Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Comparative Analysis of Pyruvate Kinases from the Hyperthermophilic Archaea Archaeoglobus fulgidus, Aeropyrum pernix, and Pyrobaculum aerophilum and the Hyperthermophilic Bacterium Thermotoga maritima

Comparative Analysis of Pyruvate Kinases from the Hyperthermophilic Archaea Archaeoglobus... THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 278, No. 28, Issue of July 11, pp. 25417–25427, 2003 © 2003 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A. Comparative Analysis of Pyruvate Kinases from the Hyperthermophilic Archaea Archaeoglobus fulgidus, Aeropyrum pernix, and Pyrobaculum aerophilum and the Hyperthermophilic Bacterium Thermotoga maritima UNUSUAL REGULATORY PROPERTIES IN HYPERTHERMOPHILIC ARCHAEA* Received for publication, October 8, 2002, and in revised form, March 21, 2003 Published, JBC Papers in Press, March 21, 2003, DOI 10.1074/jbc.M210288200 Ulrike Johnsen, Thomas Hansen, and Peter Scho ¨ nheit‡ From the Institut fu ¨ r Allgemeine Mikrobiologie, Christian-Albrechts-Universita ¨ t Kiel, Am Botanischen Garten 1–9, Kiel D-24118, Germany Pyruvate kinases (PK, EC 2.7.1.40) from three hyper- mophilic archaea and in the hyperthermophilic bacterium thermophilic archaea (Archaeoglobus fulgidus strain Thermotoga revealed that the classic Embden-Meyerhof (EM) 7324, Aeropyrum pernix, and Pyrobaculum aerophilum) pathway is operative only in Thermotoga, whereas in all ar- and from the hyperthermophilic bacterium Thermotoga chaea, the EM pathway exists in modified versions. The mod- maritima were compared with respect to their thermo- ified EM pathways contain, e.g. unusual glucokinases (GLK) philic, kinetic, and regulatory properties. PKs from the and 6-phosphofructokinases (PFK) such as ADP-dependent archaea are 200-kDa homotetramers composed of 50- GLK and ADP-dependent PFK in Pyrococcus, Thermococcus, kDa subunits. The enzymes required divalent cations, and Archaeoglobus; unusual ATP-dependent archaeal GLKs of 2 2 Mg and Mn being most effective, but were independ- the ROK (Regulators, ORFs, Kinases) protein family; non-reg- ent of K . Temperature optima for activity were 85 °C ulatory ATP-dependent PFKs in Desulfurococcus and Aeropy- (A. fulgidus) and above 98 °C (A. pernix and P. aerophi- rum; and pyrophosphate-dependent PFK in Thermoproteus.In lum). The PKs were highly thermostable up to 110 °C (A. addition, the modified EM pathways contain novel enzymes of pernix) and showed melting temperatures for thermal glyceraldehyde 3-phosphate (GAP) oxidation to 3-phosphoglyc- unfolding at 93 °C (A. fulgidus) or above 98 °C (A. pernix erate, such as GAP:ferredoxin oxidoreductase and non-phos- and P. aerophilum). All archaeal PKs exhibited sigmoi- phorylative glyceraldehyde-3-phosphate dehydrogenase, which dal saturation kinetics with phosphoenolpyruvate replace GAP dehydrogenase and phosphoglycerate kinase in (PEP) and ADP indicating positive homotropic cooper- the conventional EM pathway (2– 6). ative response with both substrates. Classic hetero- An important regulatory principle of the carbon flux in the tropic allosteric regulators of PKs from eukarya and classic EM pathway of eukarya and bacteria is the allosteric bacteria, e.g. fructose 1,6-bisphosphate or AMP, did not regulation of two key enzymes, ATP-dependent PFK and pyru- affect PK activity of hyperthermophilic archaea, sug- gesting the absence of heterotropic allosteric regula- vate kinase. Both enzymes are considered to catalyze irrevers- tion. PK from the bacterium T. maritima is also a ible reactions in vivo and have been shown to be allosterically homotetramer of 50-kDa subunits. The enzyme was in- activated or inhibited by intermediates of metabolism or by the dependent of K ions, had a temperature optimum of energy charge of the cell. To get insights into the role of allo- 80 °C, was highly thermostable up to 90 °C, and had a steric regulation of the modified EM pathways, the PFKs of melting temperature above 98 °C. The enzyme showed various hyperthermophilic archaea have been characterized. It cooperative response to PEP and ADP. In contrast to its was found that all archaeal PFKs (ADP-, ATP-, and pyrophos- archaeal counterparts, the T. maritima enzyme exhib- phate-dependent) were not allosterically regulated by classic ited the classic allosteric response to the activator AMP effectors of ATP-PFKs of eukarya and bacteria, such as ADP and to the inhibitor ATP. Sequences of hyperthermo- and PEP (7–11). Thus, PFKs appear not to be a site of allosteric philic PKs showed significant similarity to character- control in the modified EM pathways of hyperthermophilic ized PKs from bacteria and eukarya. Phylogenetic archaea. In contrast, ATP-PFK from the hyperthermophilic analysis of PK sequences of all three domains indicates bacterium Thermotoga maritima shows the classic response a distinct archaeal cluster that includes the PK from the toward the allosteric effectors, it was activated by ADP and hyperthermophilic bacterium T. maritima. inhibited by PEP (12). Thus, the ATP-PFK of Thermotoga rep- resents a site of allosteric control in the conventional EM path- way operative under hyperthermophilic conditions. To identify Hyperthermophilic prokaryotes, with an optimal growth potential allosteric sites of the modified EM pathways of hy- temperature higher than 80 °C, are considered to represent the perthermophilic archaea, we studied the regulatory properties phylogenetically most ancestral organisms (1). Recent compar- of pyruvate kinases, which catalyze the irreversible conversion ative studies of the hexose degradation pathways in hyperther- * This work was supported by grants of the Deutsche Forschungsge- meinschaft and the Fonds der Chemischen Industrie. The costs of The abbreviations used are: EM, Embden-Meyerhof pathway; GLK, publication of this article were defrayed in part by the payment of page glucokinases; PFK, 6-phosphofructokinases; GAP, glyceraldehyde charges. This article must therefore be hereby marked “advertisement” 3-phosphate; FBP, fructose 1,6-bisphosphate; MES, 4-morpho- in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. lineethanesulfonic acid; ORF, open reading frame; Ni-NTA, nickel- ‡ To whom correspondence should be addressed. Tel.: 49-431-880- nitrilotriacetic acid; CD, circular dichroism; aa, amino acid(s); PEP, 4328; Fax: 49-431-880-2194; E-mail: [email protected]. phosphoenolpyruvate; LDH, L-lactate dehydrogenase. This is an Open Access article under the CC BY license. This paper is available on line at http://www.jbc.org 25417 25418 PKs from Hyperthermophilic Archaea and Bacteria rupted by passing through a French pressure cell at 1.3  10 Pa. Cell of PEP to pyruvate, the terminal reaction of both modified and debris and unbroken cells were removed by centrifugation for 90 min at conventional EM pathways. 100,000  g at 4 °C. PKs are well characterized enzymes from many eukarya and Purification of PK from A. fulgidus—The 100,000  g supernatant bacteria (13–15). The PKs are usually homotetrameric en- was applied to a Q-Sepharose HiLoad column (22  5 cm), which had zymes of about 200 kDa composed of 50-kDa subunits; the been equilibrated with buffer A (50 mM Tris-HCl, pH 9.0, 1 mM dithio- enzymes require divalent cations for activity; many PKs were erythritol). Protein was eluted with a decreasing pH gradient from 9.0 to 7.0 in buffer A and from pH 7.0 to 6.5 in 50 mM bis-Tris-propane, pH shown to be activated by monovalent cations, K or NH . With 6.5, containing 1 mM dithioerythritol. Fractions containing the highest a few exceptions, all PKs from eukarya and bacteria are allo- PK activity were pooled and, after pH exchange (pH 9.0), applied to a sterically regulated either by intermediates of sugar metabo- Uno-Q5 column (5 ml), equilibrated with buffer A. Protein was desorbed lism, usually sugar phosphates, or by adenosine nucleotides with a NaCl gradient from 0 to 0.5 M in buffer A. Fractions containing reflecting the energy charge of the cells. Most eukaryal PKs are the highest PK activity were pooled and concentrated to a volume of 1 allosterically activated by fructose 1,6-bisphosphate (FBP). An ml by ultrafiltration. The concentrated protein solution was applied to a Superdex 200 HiLoad 16/60 gel filtration column equilibrated with 50 unusual allosteric effector, fructose 2,6-bisphosphate, has re- mM MES, pH 6.5, containing 50 mM NaCl and 1 mM dithioerythritol. cently been reported for the PKs of the protozoa Leishmania Eluted PK-containing fractions were pooled and applied to a Uno-S1 mexicana and Trypanosoma brucei (16, 17). Several bacterial column (1 ml), equilibrated with 50 mM acetate, pH 5.3, containing 1 PKs are activated by FBP, but the majority of bacterial PKs mM dithioerythritol. Protein was eluted with a linear NaCl gradient of show allosteric activation by AMP and sugar monophosphates 0to1 M. Pure enzyme was eluted at 0.25 M NaCl. (e.g. ribose 5-phosphate). Few bacteria, e.g. Escherichia coli Purification of PK from T. maritima—The 100,000  g supernatant was applied to a Q-Sepharose HiLoad column (22  5 cm), which had and Salmonella typhimurium contain two PK isoenzymes be- been equilibrated with 50 mM Tris-HCl, pH 7.0, containing 2 mM di- ing activated either by FBP or by AMP (18 –20). Few non- thioerythritol. After washing the column with 50 mM piperazine, pH allosteric PKs have also been described, e.g. M1 isoenzyme of 6.4, containing 2 mM dithioerythritol (buffer B), protein was eluted with vertebrates and the dimeric PKs from Schizosaccharomyces an increasing gradient from 0 to 2 M NaCl in buffer B. Fractions pombe and Zymomonas mobilis (15, 21, 22). More than 140 containing the highest PK activity (0.15– 0.3 M NaCl) were pooled and primary sequences of eukarya and bacteria are known. These applied to an SP-Sepharose column (75 ml) equilibrated with 50 mM MES, pH 5.5, containing 2 mM dithioerythritol. Protein was desorbed include putative PK homologs found in all available archaeal with a pH gradient from pH 5.5 to 7.5 in 50 mM MES. Fractions genomes with the exception of the methanogens Methanopyrus containing the highest PK activity were pooled and concentrated to a kandleri and Methanothermobacter thermoautotrophicus and volume of 1 ml by ultrafiltration (exclusion size, 20 kDa). The protein the hyperthermophilic sulfate reducer Archaeoglobus fulgidus solution was applied to a Superdex 200 HiLoad 16/60 gel filtration VC16 (35). Interestingly, the closely related strain A. fulgidus column equilibrated with 50 mM Tris-HCl, pH 7.0, containing 50 mM 7324 contains high pyruvate kinase activity as part of a mod- NaCl and 2 mM dithioerythritol. The eluted PK activity-containing fractions were pooled and applied to a Uno-S1 column (1 ml) equili- ified EM pathway (3). Crystal structures of PKs are available brated with 50 mM MES, pH 5.5, containing 2 mM dithioerythritol. for the enzymes from cat and rabbit muscle, yeast, and E. coli Protein was eluted with a linear gradient of 0 to 1 M NaCl. Fractions and of non-allosteric M1 PK isoenzymes from vertebrates (24 – containing the highest PK activity were pooled and applied to a Uno-Q1 27). The binding sites for PEP and for the allosteric activator column (1 ml) equilibrated with 50 mM Tris-HCl, pH 8.0, containing 1 FBP from the yeast PK were identified. mM dithioerythritol. Protein was eluted with a gradient of pH 8.0 to 5.3 To date, only two PKs from the domain of archaea have been (50 mM piperazine, 2 mM dithioerythritol). The fraction containing the highest PK activity was eluted at pH 5.5, yielding pure enzyme. biochemically characterized, from the hyperthermophile Ther- Cloning and Functional Overexpression of ORF TM0208 Coding for moproteus tenax and from the moderate thermophile Thermo- PK of T. maritima and Purification of Recombinant Enzyme—ORF plasma acidophilum. Both enzymes are homotetrameric pro- TM0208 was amplified by PCR. The PCR product was cloned into teins. For the T. tenax PK a response to heterotropic allosteric pET19b via two restriction sites (NdeI and BamHI) created with the effectors was not found, whereas the PK from T. acidophilum primers 5-CGGGGTGAACATATGCGAAGTACAAAGAT-3 and 5- has been shown to be activated by AMP (28, 29). ATCTTCATAGGGATCCCCCCTCAATCCA-3. The vector pET19b(pyk- TM0208) was transformed into E. coli BL21 codon plus(DE3)-RIL cells. In this communication we performed a comparative study on For expression, cells were grown in Luria-Bertani medium at 37 °C. The PKs, from three hyperthermophilic archaea, the crenarchaeota expression was started by inducing the promoter with isopropyl-1-thio- Aeropyrum pernix and Pyrobaculum aerophilum as well as the -D-galactopyranoside. After3hof further growth, cells were harvested euryarchaeon A. fulgidus strain 7324 and from the hyperther- by centrifugation. The pellet was suspended in buffer C (20 mM Tris- mophilic bacterium T. maritima. The thermophilic, kinetic, HCl, pH 8.2, containing 0.3 M NaCl and 4 mM imidazole). Cells were and, in particular, the regulatory properties as well as the disrupted through a French pressure cell. After centrifugation (48,000  g,4 °C, 30 min), the supernatant was heat-precipitated at phylogenetic affiliation of the PKs were analyzed. It was found 70 °C for 30 min, followed by an additional centrifugation step that all PKs from hyperthermophilic archaea and bacteria are (100,000  g,4 °C, 60 min). The heat-precipitated supernatant was homotetrameric proteins of extreme thermostability showing applied to a Ni-NTA column (7 ml) equilibrated with buffer C. Protein temperature optima up to 100 °C for catalytic activity. An was eluted with increasing imidazole concentration from 4 to 500 mM in unusual property of PKs from all hyperthermophilic archaea buffer C. Fractions containing the highest enzyme activity were pooled was the absence of regulation by classic heterotropic effectors. and concentrated to a volume of 1 ml by ultrafiltration. The protein solution was applied to a Superdex 200 HiLoad 16/60 gel filtration In contrast, the PK from the hyperthermophilic bacterium column equilibrated with 50 mM Tris-HCl, pH 7.5, containing 150 mM Thermotoga showed the classic response to allosteric effectors. NaCl and 1 mM dithioerythritol. Protein was eluted. Fractions contain- Phylogenetic analysis of PK sequences of all three domains ing the highest PK activity were pooled and applied to a Uno-S1 column indicates a distinct archaeal cluster. (1 ml) equilibrated with 50 mM MES, pH 5.5, containing 1 mM dithio- erythritol. Protein was eluted with a gradient of 0 to 1 M NaCl. The MATERIALS AND METHODS fraction containing the PK activity was recovered at 0.8 M NaCl. At this Growth of A. fulgidus and T. maritima and Preparation of Cell stage PK was essentially pure. Extracts—A. fulgidus strain 7324 and T. maritima were grown anaer- Cloning and Functional Overexpression of ORF APE0489 Coding for obically in the presence of starch as described (3, 30 –32). Cells were PK of A. pernix and Purification of Recombinant Enzyme—ORF harvested in the late exponential growth phase. Cell extracts were APE0489 was amplified and cloned into pET19b via two restriction prepared from 80 g (A. fulgidus) and 60 g (T. maritima) of frozen cells, sites (NdeI and EcoR1) created with the primers 5-TTAGAGAGGCT- which were suspended in 150 ml of 50 mM Tris-HCl, pH 7.0, containing GGCCTCATATGAGGGG-3 and 5-GATAGGAATTCAGACAGGAGC- 10 mM NaCl and 1 mM dithioerythritol and in 90 ml of 50 mM Tris-HCl, GGCTAG-3. The vector pET19b(pyk-APE0489) was transformed into pH 7.0, containing 2 mM dithioerythritol, respectively. Cells were dis- E. coli BL21 codon plus(DE3)-RIL cells. The expression and cell har- PKs from Hyperthermophilic Archaea and Bacteria 25419 vesting was performed as described above. The pellet was resuspended 0.1-mm cuvettes and corrected for the signal of the solvent (20 mM in buffer C. Cells were disrupted by passing through a French pressure sodium phosphate, pH 7.0). Secondary structure analysis and assign- cell. After centrifugation, the supernatant was heat-precipitated at ment to different secondary structure types were performed by the 77 °C for 30 min and centrifuged (100,000  g,4 °C for 60 min) again. experimentally established spectra-structure correlation using the The supernatant was applied to a Ni-NTA column (7 ml) equilibrated Varselec option of Dicroprot (33). Heat-induced unfolding of PKs was with buffer C. Protein was eluted with an imidazole gradient from 4 to analyzed in temperature gradient experiments. The protein samples 500 mM in buffer C. Fractions containing the highest PK activity were were dialyzed against 20 mM sodium phosphate buffer, pH 7.0, and the pooled, incubated at 100 °C for 15 min, and centrifuged. At this stage protein concentrations were set to 100 g/ml. The temperature of the PK was essentially pure. samples was raised at a rate of 1 °C per minute from 50 to 98 °C. Cloning and Functional Overexpression of ORF PAE0819 Coding for Protein unfolding was followed by temperature-dependent change of a PK of P. aerophilum and Purification of Recombinant Enzyme—ORF -helical ellipticity () at 221 nm. The observed ellipticity ( )ata obs PAE0819 was amplified by PCR and was cloned into pET17b via two given temperature was corrected for the temperature-dependent base- restriction sites (NdeI and BamH1) created with the primers 5-CAC- line to give  . The fraction of unfolded protein (X ) was calculated obs uf TAAAGGGCGCGGACATATGAGCGCTC-3 and 5-GTTGGGTACGC- using the temperature-corrected ellipticities of the folded () and un- CAGGATCCTCTTTTACCG-3. The vector pET17b(pyk-PAE0819) was folded ( ) states. Spectra were recorded before and after each tem- uf transformed into E. coli BL21 codon plus(DE3)-RIL cells. Expression perature gradient experiment to characterize the folded and unfolded and cell harvesting were performed as described above. The pellet was states. The fraction of unfolded protein was calculated according to the resuspended in 50 mM Tris-HCl, pH 8.5. Cells were disrupted by pass- equation, X  ( )/(  ). uf obs f uf f ing through a French pressure cell. After centrifugation, the superna- pH Dependence, Cation Specificity, and Effectors—The pH depend- tant was heat-precipitated at 75 °C for 30 min and centrifuged. After ence of the enzymes was measured between 5.5 and 8.2 at 50 °Cinthe buffer exchange by ultrafiltration, the supernatant was applied to a continuous assay using either bis-Tris (pH 5.5– 6.5), bis-Tris-propane Uno-S5 column (5 ml) equilibrated with 50 mM piperazine, pH 5.3. (pH 6.0 –7.5), or Tris-HCl (pH 7.0 – 8.2) (A. fulgidus) and at 65 °C using Protein was eluted with a gradient from 0 to 1 M NaCl. Fractions either piperazine (pH 5.5– 6.1), bis-Tris (pH 6.1– 6.5), or triethanol- containing the highest PK activity (0.3– 0.4 M NaCl) were pooled and amine (pH 6.5–7.5) (A. pernix, P. aerophilum, and T. maritima) (each concentrated to a volume of 1 ml by ultrafiltration. The protein solution 100 mM). Cation specificities were examined using the standard con- was applied to a Superdex 200 HiLoad 16/60 gel filtration column tinuous assay at 50 °C(A. fulgidus)or65 °C(A. pernix, P. aerophi- equilibrated with 50 mM Tris-HCl, pH 7.5, containing 150 mM NaCl, lum,and T. maritima) as described above by replacing MgCl for 2 2 2 2 2 2 and the eluted PK was essentially pure. alternative divalent cations (Mn ,Co ,Ca ,Zn ,Ni ,orFe ) Enzyme Assays and Determination of Kinetic Parameters—The PK at equimolar concentrations (0.1 mM,1mM,and5mM). The depend- activity was determined up to 65 °C using a continuous assay in the ence on K and NH was tested using concentrations up to 100 mM. direction of pyruvate formation. It was ensured that the auxiliary The following classic allosteric effectors of PKs, fructose 1,6- enzyme was not rate-limiting. One unit of enzyme activity is defined as bisphosphate, fructose 2,6-bisphosphate, AMP, L-alanine, ribose 1 mol of product formed per minute. The assay mixture contained for 5-phosphate, glucose 6-phosphate, fructose 6-phosphate, citrate, and A. fulgidus and T. maritima 100 mM triethanolamine, pH 7.0, 1 mM erythrose 4-phosphate (concentration range between 10 M and5mM) PEP, 2 mM ADP, 5 mM MgCl , 0.3 mM NADH, and 1 unit of LDH, and, were tested at 65 °C using the continuous assay as described above for A. pernix and P. aerophilum 100 mM bis-Tris, pH 6.2, 1 mM PEP, 2 with both PEP and ADP concentrations near their S values: A. 0.5 mM ADP, 5 mM MgCl , 0.3 mM NADH, and 1 unit of LDH. The formation fulgidus, 0.4 mM MgCl , 0.4 mM ADP, and 0.2 mM PEP; P. aerophi- of pyruvate from 65 to 98 °C was measured by using a discontinuous lum, 7.5 mM MgCl , 1.5 mM ADP, and 0.5 mM PEP; A. pernix,1 M assay. The standard assay mixture (250 l) contained 100 mM Tris-Cl, MgCl ,0.5mM ADP, and 0.3 mM PEP. In the case of A. pernix PK the 1–5mM PEP, 2 mM ADP, 5–10 mM MgCl . After preincubation, the effectors were preincubated with the protein at the respective tem- reaction was started with an aliquot of PK, incubated for 15–120 s, and perature. The assay for T. maritima PK contained 2.5 mM ADP, 2.5 stopped by rapid addition of 750 l of ice-cold buffer (100 mM Tris-HCl, mM MgCl , and 0.3 mM PEP in 0.1 M MES, pH 6.5 (65 °C). When pH 7.0, 0.6 mM NADH, 0.5 unit of LDH); the amount of pyruvate formed effectors were tested, the substrates ADP and PEP were used at the was quantified by following the oxidation of NADH at 365 nm. Kinetic highest purity available. parameters of PKs were determined at 65 °C using the continuous Sequence Handling—Sequence alignments were constructed with assay (see above). Six to eight different concentrations of the substrates the Neighbor-joining method of ClustalX (34) using the GONNET ma- PEP and ADP were used. The assay mixtures contained 0.3 mM NADH trix. Phylogenetic trees were constructed using both the Neighbor- and 1 unit of LDH for all PKs and, specifically, as follows: A. pernix, 0.1 joining option of ClustalX as well as the Maximum-likelihood method of M bis-Tris, pH 6.2, and 0 –1mM (ADP/2 MgCl ), 0.5 mM PEP or 0 –1mM PROML (Phylip, version 3.6). Confidence limits were estimated by 100 PEP, 0.5 mM ADP, 1 mM MgCl ; A. fulgidus, 0.1 M triethanolamine, pH bootstrapping replicates. 7, and 0 –3mM ADP/MgCl , 0.4 mM PEP or 0 –2mM PEP, 0.2 mM ADP, Sources of Organisms—A. fulgidus strain 7324 (DSM 8774), A. per- 1mM MgCl ; P. aerophilum, 0.1 M bis-Tris, pH 6.2, and 0 –2.5 mM nix (DSM 11879), P. aerophilum (DSM 7523), and T. maritima (DSM ADP/12.5 MgCl ,1mM PEP or 0 –1mM PEP, 1.5 mM ADP, and 7.5 mM 3109) were obtained from the Deutsche Sammlung von Mikroorganis- MgCl ; and T. maritima, 0.1 M MES, pH 6.5, and 0 –3mM ADP/MgCl , men und Zellkulturen (Braunschweig, Germany). 2 2 5mM PEP or 0 –3mM PEP, 2.5 mM ADP/MgCl . Kinetic constants and RESULTS standard errors are obtained from best-fit curves. The data points given in the figures are original measurements of one experiment; the curves Pyruvate Kinases from the Hyperthermophilic Archaea drawn represent fits to a sigmoidal model or a hyperbolic model (Fig. 4, A. pernix, P. aerophilum, and A. fulgidus Strain 7324 Thermotoga PK activity in the presence of AMP) according to non-linear TM TM regression analysis using the Microcal Origin software version 5.0. In the genomes of the hyperthermophilic crenarchaeota A. In the determination of the Hill coefficients, the best-fit lines generated pernix and P. aerophilum ORF APE0489 and PAE0819, respec- via linear regression analysis by the same software are shown (e.g. tively, were annotated as putative pyk genes coding for pyru- Fig. 1B). vate kinase. To prove their coding function, the ORFs were Temperature Dependence and Thermal Stability—The temperature cloned and functionally expressed in E. coli. The recombinant dependence of PK activities was measured between 20 and 98 °C, using proteins were characterized. In the genome of A. fulgidus the discontinuous assay, in 100 mM triethanolamine, pH 7.0 (A. fulgi- dus), or 100 mM bis-Tris, pH 6.2 (A. pernix, P. aerophilum, and strain VC16, no pyk homologous gene was identified. Because T. maritima) each containing 1 mM PEP, 2 mM ADP, and 5 mM MgCl . 2 the closely related strain A. fulgidus 7324 has been shown to Long term thermostability of PKs (0.5 gin30 l of 100 mM trietha- contain high PK activity after growth on starch (3), PK was nolamine, pH 7.0 (A. fulgidus), 1.5 gin30 l of 100 mM bis-Tris, pH 6.2 purified and characterized from this Archaeoglobus strain. (A. pernix), 1.4 gin30 l of 100 mM sodium phosphate buffer, pH 7.0 (P. aerophilum), and 1 gin30 l of 100 mM triethanolamine, pH 7.0 PK from A. fulgidus Strain 7324 (T. maritima), each at the respective temperature) were tested in sealed vials, which were incubated at temperatures between 70 and 110 °Cup Extracts of A. fulgidus grown on starch as carbon and energy to 120 min. The vials were cooled for 10 min, and the remaining activity source contained PK activity (0.13 unit/mg, 50 °C), which is was tested in a continuous assay. about 5-fold higher as compared with PK activity of lactate Circular Dichroism Spectroscopy—CD spectroscopy analyses were performed on a Jasco J-715 CD spectrometer. Spectra were recorded in grown cells (0.02– 0.04 unit/mg) indicating a catabolic function 25420 PKs from Hyperthermophilic Archaea and Bacteria TABLE I Molecular and kinetic properties of purified recombinant pyruvate kinases from A. pernix and P. aerophilum and of purified pyruvate kinase from A. fulgidus Kinetic constants were measured at 65 °C, and standard errors are given (see “Material and Methods”). A. pernix P. aerophilum A. fulgidus Apparent molecular mass of enzyme (kDa) 207  18 205  7 203  14 Apparent molecular mass of subunits (kDa) 51 348 349  3 Oligomeric structure 4 4 4 pH optimum 6.1 6.0 6.6 T (°C) 98 98 93 T (°C) 95 98 85 opt a b c Arrhenius activation energy (kJ/mol) 65 55 56 Apparent V (units/mg) 53 146  1 1000  15 max Apparent S for ADP (mM) 0.26  0.03 1.31  0.04 0.40  0.027 0.5 Hill coefficient (h) 2.10  0.17 2.69  0.25 2.11  0.153 Apparent S for PEP (mM) 0.10  0.03 0.41  0.01 0.25  0.03 0.5 Hill coefficient (h) 1.53  0.13 2.82  0.30 1.80  0.11 a b c Linear part of the Arrhenius plot: 35–95°C, 20 –98°C, 20 – 80°C. of the enzyme during sugar degradation (3). PK was purified kinetics with respect to the substrates PEP and ADP, indicat- from starch-grown cells to homogeneity using four chromato- ing a positive homotropic cooperative response to both sub- graphic steps. The enzyme was purified about 1200-fold to a strates (Fig. 1 and Table I). PK activities of all archaea require specific activity of 1000 units/mg at 80 °C with a yield of 7%. divalent cations and were not dependent on potassium. A. fulgidus PK—The purified enzyme had a specific activity Molecular Composition and N-terminal of 1000 units/mg. The apparent S values for ADP and PEP, 0.5 Amino Acid Sequence calculated from sigmoidal fit, were 0.4 and 0.25 mM, and the corresponding Hill coefficients were 2.1 and 1.8, respectively. The native enzyme had an apparent molecular mass of 203 PK activity required divalent cations. Rate dependence of Mg kDa and showed one 49-kDa band on SDS-PAGE indicating a showed sigmoidal kinetics, indicating cooperative response of homotetrameric ( ) structure of the enzyme (Table I). The PK to this cation. An apparent S value of 0.7 mM and a Hill N-terminal amino acid sequence (20 amino acids, aa) of the 0.5 coefficient of 1.3 were calculated. Alternative divalent cations subunit was determined: MQLPSHKTKIIATIGPASRQ. An were tested at concentrations of 0.1 mM,1mM,and5mM. For alignment of the N-terminal amino acid sequence from A. fulgi- most cations (except Fe )1mM concentration was not inhib- dus PK with putative PKs from hyperthermophilic archaea itory. At 1 mM concentration the highest PK activity was ob- showed the highest degree of identity with hypothetical PK served with Mg (100%  1060 units/mg at 50 °C), which from Thermococcus litoralis (18 aa identical) and Pyrococcus 2 2 could be replaced by Cu (86%) and Mn (63%) and less furiosus (15 aa identical). Surprisingly, using the N-terminal 2 2 2 efficiently by Ni (2%), Ca (7%), and Zn (6.5%). The pH sequence of PK from A. fulgidus strain 7234, no ORF could be optimum was at pH 6.6; 50% of activity was found at pH 5.5 identified in the complete sequenced genome of closely related and 7.5. A. fulgidus VC 16 (35), thus confirming the absence of a pyk P. aerophilum PK—The specific activity of PK was 46 units/ homologous gene in the A. fulgidus type strain. mg. The apparent S values for ADP and PEP were 1.3 mM 0.5 Functional Overexpression of ORF PAE0819 and and 0.4 mM; the calculated Hill coefficients were 2.7 and 2.8, APE0489 Encoding PKs from the Archaea P. aerophilum 2 respectively (Fig. 1). The highest activity was found with Mn and A. pernix and Purification of the Recombinant PKs (100%  200 units/mg at 65 °C) and Co (80%) (each at 1 mM concentration). Remarkably, the enzyme did not show signifi- ORF PAE0819 contains 1386 bp coding for a polypeptide of cant activity (about 1%) with Mg (1 mM). The relative activity 461 amino acids with a calculated molecular mass of 50.3 kDa. with Mg , however, was about 58% as compared with that The ORF was cloned and expressed in E. coli. The PK was with Mn (100%  0.75 unit/mg), when the cations were purified from E. coli by heat treatment and two chromato- tested at 0.1 mM concentration. No activity was observed with graphic steps to a specific activity of 46 units/mg at 65 °C. ORF 2 2 Ca and Zn . The pH optimum was at pH 6. APE0489, annotated as a putative pyk gene in A. pernix, con- A. pernix PK—The specific activity of PK was 53 units/mg. tains 1374 bp coding for a polypeptide of 458 amino acids with The apparent S values for ADP and PEP were 0.26 and 0.1 a calculated molecular mass of 50.5 kDa. The ORF was cloned 0.5 mM, and the calculated Hill coefficients were 2.1 and 1.5, and expressed in E. coli. The His-tagged PK was purified by respectively. PK activity required divalent cations. Highest heat treatment, chromatography on Ni-NTA agarose, and a activities (1 mM cation) were determined with Mg (100%), second heat treatment step at 100 °C to a specific activity of 53 2 2 2 2 Co (170%), and Mn (160%). Ca (14%), Zn (11%), and units/mg at 65 °C. The purified PKs from P. aerophilum and A. 2 2 Ni (14%) were less efficient. With Mg the enzyme showed pernix each showed apparent molecular masses of about 200 cooperative binding and revealed a S of 0.7 mM and a Hill kDa; SDS-PAGE revealed one subunit each with apparent mo- 0.5 coefficient of 1.4. The pH optimum of the enzyme was at lecular masses of 48 and 51 kDa, respectively, indicating a pH 6.1. homotetrameric structure of both archaeal PKs (Table I). Catalytic Properties of PKs from Temperature Optimum and Thermostability of Hyperthermophilic Archaea PKs from Hyperthermophilic Archaea The catalytic, thermophilic, and regulatory properties of the A. fulgidus PK—PK activity showed a temperature optimum PKs from A. fulgidus, P. aerophilum, and A. pernix were ana- at 85 °C. The enzyme was highly thermostable, did not lose lyzed (Table I). All archaeal PKs showed a sigmoidal saturation significant activity upon incubation at 70 °C for 120 min, and PKs from Hyperthermophilic Archaea and Bacteria 25421 P. aerophilum PK—PK activity showed an optimum at higher than 98 °C, the highest possible temperature. The en- zyme showed high stability against heat inactivation with a half-life of 220 min at 100 °C. A. pernix PK—PK activity showed a temperature optimum higher than 95 °C (Fig. 2, A and B). The enzyme showed the highest thermostabilily of all archaeal PKs. The enzyme did not lose activity upon incubation for 120 min at 100 °C. Even at 110 °C the PK showed a half-life of about 30 min (Fig. 2C). Addition of (NH ) SO , NaCl, or KCl (each 1 M) did not stabilize 4 2 4 PK activity against heat inactivation at 110 °C. Thermostability of PKs from Hyperthermophilic Archaea Analyzed by CD Spectroscopy The high stability of PKs against heat inactivation was fur- ther supported by following heat-induced unfolding of the pro- teins up to 98 °C by CD spectroscopy at 221 nm. Unfolding was observed only for PK of A. fulgidus showing a melting temper- ature (T )of93 °C. No unfolding was detected with PKs from P. aerophilum and A. pernix up to temperatures of 98 °C (Fig. 3), indicating melting temperatures higher than 100 °C. This is in accordance to the higher temperature optima for catalytic activity and the thermostabilities of the latter PKs as compared with the A. fulgidus PK. Effect of Allosteric Effectors on PKs from Hyperthermophilic Archaea The effect of classic positive allosteric effectors for PKs of most eukarya and bacteria, such as FBP and AMP, were tested at 65 °C on PK activity from A. fulgidus, P. aerophi- lum, and A. pernix (see “Materials and Methods”). There was no activation effect observed by any of the ligands tested with the PKs studied. ATP has been reported to be an allosteric inhibitor of several PKs from eukarya and bacteria, inhibi- tion being reversed by positive allosteric effectors FBP or AMP. ATP (1 mM) inhibited activities of archaeal PKs, e.g. about 55% in A. fulgidus PK (at 0.3 mM PEP, 0.2 mM ADP). Inhibition could not be reversed by the addition of FBP or AMP (1 mM each). However, ATP-induced inhibition could be reversed up to 90%, by increasing the PEP concentration from 0.3 mM to1mM or by the addition of 1 mM ADP, indicating competitive inhibition. Inhibition of activity by ATP, competitive to ADP and PEP, has also been described for other PKs (36). The apparent absence of allosteric regulation by heterotropic compounds of the PKs from hyperthermophilic archaea might be due to their hyperthermophilic nature and/or due to as yet unknown different regulatory mechanism of the modified EM pathways of archaea. Thus, for comparison we characterized the PK from the hyperthermophilic bacterium T. maritima, which uses the conventional EM pathway for glucose degrada- tion. Both the native enzyme and, for structural and functional analysis, the recombinant PK were analyzed. Pyruvate Kinase from the Hyperthermophilic Bacterium T. maritima Cell extracts of T. maritima grown on starch as carbon and FIG.1. Rate dependence of pyruvate kinase from P. aerophi- energy source contained a 5-fold higher PK activity (0.13 unit/ lum on substrate concentrations. A, PEP saturation curve; B, Hill mg, 50 °C), as compared with PK activity of cells (0.02 unit/mg) plot of the same data; and C, ADP saturation curve. grown on yeast extract indicating the induction of the enzyme during sugar catabolism. PK was purified from starch-grown showed a half-life of about 20 min at 90 °C. At 100 °C an almost cells to homogeneity in five chromatographic steps. The en- complete loss of activity was observed after 7 min. Addition of zyme was purified about 2000-fold to a specific activity of 320 1 M (NH ) SO , rather than NaCl or KCl (1 M each), effectively units/mg at 70 °C with a yield of 7%. The native enzyme had a 4 2 4 stabilized PK against heat inactivation at 100 °C, retaining molecular mass of 194 kDa and was composed of 51-kDa sub- about 50% residual activity after incubation at 120 min. units indicating a homotetrameric structure (Table II). The 25422 PKs from Hyperthermophilic Archaea and Bacteria FIG.3. Thermal induced unfolding of pyruvate kinases from the archaea A. fulgidus, A. pernix, and P. aerophilum and from the bacterium T. maritima measured by CD spectroscopy at 221 nm. Functional Overexpression of TM0208 Encoding PK from T. maritima and Purification of the Enzyme ORF TM0208 contains 1398 bp coding for a polypeptide of 466 amino acids with a calculated molecular mass of 51.9 kDa. The ORF was cloned and expressed in E. coli. The PK was purified by heat treatment in three chromatographic steps. The His-tagged PK showed a molecular mass of 210 kDa and a subunits size of 56 kDa on SDS-PAGE indicating a homotet- rameric structure. Catalytic, Thermophilic, and Regulatory Properties of Native and Recombinant PK from T. maritima The apparent V values (at 65 °C) for pyruvate formation max of the native and the recombinant PK were 320 and 580 units/ mg, respectively. The lower activity of the native enzyme is probably due to the damage during time-consuming purifica- tion procedure (2000-fold purification in five chromatographic steps). Both, native and recombinant PK were almost identical with respect to the following properties. The enzymes showed positively cooperative response to both ADP and PEP with apparent S values of 1.3 and 0.3 mM; the corresponding Hill 0.5 coefficients were 2.9 and 2.1, respectively (Table II and Fig. 4). The pH optimum was near 6.0; 30% of the activity was found at pH 5.5 and 7.0. PK activity required divalent cations; at 1 mM concentration, Mg (100%) could be efficiently replaced by 2 2 2 2 Co (120%) and Mn (35%) rather than by Ca (3%), Zn 2 2 2 (2.5%), Ni (3%), or Fe (1.5%). Mg showed cooperative response to the enzyme with a S of1mM and a Hill coefficient 0.5 of 2.3. PK activity was not dependent on monovalent cations, such as K and NH . Addition of both KCl or NH Cl (40 mM 4 4 each) resulted in a decrease of PK activity (recombinant) by 50 – 60%. Temperature Optimum and Stability PK activity (recombinant) showed an temperature opti- FIG.2. Effect of temperature on the specific activity and ther- mum at 80 °C. Both native and recombinant PK showed high mostability of pyruvate kinase from A. pernix. A, temperature dependence of the specific activity; B, Arrhenius plot of the same data; thermostability up to 85 °C; even at 100 °C the enzyme and C, thermostability at 100 °C() and at 110 °C(f). 100% activity showed a half-life of about 20 min, but an almost complete corresponded to 50 units/mg. loss of activity was observed after 120 min. Addition of (NH ) SO , rather than NaCl or KCl (each 1 M), stabilized PK 4 2 4 N-terminal amino acid sequence of the subunit (MRST- against heat inactivation at 100 °C, retaining about 40% KIVCTVGPRTD) was identical to the deduced N-terminal se- residual activity after 120-min incubation. Thermal unfold- quence of the ORF TM0208, which is annotated as a putative ing of PK, as measured by CD spectroscopy, was not observed pyk gene encoding pyruvate kinase. up to 98 °C (Fig. 3). PKs from Hyperthermophilic Archaea and Bacteria 25423 TABLE II Molecular and kinetic properties of the purified native and recombinant pyruvate kinase from T. maritima Kinetic constants were measured at 65 °C, and standard errors are given (see “Material and Methods”). Native enzyme Recombinant enzyme Apparent molecular mass of enzyme (kDa) 194  14 190  7 Apparent molecular mass of subunits (kDa) 51 356  3 Oligomeric structure 4 4 pH optimum 6.0 5.9 T (°C) ND 98 T (°C) ND 80 opt Arrhenius activation energy (kJ/mol, 30–70°C) 61.4 55.6 Apparent V (units/mg) 320  7 578  10 max ADP saturation Apparent S (mM) 1.35  0.07 1.31  0.17 0.5 Hill coefficient (h) 2.85  0.25 2.81  0.14 PEP saturation Effector Apparent S (mM) 0.26  0.03 0.23  0.06 0.5 Hill coefficient (h) 1.95  0.17 2.20  0.05 AMP Apparent S (mM) 0.09  0.06 0.08  0.05 0.5 Hill coefficient (h) 1.32  0.09 1.00  0.10 ATP Apparent S (mM) ND 0.5  0.04 0.5 Hill coefficient (h) ND 2.9  0.13 ND, not determined. Concentrations of effectors were 1 mM. FIG.5. CD spectra of pyruvate kinases from T. maritima (——) and P. aerophilum (). mained almost constant. Thus, e.g. at a PEP concentration of 0.1 mM, AMP activates PK activity up to 10-fold. Conversely, the addition of ATP resulted in an allosteric inhibition of PK FIG.4. Rate dependence of pyruvate kinase from T. maritima activity by increasing S from 0.23 to 0.5 mM; the Hill 0.5 on PEP concentration in the presence and absence of effector. coefficient increased to 2.9, and V was reduced to 70%. max No effector (f),1mM AMP (), and 1 mM ATP (Œ). Inhibition by ATP was completely reversed by the addition of the activator AMP (1 mM). Thus, in contrast to the PKs from Effect of Allosteric Effectors on PK Activity hyperthermophilic archaea, both AMP and ATP exerted their classic allosteric effects toward the hyperthermophilic PK of The effect of classic allosteric activators, such as AMP and the bacterium T. maritima. FBP, and of the allosteric inhibitor ATP was tested on PK activity at 65 °C. The rate dependence of enzyme activity on CD Spectra of Hyperthermophilic PKs from the Archaeon increasing PEP concentrations in the presence of AMP and of P. aerophilum and the Bacterium T. maritima ATP is shown in Fig. 4. In the absence of effectors, rate dependence of PK (recombinant) showed sigmoidal kinetics To get information about the secondary structure of PKs with an S value of 0.23 and a Hill coefficient of 2.2. Addi- from hyperthermophiles, CD spectra were recorded for PK from 0.5 tion of AMP, rather than of FBP, allosterically activates the Pyrobaculum and Thermotoga. The spectra of both PKs were enzyme: rate dependence on PEP changed from a sigmoidal almost superimposable (Fig. 5). For the PK from P. aerophilum kinetics to a hyperbolic, Michaelis-Menten kinetics, paral- an -helical content of 36% and a -sheet content of 25% were leled by the decrease in S for PEP from 0.23 mM to a K of estimated, which closely match the secondary structure predic- 0.5 m 0.08 mM; the Hill coefficient decreased to 1.0, and V re- tions (36% -helical and 26% -sheet) for both enzymes. The max 25424 PKs from Hyperthermophilic Archaea and Bacteria FIG.6. Multiple sequence alignment of amino acid sequences of pyruvate kinases from eukarya, bacteria, and archaea. The alignment was generated with ClustalX. Conserved residues that have been proposed to be indispensable for catalytic activity as deduced from the yeast x-ray structure (26) are indicated by asterisks. The arrow indicates conserved Glu residue essential for K dependence. The consensus pattern is indicated by a box. The predicted secondary structure of the P. aerophilum pyruvate kinase is shown above the sequences. For accession numbers see Fig. 7. secondary structure estimations were comparable to those de- ity described so far. For example, PK from P. aerophilum rived from the x-ray structures of yeast (38% -helical and 12% (optimal growth temperature, 100 °C) showed a temperature -sheet) and E. coli PK (38% -helical and 21% -sheet). optimum higher than 98 °C and was heat-resistant up to 100 °C for 2 h. In addition, thermal unfolding experiments DISCUSSION revealed extremely high melting temperatures of the PKs Molecular and Thermophilic Properties—The PKs from the near or above 100 °C. For comparison, PK from the extreme hyperthermophiles were characterized as homotetramers of thermophilic bacterium Thermus was completely heat- about 200 kDa composed of 50-kDa subunits, which is a com- inactivated in less than 10 min at 100 °C (37). mon feature of PKs from bacteria and eukarya, and of the two Kinetic and Regulatory Properties—All hyperthermophilic archaeal PKs characterized so far, from the crenarchaeon PKs require divalent cations for activity, a common property of T. tenax and the euryarchaeon T. acidophilum (28, 29). In 2 2 2 all characterized PKs; Mn ,Mg ,orCo being most effec- accordance with the optimal growth temperatures of the re- tive. The PKs from the hyperthermophilic archaea and from spective organism, the hyperthermophilic PKs of our study showed the highest temperature optimum and thermostabil- Thermotoga were not dependent on monovalent cations such as PKs from Hyperthermophilic Archaea and Bacteria 25425 FIG.7. Phylogenetic relationships of pyruvate kinases from bacteria, eukarya, and archaea. The numbers at the nodes are bootstrap- ping values according to neighbor-joining (values on top) and maximum-likelihood (values beneath). NCBI accession numbers or SwissProt identifiers: A. pernix, BAA79454; Bac.lic., Bacillus licheniformis KPYK_BACLI; Bac.ste., Bacillus stearothermophilus S29783; Cor.glu., C. glutamicum KPYK_CORGL; Dei.rad., Deinococcus radiodurans AAF12171; E. coli1, KPY1_ECOLI; E. coli2, KPY2_ECOLI; Cat, KPY1_FELCA; Hae.inf., Haemophilus influenzae KPYK_HAEIN; Human, KPY2_HUMAN; Hyd.the., Hydrogenophilus thermoluteolus BAA95686; Lac.del., Lac- tobacillus delbrueckii KPYK_LACDE; Lac.lac., Lactococcus lactis B40620; Lei.mex., L. mexicana KPYK_LEIME; M. jannaschii, Methanococcus jannaschii D64313; M. acetivorans, Methanosarcina acetivorans str. C2A AAM07241; M. mazei, Methanosarcina mazei Goe1 AAM30411; Rabbit, KPY1_RABIT; P. aerophilum, AAL63053; Pyrococcus abyssi, CAB50316; P. furiosus, AAL81312; P.hor., Pyrococcus horikoshii F71171; yeast 1, KPY1_YEAST; yeast 2, KPY2_YEAST; Sal.typ.1, S. typhimurium LT2 AAL20302; Sal.typ.2, S. typhimurium LT2 AAL20804; Str.the., Streptococ- cus thermophilus AAF25804; Str.coe., Streptomyces coelicolor T35759; Str.coe., A3 CAB70653; Sulfolobus solfataricus, AAK41255; Sulfolobus tokodaii, BAB66695; T. acidophilum, KPYK_THEAC; Thermoplasma volcanium, BAB60191; T. tenax, AAF06820; T. maritima, AAD35300; Try.bru., T. brucei brucei KPY2_TRYBB. K and NH .K -independent PKs have been reported for 1,6-bisphosphate (FBP), other sugar phosphates, or AMP. Also E. coli, Corynebacterium glutamicum, and Z. mobilis and for fructose 2,6-bisphosphate, the allosteric activator of the Leish- the archaeon T. tenax (18, 22, 29, 38). PK sequences contain mania PK (16), did not have an effect. Thus, the hyperthermo- highly conserved K sites (residues 48 –52 and 79 – 88 of the philic archaeal PKs differ from PKs of bacteria and eukarya, Pyrobaculum PK), including a glutamate in K -stimulated which are allosterically activated by these compounds. ATP has PKs, e.g. Glu-89 of the yeast PK (Fig. 6). This glutamate is been described to be an allosteric inhibitor of various eukaryal substituted in potassium-independent PKs. In accordance with and bacterial PKs, with inhibitions being reversed by the acti- the lack of the dependence on potassium, all known hyperther- vator FBP or AMP (39). In contrast, the ATP inhibition of PKs mophilic PKs have substituted glutamate at the equivalent from hyperthermophilic archaea in this study is competitive to position (e.g. by Arg, Lys, or Ser). The PK from the moderate substrates PEP and ADP and could not be reversed by AMP thermophilic archaeon T. acidophilum, which contains this or FBP. glutamate, was described to be dependent on potassium (28). An apparent absence of heterotropic allosteric regulation All hyperthermophilic PKs showed a sigmoidal rate depend- has also been reported for the PK of the hyperthermophilic ence for the substrates PEP and ADP and for Mg , indicating archaeon T. tenax (29). However, the PK from the archaeon positively homotropic cooperative response to substrates and T. acidophilum has been described to be activated by AMP (28), cations. Cooperative substrate binding has also been described indicating that the reduced regulatory capacity is probably not for a few PKs from eukarya and bacteria. However, many PKs, a general feature of all archaeal PKs. In contrast to the hyper- including the PK of the hyperthermophilic archaeon T. tenax thermophilic PKs, the PK from the hyperthermophilic bacte- (29), have been reported to show hyperbolic rate dependence on rium T. maritima showed the classic allosteric response to the ADP, suggesting a different conformational response of these allosteric regulators of bacteria. It was allosterically activated PKs to ADP binding. by AMP and inhibited by ATP. Inhibition of ATP could be An important result of this study is the apparent lack of reversed by AMP. allosteric regulation of the PKs from the hyperthermophilic The reasons for the absence of classic heterotropic regulation archaea by classic heterotropic compounds such as fructose of PKs in hyperthermophilic archaea are not understood. A 25426 PKs from Hyperthermophilic Archaea and Bacteria specific effect of high temperatures can be excluded, because ported by fairly good bootstrapping values. However, the lower the hyperthermophilic PK from Thermotoga showed the classic values of some basal nodes are probably due to the influence of allosteric response. The different regulatory behaviors in hy- several factors: phylogenetic distance, regulation, physiology, perthermophilic archaea and Thermotoga might be due to the evolutionary pressure, and temperature adaptation. The PK bacteria I cluster includes the majority of the bacterial PKs, differences in glycolytic pathways, e.g. in archaea all EM path- ways are modified, whereas in Thermotoga the classic EM and for those enzymes, which were functionally characterized, AMP has been shown to be a positive allosteric effector. Almost pathway is operative. A comparative analysis of e.g. adenine all archaeal PK sequences available form a separate cluster nucleotide pools in hyperthermophilic archaea and Thermotoga supporting the monophyletic origin of the archaeal domain. might give an answer to the different response to AMP. Interestingly, the PK from the hyperthermophilic bacterium Sequence Alignment—PKs from the hyperthermophilic or- T. maritima clusters within the archaeal sequences. This ganisms T. maritima, A. pernix, and P. aerophilum show a high might reflect lateral gene transfer of the pyk gene from an degree of similarity to characterized and putative PKs (37– 65% hyperthermophilic archaeon into the T. maritima genome, a similarity) of eukarya and bacteria. Thus, all PKs constitute a phenomenon that has been suggested to occur in T. maritima homologous family: besides minor deviations at variable posi- at high frequency (41). PKs from the archaeal cluster did not tions, all the hyperthermophilic PKs contain the PK consensus show heterotropic regulation or were regulated by AMP (Ther- pattern [LIVAC]-X-[LIVM](2)-[SAPCV]-K-[LIV]-E-[NKRST]- motoga and Thermoplasma). The PK sequence of the archaeon X-[DEQHS]-[GSTA]-[LIVM] (40). An alignment of these hyper- A. fulgidus strain 7324 is not known. N-terminal amino acid thermophilic PKs with selected homologous proteins from all sequences indicate high identity to PKs from the archaea three domains is given in Fig. 6. Available PK crystal struc- Pyrococcus and Thermococcus. The absence of a pyk homo- tures, from cat and rabbit muscle, yeast, E. coli, and L. mexi- log in the genome of the closely related strain, A. fulgidus cana revealed that each subunit is composed of four domains VC16, might be explained by a loss of this gene; alternatively, (N, A, B, and C) (eukarya) or three domains (A, B, and C) A. fulgidus 7324 might have taken up its pyk gene via lateral (bacteria). The domain N, located at the N terminus, is a short gene transfer from Thermococcales. The putative PK from the -helical stretch present in eukaryotic sequences but absent in Halobacterium NRC I, an outgroup with low bootstrapping the bacterial and archaeal homologs. The domain A, which support, was omitted. All eukaryotic PK sequences form a includes the catalytic site (residues 14 – 83 and 176 –348 of the cluster. Most of them are allosterically activated by F-1,6-BP or P. aerophilum PK, Fig. 6), constitutes a classic () barrel F-2,6-BP (PKs from protistas). However, the absence of allo- structure. The domain B (residues 84 –175) is a -sheet capping steric activation has been reported for a few eukaryal isoen- the catalytic domain. The domain C (residues 349 – 461), lo- zymes. The second bacterial PK cluster comprises PKs from cated at the C terminus, is an open twisted , structure, Gram-positives with low guanine-cytosine content and -pro- containing the FBP binding site of the yeast enzyme. The teobacteria species. Enzymes from this group show allosteric highest degree of homology is found in the catalytic domain A. response to either FBP or to AMP. The two separate bacterial In this domain, a number of residues have been identified to be PK clusters might have evolved by a gene duplication in the important for catalysis, according to the yeast PK structure early bacterial evolution. Alternatively, lateral gene transfers (26). These residues are conserved in all selected PK sequences, from eukaryotes to some Gram-positive and proteobacteria including the PKs of this study. In contrast to the domain A, might be postulated. The latter hypothesis could explain the domain B and in particular C showed significantly lower se- close clustering of the second bacterial group with the eukarya, quence homology. As deduced from the structure of the yeast as well as the allosteric regulation by FBP of isoenzymes 1 from PK, at least eight residues (Ser-402, Thr-403, Ser-404, Thr-407, E. coli and S. thyphimurium, a property found only in eukaryal Trp-452, Arg-459, Gly-475, and His-491 of the yeast PK) have PKs. been identified in the domain C to contact the allosteric effector FBP. However, these residues are not conserved among the Acknowledgments—We thank Dr. J. Gro ¨ tzinger (Kiel) for help in CD spectroscopic measurements, Dr. R. Schmid (Osnabru ¨ ck) for N-termi- FBP-regulated PKs. The presence of a conserved glutamate nal amino acid sequencing, and H. Preidel (Kiel) for mass culturing (Glu-432) has been attributed to the non-allosteric property of A. fulgidus 7324 and T. maritima. the mammalian M1 isoenzyme (26). In the PKs of hyperther- mophilic archaea, no glutamate at the equivalent position was REFERENCES found, indicating that a conserved glutamate is not a prereq- 1. Stetter, K. O. (1996) FEMS Microbiol. Rev. 18, 149 –158 2. Scho ¨ nheit, P., and Scha ¨ fer, T. (1995) World J. Microbiol. Biotechnol. 11, 26 –57 uisite for non-allosteric behavior of PKs. Thus, the absence of 3. Labes, A., and Scho ¨ nheit, P. (2001) Arch. Microbiol. 176, 329 –338 FBP regulation of the hyperthermophilic PKs cannot be ex- 4. Selig, M., Xavier, K. B., Santos, H., and Scho ¨ nheit, P. (1997) Arch. Microbiol. 167, 217–232 plained on sequence level. So far, crystal structures of both 5. De Vos, W. M., Kengen, S. W., Voorhorst, W. G., and Van der Oost, J. (1998) AMP-activated PKs and PKs from hyperthermophilic archaea Extremophiles 2, 201–205 6. Siebers, B., and Hensel, R. (1993) FEMS Microbiol. Lett. 111, 1– 8 have not been reported. Crystallization of the AMP-regulated 7. Kengen, S. W., Tuininga, J. E., Verhees, C. H., van der, O. J., Stams, A. J., and PK from T. maritima and of the archaeal PKs are in progress to de Vos, W. M. (2001) Methods Enzymol. 331, 41–53 identify the AMP binding site and to understand the structural 8. Ronimus, R. S., Koning, J., and Morgan, H. W. (1999) Extremophiles 3, 121–129 basis for reduced regulatory properties of hyperthermophilic 9. Hansen, T., and Scho ¨ nheit, P. (2001) Arch. Microbiol. 177, 62– 69 archaeal PKs. 10. Hansen, T., and Scho ¨ nheit, P. (2000) Arch. Microbiol. 173, 103–109 11. Siebers, B., Klenk, H. P., and Hensel, R. (1998) J. Bacteriol. 180, 2137–2143 Phylogenetic Affiliation—With the detailed characterization 12. Hansen, T., Musfeldt, M., and Scho ¨ nheit, P. (2002) Arch. Microbiol. 177, of the PKs from T. maritima, A. pernix, and P. aerophilum 401– 409 13. Kayne, F. J. (1973) in The Enzymes (Boyer, P. D., ed) pp. 353–382, Academic along with the recently available archaeal homologs, a suffi- Press, New York cient number of PKs from all three domains of life is available 14. Muirhead, H. (1990) Biochem. Soc. Trans. 18, 193–196 allowing phylogenetic studies with respect to both evolutionary 15. Mattevi, A., Bolognesi, M., and Valentini, G. (1996) FEBS Lett. 389, 15–19 16. Rigden, D. J., Phillips, S. E., Michels, P. A., and Fothergill-Gilmore, L. A. as well as functional aspects. As demonstrated in the phylo- (1999) J. Mol. Biol. 291, 615– 635 gram in Fig. 7, selected PK sequences are clustered in four 17. Ernest, I., Callens, M., Uttaro, A. D., Chevalier, N., Opperdoes, F. R., Muirhead, H., and Michels, P. A. (1998) Protein Expr. Purif. 13, 373–382 groups: PK bacteria I, PK archaea, PK eukarya, and PK bac- 18. Waygood, E. B., Rayman, M. K., and Sanwal, B. D. (1975) Can. J. Biochem. 53, teria II. The overall topology of the tree was achieved by both 444 – 454 neighbor-joining and maximum likelihood methods and is sup- 19. Garcia-Olalla, C., and Garrido-Pertierra, A. (1987) Biochem. J 241, 573–581 PKs from Hyperthermophilic Archaea and Bacteria 25427 20. Waygood, E. B., Mort, J. S., and Sanwal, B. D. (1976) Biochemistry 15, 277–282 33. Deleage, G., and Geourjon, C. (1993) Comput. Appl. Biosci. 9, 197–199 21. Nairn, J., Duncan, D., Gray, L. M., Urquhart, G., Binnie, M., Byron, O., 34. Thompson, J. D., Gibson, T. J., Plewniak, F., Jeanmougin, F., and Higgins, Fothergill-Gilmore, L. A., and Price, N. C. (1998) Protein Expr. Purif. 14, D. G. (1997) Nucleic Acids Res. 25, 4876 – 4882 247–253 35. Klenk, H. P., Clayton, R. A., Tomb, J. F., White, O., Nelson, K. E., Ketchum, 22. Pawluk, A., Scopes, R. K., and Griffiths-Smith, K. (1986) Biochem. J. 238, K. A., Dodson, R. J., Gwinn, M., Hickey, E. K., Peterson, J. D., Richardson, 275–281 D. L., Kerlavage, A. R., Graham, D. E., Kyrpides, N. C., Fleischmann, R. D., 23. Bateman, A., Birney, E., Durbin, R., Eddy, S. R., Howe, K. L., and Sonnhammer, Quackenbush, J., Lee, N. H., Sutton, G. G., Gill, S., Kirkness, E. F., E. L. (2001) Nucleic Acids Res. 28, 263–266 Dougherty, B. A., McKenney, K., Adams, M. D., Loftus, B., and Venter, J. C. 24. Muirhead, H., Clayden, D. A., Barford, D., Lorimer, C. G., Fothergill-Gilmore, (1997) Nature 390, 364 –370 L. A., Schiltz, E., and Schmitt, W. (1986) EMBO J. 5, 475– 481 36. Muirhead, H. (1987) in Biological Macromolecules and Assemblies: Active Sites 25. Larsen, T. M., Laughlin, L. T., Holden, H. M., Rayment, I., and Reed, G. H. of Enzymes (Jurnak, F. A., and McPherson, A., eds) pp. 143–186, John Wiley (1994) Biochemistry 33, 6301– 6309 and Sons, New York 26. Jurica, M. S., Mesecar, A., Heath, P. J., Shi, W., Nowak, T., and Stoddard, B. L. 37. Yoshizaki, F., and Imahori, K. (1979) Agric. Biol. Chem. 43, 527–536 (1998) Structure 6, 195–210 38. Jetten, M. S., Gubler, M. E., Lee, S. H., and Sinskey, A. J. (1994) Appl. 27. Mattevi, A., Valentini, G., Rizzi, M., Speranza, M. L., Bolognesi, M., and Coda, Environ. Microbiol. 60, 2501–2507 A. (1995) Structure 3, 729 –741 39. Waygood, E. B., and Sanwal, B. D. (1974) J. Biol. Chem. 249, 265–274 28. Potter, S., and Fothergill-Gilmore, L. A. (1992) FEMS Microbiol. Lett. 73, 40. Hofmann, K., Bucher, P., Falquet, L., and Bairoch, A. (1999) Nucleic Acids Res. 235–239 27, 215–219 29. Schramm, A., Siebers, B., Tjaden, B., Brinkmann, H., and Hensel, R. (2000) J. 41. Nelson, K. E., Clayton, R. A., Gill, S. R., Gwinn, M. L., Dodson, R. J., Haft, Bacteriol. 182, 2001–2009 D. H., Hickey, E. K., Peterson, J. D., Nelson, W. C., Ketchum, K. A., 30. Beeder, J., Nilsen, R. K., Rosnes, J. T., Torsvik, T., and Lien, T. (1994) Appl. McDonald, L., Utterback, T. R., Malek, J. A., Linher, K. D., Garrett, M. M., Environ. Microbiol. 60, 1227–1231 Stewart, A. M., Cotton, M. D., Pratt, M. S., Phillips, C. A., Richardson, D., 31. Huber, R., Langworthy, T. A., Ko ¨ nig, H., Thomm, M., Woese, C. R., Sleytr, U. B., and Stetter, K. O. (1986) Arch. Microbiol. 144, 324 –333 Heidelberg, J., Sutton, G. G., Fleischmann, R. D., Eisen, J. A., and Fraser, 32. Schro ¨ der, C., Selig, M., and Scho ¨ nheit, P. (1994) Arch. Microbiol. 161, 460 – 470 C. M. (1999) Nature 399, 323–329 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Biological Chemistry Unpaywall

Comparative Analysis of Pyruvate Kinases from the Hyperthermophilic Archaea Archaeoglobus fulgidus, Aeropyrum pernix, and Pyrobaculum aerophilum and the Hyperthermophilic Bacterium Thermotoga maritima

Johnsen, Ulrike; Hansen, Thomas; Schönheit, Peter
Journal of Biological Chemistry – Jul 1, 2003
Download PDF
11 pages

Loading next page...
 
/lp/unpaywall/comparative-analysis-of-pyruvate-kinases-from-the-hyperthermophilic-zAQoY6uBn4

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
Unpaywall
ISSN
0021-9258
DOI
10.1074/jbc.m210288200
Publisher site
See Article on Publisher Site

Abstract

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 278, No. 28, Issue of July 11, pp. 25417–25427, 2003 © 2003 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A. Comparative Analysis of Pyruvate Kinases from the Hyperthermophilic Archaea Archaeoglobus fulgidus, Aeropyrum pernix, and Pyrobaculum aerophilum and the Hyperthermophilic Bacterium Thermotoga maritima UNUSUAL REGULATORY PROPERTIES IN HYPERTHERMOPHILIC ARCHAEA* Received for publication, October 8, 2002, and in revised form, March 21, 2003 Published, JBC Papers in Press, March 21, 2003, DOI 10.1074/jbc.M210288200 Ulrike Johnsen, Thomas Hansen, and Peter Scho ¨ nheit‡ From the Institut fu ¨ r Allgemeine Mikrobiologie, Christian-Albrechts-Universita ¨ t Kiel, Am Botanischen Garten 1–9, Kiel D-24118, Germany Pyruvate kinases (PK, EC 2.7.1.40) from three hyper- mophilic archaea and in the hyperthermophilic bacterium thermophilic archaea (Archaeoglobus fulgidus strain Thermotoga revealed that the classic Embden-Meyerhof (EM) 7324, Aeropyrum pernix, and Pyrobaculum aerophilum) pathway is operative only in Thermotoga, whereas in all ar- and from the hyperthermophilic bacterium Thermotoga chaea, the EM pathway exists in modified versions. The mod- maritima were compared with respect to their thermo- ified EM pathways contain, e.g. unusual glucokinases (GLK) philic, kinetic, and regulatory properties. PKs from the and 6-phosphofructokinases (PFK) such as ADP-dependent archaea are 200-kDa homotetramers composed of 50- GLK and ADP-dependent PFK in Pyrococcus, Thermococcus, kDa subunits. The enzymes required divalent cations, and Archaeoglobus; unusual ATP-dependent archaeal GLKs of 2 2 Mg and Mn being most effective, but were independ- the ROK (Regulators, ORFs, Kinases) protein family; non-reg- ent of K . Temperature optima for activity were 85 °C ulatory ATP-dependent PFKs in Desulfurococcus and Aeropy- (A. fulgidus) and above 98 °C (A. pernix and P. aerophi- rum; and pyrophosphate-dependent PFK in Thermoproteus.In lum). The PKs were highly thermostable up to 110 °C (A. addition, the modified EM pathways contain novel enzymes of pernix) and showed melting temperatures for thermal glyceraldehyde 3-phosphate (GAP) oxidation to 3-phosphoglyc- unfolding at 93 °C (A. fulgidus) or above 98 °C (A. pernix erate, such as GAP:ferredoxin oxidoreductase and non-phos- and P. aerophilum). All archaeal PKs exhibited sigmoi- phorylative glyceraldehyde-3-phosphate dehydrogenase, which dal saturation kinetics with phosphoenolpyruvate replace GAP dehydrogenase and phosphoglycerate kinase in (PEP) and ADP indicating positive homotropic cooper- the conventional EM pathway (2– 6). ative response with both substrates. Classic hetero- An important regulatory principle of the carbon flux in the tropic allosteric regulators of PKs from eukarya and classic EM pathway of eukarya and bacteria is the allosteric bacteria, e.g. fructose 1,6-bisphosphate or AMP, did not regulation of two key enzymes, ATP-dependent PFK and pyru- affect PK activity of hyperthermophilic archaea, sug- gesting the absence of heterotropic allosteric regula- vate kinase. Both enzymes are considered to catalyze irrevers- tion. PK from the bacterium T. maritima is also a ible reactions in vivo and have been shown to be allosterically homotetramer of 50-kDa subunits. The enzyme was in- activated or inhibited by intermediates of metabolism or by the dependent of K ions, had a temperature optimum of energy charge of the cell. To get insights into the role of allo- 80 °C, was highly thermostable up to 90 °C, and had a steric regulation of the modified EM pathways, the PFKs of melting temperature above 98 °C. The enzyme showed various hyperthermophilic archaea have been characterized. It cooperative response to PEP and ADP. In contrast to its was found that all archaeal PFKs (ADP-, ATP-, and pyrophos- archaeal counterparts, the T. maritima enzyme exhib- phate-dependent) were not allosterically regulated by classic ited the classic allosteric response to the activator AMP effectors of ATP-PFKs of eukarya and bacteria, such as ADP and to the inhibitor ATP. Sequences of hyperthermo- and PEP (7–11). Thus, PFKs appear not to be a site of allosteric philic PKs showed significant similarity to character- control in the modified EM pathways of hyperthermophilic ized PKs from bacteria and eukarya. Phylogenetic archaea. In contrast, ATP-PFK from the hyperthermophilic analysis of PK sequences of all three domains indicates bacterium Thermotoga maritima shows the classic response a distinct archaeal cluster that includes the PK from the toward the allosteric effectors, it was activated by ADP and hyperthermophilic bacterium T. maritima. inhibited by PEP (12). Thus, the ATP-PFK of Thermotoga rep- resents a site of allosteric control in the conventional EM path- way operative under hyperthermophilic conditions. To identify Hyperthermophilic prokaryotes, with an optimal growth potential allosteric sites of the modified EM pathways of hy- temperature higher than 80 °C, are considered to represent the perthermophilic archaea, we studied the regulatory properties phylogenetically most ancestral organisms (1). Recent compar- of pyruvate kinases, which catalyze the irreversible conversion ative studies of the hexose degradation pathways in hyperther- * This work was supported by grants of the Deutsche Forschungsge- meinschaft and the Fonds der Chemischen Industrie. The costs of The abbreviations used are: EM, Embden-Meyerhof pathway; GLK, publication of this article were defrayed in part by the payment of page glucokinases; PFK, 6-phosphofructokinases; GAP, glyceraldehyde charges. This article must therefore be hereby marked “advertisement” 3-phosphate; FBP, fructose 1,6-bisphosphate; MES, 4-morpho- in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. lineethanesulfonic acid; ORF, open reading frame; Ni-NTA, nickel- ‡ To whom correspondence should be addressed. Tel.: 49-431-880- nitrilotriacetic acid; CD, circular dichroism; aa, amino acid(s); PEP, 4328; Fax: 49-431-880-2194; E-mail: [email protected]. phosphoenolpyruvate; LDH, L-lactate dehydrogenase. This is an Open Access article under the CC BY license. This paper is available on line at http://www.jbc.org 25417 25418 PKs from Hyperthermophilic Archaea and Bacteria rupted by passing through a French pressure cell at 1.3  10 Pa. Cell of PEP to pyruvate, the terminal reaction of both modified and debris and unbroken cells were removed by centrifugation for 90 min at conventional EM pathways. 100,000  g at 4 °C. PKs are well characterized enzymes from many eukarya and Purification of PK from A. fulgidus—The 100,000  g supernatant bacteria (13–15). The PKs are usually homotetrameric en- was applied to a Q-Sepharose HiLoad column (22  5 cm), which had zymes of about 200 kDa composed of 50-kDa subunits; the been equilibrated with buffer A (50 mM Tris-HCl, pH 9.0, 1 mM dithio- enzymes require divalent cations for activity; many PKs were erythritol). Protein was eluted with a decreasing pH gradient from 9.0 to 7.0 in buffer A and from pH 7.0 to 6.5 in 50 mM bis-Tris-propane, pH shown to be activated by monovalent cations, K or NH . With 6.5, containing 1 mM dithioerythritol. Fractions containing the highest a few exceptions, all PKs from eukarya and bacteria are allo- PK activity were pooled and, after pH exchange (pH 9.0), applied to a sterically regulated either by intermediates of sugar metabo- Uno-Q5 column (5 ml), equilibrated with buffer A. Protein was desorbed lism, usually sugar phosphates, or by adenosine nucleotides with a NaCl gradient from 0 to 0.5 M in buffer A. Fractions containing reflecting the energy charge of the cells. Most eukaryal PKs are the highest PK activity were pooled and concentrated to a volume of 1 allosterically activated by fructose 1,6-bisphosphate (FBP). An ml by ultrafiltration. The concentrated protein solution was applied to a Superdex 200 HiLoad 16/60 gel filtration column equilibrated with 50 unusual allosteric effector, fructose 2,6-bisphosphate, has re- mM MES, pH 6.5, containing 50 mM NaCl and 1 mM dithioerythritol. cently been reported for the PKs of the protozoa Leishmania Eluted PK-containing fractions were pooled and applied to a Uno-S1 mexicana and Trypanosoma brucei (16, 17). Several bacterial column (1 ml), equilibrated with 50 mM acetate, pH 5.3, containing 1 PKs are activated by FBP, but the majority of bacterial PKs mM dithioerythritol. Protein was eluted with a linear NaCl gradient of show allosteric activation by AMP and sugar monophosphates 0to1 M. Pure enzyme was eluted at 0.25 M NaCl. (e.g. ribose 5-phosphate). Few bacteria, e.g. Escherichia coli Purification of PK from T. maritima—The 100,000  g supernatant was applied to a Q-Sepharose HiLoad column (22  5 cm), which had and Salmonella typhimurium contain two PK isoenzymes be- been equilibrated with 50 mM Tris-HCl, pH 7.0, containing 2 mM di- ing activated either by FBP or by AMP (18 –20). Few non- thioerythritol. After washing the column with 50 mM piperazine, pH allosteric PKs have also been described, e.g. M1 isoenzyme of 6.4, containing 2 mM dithioerythritol (buffer B), protein was eluted with vertebrates and the dimeric PKs from Schizosaccharomyces an increasing gradient from 0 to 2 M NaCl in buffer B. Fractions pombe and Zymomonas mobilis (15, 21, 22). More than 140 containing the highest PK activity (0.15– 0.3 M NaCl) were pooled and primary sequences of eukarya and bacteria are known. These applied to an SP-Sepharose column (75 ml) equilibrated with 50 mM MES, pH 5.5, containing 2 mM dithioerythritol. Protein was desorbed include putative PK homologs found in all available archaeal with a pH gradient from pH 5.5 to 7.5 in 50 mM MES. Fractions genomes with the exception of the methanogens Methanopyrus containing the highest PK activity were pooled and concentrated to a kandleri and Methanothermobacter thermoautotrophicus and volume of 1 ml by ultrafiltration (exclusion size, 20 kDa). The protein the hyperthermophilic sulfate reducer Archaeoglobus fulgidus solution was applied to a Superdex 200 HiLoad 16/60 gel filtration VC16 (35). Interestingly, the closely related strain A. fulgidus column equilibrated with 50 mM Tris-HCl, pH 7.0, containing 50 mM 7324 contains high pyruvate kinase activity as part of a mod- NaCl and 2 mM dithioerythritol. The eluted PK activity-containing fractions were pooled and applied to a Uno-S1 column (1 ml) equili- ified EM pathway (3). Crystal structures of PKs are available brated with 50 mM MES, pH 5.5, containing 2 mM dithioerythritol. for the enzymes from cat and rabbit muscle, yeast, and E. coli Protein was eluted with a linear gradient of 0 to 1 M NaCl. Fractions and of non-allosteric M1 PK isoenzymes from vertebrates (24 – containing the highest PK activity were pooled and applied to a Uno-Q1 27). The binding sites for PEP and for the allosteric activator column (1 ml) equilibrated with 50 mM Tris-HCl, pH 8.0, containing 1 FBP from the yeast PK were identified. mM dithioerythritol. Protein was eluted with a gradient of pH 8.0 to 5.3 To date, only two PKs from the domain of archaea have been (50 mM piperazine, 2 mM dithioerythritol). The fraction containing the highest PK activity was eluted at pH 5.5, yielding pure enzyme. biochemically characterized, from the hyperthermophile Ther- Cloning and Functional Overexpression of ORF TM0208 Coding for moproteus tenax and from the moderate thermophile Thermo- PK of T. maritima and Purification of Recombinant Enzyme—ORF plasma acidophilum. Both enzymes are homotetrameric pro- TM0208 was amplified by PCR. The PCR product was cloned into teins. For the T. tenax PK a response to heterotropic allosteric pET19b via two restriction sites (NdeI and BamHI) created with the effectors was not found, whereas the PK from T. acidophilum primers 5-CGGGGTGAACATATGCGAAGTACAAAGAT-3 and 5- has been shown to be activated by AMP (28, 29). ATCTTCATAGGGATCCCCCCTCAATCCA-3. The vector pET19b(pyk- TM0208) was transformed into E. coli BL21 codon plus(DE3)-RIL cells. In this communication we performed a comparative study on For expression, cells were grown in Luria-Bertani medium at 37 °C. The PKs, from three hyperthermophilic archaea, the crenarchaeota expression was started by inducing the promoter with isopropyl-1-thio- Aeropyrum pernix and Pyrobaculum aerophilum as well as the -D-galactopyranoside. After3hof further growth, cells were harvested euryarchaeon A. fulgidus strain 7324 and from the hyperther- by centrifugation. The pellet was suspended in buffer C (20 mM Tris- mophilic bacterium T. maritima. The thermophilic, kinetic, HCl, pH 8.2, containing 0.3 M NaCl and 4 mM imidazole). Cells were and, in particular, the regulatory properties as well as the disrupted through a French pressure cell. After centrifugation (48,000  g,4 °C, 30 min), the supernatant was heat-precipitated at phylogenetic affiliation of the PKs were analyzed. It was found 70 °C for 30 min, followed by an additional centrifugation step that all PKs from hyperthermophilic archaea and bacteria are (100,000  g,4 °C, 60 min). The heat-precipitated supernatant was homotetrameric proteins of extreme thermostability showing applied to a Ni-NTA column (7 ml) equilibrated with buffer C. Protein temperature optima up to 100 °C for catalytic activity. An was eluted with increasing imidazole concentration from 4 to 500 mM in unusual property of PKs from all hyperthermophilic archaea buffer C. Fractions containing the highest enzyme activity were pooled was the absence of regulation by classic heterotropic effectors. and concentrated to a volume of 1 ml by ultrafiltration. The protein solution was applied to a Superdex 200 HiLoad 16/60 gel filtration In contrast, the PK from the hyperthermophilic bacterium column equilibrated with 50 mM Tris-HCl, pH 7.5, containing 150 mM Thermotoga showed the classic response to allosteric effectors. NaCl and 1 mM dithioerythritol. Protein was eluted. Fractions contain- Phylogenetic analysis of PK sequences of all three domains ing the highest PK activity were pooled and applied to a Uno-S1 column indicates a distinct archaeal cluster. (1 ml) equilibrated with 50 mM MES, pH 5.5, containing 1 mM dithio- erythritol. Protein was eluted with a gradient of 0 to 1 M NaCl. The MATERIALS AND METHODS fraction containing the PK activity was recovered at 0.8 M NaCl. At this Growth of A. fulgidus and T. maritima and Preparation of Cell stage PK was essentially pure. Extracts—A. fulgidus strain 7324 and T. maritima were grown anaer- Cloning and Functional Overexpression of ORF APE0489 Coding for obically in the presence of starch as described (3, 30 –32). Cells were PK of A. pernix and Purification of Recombinant Enzyme—ORF harvested in the late exponential growth phase. Cell extracts were APE0489 was amplified and cloned into pET19b via two restriction prepared from 80 g (A. fulgidus) and 60 g (T. maritima) of frozen cells, sites (NdeI and EcoR1) created with the primers 5-TTAGAGAGGCT- which were suspended in 150 ml of 50 mM Tris-HCl, pH 7.0, containing GGCCTCATATGAGGGG-3 and 5-GATAGGAATTCAGACAGGAGC- 10 mM NaCl and 1 mM dithioerythritol and in 90 ml of 50 mM Tris-HCl, GGCTAG-3. The vector pET19b(pyk-APE0489) was transformed into pH 7.0, containing 2 mM dithioerythritol, respectively. Cells were dis- E. coli BL21 codon plus(DE3)-RIL cells. The expression and cell har- PKs from Hyperthermophilic Archaea and Bacteria 25419 vesting was performed as described above. The pellet was resuspended 0.1-mm cuvettes and corrected for the signal of the solvent (20 mM in buffer C. Cells were disrupted by passing through a French pressure sodium phosphate, pH 7.0). Secondary structure analysis and assign- cell. After centrifugation, the supernatant was heat-precipitated at ment to different secondary structure types were performed by the 77 °C for 30 min and centrifuged (100,000  g,4 °C for 60 min) again. experimentally established spectra-structure correlation using the The supernatant was applied to a Ni-NTA column (7 ml) equilibrated Varselec option of Dicroprot (33). Heat-induced unfolding of PKs was with buffer C. Protein was eluted with an imidazole gradient from 4 to analyzed in temperature gradient experiments. The protein samples 500 mM in buffer C. Fractions containing the highest PK activity were were dialyzed against 20 mM sodium phosphate buffer, pH 7.0, and the pooled, incubated at 100 °C for 15 min, and centrifuged. At this stage protein concentrations were set to 100 g/ml. The temperature of the PK was essentially pure. samples was raised at a rate of 1 °C per minute from 50 to 98 °C. Cloning and Functional Overexpression of ORF PAE0819 Coding for Protein unfolding was followed by temperature-dependent change of a PK of P. aerophilum and Purification of Recombinant Enzyme—ORF -helical ellipticity () at 221 nm. The observed ellipticity ( )ata obs PAE0819 was amplified by PCR and was cloned into pET17b via two given temperature was corrected for the temperature-dependent base- restriction sites (NdeI and BamH1) created with the primers 5-CAC- line to give  . The fraction of unfolded protein (X ) was calculated obs uf TAAAGGGCGCGGACATATGAGCGCTC-3 and 5-GTTGGGTACGC- using the temperature-corrected ellipticities of the folded () and un- CAGGATCCTCTTTTACCG-3. The vector pET17b(pyk-PAE0819) was folded ( ) states. Spectra were recorded before and after each tem- uf transformed into E. coli BL21 codon plus(DE3)-RIL cells. Expression perature gradient experiment to characterize the folded and unfolded and cell harvesting were performed as described above. The pellet was states. The fraction of unfolded protein was calculated according to the resuspended in 50 mM Tris-HCl, pH 8.5. Cells were disrupted by pass- equation, X  ( )/(  ). uf obs f uf f ing through a French pressure cell. After centrifugation, the superna- pH Dependence, Cation Specificity, and Effectors—The pH depend- tant was heat-precipitated at 75 °C for 30 min and centrifuged. After ence of the enzymes was measured between 5.5 and 8.2 at 50 °Cinthe buffer exchange by ultrafiltration, the supernatant was applied to a continuous assay using either bis-Tris (pH 5.5– 6.5), bis-Tris-propane Uno-S5 column (5 ml) equilibrated with 50 mM piperazine, pH 5.3. (pH 6.0 –7.5), or Tris-HCl (pH 7.0 – 8.2) (A. fulgidus) and at 65 °C using Protein was eluted with a gradient from 0 to 1 M NaCl. Fractions either piperazine (pH 5.5– 6.1), bis-Tris (pH 6.1– 6.5), or triethanol- containing the highest PK activity (0.3– 0.4 M NaCl) were pooled and amine (pH 6.5–7.5) (A. pernix, P. aerophilum, and T. maritima) (each concentrated to a volume of 1 ml by ultrafiltration. The protein solution 100 mM). Cation specificities were examined using the standard con- was applied to a Superdex 200 HiLoad 16/60 gel filtration column tinuous assay at 50 °C(A. fulgidus)or65 °C(A. pernix, P. aerophi- equilibrated with 50 mM Tris-HCl, pH 7.5, containing 150 mM NaCl, lum,and T. maritima) as described above by replacing MgCl for 2 2 2 2 2 2 and the eluted PK was essentially pure. alternative divalent cations (Mn ,Co ,Ca ,Zn ,Ni ,orFe ) Enzyme Assays and Determination of Kinetic Parameters—The PK at equimolar concentrations (0.1 mM,1mM,and5mM). The depend- activity was determined up to 65 °C using a continuous assay in the ence on K and NH was tested using concentrations up to 100 mM. direction of pyruvate formation. It was ensured that the auxiliary The following classic allosteric effectors of PKs, fructose 1,6- enzyme was not rate-limiting. One unit of enzyme activity is defined as bisphosphate, fructose 2,6-bisphosphate, AMP, L-alanine, ribose 1 mol of product formed per minute. The assay mixture contained for 5-phosphate, glucose 6-phosphate, fructose 6-phosphate, citrate, and A. fulgidus and T. maritima 100 mM triethanolamine, pH 7.0, 1 mM erythrose 4-phosphate (concentration range between 10 M and5mM) PEP, 2 mM ADP, 5 mM MgCl , 0.3 mM NADH, and 1 unit of LDH, and, were tested at 65 °C using the continuous assay as described above for A. pernix and P. aerophilum 100 mM bis-Tris, pH 6.2, 1 mM PEP, 2 with both PEP and ADP concentrations near their S values: A. 0.5 mM ADP, 5 mM MgCl , 0.3 mM NADH, and 1 unit of LDH. The formation fulgidus, 0.4 mM MgCl , 0.4 mM ADP, and 0.2 mM PEP; P. aerophi- of pyruvate from 65 to 98 °C was measured by using a discontinuous lum, 7.5 mM MgCl , 1.5 mM ADP, and 0.5 mM PEP; A. pernix,1 M assay. The standard assay mixture (250 l) contained 100 mM Tris-Cl, MgCl ,0.5mM ADP, and 0.3 mM PEP. In the case of A. pernix PK the 1–5mM PEP, 2 mM ADP, 5–10 mM MgCl . After preincubation, the effectors were preincubated with the protein at the respective tem- reaction was started with an aliquot of PK, incubated for 15–120 s, and perature. The assay for T. maritima PK contained 2.5 mM ADP, 2.5 stopped by rapid addition of 750 l of ice-cold buffer (100 mM Tris-HCl, mM MgCl , and 0.3 mM PEP in 0.1 M MES, pH 6.5 (65 °C). When pH 7.0, 0.6 mM NADH, 0.5 unit of LDH); the amount of pyruvate formed effectors were tested, the substrates ADP and PEP were used at the was quantified by following the oxidation of NADH at 365 nm. Kinetic highest purity available. parameters of PKs were determined at 65 °C using the continuous Sequence Handling—Sequence alignments were constructed with assay (see above). Six to eight different concentrations of the substrates the Neighbor-joining method of ClustalX (34) using the GONNET ma- PEP and ADP were used. The assay mixtures contained 0.3 mM NADH trix. Phylogenetic trees were constructed using both the Neighbor- and 1 unit of LDH for all PKs and, specifically, as follows: A. pernix, 0.1 joining option of ClustalX as well as the Maximum-likelihood method of M bis-Tris, pH 6.2, and 0 –1mM (ADP/2 MgCl ), 0.5 mM PEP or 0 –1mM PROML (Phylip, version 3.6). Confidence limits were estimated by 100 PEP, 0.5 mM ADP, 1 mM MgCl ; A. fulgidus, 0.1 M triethanolamine, pH bootstrapping replicates. 7, and 0 –3mM ADP/MgCl , 0.4 mM PEP or 0 –2mM PEP, 0.2 mM ADP, Sources of Organisms—A. fulgidus strain 7324 (DSM 8774), A. per- 1mM MgCl ; P. aerophilum, 0.1 M bis-Tris, pH 6.2, and 0 –2.5 mM nix (DSM 11879), P. aerophilum (DSM 7523), and T. maritima (DSM ADP/12.5 MgCl ,1mM PEP or 0 –1mM PEP, 1.5 mM ADP, and 7.5 mM 3109) were obtained from the Deutsche Sammlung von Mikroorganis- MgCl ; and T. maritima, 0.1 M MES, pH 6.5, and 0 –3mM ADP/MgCl , men und Zellkulturen (Braunschweig, Germany). 2 2 5mM PEP or 0 –3mM PEP, 2.5 mM ADP/MgCl . Kinetic constants and RESULTS standard errors are obtained from best-fit curves. The data points given in the figures are original measurements of one experiment; the curves Pyruvate Kinases from the Hyperthermophilic Archaea drawn represent fits to a sigmoidal model or a hyperbolic model (Fig. 4, A. pernix, P. aerophilum, and A. fulgidus Strain 7324 Thermotoga PK activity in the presence of AMP) according to non-linear TM TM regression analysis using the Microcal Origin software version 5.0. In the genomes of the hyperthermophilic crenarchaeota A. In the determination of the Hill coefficients, the best-fit lines generated pernix and P. aerophilum ORF APE0489 and PAE0819, respec- via linear regression analysis by the same software are shown (e.g. tively, were annotated as putative pyk genes coding for pyru- Fig. 1B). vate kinase. To prove their coding function, the ORFs were Temperature Dependence and Thermal Stability—The temperature cloned and functionally expressed in E. coli. The recombinant dependence of PK activities was measured between 20 and 98 °C, using proteins were characterized. In the genome of A. fulgidus the discontinuous assay, in 100 mM triethanolamine, pH 7.0 (A. fulgi- dus), or 100 mM bis-Tris, pH 6.2 (A. pernix, P. aerophilum, and strain VC16, no pyk homologous gene was identified. Because T. maritima) each containing 1 mM PEP, 2 mM ADP, and 5 mM MgCl . 2 the closely related strain A. fulgidus 7324 has been shown to Long term thermostability of PKs (0.5 gin30 l of 100 mM trietha- contain high PK activity after growth on starch (3), PK was nolamine, pH 7.0 (A. fulgidus), 1.5 gin30 l of 100 mM bis-Tris, pH 6.2 purified and characterized from this Archaeoglobus strain. (A. pernix), 1.4 gin30 l of 100 mM sodium phosphate buffer, pH 7.0 (P. aerophilum), and 1 gin30 l of 100 mM triethanolamine, pH 7.0 PK from A. fulgidus Strain 7324 (T. maritima), each at the respective temperature) were tested in sealed vials, which were incubated at temperatures between 70 and 110 °Cup Extracts of A. fulgidus grown on starch as carbon and energy to 120 min. The vials were cooled for 10 min, and the remaining activity source contained PK activity (0.13 unit/mg, 50 °C), which is was tested in a continuous assay. about 5-fold higher as compared with PK activity of lactate Circular Dichroism Spectroscopy—CD spectroscopy analyses were performed on a Jasco J-715 CD spectrometer. Spectra were recorded in grown cells (0.02– 0.04 unit/mg) indicating a catabolic function 25420 PKs from Hyperthermophilic Archaea and Bacteria TABLE I Molecular and kinetic properties of purified recombinant pyruvate kinases from A. pernix and P. aerophilum and of purified pyruvate kinase from A. fulgidus Kinetic constants were measured at 65 °C, and standard errors are given (see “Material and Methods”). A. pernix P. aerophilum A. fulgidus Apparent molecular mass of enzyme (kDa) 207  18 205  7 203  14 Apparent molecular mass of subunits (kDa) 51 348 349  3 Oligomeric structure 4 4 4 pH optimum 6.1 6.0 6.6 T (°C) 98 98 93 T (°C) 95 98 85 opt a b c Arrhenius activation energy (kJ/mol) 65 55 56 Apparent V (units/mg) 53 146  1 1000  15 max Apparent S for ADP (mM) 0.26  0.03 1.31  0.04 0.40  0.027 0.5 Hill coefficient (h) 2.10  0.17 2.69  0.25 2.11  0.153 Apparent S for PEP (mM) 0.10  0.03 0.41  0.01 0.25  0.03 0.5 Hill coefficient (h) 1.53  0.13 2.82  0.30 1.80  0.11 a b c Linear part of the Arrhenius plot: 35–95°C, 20 –98°C, 20 – 80°C. of the enzyme during sugar degradation (3). PK was purified kinetics with respect to the substrates PEP and ADP, indicat- from starch-grown cells to homogeneity using four chromato- ing a positive homotropic cooperative response to both sub- graphic steps. The enzyme was purified about 1200-fold to a strates (Fig. 1 and Table I). PK activities of all archaea require specific activity of 1000 units/mg at 80 °C with a yield of 7%. divalent cations and were not dependent on potassium. A. fulgidus PK—The purified enzyme had a specific activity Molecular Composition and N-terminal of 1000 units/mg. The apparent S values for ADP and PEP, 0.5 Amino Acid Sequence calculated from sigmoidal fit, were 0.4 and 0.25 mM, and the corresponding Hill coefficients were 2.1 and 1.8, respectively. The native enzyme had an apparent molecular mass of 203 PK activity required divalent cations. Rate dependence of Mg kDa and showed one 49-kDa band on SDS-PAGE indicating a showed sigmoidal kinetics, indicating cooperative response of homotetrameric ( ) structure of the enzyme (Table I). The PK to this cation. An apparent S value of 0.7 mM and a Hill N-terminal amino acid sequence (20 amino acids, aa) of the 0.5 coefficient of 1.3 were calculated. Alternative divalent cations subunit was determined: MQLPSHKTKIIATIGPASRQ. An were tested at concentrations of 0.1 mM,1mM,and5mM. For alignment of the N-terminal amino acid sequence from A. fulgi- most cations (except Fe )1mM concentration was not inhib- dus PK with putative PKs from hyperthermophilic archaea itory. At 1 mM concentration the highest PK activity was ob- showed the highest degree of identity with hypothetical PK served with Mg (100%  1060 units/mg at 50 °C), which from Thermococcus litoralis (18 aa identical) and Pyrococcus 2 2 could be replaced by Cu (86%) and Mn (63%) and less furiosus (15 aa identical). Surprisingly, using the N-terminal 2 2 2 efficiently by Ni (2%), Ca (7%), and Zn (6.5%). The pH sequence of PK from A. fulgidus strain 7234, no ORF could be optimum was at pH 6.6; 50% of activity was found at pH 5.5 identified in the complete sequenced genome of closely related and 7.5. A. fulgidus VC 16 (35), thus confirming the absence of a pyk P. aerophilum PK—The specific activity of PK was 46 units/ homologous gene in the A. fulgidus type strain. mg. The apparent S values for ADP and PEP were 1.3 mM 0.5 Functional Overexpression of ORF PAE0819 and and 0.4 mM; the calculated Hill coefficients were 2.7 and 2.8, APE0489 Encoding PKs from the Archaea P. aerophilum 2 respectively (Fig. 1). The highest activity was found with Mn and A. pernix and Purification of the Recombinant PKs (100%  200 units/mg at 65 °C) and Co (80%) (each at 1 mM concentration). Remarkably, the enzyme did not show signifi- ORF PAE0819 contains 1386 bp coding for a polypeptide of cant activity (about 1%) with Mg (1 mM). The relative activity 461 amino acids with a calculated molecular mass of 50.3 kDa. with Mg , however, was about 58% as compared with that The ORF was cloned and expressed in E. coli. The PK was with Mn (100%  0.75 unit/mg), when the cations were purified from E. coli by heat treatment and two chromato- tested at 0.1 mM concentration. No activity was observed with graphic steps to a specific activity of 46 units/mg at 65 °C. ORF 2 2 Ca and Zn . The pH optimum was at pH 6. APE0489, annotated as a putative pyk gene in A. pernix, con- A. pernix PK—The specific activity of PK was 53 units/mg. tains 1374 bp coding for a polypeptide of 458 amino acids with The apparent S values for ADP and PEP were 0.26 and 0.1 a calculated molecular mass of 50.5 kDa. The ORF was cloned 0.5 mM, and the calculated Hill coefficients were 2.1 and 1.5, and expressed in E. coli. The His-tagged PK was purified by respectively. PK activity required divalent cations. Highest heat treatment, chromatography on Ni-NTA agarose, and a activities (1 mM cation) were determined with Mg (100%), second heat treatment step at 100 °C to a specific activity of 53 2 2 2 2 Co (170%), and Mn (160%). Ca (14%), Zn (11%), and units/mg at 65 °C. The purified PKs from P. aerophilum and A. 2 2 Ni (14%) were less efficient. With Mg the enzyme showed pernix each showed apparent molecular masses of about 200 cooperative binding and revealed a S of 0.7 mM and a Hill kDa; SDS-PAGE revealed one subunit each with apparent mo- 0.5 coefficient of 1.4. The pH optimum of the enzyme was at lecular masses of 48 and 51 kDa, respectively, indicating a pH 6.1. homotetrameric structure of both archaeal PKs (Table I). Catalytic Properties of PKs from Temperature Optimum and Thermostability of Hyperthermophilic Archaea PKs from Hyperthermophilic Archaea The catalytic, thermophilic, and regulatory properties of the A. fulgidus PK—PK activity showed a temperature optimum PKs from A. fulgidus, P. aerophilum, and A. pernix were ana- at 85 °C. The enzyme was highly thermostable, did not lose lyzed (Table I). All archaeal PKs showed a sigmoidal saturation significant activity upon incubation at 70 °C for 120 min, and PKs from Hyperthermophilic Archaea and Bacteria 25421 P. aerophilum PK—PK activity showed an optimum at higher than 98 °C, the highest possible temperature. The en- zyme showed high stability against heat inactivation with a half-life of 220 min at 100 °C. A. pernix PK—PK activity showed a temperature optimum higher than 95 °C (Fig. 2, A and B). The enzyme showed the highest thermostabilily of all archaeal PKs. The enzyme did not lose activity upon incubation for 120 min at 100 °C. Even at 110 °C the PK showed a half-life of about 30 min (Fig. 2C). Addition of (NH ) SO , NaCl, or KCl (each 1 M) did not stabilize 4 2 4 PK activity against heat inactivation at 110 °C. Thermostability of PKs from Hyperthermophilic Archaea Analyzed by CD Spectroscopy The high stability of PKs against heat inactivation was fur- ther supported by following heat-induced unfolding of the pro- teins up to 98 °C by CD spectroscopy at 221 nm. Unfolding was observed only for PK of A. fulgidus showing a melting temper- ature (T )of93 °C. No unfolding was detected with PKs from P. aerophilum and A. pernix up to temperatures of 98 °C (Fig. 3), indicating melting temperatures higher than 100 °C. This is in accordance to the higher temperature optima for catalytic activity and the thermostabilities of the latter PKs as compared with the A. fulgidus PK. Effect of Allosteric Effectors on PKs from Hyperthermophilic Archaea The effect of classic positive allosteric effectors for PKs of most eukarya and bacteria, such as FBP and AMP, were tested at 65 °C on PK activity from A. fulgidus, P. aerophi- lum, and A. pernix (see “Materials and Methods”). There was no activation effect observed by any of the ligands tested with the PKs studied. ATP has been reported to be an allosteric inhibitor of several PKs from eukarya and bacteria, inhibi- tion being reversed by positive allosteric effectors FBP or AMP. ATP (1 mM) inhibited activities of archaeal PKs, e.g. about 55% in A. fulgidus PK (at 0.3 mM PEP, 0.2 mM ADP). Inhibition could not be reversed by the addition of FBP or AMP (1 mM each). However, ATP-induced inhibition could be reversed up to 90%, by increasing the PEP concentration from 0.3 mM to1mM or by the addition of 1 mM ADP, indicating competitive inhibition. Inhibition of activity by ATP, competitive to ADP and PEP, has also been described for other PKs (36). The apparent absence of allosteric regulation by heterotropic compounds of the PKs from hyperthermophilic archaea might be due to their hyperthermophilic nature and/or due to as yet unknown different regulatory mechanism of the modified EM pathways of archaea. Thus, for comparison we characterized the PK from the hyperthermophilic bacterium T. maritima, which uses the conventional EM pathway for glucose degrada- tion. Both the native enzyme and, for structural and functional analysis, the recombinant PK were analyzed. Pyruvate Kinase from the Hyperthermophilic Bacterium T. maritima Cell extracts of T. maritima grown on starch as carbon and FIG.1. Rate dependence of pyruvate kinase from P. aerophi- energy source contained a 5-fold higher PK activity (0.13 unit/ lum on substrate concentrations. A, PEP saturation curve; B, Hill mg, 50 °C), as compared with PK activity of cells (0.02 unit/mg) plot of the same data; and C, ADP saturation curve. grown on yeast extract indicating the induction of the enzyme during sugar catabolism. PK was purified from starch-grown showed a half-life of about 20 min at 90 °C. At 100 °C an almost cells to homogeneity in five chromatographic steps. The en- complete loss of activity was observed after 7 min. Addition of zyme was purified about 2000-fold to a specific activity of 320 1 M (NH ) SO , rather than NaCl or KCl (1 M each), effectively units/mg at 70 °C with a yield of 7%. The native enzyme had a 4 2 4 stabilized PK against heat inactivation at 100 °C, retaining molecular mass of 194 kDa and was composed of 51-kDa sub- about 50% residual activity after incubation at 120 min. units indicating a homotetrameric structure (Table II). The 25422 PKs from Hyperthermophilic Archaea and Bacteria FIG.3. Thermal induced unfolding of pyruvate kinases from the archaea A. fulgidus, A. pernix, and P. aerophilum and from the bacterium T. maritima measured by CD spectroscopy at 221 nm. Functional Overexpression of TM0208 Encoding PK from T. maritima and Purification of the Enzyme ORF TM0208 contains 1398 bp coding for a polypeptide of 466 amino acids with a calculated molecular mass of 51.9 kDa. The ORF was cloned and expressed in E. coli. The PK was purified by heat treatment in three chromatographic steps. The His-tagged PK showed a molecular mass of 210 kDa and a subunits size of 56 kDa on SDS-PAGE indicating a homotet- rameric structure. Catalytic, Thermophilic, and Regulatory Properties of Native and Recombinant PK from T. maritima The apparent V values (at 65 °C) for pyruvate formation max of the native and the recombinant PK were 320 and 580 units/ mg, respectively. The lower activity of the native enzyme is probably due to the damage during time-consuming purifica- tion procedure (2000-fold purification in five chromatographic steps). Both, native and recombinant PK were almost identical with respect to the following properties. The enzymes showed positively cooperative response to both ADP and PEP with apparent S values of 1.3 and 0.3 mM; the corresponding Hill 0.5 coefficients were 2.9 and 2.1, respectively (Table II and Fig. 4). The pH optimum was near 6.0; 30% of the activity was found at pH 5.5 and 7.0. PK activity required divalent cations; at 1 mM concentration, Mg (100%) could be efficiently replaced by 2 2 2 2 Co (120%) and Mn (35%) rather than by Ca (3%), Zn 2 2 2 (2.5%), Ni (3%), or Fe (1.5%). Mg showed cooperative response to the enzyme with a S of1mM and a Hill coefficient 0.5 of 2.3. PK activity was not dependent on monovalent cations, such as K and NH . Addition of both KCl or NH Cl (40 mM 4 4 each) resulted in a decrease of PK activity (recombinant) by 50 – 60%. Temperature Optimum and Stability PK activity (recombinant) showed an temperature opti- FIG.2. Effect of temperature on the specific activity and ther- mum at 80 °C. Both native and recombinant PK showed high mostability of pyruvate kinase from A. pernix. A, temperature dependence of the specific activity; B, Arrhenius plot of the same data; thermostability up to 85 °C; even at 100 °C the enzyme and C, thermostability at 100 °C() and at 110 °C(f). 100% activity showed a half-life of about 20 min, but an almost complete corresponded to 50 units/mg. loss of activity was observed after 120 min. Addition of (NH ) SO , rather than NaCl or KCl (each 1 M), stabilized PK 4 2 4 N-terminal amino acid sequence of the subunit (MRST- against heat inactivation at 100 °C, retaining about 40% KIVCTVGPRTD) was identical to the deduced N-terminal se- residual activity after 120-min incubation. Thermal unfold- quence of the ORF TM0208, which is annotated as a putative ing of PK, as measured by CD spectroscopy, was not observed pyk gene encoding pyruvate kinase. up to 98 °C (Fig. 3). PKs from Hyperthermophilic Archaea and Bacteria 25423 TABLE II Molecular and kinetic properties of the purified native and recombinant pyruvate kinase from T. maritima Kinetic constants were measured at 65 °C, and standard errors are given (see “Material and Methods”). Native enzyme Recombinant enzyme Apparent molecular mass of enzyme (kDa) 194  14 190  7 Apparent molecular mass of subunits (kDa) 51 356  3 Oligomeric structure 4 4 pH optimum 6.0 5.9 T (°C) ND 98 T (°C) ND 80 opt Arrhenius activation energy (kJ/mol, 30–70°C) 61.4 55.6 Apparent V (units/mg) 320  7 578  10 max ADP saturation Apparent S (mM) 1.35  0.07 1.31  0.17 0.5 Hill coefficient (h) 2.85  0.25 2.81  0.14 PEP saturation Effector Apparent S (mM) 0.26  0.03 0.23  0.06 0.5 Hill coefficient (h) 1.95  0.17 2.20  0.05 AMP Apparent S (mM) 0.09  0.06 0.08  0.05 0.5 Hill coefficient (h) 1.32  0.09 1.00  0.10 ATP Apparent S (mM) ND 0.5  0.04 0.5 Hill coefficient (h) ND 2.9  0.13 ND, not determined. Concentrations of effectors were 1 mM. FIG.5. CD spectra of pyruvate kinases from T. maritima (——) and P. aerophilum (). mained almost constant. Thus, e.g. at a PEP concentration of 0.1 mM, AMP activates PK activity up to 10-fold. Conversely, the addition of ATP resulted in an allosteric inhibition of PK FIG.4. Rate dependence of pyruvate kinase from T. maritima activity by increasing S from 0.23 to 0.5 mM; the Hill 0.5 on PEP concentration in the presence and absence of effector. coefficient increased to 2.9, and V was reduced to 70%. max No effector (f),1mM AMP (), and 1 mM ATP (Œ). Inhibition by ATP was completely reversed by the addition of the activator AMP (1 mM). Thus, in contrast to the PKs from Effect of Allosteric Effectors on PK Activity hyperthermophilic archaea, both AMP and ATP exerted their classic allosteric effects toward the hyperthermophilic PK of The effect of classic allosteric activators, such as AMP and the bacterium T. maritima. FBP, and of the allosteric inhibitor ATP was tested on PK activity at 65 °C. The rate dependence of enzyme activity on CD Spectra of Hyperthermophilic PKs from the Archaeon increasing PEP concentrations in the presence of AMP and of P. aerophilum and the Bacterium T. maritima ATP is shown in Fig. 4. In the absence of effectors, rate dependence of PK (recombinant) showed sigmoidal kinetics To get information about the secondary structure of PKs with an S value of 0.23 and a Hill coefficient of 2.2. Addi- from hyperthermophiles, CD spectra were recorded for PK from 0.5 tion of AMP, rather than of FBP, allosterically activates the Pyrobaculum and Thermotoga. The spectra of both PKs were enzyme: rate dependence on PEP changed from a sigmoidal almost superimposable (Fig. 5). For the PK from P. aerophilum kinetics to a hyperbolic, Michaelis-Menten kinetics, paral- an -helical content of 36% and a -sheet content of 25% were leled by the decrease in S for PEP from 0.23 mM to a K of estimated, which closely match the secondary structure predic- 0.5 m 0.08 mM; the Hill coefficient decreased to 1.0, and V re- tions (36% -helical and 26% -sheet) for both enzymes. The max 25424 PKs from Hyperthermophilic Archaea and Bacteria FIG.6. Multiple sequence alignment of amino acid sequences of pyruvate kinases from eukarya, bacteria, and archaea. The alignment was generated with ClustalX. Conserved residues that have been proposed to be indispensable for catalytic activity as deduced from the yeast x-ray structure (26) are indicated by asterisks. The arrow indicates conserved Glu residue essential for K dependence. The consensus pattern is indicated by a box. The predicted secondary structure of the P. aerophilum pyruvate kinase is shown above the sequences. For accession numbers see Fig. 7. secondary structure estimations were comparable to those de- ity described so far. For example, PK from P. aerophilum rived from the x-ray structures of yeast (38% -helical and 12% (optimal growth temperature, 100 °C) showed a temperature -sheet) and E. coli PK (38% -helical and 21% -sheet). optimum higher than 98 °C and was heat-resistant up to 100 °C for 2 h. In addition, thermal unfolding experiments DISCUSSION revealed extremely high melting temperatures of the PKs Molecular and Thermophilic Properties—The PKs from the near or above 100 °C. For comparison, PK from the extreme hyperthermophiles were characterized as homotetramers of thermophilic bacterium Thermus was completely heat- about 200 kDa composed of 50-kDa subunits, which is a com- inactivated in less than 10 min at 100 °C (37). mon feature of PKs from bacteria and eukarya, and of the two Kinetic and Regulatory Properties—All hyperthermophilic archaeal PKs characterized so far, from the crenarchaeon PKs require divalent cations for activity, a common property of T. tenax and the euryarchaeon T. acidophilum (28, 29). In 2 2 2 all characterized PKs; Mn ,Mg ,orCo being most effec- accordance with the optimal growth temperatures of the re- tive. The PKs from the hyperthermophilic archaea and from spective organism, the hyperthermophilic PKs of our study showed the highest temperature optimum and thermostabil- Thermotoga were not dependent on monovalent cations such as PKs from Hyperthermophilic Archaea and Bacteria 25425 FIG.7. Phylogenetic relationships of pyruvate kinases from bacteria, eukarya, and archaea. The numbers at the nodes are bootstrap- ping values according to neighbor-joining (values on top) and maximum-likelihood (values beneath). NCBI accession numbers or SwissProt identifiers: A. pernix, BAA79454; Bac.lic., Bacillus licheniformis KPYK_BACLI; Bac.ste., Bacillus stearothermophilus S29783; Cor.glu., C. glutamicum KPYK_CORGL; Dei.rad., Deinococcus radiodurans AAF12171; E. coli1, KPY1_ECOLI; E. coli2, KPY2_ECOLI; Cat, KPY1_FELCA; Hae.inf., Haemophilus influenzae KPYK_HAEIN; Human, KPY2_HUMAN; Hyd.the., Hydrogenophilus thermoluteolus BAA95686; Lac.del., Lac- tobacillus delbrueckii KPYK_LACDE; Lac.lac., Lactococcus lactis B40620; Lei.mex., L. mexicana KPYK_LEIME; M. jannaschii, Methanococcus jannaschii D64313; M. acetivorans, Methanosarcina acetivorans str. C2A AAM07241; M. mazei, Methanosarcina mazei Goe1 AAM30411; Rabbit, KPY1_RABIT; P. aerophilum, AAL63053; Pyrococcus abyssi, CAB50316; P. furiosus, AAL81312; P.hor., Pyrococcus horikoshii F71171; yeast 1, KPY1_YEAST; yeast 2, KPY2_YEAST; Sal.typ.1, S. typhimurium LT2 AAL20302; Sal.typ.2, S. typhimurium LT2 AAL20804; Str.the., Streptococ- cus thermophilus AAF25804; Str.coe., Streptomyces coelicolor T35759; Str.coe., A3 CAB70653; Sulfolobus solfataricus, AAK41255; Sulfolobus tokodaii, BAB66695; T. acidophilum, KPYK_THEAC; Thermoplasma volcanium, BAB60191; T. tenax, AAF06820; T. maritima, AAD35300; Try.bru., T. brucei brucei KPY2_TRYBB. K and NH .K -independent PKs have been reported for 1,6-bisphosphate (FBP), other sugar phosphates, or AMP. Also E. coli, Corynebacterium glutamicum, and Z. mobilis and for fructose 2,6-bisphosphate, the allosteric activator of the Leish- the archaeon T. tenax (18, 22, 29, 38). PK sequences contain mania PK (16), did not have an effect. Thus, the hyperthermo- highly conserved K sites (residues 48 –52 and 79 – 88 of the philic archaeal PKs differ from PKs of bacteria and eukarya, Pyrobaculum PK), including a glutamate in K -stimulated which are allosterically activated by these compounds. ATP has PKs, e.g. Glu-89 of the yeast PK (Fig. 6). This glutamate is been described to be an allosteric inhibitor of various eukaryal substituted in potassium-independent PKs. In accordance with and bacterial PKs, with inhibitions being reversed by the acti- the lack of the dependence on potassium, all known hyperther- vator FBP or AMP (39). In contrast, the ATP inhibition of PKs mophilic PKs have substituted glutamate at the equivalent from hyperthermophilic archaea in this study is competitive to position (e.g. by Arg, Lys, or Ser). The PK from the moderate substrates PEP and ADP and could not be reversed by AMP thermophilic archaeon T. acidophilum, which contains this or FBP. glutamate, was described to be dependent on potassium (28). An apparent absence of heterotropic allosteric regulation All hyperthermophilic PKs showed a sigmoidal rate depend- has also been reported for the PK of the hyperthermophilic ence for the substrates PEP and ADP and for Mg , indicating archaeon T. tenax (29). However, the PK from the archaeon positively homotropic cooperative response to substrates and T. acidophilum has been described to be activated by AMP (28), cations. Cooperative substrate binding has also been described indicating that the reduced regulatory capacity is probably not for a few PKs from eukarya and bacteria. However, many PKs, a general feature of all archaeal PKs. In contrast to the hyper- including the PK of the hyperthermophilic archaeon T. tenax thermophilic PKs, the PK from the hyperthermophilic bacte- (29), have been reported to show hyperbolic rate dependence on rium T. maritima showed the classic allosteric response to the ADP, suggesting a different conformational response of these allosteric regulators of bacteria. It was allosterically activated PKs to ADP binding. by AMP and inhibited by ATP. Inhibition of ATP could be An important result of this study is the apparent lack of reversed by AMP. allosteric regulation of the PKs from the hyperthermophilic The reasons for the absence of classic heterotropic regulation archaea by classic heterotropic compounds such as fructose of PKs in hyperthermophilic archaea are not understood. A 25426 PKs from Hyperthermophilic Archaea and Bacteria specific effect of high temperatures can be excluded, because ported by fairly good bootstrapping values. However, the lower the hyperthermophilic PK from Thermotoga showed the classic values of some basal nodes are probably due to the influence of allosteric response. The different regulatory behaviors in hy- several factors: phylogenetic distance, regulation, physiology, perthermophilic archaea and Thermotoga might be due to the evolutionary pressure, and temperature adaptation. The PK bacteria I cluster includes the majority of the bacterial PKs, differences in glycolytic pathways, e.g. in archaea all EM path- ways are modified, whereas in Thermotoga the classic EM and for those enzymes, which were functionally characterized, AMP has been shown to be a positive allosteric effector. Almost pathway is operative. A comparative analysis of e.g. adenine all archaeal PK sequences available form a separate cluster nucleotide pools in hyperthermophilic archaea and Thermotoga supporting the monophyletic origin of the archaeal domain. might give an answer to the different response to AMP. Interestingly, the PK from the hyperthermophilic bacterium Sequence Alignment—PKs from the hyperthermophilic or- T. maritima clusters within the archaeal sequences. This ganisms T. maritima, A. pernix, and P. aerophilum show a high might reflect lateral gene transfer of the pyk gene from an degree of similarity to characterized and putative PKs (37– 65% hyperthermophilic archaeon into the T. maritima genome, a similarity) of eukarya and bacteria. Thus, all PKs constitute a phenomenon that has been suggested to occur in T. maritima homologous family: besides minor deviations at variable posi- at high frequency (41). PKs from the archaeal cluster did not tions, all the hyperthermophilic PKs contain the PK consensus show heterotropic regulation or were regulated by AMP (Ther- pattern [LIVAC]-X-[LIVM](2)-[SAPCV]-K-[LIV]-E-[NKRST]- motoga and Thermoplasma). The PK sequence of the archaeon X-[DEQHS]-[GSTA]-[LIVM] (40). An alignment of these hyper- A. fulgidus strain 7324 is not known. N-terminal amino acid thermophilic PKs with selected homologous proteins from all sequences indicate high identity to PKs from the archaea three domains is given in Fig. 6. Available PK crystal struc- Pyrococcus and Thermococcus. The absence of a pyk homo- tures, from cat and rabbit muscle, yeast, E. coli, and L. mexi- log in the genome of the closely related strain, A. fulgidus cana revealed that each subunit is composed of four domains VC16, might be explained by a loss of this gene; alternatively, (N, A, B, and C) (eukarya) or three domains (A, B, and C) A. fulgidus 7324 might have taken up its pyk gene via lateral (bacteria). The domain N, located at the N terminus, is a short gene transfer from Thermococcales. The putative PK from the -helical stretch present in eukaryotic sequences but absent in Halobacterium NRC I, an outgroup with low bootstrapping the bacterial and archaeal homologs. The domain A, which support, was omitted. All eukaryotic PK sequences form a includes the catalytic site (residues 14 – 83 and 176 –348 of the cluster. Most of them are allosterically activated by F-1,6-BP or P. aerophilum PK, Fig. 6), constitutes a classic () barrel F-2,6-BP (PKs from protistas). However, the absence of allo- structure. The domain B (residues 84 –175) is a -sheet capping steric activation has been reported for a few eukaryal isoen- the catalytic domain. The domain C (residues 349 – 461), lo- zymes. The second bacterial PK cluster comprises PKs from cated at the C terminus, is an open twisted , structure, Gram-positives with low guanine-cytosine content and -pro- containing the FBP binding site of the yeast enzyme. The teobacteria species. Enzymes from this group show allosteric highest degree of homology is found in the catalytic domain A. response to either FBP or to AMP. The two separate bacterial In this domain, a number of residues have been identified to be PK clusters might have evolved by a gene duplication in the important for catalysis, according to the yeast PK structure early bacterial evolution. Alternatively, lateral gene transfers (26). These residues are conserved in all selected PK sequences, from eukaryotes to some Gram-positive and proteobacteria including the PKs of this study. In contrast to the domain A, might be postulated. The latter hypothesis could explain the domain B and in particular C showed significantly lower se- close clustering of the second bacterial group with the eukarya, quence homology. As deduced from the structure of the yeast as well as the allosteric regulation by FBP of isoenzymes 1 from PK, at least eight residues (Ser-402, Thr-403, Ser-404, Thr-407, E. coli and S. thyphimurium, a property found only in eukaryal Trp-452, Arg-459, Gly-475, and His-491 of the yeast PK) have PKs. been identified in the domain C to contact the allosteric effector FBP. However, these residues are not conserved among the Acknowledgments—We thank Dr. J. Gro ¨ tzinger (Kiel) for help in CD spectroscopic measurements, Dr. R. Schmid (Osnabru ¨ ck) for N-termi- FBP-regulated PKs. The presence of a conserved glutamate nal amino acid sequencing, and H. Preidel (Kiel) for mass culturing (Glu-432) has been attributed to the non-allosteric property of A. fulgidus 7324 and T. maritima. the mammalian M1 isoenzyme (26). In the PKs of hyperther- mophilic archaea, no glutamate at the equivalent position was REFERENCES found, indicating that a conserved glutamate is not a prereq- 1. Stetter, K. O. (1996) FEMS Microbiol. Rev. 18, 149 –158 2. Scho ¨ nheit, P., and Scha ¨ fer, T. (1995) World J. Microbiol. Biotechnol. 11, 26 –57 uisite for non-allosteric behavior of PKs. Thus, the absence of 3. Labes, A., and Scho ¨ nheit, P. (2001) Arch. Microbiol. 176, 329 –338 FBP regulation of the hyperthermophilic PKs cannot be ex- 4. Selig, M., Xavier, K. B., Santos, H., and Scho ¨ nheit, P. (1997) Arch. Microbiol. 167, 217–232 plained on sequence level. So far, crystal structures of both 5. De Vos, W. M., Kengen, S. W., Voorhorst, W. G., and Van der Oost, J. (1998) AMP-activated PKs and PKs from hyperthermophilic archaea Extremophiles 2, 201–205 6. Siebers, B., and Hensel, R. (1993) FEMS Microbiol. Lett. 111, 1– 8 have not been reported. Crystallization of the AMP-regulated 7. Kengen, S. W., Tuininga, J. E., Verhees, C. H., van der, O. J., Stams, A. J., and PK from T. maritima and of the archaeal PKs are in progress to de Vos, W. M. (2001) Methods Enzymol. 331, 41–53 identify the AMP binding site and to understand the structural 8. Ronimus, R. S., Koning, J., and Morgan, H. W. (1999) Extremophiles 3, 121–129 basis for reduced regulatory properties of hyperthermophilic 9. Hansen, T., and Scho ¨ nheit, P. (2001) Arch. Microbiol. 177, 62– 69 archaeal PKs. 10. Hansen, T., and Scho ¨ nheit, P. (2000) Arch. Microbiol. 173, 103–109 11. Siebers, B., Klenk, H. P., and Hensel, R. (1998) J. Bacteriol. 180, 2137–2143 Phylogenetic Affiliation—With the detailed characterization 12. Hansen, T., Musfeldt, M., and Scho ¨ nheit, P. (2002) Arch. Microbiol. 177, of the PKs from T. maritima, A. pernix, and P. aerophilum 401– 409 13. Kayne, F. J. (1973) in The Enzymes (Boyer, P. D., ed) pp. 353–382, Academic along with the recently available archaeal homologs, a suffi- Press, New York cient number of PKs from all three domains of life is available 14. Muirhead, H. (1990) Biochem. Soc. Trans. 18, 193–196 allowing phylogenetic studies with respect to both evolutionary 15. Mattevi, A., Bolognesi, M., and Valentini, G. (1996) FEBS Lett. 389, 15–19 16. Rigden, D. J., Phillips, S. E., Michels, P. A., and Fothergill-Gilmore, L. A. as well as functional aspects. As demonstrated in the phylo- (1999) J. Mol. Biol. 291, 615– 635 gram in Fig. 7, selected PK sequences are clustered in four 17. Ernest, I., Callens, M., Uttaro, A. D., Chevalier, N., Opperdoes, F. R., Muirhead, H., and Michels, P. A. (1998) Protein Expr. Purif. 13, 373–382 groups: PK bacteria I, PK archaea, PK eukarya, and PK bac- 18. Waygood, E. B., Rayman, M. K., and Sanwal, B. D. (1975) Can. J. Biochem. 53, teria II. The overall topology of the tree was achieved by both 444 – 454 neighbor-joining and maximum likelihood methods and is sup- 19. Garcia-Olalla, C., and Garrido-Pertierra, A. (1987) Biochem. J 241, 573–581 PKs from Hyperthermophilic Archaea and Bacteria 25427 20. Waygood, E. B., Mort, J. S., and Sanwal, B. D. (1976) Biochemistry 15, 277–282 33. Deleage, G., and Geourjon, C. (1993) Comput. Appl. Biosci. 9, 197–199 21. Nairn, J., Duncan, D., Gray, L. M., Urquhart, G., Binnie, M., Byron, O., 34. Thompson, J. D., Gibson, T. J., Plewniak, F., Jeanmougin, F., and Higgins, Fothergill-Gilmore, L. A., and Price, N. C. (1998) Protein Expr. Purif. 14, D. G. (1997) Nucleic Acids Res. 25, 4876 – 4882 247–253 35. Klenk, H. P., Clayton, R. A., Tomb, J. F., White, O., Nelson, K. E., Ketchum, 22. Pawluk, A., Scopes, R. K., and Griffiths-Smith, K. (1986) Biochem. J. 238, K. A., Dodson, R. J., Gwinn, M., Hickey, E. K., Peterson, J. D., Richardson, 275–281 D. L., Kerlavage, A. R., Graham, D. E., Kyrpides, N. C., Fleischmann, R. D., 23. Bateman, A., Birney, E., Durbin, R., Eddy, S. R., Howe, K. L., and Sonnhammer, Quackenbush, J., Lee, N. H., Sutton, G. G., Gill, S., Kirkness, E. F., E. L. (2001) Nucleic Acids Res. 28, 263–266 Dougherty, B. A., McKenney, K., Adams, M. D., Loftus, B., and Venter, J. C. 24. Muirhead, H., Clayden, D. A., Barford, D., Lorimer, C. G., Fothergill-Gilmore, (1997) Nature 390, 364 –370 L. A., Schiltz, E., and Schmitt, W. (1986) EMBO J. 5, 475– 481 36. Muirhead, H. (1987) in Biological Macromolecules and Assemblies: Active Sites 25. Larsen, T. M., Laughlin, L. T., Holden, H. M., Rayment, I., and Reed, G. H. of Enzymes (Jurnak, F. A., and McPherson, A., eds) pp. 143–186, John Wiley (1994) Biochemistry 33, 6301– 6309 and Sons, New York 26. Jurica, M. S., Mesecar, A., Heath, P. J., Shi, W., Nowak, T., and Stoddard, B. L. 37. Yoshizaki, F., and Imahori, K. (1979) Agric. Biol. Chem. 43, 527–536 (1998) Structure 6, 195–210 38. Jetten, M. S., Gubler, M. E., Lee, S. H., and Sinskey, A. J. (1994) Appl. 27. Mattevi, A., Valentini, G., Rizzi, M., Speranza, M. L., Bolognesi, M., and Coda, Environ. Microbiol. 60, 2501–2507 A. (1995) Structure 3, 729 –741 39. Waygood, E. B., and Sanwal, B. D. (1974) J. Biol. Chem. 249, 265–274 28. Potter, S., and Fothergill-Gilmore, L. A. (1992) FEMS Microbiol. Lett. 73, 40. Hofmann, K., Bucher, P., Falquet, L., and Bairoch, A. (1999) Nucleic Acids Res. 235–239 27, 215–219 29. Schramm, A., Siebers, B., Tjaden, B., Brinkmann, H., and Hensel, R. (2000) J. 41. Nelson, K. E., Clayton, R. A., Gill, S. R., Gwinn, M. L., Dodson, R. J., Haft, Bacteriol. 182, 2001–2009 D. H., Hickey, E. K., Peterson, J. D., Nelson, W. C., Ketchum, K. A., 30. Beeder, J., Nilsen, R. K., Rosnes, J. T., Torsvik, T., and Lien, T. (1994) Appl. McDonald, L., Utterback, T. R., Malek, J. A., Linher, K. D., Garrett, M. M., Environ. Microbiol. 60, 1227–1231 Stewart, A. M., Cotton, M. D., Pratt, M. S., Phillips, C. A., Richardson, D., 31. Huber, R., Langworthy, T. A., Ko ¨ nig, H., Thomm, M., Woese, C. R., Sleytr, U. B., and Stetter, K. O. (1986) Arch. Microbiol. 144, 324 –333 Heidelberg, J., Sutton, G. G., Fleischmann, R. D., Eisen, J. A., and Fraser, 32. Schro ¨ der, C., Selig, M., and Scho ¨ nheit, P. (1994) Arch. Microbiol. 161, 460 – 470 C. M. (1999) Nature 399, 323–329

Journal

Journal of Biological Chemistry – Unpaywall

Published: Jul 1, 2003

There are no references for this article.