ADP-dependent Glucokinase/Phosphofructokinase, a Novel Bifunctional Enzyme from the Hyperthermophilic ArchaeonMethanococcus jannaschii *

A gene encoding an ADP-dependent phosphofructokinase homologue has been identified in the hyperthermophilic archaeon Methanococcus jannaschii via genome sequencing. The gene encoded a protein of 462 amino acids with a molecular weight of 53,361. The deduced amino acid sequence of the gene showed 52 and 29% identities to the ADP-dependent phosphofructokinase and glucokinase from Pyrococcus furiosus, respectively. The gene was overexpressed inEscherichia coli, and the produced enzyme was purified and characterized. To our surprise, the enzyme showed high ADP-dependent activities for both glucokinase and phosphofructokinase. A native molecular mass was estimated to be 55 kDa, and this indicates the enzyme is monomeric. The reaction rate for the phosphorylation of d-glucose was almost 3 times that for d-fructose 6-phosphate. The K m values for d-fructose 6-phosphate and d-glucose were calculated to be 0.010 and 1.6 mm, respectively. TheK m values for ADP were 0.032 and 0.63 mm when d-glucose and d-fructose 6-phosphate were used as a phosphoryl group acceptor, respectively. The gene encoding the enzyme is proposed to be an ancestral gene of an ADP-dependent phosphofructokinase and glucokinase. A gene duplication event might lead to the two enzymatic activities.

In general, ATP is regarded as the universal energy carrier and the most common phosphoryl group donor for kinases. However, several gluco-and phosphofructokinases have been reported to have different phosphoryl group donor specificity. The glucokinase from Mycobacterium tuberculosis can utilize both ATP and polyphosphate as the phosphoryl group donor (1). PP i -dependent phosphofructokinases have been reported to be present in several eucarya and bacteria and in the hyperthermophilic archaeon Thermoproteus tenax (2)(3)(4). Recently novel sugar kinases, ADP-dependent (AMP-forming) glucokinase (ADP-GK) 1 and phosphofructokinase (ADP-PFK), were discovered in the hyperthermophilic archaeon Pyrococcus furiosus (5). Those enzymes require ADP as the phosphoryl group donor instead of ATP and are involved in a modified Embden-Meyerhof pathway in this organism. The hyperthermophilic archaea are relatively deeply branched archaea and are considered to be phylogenetically ancient organisms. Therefore, structural analysis of the kinases from these organisms may provide abundant information for phylogenetic analysis of the sugar kinases. We cloned and sequenced the gene encoding the ADP-GKs from P. furiosus and Thermococcus litoralis (The nucleotide sequences have been submitted to the GenBank TM data bases as the genes for ADP-dependent hexokinase and are available under accession numbers E14588 and E14589.) (6). About 59% identity in amino acid sequence was observed between these two enzymes, although they did not show similarity with any ATP-dependent kinases that have been reported so far. In addition, the amino acid sequence of the P. furiosus ADP-GK showed high identity (26%) with that reported for the P. furiosus ADP-PFK (7). This suggests that those kinases belong to a novel kinase family and might have evolved from a common origin.
Recently a gene encoding the ADP-PFK homologue has been identified from genome information in the hyperthermophilic archaeon Methanococcus jannaschii (8). Verhees et al. (9) have expressed the gene in Escherichia coli and revealed that the produced enzyme has an ADP-PFK activity. They performed characterization of the enzyme with regard to the ADP-PFK activity. We have also analyzed the same gene of M. jannaschii and found that the identity (29%) in amino acid sequence between the M. jannaschii ADP-PFK homologue and P. furiosus ADP-GK is somewhat higher than that (26%) observed between P. furiosus ADP-PFK and -GK. We expressed the gene in E. coli and examined characteristics of the product. As a result, we found that the produced enzyme has high ADP-dependent activity for both glucokinase and phosphofructokinase. We show here that the enzyme is a novel type of enzyme, a bifunctional ADP-dependent glucokinase/phosphofructokinase (ADP-GK/PFK). The enzyme was proposed to be a common origin of the ADP-GK and -PFK from genome analysis. This is the first example of a kinase that catalyzes phosphorylation of both D-glucose and D-fructose 6-phosphate.

EXPERIMENTAL PROCEDURES
Overexpression and Purification of Recombinant Protein-The gene (MJ1604) encoding the ADP-PFK homologue, which shows high similarity to that of the P. furiosus ADP-PFK, has been identified in the M. jannaschii genome (8). The plasmid DNA pET15b, which carries an N-terminal His tag sequence, was obtained from Novagen, Inc. (Madi-* This work was supported in part by "The New Energy and Industrial Technology Development Organization (NEDO) Project" promoted by the Ministry of International Trade and Industry of Japan. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
¶ To whom correspondence should be addressed. Tel.: 81-88-656-7518; Fax: 81-88-656-9071; E-mail: ohshima@bio.tokushima-u.ac.jp. son, WI). The following set of oligonucleotide primers was used to amplify the ADP-PFK gene fragment by PCR: the primer (5Ј-CTAGAG-GAATAACCATATGTGTG-3Ј) introduced a unique NdeI restriction site overlapping the 5Ј-initiation codon, and the other primer (5Ј-CATGCG-GATCCTTAGCTCATTC-3Ј) introduced a unique BamHI restriction site proximal to the 3Ј-end of the termination codon. The chromosomal M. jannaschii DNA was isolated as described before (10) and used as the template. The amplified 1.4-kb fragment was digested with NdeI and BamHI and ligated with the expression vector pET15b linearized with NdeI and BamHI to generate pMJGK/PFK. The E. coli strain BL21(DE3) codon plus RIL, which was obtained from Stratagene (La Jolla, CA), was transformed with pMJGK/PFK. The transformants were cultivated at 37°C in 2 liters of Luria Bertani medium containing 50 g/ml ampicillin until the optical density at 600 nm reached 0.6. The induction was carried out by the addition of 0.4 mM isopropyl-␤-Dthiogalactopyranoside to the medium, and cultivation was continued for 3 h. Cells were harvested by centrifugation, suspended in 20 mM Tris/ HCl buffer (pH 7.5) containing 0.2 M Na 2 SO 4 , 10 mM dithiothreitol, and 1 mM phenylmethylsulfonyl fluoride, and disrupted by ultrasonication. The crude extract was heated at 80°C for 10 min, and the denatured protein was then removed by centrifugation (15,000 ϫ g for 15 min). The supernatant solution was dialyzed against 20 mM Tris/HCl buffer (pH 7.5) containing 0.5 M NaCl and 10 mM imidazole and loaded on a nickel-charged chelating Sepharose column (2.6 ϫ 10 cm, Amersham Biosciences) equilibrated with the same buffer. Protein was eluted with a 200-ml linear gradient of 0 -0.5 M imidazole in the same buffer. The active fractions were pooled, dialyzed against 20 mM Tris/HCl buffer (pH 7.5), and applied to an Uno-Q column (1.5 ϫ 10 cm) (Bio-Rad) equilibrated with the same buffer. After washing with 30 ml of the same buffer, the enzyme was eluted with a 60-ml linear gradient of 0 -0.5 M NaCl. The active fractions were pooled, dialyzed against 20 mM Tris/ HCl buffer (pH 7.5), and used as the purified enzyme preparation.
The  fructose 1-phosphate, and D-glucose 6-phosphate as substrates was tested by measuring the formation of AMP from ADP by high performance liquid chromatography (HPLC) basically as described by Koga et al. (6). The reaction mixture contained 50 mM BisTris Buffer (pH 6.5), a 20 mM concentration of each substrate, 2 mM ADP, 2 mM MgCl 2 ⅐6H 2 O, and 20 l of enzyme preparation in a total volume of 0.5 ml. After incubation for 10 min at 37°C, the reaction was stopped by cooling on ice. After 5 min, each solution was passed through a cellulose acetate filter (pore size, 0.2 m; ADVANTEC, Tokyo, Japan). An aliquot (20 l) of each filtrate was subjected to a column (7.6 mm ϫ 25 cm) of Asahipak GS320HQ (Asahi Chemical Industry Co., Ltd., Shizuoka, Japan). NaH 2 PO 4 (200 mM, pH 5.0) was used as the mobile phase at a flow rate of 1.0 ml/min. The effluent from the column was monitored by a UV detector at a wavelength of 260 nm. ATP, ADP, and AMP were separated on a column at retention times of about 7.6, 8.6, and 11.6 min, respectively. For the detection of specificity for the phosphoryl group donor, GDP, CDP, ATP, GTP, pyrophosphate, tripolyphosphate, trimetaphosphate, and phosphoenolpyruvate (each 2 mM) and polyphosphate (0.2 mg/ml) were used instead of ADP in the standard assay mixture. The divalent cation requirement was tested by the addition of 2 mM MgCl 2 , CoCl 2 , NiCl 2 , MnCl 2 , ZnCl 2 , PbCl 2 , or CaCl 2 to the standard assay mixture. The molecular mass of the enzyme was determined by gel filtration on a TSK gel column G3000SWXL (7.8 mm ϫ 30 cm) (Tosoh, Tokyo, Japan), and the subunit molecular mass of the purified enzyme was determined by SDS-PAGE as described previously (6).

RESULTS AND DISCUSSION
Expression of the Gene and Purification of the Recombinant Enzyme-The E. coli strain BL21(DE3) codon plus RIL transformed with the expression vector pMJGK/PFK exhibited high activities for both ADP-dependent glucokinase and phosphofructokinase, which were not lost by incubation at 80°C for 10 min. Hereafter we refer to the enzyme as the M. jannaschii ADP-GK/PFK. The enzyme was purified to homogeneity from the extract of E. coli cells. About 10 mg of the purified enzyme was obtained from 2 liters of the E. coli culture. The specific activity of the purified M. jannaschii ADP-GK/PFK was estimated to be 7.7 mol/min/mg at 50°C for phosphofructokinase activity at the optimum pH of 6.5. The specific activity for the ADP-PFK reaction of the M. jannaschii enzyme has been reported to be 8.2 mol/min/mg at 50°C by Verhees et al. (9), and this is compatible with that measured in this study. On the other hand, the specific activity for the ADP-GK reaction of the enzyme was about 3 times that for the ADP-PFK reaction and was estimated to be 21.5 mol/min/mg at 50°C. Upon heating at 80°C for 10 min, the enzyme retained its full activity but lost 20% of the activity at 90°C after a 10-min incubation. The purified ADP-GK/PFK showed activity only in the forward direction.
Characteristics of the M. jannaschii ADP-GK/PFK-The biochemical characteristics of the purified enzyme were determined and compared with those of the ADP-GKs from P. furiosus and T. litoralis (6) and ADP-PFK from P. furiosus (7). The deduced amino acid sequence of the M. jannaschii ADP-GK/ PFK gene showed 52, 29, and 33% identities to the P. furiosus ADP-PFK, P. furiosus ADP-GK, and T. litoralis ADP-GK, respectively. SDS-PAGE of the purified enzyme gave only one band; the subunit molecular mass was determined to be about 53 kDa, and was consistent with the molecular weight (55,524) calculated from the amino acid sequence including the His tag sequence. The native molecular mass of the enzyme determined by HPLC was about 55 kDa; this indicates the enzyme is monomeric. Although the P. furiosus ADP-PFK shows high homology with the M. jannaschii ADP-GK/PFK, it has a tetramer structure composed of four identical subunits (7), which is most common for phosphofructokinases. The Thermococcus zilligii ADP-PFK has been reported to have the same structure (8). In this regard, the M. jannaschii ADP-GK/PFK is similar to the T. litoralis ADP-GK (Table I).
ADP could be replaced by GDP to some extent for both the ADP-GK and -PFK reactions of the M. jannaschii enzyme (Table I). When D-glucose was used as a phosphoryl group acceptor, ADP could be replaced by CDP to a limited extent. ATP, GTP, pyrophosphate, tripolyphosphate, trimetaphosphate, polyphosphate, and phosphoenolpyruvate were inert. The ADP-GKs from P. furiosus and T. litoralis utilize GDP to a very limited extent as the phosphoryl group donor, although the two enzymes showed comparable activity for ADP and CDP. On the other hand, the P. furiosus ADP-PFK has some reactivity for GDP. In this respect, the M. jannaschii enzyme is similar to the P. furiosus ADP-PFK (Table I). The enzymes required divalent cations for both activities. MgCl 2 and CaCl 2 were comparatively effective, and they were able to be replaced by NiCl 2, MnCl 2 , PbCl 2 , or CoCl 2 to some extent (Table I). However, the ADP-GKs from P. furiosus and T. litoralis do not utilize CaCl 2 , and the P. furiosus ADP-PFK shows very low activity for CaCl 2 . Therefore, the reactivity for CaCl 2 is one of the remarkable characteristics of the enzyme. The ability of the enzyme to catalyze the phosphorylation of various sugars was examined. The enzyme catalyzed the phosphorylation of D-fructose and 2-deoxy-D-glucose to a limited extent in addition to D-glucose and D-fructose 6-phosphate (Table I). D-Glucosamine, D-mannose, D-galactose, D-fructose 1-phosphate, and D-glucose 6-phosphate were inert. Typical Michaelis-Menten kinetics were observed for both the phosphorylation of D-glucose and D-fructose 6-phosphate at 50°C. The apparent K m values for fructose 6-phosphate and D-glucose were calculated to be 0.010 and 1.6 mM, respectively. The K m values for ADP were 0.032 and 0.63 mM when D-glucose and D-fructose 6-phosphate were used as a phosphoryl group acceptor, respectively. To date, the ADP-PFK that utilizes D-glucose as a phosphoryl group acceptor has not been found. The ADP-GKs do not utilize D-fructose 6-phosphate (Table I). To the best of our knowledge, the ADP-GK/PFK of M. jannaschii is the only enzyme having significantly high levels of both activities. We performed inhibition studies for the glucokinase activity of the enzyme with fructose-6-phosphate as an inhibitor. The double-reciprocal plots of v versus glucose concentrations at several fixed concentrations of fructose 6-phosphate showed a typical competitive inhibition pattern (Fig. 1A). K i for fructose 6-phosphate was calculated to be 0.0047 mM from Dixon plots (Fig. 1B). This indicates that the substrate recognition site of the enzyme for glucose and fructose 6-phosphate is identical.
Genome Analysis- Fig. 2 shows an amino acid alignment of ADP-GK and -PFK homologues from M. jannaschii, P. furiosus, Pyrococcus horikoshii OT-3, and Pyrococcus abyssi. A phylogenetic tree produced using the alignment of Fig. 2 by the neighbor-joining method is shown in Fig. 3. The genome information of those organisms is available at the Kyoto Encyclopedia of Genes and Genomes (genome.ad.jp/kegg/) and the Utah Genome Center (www.genome.utah.edu.). The data bases were screened for homologues of ADP-GKs and -PFKs using blastP. As shown in Fig. 3, the ADP-GKs clustered together and are separated from another cluster of the ADP-PFKs. Interestingly both the ADP-GK and -PFK homologues are present in P. furiosus, P. horikoshii OT-3, and P. abyssi but not in M. jannaschii, except for the ADP-GK/PFK. This suggests that the gene encoding the enzyme might be an ancestral gene of the ADP-GK and -PFK, and a gene duplication event has lead to the two enzymatic activities.
Recently the crystal structure of the T. litoralis ADP-GK has been solved (11). Cocrystallization with ADP elucidated the position of the nucleotide-binding site, and the residues that interact directly with ADP were identified. In addition, the important residues involved in the binding of sugars have been predicted based on the sugar binding position of the ATP-de-pendent ribokinase family that was observed to have significant structural similarity to the ADP-GK (11). Verhees et al. (9) have reported that those putative crucial residues (Asp 42 and Gly 119 -Gly 120 of the T. litoralis ADP-GK) are well conserved in all archaeal homologues of ADP-dependent sugar kinases including the M. jannaschii ADP-PFK (ADP-GK/PFK in this paper). This indicates that those residues are common to both the ADP-GK and -PFK, and the other residues may contribute to the discrimination between D-glucose and D-fructose 6-phosphate in those kinases. As shown in Fig. 2, several residues of the M. jannaschii ADP-GK/PFK are conserved in the ADP-GKs but not in the ADP-PFKs. In particular, Lys 31 , Tyr 32 , Asp 37 , Ser 74 , Glu 82 , Leu 115 , Gln 135 , Asp 141 , Ala 192 , Lys 382 , Asn 387 , Lys 435 , and Ser 446 of the M. jannaschii enzyme are well conserved in all the ADP-GKs including the T. litoralis ADP-GK (not shown). Some or one of those residues might be responsible for the reactivity of the ADP-dependent kinases for D-glucose. To clarify the amino acid residues that are responsible for the binding of D-glucose and D-fructose 6-phosphate, x-ray reflection analysis of the M. jannaschii ADP-GK/PFK is under investigation.
The presence of a glycolytic pathway in methanogen has been proposed on the basis of enzyme analyses of several mesophilic and thermophilic methanogens (9,12) and the genome sequence of M. jannaschii, which revealed the presence of several glycolytic enzyme homologues (13). A number of methanogens have been known to synthesize glycogen intracellularly and accumulate it as a reserve polysaccharide (9,12). Under starvation, the degradation of the glycogen storage has been observed in the mesophilic archaeon Methanococcus maripaludis (12). The recent characterization of the amino acid sequence of the ADP-GKs from P. furiosus and T. litoralis (6) and that of the ADP-PFKs from P. furiosus (7) and T. zilligii (8) resulted in the identification of a homologue in the genome of M. jannaschii (8). These observations suggest that a modified Embden-Meyerhof pathway, present in P. furiosus, might also be operational in methanogens. It is therefore interesting to examine whether the production of methane from glycogen or D-glucose by methanogen itself is permitted. The observations described above and our results suggest that such a probability is not completely ruled out.