Æ PEPCase: phosphoenolpyruvate carboxylase Æ PFK and PFP: ATP- and PPi-dependent phosphofructokinase, respectively Æ PK c : cytosolic pyruvate kinase Introduction Owing to its highly predictable pattern of carbohydrate metabolism during ripening, the banana fruit represents an ideal model system in which to investigate the control of glycolysis and gluconeogenesis in vascular plants (Beaudry et al. 1989; Ball et al. 1991). The initiation of ripening in bananas is associated with a marked increase in respiration, termed the respiratory climacteric. This respiratory increase is closely followed by the massive conversion of starch, which comprises approximately 20% of the fresh weight of unripe fruit, into sucrose (Seymour 1993). Enhanced glycolytic flux and the associated rise in mitochondrial respiration at the climacteric are believed to generate ATP for the conversion of starch to sucrose and associated substrate (futile) cycles

2 114 (Hill and ap Rees 1994), as well as carbon skeletons needed for transamination reactions and other anabolic processes within this tissue (Law and Plaxton 1995; Turner and Plaxton 2000). Control of initiation of the respiratory climacteric is desirable for increasing the storage life of climacteric fruits. The initiation of the respiratory climacteric in ripening bananas has been correlated with reduced and elevated phosphoenolpyruvate (PEP) and pyruvate contents, respectively (Beaudry et al. 1989; Ball et al. 1991), indicating that activation (or de-inhibition) of cytosolic pyruvate kinase (PK c ; EC ) and/or PEP carboxylase (PEPCase; EC ) has occurred. That the control of the metabolism of fructose 6-phosphate (Fru 6-P) to fructose 1,6-bisphosphate (Fru 1,6-P 2 )by ATP- and/or PPi-dependent phosphofructokinase (PFK and PFP, respectively) is also important was indicated by the marked elevation in the Fru-1,6-P 2 :Fru-6-P concentration ratio in climacteric bananas (Beaudry et al. 1989; Ball et al. 1991). Kinetic studies of homogeneous banana PK c and PEPCase demonstrating complex allosteric control of these enzymes resulted in a model for the control of cytosolic glycolysis and PEP partitioning during banana fruit ripening (Law and Plaxton 1995, 1997; Turner and Plaxton 2000). In agreement with studies in many other plant and green algal systems (Plaxton 1996), this model suggests that primary and secondary control of glycolytic flux in ripening banana fruit is exerted at the level of PEP and Fru 6-P metabolism, respectively. Around 1980, the discovery of the strictly cytosolic PFP and its potent activation by micromolar levels of the signal metabolite fructose 2,6-bisphosphate (Fru 2,6- P 2 ) was reported in plants (Carnal and Black 1979; Sabularse and Anderson 1981). PFP catalyzes a parallel reaction to PFK, but unlike PFK is readily reversible and close to equilibrium in vivo (Stitt 1990). The discovery of PFP has added further confusion to the control of Fru 6-P metabolism in ripening fruit and to the relative roles of PFP/PFK compared to PK/PEPCase during the respiratory climacteric. In addition, some debate has centred around the maximal extractable activity of PFK during the transition of unripe to ripe bananas, which has been suggested to either increase (Salminen and Young 1975; Iyer et al. 1989a; Ball et al. 1991) or remain constant (Mertens et al. 1987). Furthermore, in contrast to other plant PFKs (Plaxton 1996), banana PFK was reported to: (i) be insensitive to activation by Pi or inhibition by PEP (Surendranathan et al. 1990), (ii) utilize ATP or ADP as a phosphoryl donor (Surendranathan and Nair 1989), and (iii) undergo subunit-dissociation (oligomer to monomer) upon fruit ripening (Iyer et al. 1989a, 1989b; Surendranathan and Nair 1989). During these studies, the purification or kinetic characterization of PFP was not undertaken and, to our knowledge, has not been presented for banana fruit elsewhere. The anomalous physical and kinetic characteristics described for banana fruit PFK and the lack of data for PFP present a problem for our current model of glycolytic control in this tissue, hypothesized to be highly dependent on the metabolism of PEP by PK c and PEPCase (Turner and Plaxton 2000). Therefore, to extend our previous characterization of key control enzymes of banana fruit glycolysis and phospate metabolism (Law and Plaxton 1995, 1997; Turner and Plaxton 2000, 2001) we have purified and characterized PFP and two isoforms of PFK from ripened banana fruit. Our results indicate that PFP and PFK are under allosteric feedback inhibition by PEP, and therefore lend support to the hypothesis that primary and secondary glycolytic flux control in ripening bananas lies in the metabolism of PEP and Fru 6-P, respectively. Materials and methods Chemicals and plant material Butyl-Sepharose fast flow resin, a Superose-6 HR 10/50 column, and M r standards were from Amersham/Pharmacia. DTT and phenylmethylsulfonylfluoride were from ICN Pharmaceuticals. Ethylene glycol, glycerol, KCl, and Fractogel EMD DEAE-650 (S) anionexchange resin were from Merck/BDH. All other chemicals were of analytical grade and obtained from Sigma Chemical Co. Rabbit anti- (potato tuber PFP) IgG and homogeneous potato tuber PFP were obtained as previously described (Podesta et al. 1994). Ethylene pre-treated green banana (Musa cavendishii L. cv. Cavendish) fruit were purchased from a local retailer and ripened in the dark in a well-ventilated room maintained at 25 C. Enzyme assays Unless otherwise indicated, assay conditions for PFK were 50 mm Hepes KOH (ph 8.0 and 7.3 for PFK I and PFK II, respectively), 5 mm Mg-acetate, 4 mm Fru 6-P, 1 mm ATP, 0.15 mm NADH, 1 unit aldolase, 10 units triose-phosphate isomerase and 1 unit glycerol 3-phosphate dehydrogenase. Standard assay conditions for PFP in the glycolytic direction were identical to those for PFK II except that 1 mm PPi replaced ATP, and 5 lm Fru 2,6-P 2 was included. PFK and PFP assays were corrected for NADH oxidase activity by respectively omitting ATP or PPi from the reaction mixture. PFP activity in the gluconeogenic direction was assayed in 50 mm Hepes KOH (ph 7.0) containing 0.5 mm Fru 1,6-P 2, 5 mm Pi, 5 lm Fru 2,6- P 2, 5 mm Mg-acetate, 0.5 mm NAD +, 2 units phosphoglucose isomerase, and 1 unit glucose 6-phosphate dehydrogenase (Leuconostoc mesenteroides). Reverse PFP assays were corrected for Fru 1,6-P 2 phosphohydrolase as previously described (Theodorou et al. 1992). All Fru 6-P and Fru 1,6-P 2 stock solutions were routinely acid pre-treated (titrated to ph 3 with HCl, incubated for 20 min, and then neutralized with KOH) to hydrolyze any contaminating traces of Fru 2,6-P 2. All assays were initiated by the addition of enzyme preparation and routinely monitored at 340 nm in a Gilford 260 recording spectrophotometer in a final volume of 1 ml, and remained linear with respect to time and concentration of enzyme assayed. Coupling enzymes were desalted before use. One unit of activity is defined as the amount of activity that catalyzes the utilization of 1 lmol of substrate min )1 at 25 C. Kinetic studies and determination of protein concentration Kinetic studies were performed using a Spectramax 250 microplate reader (Molecular Devices) in a final volume of 250 ll. V max, K m, S 0.5, Hill coefficient, K a, and I 50 values were calculated using a

3 115 computer kinetics program as previously described (Brooks 1992; Turner and Plaxton 2000). The K a and I 50 values respectively represent the concentration of activator required to achieve half-maximal activation and concentration of inhibitor required to achieve 50% inhibition of enzymatic activity. All kinetic parameters are means of three or more independent determinations and are reproducible to within ±10% (SE) of the mean value. Concentrations of free Mg 2+ were calculated based upon the respective binding to organophosphates, nucleosides, Pi, and/or acetate ions using a computer program that automatically corrects for temperature, ph, and ionic strength (Brooks and Storey 1992). PFK and PFP activities were independent of free Mg 2+ concentrations in the range mm. For kinetic studies, a minimal free Mg 2+ concentration of 4 mm was maintained by making stock solutions of nucleosides and organophosphates equimolar with MgCl 2. Substrate concentrations stated in the text refer to their total concentration in the assay medium. Protein concentration was determined using a Coomassie Blue G-250 dye-binding method using bovine c-globulin as the protein standard (Bollag et al. 1996). Buffers used in the purification of banana fruit PFK and PFP Buffer A contained 1.2 M KPi (ph 7.8), 100 mm KCl, 10 mm diethyldithiocarbamate, 10 mm thiourea, 1 mm EDTA, 1 mm DTT, 1 mm phenylmethylsulfonylfluoride, 1 mm 2,2 -dipyridyl disulfide, 10% (v/v) glycerol and 2% (w/v) polyvinyl(polypyrrolidone). Buffer B contained 1.2 M KPi (ph 7.8), 1 mm EDTA, 1 mm DTT and 20% (v/v) glycerol. Buffer C contained 50 mm KPi (ph 7.8), 4 mm Mg-acetate, 1 mm EDTA, 1 mm DTT, 1 mm Fru 6-P and 20% (v/v) ethylene glycol. Buffer D contained 50 mm Hepes KOH (ph 7.8), 4 mm Mg-acetate, 1 mm EDTA, 1 mm DTT, 1 mm Fru 6-P and 20% (v/v) glycerol. Buffer E contained 25 mm Hepes KOH (ph 7.8), 100 mm KCl, 0.5 mm EDTA, 1 mm DTT, 1 mm Mg-acetate and 20% (v/v) glycerol. Purification of banana fruit PFK and PFP All steps were carried out at 0 4 C unless otherwise noted. Peeled banana fruit (700 g) were diced and homogenized in 1.4 l of buffer A using a Waring blender followed by a Polytron. The homogenate was centrifuged at 14,000 g for 20 min and the supernatant filtered through two layers of Miracloth. The supernatant was gently stirred for 45 min with 90 ml of Butyl-Sepharose that had been pre-equilibrated in buffer B. The resin was packed into a column (22 cm long, 2.2 cm i.d.), connected to an ÄKTA fast protein liquid chromatography (FPLC) system and washed with buffer B at 4 ml min )1 until the A 280 decreased to baseline. The column was developed at 2 ml min )1 with a linear gradient (400 ml) of 0 to 100% buffer C (100 to 0% buffer B). Two peaks of PFK activity (PFK I and PFK II) eluted at approximately 50% and 80% buffer C, respectively, whereas a single peak of PFP activity eluted at 100% buffer C (Fig. 1A). Fractions of 9 ml were collected and those containing more than 20% peak PFK or PFP activity were separately pooled, concentrated to approx. 10 ml by ultrafiltration over a YM-30 membrane (Amicon), and frozen and stored in liquid N 2. All subsequent purification procedures were identical for each of the three pools of enzyme activity eluting from the Butyl-Sepharose column, and were performed sequentially on each sample. The frozen concentrates were rapidly thawed and desalted using a Sephadex G-50 column (6 cm long, 2.5 cm i.d.) that had been pre-equilibrated in buffer D. The desalted samples were loaded at 1.5 ml min )1 onto a column (5 cm long, 1 cm i.d.) of Fractogel EMD DEAE-650 (S) that had been connected to the FPLC system and pre-equilibrated with buffer D. The column was washed with buffer D until the A 280 decreased to baseline, and then developed at Fig. 1 Butyl-Sepharose FPLC elution profiles of banana (Musa cavendishii) fruit PFK I, PFK II, and PFP (A), and subsequent DEAE-Fractogel FPLC elution profiles for PFK I (B) and PFK II (C) 1 ml min )1 with a linear gradient (45 ml) of 0 to 250 mm KCl in buffer D. PFK I and PFK II (Fig. 1B, C), and PFP (not shown) each eluted as a single peak of activity at about 160, 100, and 140 mm KCl, respectively. Fractions (1 ml) containing greater than 25% peak activity were pooled and concentrated to 0.5 ml by ultrafiltration over a YM-30 membrane. Each sample was applied (0.25 ml min )1 ), in two separate 0.25-ml injections, onto a prepacked Superose-6 HR 10/50 column that was connected to the FPLC system and pre-equilibrated with buffer E. Fractions (0.75 ml) containing PFK or PFP activity were pooled and concentrated to approximately 0.5 ml using an Amicon Centricon-10 ultrafilter. The retentate was divided into 20-ll aliquots, frozen in liquid N 2, and stored at )80 C. The activity of the final preparations of PFK I, PFK II, and PFP were stable for at least 2 months when stored frozen. Estimation of native molecular mass by gel filtration FPLC This was performed during purification of each enzyme by FPLC on the Superose-6 column. Native M r was calculated from a plot of K d (partition coefficient) against log M r using the following protein standards: thyroglobulin (669 kda), ferritin (440 kda), catalase (232 kda), aldolase (158 kda), and BSA (67 kda).

4 116 Results PFK and PFP activities during banana fruit ripening The total PFK and PFP activities (measured in the glycolytic direction) in clarified extracts of ripened banana fruit pulp were equivalent. Some variation was observed between batches of fruit with specific activities ranging from to units mg )1 for PFK and to units mg )1 for PFP. The specific activities of both enzymes remained unchanged during the transition from unripe (green) to ripe (yellow) fruits (results not shown). That we were determining the maximal extractable activity of each enzyme was confirmed by mixing experiments in which equivalent portions of unripe and ripe fruit were combined and extracted. The resultant extract yielded the same PFK or PFP activity as determined in individual extracts of ripe or unripe fruit. In addition, the recovery of the activity of exogenously added homogeneous potato tuber PFP during the extraction of either unripe or ripe fruit tissue was at least 90%. Fruit that had been treated with ethylene and allowed to ripen for 5 days at 25 C were employed in all subsequent studies. Purification of PFK and PFP from banana fruit The initial Butyl-Sepharose FPLC step resolved two distinct peaks of PFK activity (PFK I and PFK II), and a single peak of PFP activity (Fig. 1A). The three enzyme pools were subjected to anion-exchange FPLC followed by gel filtration FPLC. In contrast to the Butyl- Sepharose step, the elution order of PFK I and PFK II was reversed during anion-exchange FPLC such that PFK II and PFK I eluted at 100 and 160 mm KCl, respectively (Fig. 1B, C). As shown in Table 1, PFK I and PFK II were respectively purified 375- and 344-fold to similar final specific activities of 6.0 and 5.5 units mg )1. PFP purification resulted in a final specific activity of 13.9 units mg )1, representing a 1,158-fold purification (Table 2). Instability of both PFK isoforms prevented their purification beyond gel filtration FPLC. However, although impure as judged by SDS PAGE followed by Coomassie Blue R-250 staining (results not shown), the final preparations of PFK I, PFK II, and PFP were free of the following contaminating enzyme activities: aldolase, hexose-phosphate isomerase, ATPase, PEP- Case, PEP phosphatase, and PK. Therefore, the final PFK I, PFK II, and PFP preparations were considered to be suitable for their subsequent kinetic characterization. In addition, fructose 1,6-bisphosphatase activity was undetectable in clarified banana fruit extracts, or in the final PFK and PFP preparations. The absence of extractable fructose 1,6-bisphosphatase activity in banana fruit was reported by Ball and co-workers (1991). Physical properties The native Mr values of PFK I, PFK II, and PFP were 210, 160, and 490 kda, respectively, as estimated by gel filtration FPLC on a calibrated Superose-6 column. Immunoblots of clarified extracts from unripe and ripe fruit, as well as the final PFP preparation were probed with anti-(potato tuber PFP) IgG and uniformly revealed two immunoreactive polypeptides that crossreacted with a 1:1 ratio in signal intensity, and that that respectively co-migrated with the a- (66 kda) and b- (60 kda) subunits of homogeneous potato tuber PFP (results not shown). Therefore, ripe banana fruit PFP appears to exist as an a 4 b 4 -heterooctomer. Table 1 Purificationof PFK I and PFK II from 700 g of ripe banana (Musacavendishii) fruit Step Volume Activity Protein Specific activity Purification Yield (ml) (units) (mg) (units mg )1 ) (-fold) (%) Clarified extract 1, Butyl-Sepharose PFK-I PFK-II DEAE-Fractogel PFK-I PFK-II Superose-6 PFK-I PFK-II Table 2 Purificationof PFP from 700 g of ripe banana fruit Volume Activity Protein Specific activity Purification Yield (ml) (units) (mg) (units mg )1 ) (-fold) (%) Clarified extract 1, Butyl-Sepharose DEAE-Fractogel Superose ,158 17

5 117 Kinetic properties The ph/activity profiles for PFK I and PFK II were dissimilar with maximal activity occurring at 8.0 and 7.3, respectively (Fig. 2). When assayed in the glycolytic direction and in the presence of 5 lm Fru 2,6-P 2,PFP exhibited a broad ph/activity profile centred at about ph 7.1 (Fig. 2). Substrate saturation kinetics No activity was obtained with either PFK isoform when Fru 6-P was replaced by glucose 6-phosphate, glucose 1- phosphate, or fructose 1-phosphate (4 mm each). Positive cooperativity was evident for the binding of Fru-6-P to PFK I (S 0.5 =540 lm; Hill coefficient = 1.9) and PFK II (S 0.5 =90 lm; Hill coefficient = 1.6), whereas ATP binding followed Michaelis Menten kinetics. PFK I and PFK II utilized alternative nucleoside triphosphates as substrates (Table 3). However, as indicated by the respective specificity constants (V max /K m values), ATP is the preferred substrate for both isoforms. When ADP, UDP, or PPi (1 mm each) were substituted for ATP no activity was observed with PFK I or PFK II. Table 4 lists the kinetic parameters of PFP measured in the forward (glycolytic) and reverse (gluconeogenic) directions in the presence and absence of 5 lm Fru 2,6-P 2. Hyperbolic substrate saturation kinetics were uniformly observed, except in the absence of Fru 2,6-P 2 when cooperative binding of Fru 1,6-P 2 occurred (Table 4). Metabolite effectors PEP, glycolate 2-phosphate, glycerate 2-phosphate, and glycerate 1,3-diphosphate inhibited PFK I to a greater degree than PFK II (Table 5), whereas ADP inhibition of both PFK isoforms was similar. AMP (2 mm) exerted no influence on the activity of either PFK isoform. However, Pi (2.5 mm) served as an activator of PFK I Table 3 Use of alternative nucleoside triphosphates by banana fruit PFK I and PFK II. The standard spectrophotometric assay was performed at ph 8.0 and 7.3, respectively,for PFK I and PFK II, while varying the concentration of theselected nucleoside triphosphates. n.d. Not determined Nucleoside V max K m V max /K m (units mg )1 ) (lm) (units mg )1 ÆmM )1 ) PFK I ATP UTP ITP CTP GTP 1.3 n.d. n.d. PFK II ATP UTP ITP CTP GTP Table 4 Influenceof Fru 2,6-P 2 on substrate saturation kinetics of the forward (glycolytic)and reverse (gluconeogenic) reactions of banana fruit PFP. The standard spectrophotometric assay was performed at ph 7.3, except that the concentration of Fru 6-P,PPi, Fru 1,6-P 2 or Pi was varied. The values in parentheses denotethe Hill coefficient for Fru 1,6-P 2 binding Parameter )Fru 2,6-P 2 +5 lm Fru 2,6-P 2 Forward reaction V max (units mg )1 protein) K m (Fru 6-P) (mm) K m (PPi) (lm) Reverse reaction V max (units mg )1 protein) K m (Fru 1,6-P 2 )(lm) 82 a (1.7) 25 (1.0) K m (Pi) (mm) a This value represents an S 0.5 value as thehill coefficient was >1.0 Table 5 Influence of2.5 mm Pi on the inhibition of banana fruit PFK I and PFK IIby various metabolites. Assays were conducted at ph 8.0 and ph 7.3for PFK I and PFK II, respectively, with subsaturating concentrations (approx. K m values) of substrate. For PFK I the concentrations of Fru 6P and ATP were 600 and 40 lm, respectively, while the respective values for PFK II were 100 and 20 lm Metabolite PFK I I 50 (lm) PFK II I 50 (mm) )Pi +2.5 mm Pi )Pi +2.5 mm Pi Fig. 2 Activity of banana fruit PFK I, PFK II, and PFP as a function of ph. Assays were buffered using a mixture of 20 mm Mes and 20 mm Bis-Tris-Propane, titrated to the appropriate ph with either KOH or HCl. PFP activity was determined in the presence of 5 lm Fru 2,6-P 2. All values are means of n=3 determinations and are reproducible to within ±10% (SE) of the mean value PEP Glycolate phosphate Glycerate phosphate Glycerate ,3-diphosphate ADP 1,900 2,

6 118 by relieving its sensitivity to the various inhibitors (Table 5). The K a of PFP for Fru 2,6-P 2 was 8 nm in the forward direction, with the addition of saturating (5 lm) Fru 2,6-P 2 resulting in a 24-fold increase in PFP activity. PFP activation by 5 lm Fru 2,6-P 2 in the reverse direction was much lower (approx. 2-fold). Similar results have been obtained with other plant PFPs and this has been ascribed to Fru 1,6-P 2 binding to the Fru 2,6- P 2 allosteric activator site of the enzyme (Kombrink et al. 1984; Nielsen 1995). The influence of PEP and Pi on PFP activity was also examined. Pi inhibited the forward and reverse reactions yielding I 50 values of 1.5 and 17 mm, respectively. Moreover, Pi and particularly PEP markedly influenced the enzyme s K a (Fru 2,6-P 2 ) (measured in the forward direction), with this parameter significantly increasing in the presence of progressively higher concentrations of either metabolite (Fig. 3). PK c and PEPCase were demonstrated to be under potent allosteric control by glutamate (inhibits both enzymes) and aspartate (activates PK c, inhibits PEP- Case) in a variety of plant tissues, including ripening banana fruit (Law and Plaxton 1995; Moraes and Plaxton 2000; Turner and Plaxton 2000). It was hypothesized that this regulation is involved in the coordinate control of the cytosolic PEP branchpoint as pertains to the production of Krebs cycle intermediates (i.e., oxaloacetate and 2-oxoglutarate) required for transamination reactions and/or NH 4 + -assimilation into organic form. However, neither glutamate (10 mm) nor aspartate (10 mm) influenced the activity of banana fruit PFK I, PFK II, or PFP when the respective substrates were present at concentrations equating to their approximate K m values. Fig. 3 Influence of PEP and Pi on the K a for Fru 2,6-P 2 of banana fruit PFP. Assays were conducted at ph 7.3 in the glycolytic direction using subsaturating concentrations of Fru 6-P (0.05 or 1 mm) and PPi (50 lm). All values are means of n=3 determinations and are reproducible to within ±10% (SE) of the mean value Discussion The aim of this study was to simultaneously purify and characterize banana fruit PFK and PFP. We wished to determine how these enzymes might be influenced by metabolite effectors, particularly PEP, that have been implicated in the control of glycolytic flux during banana fruit ripening (Plaxton 1996). As with banana PK and PEPCase (Ball et al. 1991; Law and Plaxton 1995; Turner and Plaxton 2000), but in contrast to the 2- to 5-fold increase in extractable PFK activity that has been described during banana fruit ripening (Saliminen and Young 1975; Iyer et al. 1989a; Ball et al. 1991), the specific activities of PFK and PFP were equivalent and remained constant throughout ripening. This indicates that the stimulation of glycolysis at the climacteric is not due to de novo biosynthesis of these enzymes. It is notable that when the extraction buffer lacked polyvinyl(polypyrrolidone), thiourea, and diethyldithiocarbamate, the PFK and PFP activities of the resultant clarified extracts were up to 80% lower than when these protective agents were present (data not shown). This indicates that banana PFK and PFP are quite susceptible to inactivation by phenols and tannins, which are particularly abundant in the unripe fruit. This provides a rationale for the previously reported low PFK activity of preclimacteric fruit, and unusual physical and kinetic properties described for banana fruit PFK (Nair and Darak 1981; Surendranathan and Nair 1989; Surendranathan et al. 1990, 1992). Phenolic- and tanninscavenging compounds, or protease inhibitors, were not included in the extraction buffers used in the aforementioned studies. Initial adsorption of proteins present in the clarified banana extract onto Butyl-Sepharose hydrophobic media quickly eliminated gelatinous and pectic substances that otherwise caused the clarified extract to gel within 30 min. When the Butyl-Sepharose column was developed with a decreasing KPi gradient, two peaks of PFK and one peak of PFP activity were eluted (Fig. 1A). All three peaks of enzyme activity were further purified using anion-exchange and gel filtration FPLC. Although impure, all three final preparations were judged to be suitable for kinetic analyses. Localization of PFK isozymes to cytosolic and plastidic compartments has been demonstrated for a variety of plant tissues (Garland and Dennis 1980a, 1980b; Botha and Small 1987; Knowles et al. 1990; Kelly and Latzko 1977). Thus, the PFK isoforms purified in this study might represent compartment-specific isozymes. If so, PFK I and PFK II likely correspond to plastid and cytosolic PFK isozymes, respectively. This reasoning is based on the following: i. The ph-activity optima of plastid and cytosolic PFK isozymes from several other plant tissues are approximately ph 8.0 and ph 7.2, respectively (Kelly and Latzko 1977; Garland and Dennis 1980b; Botha et al.

7 ) in this study PFK I and PFK II exhibited maximal activities at ph 8.0 and 7.3, respectively (Fig. 2). ii. Plant cytosolic PFK has been demonstrated to elute prior to plastid PFK during anion-exchange chromatography using a linear salt gradient (Garland and Dennis 1980b; Botha et al. 1988) in this study PFK I and PFK II eluted at 160 and 100 mm KCl, respectively (Fig. 1B, C). iii Plastid PFKs tend to be more potently inhibited by PEP as compared to cytosolic PFKs (Garland and Dennis 1980a, 1980b; Dennis and Greyson 1987) Table 5 shows that the I 50 (PEP) values of banana PFK I and PFK II were in the micromolar and millimolar range, respectively. The latter findings contrast with those of Surendranathan and co-workers (1990) who failed to observe any influence of PEP on the activity of a purified banana PFK. Further studies are required to confirm whether banana PFK I and PFK II represent cytosolic and plastidic PFK isozymes. No clear trend is apparent for either the native or subunit molecular mass of PFK isozymes from various plant species (Dennis and Greyson 1987). However, during banana fruit ripening, PFK was suggested to undergo subunit dissociation from a 390-kDa multimeric complex (unripe fruit) to a 66-kDa dimer and a subsequent 33-kDa monomer (ripe fruit;iyer et al. 1989a, 1989b). This was hypothesized to activate PFK in vivo and to be an important mechanism of metabolic control leading to the respiratory climacteric. Banana PFK I and PFK II were determined by gel filtration FPLC to have respective native molecular masses of about 280 and 160 kda. Ripening-associated changes in these values were not observed, and interconversion of either isoform did not occur upon re-chromatography on the Superose-6 column (data not shown). As described for PFP from potato tubers and Brassica nigra cell cultures (Podesta et al. 1994; Theodorou and Plaxton 1996), the native enzyme of ripe banana fruit appears to exist as a 480-kDa a 4 /b 4 -heterooctomer. Immunoblotting of time-course extracts revealed that banana PFP is uniformly composed of an equivalent ratio of a- (66 kda) and b- (60 kda) subunits throughout ripening. Our results indicate that subunit association or dissociation of PFK I, PFK II, or PFP is not likely to be a fine control mechanism that contributes to the in vivo stimulation of glycolysis and respiration characteristic of climacteric bananas. By contrast, three oligomeric forms of PFP (i.e., a 443-kDa a/b-heteromer, a 68-kDa b-doublet, and a 68-kDa b-singlet) have been isolated from unripe (green) and ripe (red) tomatoes, with the relative proportion of each isoform changing during ripening (Wong et al. 1990). The substrate saturation kinetics obtained for banana PFK I and PFK II are in good agreement with previous reports on other plant PFKs (Kelly and Latzko 1977; Garland and Dennis 1980a, 1980b; Dennis and Greyson 1987; Botha et al. 1988). Positive cooperatively was observed for the binding of Fru 6-P, whereas alternate hexose monophosphates did not serve as substrates. While ATP was the most effective phosphoryl donor, other nucleoside triphosphates, notably UTP, could also be utilized (Table 3). These results contrast with those of Surendranathan and Nair (1989) indicating that the monomeric PFK of ripe bananas utilized ATP, UTP, CTP or GTP equally effectively. No activity was observed with PFK I or PFK II when ADP or PPi substituted for ATP in the assay mixture. This also conflicts with a previous report that banana fruit PFK efficiently utilized ADP as a phosphoryl donor (Surendranathan and Nair 1989). In agreement with reports on other plant PFKs (Salminen and Young 1975; Dennis and Greyson 1987; Cawood et al. 1988), ADP actually inhibited PFK I and PFK II (Table 5). This is expected since ADP is a product of the PFK reaction. PFP substrate saturation kinetics determined in the forward and reverse directions (Table 4), and its respective activation and inhibition by Fru 2,6-P 2 and Pi were consistent with previously published findings obtained with other plant PFPs (Plaxton 1996; Theodorou and Plaxton 1996; Theodorou and Kruger 2001). The enzyme s glycolytic activity was stimulated 24-fold by saturating Fru 2,6-P 2, yielding a very low K a (Fru 2,6-P 2 ) of 8 nm. This activation was associated with enhanced binding of the substrates Fru 6-P, PPi, and Fru 1,6-P 2 (Table 4), whereas Fru-2,6-P 2 binding was progressively disrupted by increasing concentrations of PEP or Pi (Fig. 3). Our results demonstrate that Fru 6-P-metabolizing enzymes from ripening banana fruit are under potent allosteric control by various metabolites (Table 5, Fig. 3). PEP inhibited PFK I and PFK II, with Pi functioning as a PFK-I activator by relieving its inhibition by PEP (Table 5). Pi-mediated disruption of Fru- 2,6-P 2 binding to plant PFP has been demonstrated (Kombrink and Kruger 1984; Stitt 1989; Theodorou and Plaxton 1996; Theodorou and Kruger 2001) and was observed for banana fruit PFP in this study (Fig. 3). However, less well documented was the more substantive increase of PFP s K a (Fru 2,6-P 2 ) brought about by increasing PEP levels, relative to similar increases in Pi (Fig. 3). In particular, the pronounced decrease in in vivo PEP concentrations that is the initial response of glycolysis at the climacteric (Ball et al. 1991), should serve to make PFP more responsive to activation by physiologically relevant levels of Fru 2,6-P 2 that remain relatively constant at about lm 1 in the cytosol of ripening bananas. In conclusion, we have characterized the substrate saturation kinetics and allosteric control properties of PFK I, PFK II, and PFP isolated from ripening banana fruit. Our results indicate that the stimulation of 1 This concentration range was estimated from the data reported by Ball and ap Rees (1988) by assuming that Fru-2,6-P 2 is restricted to the cytosol and that the cytosol represents 10% of the cell s volume.

8 120 glycolysis that is characteristic of the respiratory climacteric is regulated by fine (allosteric) control of preexisting enzyme activity. In particular, the kinetic properties of PFK I, PFK II, and PFP (this study), and of PK c and PEPCase (Law and Plaxton 1995; Turner and Plaxton 2000), are consistent with maximal glycolytic flux control being exerted at the level of PEP utilization, with secondary control at the level of Fru 6-P metabolism. This is in accord with the quantification of metabolite changes during banana ripening (Ball et al. 1991; Hill and ap Rees 1994), and metabolic control analysis of transgenic potato tubers with altered levels of PFK (Thomas et al. 1997a, 1997b). It would be of interest to apply the tools of metabolic control analysis to carbohydrate metabolism in climacteric fruit. In addition, biochemical analyses of the relative in vivo contributions of cytosolic versus amyloplastic glycolysis, and mitochondrial pyruvate metabolism would facilitate a more thorough understanding of fruit ripening, one of the most economically important processes in agriculture. Acknowledgement Financial support was kindly provided by research and equipment grants from the Natural Sciences and Engineering Research Council of Canada (NSERC). References Ball KL, ap Rees T (1988) Fructose 2,6-bisphosphate and the climacteric in bananas. Eur J Biochem 177: Ball KL, Green JH, ap Rees T (1991) Glycolysis at the climacteric of bananas. Eur J Biochem 197: Beaudry RM, Severson RF, Black CC, Kays SJ (1989) Banana ripening: implications of changes in glycolytic intermediate concentrations, glycolytic and gluconeogenic carbon flux, and fructose 2,6-bisphosphate concentration. Plant Physiol 91: Bollag DM, Rozycki MD, Edelstein SJ (1996) Protein methods. Wiley Liss, New York, pp Botha FC, Small JGC (1987) Comparison of the activities and some properties of pyrophosphate and ATP dependent fructose 6-phosphate 1-phosphotransferase of Phaseolus vulgaris seeds. Plant Physiol 83: Botha FC, Cawood MC, Small JGC (1988) Kinetic properties of the ATP-dependent phosphofructokinase isoenzymes from cucumber seeds. Plant Cell Physiol 29: Brooks SPG (1992) A simple computer program with statistical tests for the analysis of enzyme kinetics. Biotechniques 13: Brooks SPG, Storey KB (1992) Bound and determined: a computer program for making buffers of defined ion concentrations. Anal Biochem 201: Carnal NW, Black CC (1979) Pyrophosphate dependent 6-phosphofructokinase, a new glycolytic enzyme in pineapple leaves. Biochem Biophys Res Commun 86:20 26 Cawood MC, Botha FC, Small JGC (1988) Properties of the phosphofructokinase isoenzymes from germinating cucumber seeds. J Plant Physiol 132: Dennis DT, Greyson MF (1987) Fructose 6-phosphate metabolism in plants. Physiol Plant 69: Garland WJ, Dennis DT (1980a) Plastid and cytosolic phosphofructokinase from the developing endosperm of Ricinus communis. Separation, purification and initial characterization of the isoenzymes. Arch Biochem Biophys 204: Garland WJ, Dennis DT (1980b) Plastid and cytosolic phosphofructokinase from the developing endosperm of Ricinus communis. II Comparison of the kinetic and regulatory properties of the isoenzymes. Arch Biochem Biophys 204: Hill SA, ap Rees T (1994) Fluxes of carbohydrate metabolism in ripening bananas. Planta 192:52 60 Iyer MG, Surendranathan KK, Nair PM (1989a) Correlation between increase in 6-phosphofructokinase activity and appearance of three multiple forms in ripening banana. Physiol Biochem 27: Iyer MG, Surendranathan KK, Nair PM (1989b) Evidence for the formation of lower molecular weight forms of phosphofructokinase in ripening banana by dissociation of oligomeric form. Plant Physiol Biochem 27: Kelly GJ, Latzko E (1977) Chloroplast phosphofructokinase. II. Partial purification, kinetic and regulatory properties. Plant Physiol 60: Knowles VL, Greyson MF, Dennis DT (1990) Characterization of ATP-dependent fructose 6-phosphate 1-phosphotransferase isozymes from leaf and endosperm tissues of Ricinus communis. Plant Physiol 92: Kombrink E, Kruger NJ (1984) Inhibition by metabolic intermediates of pyrophosphate:fructose 6-phosphate phosphotransferase from germinating castor bean endosperm. Z Pflanzenphysiol 114: Kombrink E, Kruger NJ, Beevers H (1984) Properties of pyrophosphate:fructose 6-phosphate phophotransferase from germinating castor bean endosperm. Plant Physiol 74: Law RD, Plaxton WC (1995) Purification and characterization of a novel phosphoenolpyruvate carboxylase from banana fruit. Biochem J 307: Law RD, Plaxton WC (1997) Regulatory phosphorylation of banana fruit phosphoenolpyruvate carboxylase by a copurifying phosphoenolpyruvate carboxylase-kinase. Eur J Biochem 247: Mertens E, Marcellin P, Van Schaftingen E, Hers H-G (1987) Effects of ethylene treatment on fructose 2,6-bisphosphate and on the activity of phosphofructokinase-2/fructose 2,6-bisphosphatase in banana. Eur J Biochem 167: Moraes TF, Plaxton WC (2000) Purification and characterization of phosphoenolpyruvate carboxylase from Brassica napus (rapeseed) suspension cell cultures. Implications for phosphoenolpyruvate carboxylase regulation during phosphate starvation and the integration of glycolysis with nitrogen assimilation. Eur J Biochem 267: Nair PM, Darak BG (1981) Identification of multiple forms of phosphofructokinase in ripening dwarf cavendish banana. Phytochemistry 20: Nielsen TH (1995) Fructose 1,6-bisphosphate is an allosteric activator of pyrophosphate:fructose 6-phosphate 1-phosphotransferase. Plant Physiol 108:69 73 Plaxton WC (1996) The organization and regulation of plant glycolysis. Annu Rev Plant Physiol Plant Mol Biol 47: Podestá FE, Moorhead GBG, Plaxton WC (1994) Potato tuber pyrophosphate-dependent phosphofructokinase: effect of thiols and polyalcohols on its intrinsic fluorescence, oligomeric structure, and activity in dilute solutions. Arch Biochem Biophys 313:50 57 Sabularse DC, Anderson RL (1981) D-fructose 2,6-bisphosphate: a naturally occurring activator for inorganic pyrophosphate:dfructose 6-phosphate 1-phosphofructotransferase. Biochem Biophys Res Commun 103: Salminen SO, Young RE (1975) The control properties of phosphofructokinase in relation to the respiratory climacteric in banana fruit. Plant Physiol 55:45 50 Seymour GB (1993) The banana. In: Seymour GB, Taylor JA, Tucker GA (eds) The biochemistry of fruit ripening. Chapman and Hall, London, pp Stitt M (1989) Product inhibition of potato tuber pyrophosphatedependent phosphofructokinase by phosphate and pyrophosphate. Plant Physiol 89:

9 121 Stitt M (1990) Fructose 2,6-bisphosphate as a regulatory molecule in plants. Annu Rev Plant Physiol Plant Mol Biol 41: Surendranathan KK, Nair PM (1989) ADP as a phosphoryl donor for banana 6-phosphofructokinase. Plant Sci 59:45 51 Surendranathan KK, Iyer MG, Nair PM (1990) Characterization of a monomeric phosphofructokinase from banana role of magnesium in its regulation. Plant Sci 72:27 35 Surendranathan KK, Iyer MG, Nair PM (1992) Mechanism of action of a dimeric phosphofructokinase from banana: role of magnesium on its kinetics and regulation. Plant Sci 81:29 36 Theodorou ME, Kruger NJ (2001) Physiological relevance of fructose 2,6-bisphosphate in the regulation of spinach leaf pyrophosphate:fructose 6-phosphate 1-phosphotransferase. Planta 213: Theodorou ME, Plaxton WC (1996) Purification and characterization of pyrophosphate-dependent phosphofructokinase from phosphate-starved Brassica nigra suspension cells. Plant Physiol 112: Theodorou ME, Cornel FA, Duff SM, Plaxton WC (1992) Phosphate starvation-inducible synthesis of the alpha-subunit of the pyrophosphate-dependent phosphofructokinase in black mustard suspension cells. J Biol Chem 267: Thomas S, Mooney PJ, Burrell MM, Fell DA (1997a) Finite change analysis of glycolytic intermediates in tuber tissue of lines of transgenic potato (Solanum tuberosum) overexpressing phosphofructokinase. Biochem J 322: Thomas S, Mooney PJ, Burrell MM, Fell DA (1997b) Metabolic control analysis of glycolysis in tuber tissue of potato (Solanum tuberosum): explanation for the low control coefficient of phosphofructokinase over respiratory flux. Biochem J 322: Turner WL, Plaxton WC (2000) Purification and characterization of cytosolic pyruvate kinase from banana fruit. Biochem J 352: Turner WL, Plaxton WC (2001) Purification and characterization of banana fruit acid phosphatase. Planta 214: Wong JH, Kiss F, Wu M-X, Buchanan BB (1990) Pyrophosphate fructose-6-p 1-phosphotransferase from tomato fruit. Evidence for change during ripening. Plant Physiol 94: