First Characterization of an Archaeal GTP-Dependent Phosphoenolpyruvate Carboxykinase from the Hyperthermophilic Archaeon Thermococcus kodakaraensis KOD1

Abstract

Phosphoenolpyruvate carboxykinase (PCK), which catalyzes the nucleotide-dependent, reversible decarboxylation of oxaloacetate to yield phosphoenolpyruvate and CO₂, is one of the important enzymes in the interconversion between C₃ and C₄ metabolites. This study focused on the first characterization of the enzymatic properties and expression profile of an archaeal PCK from the hyperthermophilic archaeon Thermococcus kodakaraensis (Pck_Tk). Pck_Tk showed 30 to 35% identities to GTP-dependent PCKs from mammals and bacteria but was located in a branch distinct from that of the classical enzymes in the phylogenetic tree, together with other archaeal homologs from Pyrococcus and Sulfolobus spp. Several catalytically important regions and residues, found in all known PCKs irrespective of their nucleotide specificities, were conserved in Pck_Tk. However, the predicted GTP-binding region was unique compared to those in other GTP-dependent PCKs. The recombinant Pck_Tk actually exhibited GTP-dependent activity and was suggested to possess dual cation-binding sites specific for Mn²⁺ and Mg²⁺. The enzyme preferred phosphoenolpyruvate formation from oxaloacetate, since the K_m value for oxaloacetate was much lower than that for phosphoenolpyruvate. The transcription and activity levels in T. kodakaraensis were higher under gluconeogenic conditions than under glycolytic conditions. These results agreed with the role of Pck_Tk in providing phosphoenolpyruvate from oxaloacetate as the first step of gluconeogenesis in this hyperthermophilic archaeon. Additionally, under gluconeogenic conditions, we observed higher expression levels of Pck_Tk on pyruvate than on amino acids, implying that it plays an additional role in the recycling of excess phosphoenolpyruvate produced from pyruvate, replacing the function of the anaplerotic phosphoenolpyruvate carboxylase that is missing from this archaeon.

Recent progress in the research on hyperthermophilic archaea has clarified the presence of unique glycolytic and gluconeogenic pathways distinct from those in mesophilic organisms. In Thermococcus and Pyrococcus spp. belonging to the order Thermococcales in Euryarchaeota, glucose is metabolized by a modified Embden-Meyerhof pathway including ADP-dependent kinases and glyceraldehyde 3-phosphate:ferredoxin oxidoreductase (21, 25). After pyruvate is formed through glycolysis, the terminal reactions of oxidative glucose degradation are the conversion of pyruvate to acetate and CO₂ as end products. Pyruvate:ferredoxin oxidoreductase oxidizes pyruvate to acetyl coenzyme A (acetyl-CoA), and acetyl-CoA synthetase (ADP forming) catalyzes acetate formation from acetyl-CoA, coupled with the phosphorylation of ADP (7, 21). In the direction of gluconeogenesis, phosphoenolpyruvate (PEP) synthase is thought to function in the supply of PEP from pyruvate. Although a wealth of knowledge about these pathways has accumulated, there is still very little information on how these metabolites are converted to C₄ compounds, and vice versa, in hyperthermophilic archaea.

In general, the metabolic interconversion among PEP, oxaloacetate (OAA), and pyruvate (the PEP-OAA-pyruvate cycle) is of importance in the replenishment of intermediates of the tricarboxylic acid cycle (anaplerosis) when they are consumed by various biosynthetic processes, as well as in their removal towards other pathways (cataplerosis), such as gluconeogenesis, when they are in excess (16). A number of enzymes involved in this cycle, such as pyruvate kinase, PEP synthase or pyruvate orthophosphate dikinase, pyruvate carboxylase, OAA decarboxylase, and malic enzyme, have been identified in a wide variety of organisms. The interconversion between PEP and OAA is made possible through PEP carboxylase and PEP carboxykinase (PCK).

PCK catalyzes the nucleotide-dependent reversible decarboxylation of OAA to yield PEP and CO₂. PCKs have traditionally been divided into two classes based on their nucleotide dependencies (8, 11). ATP-dependent enzymes are widely distributed in bacteria, yeasts, and plants, while GTP-dependent enzymes are found in animals and a small set of bacteria. Although there is no significant identity between the two classes of enzymes, ATP- and GTP-PCKs share conserved nucleoside triphosphate- and substrate-binding motifs and metal-binding residues in their primary structures (3, 10).

The physiological role of PCK depends on the presence or absence of the other enzymes in the PEP-OAA-pyruvate cycle. In animal liver and kidney, PCK catalyzes PEP formation from OAA as the first and key rate-limiting step in gluconeogenesis, so this enzyme has been well studied in terms of its participation in glucose homeostasis and type II diabetes (27, 30). In contrast, the majority of bacteria possess PEP synthase or pyruvate orthophosphate dikinase as an alternative enzyme for the first step in gluconeogenesis. Therefore, the presumed role of PCK in bacteria varies widely depending on the species. In Escherichia coli, it has been demonstrated that PEP synthase is essential for the entry of C₃ compounds into gluconeogenesis, while ATP-PCK is responsible for the entry of C₄ compounds, a function that is also performed via the concerted actions of malic enzyme and PEP synthase (22). In the l-lysine producer Corynebacterium glutamicum, GTP-PCK in glucose-grown cells has been reported to be an enzyme responsible for the recycling of anaplerotically synthesized excess OAA to PEP (17, 18). In other bacteria, as suggested for the GTP-PCK in the anaerobic ruminal bacterium Ruminococcus flaverfacience (24), PCKs have been found to fulfill an anaplerotic role when the tricarboxylic acid cycle is utilized for anabolic purposes, such as organic and amino acid synthesis.

Thermococcus kodakaraensis KOD1 is a sulfur-reducing hyperthermophilic archaeon isolated from Kodakara Island, Kagoshima, Japan (2, 13). The strain can grow with amino acids as carbon and energy sources, and elemental sulfur acts as a terminal electron acceptor to be reduced to H₂S. Moreover, T. kodakaraensis can assimilate and grow on starch or pyruvate in the absence of sulfur, providing a good tool for studying glycolysis and gluconeogenesis and their regulation in hyperthermophiles. Recently, we have been proceeding with the complete genome analysis of T. kodakaraensis KOD1, and the preliminary results have enabled us to identify one gene corresponding to an uncharacterized homolog of PCK. This study aimed to carry out the first characterization of an archaeal PCK along with an examination of its expression profile in order to estimate the physiological function of the enzyme in hyperthermophilic archaea.

MATERIALS AND METHODS

Microorganisms, plasmids, and media.

T. kodakaraensis KOD1 cells were grown anaerobically in MA-YT medium (4.8 and 26.4 g of Marine Art SF agents A and B [Senjyu Seiyaku, Osaka, Japan], respectively, per liter, 5.0 g of yeast extract per liter, and 5.0 g of tryptone per liter) with an addition of 5.0 g of sulfur or pyruvate per liter and/or 5.0 g of starch per liter. E. coli DH5α and plasmid pUC118 were used for general DNA manipulation and sequencing, and E. coli BL21-CodonPlus(DE3)-RIL (Stratagene, La Jolla, Calif.) and pET21a(+) (Novagen, Madison, Wis.) were used for gene expression. E. coli strains were cultivated in Luria-Bertani medium (10 g of tryptone per liter, 10 g of yeast extract per liter, and 5 g of NaCl [pH 7.0] per liter) at 37°C. Ampicillin was added to the medium at a concentration of 100 μg/ml when needed.

DNA manipulation and sequencing.

Plasmid DNA was purified using plasmid kits (QIAGEN, Hilden, Germany). DNA sequencing was performed with a BigDye Terminator Cycle sequencing kit (version 3.0) and a model 3100 capillary DNA sequencer (Applied Biosystems, Foster City, Calif.). Restriction and modification enzymes were purchased from Toyobo (Osaka, Japan) or Takara (Kyoto, Japan). KOD Plus (Toyobo) was used as a polymerase for PCR, and a GFX PCR DNA and gel band purification kit (Amersham Biosciences, Little Chalfont, Buckinghamshire, United Kingdom) was used to recover DNA fragments from agarose gel after the electrophoresis.

Expression of pck_Tk.

To construct an expression vector for the PCK gene from T. kodakaraensis KOD1 (pck_Tk), the gene was amplified with T. kodakaraensis genomic DNA as a template and two oligonucleotide primers (sense, 5′-AAAAAGAATTCCATATGAACGCTCTTGAACGGCTTGAAAAGC-3′, and antisense, 5′-AAAAAGAATTCTCACTCGCCCTCAAGGGCGAAGGGGC-3′ [the under-lined sequences indicate an NdeI site in the sense primer and an EcoRI site in the antisense primer]). The amplified fragment was inserted into pUC118. After the absence of unintended mutations in the insert was confirmed, the NdeI-EcoRI restriction fragment was inserted into pET21a(+) at the corresponding sites, and the resulting plasmid, pET-pck, was used to transform E. coli BL21-CodonPlus(DE3)-RIL. The recombinant E. coli strain was grown in Luria-Bertani medium containing ampicillin at 37°C, and the expression of pck_Tk was induced by the addition of 0.1 mM isopropyl-β-d-thiogalactopyranoside (IPTG). After further incubation for 2 h at 37°C, the cells were harvested by centrifugation and resuspended in 100 mM imidazole-HCl buffer (pH 7.0). The cell suspension was then sonicated and centrifuged (5,000 × g, 15 min). The soluble cell extract was heat precipitated for 20 min at 80°C, followed by centrifugation. The supernatant was applied to a Resource Q anion-exchange column (6 ml) (Amersham Biosciences), and the recombinant protein was eluted with a linear gradient of NaCl (250 to 300 mM) in 100 mM imidazole-HCl buffer (pH 7.0). The resulting solution was applied to a CHT-I hydroxyapatite column (2 ml) (Bio-Rad, Hercules, Calif.), and Pck_Tk was eluted with a linear gradient of phosphate (200 to 250 mM) in potassium phosphate buffer (pH 7.0). This solution was then applied to a Superdex 200 HR 10/30 gel filtration column (Amersham Biosciences) with a mobile phase of 100 mM imidazole-HCl buffer (pH 7.0) containing 150 mM NaCl at a flow rate of 0.25 ml/min. The molecular mass was calibrated with standard proteins of thyroglobulin (669 kDa), ferritin (440 kDa), catalase (232 kDa), aldolase (158 kDa), albumin (67 kDa), ovalbumin (43 kDa), chymotrypsinogen A (25 kDa), and RNase A (13.7 kDa). The protein concentration was determined with a protein assay kit (Bio-Rad) according to the instructions from the manufacturer by using bovine serum albumin as a standard.

Enzyme assays.

PCK activities were assayed by different procedures according to the direction of the reaction. OAA formation from PEP was generally determined by monitoring the oxidation of NADH coupled with thermostable malate dehydrogenase (MDH) from Thermus sp. (Sigma, St. Louis, Mo.) at 60°C. The decrease in absorbance at 340 nm was measured using a UV-1600 spectrophotometer (Shimadzu, Kyoto, Japan). The reaction mixture (500 μl) was composed of 50 mM NaHCO₃, 20 mM MgCl₂, 5 mM MnCl₂, 4 mM PEP, 2 mM GDP, 0.2 mM NADH, and 3 U of MDH from Thermus sp. in 100 mM imidazole-HCl buffer (pH 7.0), and the reaction was initiated by addition of the enzyme after preincubation at 60°C. A discontinuous assay was also adopted to determine PCK activities at very high temperatures, at which even MDH from Thermus sp. was inactivated (above 80°C). In this case, the PCK reaction was first performed in a mixture (100 μl) containing 50 mM NaHCO₃, 5 mM MnCl₂, 10 mM PEP, and 2 mM GDP in 100 mM imidazole-HCl buffer (pH 7.0) for 1 min at the appropriate temperatures and terminated by the addition of 100 μl of 250 mM EDTA on ice. Then 800 μl of 100 mM imidazole-HCl buffer (pH 7.0) containing 0.2 mM NADH and 3 U of yeast MDH (Oriental Yeast, Tokyo, Japan) was added to the reaction mixture. After saturation of the MDH reaction at 37°C, the absorbance at 340 nm was measured. As the product, OAA, is a thermolabile compound, we obtained beforehand the first-order constants of OAA decomposition under each condition and used them to correct the apparent quantification of the observed absorbance (6). The reverse reaction, PEP formation from OAA, was determined by following a counterreaction against OAA consumption by PCK, which tends to disturb the preestablished equilibrium of the MDH reaction between OAA and malate. The reaction mixture (500 μl), composed of 20 mM l-malate, 1 mM GTP, 20 mM MgCl₂, 5 mM MnCl₂, 12 mM NAD⁺, and 3 U of MDH from Thermus sp. in 100 mM imidazole-HCl buffer (pH 7.0), was equilibrated at 60°C, and then the enzyme was added. The reduction of NAD⁺ accompanied by the MDH-catalyzed formation of OAA compensating for the PCK reaction was continuously monitored at a wavelength of 340 nm at 60°C. Since, as we confirmed, more than 80% of the initial amount of OAA was maintained during the measurements, the thermal decomposition of OAA did not affect the accuracy of this assay. The optimum pH for Pck_Tk was determined by a discontinuous assay of OAA formation using 50 mM MES (morpholineethanesulfonic acid)-NaOH (pH 5.5 to 6.5), 100 mM imidazole-HCl (pH 6.5 to 7.5), 50 mM HEPES-NaOH (pH 7.5 to 8.0), or 50 mM bicine-NaOH (pH 8.0 to 9.0). For PCK activity measurements in T. kodakaraensis KOD1, the cells were harvested at the mid-log growth phase (optical density at 660 nm, ∼0.2 to ∼0.4), resuspended in 100 mM imidazole-HCl buffer (pH 7.0), and then sonicated. The soluble cell extract after centrifugation (5,000 × g, 15 min) was used for the continuous assay of OAA formation. The pH values of all buffers were adjusted at each assay temperature.

RNA isolation and Northern blot analysis.

Total RNA from T. kodakaraensis KOD1 was isolated from cells harvested at the early log growth phase (optical density at 660 nm, ∼0.1 to ∼0.2) with an RNeasy Midi kit (QIAGEN). For the Northern blot analysis, 15 μg of total RNA was denatured by heat treatment at 65°C for 15 min, separated by 1% agarose gel electrophoresis, and transferred onto a positively charged nylon membrane (Roche Diagnostics, Basal, Switzerland) by vacuum blotting. Labeling of DNA fragments with digoxigenin (Roche Diagnostics), hybridization, and detection of signals were performed according to the instructions from the manufacturer. A partial fragment of pck_Tk and the DNA ligase gene (lig_Tk) (15) from T. kodakaraensis KOD1 were used as probes.

Nucleotide sequence accession number.

The nucleotide sequence data reported here is available in the EMBL, GenBank, and DDBJ nucleotide sequence databases under accession number AB167819.

RESULTS

Primary structure of Pck_Tk.

In the preliminary draft of the complete genome of T. kodakaraensis strain KOD1, we identified one open reading frame (ORF) (designated pck_Tk) encoding a protein of 623 amino acids (72,036 kDa), of which the deduced amino acid sequence shared approximately 30 to 35% identities with animal and bacterial GTP-dependent PCKs. For example, the translated product showed 34, 32, and 31% overall identities to GTP-PCKs from human mitochondria (12), human cytosol (3), and C. glutamicum (1), respectively. Among the archaea whose genome sequences have been revealed, highly conserved homologs were present in the closely related genus Pyrococcus (84 to 86% identical in amino acid sequences). Moderately (50 to 51% identical) and weakly (33 to 35% identical) homologous genes could also be identified in the genera Sulfolobus and Thermoplasma, respectively, whereas no related gene was found in other archaea, including methanogens. An unrooted phylogenetic tree of GTP-PCKs together with ATP-PCKs from various sources was constructed by ClustalW. Figure 1A indicates that the archaeal homologs are included in the domain of GTP-dependent enzymes but only distantly related to their animal and bacterial counterparts. As reported previously, the PCK from the amitochondrial eucaryote Giardia intestinalis was interestingly located in the archaeal cluster (26). In contrast, PCKs from Thermoplasma spp. were positioned in the bacterial branch, and the aerophilic hyperthermophile Aeropyrum pernix was the only archaeon possessing an ATP-dependent PCK.

FIG. 1.

(A) Radial phylogenetic tree of representative bacterial, eucaryotic, and archaeal PCKs. The PCK sequences from the following organisms are from GenBank and have the indicated accession numbers: Aeropyrum pernix, Q9YG68; Arabidopsis thaliana, T05900; Ascaris suum, Q05893; Bacillus subtilis, P54418; Caenorhabditis elegans, T24168; Campylobacter jejuni, Q9PP01; chicken cytosolic, P05153; chicken mitochondrial, P21642; Chlamydia trachomatis, O84716; Chlamydophila pneumoniae, Q9Z755; Chlorobium limicola, S30180; Corynebacterium glutamicum, Q9AEM1; Drosophila melanogaster, P20007; Escherichia coli, P22259; Giardia intestinalis, EAA41306; human cytosolic, P35558; human mitochondrial, S69546; mouse cytosolic, Q9Z2V4; Mycobacterium tuberculosis, P96393; Neocallimastix frontalis, P22130; Pseudomonas aeruginosa, Q9HTZ7; Pyrococcus abyssi, Q9UY53; Pyrococcus furiosus, Q8U410; Pyrococcus horikoshii, O58058; Ralstonia solanacearum, Q8Y3G3; Rhodopseudomonas palustris, Q9ZNH4; Saccharomyces cerevisiae, P10963; Streptomyces coelicolor, Q93JL5 Sulfolobus solfataricus, Q97VS5; Sulfolobus tokodaii, BAB66087; Thermococcus kodakaraensis, AB167819; Thermoplasma acidophilum, Q9HLV2; Thermoplasma volcanium, P58306; Treponema pallidum, O83159; Trypanosoma cruzi, P51058; Vibrio parahaemolyticus, Q87TE1; Zea mays, Q9SLZ0. (B) Partial alignments of GTP-dependent PCK sequences. The gray boxes indicate amino acid residues that bound to the cations and substrates identified in human cytosolic GTP-PCK. The numbers correspond to amino acid positions in Pck_Tk human PCK. Amino acids that are fully conserved are indicated by asterisks, and those that are physicochemically similar are indicated by dots. Hyphens in the sequences represent gaps introduced to maximize alignment.

Despite the low overall identities between the primary structures of GTP- and ATP-PCKs, these enzymes have been known to contain highly conserved PCK-specific kinase 1a (P-loop) and kinase 2 regions that likely play common roles in the two classes of PCKs (11). Recently, the crystal structure of human cytosolic GTP-PCK has been elucidated (3). The structure revealed the importance of several residues within the conserved regions for binding to the substrate and metals and the presence of a unique GTP-binding pocket consisting of three Phe residues. As indicated by the partial alignments of GTP-PCKs (Fig. 1B), the archaeal PCKs possessed PCK-specific kinase 1a and kinase 2 regions similar to those in eucaryal and bacterial enzymes. The functionally important residues (Fig. 1B), such as Lys290 for phosphoryl transfer and Tyr235 for binding to the carboxylate of PEP (3), were also highly conserved among all species. In contrast, the predicted GTP-binding region in the archaeon-type enzymes was unique compared to those in bacterial and eucaryal GTP-PCKs. The most striking difference is the absence of one of the important Phe residues, Phe530 from human PCK. This Phe residue, which has been reported to sandwich the guanine ring together with Phe517 by π-electron interaction, was replaced by a nonaromatic Leu in PCKs from T. kodakaraensis, Pyrococcus spp., and Giardia or Lys in the Sulfolobus enzymes. PCKs from Thermoplasma spp. possess a GTP-binding region similar to those found in bacteria and eucarya, consistent with the result from the phylogenetic analysis.

Purification and characterization of recombinant Pck_Tk.

To characterize GTP-PCK from T. kodakaraensis, the pck_Tk gene was overexpressed in E. coli BL21-CodonPlus(DE3)-RIL under the control of the T7 promoter. The recombinant protein, expressed in a soluble form, was purified to apparent homogeneity by heat treatment at 80°C for 20 min followed by anion-exchange, hydroxyapatite, and gel filtration chromatographies (Fig. 2A). The recombinant Pck_Tk actually showed GTP-dependent activity, as was expected from the primary structure; that is, it produced OAA from PEP in the presence of NaHCO₃ and GDP with a specific activity of 11.0 U/mg at 37°C. The molecular mass of the recombinant Pck_Tk subunit was determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis to be 70 kDa, which agreed with the mass deduced from the amino acid sequence (72 kDa). Gel filtration chromatography demonstrated that the native molecular mass of Pck_Tk was 284 kDa. This mass indicated that Pck_Tk had a homotetrameric conformation, unlike other known GTP-PCKs which function as monomeric enzymes. The pH and temperature dependencies of Pck_Tk were determined by a discontinuous assay. Pck_Tk was a thermostable enzyme with an optimal pH of 7.0 and an optimal temperature of 80°C (Fig. 2B and C, respectively), and the half-life at the optimal temperature was determined to be 53 min (Fig. 2D).

FIG. 2.

Purification and pH and temperature dependencies of recombinant Pck_Tk. (A) Results of sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis of purified recombinant Pck_Tk. Lane 1, soluble fraction after sonication; lane 2, soluble fraction after heat treatment at 80°C for 20 min; lane 3, fraction after anion-exchange chromatography with Resource Q; lane 4, fraction after hydroxyapatite chromatography with Bio-Scale CHT-I. (B) Effect of pH on enzyme activity. Triangles, 50 mM MES-NaOH (pH 5.5 to 6.5); squares, 100 mM imidazole-HCl (pH 6.5 to 7.5); diamonds, 50 mM HEPES-NaOH (pH 7.5 to 8.0); circles, 50 mM Bicine-NaOH (pH 8.0 to 9.0). (C) Effect of temperature on enzyme activity. (D) Thermostability of Pck_Tk. Incubation temperatures were 80°C (filled squares), 70°C (open circles), 60°C (filled triangles), 50°C (open squares), and 40°C (filled circles).

Kinetic analysis.

We then performed a kinetic analysis of Pck_Tk activity in both directions by continuous assays at 60°C using thermostable MDH. Pck_Tk showed typical Michaelis-Menten kinetics in both directions, indicating no homotropic allosteric regulation of the activity. As shown in Table 1, the enzyme displayed a significantly lower K_m for OAA (18.1 μM) than for PEP (131 μM), while the values for GDP and GTP were similarly low (19 to 36 μM). Although the V_max for PEP formation (44.4 U/mg) was approximately half of that for OAA formation (76.9 U/mg), the preference of this enzyme for OAA suggested a physiological importance for PEP formation in vivo. As seen in the case of other GTP-PCKs (5, 14, 20), ITP and IDP acted as alternative nucleotide cofactors with similar V_max values and slightly higher K_m values than those for GTP and GDP. ATP and ADP were poor cofactors for Pck_Tk but gave relatively high V_max values (10 to 30% of those with GTP and GDP) compared to those of classical GTP-PCKs (1, 14, 20, 24). However, the K_m values for ATP and ADP were much higher than those of their guanyl counterparts.

TABLE 1.

Kinetic analysis of the reaction of recombinant Pck_Tk to various substrates

Substrate	OAA formation		Substrate	PEP formation
	Km (μM)	Vmax (U/mg)		Km (μM)	Vmax (U/mg)
PEP	131	76.9	OAA	18.1	44.4
GDP	18.5	74.6	GTP	36.1	35.8
ADP	2,230	28.0	ATP	465	4.69
IDP	71.2	72.4	ITP	71.5	33.3

We investigated further the effects of various metabolic intermediates on the OAA formation by this enzyme. ATP, ADP, pyruvate, acetate, glycolytic metabolites (glucose 6-phosphate, fructose 6-phosphate, fructose 1,6-bisphosphate, dihydroxyacetone phosphate, glyceraldehyde 3-phosphate, 3-phosphoglycerate, or 2-phosphoglycerate), tricarboxylic acid cycle metabolites (fumarate, succinate, citrate, or 2-oxoglutarate), or amino acids (glutamate or aspartate) were added to the assay mixture at a final concentration of 1 or 10 mM. As a result, no striking effects were found with any of these compounds, although we observed a weak inhibition with pyruvate (40% decrease) and 2-oxoglutarate (30% decrease) at a high concentration (10 mM). Both inhibitors are 2-oxoacids structurally resembling OAA, and indeed it has been reported that pyruvate may bind to the ATP-PCK from E. coli as a substrate analog (28) and that 2-oxoglutarate competitively inhibits GTP-PCK from rat liver (29). Although the potential competitive inhibition was also seen in the case of Pck_Tk, the inhibitory effects were weak. These results suggested that, like most GTP-PCKs, the activity of Pck_Tk was not regulated by allosteric effects.

Effect of divalent cations on enzyme activity.

PCKs from all sources absolutely require divalent metal cations for their activity, regardless of their nucleotide dependencies. Further, it has been reported that a combination of Mg²⁺ and Mn²⁺ induces a synergistic effect on the activity (4, 9). We therefore examined the divalent metal cation dependency of the archaeal PCK (Table 2). In the presence of one divalent cation alone, Mn²⁺ gave the highest activity (69.9 U/mg). Mg²⁺ and Co²⁺ also supported the reaction, although the activities were 4.5 and 21%, respectively, of that with Mn²⁺. No activity could be detected with other divalent cations, such as Ca²⁺, Zn²⁺, Cu²⁺, Ni²⁺, and Sr²⁺. When Mg²⁺ was added as a second cation, the K_m values for Mn²⁺ in both directions of the reaction were markedly decreased to 21 to 22 μM from a few hundred μM observed with Mn²⁺ alone. Such a decrease in K_m values was also seen for Co²⁺ in the presence of Mg²⁺. These results indicated that Pck_Tk, like other PCKs (28), possessed dual metal-binding sites specific to Mn²⁺ and Mg²⁺. Mn²⁺ first binds to the preferred site with high affinity (K_m, 21 to 22 μM) and then binds to a second site, available in the absence of Mg²⁺. The higher K_m that we observed for Mn²⁺ alone was probably caused by the secondary binding site's affinity for Mn²⁺ being much lower than that of the primary site.

TABLE 2.

Kinetic analysis of the reaction of recombinant Pck_Tk to various divalent cations

Divalent cation(s)	OAA formation		PEP formation
	Km (μM)	Vmax (U/mg)	Km (μM)	Vmax (U/mg)
Mn2+	263	69.9	355	35.6
Mg2+	5,360	3.13	3,780	2.30
Co2+	752	14.7	775	16.8
Mn2+ + 20 mM Mg2+	20.5	66.3	22.4	32.6
Co2+ + 20 mM Mg2+	2.79	39.2	6.77	19.9

Expression profile.

To obtain information concerning the physiological function of Pck_Tk, the transcriptional profile of the gene was investigated by Northern blot analysis. T. kodakaraensis KOD1 was cultivated on different carbon sources (amino acids, pyruvate, starch, or pyruvate plus starch), and the total RNA was hybridized with a probe specific for pck_Tk. Here, the probe for the DNA ligase gene (lig_Tk) (15) was applied as a control, because the transcription of this gene is nearly constitutive (19). Positive signals corresponding to pck_Tk transcripts were clearly detected from the RNA of cells grown under all conditions examined, and no other considerable signal could be observed. As shown in Fig. 3A, the highest levels of transcription were observed in the cells grown on pyruvate, and the signal strength was higher under gluconeogenic conditions (amino acids or pyruvate) than under glycolytic conditions (starch or starch plus pyruvate). In this experiment, the signals corresponding to the control gene were constant irrespective of the carbon source. We carried out further enzyme assays with the cell extract of T. kodakaraensis (Fig. 3B). Apparently, the activities were higher in the cells grown under gluconeogenic conditions than in those grown under glycolytic conditions. The change in activity levels depending on the carbon source showed the same tendency as the transcription profile described above; for example, the highest activity was detected in the pyruvate-grown cells. These results raise the possibility that Pck_Tk mainly plays the role of providing PEP from OAA under gluconeogenic conditions. Moreover, the accordance of transcription and activity profiles suggested that Pck_Tk activity was regulated at the transcriptional level and not at the translational or posttranslational levels.

FIG. 3.

Transcription of pck_Tk and PCK activity in T. kodakaraensis KOD1 grown on various carbon sources. (A) Northern blot analysis of RNA from T. kodakaraensis KOD1 grown on amino acids plus sulfur (lane 1), pyruvate (lane 2), starch (lane 3), and pyruvate plus starch (lane 4). The pck_Tk (upper panel) and lig_Tk (DNA ligase gene) (lower panel) from strain KOD1 were used as probes. Each lane contains 15 μg of total RNA. (B) PCK activity in T. kodakaraensis grown on amino acids plus sulfur (column 1), pyruvate (column 2), starch (column 3), and pyruvate plus starch (column 4). The activity was determined by a continuous assay as described in Materials and Methods, except that MgCl₂ was omitted. It had been confirmed that the presence or absence of MgCl₂ with a saturated concentration of MnCl₂ resulted in little effect on the activity of Pck_Tk (see Table 2).

DISCUSSION

The archaeal PCK from the hyperthermophile T. kodakaraensis was a GTP-dependent enzyme with a preference for PEP formation rather than OAA formation, and it appeared to possess no allosteric property. ITP or IDP also supported the reaction as an alternative nucleotide cofactor, but the activities were poor with ATP or ADP (Table 1). The enzyme absolutely required Mg²⁺, Mn²⁺, or Co²⁺, while the K_m values for Mn²⁺ and Co²⁺ drastically decreased in the presence of saturated concentrations of Mg²⁺ (Table 2). These properties of Pck_Tk were comparable to those of many GTP-PCKs from mammals and bacteria; however, the homotetrameric assembly was a unique feature of this archaeal enzyme, in contrast to the monomeric nature of all other known GTP-PCKs (11).

Pck_Tk and the archaeal homologs from Pyrococcus and Sulfolobus spp., interestingly, together with a Giardia homolog, occupied a branch distinct from that of the eucaryal and bacterial GTP-PCKs in the phylogenetic tree (Fig. 1A). This coincided with the presence of an altered GTP-binding region lacking one of the important Phe residues (Fig. 1B), whose π-electron system stacked onto the guanine ring in the human enzyme (3). Nevertheless, the K_m value of Pck_Tk for GDP (18.5 μM) was within levels similar to those of its eucaryal and bacterial counterparts (1, 5, 14, 20). Another aromatic residue at a presently undefined position may compensate for the function of the missing Phe residue, or a different mode of interaction may exist in the nucleotide binding of the archaeon-type enzymes. Other catalytically important regions, including the PCK-specific kinase 1a and kinase 2 regions, could be identified in the archaeal enzymes as well as in known PCKs, regardless of nucleotide dependency. It has been reported that the optimal catalytic ability of PCKs can be attained with a combination of Mg²⁺ and Mn²⁺, which participate in complexing the nucleotide substrate and in activating PEP, respectively (28). The metal-interacting residues identified in the crystal structure of human PCK, such as Thr291 in the kinase 1a region for Mg²⁺ binding, Lys244 in the PCK-specific region, Asp311 in the kinase 2 region, and His264 for Mn²⁺ binding, were also conserved in Pck_Tk. The marked decrease of K_m values for Mn²⁺ and Co²⁺ in the presence of Mg²⁺ as a second divalent cation agreed well with the existence of dual binding sites specific for Mg²⁺ and Mn²⁺ (Co²⁺). The low K_m value for Mn²⁺ and Co²⁺ also coincided with the general intracellular concentrations of these metals (28).

PEP, OAA, and pyruvate are key metabolites acting as important nodes in carbon and energy metabolisms. Based on the preliminary complete genome sequence of T. kodakaraensis, the predicted pyruvate metabolism of this strain is illustrated in Fig. 4. The genes most likely to encode pyruvate kinase, PEP synthase, PCK, and malic enzyme were identified as those involved in the PEP-OAA-pyruvate cycle, while the gene for PEP carboxylase was absent from the genome. The genome analyses also revealed the presence of one ORF encoding a protein with significant homology to the carboxyltransferase domain or subunit of pyruvate carboxylase. However, another essential gene corresponding to the biotin carboxylase component of this enzyme could not be identified. We therefore measured the pyruvate carboxylase activity in the extract of T. kodakaraensis but could not detect any considerable activity (data not shown). These results suggested that pyruvate carboxylase seems to be absent from this archaeon. Although its function is still unclear, this ORF may encode an OAA decarboxylase, because the product also shared homology with the α-subunit of OAA decarboxylase. T. kodakaraensis lacks many members of the TCA cycle, suggesting that the cyclic interconversion among C₄, C₅, and C₆ compounds is incomplete in this strain. The distribution of the genes described above is nearly the same in the Pyrococcus strains, except for Pyrococcus furiosus, which possesses a gene cluster for citrate synthase, aconitase, and isocitrate dehydrogenase that is not found in T. kodakaraensis, Pyrococcus abyssi, and Pyrococcus horikoshii. As described above, Pck_Tk kinetically preferred PEP formation rather than OAA formation. The gene transcription and intracellular activity were detected at higher levels in the cells grown on amino acids or pyruvate and were suppressed, to some extent, in the presence of starch. These results suggest that the archaeal PCK plays a role in the priming of gluconeogenesis, like other PCKs in most cases. The higher levels of PCK in cells grown on amino acids imply that PCK functions in the entry of carbons derived from Asp and Asn into gluconeogenesis. In pyruvate-grown cells, on the other hand, it can be estimated that PEP is supplied from pyruvate directly by PEP synthase and not by PCK via OAA, because OAA formation from pyruvate is not linked due to the absence of pyruvate carboxylase. Such distinguishable roles of PCK and PEP synthase in gluconeogenic PEP formation have also been reported for E. coli, which lacks pyruvate carboxylase (22). Since we adopted a complex medium containing amino acids (yeast extract and peptone) together with pyruvate, the expression of pck_Tk in the pyruvate-grown cells was initially assumed merely to reflect that the enzyme still functioned in PEP formation from the amino acids present in the medium, in addition to the major pathway from pyruvate by PEP synthase. However, the transcription and intracellular activity of PCK in the cells on pyruvate were higher than those in the cells grown on amino acids alone. This result may be explained by an anaplerotic role of Pck_Tk, which is to maintain the intracellular PEP concentration within an optimum range when excess PEP is produced from pyruvate by PEP synthase. Although the enzyme was inhibited by pyruvate and 2-oxoglutarate, the weak effects observed even with high concentrations imply that the regulatory roles of these 2-oxoacids are unlikely. Under the glycolytic conditions, we have previously reported that the expression of gluconeogenic fructose 1,6-bisphosphatase was completely suppressed in T. kodakaraensis (19). In contrast, although lower than the levels observed under gluconeogenic conditions, a considerable amount of pck_Tk expression could be detected even in the presence of starch. Along with the absence of allosteric regulation of the activity by glycolytic intermediates, these observations commonly support an anaplerotic function of PCK as an alternative to PEP carboxylase, which presumably is missing from this archaeon.

FIG. 4.

Predicted pyruvate metabolism in Thermococcus kodakaraensis KOD1. ACS, acetyl-CoA synthetase; AlaAT, alanine aminotransferase; AspAT, aspartate aminotransferase; Fd, ferredoxin; GDH, glutamate dehydrogenase; MEZ, malic enzyme; OAD, oxaloacetate decarboxylase; PCK, phosphoenolpyruvate carboxykinase; PPS, phosphoenolpyruvate synthase; POR, pyruvate:ferredoxin oxidoreductase; PYK, pyruvate kinase.

This is the first report on a PCK from the third domain of life, Archaea. The enzyme seems to play an active role in the interconversion of PEP, pyruvate, and OAA, most likely in coordination with other enzymes that await examination. Recently, a target gene disruption system for T. kodakaraensis has been successfully developed (23). Further detailed analyses of knockout strains with defined medium containing a sole carbon source will contribute to elucidating the unique metabolic features of hyperthermophilic archaea.