This Article Abstract Full Text (PDF) Other Versions of this Article: JB.00047-07v1 189/11/4108    most recent Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints and Permissions Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Jahn, U. Articles by Fuchs, G. Search for Related Content PubMed PubMed Citation Articles by Jahn, U. Articles by Fuchs, G.

Insights into the Autotrophic CO₂ Fixation Pathway of the Archaeon Ignicoccus hospitalis: Comprehensive Analysis of the Central Carbon Metabolism

Ulrike Jahn,¹ Harald Huber,^1* Wolfgang Eisenreich,² Michael Hügler,^3, and Georg Fuchs^3*

Lehrstuhl Mikrobiologie und Archaeenzentrum, Universität Regensburg, Universitätsstr. 31, D-93053 Regensburg, Germany,¹ Lehrstuhl für Organische Chemie und Biochemie, Technische Universität München, Lichtenbergstr. 4, D-85748 Garching, Germany,² Mikrobiologie, Fakultät Biologie, Universität Freiburg, Schänzlestr. 1, D-79104 Freiburg, Germany³

Received 10 January 2007/ Accepted 19 March 2007

	ABSTRACT

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Ignicoccus hospitalis is an autotrophic hyperthermophilic archaeon that serves as a host for another parasitic/symbiotic archaeon, Nanoarchaeum equitans. In this study, the biosynthetic pathways of I. hospitalis were investigated by in vitro enzymatic analyses, in vivo ¹³C-labeling experiments, and genomic analyses. Our results suggest the operation of a so far unknown pathway of autotrophic CO₂ fixation that starts from acetyl-coenzyme A (CoA). The cyclic regeneration of acetyl-CoA, the primary CO₂ acceptor molecule, has not been clarified yet. In essence, acetyl-CoA is converted into pyruvate via reductive carboxylation by pyruvate-ferredoxin oxidoreductase. Pyruvate-water dikinase converts pyruvate into phosphoenolpyruvate (PEP), which is carboxylated to oxaloacetate by PEP carboxylase. An incomplete citric acid cycle is operating: citrate is synthesized from oxaloacetate and acetyl-CoA by a (re)-specific citrate synthase, whereas a 2-oxoglutarate-oxidizing enzyme is lacking. Further investigations revealed that several special biosynthetic pathways that have recently been described for various archaea are operating. Isoleucine is synthesized via the uncommon citramalate pathway and lysine via the -aminoadipate pathway. Gluconeogenesis is achieved via a reverse Embden-Meyerhof pathway using a novel type of fructose 1,6-bisphosphate aldolase. Pentosephosphates are formed from hexosephosphates via the suggested ribulose-monophosphate pathway, whereby formaldehyde is released from C-1 of hexose. The organism may not contain any sugar-metabolizing pathway. This comprehensive analysis of the central carbon metabolism of I. hospitalis revealed further evidence for the unexpected and unexplored diversity of metabolic pathways within the (hyperthermophilic) archaea.

	INTRODUCTION

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Ignicoccus hospitalis is a recently described member of the crenarchaeal order Desulfurococcales (38). Like the two other described Ignicoccus species, Ignicoccus islandicus and Ignicoccus pacificus (24), it is a strictly anaerobic, hyperthermophilic archaeon with an optimal growth temperature of 90°C. All Ignicoccus species grow chemolithoautotrophically, using the reduction of elemental sulfur with molecular hydrogen as the sole energy source and CO₂ as the sole carbon source. They possess a unique ultrastructure of the cell envelope with an outer membrane resembling that of gram-negative bacteria (40). Despite the great similarities, I. hospitalis shows a unique feature: it is the only known host of Nanoarchaeum equitans, a representative of the archaeal kingdom Nanoarchaeota. The two organisms represent the only known purely archaeal host-parasite system (25). As the smallest archaeal genome of N. equitans (490 kbp) lacks nearly all genes for biosynthetic pathways (48), it is likely that I. hospitalis provides all essential biosynthetic building blocks, like lipids or amino acids, for N. equitans (32).

The genome of I. hospitalis has also been sequenced (http://genome.jgi-psf.org/draft_microbes/ign_k/ign_k.draft.html). It is the second smallest archaeal genome known and exhibits only 1,434 putative genes, clearly indicating a high metabolic specialization of I. hospitalis. To maintain a continuous supply of organic carbon, I. hospitalis assimilates CO₂ into low-molecular-weight building blocks for biosynthesis. At present, four pathways of CO₂ fixation are known: the Calvin-Bassham-Benson cycle, the reductive citric acid cycle, the reductive acetyl-coenzyme A (CoA) pathway, and the 3-hydroxpropionate cycle, with a modified version present in the crenarchaeal order Sulfolobales (34). Preliminary experiments with I. pacificus and I. islandicus indicated that these autotrophic members of the Desulfurococcales lack the key enzymes of all these autotrophic pathways (28), raising the question of whether the members of the genus Ignicoccus use a new or modified pathway of autotrophic CO₂ fixation.

Another central biosynthetic process is gluconeogenesis. It is believed that gluconeogenesis in Bacteria and Archaea is achieved by the reversed Embden-Meyerhof pathway; in contrast to a rather uniform gluconeogenetic pathway, Archaea use an amazing variety of different pathways for glycolysis (44). The biosynthesis of pentose phosphate in Archaea is still an open question. Three possibilities for biosynthesis of pentoses have been described. The oxidative pentose phosphate pathway is mainly restricted to bacteria, whereas the nonoxidative pentose phosphate pathway was found in several methanogens on the basis of enzymatic and genomic analyses (43, 50). In archaea lacking the genes for the nonoxidative or the oxidative pentose phosphate pathway, the ribulose-monophosphate pathway was recently suggested as an alternative for pentose phosphate biosynthesis, on the basis of genome analyses. In this pathway, pentose phosphate is synthesized by the release of formaldehyde from 3-hexulose 6-phosphate (13, 33, 45).

In this work, we investigated the autotrophic CO₂ fixation pathway of I. hospitalis and the biosynthesis of representative main cellular building blocks. Enzymatic analyses in vitro were combined with retrobiosynthetic analyses of in vivo ¹³C-labeled amino acids. The experimental data were compared to the genome data of I. hospitalis to obtain comprehensive insight into the central metabolism in the organism. We suggest that I. hospitalis uses a novel, so far unknown pathway of CO₂ fixation.

	MATERIALS AND METHODS

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Strains and culture conditions. I. hospitalis strain KIN4/I^T was obtained from our culture collection and grown in synthetic-seawater medium, pH 5.5 (26), under a gas phase of H₂-CO₂ (80%/20% [vol/vol]) at 90°C. Metallosphaera sedula strain TH2^T (for control experiments) was grown autotrophically as described previously (27).

Production of ¹³C-labeled cell masses. I. hospitalis cells were cultivated under autotrophic growth conditions in a 100-liter enamel-protected fermentor in the presence of 0.5 mM [1-¹³C]acetate (Euriso-Top, Gif sur Yvette Cedex, France) (stirring, 150 rpm; gas phase, 300 kPa without gassing).

Incorporation of organic compounds. The incorporation of organic compounds into I. hospitalis cells was investigated by adding a mixture of unlabeled and ¹⁴C-labeled compounds to the growth medium. [1-¹⁴C]acetate (2.0 GBq mmol^–1; Amersham Biosciences, Freiburg, Germany), [3-¹⁴C]pyruvate (0.6 GBq mmol^–1; American Radiolabeled Chemicals, St. Louis, MO), [1,4-¹⁴C]succinate (2.0 GBq mmol^–1; American Radiolabeled Chemicals, St. Louis, MO), or [U-¹⁴C]glucose (11.7 GBq mmol^–1; American Radiolabeled Chemicals, St. Louis, MO), respectively, was added (630, 140, 140, and 165 kBq, respectively, total radioactivity in a final concentration of 0.5 mM). Cells were grown in 1-liter serum bottles containing 250 ml medium under a gas pressure of 160 kPa (H₂-CO₂; 80/20) at 100 rpm. Samples (50 ml) were retrieved at intervals of 100 to 120 min during the exponential growth phase. The cell concentration was determined by cell counting using a Neubauer counting chamber. After a centrifugation step (40,000 x g; 30 min), the cells were washed with medium lacking the organic substrate and resuspended in 100 µl of fresh synthetic-seawater medium. Radioactivity was determined within this sample, as well as in 100 and 200 µl of the supernatant, by liquid scintillation counting.

Determination of the protein content of I. hospitalis cells. Cells grown in 20 ml culture medium to about 10⁷ cells ml^–1 (total, 2 x 10⁸ cells) were harvested, washed, and resuspended in 1 ml of double-distilled water. The cells were lysed by 30-s ultrasonic treatment of the suspension (Sonifier 250; Branson, Danbury, CT). The protein content of the sample was determined to be 6 µg per 10⁷ cells using the method of Bradford (5).

Syntheses. Malonyl-CoA, succinyl-CoA, acetyl-CoA, propionyl-CoA, L-malyl-CoA, and 3-hydroxypropionate were synthesized as described elsewhere (21), and the dry powders of the CoA esters were stored at –20°C.

Preparation of cell extracts. Cell extracts were prepared using a mixer mill (type MM2; Retsch, Haan, Germany); 0.5 ml buffer (100 mM Tris-HCl, pH 7.8, 3 mM 1,4-dithioerythritol [DTE], and 0.1 mg of DNase I per ml of buffer) and 0.5 g glass beads (diameter, 0.1 to 0.25 mm) were added to 0.1 g of wet cells in 1-ml stoppered glass vials in the anaerobic chamber. The solution was treated in the mixer mill for 6 min at 100% intensity (30 Hz). Following a centrifugation step (10 min; 12,000 x g; 4°C), the supernatant was used for enzyme tests. For larger amounts of cell extract, the cell suspension was passed through a French pressure cell at 137 MPa, and the lysate was ultracentrifuged (1 h; 100,000 x g; 4°C). The protein concentration in cell extracts was determined by the method of Bradford (5), using bovine serum albumin as a standard.

Enzyme assays. Routine enzyme assays (0.5 ml assay mixture) were performed in stoppered 0.5-ml glass cuvettes or 1-ml glass vials. For practical reasons, unless otherwise stated, the assay temperature was 75°C or 85°C, as specified in Results (compared to a 90°C growth temperature).

Assays with NADH, NADPH, NAD, or NADP. Reactions involving pyridine nucleotides were followed spectrophotometrically at 365 nm {_{365 nm}[NAD(P)H] = 3.4 x 10³ M^–1 cm^–1}.

ATP citrate lyase, malate dehydrogenase, isocitrate dehydrogenase, aconitase, pyruvate kinase, 2-oxoglutarate dehydrogenase, and pyruvate dehydrogenase, as well as the reduction of malonyl-CoA to 3-hydroxypropionate and the reduction of 3-hydroxypropionate to propionyl-CoA, were measured as described previously (28). Isocitrate dehydrogenase and aconitase were tested under oxic and anoxic conditions.

For glyceraldehyde 3-phosphate dehydrogenase, the phosphate- and glyceraldehyde 3-phosphate-dependent reduction of NAD(P)⁺ was monitored. The assay mixture (adjusted to pH 7.8) contained 100 mM Tris-HCl, 100 mM K₂HPO₄, 2 mM NAD(P)⁺, and 3 mM glyceraldehyde 3-phosphate. The addition of glyceraldehyde 3-phosphate started the reaction. The presence of phosphate-independent glyceraldehyde 3-phosphate dehydrogenase was tested in an assay lacking phosphate.

Assays with viologen dyes. Reactions involving methyl viologen (MV) or benzyl viologen (BV) were followed spectrophotometrically at 578 nm [₅₇₈ (MV) = 9.8 x 10³ M^–1 cm^–1; ₅₇₈ (BV) = 8.6 x 10³ M^–1 cm^–1] (7). The following anoxic assay mixture was used with N₂ gas as the headspace: 100 mM Tris-HCl, pH 7.8, 4 mM DTE, 2 mM MgCl₂, and 4 mM MV or 2 mM BV. Dithionite was added by syringe from 5 mM stock solution until a permanent faint bluish color was obtained.

For 2-oxoglutarate- and pyruvate-acceptor oxidoreductases, the assay mixture contained buffer, 1 mM CoA, and 3 mM 2-oxoglutarate or pyruvate. The reaction was started by the addition of 2-oxoglutarate or pyruvate. As a control, the assays were repeated by adding 0.5 mM thiamine-diphosphate to the reaction mixture.

For formate dehydrogenase, the assay mixture contained buffer and 3 mM formate; the reaction was started by the addition of formate.

For fumarate reductase, before the reaction was started, MV was reduced to an optical density at 578 nm of about 2.0 with dithionite. The addition of fumarate started the reaction.

Carbon monoxide dehydrogenase was measured according to a previously described method (30).

Other spectrophotometric assays. L-Malyl-CoA lyase, isocitrate lyase, fumarate hydratase, and citrate synthase were measured as described by Hügler et al. (28). All enzyme activities were tested under oxic and anoxic conditions.

Radioactive assays. Carboxylation reactions with ¹⁴CO₂ were followed by measuring the substrate-dependent incorporation of ¹⁴C from [¹⁴C]-bicarbonate into acid-stable products.

Ribulose 1,5-bisphosphate carboxylase/oxygenase (RubisCO), acetyl-CoA carboxylase, propionyl-CoA carboxylase, phosphoenolpyruvate (PEP) carboxylase, pyruvate carboxylase, PEP carboxykinase, pyruvate water dikinase (PEP synthase), and pyruvate phosphate dikinase were measured according to the method of Hügler et al. (28).

For 2-oxoglutarate- and pyruvate-acceptor oxidoreductases, the exchange of ¹⁴CO₂ into the carboxyl group of the 2-oxoacids in the absence of electron acceptors is characteristic of reversible 2-oxoacid synthases. The following anoxic assay mixture was used with N₂ gas as the headspace: 100 mM Tris-HCl, pH 7.8, 4 mM DTE, 2 mM MgCl₂, 10 mM NaHCO₃, 37 kBq [¹⁴C]Na₂CO₃ (specific radioactivity, 2.5 Bq nmol^–1), and 3 mM 2-oxoglutarate or pyruvate. In some experiments, 1 mM CoA or 0.5 mM thiamine PP_i was included, but without effect. The reaction was started by the addition of 2-oxoglutarate or pyruvate.

Enzymatic assays coupled to phosphate determination. Reactions resulting in a substrate-dependent release of inorganic phosphate from the phosphorylated substrates were followed by measuring the substrate- and cell extract-dependent increase of inorganic phosphate in the assay (total volume, 0.25 ml). The buffer contained 100 mM Tris-HCl, pH 7.8, and 5 mM MgCl₂. Reaction mixtures were incubated at 85°C for 0, 2, and 5 min and stopped by rapid cooling on ice. Experiments in which substrate or cell extract, respectively, were omitted served as blanks and controls.

For fructose 1,6-bisphosphate phosphatase, the reaction was started by the addition of 4 mM fructose 1,6-bisphosphate to the reaction mixture.

For fructose 1,6-bisphosphate aldolase, the enzyme activity was determined in the direction of fructose bisphosphate formation by coupling the reaction to endogenous fructose 1,6-bisphosphate phosphatase activity, which dephosphorylates fructose 1,6-bisphosphate. Glyceraldehyde 3-phosphate (8 mM) was incubated for 10 min at 25°C in the presence of 5 U triosephosphate isomerase. Afterward, the assay mixture was heated to 85°C and the test was started by the addition of cell extract. The release of inorganic phosphate from fructose 1,6-bisphosphate was monitored.

The glyceraldehyde-3-phosphate dehydrogenase and phosphoglycerate kinase assay detected both activities, although the specific activity observed was that of the rate-limiting step. The NAD(P)H- and 3-phosphoglycerate-dependent release of inorganic phosphate from MgATP was followed. The reaction was coupled to the endogenous phosphoglycerate kinase activity, which forms 1,3-bisphosphoglycerate from 3-phosphoglycerate and ATP. The reaction mixture contained 0.5 mM NAD(P)H, 5 mM ATP, and 5 mM 3-phosphoglycerate. The reaction was started by the addition of 3-phosphoglycerate. The blank reaction in the absence of ATP was zero, and the blank value obtained in the presence of ATP and in the absence of NAD(P)H (20 nmol min^–1 mg^–1) was subtracted from the measured values.

Determination of acid-stable ¹⁴C. The fixation of ¹⁴C from [¹⁴C]bicarbonate into nonvolatile, acid-stable products was measured in samples (0.1 ml) incubated for 1, 5, and 10 min (see above). The addition of 20 µl of 6 M HCl (to pH 1 to 2) stopped the reaction. Volatile ¹⁴CO₂ (nonfixed) was removed by vigorously shaking the sample in open scintillation vials for 3 h. The remaining radioactivity in the liquid sample was then determined. The radioactivities in samples of two control experiments, in which the substrate and extract, respectively, were omitted, served as blanks and controls.

Determination of ¹⁴C. The amount of ¹⁴C in liquid samples (up to 0.2 ml) was determined by liquid scintillation counting using 3 ml of Rotiszint eco plus scintillation cocktail (Roth, Karlsruhe, Germany). The counting efficiency (75 to 85%) was determined via the channel ratio method.

Determination of inorganic phosphate. Inorganic phosphate released in enzymatic assays was determined using the modified malachite green-ammonium molybdate assay (22).

Detection of biotinylated proteins. Biotinylated proteins in cell extracts were detected with peroxidase-conjugated avidin as described previously (12, 35).

Fractionation of cell material. Cells (3 g [wet weight]) were suspended in 3 ml double-distilled water, and the suspension was treated with a sonifier (Sonifier 250; Branson, Danbury, CT). After cell disruption, 3 ml of 70% (vol/vol) HClO₄ was added to precipitate the protein. After 30 min of incubation on ice, a centrifugation step (15 min; 48,000 x g; 4°C) followed. To remove low-molecular-weight substances, the resulting pellet was washed three times with 10 ml 0.5 M HClO₄, followed by centrifugation (5 min; 48,000 x g). To extract nucleic acids, 10 ml 0.5 M HClO₄ was added to the pellet, and the solution was incubated at 70°C for 20 min. After centrifugation (15 min; 20,000 x g; 4°C), nucleic acids stayed in the supernatant and proteins in the pellet. To extract lipids, the pellet was resuspended in a mixture of ethanol and diethylether (3:1 [vol/vol]) and incubated at 40°C for 1 h. A centrifugation step (15 min; 20,000 x g; 4°C) separated the protein pellet from lipids in the supernatant. After a wash with 10 ml diethylether and centrifugation (15 min; 20,000 x g; 4°C), the protein pellet was dried by lyophilization (Celvinator Beta 1-16; Christ, Osterode, Germany). For protein hydrolysis, 3 ml 6 M HCl was added to the protein pellet. The suspension was incubated in sealed and evacuated glass ampoules at 112°C for 24 h. HCl was removed by three evaporation cycles, with the addition of 20 ml of double-distilled water after each cycle.

Separation of amino acids. Amino acids were isolated by chromatographic procedures, as described previously (10).

NMR spectroscopy. ¹H and ¹³C nuclear magnetic resonance (NMR) spectra were recorded at 500.13 and 125.76 MHz, respectively, with a DRX500 spectrometer (Bruker Biospin, Rheinstetten, Germany) as described previously (9). Tyr and Asp were dissolved in D₂O containing 0.1 M NaOD (pH 13); the other amino acids were dissolved in D₂O containing 0.1 M DCl (pH 1). ¹³C enrichments were determined for individual positions by quantitative NMR spectroscopy (9). Briefly, ¹³C NMR spectra of the ¹³C-labeled compound under study and of natural-abundance material were recorded under the same experimental conditions. Integrals were determined for each ¹³C NMR signal, and the signal integral for each carbon atom in the labeled compound was referenced to that of the natural-abundance material, thus affording relative ¹³C abundances for each position in the labeled molecular species. Relative abundances were normalized by assigning a value of 1.1% to the carbon atom with the lowest ¹³C enrichment. The validity of this approach was confirmed for certain amino acids by analysis of ¹³C coupling satellites in ¹H NMR spectra.

Database search. The genome sequence of I. hospitalis was obtained from http://genome.jgi-psf.org/draft_microbes/ign_k/ign_k.draft.html. Blast searches for genes encoding metabolic enzymes were performed using the program Clone Manager (Scientific & Educational Software). Query sequences were obtained from the Entrez gene database (http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?CMD=search&DB=gene).

	RESULTS

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Search for key enzymes of autotrophic pathways. Cell extracts of I. hospitalis were screened for the presence of enzyme activities that were considered representative of the four known autotrophic CO₂ fixation pathways. Autotrophic growth under optimal conditions (90°C; pH 6.0; 1% S⁰; H₂/CO₂ ratio, 80%/20% [vol/vol] at 250 kPa) occurred at a generation time of 1 h, corresponding to a specific growth rate (µ) of 0.011 min^–1. The cells that were studied grew with a generation time of approximately 2 h. Assuming a carbon content of 50% of the dry cell mass, the rate of carbon fixation under these conditions amounted to 0.4 µmol CO₂ fixed per min per mg of protein. This estimate is based on the specific substrate (S) consumption (dS) per time unit (dt): dS/dt = (µ/Y)X, where Y represents the growth yield (1 g of dry cell mass formed per 0.5 g of carbon fixed) and X represents the cell dry mass in g (1 g of cell dry mass, corresponding to approximately 0.5 g of protein).

The activity of RubisCo, the key enzyme of the Calvin cycle, was tested as ribulose 1,5-bisphosphate-dependent fixation of ¹⁴CO₂ into acid-stable products. The specific activity at 80°C was below 1 nmol min^–1 mg protein^–1, indicating that the Calvin cycle was not operating. Also, neither carbon monoxide dehydrogenase nor formate dehydrogenase activity—enzyme activities characteristic of the reductive acetyl-CoA pathway—was detectable (both were tested with MV as the artificial electron acceptor), indicating that the reductive acetyl-CoA pathway was not present, either.

In members of the Sulfolobales, acetyl-CoA/propionyl-CoA carboxylase is considered a key enzyme of a modified 3-hydroxypropionate cycle (28). Therefore, cell extracts were tested for the acetyl-CoA- or propionyl-CoA- plus Mg-ATP-dependent fixation of ¹⁴CO₂ into acid-stable products, but no activity was detected at 80°C. In addition, no malonyl-CoA reductase or propionyl-CoA synthase activity could be detected (<1 nmol min^–1 mg protein^–1). These enzymes are characteristic of the 3-hydroxypropionate cycle in Chloroflexus aurantiacus, and their activities could also be measured in Sulfolobales (1, 2, 28, 29, 31a). The presence of biotin proteins, such as the biotin carrier protein of acetyl-CoA carboxylase or pyruvate carboxylase, was investigated using the avidin technique. Extracts of I. hospitalis showed no detectable biotin-containing protein, whereas the small 18.6-kDa biotin carrier protein of acetyl-CoA/propionyl-CoA carboxylase in M. sedula extract, which served as a positive control, stained extremely strongly (not shown). Also, no band with a molecular mass around 65 kDa was visible. This would be the approximate size of the biotin-containing subunit of pyruvate carboxylase (e.g., in Methanocaldococcus jannaschii or Mycobacterium smegmatis, (36, 37). These results exclude the operation of a 3-hydroxypropionate cycle.

Members of the Thermoproteales exhibit a reductive citric acid cycle for autotrophic CO₂ fixation. Cell extracts of I. hospitalis did not catalyze the 2-oxoglutarate- and CoA-dependent reduction of viologen dyes or the exchange of ¹⁴CO₂ into 2-oxoglutarate in the presence or absence of CoA (<1 nmol min^–1 mg protein^–1 at 75°C or 85°C), which is characteristic of 2-oxoglutarate-acceptor oxidoreductase, a key enzyme of this cycle. Since pyruvate-acceptor oxidoreductase activity was clearly detectable by the same methods (see below), the failure to detect 2-oxoglutarate-acceptor oxidoreductase was not due to experimental failure but was most likely due to the absence of this enzyme of the reductive citric acid cycle. In addition, ATP citrate lyase activity was not detectable (<1 nmol min^–1 mg protein^–1 at 75°C), excluding the reductive citric acid cycle as a potential carbon fixation pathway.

In summary, no evidence for the operation of one of the known CO₂ fixation pathways was obtained. However, two possible CO₂-fixing enzyme activities were easily detected: pyruvate-acceptor oxidoreductase activity was observed both in the spectrophotometric assay with MV as the electron acceptor (115 ± 15 nmol min^–1 mg protein^–1 at 75°C) and in the radioactive assay following the exchange of ¹⁴CO₂ into the carboxyl group (145 ± 15 nmol min^–1 mg protein^–1 at 85°C). In contrast, no pyruvate dehydrogenase or 2-oxoglutarate dehydrogenase activity was observed with NAD⁺ or NADP⁺ as the electron acceptor (<1 nmol min^–1 mg protein^–1 at 75°C). Furthermore, PEP carboxylase was present and highly active (202 ± 20 nmol min^–1 mg protein^–1 at 85°C).

Enzyme activities in extracts of autotrophically grown cells. The presence of pyruvate-acceptor oxidoreductases suggests that acetyl-CoA is reductively carboxylated to pyruvate. Therefore, enzyme activities possibly involved in biosynthesis starting from pyruvate were investigated at 75°C or 85°C in cell extracts of I. hospitalis. The results of the measurement are presented in Table 1. Figure 1 gives an overview of the reactions involved.

View larger version (17K): [in this window] [in a new window]   FIG. 1. Biosynthetic capacity of I. hospitalis, according to genome data. Circles, putative genes of enzymes that were detected (E value < e–11); squares, genes of enzymes that were not detected in the I. hospitalis genome. The numbers of the putative genes of the I. hospitalis genome are shown: 1a, citrate synthase; 1b, ATP citrate lyase; 2, aconitase; 3, isocitrate dehydrogenase; 4a, malate dehydrogenase; 4b, malate quinone oxidoreductase; 5, fumarate hydratase; 6a, fumarate reductase; 6b, succinate dehydrogenase; 7, succinate-CoA ligase (ADP forming); 8a, 2-oxoacid-ferredoxin oxidoreductase; 8b, acetyl-CoA carboxylase; 8c, pyruvate dehydrogenase; 9a, pyruvate-water dikinase; 9b, pyruvate-phosphate dikinase; 9c, pyruvate kinase; 10a, PEP carboxylase; 10b, PEP carboxykinase; 10c, PEP carboxytransphosphorylase; 11, pyruvate carboxylase; 12, enolase; 13, phosphoglycerate mutase; 14, phosphoglycerate kinase; 15, glyceraldehyde 3-phosphate dehydrogenase; 16, triosephosphate isomerase; 17, fructose 1,6-bisphosphate aldolase; 18, fructose 1,6-bisphosphate phosphatase; 19, phosphoglucose isomerase (cupin type); 20, 3-hexulose 6-phosphate isomerase; 21, 3-hexulose 6-phosphate synthase.

PEP formation from pyruvate was tested by coupling PEP conversion to malate catalyzed by endogenous PEP carboxylase and NADH-dependent malate dehydrogenase. Extracts fixed ¹⁴CO₂ in an Mg-ATP- and pyruvate-dependent reaction in the presence of NADH with a specific activity of 210 ± 10 nmol min^–1 mg protein^–1. The addition of phosphate did not stimulate excluding the activity of pyruvate-phosphate dikinase (pyruvate + ATP + P_i PEP + AMP + PP_i). The reaction was also not inhibited by avidin, which excludes oxaloacetate formation via biotin-dependent pyruvate carboxylase (pyruvate + HCO₃^– + ATP oxaloacetate + ADP + P_i). These results indicate the presence of pyruvate-water dikinase (PEP synthetase) (pyruvate + ATP + H₂O PEP + AMP + P_i). Pyruvate kinase activity (PEP + ADP pyruvate + ATP) could not be detected.

The level of NAD(P)⁺-dependent malate dehydrogenase activity was high when oxaloacetate reduction with NADPH (745 ± 100 nmol min^–1 mg protein^–1) and NADH (1,190 ± 240 nmol min^–1 mg protein^–1) were tested. Fumarase activity, measured in the direction of fumarate hydration, was 895 ± 45 nmol min^–1 mg protein^–1. Fumarate reductase activity tested with reduced MV as the electron donor was 840 ± 45 nmol min^–1 mg protein^–1.

Extracts catalyzed the formation of citrate from oxaloacetate and acetyl-CoA with a specific activity of 21 ± 4 nmol min^–1 mg protein^–1. This activity was oxygen sensitive and detectable only under anoxic conditions. Since ATP citrate lyase activity was not detectable (the observed reaction did not require ADP and phosphate), this indicates the presence of a citrate synthase. Aconitase activity turned out also to be oxygen sensitive (30 ± 2 nmol min^–1 mg protein^–1). Isocitrate dehydrogenase was NADP⁺ specific (238 ± 22 nmol min^–1 mg^–1). As indicated above, both NAD(P)⁺ dependent 2-oxoglutarate dehydrogenase and 2-oxoglutarate-acceptor oxidoreductase activities were lacking. L-Malyl-CoA lyase activity was not detected.

Characteristic enzymes of gluconeogenesis were found. 3-Phosphoglycerate kinase and glyceraldehyde 3-phosphate dehydrogenase activities were measured in one assay and were clearly present (270 ± 10 nmol min^–1 mg protein^–1), with a preference of the dehydrogenase for NADPH. Similarly, fructose 1,6-bisphosphate aldolase and fructose 1,6-bisphosphate phosphatase were measured in one assay, and both enzymes were clearly detected (116 ± 16 nmol min^–1 mg protein^–1 and 146 ± 4 nmol min^–1 mg protein^–1, respectively, at 85°C).

Database search. The available genome database of I. hospitalis (http://genome.jgi-psf.org/draft_microbes/ign_k/ign_k.draft.html) was screened for the presence of putative genes coding for autotrophic pathways (Table 1). The results are summarized and visualized in Fig. 1.

Citric acid cycle. All putative genes for the citric acid cycle were detected, except for an (si)-specific citrate synthase and a 2-oxoglutarate dehydrogenase (Table 1). Probes for (si)-specific citrate synthase used for screening the database were from Thermus aquaticus, Pyrococcus furiosus, and Escherichia coli, but all E values were >e–5. However, since citrate synthase activity could be clearly detected, the gene encoding a putative (re)-specific citrate synthase must be present. Its gene is not known, and any annotation is therefore speculative. The genes for isocitrate lyase, L-malate synthase, L-malyl-CoA lyase, ATP citrate lyase, citrate lyase, and malic enzyme were not detected. The genes for succinate dehydrogenase were most similar to those of Sulfolobus acidocaldarius. However, it is unclear whether fumarate reduction is catalyzed by the annotated succinate dehydrogenase or by a separate fumarate reductase. It should be mentioned that two gene clusters containing the four subunits of putative 2-oxoacid-acceptor oxidoreductases were present. However, 2-oxoglutarate dehydrogenase or 2-oxoglutarate-acceptor oxidoreductase activity was not detectable, in contrast to pyruvate-acceptor oxidoreductase activity. These data taken together suggest the presence of an incomplete citric acid cycle that serves solely anabolic purposes. However, the nature of citrate synthase remains elusive.

Pyruvate and PEP synthesis and C-3 and C-4 interconversion. The putative genes for pyruvate-acceptor oxidoreductase (pyruvate synthase) were found; in contrast, pyruvate dehydrogenase genes were not found. Furthermore, a putative gene for pyruvate-water dikinase (PEP synthetase) that forms PEP was detected. The corresponding genes for pyruvate-phosphate dikinase or pyruvate kinase were not seen. The putative gene for PEP carboxylase was present. The alternative routes of oxaloacetate synthesis from pyruvate via pyruvate carboxylase, as well as PEP carboxykinase and PEP carboxytransphosphorylase, seem not to be present, since these genes were not detected. This suggests that acetyl-CoA is reductively carboxylated to pyruvate by pyruvate synthase and then converted to PEP by PEP synthetase and that PEP then gives rise to oxaloacetate via PEP carboxylase.

Gluconeogenesis from PEP and pentosephosphate synthesis. Gluconeogenesis from PEP via fructose 1,6-bisphosphate and fructose 6-phosphate seems to proceed in the conventional way, with some remarkable exceptions regarding the nature of the enzymes. The putative genes coding for enolase, phosphoglycerate mutase, phosphoglycerate kinase, glyceraldehyde 3-phosphate dehydrogenase, and triosephosphate isomerase were found. A gene coding for fructose 1,6-bisphosphate aldolase was not detected. The same was true for the classical hexosephosphate isomerase. However, a 1-phosphoglucose isomerase of the cupin superfamily was found, although with only a low E value (e–11) for the corresponding gene from Thermococcus litoralis; this gene product is thought to act as hexosephosphate isomerase. A putative fructose 1,6-bisphosphate phosphatase gene was found. In contrast, the genes for the classical enzymes of the pentose phosphate pathway, glucose 6-phosphate dehydrogenase, 6-phosphogluconolactone hydrolase, and 6-phosphogluconate dehydrogenase, were lacking. Instead, the two genes for the enzymes of the recently proposed ribulose monophosphate pathway of pentosephosphate synthesis, 3-hexulose 6-phosphate isomerase and 3-hexulose 6-phosphate synthase, were present.

Glycolysis. Genes for enzymes of the nonphosphorylating or phosphorylating Entner-Doudoroff pathway could not be found. Also, the genes coding for glucokinase and phosphofructokinase of a (modified) Embden-Meyerhof pathway were not detected. The gene for pyruvate-kinase was also not detected. However, it should be mentioned that in Archaea, PEP synthase can function as pyruvate kinase, as well (31). The genes coding for glyceraldehyde 3-phosphate-ferredoxin oxidoreductase (EC 1.2.7.6) and nonphosphorylative NAD⁺- or NADP⁺-dependent glyceraldehyde 3-phosphate dehydrogenase (EC 1.2.1.9 or EC 1.2.1.12, respectively) were missing. However, the gene for NAD(P)⁺-dependent phosphorylative D-glyceraldehyde 3-phosphate dehydrogenase (EC 1.2.1.59) was present (see above) and likely functions in gluconeogenesis.

Incorporation of ¹⁴C-labeled acetate, pyruvate, succinate, and glucose. I. hospitalis was grown autotrophically under anaerobic conditions with hydrogen as the electron donor and elemental sulfur as the electron acceptor (25). To test the incorporation of organic substrates, ¹⁴C-labeled acetate, pyruvate, succinate, or glucose was added to the medium at a concentration of 0.5 mM (Table 2). The generation time of I. hospitalis was about 60 min in all experiments, and cells were harvested at the end of the exponential growth phase. Figure 2 shows the accumulation of radioactivity from each labeled compound in the cell pellet and the decrease of radioactivity in the cell-free medium during growth. The organic compounds contributed to different extents to the cell material. Approximately 5.2% of the proffered labeled acetate, 3.5% of the pyruvate, 1.9% of the succinate, and 0% of the glucose were incorporated. In the experiments with pyruvate and succinate, losses of total radioactivity of 31% and 44%, respectively, occurred, which were most probably caused by the release of ¹⁴CO₂ produced from the labeled compounds.

View larger version (26K): [in this window] [in a new window]   FIG. 2. Incorporation of 14C-labeled substrates by growing I. hospitalis cells. (A) [1-14C]acetate. (B) [1,4-14C]succinate. (C) [3-14C]pyruvate. (D) [U-14C]glucose. The total radioactivities in the assay (), in the supernatant/cell-free medium (), and in the cells () in relation to the dry cell mass (x) of the growing I. hospitalis culture are shown.

This obligate autotrophic archaeon, like other obligate autotrophic microorganisms, obviously is able to assimilate to a certain extent external carbon sources, such as acetate. Assimilation of acetate requires an acetate-activating enzyme, and the gene coding for an acetate-CoA ligase (AMP forming) was detected in the genome, whereas genes for acetate kinase and phosphotransacetylase were not found. Succinate and pyruvate assimilation do not need activation, and uptake may proceed by passive diffusion of the protonated form of the carboxylic acids. Glucose apparently was not taken up.

Labeling of growing cells with ¹³C-labeled compounds and distribution of ¹³C in individual building blocks. Cells were grown for at least six generations in the presence of 0.5 mM [1-¹³C]acetate, and then amino acids were isolated from the protein fraction and the ¹³C distributions in the individual carbon positions of the amino acids were determined by ¹³C NMR spectroscopy (Table 3). The values indicate that in this experiment up to 20% of the basic building blocks of amino acids were derived from carbons of the exogenous acetate precursor.

View larger version (15K): [in this window] [in a new window]   FIG. 3. Proposed incomplete pathway of CO2 fixation in I. hospitalis. The proposed pathway was deduced from the labeling pattern and the complement of enzymes and genes of I. hospitalis. The circles represent highly 13C-enriched carbon atoms from [1-13C]acetate. Partial randomization of labeling in oxaloacetate is explained by reversible reactions between oxaloacetate, malate, fumarate, and succinate. A reductive carboxylation of succinyl-CoA implies a lack of 13C enrichment at C-1 of 2-oxoglutarate. This is at odds with the labeling data for glutamate, proline, and arginine, which all display high 13C enrichment at C-1. The compounds in boxes were analyzed.

All of the data can be integrated into a hypothetical pathway of CO₂ fixation in I. hospitalis (Fig. 3) that is still incomplete. The labeling of pyruvate predominantly at C-2 is consistent with its synthesis from acetyl-CoA and CO₂ via pyruvate synthase. Oxaloacetate was strongly labeled at C-2, which is also consistent with its biosynthesis from pyruvate via PEP. The partial acquisition of label in C-3 is probably due to the randomization that occurs when oxaloacetate partly equilibrates with symmetric fumarate or succinate.

Interestingly, the labeling of 2-oxoglutarate preferentially at C-1 and C-3 is consistent with its synthesis via citrate and isocitrate involving a (re)-specific citrate synthase. The operation of a (si)-specific enzyme would result in the formation of C-3- and C-5-labeled 2-oxoglutarate. In any case, the reductive carboxylation of succinyl-CoA can be excluded, since a random labeling of C-3 and C-4 would be expected, and in no case should C-1 be labeled.

The labeling of ribose 5-phosphate preferentially at C-1 and C-4 (deduced from the labeling of histidine) indicates that hexose phosphate labeled at C-2 and C-5 was presumably an intermediate (the biosynthetic pathway is shown in Fig. 4). This pattern is consistent with the synthesis of fructose 1,6-bisphosphate from two molecules of C-2-labeled pyruvate via gluconeogenesis and conversion of hexosephosphate to pentosephosphate via the ribulose-monophosphate pathway (see "Special biosynthetic pathways" below).

View larger version (9K): [in this window] [in a new window]   FIG. 4. Biosynthesis of histidine via fructose 6-phosphate from two molecules of PEP. (Left column) Formation of ribulose 5-phosphate by oxidative decarboxylation. (Right column) Formation of ribulose 5-phosphate by the ribulose monophosphate pathway. PHI, 3-hexulose 6-phosphate isomerase; HPS, 3-hexulose 6-phosphate synthase. The circles represent 13C-enriched carbon atoms from [1-13C]acetate. The observed labeling pattern in histidine is shown in the box. The lowercase characters in histidine and ribose phosphate indicate biosynthetically equivalent carbon atoms.

Biosynthesis of aromatic compounds may proceed either via the classical pathway that starts from erythrose 4-phosphate and PEP or via the novel pathway that starts from L-aspartate semialdehyde and 6-deoxy-5-keto-fructose-1-phosphate (49). Coincidentally, the labeling patterns that can be predicted to arise from [1-¹³C]acetate via the two alternatives are very similar (Fig. 5). The apparent enrichment detected for the ortho ring positions in phenylalanine and tyrosine, albeit weak, might suggest the involvement of the aspartate semialdehyde pathway. In the genome, however, both genes for the initial two enzymes of the classical pathway, 3-deoxy-D-arabino-heptulosonate 7-phosphate synthase and 3-dehydroquinate synthase, were present, whereas genes with similarity to the two genes for the new pathway were not found. In addition, a transketolase gene, characteristic of the "classical" pathway of biosynthesis of aromatic amino acids, was found in the genome of I. hospitalis. All known organisms using the aspartate semialdehyde pathway lack the gene for this enzyme in their genomes (45).

View larger version (12K): [in this window] [in a new window]   FIG. 5. Biosynthesis of phenylalanine via fructose 6-phosphate. (Left column) Synthesis of erythrose 4-phosphate by reaction of the nonoxidative pentose phosphate pathway. TK, transketolase; TA, transaldolase. The labeling pattern of dehydroshikimate is constructed from erythrose 4-phosphate and PEP on the basis of the conventional shikimate pathway. (Right column) Synthesis via 6-deoxy-5-ketofructose 1-phosphate and aspartate semialdehyde. The labeling pattern of aspartate semialdehyde was predicted from that of aspartate.

View larger version (18K): [in this window] [in a new window]   FIG. 6. Biosynthesis of lysine. (Left column) Formation of lysine via the common pathway starting from aspartate and pyruvate and leading to diaminopimelate. (Right column) Formation of lysine via the uncommon pathway starting from 2-oxoglutarate and acetyl-CoA and leading to -aminoadipate. The circles represent 13C-enriched carbon atoms from [1-13C]acetate. The observed labeling pattern in lysine is shown in the box. The labeling pattern of 2-oxoglutarate is reconstructed from the observed pattern in glutamate, proline, and arginine.

View larger version (18K): [in this window] [in a new window]   FIG. 7. Biosynthesis of isoleucine. (Left column) Formation of isoleucine via the conventional pathway in which 2-oxobutyrate is formed from threonine. The labeling pattern of threonine corresponds to that of aspartate from which it is formed (Table 3). (Right column) Formation of isoleucine via the unconventional pathway in which 2-oxobutyrate is formed from pyruvate and acetyl-CoA via citramalate and 3-methyloxaloacetate, which is decarboxylated to 2-oxobutyrate. Note that the labeling of 2-oxobutyrate differs in the two pathways. The subsequent steps from 2-oxobutyrate to isoleucine are identical. The circles represent 13C-enriched carbon atoms from [1-13C]acetate. The observed labeling patterns in threonine and isoleucine are shown in boxes. The labeling pattern of pyruvate is reconstructed from the observed pattern in alanine.

	DISCUSSION

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Postulated CO₂-fixing enzymes. Extracts contained pyruvate-acceptor oxidoreductase and PEP carboxlase activities. The specific activities at 85°C were approximately 200 nmol min^–1 mg protein^–1. The required specific CO₂ fixation rate at a generation time of 2 h is around 400 nmol CO₂ fixed min^–1 mg protein^–1. If the new pathway comprises two CO₂ fixation steps, as indicated, the minimal specific enzyme activity of each of the two carboxylating enzymes should be on the order of 200 nmol min^–1 mg protein^–1. It should be noted that the enzyme assays have not been optimized and the activity was tested in the direction of pyruvate oxidation with MV as the electron acceptor or by following the incorporation of label from ¹⁴CO₂ into the carboxyl group via the isotope exchange reaction. In growing cells, the reduction of ferredoxin may be brought about by energy-driven reverse electron transport from NAD(P)H, allowing an effective reductive carboxylation of acetyl-CoA (20). The acetate-labeling experiment clearly showed that pyruvate synthase indeed functions in the reductive-carboxylation direction. A similar phenomenon has been observed in Methanobacterium thermoautotrophicum strain Marburg (renamed Methanothermobacter marburgensis), where acetyl-CoA is converted unidirectionally into pyruvate (11, 46). PEP carboxylation to oxaloacetate by PEP carboxylase is also supported by the labeling pattern of aspartate, where C-4 is derived from CO₂.

Proposed outline of biosynthetic pathways starting from acetyl-CoA. We propose that the autotrophic CO₂ fixation pathway starts from acetyl-CoA. Therefore, we analyzed the assimilation of acetate into cellular building blocks under conditions where 80% of the cell carbon is still derived from CO₂. The observed labeling pattern, together with positive or negative enzyme activity assays and supported by a database search of the genome, are consistent with the following scheme (if not otherwise mentioned, the putative corresponding genes were detected in the genome) (Fig. 1). Acetyl-CoA is reductively carboxylated to pyruvate by pyruvate synthase (see above). Pyruvate is converted into PEP by pyruvate-water dikinase (PEP synthase), and carboxylation of PEP by PEP carboxylase produces oxaloacetate. The citric acid cycle is incomplete and serves only anabolic functions. 2-Oxoglutarate oxidation is missing. Interestingly, the labeling experiments clearly showed that citrate synthase is citrate (re)-synthase (EC 2.3.3.3.) rather than the common (si)-synthase (EC 2.3.3.1.). These two enzymes differ with respect to the side to which acetyl-CoA is added to C-2 of oxaloacetate. They can be discriminated if carbon isotope-labeled acetyl-CoA is used, resulting in chiral citrate (use of unlabeled acetate results in prochiral citrate). Chiral citrate is then converted asymmetrically to 2-oxoglutarate, resulting in differently labeled glutamate. In line with this finding, no citrate (si)-synthase gene was found in the genome.

Gluconeogenesis from PEP appears to be conventional, including a NAD(P)-dependent glyceraldehyde 3-phosphate dehydrogenase. A fructose 1,6-bisphosphate aldolase gene was not detected; however, cell extracts exhibited fructose 1,6-bisphosphate aldolase activity. Fructose 1,6-bisphosphate phosphatase was active in cell extracts, and a putative gene was detected in the genome (see below).

Pentosephosphates are likely derived from hexosephosphate by loss of C-1. The database suggests that this is brought about by enzymes of the ribulose monophosphate pathway (formaldehyde released from C-1) rather than by enzymes of the oxidative pentose phosphate pathway (CO₂ released from C-1).

Catabolic potential. Cultures of I. hospitalis cannot grow on sugars (38), and this inability is reflected by the gene outfit and the enzyme pattern. Genes for key enzymes of glycolytic pathways (44) in Archaea are apparently not present. These include genes for enzymes of the phosphorylating and nonphosphorylating Entner-Doudoroff pathway and genes encoding glucokinase and phosphofructokinase of a (modified) Embden-Meyerhof pathway, as well as pyruvate kinase and pyruvate dehydrogenase. In accordance with this, neither pyruvate kinase nor pyruvate dehydrogenase activity was detected in cell extracts of I. hospitalis. Also, we showed that growing I. hospitalis cells are obviously unable to incorporate glucose. Taken together, the data suggest that I. hospitalis is unable to gain energy or low-molecular-weight building blocks from oxidizing sugars.

Special enzymes and biosynthetic pathways. The central carbon metabolism of I. hospitalis has some special features that are only partly common to Archaea. Citrate synthase is (re)-specific. The corresponding gene has not been detected yet; a gene with similarity to (si)-specific citrate synthase is missing, indicating that the re-specific enzyme is completely different. Citrate (re)-synthase has been reported in Desulfovibrio vulgaris, Desulfovibrio desulfuricans, Clostridium kluyveri, Clostridium acidiurici, and Clostridium cylindrosporum (14-17). The gene for citrate (re)-synthase has been identified as a member of the isopropylmalate synthase enzyme family by expressing the Ignicoccus gene in E. coli and studying the reaction catalyzed by the enzyme (G. Fuchs and R. Say, unpublished results).

PEP carboxylase is a member of the archaeal-type carboxylase (39). Fructose 1,6-bisphosphate aldolase appears to be different from the known enzyme, since activity was detected but a database search did not yield any hit. Glucose 6-phosphate isomerase seems to be a member of the CUPIN superfamily (18, 19, 47), an archaeal enzyme type that was found in several members of the Euryarchaeotes. The new pathway of pentose phosphate synthesis (13, 33, 45), rather than the common oxidative pentose phosphate pathway, seems to be operative, based on the gene outfit and the labeling pattern, which is consistent with this pathway.

Lysine is synthesized via the unconventional -aminoadipate pathway, rather than via the conventional diaminopimelate pathway. The -aminoadipate pathway is common in eukaryotes but was also found in Thermoproteus neutrophilus within the prokaryotes (42). The biosynthesis of the aromatic amino acids seems to proceed via the shikimate pathway, starting from erythrose 4-phosphate and PEP. Isoleucine is made from -ketobutyrate, which seems to be assembled from pyruvate and acetyl-CoA but not from threonine. This citramalate pathway (6) is also found in some other Archaea (3, 8, 23, 42).

	ACKNOWLEDGMENTS

 
We thank Rafael Say, Freiburg, Germany, for valuable database and literature searches. We thank Christine Schwarz, Garching, Germany, for help with NMR spectroscopy and Fritz Wendling, Garching, Germany, for help with preparation of the manuscript. We acknowledge that the valuable genome database of I. hospitalis (http://genome.jgi-psf.org/draft_microbes/ign_k/ign_k.draft.html) was made available to us.

This work was supported by the Deutsche Forschungsgemeinschaft (grant number HU 703/1-2, FU 118/15) and the Fonds der Chemischen Industrie.

	FOOTNOTES

 
* Corresponding author. Mailing address for H. Huber: Lehrstuhl Mikrobiologie und Archaeenzentrum, Universität Regensburg, Universitätsstrasse 31, D-93053 Regensburg, Germany. Phone: 49 941 9433185. Fax: 49 941 9432403. E-mail: Harald.huber{at}biologie.uni-regensburg.de. Mailing address for G. Fuchs: Mikrobiologie, Biologie, Schaenzlestrasse 1, D-79104 Freiburg, Germany. Phone: 49 761 2032649. Fax: 9 761 2032626. E-mail: georg.fuchs{at}biologie.uni-freiburg.de

Published ahead of print on 30 March 2007.

Present address: Leibniz-Institut für Meereswissenschaften, IFM-GEOMAR, Kiel, Germany.

	REFERENCES

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

1. Alber, B., M. Olinger, A. Rieder, D. Kockelkorn, B. Jobst, M. Hügler, and G. Fuchs. 2006. Malonyl-coenzyme A reductase in the modified 3-hydroxypropionate cycle for autotrophic carbon fixation in archaeal Metallosphaera and Sulfolobus spp. J. Bacteriol. 188:8551-8559.[Abstract/Free Full Text]
2. Alber, B. E., and G. Fuchs. 2002. Propionyl-coenzyme A synthase from Chloroflexus aurantiacus, a key enzyme of the 3-hydroxypropionate cycle for autotrophic CO₂ fixation. J. Biol. Chem. 277:12137-12143.[Abstract/Free Full Text]
3. Bacher, A., C. Rieder, D. Eichinger, D. Arigoni, G. Fuchs, and W. Eisenreich. 1999. Elucidation of novel biosynthetic pathways and metabolic flux pattern by retrobiosynthetic NMR analysis. FEMS Microbiol. Rev. 22:567-598.[CrossRef]
4. Bhattacharjee, J. K. 1985. -Aminoadipate pathway for the biosynthesis of lysine in lower eukaryotes. Crit. Rev. Microbiol. 12:131-151.[Medline]
5. Bradford, M. M. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72:248-254.[CrossRef][Medline]
6. Charon, N. W., R. Johnson, and D. Peterson. 1974. Amino acid biosynthesis in the spirochete Leptospira: evidence for a novel pathway of isoleucine biosynthesis. J. Bacteriol. 117:203-211.[Abstract/Free Full Text]
7. Dawson, R. M., D. C. Elliot, W. H. Elliot, and K. M. Jones. 1986. Data for biochemical research, 3rd ed. Clarendon Press, Oxford, United Kingdom.
8. Eikmanns, B., D. Linder, and R. K. Thauer. 1983. Unusual pathway of isoleucine biosynthesis in Methanobacterium thermoautotrophicum. Arch. Microbiol. 136:111-113.[CrossRef]
9. Eisenreich, W., and A. Bacher. 2000. Elucidation of biosynthetic pathways by retrodictive/predictive comparison of isotopomer patterns determined by NMR spectroscopy. Genet. Eng. 22:121-153.
10. Eisenreich, W., B. Schwarzkopf, and A. Bacher. 1991. Biosynthesis of nucleotides, flavins, and deazaflavins in Methanobacterium thermoautotrophicum. J. Biol. Chem. 266:9622-9631.[Abstract/Free Full Text]
11. Fuchs, G., and E. Stupperich. 1980. Acetyl-CoA, a central intermediate of autotrophic CO₂ fixation in Methanobacterium thermoautotrophicum. Arch. Microbiol. 127:267-272.[CrossRef]
12. Gallagher, S., S. E. Winston, S. A. Fuller, and J. G. R. Hurrell. 1997. Immunoblotting and immunodetection, p. 10.8.1-10.8.21. In F. M. Ausubel et al. (ed.), Current protocols in molecular biology. John Wiley & Sons, Inc., New York, NY.
13. Goenrich, M., R. K. Thauer, H. Yurimoto, and N. Kato. 2005. Formaldehyde activating enzyme (Fae) and hexulose-6-phosphate synthase (Hps) in Methanosarcina barkeri: a possible function in ribose-5-phosphate biosynthesis. Arch. Microbiol. 184:41-48.[CrossRef][Medline]
14. Gottschalk, G. 1968. The stereospecificity of the citrate synthase in sulfate-reducing and photosynthetic bacteria. Eur. J. Biochem. 5:346-351.[Medline]
15. Gottschalk, G. 1969. Partial purification and some properties of the (R)-citrate synthase from Clostridium acidiurici. Eur. J. Biochem. 7:301-306.[Medline]
16. Gottschalk, G., and H. A. Barker. 1966. Synthesis of glutamate and citrate by Clostridium kluyveri. A new type of citrate synthase. Biochemistry 5:1125-1133.[CrossRef][Medline]
17. Gottschalk, G., and H. A. Barker. 1967. Presence and stereospecificity of citrate synthase in anaerobic bacteria. Biochemistry 6:1027-1034.[CrossRef][Medline]
18. Hansen, T., M. Oehlmann, and P. Schönheit. 2001. Novel type of glucose-6-phosphate isomerase in the hyperthermophilic archaeon Pyrococcus furiosus. J. Bacteriol. 183:3428-3435.[Abstract/Free Full Text]
19. Hansen, T., B. Schlichting, M. Felgendreher, and P. Schönheit. 2005. Cupin-type phosphoglucose isomerases (Cupin-PGIs) constitute a novel metal-dependent PGI family representing a convergent line of PGI evolution. J. Bacteriol. 187:1621-1631.[Abstract/Free Full Text]
20. Hedderich, R. 2004. Energy-converting [NiFe] hydrogenases from archaea and extremophiles: ancestors of complex I. J. Bioenerg. Biomembr. 36:65-75.[CrossRef][Medline]
21. Herter, S., J. Farfsing, N. Gad'On, C. Rieder, W. Eisenreich, A. Bacher, and G. Fuchs. 2001. Autotrophic CO₂ fixation by Chloroflexus aurantiacus: study of glyoxylate formation and assimilation via the 3-hydroxypropionate cycle. J. Bacteriol. 183:4305-4316.[Abstract/Free Full Text]
22. Howard, J. L., and S. M. Ridley. 1990. Acetyl-CoA carboxylase: a rapid novel assay procedure used in conjunction with the preparation of enzyme from maize leaves. FEBS Lett. 261:261-264.[CrossRef]
23. Howell, D. M., H. Xu, and R. H. White. 1999. (R)-citramalate synthase in methanogenic archaea. J. Bacteriol. 181:331-333.[Abstract/Free Full Text]
24. Huber, H., S. Burggraf, T. Mayer, I. Wyschkony, R. Rachel, and K. O. Stetter. 2000. Ignicoccus gen. nov., a novel genus of hyperthermophilic, chemolithoautotrophic Archaea, represented by two new species, Ignicoccus islandicus sp. nov. and Ignicoccus pacificus sp. nov. Int. J. Syst. Evol. Microbiol. 50:2093-2100.[Abstract]
25. Huber, H., M. J. Hohn, R. Rachel, T. Fuchs, V. C. Wimmer, and K. O. Stetter. 2002. A new phylum of Archaea represented by a nanosized hyperthermophilic symbiont. Nature 417:63-67.[CrossRef][Medline]
26. Huber, H., M. J. Hohn, K. O. Stetter, and R. Rachel. 2003. The phylum Nanoarchaeota: present knowledge and future perspectives of a unique form of life. Res. Microbiol. 154:165-171.[Medline]
27. Huber, G., C. Spinnler, A. Gambacorta, and K. O. Stetter. 1989. Metallosphaera sedula, gen. and sp. nov. represents a new genus of aerobic, metal-mobilizing, thermoacidophilic archaebacteria. Syst. Appl. Microbiol. 12:38-47.[Medline]
28. Hügler, M., H. Huber, K. O. Stetter, and G. Fuchs. 2003. Autotrophic CO₂ fixation pathways in archaea (Crenarchaeota). Arch. Microbiol. 179:160-173.[Medline]
29. Hügler, M., C. Menendez, H. Schägger, and G. Fuchs. 2002. Malonyl-coenzyme A reductase from Chloroflexus aurantiacus, a key enzyme of the 3-hydroxypropionate cycle for autotrophic CO₂ fixation. J. Bacteriol. 184:2404-2410.[Abstract/Free Full Text]
30. Hügler, M., C. O. Wirsen, G. Fuchs, C. D. Taylor, and S. M. Sievert. 2005. Evidence for autotrophic CO₂ fixation via the reductive tricarboxylic acid cycle in members of the -Proteobacteria. J. Bacteriol. 187:3020-3027.[Abstract/Free Full Text]
31. Imanaka, H., A. Yamatsu, T. Fukui, H. Atomi, and T. Imanaka. 2006. Phosphoenol-pyruvate synthase plays an essential role for glycolysis in the modified Embden-Meyerhof pathway in Thermococcus kodakarensis. Mol. Microbiol. 61:898-909.[CrossRef][Medline]
31. Ishii, M., T. Miyake, T. Satoh, H. Sugiyama, Y. Oshima, T. Kodama, and Y. Igarashi. 1996. Autotrophic carbon dioxide fixation in Acidianus brierleyi. Arch. Microbiol. 166:368-371.[CrossRef][Medline]
32. Jahn, U., R. Summons, H. Sturt, E. Grosjean, and H. Huber. 2004. Composition of the lipids of Nanoarchaeum equitans and their origin from its host Ignicoccus sp. strain KIN4/I. Arch. Microbiol. 182:404-413.[CrossRef][Medline]
33. Kato, N., H. Yurimoto, and R. K. Thauer. 2006. The physiological role of the ribulose monophosphate pathway in bacteria and archaea. Biosci. Biotechnol. Biochem. 70:10-21.[CrossRef][Medline]
34. Madigan, M., and J. Martinko. 2006. Brock biology of microorganisms, 11th ed. Prentice Hall, Upper Saddle River, NJ.
35. Menendez, C., Z. Bauer, H. Huber, N. Gad'on, K. O. Stetter, and G. Fuchs. 1999. Presence of acetyl coenzyme A (CoA) carboxylase and propionyl-CoA carboxylase in autotrophic Crenarchaeota and indication for operation of a 3-hydroxypropionate cycle in autotrophic carbon fixation. J. Bacteriol. 181:1088-1098.[Abstract/Free Full Text]
36. Mukhopadhyay, B., V. J. Patel, and R. S. Wolfe. 2000. A stable archaeal pyruvate carboxylase from the hyperthermophile Methanococcus jannaschii. Arch. Microbiol. 174:406-414.[CrossRef][Medline]
37. Mukhopadhyay, B., and E. Purwantini. 2000. Pyruvate carboxylase from Mycobacterium smegmatis: stabilization, rapid purification, molecular and biochemical characterization and regulation of the cellular level. Biochim. Biophys. Acta 1475:191-206.[Medline]
38. Paper, W., U. Jahn, M. J. Hohn, M. Brandl, D. J. Näther, T. Burghardt, R. Rachel, K. O. Stetter, and H. Huber. 2007. Ignicoccus hospitalis sp. nov., the host of Nanoarchaeum equitans. Int. J. Syst. Evol. Microbiol. 57:803-808.[Ab