Catalytic residues Lys197 and Arg199 of Bacillus subtilis phosphoribosyl diphosphate synthase Alanine-scanning mutagenesis of the flexible catalytic loop Bjarne Hove-Jensen, Ann-Kristin K. Bentsen and Kenneth W. Harlow* Department of Biological Chemistry, Institute of Molecular Biology and Physiology, University of Copenhagen, Denmark The compound 5-phospho-d-ribosyl a-1-diphosphate (PRibPP) is an important component of the metabo- lism of most organisms. PRibPP is a precursor for the biosynthesis of purine, pyrimidine and pyridine nucleo- tides, as well as of the amino acids tryptophan and his- tidine [1,2]. Microorganisms like Bacillus subtilis and Escherichia coli typically contain 10 enzymes that use PRibPP as a substrate [3]. In addition, methanogenic archaea utilize PRibPP for the biosynthesis of metha- nopterin, a folate analogue involved in C1 metabolism [4], and Methanocaldococcus jannaschii utilizes PRibPP as a precursor of ribose 1,5-bisphosphate and subse- quently, ribulose 1,5-bisphosphate [5]. Finally, myco- bacteria utilize PRibPP for the biosynthesis of polyprenylphosphate-pentoses [6]. PRibPP is synthes- ized by transfer of the b,c-diphosphoryl group of ATP to the C-1 hydroxyl of ribose 5-phosphate (Rib5P), in a reaction which is catalysed by PRibPP synthase (ATP:d-ribose 5-phosphate pyrophosphotransferase, EC 2.7.6.1) [7,8]: Rib5P + ATP fi PRibPP + AMP. PRibPP synthase is encoded by the prs gene [9,10]. Several crystal forms of B. subtilis PRibPP synthase have been obtained and the structure was solved to high resolution. The analysis revealed a two-domain subunit structure, which assembles to form a hexamer. Each domain contains a central five-stranded parallel b-sheet surrounded by a-helices, and thus, the overall folds of the domains resemble those of type I phospho- ribosyltransferases. Initially the flexible catalytic loop, KRRPRPNVAEVM(197–208), which contains several charged residues, remained unresolved, except for Lys197, Arg198 and Met208 [11]. Among the amino Keywords flexible loop; nucleotide metabolism; PRPP Correspondence B. Hove-Jensen, Department of Biological Chemistry, Institute of Molecular Biology and Physiology, University of Copenhagen, DK-1307 Copenhagen K, Denmark Fax: +45 3532 2040 Tel: +45 3532 2027 E-mail: hove@mermaid.molbio.ku.dk Website: http://www.imbf.ku.dk *Present address Novo Nordisk A ⁄ S, Drug Metabolism, Novo Nordisk Park, DK-2760 Ma ˚ løv, Denmark (Received 7 April 2005, revised 12 May 2005, accepted 20 May 2005) doi:10.1111/j.1742-4658.2005.04785.x Eleven of the codons specifying the amino acids of the flexible catalytic loop [KRRPRPNVAEVM(197–208)] of Bacillus subtilis phosphoribosyl diphosphate synthase have been changed individually to specify alanine. The resulting variant enzyme forms, as well as the wildtype enzyme, were produced in an Escherichia coli strain lacking endogenous phosphoribosyl diphosphate synthase activity and purified to near homogeneity. The B. subtilis phosphoribosyl diphosphate synthase mutant variants K197A and R199A were studied in detail. The physical properties of the two enzymes were similar to those of the wildtype enzyme. Kinetic characteriza- tion showed that the V max values of the K197A and R199A mutant enzymes were more than 30 000- and more than 24 000-fold reduced, respectively, compared to the wildtype enzyme. The K m values for ATP and ribose 5-phosphate of the two mutant enzymes were essentially unchanged. V app values of the remaining mutant enzymes were much less affected, ranging from 20 to 100% of the V max value of the wildtype enzyme. The data presented show that Lys197 and Arg199 are important in stabilization of the transition state. Abbreviations AMP, adenosine 5¢-monophosphate; PRibPP, 5-phospho- D-ribosyl a-1-diphosphate; Rib5P, ribose 5-phosphate. FEBS Journal 272 (2005) 3631–3639 ª 2005 FEBS 3631 acids of the loop, Lys197 and Arg199 are highly con- served, whereas the remaining 10 amino acids are only moderately conserved. Crystallization of the enzyme in the presence of the transition state analogue AlF 3 , the substrate Rib5P and the product AMP resulted in a structure with a bend arrangement of AMP-AlF 3 -AMP and with Rib5P attached in a manner that is believed to resemble the transition state with the exception of the addition of an adenosyl group [12]. In this crystal form the flexible catalytic loop is fixed in a closed con- formation that appears to shield the transition state analogue AlF 3 from the solvent. Closure of the loop is stabilized by interaction of Lys197 through the e-amino group and Arg199 through the guanidino group with two of the fluoride atoms, which are analogous to oxy- gen atoms of the b-phosphorus of the substrate ATP. A transient negative charge that may develop on the b- phosphoryl oxygen atoms could be stabilized by Lys197 and Arg199 [12]. Furthermore, analysis of a crystal form with the inhibitor analogue a,b-methylene GDP bound at the allosteric site as well as the substrate Rib5P and the reaction-inert substrate analogue a,b- methylene ATP present revealed a tight interface between two subunits. This interface is primarily formed by reciprocal interaction of hydrophobic amino acids of b-strands located on either side of the flexible catalytic loop. The tightly packed interface prevents the closure of the loop, and thus, prevents the interaction of Lys197 and Arg199 with the phosphate chain of ATP. Release of this tight interaction following release of allosteric inhibitor binding causes a 7 A ˚ displace- ment of the b-strands, which is expected to allow the closure of the loop followed by catalysis. Closure of the loop is furthermore stabilized by hydrogen bonds formed between Asn203 and substrate-bound water molecules [12]. Finally, a role of the flexible catalytic loop residue Arg198 in allosteric regulation was sugges- ted. This arginine residue interacts primarily with Asp196 of the same subunit, but also with Asp186 of a neighbouring subunit, and in doing so assists in main- taining the tightly packed interface that prevents the closure of the flexible catalytic loop [12]. The flexible catalytic loop of PRibPP synthase is topologically and functionally equivalent to a loop, variously designated the flexible loop, the catalytic loop, or loop II of the class I phosphoribosyltransferases. This loop is involved in the catalytic function of these enzymes by closing down on the bound substrates and thereby forming the active site (reviewed in [13,14]). Evidence for the importance of Lys197 in catalysis also comes from results of chemical modification of PRibPP synthase from E. coli with the substrate analogue 2¢,3¢-dialdehyde-ATP. Lysine residues 181, 193 and 230 were identified as possible reactive site residues. Of these, Lys193 is homologous to B. subtilis PRibPP synthase Lys197. It was suggested that this lysine residue might interact with the triphosphate chain of ATP either directly or indirectly through the Mg 2+ ion chelated to the phosphate chain of the MgATP complex or it might form hydrogen bonds with the C-2¢ or the C-3¢ hydroxyl group of ATP [15]. We report here the characterization of B. subtilis PRibPP synthase mutant forms with the individual amino acids of the flexible catalytic loop altered to alanine with emphasis on the two catalytic residues Lys197 and Arg199. Results Complementation of Dprs by mutant alleles specifying alanine substitutions of the flexible catalytic loop The codons of the flexible catalytic loop were altered individually to specify alanine as described in Experi- mental procedures. Each of the plasmids harbouring a prs-allele specifying an alanine variant was transformed to strain HO1088 (Dprs) and the growth of the trans- formants were analysed (Table 1). The results showed Table 1. Complementation of Dprs by various prs mutant alleles. Complementation was performed as described in Experimental pro- cedures. Minimal medium contained glucose as the carbon source, thiamine and the indicated compounds. Tetracycline and ampicillin were added to maintain episome and plasmid, respectively. Growth was recorded after 24 h of incubation at 37 °C. +, growth; –, lack of growth. Plasmid PRibPP synthase Growth in minimal medium with supplements None Guanosine, uridine, histidine, tryptophan Guanosine, uridine, histidine, tryptophan, NAD pAB600 Wildtype + + + pAB700 K197A – – + pAB701 R198A + + + pHO377 R199A – – + pHO378 P200A + + + pHO379 R201A + + + pHO380 P202A + + + pHO392 N203A + + + pHO393 V204A + + + pHO382 E206A + + + pHO383 V207A + + + pHO385 M208A + + + pUHE23-2 None – – + PRibPP synthase flexible catalytic loop B. Hove-Jensen et al. 3632 FEBS Journal 272 (2005) 3631–3639 ª 2005 FEBS that the mutant alleles specifying K197A and R199A were unable to complement Dprs, as the transformants HO1088 (Dprs) ⁄ pAB700 (specifying K197A) and HO1088 (Dprs) ⁄ pHO377 (specifying R199A) grew only with all of the components of the PRibPP-consuming pathways present. This indicates the acquisition of PRibPP synthase with no or very low activity. In con- trast, all of the remaining mutant alleles complemented the Dprs allele as the transformants grew without any of the compounds present. Kinetic analysis of K197A and R199A PRibPP synthases Each mutant variant enzyme was purified to near homogeneity, i.e., more than 98% purity as evaluated by SDS ⁄ PAGE. Their kinetic constants were deter- mined together with those of wildtype PRibPP. Initial velocity vs the concentration of either ATP or Rib5P followed typical Michaelis–Menten kinetics. Kinetic parameters of the forward reaction of wildtype, K197A and R199A PRibPP synthases were obtained by measuring initial reaction rates under conditions where ATP and Rib5P were varied against each other. Double reciprocal plots of initial velocity vs ATP and fixed concentrations of Rib5P and vice versa showed a series of intersecting lines for both the wildtype enzyme, the K197A enzyme and the R199A enzyme. The K-values, obtained by fitting the data to Eqn (1), and assuming the binding of ATP before Rib5P (see below), are presented in Table 2. The two mutant enzymes displayed a large decrease in the maximal velocity. The V max value of the K197A enzyme was more than 30 000-fold reduced compared to that of the wildtype enzyme, whereas that of the R199A enzyme was more than 24 000-fold reduced. In con- trast, K ATP and K Rib5P values of the mutant enzymes were much less altered, if at all, compared to those of the wildtype enzyme. The K i(ATP) value of the K197A mutant was reduced six-fold compared to that of the wildtype enzyme, whereas that of the R199A enzyme was three-fold reduced. Inhibition of K197A, R198A and R199A PRibPP synthases The mode of inhibition by ADP of wildtype PRibPP synthase was analysed. Double reciprocal plots of activity vs ATP concentration at fixed concentration of ADP were used to calculate values of slopes and intercepts. Fitting of the slope and intercept values as parabolic inhibition [16] failed to give a reasonable description of the data. Instead the data were fitted to Eqn (5). Replots of intercepts and slopes revealed parabolic curves which are characteristic for binding of ADP to both an active site and to an independent, allosteric site of PRibPP synthase [17,18]. A similar analysis was performed with the K197A, R199A, and R198A mutant enzymes. The Arg198 residue has been proposed to be involved in allosteric regulation due to its involvement in subunit–subunit interactions [12]. The inhibition constants obtained from this ana- lysis are given in Table 3. Although the inhibition pattern appeared similar for the wildtype, the R198A and the R199A enzyme species, i.e., the inhibition by ADP was nonlinear noncompetitive with respect to ATP, the inhibition of the K197A mutant PRibPP synthase was altered. The K¢ is value of the K197A mutant enzyme was only three-fold higher than that of the wildtype enzyme, but the K¢ ii value of the mutant enzyme was 12-fold higher than that of the wildtype enzyme. The apparent Hill coefficient n indi- cates that ADP binds to nonidentical sites on the Table 2. Kinetic constants of wildtype, K197A and R199A PRibPP synthases. The data were fitted to Eqn (1). Standard errors are those given by the computer program. Wildtype, the concentration of ATP was varied from 0.10 to 2.0 m M in the presence of 0.25– 2.0 m M Rib5P and 10 mM MgCl 2 . K197A, the concentration of ATP was varied from 0.05 to 2.0 m M in the presence of 0.10–5.0 mM Rib5P and 10 mM MgCl 2 . R199A, the concentration of ATP was varied from 0.05 to 1.8 m M in the presence of 0.10–3.2 mM Rib5P and 5.0 m M MgCl 2 . Enzyme V max (lmolÆmin )1 Æ mg protein )1 ) K ATP (lM) K Rib5P (lM) K i(ATP) (lM) Wildtype 108 ± 3.9 191 ± 33 230 ± 43 1089 ± 322 K197A 0.003 ± 0.0002 120 ± 24 217 ± 48 167 ± 99 R199A 0.004 ± 0.0002 57 ± 16 106 ± 27 323 ± 146 Table 3. Kinetic constants of ADP inhibition vs ATP of wildtype and mutant PRibPP synthases. The concentration of Rib5P and MgCl 2 was 5.0 and 6.0 mM, respectively. Results of activity determina- tions were plotted as double reciprocal plots, and slopes and inter- cepts were determined. The data were fitted to Eqn (5). Standard errors are those given by the computer program. Wildtype, the con- centration of ATP was varied from 0.1 to 2.0 m M in the presence of 0–0.3 m M ADP. K197A, the concentration of ATP was varied from 0.1 to 2.0 m M in the presence of 0–1.0 mM ADP. R198A ⁄ 199A, the concentration of ATP was varied from 0.1 to 0.8 m M in the presence of 0–0.75 mM ADP. Enzyme K ¢ is (lM) nK¢ ii (lM) n Wildtype 99 ± 7.0 3.32 ± 0.07 116 ± 3.0 4.47 ± 0.34 K197A 296 ± 34 2.77 ± 0.12 1490 ± 430 1.93 ± 1.39 R198A 56 ± 1.0 2.31 ± 0.02 190 ± 54 4.05 ± 0.9 R199A 224 ± 30 3.91 ± 0.08 186 ± 83 2.87 ± 0.25 B. Hove-Jensen et al. PRibPP synthase flexible catalytic loop FEBS Journal 272 (2005) 3631–3639 ª 2005 FEBS 3633 enzyme, consistent with ADP binding both to the act- ive and the allosteric site. From the data it appears that the cooperativity of ADP binding is reduced for the K197A mutant enzyme. Stability of K197A and R199A PRibPP synthases To see if the mutations had an influence on the qua- ternary structure, the K197A and R199A mutant as well as wildtype PRibPP synthases were subjected to electrophoresis in nondenaturing gels. All three PRibPP synthases migrated as single species, with a slight increase in mobility of the K197A enzyme relat- ive to that of the wildtype enzyme, and with a slight increase in mobility of the R199A enzyme relative to that of the K197A enzyme. This is the result expected from the loss of six positively charged lysine or argin- ine residues per hexamer (data not shown). In addi- tion, the temperature of irreversible denaturation of the three enzymes was analysed by differential scan- ning calorimetry. This analysis revealed that the major transition temperature of the wildtype and the two mutant enzymes was quite similar, as the apparent transition temperature was determined as 62.8 °C for the wildtype enzyme, 61.2 °C for the K197A enzyme, and 62.6 °C for the R199A enzyme (data not shown). Altogether the results of native gel electrophoresis and differential scanning calorimetry show that the pres- ence of alanine rather than lysine at position 197, or alanine rather than arginine at position 199 appeared to have no effect on either structure or stability of the enzyme. Kinetic analysis of nine other mutants of the flexible catalytic loop The remaining mutants were less thoroughly analysed. Values of V app as well as K m for ATP were determined by varying the ATP concentration at a fixed Rib5P concentration (Table 4). In general the V app values were reduced compared to the V max value of the wild- type enzyme. However, the reduction was much less severe than that determined for the K197A or R199A mutant enzymes. The V app values ranged between 20 and 100% of the V max value of the wildtype enzyme. The R201A, N203A, E206A and M208A enzymes had V app values similar to the V max value of the wildtype. The K m values for ATP varied somewhat, but the values were not significantly different from the K ATP value of the wildtype enzyme reported above, with the exception of the R201A, N203A and E206A enzymes, which revealed an approximate 3.0-, 4.5- and 2.5-fold increase, respectively, in apparent K m . Reaction mechanism of B. subtilis PRibPP synthase As mentioned above, the double reciprocal plots of ini- tial velocity vs ATP and fixed concentrations of Rib5P or vice versa showed intersecting lines, which indicated a sequential mechanism. To determine if the binding of the substrates was ordered or random, product inhi- bition was analysed. Measurements were made with both of the products, PRibPP and AMP, varied against different concentrations of ATP at fixed Rib5P concentrations and vice versa. Inhibition by AMP was noncompetitive with respect to Rib5P and competitive with respect to ATP. Inhibition by P RibPP was com- petitive with respect to ATP and noncompetitive with respect to Rib5P. The calculated inhibition constants are given in Table 5. These are the results predicted for an ordered binding of the substrates with ATP binding first and a random release of the products [16]. Discussion The kinetic analysis of the K197A and R199A mutant PRibPP synthases indicates that these amino acids are vitally important for catalysis because more than 30 000- and 24 000-fold reductions in V max values were determined in the absence of significant effects on substrate binding, respectively. The K ATP and K Rib5P values of both mutant enzymes were similar to those of the wildtype PRibPP synthase. These results strongly support the hypothesis that Lys197 and Arg199 are involved in catalysis by interacting with Table 4. Kinetic constants of wildtype and mutant PRibPP syn- thases. The concentration of Rib5P was 5.0 m M. The ATP concen- tration was varied from 0.025 to 0.8 m M. The MgCl 2 concentration exceeded the ATP concentration by at least 2.0 m M. The data were fitted to Eqn (4). Standard errors are those given by the computer program. Enzyme V app (lmolÆmin )1 Æ mg protein )1 ) K m(ATP) (lM) Wildtype 108 ± 3.9 a 191 ± 33 b R198A 52 ± 4.6 190 ± 45 P200A 48 ± 10 130 ± 82 R201A 83 ± 14 560 ± 200 P202A 76 ± 8.0 290 ± 110 N203A 130 ± 20 870 ± 210 V204A 24 ± 2.1 190 ± 36 E206A 110 ± 20 480 ± 150 V207A 32 ± 3.0 160 ± 44 M208A 88 ± 8.0 230 ± 55 a V max value from Table 2. b K ATP value from Table 2. PRibPP synthase flexible catalytic loop B. Hove-Jensen et al. 3634 FEBS Journal 272 (2005) 3631–3639 ª 2005 FEBS the b-phosphorus oxygen atoms of the ATP triphos- phate chain after formation of the enzyme–substrate complex. According to this hypothesis, this interaction results in a closure of the flexible catalytic loop, a requisite for stabilization of the transition state, which causes a large spatial displacement of the two amino acids [12]. The remaining nine mutant variants of the flexible catalytic loop were much less affected and were of two types. The first type, represented by R201A, N203A and E206A, and possibly also P202A and M208A, had V app values similar to the V max of the wildtype, whereas their K m(ATP) values were signifi- cantly increased compared to the K ATP value of the wildtype enzyme. This result seems to indicate an involvement of these residues in the binding of the substrate ATP. The second type, represented by R198A, P200A, V204A, and V207A, had reduced V app values compared to the V max value of the wildtype enzyme, whereas their K m(ATP) values were similar to the K ATP value of the wildtype enzyme. A flexible loop is also found in several phosphoribo- syltransferases [19–21]. The conformation of the flag region, of which the flexible loop is a part, varies con- siderably; the loop is either completely unresolved or in a very flexible state [13,14,22]. Furthermore, the act- ive site of Salmonella enterica serovar Typhimurium orotate phosphoribosyltransferase is rather solvent- exposed in the crystal structure [23]. Analysis of mutant variants of a lysine residue located in this dis- ordered loop resulted in an enzyme that had a 600- to 1000-fold reduction in the k cat value but only a minor increase of the K m values for the substrates, orotate and PRibPP [24]. Characterization of alanine variants of the remaining amino acid residues of the flexible loop revealed an effect on the K m values, whereas the effect on the k cat values was minor [22]. It is possible therefore that Lys197 or Arg199 in the PRibPP syn- thase-catalysed reaction may play a role similar to that of this lysine residue in the orotate phosphoribosyl- transferase-catalysed reaction. ADP is the primary allosteric inhibitor of PRibPP synthase and also acts as a competitive inhibitor of the enzyme from several bacterial species [18,25–27]. ADP competes with P i for binding to the allosteric site of E. coli PRibPP synthase [28]. Consistent with this pat- tern of inhibition the K197A and R199A mutant and the wildtype PRibPP synthases all displayed a noncom- petitive pattern of ADP inhibition when varied against ATP. The secondary inhibition plots of the slopes and intercepts vs the ADP concentration for both enzymes, which yielded parabolic functions, is consistent with a complex inhibition mode including a competitive and a noncompetitive component. However, the K197A mutant enzyme appeared to be less sensitive to ADP, because higher ADP concentration was required to produce the same degree of inhibition as the wildtype enzyme. This was further confirmed by the kinetic con- stants for the K197A mutant enzyme, where K ii for the mutant enzyme was increased 12-fold and K is was less altered. These results indicate that the competitive inhi- bition mechanism is more or less unchanged, whereas the noncompetitive inhibition mechanism is altered. Furthermore, the apparent degree of cooperativity of the K197A mutant enzyme in response to ADP in the allosteric site is decreased, and Lys197 appears to be also involved in the allosteric mechanism, which could be a consequence of an altered structure of the flexible catalytic loop. Thus, the inhibition mode of ADP (vs ATP) for the K197A mutant enzyme may involve an unchanged competitive and an altered noncompetitive mechanism. The K¢ is value for the R199A mutant enzyme was increased similarly to that of the K197A enzyme, whereas the K¢ ii value resembled that of the wildtype enzyme. Consequently the R199A enzyme appeared to be little or not at all altered with respect to inhibition by ADP, although a minor reduction in cooperativity was observed. The R198A enzyme also resembled the wildtype enzyme in properties. Thus, the lack of interaction in the R198A enzyme of Arg198 with the Asp186 and Asp196, revealed by the three- dimensional structure of the wildtype enzyme [12], apparently did not weaken the tight interactions between the subunits. Possibly other amino acids might substitute Arg198, with Arg182 being a candidate for this interaction. Therefore, in the R198A mutant enzyme stabilization might be provided by interaction of Arg182 with Asp186 of the same chain and with Asp196 of another chain. Table 5. Mode of inhibition and inhibition constants of wildtype PRibPP synthase. Inhibition constants were determined as des- cribed in Experimental procedures. The concentration of MgCl 2 was 10 mM. The data were fitted to Eqn (2) or (3). Standard errors are those given by the computer program. Inhibitor Substrate Mode of inhibition K is (lM) K ii (lM) AMP ATP a Competitive 131 ± 35 – Rib5P b Noncompetitive 353 ± 108 2365 ± 709 PRibPP ATP c Competitive 157 ± 50 – Rib5P d Noncompetitive 2877 ± 763 1548 ± 680 a The concentration of ATP was varied from 0.1 to 2.0 mM in the presence of 0–1.0 m M AMP and 5.0 mM Rib5P. b The concentration of Rib5P was varied from 0.1 to 5.0 m M in the presence of 0–1.0 m M AMP and 1.5 mM ATP. c The concentration of ATP was varied from 0.1 to 2.0 m M in the presence of 0–1.0 mM PRibPP and 5.0 m M Rib5P. d The concentration of Rib5P was varied from 0.1 to 5.0 m M in the presence of 0–1.0 mM PRibPP and 1.5 mM ATP. B. Hove-Jensen et al. PRibPP synthase flexible catalytic loop FEBS Journal 272 (2005) 3631–3639 ª 2005 FEBS 3635 The kinetic constants obtained here for the wildtype B. subtilis PRibPP synthase are somewhat lower than those reported earlier, which were 660 lm for K m(ATP) , 480 lm for K m(Rib5P) and 250 lmolÆmin )1 Æmg protein )1 for V app [26]. The kinetic mechanism of B. subtilis PRibPP synthase was shown to be sequential as the products PRibPP and AMP were competitive inhibi- tors of the enzyme with respect to ATP and noncom- petitive inhibitors with respect to Rib5P. These results indicate that binding of the substrates is ordered with initial binding of ATP followed by binding of Rib5P. In contrast, the leaving order of the products appears to be random. A possible sequential mechanism for B. subtilis PRibPP synthase is summarized in Scheme 1. Experimental procedures Site-directed mutagenesis Overlap extension PCR was employed to alter the desired codons using AmpliTaq DNA polymerase (PE Applied Bio- systems, Foster City, CA) and the four deoxyribonucleotides (Amersham Biosciences, Hillerød, Denmark) in a Trio- Thermoblock (Biometra, Go ¨ ttingen, Germany) with stand- ard cycle settings [29]. The template was DNA of pAB600 harbouring a wildtype B. subtilis prs gene [30]. Two specific oligodeoxyribonucleotides were used as primers in PCR to produce a mutant allele. With K197A PRibPP synthase as an example, these oligodeoxyribonucleotides were 5¢-CGAT TATCGAT GCACGCCG (designated K197Ap) and the complementary 5¢-CGGCG TGCATCGATAATCG (desig- nated K197Am) where the underlined codons represent the altered codon. Two additional oligodeoxyribonucleotides were used for all mutations: 5¢-CGTTCTGAACAAATC CAGATGG (designated UHE3) and 5¢-CACACAGAATT CTCTAGAGG (designated UHE51), which anneal outside of the region of mutagenesis. PCR was performed with the oligodeoxyribonucleotides UHE51 and K197Am and with K197Ap and UHE3. The resulting two DNA fragments were purified (Qiagen, Hilden, Germany) following agarose gel electrophoresis and were used as template in a second round of PCR with the oligodeoxyribonucleotides UHE51 and UHE3 as primers. The DNA fragment resulting from this second round of PCR was digested by restriction endo- nucleases EcoRV and NsiI and ligated by T4 DNA ligase (Amersham Biosciences) to similarly digested DNA of pAB600. The ligated DNA was transformed to strain HO1088 [araC am araD D(lac)U169 trp am mal am rpsL relA thi supF deoD gsk-3 udp Dprs-4] [31] and the resulting plasmid, pAB700, was isolated and the insert sequenced and com- pared to the nucleotide sequence of the wildtype prs gene [32], which confirmed introduction of the mutation. DNA sequencing was performed using a Thermo Sequenase dye terminator cycle premix kit (Amersham Biosciences), fol- lowed by analysis in an Abi Prism 377 DNA Sequencer (PE Applied Biosystems). A similar procedure was used for each mutation. The specific primers were 5¢-GCGATTATCGA TAAA GCGCGTCCGCGTCC and 5¢-GGACGCGGACG CGCTTTATCGATAATCGC for R198A, which resulted in pAB701, 5¢-CGATAAACGC GCGCCGCGTCC and 5¢- GGACGCGG CGCGCGTTTATCG for R199A (pHO377), 5¢-CGCCGT GCGCGTCCAAACGTGG and 5¢-CCACGT TTGGACG CGCACGGCG for P200A (pHO378), 5¢-CGC CGTCCG GCACCAAACGTGG and 5¢-CCACGTTTGG TGCCGGACGGCG for R201A (pHO379), 5¢-CGTCCG CGT GCAAACGTGG and 5¢-CCACGTTTGCACGCGG ACG for P202A (pHO380), 5¢-CGCCGTCCGCGTCCA GCTGTGGCGGAAGTCATGAATATTGTAGGTAACATC GAAGGG and 5¢-CCCTTCGATGTTACCTACAATATTC ATGACTTCCGCCAC AGCTGGACGCGGACGGCG for N203A (pHO392), 5¢-CGCCGTCCGCGTCCAAAC GCCG CGGAAGTCATGAATATTGTAGGTAACATCGAAGGG and 5¢-CCCTTCGATGTTACCTACAATATTCATGACT TCCGC GGCGTTTGGACGCGGACGGCG for V204A (pHO393), 5¢-CGTGGCG GCGGTCATGAATATTGTAGG and 5¢-CCTACAATATTCATGAC CGCCGCCACG for E206A (pHO382), 5¢-CGTGGCGGAA GCGAGTAGG and 5¢-CCTACAATATTCAT CGCTTCCGCCACG for V207A (pHO383), and 5¢-CGTGGCGGAAGTC GCGAATATTG TAGG and 5¢-CCTACAATATT CGCGACTTCCGCCACG for M208A (pHO385). Oligodeoxyribonucleotides were pur- chased from Hobolt DNA Syntese (Hillerød, Denmark). Complementation E. coli strain HO1088 harbours a deletion of the chromoso- mal prs gene and therefore requires compounds that may be converted to the products of the PRibPP-consuming pathways, guanosine, uridine, histidine, tryptophan and NAD. The acquisition in strain HO1088 of a prs gene specifying active PRibPP synthase relieves one or more of these requirements. Complementation was analysed by pla- ting transformed cells of strain HO1088 in AB minimal medium [33] with glucose (0.2%, w ⁄ v) as the carbon source. Supplements were used at the following concentrations: E E E ATP [E Rib5P ATP E] PRibPP AMP E PRibPP E AMP Scheme 1. Possible reaction mechanism of B. subtilis PRibPP syn- thase with ordered sequential binding of substrates and random release of products. ‘E’ represents the enzyme. PRibPP synthase flexible catalytic loop B. Hove-Jensen et al. 3636 FEBS Journal 272 (2005) 3631–3639 ª 2005 FEBS guanosine, 30 mgÆL )1 ; uridine, 20 mgÆL )1 ; histidine, 40 mgÆL )1 ; tryptophan, 40 mgÆL )1 ; NAD, 40 mgÆL )1 ; thi- amine, 1.0 mgÆL )1 . Tetracycline and ampicillin were added to 10 and 100 mgÆL )1 , respectively. Gene expression and enzyme purification Procedures for expression of the B. subtilis prs gene as well as purification of PRibPP synthase have been described previously [30]. E. coli strain HO1088 harbouring the var- ious plasmids was grown at 37 °C in NZY medium supple- mented with ampicillin and tetracycline [34]. Cultures of cells harbouring pAB700 or pHO377 also contained NAD. At an attenuance at 436 nm of 1–2, prs gene expression was induced by the addition of isopropyl thio-b-d-galacto- side (50 lm) and cultures were incubated overnight with shaking. An attenuance at 436 nm of 1 (1-cm path length) corresponds to approximately 3 · 10 11 cellsÆL )1 . Cells were harvested and stored frozen at )20 °C. Purification of the mutant variants was performed essentially as previously described [30], except that the final anion-exchange chroma- tographic step was modified as follows: the protein solution in 50 mm Na 2 HPO 4 ⁄ NaH 2 PO 4 , pH 7.5 was applied to a 20 mL anion exchange Hiload Q-Sepharose column (Amer- sham Biosciences), previously equilibrated with the same buffer, at a rate of 1.0 mLÆmin )1 and washed with five vol- umes of the same buffer. PRibPP synthase was eluted by applying a salt gradient of 0–100% Salt Buffer (1.0 m NaCl in 50 mm Na 2 HPO 4 ⁄ NaH 2 PO 4 , pH 7.5) at a rate of 2.0 mLÆmin )1 over 60 min. The gradient was an initial lin- ear increase from 0 to 20% Salt Buffer, followed by a hold for two column volumes. This was followed by an increase to 35% Salt Buffer over approximately six column volumes. The salt concentration was then raised to 100% Salt Buffer. The enzyme eluted at an NaCl concentration of approxi- mately 0.30 m. Fractions were collected across the peak of PRibPP synthase activity and analysed by SDS ⁄ PAGE. The fractions with highest purity were pooled and dialysed against multiple changes of 50 mm Na 2 HPO 4 ⁄ NaH 2 PO 4 , pH 7.5. The enzyme was stored refrigerated at protein con- centrations of 5–10 gÆL )1 . Analysis of physical properties of PRib PP synthases Hydrodynamic properties of K197A and R199A mutant PRibPP synthases were analysed by nondenaturing gel elec- trophoresis in 7.5% (w ⁄ v) polyacrylamide gels (pH 8.8) pre- pared as described previously with the omission of SDS [35]. Samples containing 5–10 lgofPRibPP synthase in 50 mm Na 2 HPO 4 ⁄ NaH 2 PO 4 , pH 7.5, were loaded in the gel. The loading dye contained 50 mm Tris ⁄ HCl, pH 6.8, 10% (v ⁄ v) glycerol and 0.025% (v ⁄ v) bromophenol blue. Electrophoresis was performed at 60 V at room tempera- ture. The stability of PRibPP synthase was analysed by differential scanning calorimetry (Microcal, Northampton, MA) with cell volumes of 1.16 mL. The temperature range was 15–90 °C, the scan rate was 1 °C per min. Before meas- urement a protein sample (at a concentration of  1gÆL )1 ) was dialysed against 50 mm Na 2 HPO 4 ⁄ NaH 2 PO 4 , pH 7.5 and degassed by stirring in an evacuated chamber for 5–10 min at room temperature and then immediately loa- ded into the calorimeter. Buffer (50 mm Na 2 HPO 4 ⁄ NaH 2 PO 4 , pH 7.5) was loaded in the reference cell. A pres- sure of two bars of nitrogen gas was maintained over the liquids in the cells throughout the scans to prevent degas- sing during heating. The program origin 3.1 (Microcal) was used to evaluate the scans. Assay of PRibPP synthase activity PRibPP synthase activity was assayed at 37 °C as previously described [36]. The amount of radioactivity in ATP and PRibPP spots on thin-layer chromatograms was quantitated with an Instant Imager model 2024 or with the Cyclone Storage Phosphor System (PerkinElmer, Wellesley, MA). The assay components were 50 mm Tris ⁄ HCl pH 8.0 (adjus- ted at 37 °C), 50 mm NaH 2 PO 4 ⁄ Na 2 HPO 4 , 2.0 mm EGTA and MgCl 2 , which exceeded the nucleotide concentration by at least 2.0 mm. The ATP and Rib5P concentrations were varied appropriately. Standard concentrations were 3.0 mm for ATP and 5.0 mm for Rib5P. Reaction was initiated by the addition of Rib5P or enzyme, which was appropriately diluted in 50 mm Tris ⁄ HCl pH 8.0 (adjusted at 37 °C), 50 mm NaH 2 PO 4 ⁄ Na 2 HPO 4 , 2.0 mm EGTA and 1.0 gÆL )1 BSA, to a reaction volume of 100 lL. Protein content was determined by the bicinchoninic acid procedure (Pierce Chemical Company Rockford, IL) with BSA as the standard. Enzyme activity is expressed as lmolÆmin )1 Æmg protein )1 . Kinetic analysis In kinetic experiments the concentration of substrates was varied as indicated in Results. Initial velocities are averages of at least duplicate determinations. For each experiment determination of initial velocity parameters of the wildtype and mutant enzymes were analysed by fitting the data to a sequential mechanism with ATP binding as the first sub- strate, Eqn (1) by use of the programs hyper [37] or ultrafit (version 3.0; Biosoft, Cambridge, UK): v ¼ V max ½ATP½Rib5P =ðK iðATPÞ K Rib5P þ K ATP ½Rib5P þ K Rib5P ½ATPþ½ATP½Rib5PÞ ð1Þ where v is the initial velocity, V max is the maximal velocity, K i(ATP) is the dissociation constant for the substrate ATP, and K ATP and K Rib5P are the Michaelis–Menten constants for the substrates ATP and Rib5P, respectively. For noncompetitive inhibition or competitive inhibition the initial velocities were fitted to Eqn (2) or (3), respectively. B. Hove-Jensen et al. PRibPP synthase flexible catalytic loop FEBS Journal 272 (2005) 3631–3639 ª 2005 FEBS 3637 Equation (4) is the Michaelis–Menten equation for hyper- bolic substrate saturation kinetics: v ¼ V app ½S=fK m ð1 þ½I=K is Þþ½Sð1 þ½I=K ii Þg ð2Þ v ¼ V app ½S=fK m ð1 þ½I=K is Þþ½Sg ð3Þ v ¼ V app ½S=ðK m þ½SÞ ð4Þ where V app is the apparent maximal velocity and K m is the apparent Michaelis–Menten constant for the varied sub- strate S. K ii and K is are the inhibitor constants for the inhibitor I obtained from the effect on intercepts and slopes, respectively. When ATP was varied against different concentrations of ADP, a secondary plot was constructed with slopes and intercepts from straight lines obtained from double reciprocal plots. The reciprocal initial velocities were weighted assuming relative errors [16]. The slopes and inter- cepts vs ADP concentration were fitted without weighting to Eqn (5): Slope ¼ Qf1 þð½I=K 0 is Þ n gð5Þ where Q is the intercept (here representing K m ⁄ V app ), K¢ is is the concentration of inhibitor that causes a doubling of the slope value and n is the apparent Hill coefficient [17]. The intercepts vs ADP concentration were fitted without weight- ing to Eqn (5), replacing K¢ is with K¢ ii and Q representing 1 ⁄ V app . In the experiments above, data were analysed with ultrafit. Acknowledgements We are grateful to Robert L. Switzer and Martin Willemoe ¨ s for careful reading of the manuscript and to Bent W. Sigurskjold for performing the scanning dif- ferential calorimetry. We thank Martin Willemoe ¨ s for invaluable discussions and Tonny D. Hansen for excel- lent technical assistance. This work was supported by the Danish Natural Science Research Council. References 1 Hove-Jensen B (1988) Mutation in the phosphoribosyl- pyrophosphate synthetase gene (prs) that results in sim- ultaneous requirements for purine and pyrimidine nucleosides, nicotinamide nucleotide, histidine, and tryp- tophan in Escherichia coli. J Bacteriol 170, 1148–1152. 2 Hove-Jensen B (1989) Phosphoribosylpyrophosphate (PRPP)-less mutants of Escherichia coli. Mol Microbiol 3, 1487–1492. 3 Jensen KF (1983) Metabolism of 5-phosphoribosyl 1-pyrophosphate (PRPP) in Escherichia coli and Salmonella typhimurium.InMetabolism of Nucleotides, Nucleosides and Nucleobases (Munch-Petersen A, ed.), pp. 1–25. Academic Press, London. 4 White RH (1996) Biosynthesis of methanopterin. Biochemistry 35, 3447–3456. 5 Finn MW & Tabita FR (2004) Modified pathway to synthesize ribulose 1,5-bisphosphate in methanogenic archaea. J Bacteriol 186, 6360–6366. 6 Scherman MS, Kalbe-Bournonville L, Bush D, Xin Y, Deng L & McNeil M (1996) Polyprenylphosphate-pen- toses in mycobacteria are synthesized from 5-phospho- ribose pyrophosphate. J Biol Chem 271, 29652–29658. 7 Kornberg A, Lieberman I & Simms ES (1955) Enzy- matic synthesis and properties of 5-phosphoribosylpyro- phosphate. J Biol Chem 215, 389–402. 8 Khorana HG, Fernandes JF & Kornberg A (1958) Pyrophosphorylation of ribose 5-phosphate in the enzy- matic synthesis of 5-phosphorylribose 1-pyrophosphate. J Biol Chem 230, 941–948. 9 Hove-Jensen B & Nygaard P (1982) Phosphoribosyl- pyrophosphate synthetase of Escherichia coli. Identifica- tion of a mutant enzyme. Eur J Biochem 126, 327–333. 10 Nilsson D & Hove-Jensen B (1987) Phosphoribosylpyro- phosphate synthetase of Bacillus subtilis. Cloning, char- acterization and chromosomal mapping of the prs gene. Gene 53, 247–255. 11 Eriksen TA, Kadziola A, Bentsen AK, Harlow KW & Larsen S (2000) Structural basis for the function of Bacillus subtilis phosphoribosyl-pyrophosphate synthe- tase. Nat Struct Biol 7, 303–307. 12 Nygaard FB (2001) The molecular mechanism of cata- lysis and allosteric regulation in the phosphoribosyl- diphosphate synthase from Bacillus subtilis. Thesis, University of Copenhagen, Denmark. 13 Smith JL (1999) Forming and inhibiting PRT active sites. Nat Struct Biol 6, 502–504. 14 Sinha SC & Smith JL (2001) The PRT protein family. Curr Opin Struct Biol 11, 733–739. 15 Hilden I, Hove-Jensen B & Harlow KW (1995) Inacti- vation of Escherichia coli phosphoribosylpyrophosphate synthetase by the 2¢,3¢-dialdehyde derivative of ATP. Identification of active site lysines. J Biol Chem 270, 20730–20736. 16 Cleland WW (1963) The kinetics of enzyme-catalyzed reactions with two or more substrates or products. II. Inhibition: Nomenclature and theory. Biochim Bio- phys Acta 67, 173–187. 17 Willemoe ¨ s M & Hove-Jensen B (1997) Binding of diva- lent magnesium by Escherichia coli phosphoribosyl diphosphate synthetase. Biochemistry 36, 5078–5083. 18 Switzer RL & Sogin DC (1973) Regulation and mechan- ism of phosphoribosylpyrophosphate synthetase. V. Inhibition by end products and regulation by adeno- sine diphosphate. J Biol Chem 248, 1063–1073. 19 Eads JC, Scapin G, Xu Y, Grubmeyer C & Sacchettini JC (1994) The crystal structure of human hypoxanthine- guanine phosphoribosyltransferase with bound GMP. Cell 78, 325–334. PRibPP synthase flexible catalytic loop B. Hove-Jensen et al. 3638 FEBS Journal 272 (2005) 3631–3639 ª 2005 FEBS 20 Scapin G, Grubmeyer C & Sacchettini JC (1994) Crystal structure of orotate phosphoribosyltransferase. Biochem- istry 33, 1287–1294. 21 Henriksen A, Aghajari N, Jensen KF & Gajhede M (1996) A flexible loop at the dimer interface is a part of the active site of the adjacent monomer of Escherichia coli orotate phosphoribosyltransferase. Biochemistry 35, 3803–3809. 22 Schramm VL & Grubmeyer C (2004) Phosphoribosyl- transferase mechanism and roles in nucleic acid metabolism. Prog Nucleic Acid Res Mol Biol 75, 261–304. 23 Ozturk DH, Dorfman RH, Scapin G, Sacchettini JC & Grubmeyer C (1995) Structure and function of Salmo- nella typhimurium orotate phosphoribosyltransferase: Protein complementation reveals shared active sites. Biochemistry 34, 10755–10763. 24 Scapin G, Ozturk DH, Grubmeyer C & Sacchettini JC (1995) The crystal structure of the orotate phosphoribo- syltransferase complexed with orotate and a-d-5-phos- phoribosyl-1-pyrophosphate. Biochemistry 34, 10744– 10754. 25 Hove-Jensen B, Harlow KW, King CJ & Switzer RL (1986) Phosphoribosylpyrophosphate synthetase of Escherichia coli. Properties of the purified enzyme and primary structure of the prs gene. J Biol Chem 261, 6765–6771. 26 Arnvig K, Hove-Jensen B & Switzer RL (1990) Purifica- tion and properties of phosphoribosyl-diphosphate synthetase from Bacillus subtilis. Eur J Biochem 192, 195–200. 27 Hove-Jensen B & McGuire JN (2004) Surface exposed amino acid differences between mesophilic and thermo- philic phosphoribosyl diphosphate synthase. Eur J Biochem 271, 4526–4533. 28 Willemoe ¨ s M, Hove-Jensen B & Larsen S (2000) Steady state kinetic model for the binding of substrates and allosteric effectors to Escherichia coli phosphoribosyl- diphosphate synthase. J Biol Chem 275, 35408–35412. 29 Ho SN, Hunt HD, Horton RM, Pullen JK & Pease LR (1989) Site-directed mutagenesis by overlap extension using the polymerase chain reaction. Gene 77, 51–59. 30 Bentsen AK, Larsen TA, Kadziola A, Larsen S & Harlow KW (1996) Overexpression of Bacillus subtilis phosphoribosylpyrophosphate synthetase and crystalli- zation and preliminary x-ray characterization of the free enzyme and its substrate-effector complex. Proteins 24, 238–246. 31 Krath BN & Hove-Jensen B (2001) Class II recombi- nant phosphoribosyl diphosphate synthase from spi- nach. Phosphate independence and diphosphoryl donor specificity. J Biol Chem 276, 17851–17856. 32 Nilsson D, Hove-Jensen B & Arnvig K (1989) Primary structure of the tms and prs genes of Bacillus subtilis. Mol Gen Genet 218, 565–571. 33 Clark DJ & Maaløe O (1967) DNA replication and the division cycle in Escherichia coli. J Mol Biol 23, 99–112. 34 Hove-Jensen B & Maigaard M (1993) Escherichia coli rpiA gene encoding ribose phosphate isomerase A. J Bacteriol 175, 6852–6865. 35 Laemmli UK (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680–685. 36 Jensen KF, Houlberg U & Nygaard P (1979) Thin-layer chromatographic methods to isolate 32 P-labeled 5-phos- phoribosyl-a-1-pyrophosphate (PRPP): Determination of cellular PRPP pools and assay of PRPP synthetase activity. Anal Biochem 98, 254–263. 37 Cleland WW (1979) Statistical analysis of enzyme kin- etic data. Methods Enzymol 63 , 103–138. B. Hove-Jensen et al. PRibPP synthase flexible catalytic loop FEBS Journal 272 (2005) 3631–3639 ª 2005 FEBS 3639 . Catalytic residues Lys197 and Arg199 of Bacillus subtilis phosphoribosyl diphosphate synthase Alanine-scanning mutagenesis of the flexible catalytic loop Bjarne. designated the flexible loop, the catalytic loop, or loop II of the class I phosphoribosyltransferases. This loop is involved in the catalytic function of these