Class II Recombinant Phosphoribosyl Diphosphate Synthase from Spinach

A recombinant form of spinach (Spinacia oleracea) phosphoribosyl diphosphate (PRPP) synthase isozyme 3 resembling the presumed mature enzyme has been synthesized in an Escherichia coli strain in which the endogenous PRPP synthase gene was deleted, and has been purified to near homogeneity. Contrary to other PRPP synthases the activity of spinach PRPP synthase isozyme 3 is independent of Pi, and the enzyme is inhibited by ribonucleoside diphosphates in a purely competitive manner, which indicates a lack of allosteric inhibition by these compounds. In addition spinach PRPP synthase isozyme 3 shows an unusual low specificity toward diphosphoryl donors by accepting dATP, GTP, CTP, and UTP in addition to ATP. The kinetic mechanism of the enzyme is an ordered steady state Bi Bi mechanism with K ATPand K Rib-5-P values of 170 and 110 μm, respectively, and a V maxvalue of 13.1 μmol (min × mg of protein)−1. The enzyme has an absolute requirement for magnesium ions, and maximal activity is obtained at 40 °C at pH 7.6.

A recombinant form of spinach (Spinacia oleracea) phosphoribosyl diphosphate (PRPP) synthase isozyme 3 resembling the presumed mature enzyme has been synthesized in an Escherichia coli strain in which the endogenous PRPP synthase gene was deleted, and has been purified to near homogeneity. Contrary to other PRPP synthases the activity of spinach PRPP synthase isozyme 3 is independent of P i , and the enzyme is inhibited by ribonucleoside diphosphates in a purely competitive manner, which indicates a lack of allosteric inhibition by these compounds. In addition spinach PRPP synthase isozyme 3 shows an unusual low specificity toward diphosphoryl donors by accepting dATP, GTP, CTP, and UTP in addition to ATP. The kinetic mechanism of the enzyme is an ordered steady state Bi Bi mechanism with K ATP and K Rib-5-P values of 170 and 110 M, respectively, and a V max value of 13.1 mol (min ؋ mg of protein) ؊1 . The enzyme has an absolute requirement for magnesium ions, and maximal activity is obtained at 40°C at pH 7.6.
The compound 5-phospho-D-ribosyl ␣-1-diphosphate (PRPP) 1 is a precursor in several important metabolic pathways, including purine, pyrimidine, and pyridine nucleotide de novo and salvage synthesis, and histidine and tryptophan synthesis (1,2). Certain microorganisms also utilize PRPP for the synthesis of methanopterin or polyprenylphosphate pentoses (3,4). The synthesis of PRPP is catalyzed by PRPP synthase: Rib-5-P ϩ ATP 3 PRPP ϩ AMP (5). PRPP synthase is encoded by the PRS gene(s). Eukaryotic organisms usually contain more than one PRS gene. Analysis of a cDNA library as well as genome sequencing have revealed the presence of at least five PRS genes in the flowering plant Arabidopsis thaliana (Ref. 6, GenBank TM /EBI database accession no. AC004521). Spinach (Spinacia oleracea) appears to contain at least four PRS genes, coding for PRPP synthase isozyme 1 to 4, all of which have enzymatic activity. This was established by analysis of a cDNA library for complementation of an Escherichia coli ⌬prs allele (7). Spinach PRPP synthase isozyme 2 and 3 contain 76 and 87 additional amino acids at their N-terminal ends, respectively, compared with spinach PRPP synthase isozyme 4 and PRPP synthases from E. coli (8), Bacillus subtilis (9), or man (10). These transit peptides of the spinach PRPP synthases contain consensus motifs for maturation of the polypeptides during entry into organelles. Experimental evidence and amino acid sequence comparison revealed that spinach PRPP synthase isozyme 2 was found within the chloroplast, whereas isozyme 3 appeared to be located within the mitochondrion and isozyme 4 in the cytosol. The location of spinach PRPP synthase isozyme 1 is unknown. The deduced N-terminal sequence of the presumed mature isozyme 3 polypeptide is Asn-Ser-Val-Glu-Phe; the polypeptide contains 320 amino acids, and has a calculated molecular mass of 35,497 Da. This value is similar to that for PRPP synthases from most other organisms (7).
In the present work we report the properties of spinach PRPP synthase isozyme 3. The enzyme was synthesized in E. coli and was specified by a gene, which had been manipulated to make the gene product resemble the presumed mature polypeptide, i.e. without a transit peptide. Spinach PRPP synthase isozyme 3 and the "classical" PRPP synthases from E. coli, B. subtilis, and man differ greatly in their specificity for diphosphoryl donor, dependence of P i for activity, and in their allosteric properties. We show that spinach PRPP synthase isozyme 3 belongs to a novel class of PRPP synthases, Class II.
Manipulation of Spinach PRPP Synthase Isozyme 3-Specifying cDNA-To construct pBK862, DNA of pBK843 (7) was digested by restriction endonucleases BstBI and EcoRV. The resulting DNA fragments were subjected to electrophoresis in a 1% agarose gel and the 1048-base pair DNA fragment containing PRS3 was cut out and purified (Qiagen). Ten picomoles of each of the oligodeoxyribonucleo-tides 5Ј-AATTCATTAAAGAGGAGAAATTAACTATGAATTCCGTCGA-GTTTT (MNS) and 5Ј-CGAAAACTCGACGGAATTCATAGTTAATTTC-TCCTCTTTAATG (MNS-komp) were annealed by heating to 95°C and slowly cooled to 0°C. This annealing resulted in the formation of a double-stranded DNA fragment with EcoRI and BstBI overhangs. The underlined nucleotides of the MNS oligodeoxyribonucleotide indicate the codons specifying the N-terminal end of the presumed mature spinach PRPP synthase isozyme 3 polypeptide. DNA of pUHE23-2 2 was digested by restriction endonucleases PvuII and EcoRI. The three DNA species were mixed and ligated by T4 DNA ligase (Promega), transformed (14) to E. coli strain HO1088 followed by a selection for prs ϩ (2). The nucleotide sequence of the resulting plasmid (pBK862) was confirmed by sequencing using an Abi Prism 310 DNA Sequencer as recommended by the supplier (PerkinElmer).
Purification of PRPP Synthase and Protein Methods-Two liters of NZY broth supplemented with ampicillin, tetracycline, and isopropyl 1-thio-␤-D-galactopyranoside were inoculated with 100 ml of overnight culture of strain HO1088/pBK862 and incubated with shaking for 20 h. The following procedures were carried out at 4°C. Cells (6 g of wet weight) were harvested by centrifugation (Sorvall, GS3 rotor) at 6000 rpm for 12 min, washed in 0.9% NaCl, collected by centrifugation (Sorvall, SS34 rotor), resuspended in 30 ml of 50 mM Tris/HCl, pH 7.6 and homogenized for 6 ϫ1 min in a Soniprep ultrasonic disintegrator (Measuring and Scientific Equipment, model 150). Debris was removed by centrifugation at 10,000 rpm for 15 min. Streptomycin sulfate (10% w/v in 50 mM Tris/HCl, pH 7.6) was added to a final concentration of 1%, and the precipitate was removed by centrifugation at 10,000 rpm for 15 min. The supernatant fluid was 45% saturated with solid (NH 4 ) 2 SO 4 . The precipitate was removed by centrifugation, and the resulting supernatant fluid was 65% saturated with solid (NH 4 ) 2 SO 4 . After centrifugation, the precipitate was dissolved in 50 mM Tris/HCl, pH 7.6, and an equal volume of a solution of 10% (w/v) polyethylene glycol 6000 was added. The precipitate was collected by centrifugation, dissolved in 50 mM Tris/HCl, pH 7.6 and applied to a column (1.0 ϫ 25 cm) of Dyematrex Gel Blue B (Millipore). After washing the column with three volumes of 50 mM Tris/HCl, pH 7.6 protein was eluted at 1 ml min Ϫ1 with a 50-ml linear 0.0 -0.5 M KCl gradient in 50 mM Tris/HCl, pH 7.6. Fractions containing PRPP synthase activity were combined and applied to a DE52 anion exchange column (Whatman; 1.0 ϫ 30 cm). Protein was eluted at 1 ml min Ϫ1 with a 100-ml linear 0.0 -0.5 M NaCl gradient in 50 mM Tris/HCl, pH 7.6. Fractions containing PRPP synthase activity were combined, concentrated by (NH 4 ) 2 SO 4 precipitation, dialyzed against 50 mM Tris/HCl, pH 7.6 containing 50% glycerol, and stored at Ϫ20°C. The purity was evaluated by SDS-PAGE and Coomassie Brilliant Blue staining (15). Protein content was determined by the bicinchoninic acid procedure with chemicals provided by Pierce (16). Bovine serum albumin was used as the standard. Amino acid sequencing by automated Edman degradation was performed by the Department of Biochemistry and Nutrition, Technical University of Denmark.
Kinetic Analysis-Results of initial velocity determinations and of product inhibition studies, both of which were the average of at least three determinations, were fitted to the following equations using the program UltraFit, (Biosoft, version 3.01). Equation 1 is the Michaelis-Menten equation for hyperbolic substrate saturation kinetics, whereas Equation 2 is the rate equation for a sequential Bi Bi mechanism. For competitive inhibition, noncompetitive inhibition, and substrate inhibition the initial velocities were fitted to Equations 3, 4, and 5, respectively (19,20).
where v is the initial velocity, V app is the apparent maximal velocity, K m is the apparent MichaelisϪMenten constant for the varied substrate S, V max is the maximal velocity, K ATP and K Rib-5-P are the MichaelisϪMenten constants for ATP and Rib-5-P, respectively. K iATP is the dissociation constant for ATP, K is and K ii are inhibitor constants for the inhibitor I obtained from the effect on slopes and intercept, respectively, and K i is the inhibitor constant for the substrate S (19,20). Calculation of the free Mg 2ϩ concentration was performed as previously described (21).

Synthesis and Purification of Spinach PRPP Synthase Isozyme 3-
The DNA encoding the 87 codons specifying the presumed transit peptide of spinach PRPP synthase isozyme 3 were removed and replaced by an ATG triplet as described in "Experimental Procedures." Nucleotide sequencing of pBK862 as well as amino acid sequencing of the purified recombinant PRPP synthase demonstrated its N-terminal amino acid sequence to be Met-Asn-Ser-Val-Glu-Phe. This was identical to the sequence of the presumed mature polypeptide with an additional methionine, which served as the translation initiation residue. Induction of expression of the resulting PRS3 allele, harbored in pBK862, resulted in overproduction of the enzyme. PRPP synthase was purified by (NH 4 ) 2 SO 4 and polyethylene glycol precipitations, followed by affinity chromatography and anion exchange chromatography. The yield was 7.2 mg of 95% pure enzyme with a specific activity of 13.1 mol (min ϫ mg of protein) Ϫ1 .
Characteristics of Spinach PRPP Synthase Isozyme 3-The enzyme required the presence of Mg 2ϩ for activity. When ATP and Mg 2ϩ were added at equimolar concentrations, the saturation curve for ATP appeared sigmoid. When Mg 2ϩ was added at 2 mM excess of ATP an increase in enzyme activity was observed compared with the activity with equimolar concentrations of ATP and Mg 2ϩ , and the saturation curve for ATP appeared hyperbolic with a decrease in the K m for ATP. These results indicate that Mg 2ϩ was required as a free cation to activate the enzyme and that the true substrate for the reaction was MgATP. The enzyme accepted Mn 2ϩ as an alternative cation with an activity of 15% of that obtained with Mg 2ϩ . The enzyme was inactive with the following divalent cations: Ca 2ϩ , Cd 2ϩ , and Co 2ϩ . The activity of the enzyme was independent of P i , and in addition no activation by P i was observed as the activity of the enzyme remained at 11 mol (min ϫ mg of protein) Ϫ1 at P i concentrations between 0 and 80 mM. The effect of pH on the activity was determined in the range 6.5-9.0 with Tris/HCl as buffer. The enzyme showed maximal activity at pH 7.6. The activity increased steeply up to pH 7.2 and decreased with a shallower slope above pH 7.6. The shape of the curve for the dependence of temperature on the activity was symmetrical with maximal activity at 40°C. As a result of these analyses we chose to assay the activity of the enzyme at 37°C, at pH 7.6 and at a free Mg 2ϩ concentration of at least 1.2 mM.
Kinetic Mechanism-Initial velocity versus the concentration of either ATP or Rib-5-P was found to follow typical Michaelis-Menten kinetics. To determine whether the reaction occurred by a sequential or a ping-pong mechanism, the effect of the ATP concentration on the initial velocity was measured at different concentrations of Rib-5-P and vice versa. In double reciprocal plots of the data, intersecting lines indicated that the reaction followed a sequential mechanism (Fig. 1). The data were fitted to Equation 2 and resulted in a K ATP value of 170 M, a K Rib-5-P value of 110 M, and a V max value of 13.1 mol (min ϫ mg of protein) Ϫ1 .
The binding order was analyzed by product inhibition studies. Rib-5-P or ATP was varied at different fixed concentrations of the products PRPP or AMP. Inhibition by PRPP was competitive with respect to ATP ( Fig. 2A) and noncompetitive with respect to Rib-5-P (Fig. 2B). Inhibition by AMP was noncompetitive with respect to both ATP and Rib-5-P (Fig. 2, C and D). The calculated inhibition constants are given in Table I. The data agree with a steady-state ordered Bi Bi mechanism in which ATP binds to the enzyme first, followed by Rib-5-P and a leaving order of products where AMP release is followed by release of PRPP.
Alternate Diphosphoryl Donors-Spinach PRPP synthase isozyme 3 readily accepted nucleoside triphosphates other than ATP as diphosphoryl donors. Saturation with dATP, CTP, or UTP followed Michaelis-Menten kinetics, and data were fitted to Equation 1. With GTP, substrate inhibition was observed   (Fig. 3). The calculated kinetic constants, K m , V app , and K i for GTP, are shown in Table II. The K m values for dATP, CTP, and UTP varied little from K ATP , whereas the K m value for GTP was 3-to 4-fold higher. The V app values for CTP and UTP were 10% of V max (Table II). As a result dATP and ATP were essentially equally effective as diphosphoryl donors, whereas GTP, CTP, and UTP were less efficient compared with ATP.
In contrast to the diphosphoryl donor, the enzyme was strictly specific for the ribose moiety. Thus, activity of the enzyme with deoxyribose 5-phosphate as substrate was undetectable.
Inhibition by Ribonucleotides-ADP and in some cases GDP, are potent inhibitors of PRPP synthases from a variety of organisms. Both of these ribonucleotides also inhibited spinach PRPP synthase isozyme 3. However, both ADP ( Fig. 4A and Table I) and GDP (data not shown) were linear competitive inhibitors of ATP saturation. Furthermore inhibition by ADP was noncompetitive with respect to Rib-5-P (Fig. 4B and Table  I). The inhibitory effect of GTP on PRPP synthase activity was analyzed further (Fig. 5). The results are interpreted as    simple competitive inhibition of binding of ATP by GTP. The K iGTP value at a given ATP concentration can be estimated as [ATP] ϫ K mGTP /K ATP . At 1 mM ATP this value is 3.8 mM, whereas at 3 mM ATP the value is 11.5 mM. Both values are consistent with those extrapolated from the data of Fig. 5. Thus, ATP relieved the inhibition by GTP. This indicates that the cause of substrate inhibition by GTP most likely resulted from the formation of a catalytically noncompetent complex of two or more GTP molecules bound in the active site.
Neither tryptophan nor histidine (up to 8 mM, each), which are also products of the PRPP consuming pathways, inhibited the activity of spinach PRPP synthase isozyme 3. DISCUSSION We here show that spinach PRPP synthase isozyme 3 had properties very different from other well characterized PRPP synthases. These properties were independence of P i for activity and stability, lack of allosteric inhibition, and relaxed specificity for diphosphoryl donor. The classical PRPP synthases share common properties, among others, of being dependent on P i for maximal activity and being allosterically inhibited by ADP. In fact evidence has accumulated, which indicate that P i and ADP both bind to the same allosteric site. This evidence comes from results of characterization of six point mutations of human PRPP synthase 1. These mutant variants are associated with gouty arthritis because of overproduction of purine compounds. The mutant forms are less sensitive or insensitive to ADP inhibition and simultaneously are activated by lower concentrations of P i than the normal enzyme (22). In addition the crystal structure of B. subtilis PRPP synthase has been obtained either in complex with ␣, ␤-methylene ADP, an analog of ADP, or with SO 4 2Ϫ , an analog of P i . In these two complexes the ␤-phosphorus of ␣, ␤-methylene ADP and SO 4 2Ϫ occupy the same position (23). Finally, analysis of ADP inhibition of P i activation of E. coli PRPP synthase suggests that the two compounds bind to the same site (24). The present work demonstrates that the activity of spinach PRPP synthase isozyme 3 was independent of P i and that the enzyme was inhibited by ADP only in a manner competitive with ATP. This is in contrast to the complex ADP inhibition found for the Salmonella typhimurium PRPP synthase (25). Together these observations suggest that spinach PRPP synthase isozyme 3 lacks a functional allosteric site for ribonucleoside diphosphates. GTP also inhibited spinach PRPP synthase isozyme 3. This effect appeared to be caused by binding of two molecules of GTP to the active site. Similar observations have been done with various nucleotides inhibiting PRPP synthase from S. typhimurium (25). From the data in Fig. 5, we conclude that there seems to be no alternative allosteric binding site to which binding of GTP will inhibit the enzyme activity. In addition, the classical PRPP synthases use ATP exclusively as diphosphoryl donor, except for the mammalian enzymes, which may also use dATP (26,27). Contrary to this, spinach PRPP synthase isozyme 3 had a much broader specificity toward diphosphoryl donor by using, in addition to ATP or dATP, GTP or the pyrimidine ribonucleoside triphosphates CTP or UTP.
The amino acid sequence identity of spinach PRPP synthase isozyme 3 with the classical PRPP synthases from E. coli, S. typhimurium, B. subtilis, and mammalian sources is low. Thus, the identity of spinach PRPP synthase isozyme 3 with E. coli and B. subtilis PRPP synthases and human PRPP synthase 1 is 23-25%, whereas the identity of E. coli PRPP synthase with B. subtilis PRPP synthase and human PRPP synthase 1 is 52 and 49%, respectively. Based on amino acid sequence comparison we previously suggested a distinction between Class I (i.e. classical) PRPP synthases and Class II PRPP synthases (6). This low sequence identity of spinach PRPP synthase isozyme 3 with the classical PRPP synthases is reflected in the dramatic differences in enzymatic properties between spinach PRPP synthase isozyme 3 and the classical PRPP synthases reported in the present work. Class II appears to be specific for plants and contains a few other members: spinach PRPP synthase isozyme 4, A. thaliana PRPP synthase isozymes 3 and 4. These PRPP synthases have been shown to be independent of P i for activity (6,7). Six additional partial sequences with high homology to spinach PRPP synthase isozyme 3 have been identified by database search. All of these sequences originate from plant sources (GenBank TM /EBI accession no. AI973393 (maize), AW620825 (soybean), AW696707 (barrel medic), BE041040 (rice), BE341210 (potato), and H75122 (loblolly pine)). Spinach and A. thaliana contain at least two and three additional PRPP synthases, respectively. Enzymatic analysis or amino acid sequence comparison showed that these enzymes belong to Class I (6, 7).
Spinach PRPP synthase isozyme 3 also shares common properties with the classical PRPP synthases. The kinetic mechanism of spinach PRPP synthase isozyme 3 was steady-state ordered Bi Bi with MgATP binding first and PRPP leaving last similar to that shown previously for S. typhimurium and E. coli PRPP synthases (21,28). In addition, spinach PRPP synthase isozyme 3 required free Mg 2ϩ as an activator, similar to what has been shown for the bacterial and human PRPP synthases (8,21,26,29).
Previous reports on plant PRPP synthases describe some properties of an enzyme purified 14-fold from spinach leaves (30) or an enzyme purified from rubber tree latex (31). The latter enzyme apparently is different from all other PRPP synthases in size and specific activity. The molecular mass of the subunit is estimated as 57,000 Da, compared with 35,497 Da of spinach PRPP synthase isozyme 3. The activity of the purified rubber tree latex enzyme is 0.7 mol (min ϫ mg of protein) Ϫ1 , which is much lower than the 13 and 170 mol (min ϫ mg of protein) Ϫ1 determined for the spinach PRPP synthase isozyme 3 and E. coli PRPP synthase (28), respectively. It is very likely that these preparations of PRPP synthases from spinach leaves and from rubber tree latex are composed of more than one subunit species. In contrast, we used a highly purified enzyme specified by an allele of the spinach PRS3 cDNA in the present work. careful reading of the manuscript, and Tonny D. Hansen for excellent technical assistance.