|Home | About | Journals | Submit | Contact Us | Français|
The pathway of glucose degradation in the thermoacidophilic euryarchaeon Picrophilus torridus has been studied by in vivo labeling experiments and enzyme analyses. After growth of P. torridus in the presence of [1-13C]- and [3-13C]glucose, the label was found only in the C-1 and C-3 positions, respectively, of the proteinogenic amino acid alanine, indicating the exclusive operation of an Entner-Doudoroff (ED)-type pathway in vivo. Cell extracts of P. torridus contained all enzyme activities of a nonphosphorylative ED pathway, which were not induced by glucose. Two key enzymes, gluconate dehydratase (GAD) and a novel 2-keto-3-deoxygluconate (KDG)-specific aldolase (KDGA), were characterized. GAD is a homooctamer of 44-kDa subunits, encoded by Pto0485. KDG aldolase, KDGA, is a homotetramer of 32-kDa subunits. This enzyme was highly specific for KDG with up to 2,000-fold-higher catalytic efficiency compared to 2-keto-3-deoxy-6-phosphogluconate (KDPG) and thus differs from the bifunctional KDG/KDPG aldolase, KD(P)GA of crenarchaea catalyzing the conversion of both KDG and KDPG with a preference for KDPG. The KDGA-encoding gene, kdgA, was identified by matrix-assisted laser desorption ionization-time of flight (MALDI-TOF) mass spectrometry (MS) as Pto1279, and the correct translation start codon, an ATG 24 bp upstream of the annotated start codon of Pto1279, was determined by N-terminal amino acid analysis. The kdgA gene was functionally overexpressed in Escherichia coli. Phylogenetic analysis revealed that KDGA is only distantly related to KD(P)GA, both enzymes forming separate families within the dihydrodipicolinate synthase superfamily. From the data we conclude that P. torridus degrades glucose via a strictly nonphosphorylative ED pathway with a novel KDG-specific aldolase, thus excluding the operation of the branched ED pathway involving a bifunctional KD(P)GA as a key enzyme.
Comparative analyses of sugar-degrading pathways in members of the domain Archaea revealed that all species analyzed so far degrade glucose and glucose polymers to pyruvate via modification of the classical Embden-Meyerhof (EM) and Entner-Doudoroff (ED) pathways found in bacteria and eukarya. Modified EM pathways were reported for hyperthermophilic archaea, including, e.g., the strictly fermentative Thermococcales and Desulfurococcales, the sulfur-reducing Thermoproteus tenax, and the microaerophilic Pyrobaculum aerophilum. These pathways differ from the classical EM pathway by the presence of several novel enzymes and enzyme families, catalyzing, e.g., the phosphorylation of glucose and fructose-6-phosphate, isomerization of glucose-6-phosphate, and oxidation of glyceraldehyde-3-phosphate (18, 22, 25).
Modified ED pathways have been proposed for aerobic archaea, including halophiles, and thermoacidophilic crenarchaea, such as Sulfolobus species, and the euryarchaea Thermoplasma acidophilum and Picrophilus torridus. The anaerobic Thermoproteus tenax, which degrades glucose predominantly via a modified EM pathway, also utilizes—to a minor extent (<20%)—a modified ED pathway for glucose degradation. The following ED pathway modifications have been reported in archaea (25). A semiphosphorylative ED pathway was reported in halophilic archaea. Accordingly, glucose is converted to 2-keto-3-deoxy-6-gluconate (KDG) via glucose dehydrogenase and gluconate dehydratase. KDG is then phosphorylated by KDG kinase to 2-keto-3-deoxy-6-phosphogluconate (KDPG), which is split by KDPG aldolase to pyruvate and glyceraldehyde-3-phosphate (GAP). GAP is further converted to form another pyruvate via common reactions of the EM pathway, i.e., phosphorylative GAP dehydrogenase, phosphoglycerate kinase, phosphoglycerate mutase, enolase, and pyruvate kinase. The net ATP yield of this pathway is 1 ATP/mol glucose.
From initial enzyme studies of the thermoacidophilic archaea Sulfolobus solfataricus, Thermoplasma acidophilum, and Thermoproteus tenax, a nonphosphorylative ED pathway was proposed (25). In this modification of the ED pathway, glucose is converted to KDG via glucose dehydrogenase and gluconate dehydratase, as in the semiphosphorylative pathway, but then the steps differ as follows: KDG is cleaved into pyruvate and glyceraldehyde via 2-keto-3-deoxygluconate-specific aldolase (KDGA). The subsequent oxidation of glyceraldehyde to glycerate involves either NAD(P)+-dependent dehydrogenases or oxidoreductases. Glycerate is then phosphorylated by a specific kinase to 2-phosphoglycerate, which is finally converted to pyruvate via enolase and pyruvate kinase. This modification of the ED pathway was called “nonphosphorylative” since it is not coupled with net ATP synthesis.
However, recent comparative genomic studies and refined enzyme analyses suggest that the crenarchaea Sulfolobus and Thermoproteus utilize a so-called branched ED pathway, in which a semiphosphorylated route is simultaneously operative in addition to the nonphosphorylative route (25, 32). Accordingly, the semiphosphorylated route involves—via KDG kinase—the phosphorylation of KDG to KDPG, which is then cleaved to pyruvate and GAP by means of a bifunctional KDG/KDPG aldolase, KD(P)GA. GAP is then converted to another pyruvate via nonphosphorylative GAP dehydrogenase (GAPN), phosphoglycerate mutase, enolase, and pyruvate kinase. The net ATP yield of the branched ED pathway is zero. In support of this pathway, the genes encoding gluconate dehydratase, bifunctional KD(P)GA, KDG kinase, and GAPN were found to be clustered in Sulfolobus solfataricus (see Discussion) and Thermoproteus tenax. The key enzyme of the proposed branched ED pathway is the bifunctional KD(P)GA, which catalyzes the cleavage of KDG to pyruvate and glyceraldehyde and cleavage of KDPG to pyruvate and glyceraldehyde-3-phosphate. This bifunctional aldolase, which has been characterized from S. solfataricus, was found to be identical to a previously described KDG aldolase of the same organism; however, its catalytic property to also utilize KDPG as a substrate has been recognized only recently. In fact, the bifunctional KD(P)GA showed a higher catalytic efficiency for KDPG than for KDG (1, 14). Crystal structures of bifunctional KD(P)GAs of S. solfataricus and T. tenax have been reported (16, 27, 30; G. Taylor [United Kingdom], unpublished data).
The branched ED pathway in S. solfataricus has been reported to be promiscuous and therefore represents an equivalent degradation route for both glucose and its C-4 epimer, galactose. Accordingly, glucose dehydrogenase, gluconate dehydratase, KDG kinase, and bifunctional KD(P)GA were found to catalyze the conversion of both glucose and galactose and the corresponding subsequent intermediates, i.e., gluconate/galactonate, KDG/KDGal (KDGal stands for 2-keto-3-deoxygalactonate), and KDPG/KDPGal (KDPGal stands for 2-keto-3-deoxy-6-phosphogalactonate) (4, 12-14).
In contrast to crenarchaea, the modified ED pathway in the thermoacidophilic euryarchaea Thermoplasma acidophilum and Picrophilus torridus has not been studied in detail. Enzyme measurements in cell extracts and the characterization of few enzymes suggest the operation of a nonphosphorylative ED pathway in these organisms (2, 3, 17, 19, 25). However, in vivo evidence for the operation of an ED-type pathway, e.g., by 13C-labeling experiments with growing cultures, has not been provided yet. Furthermore, the KDG aldolase activity measured in cell extracts of P. torridus and T. acidophilum has not been purified and characterized, in particular with respect to substrate specificity, and the genes encoding these enzymes have not been identified. The biochemical analysis of this aldolase is crucial to define the enzyme as a KDG-specific aldolase, indicative of a nonphosphorylative ED pathway, or as bifunctional KD(P)GA, indicative of the branched ED pathway as proposed for the crenarchaea Sulfolobus and Thermoproteus.
In this communication we studied the sugar-degrading pathway in P. torridus by in vivo labeling experiments with [13C]glucose, by enzyme measurements, and by characterization of two key enzymes, gluconate dehydratase and KDG aldolase. The data indicate that P. torridus utilizes a strict nonphosphorylative ED pathway, involving a novel KDG-specific aldolase as a key enzyme, and thus exclude the operation of a branched ED pathway, as in crenarchaea involving a bifunctional KD(P)GA as a key enzyme.
Picrophilus torridus (7, 21) was routinely grown aerobically at pH 0.9 and 60°C in 100-ml Erlenmeyer flasks filled with 20 ml medium containing 25 mM glucose and 0.2% yeast extract (19, 24) and shaken at 150 rpm. For mass culturing, cells were grown in a 8-liter fermentor (FairmenTec, Germany) (stirred at 200 rpm) filled with 5 liters of medium. Sulfolobus acidocaldarius was grown at 70°C on a synthetic medium with 25 mM glucose as a substrate as described previously (23). Growth was monitored by measuring the optical density at 600 nm (OD600). Glucose consumption was determined enzymatically with hexokinase and glucose-6-phosphate dehydrogenase.
To identify the glucose degradation pathways in vivo, experiments with [d-13C]glucose were performed with growing cultures in 100-ml Erlenmeyer flasks filled with 20 ml medium. In the case of P. torridus, the medium contained [1-13C]- or [3-13C]glucose (25 mM each) and yeast extract (0.1%). The pH was adjusted to pH 0.3. Cells were harvested in late log phase, when significant amounts of glucose had been consumed. In the case of S. acidocaldarius, the synthetic medium (pH 2.5) contained [1-13C]- or [3-13C]glucose (25 mM each). Cell aliquots were harvested in mid-exponential growth phase. For both P. torridus and S. acidocaldarius, 2-ml portions of culture broth were centrifuged at 8,000 × g and 10°C for 10 min. The dry biomass pellet was hydrolyzed in 1.5 ml of 6 M HCl for 24 h at 110°C in a sealed 2-ml Eppendorf tube and desiccated overnight in a heating block at 85°C under a constant air stream. The hydrolysate was dissolved in 50 μl of 99.8% dimethyl formamide and transferred into a new Eppendorf cup within a few seconds. For derivatization, 30 μl of N-methyl-N-(tert-butyldimethylsilyl)-trifluoroacetamide was added, which readily silylates hydroxyl groups, thiols, primary amines, amides, and carboxyl groups (5), and the mixture was incubated at 550 rpm and 85°C for 60 min. One microliter of the derivatized sample was injected into a 6890N Network gas chromatograph (GC) system, combined with a 5975 inert XL mass selective detector (Agilent Technologies) and analyzed as described earlier (6, 31). The GC temperature profile was 160°C for 1 min, increase to 320°C at 20°C per minute, and hold at 320°C for 1 min. The injector temperature was set at 230°C, the split ratio was 1:10, the flow rate was 1.5 ml/min, and the carrier gas was helium in a HP-5MS column (30 m by 0.25 mm; 0.25 μm coated) (Agilent Technologies). The mass spectra of the derivatized amino acid alanine were corrected for the natural abundance of all stable isotopes and unlabeled biomass from inoculum. The labeling pattern of alanine is a direct and quantitative evidence for metabolic pathways leading from glucose to pyruvate.
Enzyme activities were assayed spectrophotometrically in 1 ml of assay mixture containing cell extracts of P. torridus grown on glucose and yeast extract. Cells were harvested in the late exponential growth phase at an OD600 of about 1.6 (see Fig. Fig.1).1). To determine a possible glucose-specific induction, the specific activities were also measured in cells after growth in the absence of glucose on medium containing yeast extract (0.2%). Exponentially grown cells, harvested at an OD600 of 0.9, were used. One unit of enzyme activity is defined as 1 μmol substrate consumed or product formed per min. KDG was prepared enzymatically from gluconate with purified gluconate dehydratase from P. torridus. KDG was quantified by the thiobarbituric acid assay (29). Glucose dehydrogenase was tested by the method of Reher and Schönheit (19). Gluconate dehydratase was determined at 60°C by measuring the conversion of gluconate to KDG, which was quantified as described previously (29). The assay mixture contained 50 mM sodium phosphate buffer (pH 6.2), 7.5 mM gluconate, and 20 mM MgCl2. KDG aldolase was tested at 60°C both in the direction of aldol (KDG) cleavage and aldol (KDG) synthesis: KDG cleavage was analyzed by measuring pyruvate formation from KDG using lactate dehydrogenase as described previously (12). The assay mixture contained 50 mM sodium phosphate buffer (pH 6.2), 2 to 5 mM KDG, 0.3 mM NADH, and 5.5 U lactate dehydrogenase. The formation of KDG from pyruvate and d/l-glyceraldehyde was monitored at 60°C over a period of 20 min and quantified in the thiobarbituric acid assay at 546 nm (29). The assay mixture contained 50 mM sodium phosphate buffer (pH 6.2), 20 mM pyruvate, and 20 mM glyceraldehyde. KDPG aldolase was determined as described above for KDG aldolase, except that KDG was replaced by KDPG in the cleavage direction and glyceraldehyde was replaced by glyceraldehyde-3-phosphate in the direction of aldol synthesis. KDG aldolase and KDPG aldolase activities were also determined with cell extracts of glucose-grown T. acidophilum and S. solfataricus. Glyceraldehyde dehydrogenase, enolase, and pyruvate kinase were determined as described in reference 19.
For gluconate dehydratase (GAD) purification, fermentor-grown cells (18 g [wet weight]) were harvested in the late exponential growth phase by centrifugation, suspended in 50 mM Tris-HCl (pH 8.1) and 10 mM MgCl2, and disrupted by passage through a French pressure cell at 1.3 × 108 Pa. Cell debris was removed by centrifugation for 90 min at 50,000 × g. The supernatant was adjusted to 3 M (NH4)2SO4 and incubated at 4 to 8°C for 17 h. The precipitate was removed by centrifugation at 50,000 × g for 90 min. The supernatant was adjusted to a pH of 8 and to 2 M (NH4)2SO4 and applied to a phenyl-Sepharose 26/10 column equilibrated with 100 mM Tris-HCl adjusted to a pH of 8.1 with 2 M (NH4)2SO4 and 10 mM MgCl2. Protein was desorbed by a linear gradient from 2 M to 0 M (NH4)2SO4. The fractions with the highest GAD activity eluting at 1.3 to 0.7 M (NH4)2SO4 were pooled, dialyzed against 100 mM Tris-HCl (pH 8) and 10 mM MgCl2, and applied to a UnoQ1 column equilibrated with the same buffer. The protein was eluted with a increasing gradient from 0 to 2 M NaCl. The fractions containing the highest GAD activity (0.2 to 0.4 M NaCl) were pooled and concentrated to 1,000 μl by ultrafiltration (cutoff, 10 kDa). The concentrated protein solution was applied to a Superdex 200 HiLoad 16/60 column equilibrated with 50 mM Tris-HCl (pH 7.1) and 150 mM NaCl. The protein was eluted at a flow rate of 1 ml/min. At this stage, GAD was essentially pure as judged by SDS-PAGE.
The gene encoding GAD was identified by matrix-assisted laser desorption ionization-time of flight (MALDI-TOF) mass spectrometry (MS) of the purified protein (44-kDa band by SDS-PAGE) as reported previously (17, 20). The pH dependence of gluconate dehydratase was measured at 60°C between pH 4.0 and pH 7.5 using either 0.1 M sodium acetate (pH 4.0 to 6.0) or 0.1 M sodium phosphate (pH 6.2 to 7.5) as a buffer. The temperature dependence was determined between 49°C and 83°C. The long-term thermostability was tested in sealed vials containing 6 μg protein in 100 μl sodium phosphate buffer (pH 6.2) with 50 mM MgCl2, which were incubated at 60°C, 70°C, 80°C, and 90°C for up to 120 min. The vials were cooled for 10 min, and the remaining activity was tested. The substrate specificity was tested at 60°C in 50 mM sodium phosphate buffer (pH 6.2). Kinetic constants were determined for gluconate and galactonate in the presence of 20 mM MgCl2, and kinetic constants for MgCl2 were deternined in the presence of 10 mM gluconate.
For KDGA purification, glucose-grown cells were harvested in the late exponential growth phase by centrifugation. The cells (15 g [wet weight]) were suspended in 50 mM Tris-HCl (pH 8.1) and disrupted by passage through a French pressure cell. Cell debris was removed by centrifugation at 50,000 × g for 90 min. The supernatant was adjusted to 1 M (NH4)2SO4 and applied to a phenyl-Sepharose 26/10 column equilibrated with 100 mM Tris-HCl (pH 8) and 1 M (NH4)2SO4. Protein was eluted by a linear gradient from 1 M to 0 M (NH4)2SO4. Fractions containing the highest KDGA activity [eluting at 0.3 to 0.1 M (NH4)2SO4] were diluted 30-fold in 50 mM Tris-HCl (pH 7.6), and applied to a UnoQ5 column (5 ml), which was equilibrated with the same buffer. Bound protein was eluted with a linear gradient up to 1 M NaCl. Fractions containing the KDGA activity (eluting at 0.05 to 0.15 M NaCl) were concentrated by ultrafiltration (cutoff, 10 kDa) and applied to a Superdex 200 HiLoad 16/60 column equilibrated with 50 mM Tris-HCl (pH 7.1) containing 150 mM NaCl. The protein was eluted with the same buffer. Fractions containing the highest KDGA activity were diluted in 0.1 M Tris-HCl (pH 8.0) and applied to a UnoQ1 column equilibrated with 100 mM Tris-HCl (pH 8). KDGA was eluted with an increasing gradient up to 0.5 M NaCl. At this stage, KDGA was essentially pure as judged by SDS-PAGE, yielding a single protein band at 32 kDa. After in-gel digestion of this band with trypsin, the eluted peptides were analyzed by MALDI-TOF mass spectrometry and used to identify the KDGA-encoding gene as described previously (17, 20). In addition, N-terminal amino acid sequencing of the purified enzyme was performed by the method of Meyer et al. (15).
The pH dependence and temperature dependence of KDGA were determined using the same assay conditions as described for GAD. KDGA activity was analyzed in the direction of KDG formation. The substrate specificity of KDG aldolase was tested at 60°C in 50 mM sodium phosphate buffer (pH 6.2). Kinetic constants for KDG and KDPG were determined using substrate concentrations up to 1 mM and 32 mM, respectively. In the direction of aldol formation, apparent Km and Vmax values were determined for glyceraldehyde and glyceraldehyde-3-phosphate with 25 mM pyruvate, and the values were determined for pyruvate at 10 mM glyceraldehyde. Other aldehyde substrates, i.e., glycolaldehyde, d-ribose, d-xylose, l-arabinose, d-arabinose, acetaldehyde, and crotonaldehyde, were each tested at 25 mM in the presence of 25 mM pyruvate. The (aldol) condensation product formed by KDGA from d- or l-glyceraldehyde and pyruvate was identified by the method of Lamble et al. (12), with the following modifications: 80 mM (each) d-glyceraldehyde or l-glyceraldehyde was mixed with 160 mM sodium pyruvate in 250 μl of water containing 5 μg P. torridus KDGA. The reaction mixture was heated at 50°C overnight in a shaking incubator. Samples were analyzed by high-performance liquid chromatography (HPLC) using an Aminex HPX-87H column (Bio-Rad) using 1 M formic acid as the eluent and linked to a refractive index detector.
On the basis of MALDI-TOF MS analysis and of the N-terminal amino acid sequence, a single open reading frame (ORF), Pto1279 (see Results), was identified in the sequenced genome of P. torridus. The ORF was characterized as the kdgA gene, encoding 2-keto-3-deoxygluconate aldolase, by its functional overexpression in E. coli as follows. The gene was amplified from genomic DNA of P. torridus by PCR and cloned into pET19b via two restriction sites (NdeI and BamHI) created with the primers 5′-GAATTCATATGTACAAGGGTATAGTATG-3′ and 5′-GAT TAGGATCCAAAATATTAATTTATATTTCAA-3′ (restriction sites are underlined). The recombinant plasmid pET19b-kdgA was transferred into E. coli BL21 CodonPlus (DE3)-RIL cells. Transformed cells were grown in Luria-Bertani medium at 37°C, and kdgA expression was induced by the addition of 1 mM isopropyl-1-thio-β-d-galactopyranoside (IPTG). After 4 h, cells were harvested by centrifugation, followed by resuspension in 50 mM Tris-HCl (pH 8.2) containing 300 mM NaCl and 5 mM imidazole. Cells were disrupted by passage through a French pressure cell. After centrifugation, the supernatant was incubated at 58°C for 45 min and centrifuged at 100,000 × g for 1 h. The supernatant, exhibiting KDG aldolase activity, was applied to a Ni-nitrilotriacetic acid (Ni-NTA) column equilibrated with 50 mM Tris-HCl (pH 8.2) containing 300 mM NaCl and 5 mM imidazole. Bound protein was specifically eluted with increasing imidazole concentrations, yielding pure KDGA as judged by a single protein band at 33 kDa on SDS-polyacrylamide gels.
Growth of P. torridus on glucose requires low concentrations of yeast extract. With 25 mM glucose and 0.2% yeast extract and at pH 0.9, the cells grew with a doubling time of about 10 h up to a final cell density with an OD600 of 1.6. In the absence of glucose, the cells grew on yeast extract with a similar doubling time to a cell density with an OD600 of 1. As shown in Fig. Fig.1,1, significant glucose consumption occurred only in the later growth phase. This indicates that initial growth was based almost exclusively on components of the yeast extract which apparently prevented glucose consumption.
To analyze the sugar-degrading pathway of P. torridus by in vivo 13C-labeling experiments (31), cells were grown on 13C-labeled glucose at pH 0.3 at a reduced yeast extract concentration of 0.1%. Under these conditions, the cells grew on yeast extract up to an OD600 of 0.3 followed by the phase of glucose consumption up to an OD600 of 0.8 (not shown). Cells were harvested in the late exponential phase at an OD600 of 0.8, i.e., when significant amounts of glucose have been consumed. l-Alanine obtained by protein hydrolysis after growth on natural, [1-13C]- or [3-13C]glucose was analyzed by GC-MS. When [1-13C]- or [3-13C]glucose was used, 13C label was found exclusively in the C-1 or C-3 position of alanine, respectively, derived directly from pyruvate, while phosphoenolpyruvate (PEP) was unlabeled (not shown). These labeling patterns clearly show that glucose was exclusively catabolized through an ED-like pathway in vivo. The activity of an Embden-Meyerhof pathway was excluded due to the absence of 13C label at the C-1 or C-3 position of alanine when [1-13C]- or [3-13C]glucose was used as the carbon source, respectively (Table (Table1).1). From the fact that about 30% of the alanine contained 13C label, it is concluded that about 40% of the alanine is taken up by the cell from the yeast extract in the medium and 60% (two times 30% due to the unlabeled PEP) originates from glucose with pyruvate as the immediate precursor (Table (Table11).
For comparison, similar in vivo 13C-labeling experiments were performed with S. acidocaldarius, for which a branched ED pathway was proposed. The labeling distribution in protein-derived l-alanine was analyzed after growth of S. acidocaldarius in a minimal medium with [1-13C]- or [3-13C]glucose as the sole carbon source. Similar to P. torridus, the labeling pattern indicates the exclusive operation of an ED pathway in vivo with more than 50% alanine containing 13C label (Table (Table1).1). Assuming that alanine was produced from pyruvate originating exclusively from the ED pathway, a fractional label of maximally 50% can be expected with the other 50% synthesized from unlabeled PEP. The slightly higher observed 60% label might be due to additional label introduced by other pathways, presumably tricarboxylic acid (TCA) cycle and gluconeogenesis.
In accordance with previous data (19), extracts of P. torridus grown on glucose and yeast extract contained activities of enzymes of a nonphosphorylative ED pathway, including glucose dehydrogenase, gluconate dehydratase, KDG aldolase, glyceraldehyde dehydrogenase, glycerate kinase, (2-phosphoglycerate-forming) enolase, and pyruvate kinase. To test the possible induction of these enzymes by glucose, the activities were also measured in cells after growth with yeast extract in the absence of glucose. The specific activities of all ED enzymes were similar both in the presence and absence of glucose (Table (Table2),2), suggesting that the enzymes were not inducible by glucose but constitutively expressed. Accordingly, the activities of ED enzymes did not change in growing cultures after transition from growth on yeast extract components to growth on glucose (data not shown).
Cell extracts of P. torridus exhibited aldolase activity catalyzing the cleavage of KDG (0.07 U/mg; Km of 0.4 mM) at a 25-fold-higher rate compared to the cleavage of KDPG (0.003 U/mg). This high preference for KDG over KDPG suggests that in P. torridus, a KDG-specific aldolase (KDGA) is the relevant aldolase for glucose degradation in vivo as part of the nonphosphorylated ED pathway. Similar results were obtained with the thermoacidophilic euryarchaeon Thermoplasma acidophilum. Extracts of glucose-grown cells converted KDG (0.26 U/mg; Km 0.28 mM) at 130-fold-higher activities compared to KDPG (0.002 U/mg), indicating that as in P. torridus, a KDG-specific aldolase is operative in vivo in T. acidophilum.
For comparison, aldolase activities were also measured in the thermoacidophilic crenarchaeon Sulfolobus acidocaldarius, for which a branched ED pathway was proposed with a bifunctional KD(P)G aldolase as a key enzyme. Extracts of glucose-grown cells of S. acidocaldarius catalyzed the cleavage of KDG (0.18 U/mg; Km of 2.6 mM) and KDPG (0.2 U/mg; Km of 0.19 mM) at similar rates, which is in accordance with an in vivo function of a bifunctional KD(P)G aldolase.
The enzyme activities in cell extracts of P. torridus, in particular the unusual KDG-specific aldolase, suggest the operation of a nonphosphorylative ED pathway in this euryarchaeon. To further characterize the key enzymes of this pathway, the first euryarchaeal gluconate dehydratase and the KDG-specific aldolase activity were purified and characterized, and the genes encoding them were identified.
Gluconate dehydratase (GAD) was purified 190-fold to apparent homogeneity involving five purification steps (Table (Table3).3). SDS-PAGE of the purified enzyme revealed one subunit at 44 kDa (Fig. (Fig.2A).2A). The molecular mass as estimated by gel filtration was 340 kDa, indicating a homooctameric structure of the native enzyme. By peptide mass fingerprinting of the purified enzyme, a single ORF, Pto0485, was identified in the genome of P. torridus; the matched peptides cover 34% of the protein. Thus, Pto0485 represents the gad gene encoding gluconate dehydratase in P. torridus. GAD catalyzed the conversion of gluconate to KDG following Michaelis-Menten kinetics with Vmax and Km values of 15 U/mg and 2.5 mM, respectively. Gluconate dehydratase also catalyzed dehydration of the d-galactonate at about 1 U/mg with a Km for galactonate of 2 mM. Xylonate (10 mM) was not utilized. The enzyme required Mg2+ with an apparent Km of 4.2 mM. The pH optimum of GAD was at pH 6, and the enzyme showed 50% of activity at pH 8.1 and at pH 4.5, which corresponds to the measured internal pH of P. torridus of pH 4.6 (28). GAD showed moderately thermophilic properties with a temperature optimum at 70°C and a substantial thermostability at 60°C, not losing activity upon incubation for 2 h. At 70°C, its half-life was about 15 min.
2-Keto-3-deoxygluconate aldolase (KDGA) activity from P. torridus was purified 340-fold to apparent homogeneity by four chromatographic steps (Table (Table4).4). SDS-PAGE of the purified enzyme revealed one subunit at 32 kDa (Fig. (Fig.2B).2B). The molecular mass as estimated by gel filtration was 120 kDa, indicating a homotetrameric structure of the native enzyme. The pH optimum of KDGA was at pH 5.5, and 50% of activity was found at pH 4.5 and 7.5. The enzyme has a temperature optimum at 65°C and showed high thermostability. At 70°C, the enzyme did not lose activity upon incubation for 2 h. The half-lives of the enzyme at 80°C and 90°C were 20 min and 15 min, respectively.
KDGA catalyzed the cleavage of KDG to pyruvate and glyceraldehyde with apparent Vmax and Km values of 50 U/mg and 0.3 mM, respectively. The enzyme also catalyzed the conversion of KDPG to pyruvate and glyceraldehyde-3-phosphate with apparent Vmax and Km of 0.63 U/mg and 8 mM, respectively. Thus, the catalytic efficiency (kcat/Km) for KDG was almost 2,000-fold higher than that for KDPG, indicating that KDGA from P. torridus is highly specific for KDG, making the use of KDPG as a physiological substrate highly unlikely (Table (Table55).
KDGA catalyzed the aldol condensation reaction, i.e., KDG formation from d- or l-glyceraldehyde and pyruvate with apparent Vmax values of 67 U/mg (d form) and 59 U/mg (l form) and apparent Km values of 4.6 mM (glyceraldehyde) and 2.7 mM (pyruvate). HPLC analysis of the aldol formed from d-glyceraldehyde and pyruvate revealed the formation of both KDG and its C-4 epimer 2-keto-3-deoxygalactonate (KDGal) at a ratio of 60% and 40%, respectively. Almost identical data were obtained with l-glyceraldehyde and pyruvate. The data indicate that KDGA from P. torridus lacks facial stereoselectivity of aldol formation as has been reported previously for the bifunctional KD(P)G aldolase from Sulfolobus solfataricus (12). Besides glyceraldehyde (100%), glycolaldehyde (13%), d-ribose (4%), and d-xylose (4%) were accepted as aldehyde substrates with pyruvate. KDGA enzyme also catalyzed at low activity the condensation of glyceraldehyde-3-phosphate and pyruvate. The reaction showed a pronounced substrate inhibition above 4 mM GAP. No activity was found at 15 mM GAP. At 4 mM GAP and 20 mM pyruvate, the specific activity was 7 U/mg.
The gene encoding KDGA, kdgA, was identified in the genome of P. torridus by peptide mass fingerprinting of the purified enzyme. A single ORF, Pto1279, was detected with matching peptides covering 51% of the encoded protein. This ORF, previously annotated as dihydrodipicolinate synthase, codes for a protein of 266 amino acids with a calculated molecular mass of 30.2 kDa. The determination of N-terminal amino acid sequence (MYKGIVCPMITPLDAHGNIDYNATN) of KDGA purified from P. torridus revealed that the translation start codon was an ATG 24 bp upstream of the annotated ATG start codon of Pto1279. Thus, the corrected ORF Pto1279 encodes a larger protein (274 amino acids, 31.3 kDa) containing eight additional amino acids at the N terminus.
The function encoded by Pto1279 (kdgA gene) was determined by its functional overexpression in E. coli. The ORF was cloned into a pET vector and expressed in E. coli as a His-tagged fusion protein, which was purified by a heat step and Ni-NTA affinity chromatography. The purified recombinant KDGA was characterized as a 120-kDa homotetramer showing catalytic properties similar to those of the enzyme purified from P. torridus (Table (Table55).
In the present communication, the glucose degradation pathway in P. torridus was analyzed by in vivo labeling experiments and by detailed enzyme studies. The data indicate that P. torridus utilizes a strictly nonphosphorylative Entner-Doudoroff pathway with a novel KDG-specific aldolase as a key enzyme. The nonphosphorylative ED pathway of P. torridus and its key enzymes, gluconate dehydratase (GAD) and KDGA, will be discussed in comparison with the branched ED pathway of Sulfolobus.
The in vivo 13C-labeling pattern of protein-derived l-alanine after growth of P. torridus with specifically 13C-labeled glucose clearly indicates the exclusive operation of an ED-like pathway. Cell extracts contained all enzyme activities of a nonphosphorylative ED pathway. The two key enzymes, gluconate dehydratase (GAD) and a KDG-specific aldolase (KDGA), were characterized, and the KDGA-encoding genes were identified (see below).
The enzymes of the nonphosphorylative ED pathway in P. torridus were found not to be regulated by glucose. This constitutive expression might be explained since the organism uses different pathways for glucose degradation and gluconeogenesis, i.e., modified ED pathway and reversed EM pathway, respectively. Theses pathways do not share common intermediates and reactions which might cause a futile cycle, and thus, they can exist in parallel without being regulated. In accordance, we measured several enzymes of the reversed EM pathway in P. torridus, i.e., phosphoglycerate mutase (0.07), phosphoglycerate kinase (0.03), glyceraldehyde-3-phosphate dehydrogenase (0.03), triosephosphate isomerase (0.64), and phosphoglucose isomerase (0.03) and found that the specific activities (given in U/mg at 55°C) were also constitutive and not regulated by glucose. Despite the constitutive formation of ED enzymes in P. torridus, glucose was not significantly metabolized in the first growth phase (Fig. (Fig.1).1). This indicates that components of the yeast extract are used as carbon sources in this phase and that the presence of these components apparently prevents glucose utilization, e.g., by inhibition of glucose transport into the cell.
GAD was characterized as a 340-kDa homooctameric protein encoded by Pto0485. The enzyme showed high sequence identity (44%) to GAD (SSO3198) from S. solfataricus. Both GADs share a similar subunit size and an octameric oligomeric structure (11). However, a monomeric structure of the S. solfataricus GAD has also been reported (13). GADs from both organisms showed similar catalytic properties, including Mg2+ dependence of activity and the utilization of galactonate in addition to gluconate, indicating substrate promiscuity. The ratio of catalytic efficiency for gluconate and galactonate utilization was about 10 to 1, which is on the order of that reported for the promiscuous GAD from S. solfataricus (ratio of 6:1) (13).
Orthologs of P. torridus GAD with high sequence identity were also found in Thermoplasma species and in the crenarchaeon Thermoproteus tenax (1). GAD from P. torridus and other archaeal GADs belong to the mandelate racemase (MR) subfamily of the enolase superfamily; they contain the characteristic signature patterns of this family, including conserved glutamate residues as ligands for Mg2+ and conserved arginine and aspartate residues involved in general acid base catalysis in the dehydration mechanism (8, 14). The archaeal GADs do not show similarities to the bacterial GAD from Achromobacter xylosoxidans (10) and to 6-phosphogluconate dehydratases of the classical ED pathway in bacteria; these bacterial enzymes belong to the dihydroxyacid dehydratase/6-phosphogluconate dehydratase (ILVD/EDD) superfamily.
Cell extracts of P. torridus contained an aldolase activity which shows a high preference for KDG over KDPG, giving the first indication of a KDG-specific aldolase to be operative in glucose degradation in vivo. The purified KDG aldolase (KDGA) from P. torridus was characterized as a novel KDG-specific aldolase. KDGA is a homotetrameric protein composed of 32-kDa subunits. This molecular composition was also reported for bifunctional KD(P)GA from Sulfolobus species and Thermoproteus tenax. However, KDGA differs from KD(P)GA with respect to substrate specificity for KDG and KDPG and to phylogenetic affiliation.
KDGA from P. torridus cleaved KDG at a 1,000- to 2,000-fold-higher catalytic efficiency compared to that of the phosphorylated aldol KDPG. Thus, KDGA represents an aldolase with a novel substrate specificity being highly specific for (nonphosphorylated) KDG. Also, in the direction of aldol synthesis, KDGA catalyzed the formation of KDG from pyruvate and glyceraldehyde with high preference over KDPG synthesis from pyruvate and glyceraldehyde-3-phosphate.
In contrast, KD(P)GA from S. solfataricus cleaves both KDPG and KDG; however, it has a high preference for KDPG (14) (Table (Table5).5). In the direction of aldol formation, KD(P)GA from S. acidocaldarius and from Thermoproteus tenax showed a preferred formation of KDPG over KDG (1, 30).
KDGA from P. torridus showed substrate promiscuity, as measured in the direction of aldol synthesis. The enzyme catalyzed the formation of both KDG and its C-4 epimer 2-keto-3-deoxygalactonate (KDGal) from d- or l-glyceraldehyde and pyruvate at a similar ratio, indicating lack of stereospecific control in formation and cleavage of KDG and KDGal. In this respect, KDGA from P. torridus is similar to KD(P)GA from Sulfolobus solfataricus.
By mass spectrometry analysis of purified P. torridus KDGA, a single ORF, Pto1279, originally annotated as dihydrodipicolinate synthase, was identified as the gene (kdgA) encoding KDGA. The start codon of Pto1279 was incorrectly annotated as identified by N-terminal amino acid sequencing of the enzyme purified from P. torridus. The correct Pto1279 gene encodes a protein that is 8 amino acids longer. Heterologous overexpression of Pto1279 yielded a recombinant KDGA with kinetic properties similar to those of the KDGA purified from P. torridus.
KDGA from P. torridus (Pto1279) showed about 50% sequence identity to putative homologs from Thermoplasma species, T. acidophilum (Ta1157), and T. volcanium (TVN1228), suggesting the presence of functional KDG-specific aldolases in these euryarchaea. This is supported by enzymatic measurements in T. acidophilum (this paper), showing aldolase activity in vivo with high preference for KDG over KDPG, suggesting the operation of a KDG-specific aldolase encoded by Ta1157. KDGA from P. torridus showed only low sequence identities (20 to 25%) to characterized bifunctional KD(P)GA from Sulfolobus species and T. tenax. Both types of aldolases, KDGA and KD(P)GA, are class I aldolases, which belong to the dihydropicolinate synthase (DHPDS)-like superfamily, which also include DHPDSs and N-acetylneuraminate lyases (NAL). A multiple-sequence alignment of KDGA from P. torridus and putative homologs from Thermoplasma species, from characterized bifunctional KD(P)GAs from Sulfolobus species and T. tenax, and from selected members of DHPDS and NAL families is given in Fig. Fig.3.3. Both KDGA and KD(P)GA contain several conserved amino acids typical of enzymes in the DHPDS-like family, including conserved lysine residues indicative of class I aldolases forming a Schiff base intermediate in the catalytic cycle. On the basis of crystal structures of Sulfolobus and Thermoproteus KD(P)GAs, amino acids for substrate binding were identified, including two conserved arginines and a conserved tyrosine (Fig. (Fig.3),3), which were proposed to form a putative phosphate binding pocket (16, 30; Gary Taylor [United Kingdom], unpublished results). These conserved amino acids are absent in KDGA from P. torridus, which might reflect its preference for nonphosphorylated substrate KDG.
The phylogenetic relationships of KDGA from P. torridus, characterized bifunctional KDG/KDPG aldolases, KD(P)GA, and other families (dihydrodipicolinate synthase [DHDPS] and N-acetylneuraminate lyase [NAL]) of the DHDPS superfamily are shown in Fig. Fig.4.4. In accordance with its unique substrate specificity being highly specific for KDG, KDGA from P. torridus (Pto1279) and putative homologs from Thermoplasma volcanium (TVN1228), and T. acidophilum (Ta1157), form a distinct family within the DHDPS superfamily. The KDGA family is largely separated by strong bootstrap support from bifunctional KD(P)GA, which also form a distinct cluster. Besides characterized KD(P)GAs (underlined), the cluster contained putative homologs in Pyrobaculum and Methanosphaera. Distant homologs with less sequence identity were found in Ferroplasma acidarmanus, P. torridus (Pto1026), T. acidophilum (Ta0619), and T. volcanii (TVN0669). The function of the encoded proteins is not known. A functional involvement in glucose degradation appears unlikely as concluded from a proteomic study of soluble proteins in glucose-grown Thermoplasma acidophilum (26). In this study, no gene product of ORF Ta0619 was detected, whereas the Ta1157-encoded gene product, which is a close homolog to P. torridus KDGA, was present. The data suggest that in T. acidophilum, a KDG-specific aldolase encoded by Ta1157 is operative in glucose degradation, which is in accordance with the KDG-specific aldolase activity, measured in cell extracts (see above).
In summary, the data reported here and in previous work for P. torridus (2, 17, 19) present the first comprehensive description of a strictly nonphosphorylative ED pathway in archaea. A similar pathway is probably operative in closely related Thermoplasma species based on enzyme and genomic studies. A comparison of the nonphosphorylative ED pathway in P. torridus and the branched ED pathway of S. solfataricus is given in Fig. Fig.5.5. For a recent article on the branched ED pathway in Thermoproteus tenax, see reference 32.
In both the nonphosphorylative ED pathway and branched ED pathway in Picrophilus and Sulfolobus, the formation of KDG is catalyzed by homologous glucose dehydrogenases and gluconate dehydratases. However, the subsequent routes of gluconate degradation to 2-phosphoglycerate differ as follows. In the nonphosphorylative ED pathway of Picrophilus, KDG is cleaved by the novel KDG-specific aldolase to pyruvate and glyceraldehyde, which in turn is oxidized to glycerate via NADP-specific glyceraldehyde dehydrogenase, a novel enzyme of the aldehyde dehydrogenase superfamily (19). Glycerate is then phosphorylated by a specific 2-phosphoglycerate-forming kinase (17).
In the branched ED pathway, two routes of KDG conversion to 2-phosphoglycerate have been proposed (25). In the nonphosphorylative route, KDG is split to glyceraldehyde and pyruvate via bifunctional KDGA. The oxidation of glyceraldehyde is catalyzed by an oxidoreductase (9), rather than by the NADP+-dependent glyceraldehyde dehydrogenase as in P. torridus. Glycerate phosphorylation to 2-phosphoglycerate is catalyzed by a specific kinase homologous to the P. torridus enzyme. The semiphosphorylated route involves phosphorylation of KDG—via KDG kinase—to KDPG, which is cleaved to pyruvate and glyceraldehyde-3-phosphate (GAP) by bifunctional KDPGA. GAP is then oxidized by nonphosphorylative GAPN, forming 3-phosphoglycerate, which is converted to 2-phosphoglycerate by means of a phosphoglycerate mutase. Finally, the conversion of 2-phosphoglycerate to pyruvate in both the nonphosphorylative and branched ED pathways is catalyzed by conventional enolase and pyruvate kinase.
The nonphosphorylative ED pathway in P. torridus contained promiscuous glucose/galactose dehydrogenase (2), gluconate/galactonate dehydratase, and KDG/KDGal aldolase (this work). These data suggest that the nonphosphorylative ED pathway in P. torridus provides an equivalent route for the degradation of both glucose and galactose, as first proposed for the branched ED pathway (Fig. (Fig.5)5) in S. solfataricus (4).
The genes encoding all enzymes of the nonphosphorylative ED pathway in Picrophilus were found to be scattered along the chromosome, whereas several genes encoding enzymes of the branched ED pathway in Sulfolobus (and also in Thermoproteus), i.e., GAD, KD(P)GA, KDG kinase, and GAPN, are clustered (see the ORF numbers in Fig. Fig.5).5). The latter finding is in accordance with the in vivo operation of the branched ED pathway in the crenarchaea. Furthermore, in the P. torridus genome, homologous genes encoding KDG kinase and GAPN were absent, which is in accordance with the proposed strict nonphosphorylative ED pathway in P. torridus.
We thank U. Sauer for financial support of T. Fuhrer.
We thank U. Sauer for use of GC-MS. Further, we thank R. Schmid (Osnabrück, Germany) for N-terminal amino acid analyses, S. Anemüller (Lübeck, Germany) for providing cell mass of T. acidophilum and H. Preidl and A. Brandenburger for expert technical assistance. KDPG was a gift from E. Toone (Durham, NC). The analysis of KDG and KDGal formation by KDGA from P. torridus was performed by D. Hough and M. Danson (Bath, United Kingdom). Finally, we thank U. Johnsen for measuring the kinetic constants of GAD and for help preparing the manuscript.
Published ahead of print on 18 December 2009.