Roles of Two ATPase-Motif-containing Domains in Cyanobacterial Circadian Clock Protein KaiC
2004; Elsevier BV; Volume: 279; Issue: 50 Linguagem: Inglês
10.1074/jbc.m406604200
ISSN1083-351X
AutoresFumio Hayashi, N. Itoh, Tatsuya Uzumaki, Ryo Iwase, Yuka Tsuchiya, Hisanori Yamakawa, Megumi Morishita, Kiyoshi Onai, Shigeru Itoh, Masahiro Ishiura,
Tópico(s)Circadian rhythm and melatonin
ResumoCyanobacterial clock protein KaiC has a hexagonal, pot-shaped structure composed of six identical dumbbell-shaped subunits. Each subunit has duplicated domains, and each domain has a set of ATPase motifs. The two spherical regions of the dumbbell are likely to correspond to two domains. We examined the role of the two sets of ATPase motifs by analyzing the in vitro activity of ATPγS binding, AMPPNP-induced hexamerization, thermostability, and phosphorylation of KaiC and by in vivo rhythm assays both in wild type KaiC (KaiCWT) and KaiCs carrying mutations in either Walker motif A or deduced catalytic Glu residues. We demonstrated that 1) the KaiC subunit had two types of ATP-binding sites, a high affinity site in N-terminal ATPase motifs and a low affinity site in C-terminal ATPase motifs, 2) the N-terminal motifs were responsible for hexamerization, and 3) the C-terminal motifs were responsible for both stabilization and phosphorylation of the KaiC hexamer. We proposed the following reaction mechanism. ATP preferentially binds to the N-terminal high affinity site, inducing the hexamerization of KaiC. Additional ATP then binds to the C-terminal low affinity site, stabilizing and phosphorylating the hexamer. We discussed the effect of these KaiC mutations on circadian bioluminescence rhythm in cells of cyanobacteria. Cyanobacterial clock protein KaiC has a hexagonal, pot-shaped structure composed of six identical dumbbell-shaped subunits. Each subunit has duplicated domains, and each domain has a set of ATPase motifs. The two spherical regions of the dumbbell are likely to correspond to two domains. We examined the role of the two sets of ATPase motifs by analyzing the in vitro activity of ATPγS binding, AMPPNP-induced hexamerization, thermostability, and phosphorylation of KaiC and by in vivo rhythm assays both in wild type KaiC (KaiCWT) and KaiCs carrying mutations in either Walker motif A or deduced catalytic Glu residues. We demonstrated that 1) the KaiC subunit had two types of ATP-binding sites, a high affinity site in N-terminal ATPase motifs and a low affinity site in C-terminal ATPase motifs, 2) the N-terminal motifs were responsible for hexamerization, and 3) the C-terminal motifs were responsible for both stabilization and phosphorylation of the KaiC hexamer. We proposed the following reaction mechanism. ATP preferentially binds to the N-terminal high affinity site, inducing the hexamerization of KaiC. Additional ATP then binds to the C-terminal low affinity site, stabilizing and phosphorylating the hexamer. We discussed the effect of these KaiC mutations on circadian bioluminescence rhythm in cells of cyanobacteria. Circadian rhythms, 24-h biological oscillations of metabolic and behavioral activities observed ubiquitously in prokaryotes and eukaryotes, are endogenously regulated by circadian clocks. Several clock and clock-related genes from Drosophila, Neurospora, Arabidopsis, mice, and cyanobacteria have been cloned and analyzed (1Dunlap J.C. Cell. 1999; 96: 271-290Abstract Full Text Full Text PDF PubMed Scopus (2391) Google Scholar). Cyanobacteria are the simplest organisms that exhibit circadian rhythms. We previously cloned and analyzed the kaiABC circadian clock gene cluster in the cyanobacterium Synechococcus sp. strain PCC 7942 (hereafter called Synechococcus) that is essential for the generation of circadian rhythms in cyanobacteria. The gene cluster consists of two operons: kaiA, whose product (probably indirectly) enhances kaiBC promoter activity, and kaiBC, which is repressed by KaiC (probably indirectly) protein itself (2Ishiura M. Kutsuna S. Aoki S. Iwasaki H. Andersson C.R. Tanabe A. Golden S.S. Johnson C.H. Kondo T. Science. 1998; 281: 1519-1523Crossref PubMed Scopus (594) Google Scholar). KaiA, KaiB, and KaiC interact in any combination (3Iwasaki H. Taniguchi Y. Ishiura M. Kondo T. EMBO J. 1999; 18: 1137-1145Crossref PubMed Scopus (146) Google Scholar), and KaiC also interacts with SasA, a sensory histidine kinase that enhances kaiBC promoter activity (4Iwasaki H. Williams S.B. Kitayama Y. Ishiura M. Golden S.S. Kondo T. Cell. 2000; 101: 223-233Abstract Full Text Full Text PDF PubMed Scopus (223) Google Scholar). To elucidate the mechanism of Kai protein-based generation of circadian oscillations in cyanobacteria, we analyzed the proteins derived from the thermophilic cyanobacterium Thermosynechococcus elongatus BP-1 by biochemical (5Hayashi F. Itoh H. Fujita M. Iwase R. Uzumaki T. Ishiura M. Biochem. Biophys. Res. Commun. 2004; 316: 195-202Crossref PubMed Scopus (51) Google Scholar, 6Hayashi F. Suzuki H. Iwase R. Uzumaki T. Miyake A. Shen J.R. Imada K. Furukawa Y. Yonekura K. Namba K. Ishiura M. Genes to Cells. 2003; 8: 287-296Crossref PubMed Scopus (111) Google Scholar, 7Uzumaki T. Fujita M. Nakatsu T. Hayashi F. Shibata H. Itoh N. Kato H. Ishiura M. Nat. Struct. Mol. Biol. 2004; 11: 623-631Crossref PubMed Scopus (69) Google Scholar), structural (6Hayashi F. Suzuki H. Iwase R. Uzumaki T. Miyake A. Shen J.R. Imada K. Furukawa Y. Yonekura K. Namba K. Ishiura M. Genes to Cells. 2003; 8: 287-296Crossref PubMed Scopus (111) Google Scholar, 7Uzumaki T. Fujita M. Nakatsu T. Hayashi F. Shibata H. Itoh N. Kato H. Ishiura M. Nat. Struct. Mol. Biol. 2004; 11: 623-631Crossref PubMed Scopus (69) Google Scholar), and biophysical techniques (7Uzumaki T. Fujita M. Nakatsu T. Hayashi F. Shibata H. Itoh N. Kato H. Ishiura M. Nat. Struct. Mol. Biol. 2004; 11: 623-631Crossref PubMed Scopus (69) Google Scholar). T. elongatus cells grow at over 50 °C, and their stable heat-tolerant proteins are suitable for analyses. KaiC has a duplicative structure (2Ishiura M. Kutsuna S. Aoki S. Iwasaki H. Andersson C.R. Tanabe A. Golden S.S. Johnson C.H. Kondo T. Science. 1998; 281: 1519-1523Crossref PubMed Scopus (594) Google Scholar, 6Hayashi F. Suzuki H. Iwase R. Uzumaki T. Miyake A. Shen J.R. Imada K. Furukawa Y. Yonekura K. Namba K. Ishiura M. Genes to Cells. 2003; 8: 287-296Crossref PubMed Scopus (111) Google Scholar), and each half has a set of ATPase motifs (Walker motifs A and B and a pair of deduced catalytic carboxylate Glu residues (named CatEs)) (2Ishiura M. Kutsuna S. Aoki S. Iwasaki H. Andersson C.R. Tanabe A. Golden S.S. Johnson C.H. Kondo T. Science. 1998; 281: 1519-1523Crossref PubMed Scopus (594) Google Scholar, 6Hayashi F. Suzuki H. Iwase R. Uzumaki T. Miyake A. Shen J.R. Imada K. Furukawa Y. Yonekura K. Namba K. Ishiura M. Genes to Cells. 2003; 8: 287-296Crossref PubMed Scopus (111) Google Scholar). We determined the three-dimensional structure of KaiC by single particle analysis of cryoelectron microscopic images and demonstrated that the hexamer has a hexagonal, pot-shaped structure composed of six identical dumbbell-shaped subunits (6Hayashi F. Suzuki H. Iwase R. Uzumaki T. Miyake A. Shen J.R. Imada K. Furukawa Y. Yonekura K. Namba K. Ishiura M. Genes to Cells. 2003; 8: 287-296Crossref PubMed Scopus (111) Google Scholar). The two spherical regions of the dumbbell-shaped structure probably correspond to the two domains of KaiC (6Hayashi F. Suzuki H. Iwase R. Uzumaki T. Miyake A. Shen J.R. Imada K. Furukawa Y. Yonekura K. Namba K. Ishiura M. Genes to Cells. 2003; 8: 287-296Crossref PubMed Scopus (111) Google Scholar). The hexamerization of KaiC depends on ATP, probably on its binding to each KaiC subunit (6Hayashi F. Suzuki H. Iwase R. Uzumaki T. Miyake A. Shen J.R. Imada K. Furukawa Y. Yonekura K. Namba K. Ishiura M. Genes to Cells. 2003; 8: 287-296Crossref PubMed Scopus (111) Google Scholar). The hexamerization does not require ATP hydrolysis, because AMPPNP, 1The abbreviations used are: AMPPNP, 5′-adenylylimidodiphosphate; ATPγS, adenosine 5′-O-(thiotriphosphate); GST, glutathione S-transferase. an unhydrolyzable analog of ATP, as well as ATP, can induce the hexamerization (6Hayashi F. Suzuki H. Iwase R. Uzumaki T. Miyake A. Shen J.R. Imada K. Furukawa Y. Yonekura K. Namba K. Ishiura M. Genes to Cells. 2003; 8: 287-296Crossref PubMed Scopus (111) Google Scholar). ADP, AMP, and inorganic phosphate do not induce hexamerization (6Hayashi F. Suzuki H. Iwase R. Uzumaki T. Miyake A. Shen J.R. Imada K. Furukawa Y. Yonekura K. Namba K. Ishiura M. Genes to Cells. 2003; 8: 287-296Crossref PubMed Scopus (111) Google Scholar). Thus, ATP is a necessary cofactor for KaiC hexamerization. In the presence of ATP, KaiC is phosphorylated by KaiC itself, and the phosphorylation is enhanced by KaiA (7Uzumaki T. Fujita M. Nakatsu T. Hayashi F. Shibata H. Itoh N. Kato H. Ishiura M. Nat. Struct. Mol. Biol. 2004; 11: 623-631Crossref PubMed Scopus (69) Google Scholar, 8Iwasaki H. Nishiwaki T. Kitayama Y. Nakajima M. Kondo T. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 15788-15793Crossref PubMed Scopus (245) Google Scholar, 9Williams S.B. Vakonakis I. Golden S.S. LiWang A.C. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 15357-15362Crossref PubMed Scopus (175) Google Scholar, 10Xu Y. Mori T. Johnson C.H. EMBO J. 2003; 22: 2117-2126Crossref PubMed Scopus (201) Google Scholar). We demonstrated that two molecules of KaiA dimer can interact with one molecule of KaiC hexamer and that one molecule of KaiA dimer interacting with one molecule of KaiC hexamer is enough to enhance KaiC phosphorylation (5Hayashi F. Itoh H. Fujita M. Iwase R. Uzumaki T. Ishiura M. Biochem. Biophys. Res. Commun. 2004; 316: 195-202Crossref PubMed Scopus (51) Google Scholar). By x-ray crystal structure analysis, in vitro biochemical analysis, and in vivo rhythm assays, we demonstrated that the His270 residue in the C-terminal clock-oscillator domain of KaiA is essential to clock oscillation in vivo and important for both the binding of KaiA to KaiC and the enhancement of KaiC phosphorylation by KaiA (7Uzumaki T. Fujita M. Nakatsu T. Hayashi F. Shibata H. Itoh N. Kato H. Ishiura M. Nat. Struct. Mol. Biol. 2004; 11: 623-631Crossref PubMed Scopus (69) Google Scholar). Although ATP, not ADP, is required for the hexamerization of KaiC, ATP is hydrolyzed into ADP and Pi in the KaiC phosphorylation reaction. How does KaiC enable the contradictory reactions mediated by ATP? Two different domains in the KaiC subunit (one at the N-terminal and one at the C-terminal) may be involved. To clarify the roles of these ATPase motifs, we prepared four mutant KaiCs, each carrying a mutation on Walker motif A or a pair of CatEs in each domain, and examined their ATP binding, hexamerization, stabilization, and phosphorylation in vitro. We prepared cyanobacterial cells expressing the mutant KaiCs and examined the effects of the mutations on circadian oscillations in vivo. Plasmid Construction—The plasmids that express wild type KaiC derived from T. elongatus (hereafter called KaiCWT) and the KaiCK53H and KaiCK294H mutants were described previously (6Hayashi F. Suzuki H. Iwase R. Uzumaki T. Miyake A. Shen J.R. Imada K. Furukawa Y. Yonekura K. Namba K. Ishiura M. Genes to Cells. 2003; 8: 287-296Crossref PubMed Scopus (111) Google Scholar). We mutated the Glu78 and Glu79 residues in the N-terminal domain of KaiC and the Glu318 and Glu319 residues in its C-terminal domain to Gln by PCR-mediated site-directed mutagenesis of the T. elongatus kaiC gene (the resulting KaiCE78Q&E79Q protein was named KaiCCatE1- and the resulting KaiCE318Q&E319Q protein was named KaiCCatE2-). We conducted two PCRs using T. elongatus genomic DNA as a template and two sets of primers: 1 and 3, and 2 and 4 Supplemental Data). We conducted the third PCR using primers 1 and 2 and the two PCR products as a template, and we cloned the PCR product into pGEX-6P-1 (Amersham Biosciences). We named the resulting plasmid pTeCatE1-. We similarly constructed plasmid pTeCatE2-, using four primers: 1, 2, 5, and 6 (Supplemental Data). We introduced the plasmids into Escherichia coli DH 5α and BL21 and grew the cells in Luria-Bertani broth (LB) and Terrific Broth and on LB plates containing 1.5% agar in LB. Production of KaiCs in E. coli and Their Purification—Production of KaiCs (KaiCWT, KaiCK53H, KaiCK294H, KaiCCatE1-, and KaiCCatE2-) in E. coli BL21, their purification, protein determination, and SDS-PAGE were carried out as described previously (5Hayashi F. Itoh H. Fujita M. Iwase R. Uzumaki T. Ishiura M. Biochem. Biophys. Res. Commun. 2004; 316: 195-202Crossref PubMed Scopus (51) Google Scholar). Briefly, KaiCs were overproduced in E. coli cells as soluble GST fusion proteins and trapped with glutathione-Sepharose-4B. The KaiCs were excised from the GST fusion proteins by PreScission Protease (Amersham Biosciences) and purified by ion exchange chromatography on a Mono Q HR5/5 column (Amersham Biosciences) followed by gel filtration chromatography on a Superdex200 HR10/30 column (Amersham Biosciences). Each protein eluted as a single symmetric peak at the position expected for a globular protein of about 60 kDa (data not shown). ATPγS Filter Binding Assay and Hill Plot Analysis—The amount of ATPγS (Amersham Biosciences) bound to KaiCs was measured by adsorption to nitrocellulose membrane filters (S & S, Arthur H. Thomas, Philadelphia, PA; 0.45-μm pore size, 25-mm diameter) as described previously (6Hayashi F. Suzuki H. Iwase R. Uzumaki T. Miyake A. Shen J.R. Imada K. Furukawa Y. Yonekura K. Namba K. Ishiura M. Genes to Cells. 2003; 8: 287-296Crossref PubMed Scopus (111) Google Scholar). KD values for ATPγS were calculated by Scatchard plot analysis. Briefly, a reaction mixture (50 μl) containing 20 mm Tris-HCl (pH 7.5), 1 mm dithiothreitol, 10 mm MgCl, [35S]ATPγS, and 12.5 pmol of KaiCs was incubated at 25 °C for 30 min. Then, 40 μl of the mixture was filtered through a nitrocellulose membrane presoaked in a washing buffer containing 1 mm dithiothreitol and 10 mm MgCl2 in 20 mm Tris-HCl (pH 7.5). The filter was washed with 6 ml of washing buffer, dried, and then soaked in a toluene-based scintillation mixture. The radioactivity of [35S]ATPγS retained on the filter was measured with a liquid scintillation counter. Assay for the Hexamerization of KaiCs—KaiC hexamerization, Native-PAGE, and determination of the Hill constants and KH value (the concentration giving half-saturation of the amount of KaiC hexamer induced) for AMPPNP were carried out as described previously (6Hayashi F. Suzuki H. Iwase R. Uzumaki T. Miyake A. Shen J.R. Imada K. Furukawa Y. Yonekura K. Namba K. Ishiura M. Genes to Cells. 2003; 8: 287-296Crossref PubMed Scopus (111) Google Scholar). Briefly, KaiCs were incubated with various concentrations of AMP-PNP at 25 °C for 30 min and immediately put on ice, and 4-μl aliquots of the mixtures were electrophoresed on PhastGel Gradient 8-25% polyacrylamide gels, which were then stained with PhastGel Blue R. We determined the amounts of KaiC hexamer by densitometry and analyzed the Native-PAGE data by Hill plot analysis. Assay for the Thermostability of Hexameric KaiCs by Native-PAGE—To induce the hexamerization of KaiC, we incubated KaiC monomer (11 μm) with 0.5 or 20 mm AMPPNP in 20 mm Tris-HCl buffer (pH 7.5) containing 5 mm MgCl2 and 150 mm NaCl at 4 °C overnight, and confirmed the hexamerization by Native-PAGE as described previously (6Hayashi F. Suzuki H. Iwase R. Uzumaki T. Miyake A. Shen J.R. Imada K. Furukawa Y. Yonekura K. Namba K. Ishiura M. Genes to Cells. 2003; 8: 287-296Crossref PubMed Scopus (111) Google Scholar). We determined the thermostability of the KaiC monomers and hexamers as follows: 10-μl aliquots of a reaction mixture containing KaiC monomer or hexamer were transferred to 1.5-ml plastic microcentrifuge tubes, overlaid with 30 μl of mineral oil, and incubated at various temperatures for 30 min. We then subjected the aliquots to Native-PAGE on PhastGel Gradient 8-25% gels and stained the gels with 0.1% PhastGel Blue R (Amersham Biosciences) solution in 30% methanol and 10% acetic acid using a PhastSystem (Amersham Biosciences). We used a densitograph system (AE-6920V-FX, ATTO) to estimate the densities of the bands that corresponded to native KaiC monomers and hexamers. We defined the relative amount of KaiC hexamers formed after incubation at 4 °C overnight as 100%. Assay for the Phosphorylation of KaiCs in the Presence or Absence of KaiA—We assayed KaiC phosphorylation as described previously (5Hayashi F. Itoh H. Fujita M. Iwase R. Uzumaki T. Ishiura M. Biochem. Biophys. Res. Commun. 2004; 316: 195-202Crossref PubMed Scopus (51) Google Scholar). Briefly, KaiC monomer (7.3 pmol in hexamer) was incubated with 1 mm ATP in 30 μl of 20 mm Tris-HCl buffer (pH 7.5) containing 5 mm MgCl2 at 50 °C for various times in the presence or absence of 43.8 pmol of KaiA (in dimer). We electrophoresed 15 μl of the reaction mixtures on SDS-polyacrylamide gels and stained the gels with Simply Blue™ SafeStain (Invitrogen). We estimated the densities of the bands corresponding to phosphorylated KaiC (p-KaiC) and non-phosphorylated KaiC by densitometry, as described above. Assay for the Phosphorylation of KaiCs Using [γ-32P]ATP—We assayed KaiC phosphorylation using [γ-32P]ATP as described previously (5Hayashi F. Itoh H. Fujita M. Iwase R. Uzumaki T. Ishiura M. Biochem. Biophys. Res. Commun. 2004; 316: 195-202Crossref PubMed Scopus (51) Google Scholar) with minor modifications. Briefly, 43.8 pmol of KaiC monomer was incubated with 1 mm [γ-32P]ATP (0.33 pCi/mm, Amersham Biosciences) in 30 μl of 20 mm Tris-HCl buffer (pH 7.5) containing 5 mm MgCl2 and 43.8 pmol of KaiA (in dimer) at 50 °C for 2 h. We electrophoresed 15 μl of the reaction mixture on 7.5% SDS-polyacrylamide gels. We used a BAS2000 bioimaging analyzer (Fuji) to visualize the radioactivity incorporated into KaiC bands. In Vitro Mutagenesis and in Vivo Rhythm Assay—We mutated the Synechococcus kaiC gene in a kaiBC operon segment by in vitro mutagenesis, transferred the segment into a specific targeting site (TS2) in the genome of a PkaiBC::luxAB bioluminescence reporter strain of Synechococcus with a deletion of the kaiBC operon, and we monitored the bioluminescence rhythms of the strain as described previously (7Uzumaki T. Fujita M. Nakatsu T. Hayashi F. Shibata H. Itoh N. Kato H. Ishiura M. Nat. Struct. Mol. Biol. 2004; 11: 623-631Crossref PubMed Scopus (69) Google Scholar, 11Kutsuna S. Kondo T. Aoki S. Ishiura M. J. Bacteriol. 1998; 180: 2167-2174Crossref PubMed Google Scholar). We constructed a plasmid disrupting the kaiBC operon, pDkaiBC, by deleting a 1832-bp AflII-EcoRI fragment carrying both the whole coding region of the kaiB gene and a 1405-bp segment carrying the 5′-coding region of the kaiC gene in pCkaiABC (2Ishiura M. Kutsuna S. Aoki S. Iwasaki H. Andersson C.R. Tanabe A. Golden S.S. Johnson C.H. Kondo T. Science. 1998; 281: 1519-1523Crossref PubMed Scopus (594) Google Scholar). We then transformed PkaiBC::luxAB cells with pDkaiBC to construct strain PkaiBC::luxAB with a deletion of the kaiBC operon in the genome as a host strain and selected with spectinomycin dihydrochloride (40 μg/ml) (strain PkaiBC::luxABΔkaiBC). We constructed targeting vector pTS2kaiBC carrying a kaiBC operon segment (nucleotides -547 to +1918; the first nucleotide A of the translation initiation codon of the kaiB gene is numbered +1) by inserting a 2464-bp NcoI-NheI fragment of pCkaiABC into the BamHI site of targeting vector pTS2K (11Kutsuna S. Kondo T. Aoki S. Ishiura M. J. Bacteriol. 1998; 180: 2167-2174Crossref PubMed Google Scholar). Lys52 and Lys294 in Synechococcus KaiC correspond to Lys53 and Lys294 in T. elongatus KaiC, respectively. Glu77,78 and Glu318,319 in Synechococcus KaiC correspond to Glu78,79 and Glu318,319 in T. elongatus KaiC, respectively. We mutated the Synechococcus kaiC gene in the kaiBC operon segment by PCR mediated site-directed mutagenesis using Synechococcus genomic DNA as template DNA. We mutated Lys52 of Synechococcus KaiC to a histidine residue by two PCRs using primer sets 7 and 8, and 9 and 10 (Supplemental Data). Then, we carried out a third PCR using the resulting two PCR products as templates and primer set 7 and 10. The PCR product was digested with HpaI and EcoRI, and a 1450-bp HpaI-EcoRI fragment was inserted into the HpaI-EcoRI site of pTS2kaiBC (named pTS2kaiBCK52H). We similarly mutated Lys294 of Synechococcus KaiC to a histidine residue using the two primer sets 7 and 11, and 12 and 10 (pTS2kaiBCK294H), and we mutated the Glu77 and Glu78 residues of Synechococcus KaiC to glutamine residues using the two primer sets 7 and 13, and 14 and 10, respectively (pTS2kaiBCCatE1-). We mutated both the Glu318 and Glu319 residues to glutamine residues using primer sets 7 and 15, and 16 and 10, respectively (pTS2kaiBCCatE2-). To construct a PkaiBC::luxAB/ΔkaiBC reporter strain expressing KaiCWT and four mutant KaiCs (KaiCK52H, KaiCK294H, KaiCCatE1-, and KaiCCatE2-), we transformed cells of strain PkaiBC::luxAB/ΔkaiBC with five kaiBC constructs (pTS2kaiBC, pTS2kaiBCK52H, pTS2kaiBCK294H, pTS2kaiBCCatE1-, and pTS2kaiBCCatE2-) to target the constructs into TS2 and selected the cells with kanamycin sulfate (30 μg/ml). Design of Mutants—KaiC is composed of an N-terminal domain (residues 1-268) and a C-terminal domain (residues 269-518) (6Hayashi F. Suzuki H. Iwase R. Uzumaki T. Miyake A. Shen J.R. Imada K. Furukawa Y. Yonekura K. Namba K. Ishiura M. Genes to Cells. 2003; 8: 287-296Crossref PubMed Scopus (111) Google Scholar). The two domains showed an identity of 23% and a similarity of 13%. Each domain contains a set of ATPase motifs (one Walker motif A and a pair of CatEs and one Walker motif B) (Fig. 1A and Ref. 6Hayashi F. Suzuki H. Iwase R. Uzumaki T. Miyake A. Shen J.R. Imada K. Furukawa Y. Yonekura K. Namba K. Ishiura M. Genes to Cells. 2003; 8: 287-296Crossref PubMed Scopus (111) Google Scholar). To clarify the roles of the two sets of ATPase motifs, we designed four T. elongatus KaiC mutants, each carrying a mutation on Walker motif A or the CatE pair in one domain (KaiCK53H, KaiCK294H, KaiCCatE1-, and KaiCCatE2-) (Fig. 1, B and C). Although we tried to prepare several other mutant KaiCs carrying different amino acid substitutions located on the two Walker motif As from those described above (G47TSGTG52-K53T54 → A47TSGTA52A53A54, G288ATGTG293K294T295 → A288ATGTGA293A294A295, K53M, K294M), we could not obtain proteins suitable for biochemical analysis because they became denatured during preparation. Binding of ATPγS to KaiCWT and Mutant KaiCs—To clarify the significance of residues Glu78,79, Glu318,319, Lys53, and Lys294 in the binding of ATP to KaiC, we determined the number of nucleotide-binding sites per KaiC subunit and KD values for ATPγS by the filter binding assay using [35S]ATPγS, and we applied Scatchard plot analysis to the results (Fig. 2 and Table I).Table IScatchard plot analysis of the binding of ATPγS to KaiCsKaiCWTKaiCCatE1-KaiCCatE2-KaiCK53HKaiCK294HNumber of ATPγS bound per KaiC subunit22222KD (μm)aValues represent mean ± S.D. of triplicate assays and were calculated using the Scatchard plot analysis shown in Fig. 234 ± 772 ± 1448 ± 22166 ± 4799 ± 47a Values represent mean ± S.D. of triplicate assays and were calculated using the Scatchard plot analysis shown in Fig. 2 Open table in a new tab We found that two ATP molecules bound to one molecule of the KaiCWT subunit, suggesting that both ATPase motifs on KaiC were functional nucleotide-binding sites. All the mutant KaiC subunits we examined also bound two molecules of ATPγS. The apparent KD values of mutant KaiCs carrying mutations on N-terminal ATPase motifs (KaiCK53H and KaiCCatE1-) were 1.5 to 1.7 times higher than those of the mutant KaiCs carrying mutation on C-terminal ATPase motifs (KaiCK294H and KaiCCatE2-), suggesting that the affinity for ATP is higher for the N-terminal motifs than for the C-terminal motifs. Thus, ATP binding to the N-terminal motifs is favored. The apparent KD values of KaiCK53H and KaiCK294H were 4.9 and 2.9 times higher than that of KaiCWT, respectively, whereas those of KaiCCatE1- and KaiCCatE2- were 2.1 and 1.4 times higher than that of KaiCWT, respectively. These results suggest that Lys53, Lys294, Glu78,79, and Glu318,319 are involved in the binding of ATP, with Lys53 and Lys294 being more significantly involved. Hexamerization of KaiCWT and Mutant KaiCs—We examined which N-terminal or C-terminal ATPase motifs are responsible for KaiC hexamerization. We determined the KH values of KaiCWT, KaiCCatE1-, KaiCCatE2-, KaiCK53H, and KaiCK294H by Native-PAGE (Fig. 3A) and the Hill plot analysis of the Native-PAGE data (Fig. 3B and Table II).Table IIHill plot analysis of the hexamerization of KaiCsKaiCWTKaiCCatE1-KaiCCatE2-KaiCK53HKaiCK294HHill constant1.8 ± 0.21.7 ± 0.30.9 ± 0.20.9 ± 0.30.9 ± 0.1KH (μm)aValues represent mean ± S.D. of triplicate assays and were calculated using the Hill plot analysis shown in Fig. 310 ± 0.486 ± 1341 ± 6.61,183 ± 120135 ± 11a Values represent mean ± S.D. of triplicate assays and were calculated using the Hill plot analysis shown in Fig. 3 Open table in a new tab Consistent with the order of increasing KD values (Table I), the apparent KH values for AMPPNP required for KaiC hexamerization increased in the order KaiCWT < KaiCCatE2- < KaiCCatE1- < KaiCK294H < KaiCK53H (Table II). The lower the dissociation constant for ATPγS were, the lower were the KH values for the AMPPNP required for the hexamerization. The apparent KH values of KaiCK53H and KaiCK294H for AMPPNP required for the hexamerization were 118 and 13 times higher than KH value of KaiCWT, suggesting that Lys53 in N-terminal Walker motif A is more critical for hexamerization than Lys294 in C-terminal Walker motif A. The apparent KH value of KaiCCatE1- was 2.3 times higher than the KaiCCatE2- value, suggesting that residues Glu78,79 (N-terminal CatEs) are more significantly involved in the hexamerization than residues Glu318,319 (C-terminal CatEs). Thus, such mutations on N-terminal Walker motif A and CatEs affected hexamerization more significantly than their C-terminal counterparts. The apparent KH values of KaiCK294H and KaiCCatE2- were 4 and 13 times higher than that of KaiCWT, respectively, suggesting that the latter Walker motif A and CatEs are also responsible for hexamerization. We concluded, therefore, that, although all residues examined (Glu78,79, Glu318,319, Lys53, and Lys294) contribute, it is Glu78,79 and Lys53 in N-terminal ATPase motifs that are mainly responsible for KaiC hexamerization. The Hill constant values of KaiCCatE1- (1.7) and KaiCWT (1.8) were similar, suggesting that residues Glu78,79 did not affect the cooperativity of the KaiC hexamerization induced by AMP-PNP binding. The Hill constant values of KaiCCatE2-, KaiCK53H, and KaiCK294H, on the other hand, were 0.9, indicating that residues Glu318,319, Lys53, and Lys294 are involved in the cofactor (AMPPNP) cooperativity in hexamerization. Thermostabilities of KaiCWT and Mutant KaiCs—To examine the role of KaiC residues Glu78,79, Glu318,319, Lys53, and Lys294 on the thermostability of KaiC hexamer, we quantitatively determined the thermostabilities of KaiCWT and the mutant KaiCs by Native-PAGE and estimated their Tm values from heat denaturation curves (Fig. 4). We prepared hexamers of KaiCWT and the mutant KaiCs by incubating the monomers with 0.5 or 20 mm AMPPNP at 4 °C overnight. We found that 20 mm AMPPNP was sufficient to induce complete hexamerization of all KaiCs and that 0.5 mm AMPPNP was enough to induce complete hexamerization of KaiCTWT, KaiCCatE1-, KaiCCatE2-, and KaiCK294H, but not KaiCK53H, because the apparent AMPPNP KH value for KaiCK53H was 1.2 mm (Table II). We assayed the thermostability of the above hexamers. The Tm values of KaiCWT were similar in the presence of 0.5 mm and 20 mm AMPPNP and were similar to the values of KaiCCatE1- (Table III). The Tm values of KaiCCatE2- obtained in the presence of 0.5 and 20 mm AMPPNP, in contrast, differed and were 12-30 °C lower than those of KaiCWT and KaiCCatE1-. These results indicate that Glu318,319 in the C-terminal CatEs but not Glu78,79 in the N-terminal CatEs are involved in KaiC hexamer thermostability. The heat denaturation of KaiCCatE2- probably resulted from the denaturation of the hexamers, not of the monomers that might have been dissociated from the hexamers by the release of AMPPNP, because of its lower KH values for AMPPNP (0.14 mm) and because KaiCCatE1- was as thermostable as KaiCWT despite its higher KH value (1.18 mm). The Tm values of KaiCK294H were as low as those of KaiCCatE2-, suggesting that residue Lys294 in C-terminal Walker motif A is also involved in the thermostability of the KaiC hexamer.Table IIITm values for the heat-denaturation of KaiCWT and mutant KaiCsAMPPNPTmaValues represent mean ± S.D. of triplicate assays and were calculated using the heat-denaturation curves shown in Fig. 4KaiCWTKaiCCatE1-KaiCCatE2-KaiCK53HKaiCK294Hmm°C0.563 ± 1.467 ± 1.137 ± 1.443 ± 0.337 ± 0.12068 ± 0.372 ± 0.456 ± 0.366 ± 0.441 ± 0.4a Values represent mean ± S.D. of triplicate assays and were calculated using the heat-denaturation curves shown in Fig. 4 Open table in a new tab The denaturation curve of KaiCK53H in the presence of 20 mm AMPPNP, a concentration much higher than its apparent KH value (1.18 mm), probably reflected the heat denaturation of the hexamer (Fig. 4), and the Tm value (66 °C) of KaiCK53H was almost the same as that of KaiCWT (68 °C). On the other hand, when we used the KaiCK53H incubated with 0.5 mm AMPPNP, the Tm value was 43 °C, which was much lower than the value obtained after incubation with 20 mm AMPPNP (66 °C). This reduction of Tm value in the presence of 0.5 mm AMPPNP was probably reflected by the depolymerization of its hexamer. Because the KaiCK53HKH value for AMPPNP was 1.18 mm (Table II), 0.5 mm AMPPNP was not sufficient to induce complete hexamerization of KaiCK53H. Because the Tm value of KaiC monomer was 32 °C (data not shown), monomers were denatured earlier than hexamers in the denaturation reaction. The concentration of native KaiCK53H monomer decreased irreversibly by heat denaturation, resulting in the depolymerization of KaiCK53H hexamer. Consequently, the population of KaiCK53H hexamers decreased (Fig. 4B). We concluded, therefore, that residue Lys53 in N-terminal Walker motif A is not significantly involved in the thermostability of the KaiC hexamer. The population of hexamers for KaiCK53H increased between 4 and 20 °C. The reason for this hexamerization is unclear at present. In conclusion, Glu318,319 and Lys294 in C-terminal ATPase motifs are responsible for the stabilization of KaiC hexamer. Phosphorylation of KaiCWT and Mutant KaiCs—KaiC prepared from GST-KaiC produced in E. coli cells shows a doublet band on SDS-polyacrylamide gels, and the upper band corresponds to p-KaiC formed by the phosphorylation reaction catalyzed by KaiC itself (7Uzumaki T. Fujita M. Nakatsu T. Hayashi F. Shibata H. Itoh N. Kato H. Ishiura M. Nat. Struct. Mol. Biol. 2004; 11: 623-631Crossref PubMed Scopus (69) Google Scholar). In contrast, when prepared and assayed similarly, KaiCK294H and KaiCCatE2- showed single bands that did not correspond to p-KaiC (data not shown). We examined the phosphorylation of KaiCWT and four mutant KaiCs in the presence or absence of KaiA, which enhances KaiC phosphorylation (Figs. 5, A and B). Following a 2-h incubation with 1 mm ATP, the relative amount of the band corresponding to p-KaiCWT increased from 34 to 43% in the absence of KaiA and 56% in its presence. The relative amount of the p-KaiCK53H, p-KaiCCatE1-, and p-KaiC WT band increased similarly, but we could not detect any p-KaiCK294H or p-KaiCCatE2- bands even in the presence of KaiA (Fig. 5, A and B). Nor could we detect incorporation of γ-32P from [γ-32P]ATP into KaiCK294H and KaiCCatE2-, whereas γ-32P was significantly incorporated into KaiCWT, KaiCK53H, and KaiCCatE1- (Fig. 5C), indicating that KaiCK294H and KaiCCatE2- lacked KaiC phosphorylation activity. Therefore, we concluded that Glu318,319 and Lys294 in C-terminal ATPase motifs are responsible for KaiC phosphorylation activity. In Vivo Bioluminescence Rhythms—We recently demonstrated that T. elongatus has a circadian clock and developed a circadian bioluminescence rhythm system in T. elongatus (12Onai K. Morishita M. Itoh S. Okamoto K. Ishiura M. J. Bacteriol. 2004; 186: 4972-4977Crossref PubMed Scopus (45) Google Scholar). We have not refined the system enough to use it for detailed in vivo rhythm analysis such as that described below, but because Synechococcus KaiC is not stable enough to be characterized in vitro, we used T. elongatus KaiC. We examined the functional importance of the two pairs of CatEs and each lysine residue on the two Walker motif As in KaiC by in vitro mutagenesis and in vivo rhythm assays using a PkaiBC::luxAB bioluminescence reporter strain of Synechococcus expressing wild type or mutant Synechococcus KaiCs (Fig. 6). Mutations in the N-terminal and C-terminal CatEs (Fig. 6, KaiCCatE1- and KaiCCatE2-) and those of the lysine residue in the N-terminal Walker motif A (Fig. 6, KaiCK52H) caused an almost complete disruption of circadian bioluminescence rhythms, indicating that these ATPase motifs are critical to clock oscillation. The substitution of histidine for Lys294 in C-terminal Walker motif A caused rhythm irregularities (Fig. 6, panel KaiCK294H): 1) the bioluminescence of the cells decreased transiently after their transfer to constant light conditions, 2) the appearance of the first bioluminescence peak was delayed (28.6 ± 2.4 h, n = 124) compared with the strain expressing Synechococcus KaiCWT (10.8 ± 0.8 h, n = 65), and 3) the period lengthened in each cycle (first to the second peak = 29.5 ± 6.8 h, n = 124; second to the third peak = 39.9 ± 15.5 h, n = 124; third to the fourth peak = 45.3 ± 14.6 h, n = 59). This indicates that Lys294 in the C-terminal Walker motif A is important to normal circadian oscillation. Substituting histidine for the lysine residue in either the N- or C-terminal Walker motif A increased the apparent KD values for ATPγS more significantly than did substituting glutamine for either of the CatEs (Fig. 2 and Table I). This observation is consistent with previous data showing that the conserved Lys residue on Walker motif A in the Walker-type ATPase (13Leipe D.D. Wolf Y.I. Koonin E.V. Aravind L. J. Mol. Biol. 2002; 317: 41-72Crossref PubMed Scopus (859) Google Scholar), such as RecA (14Story R.M. Weber I.T. Steitz T.A. Nature. 1992; 355: 318-325Crossref PubMed Scopus (682) Google Scholar) and F1-ATPase (15Abrahams J.P. Leslie A.G. Lutter R. Walker J.E. Nature. 1994; 370: 621-628Crossref PubMed Scopus (2764) Google Scholar), binds ATP by interacting with β- and γ-phosphate groups of ATP, whereas CatE binds ATP weakly by interacting with the γ-phosphate group of ATP via a water molecule. Recently, the crystal structure of Synechococcus sp. strain PCC 7942 KaiC hexamer with 12 molecules of ATPγS was determined (16Pattanayek R. Wang J. Mori T. Xu Y. Johnson C.H. Egli M. Mol. Cell. 2004; 15: 375-388Abstract Full Text Full Text PDF PubMed Scopus (153) Google Scholar). Consistent with our pot-shaped structure of T. elongatus KaiC hexamer determined by single particle analysis (6Hayashi F. Suzuki H. Iwase R. Uzumaki T. Miyake A. Shen J.R. Imada K. Furukawa Y. Yonekura K. Namba K. Ishiura M. Genes to Cells. 2003; 8: 287-296Crossref PubMed Scopus (111) Google Scholar), the structure was relatively wide and open at the N-terminal end and narrower for a part of the C-terminal domain. Here, we showed the different roles of the two domains: hexamerization in N-terminal domain and stabilization and phosphorylation in C-terminal domain (Fig. 7). In their structure, ATP binding modes are different between N-terminal and C-terminal domains, and different roles of the two domains were anticipated (16Pattanayek R. Wang J. Mori T. Xu Y. Johnson C.H. Egli M. Mol. Cell. 2004; 15: 375-388Abstract Full Text Full Text PDF PubMed Scopus (153) Google Scholar). Based on the different roles of two sets of ATPase motifs demonstrated here, we propose the following reaction mechanism (Figs. 7 and 8): 1) A KaiC subunit has two types of ATP-binding sites, a high affinity site in the N-terminal domain and a low affinity site in the C-terminal domain; 2) ATP preferentially binds to the N-terminal high affinity site, inducing KaiC to form an unstable hexamer; 3) ATP binds to the C-terminal low affinity site, inducing stabilization of the hexamer. Alternatively, the presence of the C-terminal ATPase motifs themselves induce stabilization, as suggested by the fact that the hexamers of KaiCK294H and KaiCCatE2-, which showed lower KD values (higher affinity for ATP) than KaiCK53H and KaiCCatE1- (Fig. 2 and Table I), were unstable; and 4) The phosphorylation reaction (the transfer of the γ-phosphate group from ATP bound to the C-terminal low affinity site to the recently identified phosphorylation sites (Ser431 and Thr432) 2F. Hayashi and M. Ishiura, unpublished data. in KaiC occurs in the C-terminal ATPase motifs. This mechanism would maintain the hexameric structure even if the ATP bound to the C-terminal domain of KaiC hexamer were hydrolyzed. The ATP-induced hexamerization of KaiC is crucial for the generation of clock oscillation, as evidenced by the almost complete disruption of circadian bioluminescence rhythms in vivo when the ATP binding site in the N-terminal Walker motif A was mutated (KaiCK53H in Fig. 3 and Table II). The stabilization and/or phosphorylation of KaiC hexamer may be involved in the regulation of the period length, phase, and amplitude of clock oscillations, but not in the generation of the oscillations, as evidenced by the unique and unstable circadian bioluminescence rhythms produced when the ATP-binding site in C-terminal Walker motif A was mutated (KaiCK294H in Figs. 4 and 5, and Table III). Although KaiCK294H and KaiCCatE2- showed similar characteristics in in vitro stabilization and hexamer phosphorylation (Table III and Figs. 4 and 5), the aberrant rhythms they expressed differed (Fig. 6), suggesting other unknown crucial biochemical functions. Although the contributions of the C-terminal CatEs to ATP binding and KaiC hexamerization were the lowest among the ATPase motifs we examined (Tables II and III and Figs. 2 and 3), mutations at those sites disrupted clock oscillation almost completely (Fig. 6), suggesting important but unidentified roles. We thank Dr. Miriam Bloom (SciWrite Biomedical Writing & Editing Services) for professional editing and Satoko Ogawa and Kumiko Tanaka for technical support. Download .pdf (.18 MB) Help with pdf files
Referência(s)