Portal de Periódicos da CAPES

Presence of a Structurally Novel Type Ribulose-bisphosphate Carboxylase/Oxygenase in the Hyperthermophilic Archaeon,Pyrococcus kodakaraensis KOD1

1999; Elsevier BV; Volume: 274; Issue: 8 Linguagem: Inglês

10.1074/jbc.274.8.5078

1083-351X

Satoshi Ezaki, Norihiro Maeda, Tsukuru Kishimoto, Haruyuki Atomi, Tadayuki Imanaka,

Hemoglobin structure and function

We have characterized the gene encoding ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco) of the hyperthermophilic archaeon, Pyrococcus kodakaraensis KOD1. The gene encoded a protein consisting of 444 amino acid residues, corresponding in size to the large subunit of previously reported Rubiscos. Rubisco of P. kodakaraensis KOD1 (Pk-Rubisco) showed only 51.4% similarity with the large subunit of type I Rubisco from spinach and 47.3% with that of type II Rubisco from Rhodospirillum rubrum, suggesting that the enzyme was not a member of either type. Active site residues identified from type I and type II Rubiscos were conserved. We expressed the gene in Escherichia coli, and we obtained a soluble protein with the expected molecular mass and N-terminal amino acid sequence. Purification of the recombinant protein revealed thatPk-Rubisco was an L8 type homo-octamer.Pk-Rubisco showed highest specific activity of 19.8 × 103 nmol of CO2 fixed per min/mg, and a τ value of 310 at 90 °C, both higher than any previously characterized Rubisco. The optimum pH was 8.3, and the enzyme possessed extreme thermostability, with a half-life of 15 h at 80 °C. Northern blot analysis demonstrated that the gene was transcribed inP. kodakaraensis KOD1. Furthermore, Western blot analysis with cell-free extract of P. kodakaraensisKOD1 clearly indicated the presence of Pk-Rubisco in the native host cells. We have characterized the gene encoding ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco) of the hyperthermophilic archaeon, Pyrococcus kodakaraensis KOD1. The gene encoded a protein consisting of 444 amino acid residues, corresponding in size to the large subunit of previously reported Rubiscos. Rubisco of P. kodakaraensis KOD1 (Pk-Rubisco) showed only 51.4% similarity with the large subunit of type I Rubisco from spinach and 47.3% with that of type II Rubisco from Rhodospirillum rubrum, suggesting that the enzyme was not a member of either type. Active site residues identified from type I and type II Rubiscos were conserved. We expressed the gene in Escherichia coli, and we obtained a soluble protein with the expected molecular mass and N-terminal amino acid sequence. Purification of the recombinant protein revealed thatPk-Rubisco was an L8 type homo-octamer.Pk-Rubisco showed highest specific activity of 19.8 × 103 nmol of CO2 fixed per min/mg, and a τ value of 310 at 90 °C, both higher than any previously characterized Rubisco. The optimum pH was 8.3, and the enzyme possessed extreme thermostability, with a half-life of 15 h at 80 °C. Northern blot analysis demonstrated that the gene was transcribed inP. kodakaraensis KOD1. Furthermore, Western blot analysis with cell-free extract of P. kodakaraensisKOD1 clearly indicated the presence of Pk-Rubisco in the native host cells. Ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco 1The abbreviations Rubiscoribulose-1,5-bisphosphate carboxylase/oxygenaseBicineN,N-bis[2-hydroxyethyl]glycineORFopen reading framePk-RubiscoRubisco from Pyrococcus kodakaraensis KOD1; 1The abbreviations Rubiscoribulose-1,5-bisphosphate carboxylase/oxygenaseBicineN,N-bis[2-hydroxyethyl]glycineORFopen reading framePk-RubiscoRubisco from Pyrococcus kodakaraensis KOD1EC 4.1.1.39) catalyzes the covalent addition of CO2 to ribulose 1,5-bisphosphate producing two molecules of 3-phosphoglycerate (reviewed in Ref. 1Hartman F.C. Harpel M.R. Annu. Rev. Biochem. 1994; 63: 197-234Crossref PubMed Google Scholar). This reaction is the first step in the fixation of atmospheric CO2 into the food chain. A competing side reaction occurs in the presence of oxygen, leading to one molecule of 3-phosphoglycerate and one molecule of 2-phosphoglycolate. ribulose-1,5-bisphosphate carboxylase/oxygenase N,N-bis[2-hydroxyethyl]glycine open reading frame Rubisco from Pyrococcus kodakaraensis KOD1 ribulose-1,5-bisphosphate carboxylase/oxygenase N,N-bis[2-hydroxyethyl]glycine open reading frame Rubisco from Pyrococcus kodakaraensis KOD1 Rubisco is found predominantly in higher plants, algae, cyanobacteria, and other photosynthetic organisms (reviewed in Ref. 2Tabita F.R. Blankenship R.E. Madigan M.T. Bauer C.E. Anoxygenic Photosynthetic Bacteria. 2. Kluwer Academic Publishers Group, Drodrecht, Netherlands1995: 885-914Google Scholar). In most of these organisms, Rubisco is a hexadecamer consisting of eight large (L) and eight small (S) subunits in an L8S8arrangement (type I or form I). Another type of Rubisco, found in some photosynthetic β-purple bacteria like Rhodospirillum rubrum, is a homodimer of L subunits only (type II or form II) (3Hartman F.C. Stringer C.D. Lee E.H. Arch. Biochem. Biophys. 1984; 232: 280-295Crossref PubMed Scopus (33) Google Scholar). The nonphotosynthetic chemoautotrophic β-purple bacteriumThiobacillus denitrificans also harbors a type II Rubisco, along with a distinct type I Rubisco (4Hernandez J.M. Baker S.H. Lorbach S.C. Shively J.M. Tabita F.R. J. Bacteriol. 1996; 178: 347-356Crossref PubMed Google Scholar). A new classification of Rubisco enzymes, based on their primary structure relatedness, has also recently been proposed by Tabita (2Tabita F.R. Blankenship R.E. Madigan M.T. Bauer C.E. Anoxygenic Photosynthetic Bacteria. 2. Kluwer Academic Publishers Group, Drodrecht, Netherlands1995: 885-914Google Scholar). A detailed phylogenetic analysis of Rubiscos has also been conducted (5Watson G.M.F. Tabita F.R. FEMS Microbiol. Lett. 1997; 146: 13-22Crossref PubMed Google Scholar), and from these studies, type I enzymes can first be divided into two major branches, the "Green-like" and "Red-like" Rubiscos. The two groups can be further divided into two subgroups, respectively. Type II Rubiscos form a uniform class of enzymes including those from R. rubrum,Rhodobacter sphaeroides, and T. denitrificans. The incongruency of Rubisco phylogenies with those derived from 16 S rRNA sequences (6Woese C.R. Microbiol. Rev. 1987; 51: 221-271Crossref PubMed Google Scholar), or other macromolecules, has suggested the occurrence of horizontal gene transfer during the evolution of Rubisco and the organisms that harbor the enzyme (5Watson G.M.F. Tabita F.R. FEMS Microbiol. Lett. 1997; 146: 13-22Crossref PubMed Google Scholar). Concerning archaea, the Rubisco from halobacterium Haloferax mediterranei is the only enzyme that has been characterized (7Rajagopalan R. Altekar W. Eur. J. Biochem. 1994; 221: 863-869Crossref PubMed Scopus (16) Google Scholar).H. mediterranei Rubisco is composed of L and S subunits and, from this viewpoint, corresponds to type I Rubisco. Isolation of the genes encoding the subunits has not been characterized as yet, and therefore the primary structures have not been determined. Although there have been no reports of a Rubisco from a hyperthermophilic archaeon, putative Rubisco genes have been reported through genome analysis of the archaea Methanococcus jannaschii (8Bult C.J. White O. Olsen G.J. Zhou L. Fleischmann R.D. Sutton G.G. Blake J.A. FitzGerald L.M. Clayton R.A. Gocayne J.D. Kerlavage A.R. Dougherty B.A. Tomb J.F. Adams M.D. Reich C.I. Overbeek R. Kirkness E.F. Weinstock K.G. Merrick J.M. Glodek A. Scott J.L. Geoghagen N.S.M. Weidman J.F. Fuhrmann J.L. Nguyen D.T. Utterback T.R. Kelley J.M. Peterson J.D. Sadow P.W. Hanna M.C. Cotton M.D. Roberts K.M. Hurst M.A. Kaine B.P. Borodovsky M. Klenk H.P. Fraser C.M. Smith H.O. Woese C.R. Venter J.C. Science. 1996; 273: 1058-1073Crossref PubMed Scopus (2284) Google Scholar) andArchaeoglobus fulgidus (9Klenk H.P. Clayton R.A. Tomb J.F. White O. Nelson K.E. Ketchum K.A. Dodson R.J. Gwinn M. Hickey E.K. Peterson J.D. Richardson D.L. Kerlavage A.R. Graham D.E. Kyrpides N.C. Fleischmann R.D. Quackenbush J. Lee N.H. Sutton G.G. Gill S. Kirkness E.F. Dougherty B.A. McKenney K. Adams M.D. Loftus B. Peterson S. Reich C.I. McNeil L.K. Badger J.H. Glodek A. Zhou L. Overbeek R. Gocayne J.D. Weidman J.F. McDonald L. Utterback T. Cotton M.D. Spriggs T. Artiach P. Kaine B.P. Sykes S.M. Sadow P.W. D'Andrea K.P. Bowman C. Fujii C. Garland S.A. Mason T.M. Olsen G.J. Fraser C.M. Smith H.O. Woese C.R. Venter J.C. Nature. 1997; 390: 364-370Crossref PubMed Scopus (1201) Google Scholar). The deduced amino acid sequences indicate that they do not belong to the type I and type II Rubiscos. However, the protein products of the genes have not been identified nor characterized, and evidence that they actually harbor Rubisco activity is an urgent subject of Rubisco research (5Watson G.M.F. Tabita F.R. FEMS Microbiol. Lett. 1997; 146: 13-22Crossref PubMed Google Scholar). Pyrococcus kodakaraensis KOD1 is a hyperthermophilic archaeon that was isolated from a solfatara at a wharf of Kodakara Island, Kagoshima, Japan (10Morikawa M. Izawa Y. Rashid N. Hoaki T. Imanaka T. Appl. Environ. Microbiol. 1994; 60: 4559-4566Crossref PubMed Google Scholar). P. kodakaraensis KOD1 shows optimum growth at 95° C, and enzymes characterized from the strain have all shown extreme thermostability. We have been conducting genome analysis and have identified and characterized genes such as those encoding chaperonin (11Yan Z. Fujiwara S. Kohda K. Takagi M. Imanaka T. Appl. Environ. Microbiol. 1997; 63: 785-789Crossref PubMed Google Scholar), DNA polymerase (12Takagi M. Nishioka M. Kakihara H. Kitabayashi M. Inoue H. Kawakami B. Oka M. Imanaka T. Appl. Environ. Microbiol. 1997; 63: 4504-4510Crossref PubMed Google Scholar), ribose phosphate pyrophosphokinase (13Rashid N. Morikawa M. Imanaka T. J. Ferment. Bioeng. 1997; 83: 412-418Crossref Scopus (10) Google Scholar), DNA recombination protein (14Rashid N. Morikawa M. Nagahisa K. Kanaya S. Imanaka T. Nucleic Acids Res. 1997; 25: 719-726Crossref PubMed Scopus (24) Google Scholar), and indolepyruvate ferredoxin oxidoreductase (15Siddiqui M.A. Fujiwara S. Imanaka T. Mol. Gen. Genet. 1997; 254: 433-439PubMed Google Scholar). Many of the proteins from P. kodakaraensis KOD1 show distinct structures and characteristics compared with proteins from eukaryotes and bacteria (11Yan Z. Fujiwara S. Kohda K. Takagi M. Imanaka T. Appl. Environ. Microbiol. 1997; 63: 785-789Crossref PubMed Google Scholar, 12Takagi M. Nishioka M. Kakihara H. Kitabayashi M. Inoue H. Kawakami B. Oka M. Imanaka T. Appl. Environ. Microbiol. 1997; 63: 4504-4510Crossref PubMed Google Scholar, 13Rashid N. Morikawa M. Imanaka T. J. Ferment. Bioeng. 1997; 83: 412-418Crossref Scopus (10) Google Scholar, 14Rashid N. Morikawa M. Nagahisa K. Kanaya S. Imanaka T. Nucleic Acids Res. 1997; 25: 719-726Crossref PubMed Scopus (24) Google Scholar, 15Siddiqui M.A. Fujiwara S. Imanaka T. Mol. Gen. Genet. 1997; 254: 433-439PubMed Google Scholar). Through sequence analysis of anAscI–AscI fragment (200 kilobase pairs) of the genome, we found an open reading frame that displayed similarity with previously reported Rubisco genes. Here we report that the gene (Pk-rbc1) actually encodes an extremely thermostable and highly carboxylase specific Rubisco and that the primary structure is distinct from type I and type II Rubiscos. P. kodakaraensisKOD1 was cultivated using a medium described in Ref. 10Morikawa M. Izawa Y. Rashid N. Hoaki T. Imanaka T. Appl. Environ. Microbiol. 1994; 60: 4559-4566Crossref PubMed Google Scholar with additional saturation of carbon dioxide by bubbling with dry ice.Escherichia coli JM109 and the vector pUC19 were used for cloning and gene manipulation. E. coli BL21(DE3) (Stratagene, La Jolla, CA) and the vector pET21a(+) (Novagen Inc., Madison, WI) were used for overexpression of Pk-rbc1. Pk-rbc1 was originally identified on a DNA fragment inserted into λ EMBL4 phage vector. LB medium was used for cultivation of E. coli and NZYM medium for amplification of phage (16Maniatis T. Fritsch E.F. Sambrook J. Molecular Cloning: A Laboratory Manual. 2nd Ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1989Google Scholar). For isolation of plasmid DNA, Plasmid Mini-, Midi- and Lambda Kits (Qiagen, Hilden, Germany) were used along with the alkaline extraction method (16Maniatis T. Fritsch E.F. Sambrook J. Molecular Cloning: A Laboratory Manual. 2nd Ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1989Google Scholar). Restriction enzymes and modifying enzymes were purchased from Toyobo (Osaka, Japan), Takara (Kyoto, Japan), and Boehringer Mannheim (Mannheim, Germany). DNA sequencing on both strands of DNA was conducted using the ABI PRISM kit and Model 310 capillary DNA sequencer (Perkin-Elmer). The multiple alignment of protein sequences and the number of similarity between sequences were obtained with the program ALIGN contained within the ClustalW program provided by DNA Data Bank Japan (DDBJ). After insertion of the Pk-rbc1 into pET21a(+), the plasmid was transformed intoE. coli BL21(DE3). Transformants were cultivated until optical density at 660 nm reached 0.5. Isopropyl–d-thiogalactopyranoside (0.1 mm) was added to induce gene expression for 4 h, when cells were harvested and disrupted by sonication. After centrifugation (7,500 ×g, 10 min), the soluble cell-free extract was heat-precipitated for 30 min at 85° C. Samples were centrifuged (17,000 × g, 10 min), and the supernatant was applied to a ResourceQ column (Amersham Pharmacia Biotech, Uppsala, Sweden) and eluted with 0 to 1.0 m NaCl gradient in 100 mmBicine/10 mm MgCl2 buffer using AKTA explorer 10S (Amersham Pharmacia Biotech). Purified Rubisco was applied to Superdex 200HR 10/30 column in 50 mm sodium phosphate, 0.15 m NaCl buffer (pH 7.0). 3.68 μg of enzyme in 40 μl of 100 mm Bicine/KOH (pH 8.3), 10 mmMgCl2 buffer, and 10 μl of 15 mm ribulose 1,5-bisphosphate (Sigma) in 100 mm Bicine/KOH (pH 8.3) buffer were incubated under a 100% CO2 atmosphere in the experiments described in Fig. 4. Enzyme and substrate were incubated in air in order to determine the τ values in Table I. All buffers were prepared at the temperatures at which activity measurements were conducted. Enzyme and substrate were incubated at the desired temperature for 4 and 1 min, respectively, in closed vials. After incubation, the two solutions were mixed, and the enzymatic reaction was carried out at the same temperature for 5, 10, and 15 min. The products of carboxylase and oxygenase reaction (3-phosphoglycerate or 3-phosphoglycerate and phosphoglycolate for the carboxylase and oxygenase reaction, respectively) were detected and quantified using ASAHI PAK GS-220HQ column (Showa Denko K.K., Tokyo, Japan) by UV absorbance at 210 nm following high performance liquid chromatography (Shimadzu, Kyoto, Japan). Standards of 3-phosphoglycerate and phosphoglycolate were added after each measurement, and an additional run was conducted in order to confirm the retention time of each compound. Thermostability measurements were conducted by incubating the enzyme in 100 mm Bicine/KOH (pH 8.3), 10 mmMgCl2 buffer for the desired time prior to saturation with CO2. Activity without heat treatment and residual activity after heat treatment to examine thermostability were measured at 60° C.Table Iτ values of Pk-Rubisco at various temperaturesTemperatureτ value°C507060140702508029090310 Open table in a new tab A probe corresponding to the complete open reading frame of Pk-rbc1 was constructed using the DIG DNA Labeling and Detection Kit (Boehringer Mannheim). Hybond-N+ membranes (Amersham Pharmacia Biotech) were used for blotting. 25 μg of total RNA was applied to agarose gel electrophoresis. Polyclonal rabbit antibodies were raised against purified recombinant Pk-Rubisco. Antisera diluted −500-fold were used to detect Pk-Rubisco in cell-free extracts of P. kodakaraensis KOD1. Polyvinylidene difluoride membranes from ATTO (Tokyo, Japan) were used for blotting. Through sequence analysis of the genome ofP. kodakaraensis KOD1, we found an open reading frame of 1332 base pairs (Pk-rbc1), which showed similarity with previously characterized genes encoding Rubisco. The deduced amino acid sequence consisted of 444 residues, corresponding in size to the large subunit of previously characterized Rubiscos (Fig.1). We compared the amino acid sequence of Pk-Rubisco with other previously reported enzymes. Type I and type II enzymes showed high similarity when sequences were compared among Rubiscos of the same type. Rubisco from spinach (Spinacia oleracea) (17Zurawski G. Perrot B. Bottomley W. Whitfeld P.R. Nucleic Acids Res. 1981; 9: 3251-3270Crossref PubMed Scopus (273) Google Scholar) showed 83.7 and 71.6% similarity with the type I enzymes from T. denitrificans (4Hernandez J.M. Baker S.H. Lorbach S.C. Shively J.M. Tabita F.R. J. Bacteriol. 1996; 178: 347-356Crossref PubMed Google Scholar) and R. sphaeroides (18Gibson J.L. Falcone D.L. Tabita F.R. J. Biol. Chem. 1991; 266: 14646-14653Abstract Full Text PDF PubMed Google Scholar), respectively. The type II enzyme from R. rubrum (19Nargang F.E. McIntosh L. Somerville C. Mol. Gen. Genet. 1984; 193: 220-224Crossref Scopus (95) Google Scholar) displayed 88.2 and 82.9% similarity with the type II enzymes from R. sphaeroides (20Chen J.H. Gibson J.L. McCue L.A. Tabita F.R. J. Biol. Chem. 1991; 266: 20447-20452Abstract Full Text PDF PubMed Google Scholar) and T. denitrificans (4Hernandez J.M. Baker S.H. Lorbach S.C. Shively J.M. Tabita F.R. J. Bacteriol. 1996; 178: 347-356Crossref PubMed Google Scholar), respectively. However, spinach Rubisco showed only 43.3% similarity with the type II enzyme from R. sphaeroides, indicating that type I and type II enzymes are distinct in primary structure. In Fig. 2, we compared Pk-Rubisco with Rubisco from spinach as a typical type I enzyme, and the type II Rubisco from the α-purple bacterium R. rubrum. As P. kodakaraensis KOD1 is a hyperthermophilic archaeon, we also included three putative Rubisco sequences from archaea. One from M. jannaschii showed 58.8% similarity to the large subunit of Synechococcus PCC6301 Rubisco (8Bult C.J. White O. Olsen G.J. Zhou L. Fleischmann R.D. Sutton G.G. Blake J.A. FitzGerald L.M. Clayton R.A. Gocayne J.D. Kerlavage A.R. Dougherty B.A. Tomb J.F. Adams M.D. Reich C.I. Overbeek R. Kirkness E.F. Weinstock K.G. Merrick J.M. Glodek A. Scott J.L. Geoghagen N.S.M. Weidman J.F. Fuhrmann J.L. Nguyen D.T. Utterback T.R. Kelley J.M. Peterson J.D. Sadow P.W. Hanna M.C. Cotton M.D. Roberts K.M. Hurst M.A. Kaine B.P. Borodovsky M. Klenk H.P. Fraser C.M. Smith H.O. Woese C.R. Venter J.C. Science. 1996; 273: 1058-1073Crossref PubMed Scopus (2284) Google Scholar), and two from the chemoautotrophic sulfate-reducing archaeon A. fulgidus (rbcL-1 andrbcL-2) showed 41 and 45% identity with the M. jannaschii Rubisco (9Klenk H.P. Clayton R.A. Tomb J.F. White O. Nelson K.E. Ketchum K.A. Dodson R.J. Gwinn M. Hickey E.K. Peterson J.D. Richardson D.L. Kerlavage A.R. Graham D.E. Kyrpides N.C. Fleischmann R.D. Quackenbush J. Lee N.H. Sutton G.G. Gill S. Kirkness E.F. Dougherty B.A. McKenney K. Adams M.D. Loftus B. Peterson S. Reich C.I. McNeil L.K. Badger J.H. Glodek A. Zhou L. Overbeek R. Gocayne J.D. Weidman J.F. McDonald L. Utterback T. Cotton M.D. Spriggs T. Artiach P. Kaine B.P. Sykes S.M. Sadow P.W. D'Andrea K.P. Bowman C. Fujii C. Garland S.A. Mason T.M. Olsen G.J. Fraser C.M. Smith H.O. Woese C.R. Venter J.C. Nature. 1997; 390: 364-370Crossref PubMed Scopus (1201) Google Scholar). From sequence comparison, active site residues, namely Lys201, Asp203, His294, and Lys334 of the enzyme from spinach, were conserved among all protein sequences (Fig. 2). As forPk-Rubisco, 51.4% similarity was observed with spinach Rubisco and 47.3% similarity with the type II enzyme from R. rubrum, suggesting that Pk-Rubisco was not a member of either type. Interestingly, Pk-Rubisco showed 81.9% similarity with the putative Rubisco RBCL-2 of A. fulgidus. Pk-Rubisco showed lower similarities, 61.3 and 50.1%, withM. jannaschii Rubisco and RBCL-1, respectively. Amino acid residues showing similarity among archaeal Rubiscos were also well conserved in the type I and type II Rubisco enzymes. There were no notable regions where high similarity was found only between archaeal enzymes and type I or type II enzymes.Figure 2Amino acid sequence alignment ofPk-Rubisco with various Rubiscos from other sources. Sequences compared were type I Rubisco from spinach (Sp), type II enzyme from R. rubrum(Rr), three putative Rubisco sequences from archaea, RBCL-1 and RBCL-2 from A. fulgidus (Af1 andAf2) and M. jannaschii (Mj). Residues with similarity among archaeal enzymes are indicated withbars. Conserved residues among archaeal and type I or type II enzymes are indicated with an asterisk. Active site residues referred to in the text are indicated with #. Gaps are indicated with a minus.View Large Image Figure ViewerDownload (PPT) As the structure ofPk-Rubisco was distinct from type I and type II Rubiscos, we expressed Pk-rbc1 in E. coli in order to examine its catalytic activity. High levels of a protein with a molecular mass of approximately 48 kDa was observed on SDS gels (Fig.3, lane 2). Following heat precipitation of the cell-free extract, we were able to purify the protein by ion exchange and gel filtration chromatography (Fig. 3,lane 5). The N-terminal sequence of the protein was Val-Glu-Lys-Phe-Asp-Thr-Ile-Tyr-Asp-Tyr-Tyr, which was identical to the predicted sequence beginning with the second amino acid residue deduced from the open reading frame. We also investigated the subunit composition of the enzyme by gel filtration chromatography. The native enzyme had a molecular mass of 450 kDa, indicating thatPk-Rubisco was an L8 type homo-octamer (data not shown). We next investigated the enzymatic activity of the purified Pk-Rubisco. Measurements of carboxylase activity were conducted at various temperatures. We could clearly detect significant carboxylase activity from 40 to 100° C, and the optimum temperature was extremely high, at 90° C (Fig. 4 A). The optimum pH of the enzyme was 8.3 (data not shown). The enzyme also showed high thermostability, with a half-life of 2.5 h at 100° C and 15 h at 80° C (Fig. 4 B). The specific activity ofPk-Rubisco at 90° C was 19.8 × 103 nmol of CO2 fixed per min/mg (Fig. 4 A). This specific activity level is higher than any previously characterized Rubisco. Another important parameter in the case of Rubisco is the τ value, which is the ratio of carboxylase activity to oxygenase activity. The τ value of Pk-Rubisco was 70 at 50° C, 140 at 60° C, 250 at 70° C, 290 at 80° C, and 310 at 90° C (TableI). The increase in τ value ofPk-Rubisco at elevated temperatures is totally distinct from previously characterized Rubiscos, whose τ values decrease with increase in temperature (21Uemura K. Anwaruzzaman Miyachi S. Yokota A. Biochem. Biophys. Res. Commun. 1997; 233: 568-571Crossref PubMed Scopus (135) Google Scholar). In order to determine whether Pk-rbc1 was actually expressed in the host cells, we conducted Northern blot analysis using a DNA probe corresponding to the complete open reading frame of Pk-rbc1. An mRNA with a length of 1,400 bases specifically hybridized with the Pk-Rubisco probe (Fig.5 A), confirming thatPk-rbc1 is transcribed in the native host. The length of the mRNA strongly indicates that Pk-rbc1 is transcribed as a single gene and not a member of a multi-gene operon. In order to conduct Western blot analysis, polyclonal antibodies were raised against the recombinant Pk-Rubisco. A specific band with a molecular mass identical with the recombinant protein and corresponding well with the size of Pk-Rubisco deduced from its amino acid sequence (49,710 Da) was observed (Fig. 5 B). The above results prove that Pk-rbc1 is transcribed and translated and that the protein product, Pk-Rubisco, is present in the native host, P. kodakaraensis KOD1. We have identified and characterized a Rubisco gene from the hyperthermophilic archaeon P. kodakaraensis KOD1. By expression of the gene and activity measurements of the protein product, we have demonstrated that the gene actually encodes an active Rubisco enzyme. Although putative Rubisco genes from the archaeaM. jannaschii and A. fulgidus have also been recently reported through genome analysis (8Bult C.J. White O. Olsen G.J. Zhou L. Fleischmann R.D. Sutton G.G. Blake J.A. FitzGerald L.M. Clayton R.A. Gocayne J.D. Kerlavage A.R. Dougherty B.A. Tomb J.F. Adams M.D. Reich C.I. Overbeek R. Kirkness E.F. Weinstock K.G. Merrick J.M. Glodek A. Scott J.L. Geoghagen N.S.M. Weidman J.F. Fuhrmann J.L. Nguyen D.T. Utterback T.R. Kelley J.M. Peterson J.D. Sadow P.W. Hanna M.C. Cotton M.D. Roberts K.M. Hurst M.A. Kaine B.P. Borodovsky M. Klenk H.P. Fraser C.M. Smith H.O. Woese C.R. Venter J.C. Science. 1996; 273: 1058-1073Crossref PubMed Scopus (2284) Google Scholar, 9Klenk H.P. Clayton R.A. Tomb J.F. White O. Nelson K.E. Ketchum K.A. Dodson R.J. Gwinn M. Hickey E.K. Peterson J.D. Richardson D.L. Kerlavage A.R. Graham D.E. Kyrpides N.C. Fleischmann R.D. Quackenbush J. Lee N.H. Sutton G.G. Gill S. Kirkness E.F. Dougherty B.A. McKenney K. Adams M.D. Loftus B. Peterson S. Reich C.I. McNeil L.K. Badger J.H. Glodek A. Zhou L. Overbeek R. Gocayne J.D. Weidman J.F. McDonald L. Utterback T. Cotton M.D. Spriggs T. Artiach P. Kaine B.P. Sykes S.M. Sadow P.W. D'Andrea K.P. Bowman C. Fujii C. Garland S.A. Mason T.M. Olsen G.J. Fraser C.M. Smith H.O. Woese C.R. Venter J.C. Nature. 1997; 390: 364-370Crossref PubMed Scopus (1201) Google Scholar), identification of the translation products has not been reported as of yet. Therefore, the question remains whether they actually encode active Rubiscos. Pk-Rubisco proved to be an extremely thermostable enzyme, and its specific activity at 90° C was the highest among all previously characterized Rubiscos. This may primarily be due to the high temperature where activity tests were carried out. Kinetic studies indicated that Pk-Rubisco had a τ value of 310, highly specific toward the carboxylase reaction. Type I Rubiscos from plants have a τ value of approximately 90, while those fromChromatium or Synechococcus share a value of approximately 45 (22Uemura K. Suzuki Y. Shikanai T. Wadano A. Jensen R.G. Chmara W. Yokota A. Plant Cell Physiol. 1996; 37: 325-331Crossref Scopus (50) Google Scholar). Recently, a Rubisco from thermophilic red algae was reported to possess a high τ value of 238 at 25° C, which decreased when temperature was elevated (21Uemura K. Anwaruzzaman Miyachi S. Yokota A. Biochem. Biophys. Res. Commun. 1997; 233: 568-571Crossref PubMed Scopus (135) Google Scholar). In contrast, the τ value of Pk-Rubisco increased as temperatures were elevated (Table I). Future studies on the kinetics and three-dimensional structure of the protein are necessary to understand the structural characteristics of Pk-Rubisco which lead to the extremely high activity and τ value. This should also identify the presence of ion pair networks involved in the high thermostability ofPk-Rubisco. We have not been able to identify an ORF with similarity to the small subunit of Rubisco. Results of Northern blot analysis indicated thatPk-rbc1 was not a member of a multi-gene operon and the most nearby ORFs encoded histone and enzymes involved in tryptophan biosynthesis (data not shown). At present, considering thatPk-Rubisco displayed high carboxylase activity as a homo-octamer, we expect Pk-Rubisco be composed of only large subunits. In support, ORFs showing similarity with the Rubisco small subunit could not be found in the whole genome of M. jannaschii and A. fulgidus. Western blot analysis has proven that Pk-Rubisco is present in P. kodakaraensis KOD1. Therefore, identification of the physiological role of Pk-Rubisco, especially its contribution to carbon dioxide fixation, will be an attractive target in future research. P. kodakaraensis KOD1 was isolated from a solfatara deep under sea level, and at present we have no evidence that the strain can grow chemolithotrophically. We found high structural similarity of Pk-Rubisco with the enzyme fromA. fulgidus, and A. fulgidus can grow lithoautotrophically on hydrogen, thiosulfate, and carbon dioxide. Genome analysis of A. fulgidus has revealed the presence of putative ribose-5-phosphate isomerase, phosphoglycerate kinase, glyceraldehyde-3-phosphate dehydrogenase, and triose-phosphate isomerase genes of the Calvin cycle (9Klenk H.P. Clayton R.A. Tomb J.F. White O. Nelson K.E. Ketchum K.A. Dodson R.J. Gwinn M. Hickey E.K. Peterson J.D. Richardson D.L. Kerlavage A.R. Graham D.E. Kyrpides N.C. Fleischmann R.D. Quackenbush J. Lee N.H. Sutton G.G. Gill S. Kirkness E.F. Dougherty B.A. McKenney K. Adams M.D. Loftus B. Peterson S. Reich C.I. McNeil L.K. Badger J.H. Glodek A. Zhou L. Overbeek R. Gocayne J.D. Weidman J.F. McDonald L. Utterback T. Cotton M.D. Spriggs T. Artiach P. Kaine B.P. Sykes S.M. Sadow P.W. D'Andrea K.P. Bowman C. Fujii C. Garland S.A. Mason T.M. Olsen G.J. Fraser C.M. Smith H.O. Woese C.R. Venter J.C. Nature. 1997; 390: 364-370Crossref PubMed Scopus (1201) Google Scholar). The corresponding genes have also been found in M. jannaschii (8Bult C.J. White O. Olsen G.J. Zhou L. Fleischmann R.D. Sutton G.G. Blake J.A. FitzGerald L.M. Clayton R.A. Gocayne J.D. Kerlavage A.R. Dougherty B.A. Tomb J.F. Adams M.D. Reich C.I. Overbeek R. Kirkness E.F. Weinstock K.G. Merrick J.M. Glodek A. Scott J.L. Geoghagen N.S.M. Weidman J.F. Fuhrmann J.L. Nguyen D.T. Utterback T.R. Kelley J.M. Peterson J.D. Sadow P.W. Hanna M.C. Cotton M.D. Roberts K.M. Hurst M.A. Kaine B.P. Borodovsky M. Klenk H.P. Fraser C.M. Smith H.O. Woese C.R. Venter J.C. Science. 1996; 273: 1058-1073Crossref PubMed Scopus (2284) Google Scholar). However, a putative ribulose-5-phosphate kinase, whose product should be the substrate of Rubisco, has not been identified in either strain. Further metabolic studies of P. kodakaraensis KOD1 are necessary to identify the pathway to which Pk-Rubisco mainly contributes. The primary structure of Pk-Rubisco was clearly distinct from those of type I and type II enzymes. The τ value was also significantly higher than those of type I and type II Rubiscos. As we have proven that the protein actually harbors Rubisco activity, we regard Pk-Rubisco as a structurally novel Rubisco. The putative Rubisco RBCL-2 of the archaeon A. fulgidus was the most closely related enzyme, showing 81.9% similarity. Furthermore, phylogenetic analysis of various Rubiscos indicated that the archaeal enzymes, although widely diverse, followed a distinct pathway of evolution (data not shown). These results suggest that archaeal Rubiscos, if others are expressed and show activity as in the case ofPk-Rubisco, may constitute a structurally distinct type of Rubisco.