Mycoplasma pneumoniae HPr kinase/phosphorylase Assigning functional roles to the P-loop and the HPr kinase/phosphorylase signature sequence motif Matthias Merzbacher, Christian Detsch, Wolfgang Hillen and Jo¨rg Stu¨ lke* Lehrstuhl fu ¨ r Mikrobiologie, Institut fu ¨ r Mikrobiologie, Biochemie und Genetik der Friedrich-Alexander-Universita ¨ t Erlangen-Nu ¨ rnberg, Germany HPr kinase/phosphorylase (HPrK/P) is the key regulator of carbon metabolism in many Gram-positive bacteria. It phosphorylates/dephosphorylates the HPr protein of the bacterial phosphotransferase system on a regulatory serine residue in response to the nutrient status of the cell. In Mycoplasma pneumoniae, HPrK/P is one of the very few regulatory proteins encoded in the genome. The regulation of this enzyme by metabolites is unique among HPrK/P proteins studied so far: it is active as a kinase at low ATP concentrations, whereas the proteins from other bacteria need high ATP concentrations as an indicator of a good nutrient supply for kinase activity. We studied the inter- action of M. pneumoniae HPrK/P with ATP, Fru1,6P 2 and P i by fluorescence spectroscopy. In agreement with the pre- viously observed unique regulation, we found a very high affinity for ATP (K d ¼ 5.4 l M ) compared with the HPrK/P proteins from other bacteria. The K d for Fru1,6P 2 was three orders of magnitude higher, which explains why Fru1,6P 2 has only a weak regulatory effect on M. pneumoniae HPrK/ P. Mutations of two important regions in the active site of HPrK/P, the nucleotide binding P-loop and the HPrK/P family signature sequence, had different effects. P-loop region mutations strongly affect ATP binding and thus all enzymatic functions, whereas the signature sequence motif seems to be important for the catalytic mechanism rather than for nucleotide binding. Keywords: Gram-positive bacteria; HPr kinase/phosphory- lase; Mycoplasma pneumoniae; nutrients; regulation. All organisms need to compete for scarce resources of nutrients and energy. Therefore, it is essential to have not only highly efficient metabolic pathways but also sophis- ticated regulatory systems that allow rapid adaptation to changing environmental conditions. In bacteria, the ability to live in many different ecosystems is directly related to the genetic equipment with regulatory systems. Bacteria that are metabolically versatile and able to live in a wide variety of different habitats such as Pseudomonas aerugi- nosa reserve about 10% of their genomes for regulatory genes [1]. At the other extreme, mycoplasmas, which depend on nutrient-rich habitats, contain the smallest genomes of all self-replicating organisms known so far and encode very few regulatory proteins [2]. The genome of Mycoplasma pneumoniae encodes only nine regulatory proteins, among them no alternative sigma factors and no two-component system. However, M. pneumoniae and the other mycoplasmas studied so far possess one of the key regulatory proteins of carbon metabolism in Gram- positive bacteria, the HPr kinase/phosphorylase (HPrK/P) [3–6]. HPrK/P is a metabolite-sensitive enzyme that phos- phorylates/dephosphorylates the HPr protein of the bacterial phosphoenolpyruvate–sugar phosphotransferase system (PTS) on a serine residue [7–9]. It is present in many but not all Gram-positive and Gram-negative bacteria. However, it is absent from enteric bacteria such as Escherichia coli and their close relatives [7,8]. Phos- phorylation of HPr by HPrK/P has different conse- quences. (a) The protein can no longer take part in the phosphotransfer reactions of the PTS because of a greatly reduced affinity for the PTS phosphoryl donor, enzyme I; this leads to inhibition of PTS-dependent sugar transport activity [10,11]. (b) In Gram-positive bacteria with a low GC content such as Bacillus subtilis, HPr(Ser-P) is a cofactor of the pleiotropic transcriptional regulator CcpA which controls the expression of about 100 genes involved in carbon metabolism [12,13]. (c) HPr(HisP) is required to phosphorylate and thereby stimulate the activity of several transcriptional regulators, enzymes and transport- ers. This stimulation cannot occur if HPr becomes phosphorylated by HPrK/P [14]. Although the function of HPrK/P is well established in B. subtilis and its relatives, much less is known about its role in mycoplasmas and Correspondence to J. Stu ¨ lke, Abteilung fu ¨ r Allgemeine Mikrobiologie, Institut fu ¨ r Mikrobiologie und Genetik, Georg-August – Universita ¨ t Go ¨ ttingen, Grisebachstr. 8, D-37077 Go ¨ ttingen, Germany. Fax: + 49 551 393808, Tel.: + 49 551 393781, E-mail: jstuelk@gwdg.de Abbreviations: HPrK/P, HPr kinase/phosphorylase; PTS, phos- phoenolpyruvate-dependent sugar phosphotransferase system; HPr, histidine-containing phosphocarrier protein of the PTS. *Present address: Abteilung fu ¨ r Allgemeine Mikrobiologie, Institut fu ¨ r Mikrobiologie und Genetik, Georg-August – Universita ¨ tGo ¨ ttingen, Grisebachstr. 8, D-37077 Go ¨ ttingen, Germany. (Received 26 September 2003, revised 11 November 2003, accepted 20 November 2003) Eur. J. Biochem. 271, 367–374 (2004) Ó FEBS 2003 doi:10.1046/j.1432-1033.2003.03935.x Gram-negative bacteria. Recently, it was proposed that HPrK/P might control the activity of certain two-compo- nent regulatory systems and the alternative sigma factor, r N , in Gram-negative proteobacteria [15]. As neither two- component systems nor alternative sigma factors are present in M. pneumoniae, HPrK/P must fulfil other functions in this bacterium. To determine the role of HPrK/P in M. pneumoniae, we analyzed the biochemical activity, its regulation and the structure of this enzyme and compared it with the HPrK/P from other organisms. All HPrK/Ps have the same basic activity: they phosphorylate/dephosphorylate HPr. How- ever, enzymes from different organisms differ in their regulation: the HPrK/P from B. subtilis and related bacteria is dependent on high ATP concentrations for activity, and this is stimulated in a co-operative manner by Fru1,6P 2 . The phosphorylase activity is stimulated by P i [16]. Thus, the kinase activity may prevail under conditions of good nutrient supply, whereas the phosphorylase activity is dominant if carbon and energy sources become limiting. In contrast, HPrK/P from M. pneumoniae already exhibits kinase activity at low ATP concentrations and is not stimulated by Fru1,6P 2 . As observed in other bacteria, P i triggers phosphorylase activity, and Fru1,6P 2 antagon- izes the action of P i [4]. This unique mode of control of HPrK/P activity was proposed to reflect the lifestyle of M. pneumoniae.WhereasM. pneumoniae is strictly adapted to its ecological niche on nutrient-rich human mucous membranes, B. subtilis typically faces nutrient limitations in its natural environments. Thus, the default state of HPrK/P indicates a good (M. pneumoniae, kinase activity) or a poor (B. subtilis, phosphorylase activity) nutrient supply. Recently, the HPrK/P proteins of Lactobacillus casei, Staphylococcus xylosus and M. pneumoniae were crystal- lized and their structures determined [17–20]. All three enzymes form homohexamers which are arranged as bilayered trimers. The monomers are composed of two domains. The N-terminal domains are exposed to the environment, whereas the C-terminal domains form the compact core of the protein. The C-terminal domains alone are sufficient for the formation of hexamers and for all known enzymatic activities. The active site of the protein contains the well-conserved ATP-binding P-loop motif, a so-called Ôkinase-2 motifÕ composed of two neighbouring aspartate residues and the HPrK/P family signature sequence motif [19–21]. The function of the N-terminal domains is so far unknown. Analysis of the structure did not reveal any specific features of the M. pneumoniae enzyme that would explain how its activity is regulated differently from the other enzymes [19]. In this work, we studied the interaction of the M. pneu- moniae HPrK/P with its low molecular mass effectors using fluorescence spectroscopy. We found that the enzyme has a very high affinity for ATP, explaining kinase activity even at low ATP concentrations. Binding of ATP is affected in the presence of Fru1,6P 2 allowing a more rapid phosphoryl transfer to HPr. Fru1,6P 2 and ATP must be bound by different sites of the enzyme, as proposed for the B. subtilis HPrK/P by Pompeo et al. [29], as the P-loop is required for binding of ATP but not Fru1,6P 2 . Materials and methods Bacterial strains and growth conditions E.coli DH5a and BL21(DE3)/pLysS [22] were used for cloning experiments and overexpression of recombinant proteins, respectively. The cells were grown in Luria– Bertani medium, and transformants were selected on plates containing ampicillin (100 lgÆmL )1 ). DNA manipulation and plasmid constructions Transformation of E.coli and plasmid DNA extraction were performed using standard procedures [22]. Restriction enzymes, T4 DNA ligase and DNA polymerase were used as recommended by the manufacturers. DNA fragments were purified from agarose gels using a Nucleospin Gel Extraction kit (Macherey & Nagel, Du ¨ ren, Germany). DNA sequences were determined using the dideoxy chain termination method [22]. Plasmids pGP204 and pGP217 [4] were used for the overexpression of His-tagged forms of M. pneumoniae HPrK/P and HPr, respectively. To express a variant of HPrK/P with a single Trp residue (F200W), we constructed plasmid pGP628 by a two-step PCR protocol as described previously using the muta- genic primer MZ6 (5¢-GTAGCGAAAAAGT GGATGGA AATCCGT; the mutation is underlined) and the cloning primers KS9 and KS10 [4]. The resulting PCR product was cloned between the SalIandHindIII sites of the expression vector pWH844 [23]. To attach a Strep-tag to recombinant proteins, we used the vector pGP172. This plasmid was obtained by cloning a fragment encoding the Strep-tagÒ II from pASK-IBA5 (IBA, Go ¨ ttingen, Germany) into the expression vector pET3c (Novagen). For this purpose, the fragment encoding the Strep-tag II was amplified using the primers CD13 (5¢-AAA CATATGGCTAGCTGGAGCCACCCGCAG TTC, a NdeI site introduced by the primer is underlined) and CD14 (5¢-AAGCTTAGTTAGATATCAGAGACC ATG). The PCR product was digested with NdeIand BamHIandclonedintopET3ccutwiththesameenzymes. The wild-type and F200W forms of HPrK/P were tagged by amplifying the hprK alleles of pGP204 and pGP628 using the primers SH1 (5¢-AAA CCGCGGCAATGAAAAAG TTATTAGTCAAGGAG) and SH3 (5¢-AAA GGATCC GGTCTGCTACTAACACTAGGATTCATC). The PCR fragments were cut with SacII and BamHI and cloned into pGP172 linearized with the same enzymes. The resulting plasmids were pGP611 and pGP612 for the wild-type and F200W forms of hprK, respectively. Mutagenesis of the hprK-F200W allele was performed by two-step PCR as described [4]. Protein purification E.coliBL21(DE3)/pLysS was used as host for the over- expression of recombinant proteins. Expression was indu- ced by the addition of isopropyl thio-b- D -galactoside (final concentration 1 m M ) to exponentially growing cultures (A 600 ¼ 0.8). Cells were lysed using sonication (6 · 30 s, 4 °C, 50 W). After lysis the crude extracts were centrifuged 368 M. Merzbacher et al.(Eur. J. Biochem. 271) Ó FEBS 2003 at 15 000 g for 30 min. For purification of His-tagged proteins, the resulting supernatants were passed over a Ni 2+ HiTrap chelating column (5 mL bed volume; Pharmacia) followed by elution with an imidazole gradient (from 0 to 500 m M imidazole in buffer containing 10 m M Tris/HCl, pH 7.5, 600 m M NaCl, 10 m M 2-mercaptoethanol) over 30 mL at a flow rate of 0.5 mLÆmin )1 . For HPrK/P carrying an N-terminal Strep-tag, the crude extract was passed over a Streptactin column (IBA). The recombinant protein was eluted with desthiobiotin (Sigma; final concentration 2.5 m M ). For the recombinant HPr protein, the overproduced protein was purified from the pellet fraction of the lysate by urea extraction and renatured as described previously [4]. After elution, the fractions were tested for the desired protein using SDS/12.5% (w/v) polyacrylamide gels for HPrK/P and 10% (v/v) Tris/Tricine gels [24] for HPr. The relevant fractions were combined and dialysed overnight. Purified proteins were concentrated using Microsep TM Microconcentrators with a molecular mass cut-off of 3 kDa and 10 kDa for HPr and HPrK/P, respectively (Pall Filtron, Northborough, MA, USA). Protein concentration was determined by the method of Bradford [25] using the Bio-rad dye-binding assay with BSA as standard. Assay of HPrK/P activity Activity assays were carried out with purified HPrK/P in assay buffer (10 m M MgCl 2 , 25 m M Tris/HCl, pH 7.6, 1m M dithiothreitol) using purified (His 6 )HPr or (His 6 )HPr- Ser-P. ATP, potassium phosphate and Fru1,6P 2 were added as indicated. The assays were carried out at 37 °Cfor 15 min followed by thermal inactivation of the enzyme (4 min at 95 °C). The assay mixtures were analyzed on 10% native polyacrylamide gels as described previously [26]. Proteins were visualized by Coomassie staining. (His 6 )HPr-Ser-P of M. pneumoniae was prepared as described previously [4]. Fluorescence measurements All experiments were performed at 25 ± 0.1 °Cusinga Fluorolog-3 Jobin Yvon–Spex spectrofluorimeter. All spec- tra were corrected for buffer fluorescence. Fluorescence measurements were carried out after dilution of HPrK/P (170 n M final concentration) and equilibration for 30 min in 2 mL buffer containing 200 m M NaCl, 10 m M Tris/HCl, pH 7.5 and 5 m M dithiothreitol. After each titration step with ATP/MgCl 2 , Fru1,6P 2 or P i , the mixture was stirred for 1 min and equilibrated for 4 min at 25 ± 0.1 °C. After specific excitation of the tryptophan residue at 295 nm, the fluorescence emission was recorded at 300–450 nm. Binding of ligands was monitored by the variation in intrinsic tryptophan fluorescence after addition of increasing con- centrations of effectors and corrections for the variation in volume. For quantitative analysis, all fluorescence peaks were integrated between 320 and 380 nm, and the formation of the HPrK/P–ligand complex was described as saturation f ¼ (f x ) f 0 )/(f f ) f 0 ), where f x is the integrated relative fluorescence intensity between 320 and 380 nm at each single titration step, f 0 before addition of any ligand, and f f at the end of the titration. Fitting of the curves was performed using the GRAPHPAD PRISM software (GraphPad Software, Inc.) and a one-site binding model f ¼ [mK A c (ligand) ]/[1 + K A c (ligand) ], where m is the overall concentration of binding sites and K A is an equilibrium association constant [27]. The equilibrium dissociation constants K d were calculated from K d ¼ 1/K A . Results Construction and characterization of an HPrK/P variant with a single tryptophan fluorescence probe ATP is the substrate of HPrK/P, but at the same time it is an effector molecule of HPrK/P from B. subtilis and related bacteria [16]. It was therefore interesting to study the interaction of HPrK/P with ATP and other potential effectors. Changes in intrinsic fluorescence of tryptophan residues in proteins can provide a sensitive assay for such interactions. However, the M. pneumoniae HPrK/P does not contain any Trp residues. Therefore, we constructed, expressed, and purified a variant of HPrK/P with a single Trp residue. We chose to place the Trp residue at position 200 of HPrK/P in order to have it close to the P-loop ATP-binding motif and the signature sequence of HPrK/P. Moreover, the Trp replaces another aromatic amino acid, Phe200, in this mutant [19]. The mutant alleles were obtained as described in Materials and methods, and the proteins were purified as versions with aHis-oraStrep-tag. In a previous study, we determined the structure of the His-tagged HPrK/P from M. pneumo- niae [19], which thus serves as the standard. However, some mutant variants precipitated during purification if a His-tag was present. These proteins were studied using the Strep-tag. Kinase and phosphorylase activities of the single Trp mutants were assayed and found to be very similar to those of the wild-type protein [4] (data not shown). Thus, these proteins could be used to determine ligand affinities. Binding of ATP to M. pneumoniae HPrK/P The HPrK/P from B. subtilis is active as a kinase only at high ATP concentrations or at a low ATP concentration in the presence of Fru1,6P 2 [16]. In contrast, the protein of M. pneumoniae already exhibits kinase activity at very low ATP concentrations. We were therefore interested to study the ATP-binding characteristics of the M. pneumoniae enzyme by fluorescence measurements. HPrK/P(F200W) was excited at 295 nm, and the emission spectrum was recorded. As shown in Fig. 1A, fluorescence reached a maximum at 350 nm. The emission was constant over more than an hour (data not shown), indicating that the protein was sufficiently stable for analysis of ATP binding. HPrK/ P(F200W) was incubated with increasing concentrations of ATP, and the changes in fluorescence were recorded (Fig. 1A). With increasing ATP concentrations, the fluor- escence intensities decreased until a minimum was reached. To be sure that the observed effects were indeed specific, we tested the changes in fluorescence on addition of AMP. No decrease in fluorescence was detected. Moreover, we incubated wild-type HPrK/P carrying a Strep-tag with ATP. Again, no changes in fluorescence were observed. Ó FEBS 2003 M. pneumoniae HPr kinase/phosphatase (Eur. J. Biochem. 271) 369 Thus, the Trp residue is placed at a suitable position for specifically detecting interaction with ATP. To quantify the binding of ATP to HPrK/P, the saturation (f) of HPrK/P with ATP was plotted against the ATP concentrations (Fig. 1B). In contrast with the results with the B. subtilis HPrK/P, no sigmoidal curve indicative of co-operative binding was obtained. The dissociation constant K d was calculated to be 5.4 ± 1.3 l M . For the HPrK/P of B. subtilis, it was shown that P i inhibits binding of nucleotides [28]. To test whether this applies also to the M. pneumoniae enzyme, we analyzed the effect of increasing P i concentrations on the binding of HPrK/P to ATP. HPrK/P(F200W) was incubated with a saturating ATP concentration (50 l M )andP i was added. As shown in Fig. 2, the fluorescence intensity was substantially decreased in the presence of ATP if no additional ligand was present (ligand concentration 0 m M ). However, with increasing concentrations of P i , an increase in fluorescence intensity was detected, suggesting that P i can inhibit ATP binding to M. pneumoniae HPrK/P. Incubation of HPrK/ P(F200W) with P i did not result in any changes in fluorescence intensity. Thus, the Trp at position 200 may not be suitable for detecting phosphate binding. Moreover, we studied the effect of the chloride anion on ATP binding. No effect was observed, even on addition of 150 m M chloride. Thus, the negative effect of phosphate on ATP binding results from a specific interaction with HPrK/P. Interaction of M. pneumoniae HPrK/P with Fru1,6 P 2 High concentrations of Fru1,6P 2 indicate high glycolytic activity and stimulate kinase activity of HPrK/P in B. sub- tilis and related bacteria [16]. In M. pneumoniae, Fru1,6P 2 is involved in the control of phosphorylase rather than kinase activity [4]. We wished therefore to study the binding of Fru1,6P 2 to M. pneumoniae HPrK/P by fluorescence measurements. As described for ATP-binding assays, HPrK/P(F200W) was incubated with increasing amounts of Fru1,6P 2 , and changes in fluorescence were recorded. Fig. 1. ATP binding to HPrK/P. (A) Increasing concentrations of ATP/MgCl 2 from 0 to 200 l M were added to 2 mL 170 n M HPrK/P in buffer containing 200 m M NaCl, 10 m M Tris/HCl, pH 7.5, and 5 m M dithiothreitol. The fluorescence intensity was recorded from 300 to 450 nm after each titration step. From the upper to the lower curves, the concentration of ATP was 0, 5, 10, 20, 50, 100 and 200 l M , respectively. (B) All fluorescence curves were integrated between 320 and 380 nm, and the saturation f of HPrK/P with ATP was plotted against the concentration of the ligand (s). The continuous curve represents the fitted values of the saturation f. Fig. 2. Influence of P i on ATP bound to HPrK/P. HPrK/P (170 n M ) was incubated in 2 mL buffer containing 200 m M NaCl, 10 m M Tris/ HCl, pH 7.5, and 5 m M dithiothreitol. After 30 min incubation at 25 °C, the fluorescence of the sample was recorded from 300 to 450 nm (a). Then 50 l M ATP/MgCl 2 was added to the mixture and the fluorescence of the sample was measured (b). The titrations with K 2 HPO 4 or NaCl were performed between 10 and 150 m M ligand, and, after each titration step, the fluorescence was recorded. All curves were integrated between 320 and 380 nm and the relative fluorescence intensity was plotted against the concentration of the ligand. 370 M. Merzbacher et al.(Eur. J. Biochem. 271) Ó FEBS 2003 Again, a reduction in fluorescence intensity was observed with increasing concentrations of Fru1,6P 2 . The quantifi- cation of Fru1,6P 2 binding revealed a K d of 4.5 ± 1 m M . To test the specificity of binding, we studied the changes in fluorescence after addition of Fru6P. No effect was observed. Effect of Fru1,6 P 2 on ATP binding by HPrK/P Fru1,6P 2 is not required for kinase activity of M. pneumo- niae HPrK/P but it inhibits the stimulation of phosphory- lase activity by P i [4]. As P i itself antagonizes the binding of ATP, it seemed reasonable to assume an effect of Fru1,6P 2 on ATP binding as well. Indeed, Fru1,6P 2 was able to attenuate the decrease in fluorescence intensity of HPrK/ P(F200W) seen after addition of ATP and MgCl 2 (Fig. 3). In the presence of 1 m M Fru1,6P 2 , the apparent K d for ATP was increased to 31 ± 4 l M , as compared with 5.4 l M in its absence. We may thus conclude that Fru1,6P 2 influences the binding of ATP to HPrK/P. The specificity of this effect was addressed by testing whether Fru6P would give the same result. In agreement with the observation that Fru6P does not bind to HPrK/P (see above), it had also no effect on ATP binding (Fig. 3). If Fru1,6P 2 interferes with ATP binding, it might also affect HPrK/P activity. In our previous experiments, we detected an effect of Fru1,6P 2 on phosphorylase but not on kinase activity [4]. To test the role of Fru1,6P 2 in kinase activity more rigorously, we studied the time course of the kinase reaction in the presence and absence of Fru1,6P 2 . Under the conditions used in this assay, about 50% of HPr was phosphorylated after 4 min in the absence of Fru1,6P 2 (Fig. 4A). In the presence of Fru1,6P 2 , the reaction was somewhat accelerated: significant HPr phosphorylation was already detected after 2 min. The quantification of this reaction and the values for the slope of the graphs in the linear part between 0 and 4 min (without Fru1,6P 2 , 12.4 ± 2.4; with Fru1,6P 2 , 20.1 ± 0.3; Fig. 4B) indicate that the kinase reaction is indeed faster in the presence of Fru1,6P 2 . Although this was only a slight effect, it was reproducible and is therefore considered to be relevant (see the discussion). Role of specific residues in the P-loop and the signature sequence in the interaction with ATP and Fru1,6 P 2 Of the three strongly conserved regions in HPrK/P, the kinase-2 motif was implicated in the catalytic mechanism [21]. The P-loop is involved in nucleotide binding whereas the function of the signature sequence is so far unknown. Previous mutation analyses with M. pneumoniae HPrK/P revealed that both regions were essential for enzymatic Fig. 3. Effect of Fru1,6P 2 on ATP binding. Increasing concentrations of ATP/MgCl 2 were added to the 2 mL mixture containing 170 n M HPrK/P and in addition 1 m M Fru1,6P 2 or as a control 5 m M Fru6P. The fluorescence intensity was measured after each titration, and the decrease in fluorescence plotted against the ligand concentration. Fig. 4. Time-dependent HPr phosphorylation in the presence or absence of Fru1,6P 2 . (A) 20 l M HPr was incubated with 500 n M HPrK/P and 100 l M ATP in reaction buffer in the presence or absence of 5 m M Fru1,6P 2 . Before and 1, 2, 4 and 6 min after the start of the reaction, 20 lLmixturewastransferredtoafreshreactiontubeandthereaction was stopped (96 °C, 4 min). The fractions were then separated on a native polyacrylamide gel, analyzed using the TINA quantification software (raytest; Isotopenmessgera ¨ te GmbH, Straubenhardt, Ger- many), and the amount of remaining unphosphorylated HPr was plotted against time (B). Ó FEBS 2003 M. pneumoniae HPr kinase/phosphatase (Eur. J. Biochem. 271) 371 activity. However, some mutations strongly reduced or abolished the phosphorylase activity without affecting kinase activity [4]. It was therefore interesting to study these mutants in more detail. For this purpose, a series of previously described mutations was introduced into the Strep-tagged variant of HPrK/P(F200W). This allows easy purification of the mutant proteins and analysis of the effects of the mutations by fluorescence measurements. Binding of ATP was studied as described for the wild-type protein. As observed previ- ously, the fluorescence intensity of the wild-type protein decreased greatly after ATP addition (Fig. 5). Similarly, the proteins with mutations at positions S161T, R204 and G207 showed good ATP binding. The G154A mutant protein exhibited greatly reduced ATP binding with K d about sevenfold increased (Fig. 5, Table 1). In contrast, the G159A and K160A proteins were not able to bind ATP. This finding is in good agreement with the observation that these proteins had lost all enzymatic activity [4]. Binding of Fru1,6P 2 was not significantly altered in the mutant proteins, as judged from the dissociation constants (Table 1). In the presence of Fru1,6P 2 , the apparent K d for ATP was greatly increased for the wild-type protein. Similar effects were found for the S161T, R204K, and G207A mutant proteins. In contrast, the apparent K d values for ATP were increased even more for the G154A and the S161A proteins (Table 1). These two proteins exhibit reduced kinase activity and have lost their phosphorylase activity [4]. Discussion In contrast with its counterparts from B. subtilis and other low-GC Gram-positive bacteria, the M. pneumoniae HPrK/ P is already active at very low ATP concentrations, suggesting that ATP does not control the activity of the enzyme in this organism. In this work, we demonstrate that the K d of M. pneumoniae HPrK/P for ATP is indeed very low (5.4 l M ) compared with the value determined for the B. subtilis enzyme (100 … 300 l M ) [16,21,29]. Thus, the B. subtilis enzyme requires an ATP concentration that is only present in the cell at high metabolic activity. Alternat- ively, Fru1,6P 2 as an intracellular signal of glycolytic activity, stimulates activity of B. subtilis HPrK/P at low ATP concentrations. Although binding of Fru1,6P 2 does not affect the affinity of B. subtilis HPrK/P for ATP [16,29], an altered affinity for ATP was observed in this study for the M. pneumoniae enzyme. The presence of Fru1,6P 2 may be important for the reaction rate of M. pneumoniae HPrK/P: a weak but reproducible acceleration of the kinase reaction was observed if Fru1,6P 2 was added to the reaction. Similarly, an acceleration of HPrK/P kinase activity was recently demonstrated for the enzyme from Streptococcus salivarius. As found here for M. pneumoniae HPrK/P, this enzyme exhibits kinase activity in the absence of Fru1,6P 2 , although much higher ATP concentrations are required [30]. It is of interest to investigate the structural determinants of the altered regulation of M. pneumoniae HPrK/P. With the crystallization and solution of the 3D structure of the protein to a resolution of 2.5 A ˚ , it is now possible to directly compare structures of the M. pneumoniae, S. xylosus and L. casei HPrK/P proteins [17–19,31]. However, such a comparison did not reveal any obvious structural differ- ences that may be responsible for the observed difference in regulation and nucleotide affinity [19,20]. Therefore, we extended a previous mutational analysis by quantitatively analysing the mutants. In the P-loop, we studied the roles of the conserved amino acids G154, G159, K160, and S161 [4,21,32]. Residue G154 has been proposed to stabilize the P-loop of adenylate kinase by forming a hydrogen bond with K160 [33]. In the structure of M. pneumoniae HPrK/P, a water molecule forms a bridge between these two amino Fig. 5. ATP binding of point mutated HPrK/P. Increasing concentra- tions of ATP/MgCl 2 from 0 to 200 l M wereaddedto2mL170n M HPrK/P in buffer containing 200 m M NaCl, 10 m M Tris/HCl pH 7.5 and 5 m M dithiothreitol. The fluorescence intensity was measured after each titration step, and the decrease in fluorescence plotted against the concentration of the ligand. Table 1. K d values for binding of ATP and Fru1,6P 2 to the HPrK/P mutants. All values were determined in duplicate. Mean values ± SD are shown. n.d., No change in fluorescence was detectable during titration. Wild-type G154A G159A K160A S161A S161T R204K G207A ATP (l M ) 5.4 ± 1.3 39 ± 4.1 n.d. n.d. 11 ± 0.9 6.8 ± 0.5 6.9 ± 0.4 6.7 ± 0.3 ATP (l M ) a 31 ± 4.0 80 ± 9.8 n.d. n.d. 61 ± 1.5 38 ± 1.8 34 ± 2.8 37 ± 2.6 Fru1,6P 2 (m M ) 4.5 ± 1.0 3.6 ± 0.8 4.4 ± 0.9 4.6 ± 1.3 2.7 ± 0.9 2.4 ± 0.7 3.8 ± 1.3 4.8 ± 1.7 a The HPrK/P solution contained 1 m M Fru1,6P 2 . Values given are apparent dissociation constants. 372 M. Merzbacher et al.(Eur. J. Biochem. 271) Ó FEBS 2003 acids [19]. The mutational analysis supports the idea that G154 plays a structural rather than an enzymatic role: the mutant G154A is still active as a kinase, suggesting that it can bind ATP. Indeed, the K d of this mutant protein for ATP was increased about eightfold, which may explain the reduced kinase activity. Mutational studies of the corresponding residue with the B. subtilis HPrK/P revea- led conflicting results: Hanson et al. [26] found decreased kinase and phosphorylase activity whereas Pompeo et al. [29] reported the complete loss of both enzymatic activities and did not detect any nucleotide binding. The reason for this discrepancy is so far unknown. The three amino acids G159, K160, and S161 are involved in binding of the triphosphate moiety of ATP or GTP [19,32]. The muta- tions G159A and K160A completely abolish the binding of ATP (Table 1, Fig. 5). Accordingly, neither mutant enzyme exhibited any kinase or phosphorylase activities [4]. S161 is not only involved in nucleotide binding but also in the co-ordination of a magnesium cation necessary for catalytic activity [19,32,34]. Two different mutations of S161 were studied: replacement with the structur- ally similar amino acid threonine and replacement with alanine. In many P-loop proteins, a threonine is present and can functionally replace the serine residue at this position [32]. Indeed, the S161T mutant protein binds ATP with high affinity and has both enzymatic activities. Similar results were reported for the B. subtilis HPrK/P [26,29]. Functional interchangeability of serine and threo- nine residues of the P-loop was also detected for the corresponding position of the liver 6-phosphofructo-2- kinase [35]. In contrast, the mutation S161A resulted in greatly impaired kinase activity and loss of phosphorylase activity [4]. However, the affinity of this protein for ATP is only slightly affected. This is in good agreement with the proposal that S161 is involved in the catalytic activity, in addition to its role in triphosphate binding. In the HPrK/P of B. subtilis, the corresponding S160A mutation resulted in loss of ATP binding and enzymatic activities [29]. The function of the HPrK/P family signature sequence [8] has so far not been elucidated. The two mutant proteins studied in this work had different properties: the R204K enzyme had wild-type kinase and a reduced phosphorylase activity. In contrast, the G207A mutant protein had lost all activities [4]. However, both mutant proteins bound ATP as effectively as the wild-type protein. Thus, the signature sequence motif seems to be involved in the catalytic mechanism rather than the inter- action with nucleotides. It was proposed that the guanid- inium group of arginine may be involved in the stabilization of the transition state complex of HPr(Ser-P) during its dephosphorylation by HPrK/P [19]. An arginine residue corresponding to R204 is also present in the E.coliPEP carboxykinase (R333). In this enzyme, R333 makes contact with an oxygen atom of the c-phosphate of ATP [36], and it has been proposed that the lysine present in the HPrK/P-R204K mutant could also provide a salt bridge with the ATP [19]. Replacement of the R204 residue of B. subtilis HPrK/P with an alanine resulted in a large reduction in kinase activity, although the affinity for ATP was not significantly affected [29]. G207 is located in asmallb-sheet, and it was proposed that this amino acid cannot be replaced at this position of the M. pneumoniae HPrK/P for structural reasons. Indeed, the G207A mutation abolished both enzymatic activities without affecting ATP binding. In B. subtilis, the corresponding mutation results in reduced kinase activity and loss of phosphorylase activity [26]. Taken together, the results of this and previous studies on the signature sequence motif of the M. pneumoniae and B. subtilis HPrK/P indicate that this motif is important for the catalytic mechanism of both phosphorylation and dephosphorylation. The close proximity of the signature sequence motif to the P-loop seen in all HPrK/P crystal structures [17–19] supports this idea. Future work on M. pneumoniae HPrK/P will address two major questions: (a) the structural reasons for the high affinity for ATP and thus the altered regulation; (b) the regulatory role of HPrK/P in the physiology of M. pneu- moniae. Acknowledgements Serkan Halici is acknowledged for help with some experiments. We are grateful to Marco Diel for helpful discussions. This work was supported by grants from the Deutsche Forschungsgemeinschaft and the Fonds der Chemischen Industrie to J.S. and W.H. References 1. Greenberg, E.P. (2000) Bacterial genomics: pump up the versa- tility. Nature (London) 406, 947–948. 2. Razin, S., Yogev, D. & Naot, Y. (1998) Molecular biology and pathogenicity of mycoplasmas. Microbiol. Mol. Biol. Rev. 62, 1094–1156. 3. Himmelreich,R.,Hilbert,H.,Plagens,H.,Pirkl,E.,Li,B C.& Herrmann, R. (1996) Complete sequence analysis of the genome of the bacterium Mycoplasma pneumoniae. Nucleic Acids Res. 24, 4420–4449. 4. Steinhauer, K., Jepp, K., Hillen, W. & Stu ¨ lke, J. (2002) A novel mode of control of Mycoplasma pneumoniae HPr kinase/phos- phatase activity reflects its parasitic life style. Microbiology 148, 3277–3284. 5. Hoischen, C., Dijkstra, A., Rottem, S., Reizer, J. & Saier, M.H. (1993) Presence of protein constituents of the gram-positive bacterial phosphotransferase regulatory system in Acholeplasma laidlawii. J. Bacteriol. 175, 6599–6604. 6. Zhu, P P., Nosworthy, N., Ginsburg, A., Miyata, M., Seok, Y J. & Peterkofsky, A. (1997) Expression, purification, and char- acterization of Enzyme IIA glc of the phosphoenolpyruvate:sugar phosphotransferase system of Mycoplasma capricolum. Biochem- istry 36, 6947–6953. 7. Galinier, A., Kravanja, M., Engelmann, R., Hengstenberg, W., Kilhoffer, M C., Deutscher, J. & Haiech, J. (1998) New protein kinase and protein phosphatase families mediate signal transduc- tion in bacterial catabolite repression. Proc. Natl Acad. Sci. USA 95, 1823–1828. 8. Reizer, J., Hoischen, C., Titgemeyer, F., Rivolta, C., Rabus, R., Stu ¨ lke, J., Karamata, D., Saier, M.H. & Hillen, W. (1998) A novel protein kinase that controls carbon catabolite repression in bac- teria. Mol. Microbiol. 27, 1157–1169. 9. Kravanja, M., Engelmann, R., Dossonnet, V., Blu ¨ ggel, M., Meyer, H.E., Frank, R., Galinier, A., Deutscher, J., Schnell, N. & Hengstenberg, W. (1999) The hprK gene of Enterococcus faecalis encodes a novel bifunctional enzyme: the HPr kinase/phosphatase. Mol. Microbiol. 31, 59–66. Ó FEBS 2003 M. pneumoniae HPr kinase/phosphatase (Eur. J. Biochem. 271) 373 10. Deutscher, J., Kessler, U., Alpert, C.A. & Hengstenberg, W. (1984) Bacterial phosphoenolpyruvate-dependent phosphotrans- ferase system: P-ser-HPr and its possible regulatory function. Biochemistry 23, 4455–4460. 11. Ludwig, H., Rebhan, N., Blencke, H M., Merzbacher, M. & Stu ¨ lke, J. (2002) Control of the glycolytic gapA operon by the catabolite control protein A in Bacillus subtilis: a novel mechanism of CcpA-mediated regulation. Mol. Microbiol. 45, 543–553. 12. Moreno, M.S., Schneider, B.L., Maile, R.R., Weyler, W. & Saier, M.H. (2001) Catabolite repression mediated by CcpA protein in Bacillus subtilis: novel modes of regulation revealed by whole- genome analyses. Mol. Microbiol. 39, 1366–1381. 13. Blencke, H M., Homuth, G., Ludwig, H., Ma ¨ der, U., Hecker, M. &Stu ¨ lke, J. (2003) Transcriptional profiling of gene expression in response to glucose in Bacillus subtilis: regulation of the central metabolic pathways. Metab. Engn. 5, 133–149. 14. Gunnewijk, M.G.W., van den Bogaard, P.T.C., Venhoff, L.M., Heuberger, E.H., de Vos, W.M., Kleerebezem, M., Kuipers, O.P. & Poolman, B. (2001) Hierarchical control versus autoregulation of carbohydrate utilization in bacteria. J. Mol. Microbiol. Bio- technol. 3, 401–413. 15. Boe ¨ l, G., Mijakovic, I., Maze ´ , A., Poncet, S., Taha, M.K., Laribe, M., Darbon, E., Khemiri, A., Galinier, A. & Deutscher, J. (2003) Transcription regulators potentially controlled by HPr kinase/ phosphorylase in Gram-negative bacteria. J. Mol. Microbiol. Biotechnol. 5, 206–215. 16. Jault, J.M., Fieulaine, S., Nessler, S., Gonzalo, P., Di Pietro, A., Deutscher, J. & Galinier, A. (2000) The HPr kinase from Bacillus subtilis is a homo-oligomeric enzyme which exhibits strong posi- tive cooperativity for nucleotide and fructose 1,6-bisphosphate binding. J. Biol. Chem. 275, 1773–1780. 17. Fieulaine, S., Morera, S., Poncet, S., Monedero, V., Gueguen- Chaignon, V., Galinier, A., Janin, J., Deutscher, J. & Nessler, S. (2001) X-ray structure of HPr kinase: a bacterial protein with a P-loop nucleotide-binding domain. EMBO J. 20, 3917–3927. 18. Ma ´ rquez, J.A., Hasenbein, S., Koch, B., Fieulaine, S., Nessler, S., Russell, R.B., Hengstenberg, W. & Scheffzek, K. (2002) Structure of the full-length HPr kinase/phosphatase from Staphylococcus xylosus at 1.95 A ˚ resolution: mimicking the product/substrate of the phosphotransfer reactions. Proc.NatlAcad.Sci.USA99, 3458–3463. 19. Allen, G.S., Steinhauer, K., Hillen, W., Stu ¨ lke, J. & Brennan, R.G. (2003) Crystal structure of HPr kinase/phosphatase from Myco- plasma pneumoniae. J. Mol. Biol. 326, 1203–1217. 20. Nessler, S., Fieulaine, S., Poncet, S., Galinier, A., Deutscher, J. & Janin, J. (2003) HPr kinase/phosphorylase, the sensor enzyme of catabolite repression in Gram-positive bacteria: structural aspects of the enzyme and the complex with its protein substrate. J. Bacteriol. 185, 4003–4010. 21. Galinier, A., Lavergne, J P., Geourjon, C., Fieulaine, S., Nessler, S. & Jault, J M. (2002) A new family of phosphotransferases with aP-loopmotif.J. Biol. Chem. 277, 11362–11367. 22. Sambrook, J., Fritsch, E.F. & Maniatis, T. (1989) Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. 23. Schirmer, F., Ehrt, S. & Hillen, W. (1997) Expression, inducer spectrum, domain structure, and function of MopR, the regulator of phenol degradation in Acinetobacter calcoaceticus NCIB8250. J. Bacteriol. 179, 1329–1336. 24. Scha ¨ gger, H. & von Jagow, G. (1987) Tricine-sodium dodecyl sulfate-polyacrylamide gel electrophoresis for the separation of proteins in the range from 1 to 100 kDa. Anal. Biochem. 166, 368–379. 25. Bradford, M.M. (1976) A rapid sensitive method for the quanti- fication of microgram quantities of protein utilizing the principle of protein dye binding. Anal. Biochem. 72, 248–254. 26. Hanson, K.G., Steinhauer, K., Reizer, J., Hillen, W. & Stu ¨ lke, J. (2002) HPr kinase/phosphatase of Bacillus subtilis: expression of the gene and effects of mutations on enzyme activity, growth and carbon catabolite repression. Microbiology 148, 1805–1811. 27. Winzor, D.J. & Sawyer, W.H. (1995) Quantitative Characterisa- tion of Ligand Binding, New York: John Wiley & Sons. 28. Lavergne, J P., Jault, J M. & Galinier, A. (2002) Insights into the functioning of Bacillus subtilis HPr kinase/phosphatase: affinity for its protein substrates and role of cations and phosphate. Bio- chemistry 41, 6218–6225. 29. Pompeo, F., Granet, Y., Lavergne, J P., Grangeasse, C., Nessler, S., Jault, J M. & Galinier, A. (2003) Regulation and mutational of the HPr kinase/phosphorylase from Bacillus subtilis. Biochemistry 42, 6762–6771. 30. Frey, N., Nessler, S., Fieulaine, S., Vaillancourt, K., Frenette, M. & Vadeboncoeur, C. (2003) The HPr (Ser) kinase of Streptococcus salivarius: a hexameric bifunctional enzyme controlled by glyco- lytic intermediates and inorganic phosphate. FEMS Microbiol. Lett. 224, 67–72. 31. Steinhauer, K., Allen, G.S., Hillen, W., Stu ¨ lke, J. & Brennan, R.G. (2002) Crystallization, preliminary X-ray analysis and biophysical characterization of HPr kinase/phosphatase of Mycoplasma pneumoniae. Acta Crystallogr. Sect. D 58, 515–518. 32. Saraste, M., Sibbald, P.R. & Wittinghofer, A. (1990) The P-loop: a common motif in ATP- and GTP-binding proteins. Trends Bio- chem. Sci. 15, 430–434. 33. Tian, G., Yan, H., Jiang, R T., Kishi, F., Nakazawa, A. & Tsai, M D. (1990) Mechanism of adenylate kinase: are the essential lysines essential? Biochemistry 29, 4296–4304. 34. Ramstro ¨ m, H., Sanglier, S., Leize-Wagner, E., Philippe, C., van Dorsselaer, A. & Haiech, J. (2003) Properties and regulation of the bifunctional enzyme HPr kinase/phosphatase in Bacillus subtilis. J. Biol. Chem. 278, 1174–1185. 35. Vertommen, D., Bertrand, L., Sontag, B., Di Pietro, A., Louckx, M.P., Vidal, H., Hue, L. & Rider, M.H. (1996) The ATP-binding site in the 2-kinase domain of liver 6-phosphofructo- 2-kinase/fructose-2,6-bisphosphatase. J. Biol. Chem. 271, 17875– 17880. 36. Matte, A., Tari, L.W., Goldie, H. & Delbaere, L.T. (1997) Structure and mechanism of phosphoenolpyruvate carboxykinase. J. Biol. Chem. 272, 8105–8108. 374 M. Merzbacher et al.(Eur. J. Biochem. 271) Ó FEBS 2003 . Mycoplasma pneumoniae HPr kinase/phosphorylase Assigning functional roles to the P-loop and the HPr kinase/phosphorylase signature sequence motif Matthias. place the Trp residue at position 200 of HPrK/P in order to have it close to the P-loop ATP-binding motif and the signature sequence of HPrK/P. Moreover, the