• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of embojLink to Publisher's site
EMBO J. Apr 21, 2004; 23(8): 1688–1698.
Published online Apr 8, 2004. doi:  10.1038/sj.emboj.7600190
PMCID: PMC394244

Anabaena circadian clock proteins KaiA and KaiB reveal a potential common binding site to their partner KaiC

Robert G Garces,1,3 Ning Wu,2,3 Wanda Gillon,2,3 and Emil F Pai1,2,3,4,5,a

Abstract

The cyanobacterial clock proteins KaiA and KaiB are proposed as regulators of the circadian rhythm in cyanobacteria. Mutations in both proteins have been reported to alter or abolish circadian rhythmicity. Here, we present molecular models of both KaiA and KaiB from the cyanobacteria Anabaena sp PCC7120 deduced by crystal structure analysis, and we discuss how clock-changing or abolishing mutations may cause their resulting circadian phenotype. The overall fold of the KaiA monomer is that of a four-helix bundle. KaiB, on the other hand, adopts an alpha–beta meander motif. Both proteins purify and crystallize as dimers. While the folds of the two proteins are clearly different, their size and some surface features of the physiologically relevant dimers are very similar. Notably, the functionally relevant residues Arg 69 of KaiA and Arg 23 of KaiB align well in space. The apparent structural similarities suggest that KaiA and KaiB may compete for a potential common binding site on KaiC.

Keywords: Anabaena sp PCC7120, circadian rhythm, kaiA, kaiB, kaiC, X-ray crystallographic structure

Introduction

Circadian rhythms, biological oscillations with a period of approximately 24 h, have been observed in a wide variety of organisms from cyanobacteria to green plants to humans (Dunlap et al, 1999; Mori and Johnson, 2001; Young and Kay, 2001). In order for biological oscillations to be denoted as a circadian ‘clock', they have to fulfil three criteria. They must: (1) persist in the absence of external time cues; (2) reset their phase by exposure to light or dark; and (3) compensate for temperature changes during the circadian period. It has been postulated that the circadian clock controls the rhythm that is fundamental for cells to adapt to the alternation between day and night (Pittendrigh, 1993). Physiological studies have identified several clock-related genes in cyanobacteria, Neurospora, Arabidopsis, Drosophila, and mammals (Dunlap, 1999; Young and Kay, 2001).

Cyanobacteria are the simplest organisms known to possess a circadian clock. They are well suited as genetic and physiological model systems for the study of circadian rhythms (Iwasaki and Kondo, 2000; Mori and Johnson, 2001). Various cyanobacterial strains, containing a bioluminescence reporter gene, have been used to monitor circadian rhythms and identify key genes involved in its regulation and maintenance (Kondo et al, 1993; Ishiura et al, 1998). Using genetic complementation of various rhythm mutants, a gene cluster kai composed of three genes (kaiA, kaiB, and kaiC) was identified and cloned from the cyanobacterium Synechococcus sp PCC7942 (Ishiura et al, 1998). Kai genes, their name derived from the Japanese word for cycle, are found ubiquitously among cyanobacteria (Lorne et al, 2000). Amino-acid sequence analyses have shown that the Kai proteins display no homology to clock proteins identified in other organisms (Dunlap et al, 1999). However, circadian rhythms can be altered or even abolished through mutations in any of the kai genes and, as a corollary, re-introduction of the kai gene cluster will restore lost rhythmicity (Ishiura et al, 1998). Evidence gathered so far indicates that all the three Kai proteins play a crucial role for circadian rhythm in cyanobacteria.

How exactly the kai genes generate and maintain rhythm in cyanobacteria is only slowly being unravelled. Transcriptional analysis of the kai genes has provided early clues on how this system functions (Ishiura et al, 1998). Two promoters control kai gene transcription: PkaiA for kaiA and PkaiBC for kaiB and kaiC. Both promoters regulate the rhythm of the same phase for their respective mRNAs. Overexpression of the kaiC gene will immediately and completely repress its own PkaiBC promoter. This strong negative autoregulation suggests that KaiC functions as the key component of the circadian oscillation in cyanobacteria. However, inactivation of PkaiA reduces PkaiBC activity, while overexpression of KaiA dramatically increases PkaiBC activity. As PkaiA is in phase with PkaiBC, it appears that KaiA regulates the expression of kaiBC in a positive manner.

Recently, the structure of the N-terminal domain of KaiA, termed KaiA 135N, from Synechococcus elongatus sp PCC 7942 has been determined using NMR techniques (Williams et al, 2002). KaiA 135N adopts the structure of a pseudo-receiver domain and is postulated to act as a regulatory component of KaiA activity. It should be noted that the sequence of kaiA is the one least conserved among the kai genes. KaiA protein from Anabaena sp PCC 7120, one of the objects of this study, is highly homologous to the C-terminal part of the S. elongatus KaiA, but completely lacks the region corresponding to KaiA 135N. The conserved portion of KaiA has been reported to stimulate KaiC phosphorylation (Iwasaki et al, 2002; Williams et al, 2002; Kitayama et al, 2003). Also KaiA mutations in S. elongatus appear to lower PkaiBC expression (Ishiura et al, 1998; Nishimura et al, 2002), suggesting a dual role for KaiA in regulating the length of the period of the circadian oscillation.

In contrast, KaiB, which shares its promoter with KaiC, has been identified as an attenuator of KaiC phosphorylation (Kitayama et al, 2003). Again setting it apart from KaiA, the primary sequence of KaiB is well conserved among cyanobacterial species. Both KaiB and KaiC protein levels cycle rhythmically even under constant light conditions (Xu et al, 2000). As KaiB and KaiC are co-transcribed, the phasing of their protein levels suggests that both proteins have similar turnover rates. While KaiB–KaiC interactions are well documented, the evidence for KaiA–KaiB interactions is much weaker (Iwasaki et al, 1999). KaiB point mutations show a shortened circadian period likely arising from its diminished ability to reduce KaiC phosphorylation (Ishiura et al, 1998).

Based on a genomic survey, KaiC from S. elongatus sp PCC7942 was proposed to be a member of the bacterial RecA/DnaB family (Leipe et al, 2000). Also, analysis of the amino-acid sequence of KaiC shows a significant similarity between the first and second halves of the sequence, with each half containing a trinucleotide-binding Walker A motif (Ishiura et al, 1998; Walker et al, 2000). Mutations of these Walker A motifs abolish nucleotide binding and rhythmicity is severely disrupted or abolished (Nishiwaki et al, 2000; Mori and Johnson, 2001). Characterized by biophysical methods and electron microscopy (EM), KaiC is reported to form hexameric rings in the presence of ATP (Mori et al, 2002). In line with its apparent negative feedback on PkaiBC, KaiC was able to bind onto forked DNA substrates despite lacking any known DNA-binding motifs (Mori et al, 2002). In addition, KaiC undergoes autophosphorylation in vitro (Nishiwaki et al, 2000), a feature that may play a role in its functional regulation and/or degradation, as is seen for eukaryotic clock proteins FRQ and PER (Young and Kay, 2001). Point mutations in KaiC can cause a wide variety of period changes including shortened and extended periods as well as complete loss of rhythmic cycling (Ishiura et al, 1998). While KaiC appears to be the time-keeping protein for the circadian rhythm, its partners, KaiA and KaiB, seem to be crucial for adjusting the periodicity of the level of KaiC protein, thereby compensating for length changes in day/night cycles.

In vitro, S. elongatus Kai proteins form oligomeric complexes in all possible combinations (Iwasaki et al, 1999). In vivo, homodimeric interactions between KaiB molecules have been demonstrated using a bioluminescence resonance energy transfer system (Xu et al, 1999). Pull-down assays demonstrated that each Kai protein was able to interact with each of the other two Kai proteins, although the interaction between KaiA and KaiB was quite weak (Iwasaki et al, 1999). Moreover, KaiC appears to have dual sites (one in each of the internal repeats) that have the ability to bind onto KaiA and KaiB proteins (Iwasaki et al, 1999). Further in vivo studies support the formation of higher order complexes composed of all three Kai proteins, which vary in size depending on subjective day and night times (Kageyama et al, 2003).

Although our knowledge of the regulation and binding interactions of the Kai proteins has been growing recently, very little is known about the molecular structures of the individual proteins. In an attempt to further the structural characterization of the Kai system, we now report the three-dimensional structures of both KaiA and KaiB from the cyanobacterium Anabaena sp PCC7120 and discuss their potential interactions and functions.

Results

Determination of the oligomerization state of KaiA and KaiB

Following purification, both recombinant KaiA and KaiB proteins were loaded in separate experiments onto a Superdex 200 column for size fractionation. KaiA eluted within a single peak of about ~27 kDa, which is close to the estimated size of the KaiA dimer. On reducing SDS–PAGE gels, KaiA ran as a monomer showing a band around 14 kDa in size, and no contaminating bands could be detected by Coomassie Blue staining. KaiB eluted from the column in fractions corresponding to a molecular weight of about ~30 kDa, which again is close to the estimated size of the KaiB dimer. On a reducing SDS–PAGE gel, peak fractions ran as monomers around 15 kDa in size, with no other bands visible. These findings are consistent with reports that KaiA and KaiB form homodimers in vivo (Kageyama et al, 2003).

Crystallization of native KaiA, native KaiB, and KaiB mutants

Native KaiA crystals were reproducibly grown in 100 mM MES (pH 6.0) and using 1.1 M MgSO4 as a precipitant. They were strongly birefringent and displayed apparent six-fold symmetry along their major axis. The crystals belonged to space group P6522 with cell axes a=b=57.3 Å, c=137.5 Å and one KaiA monomer per asymmetric unit. The crystal axes of the bromine soaked crystal changed slightly to a=b=57.1 Å, c=137.5 Å.

Native KaiB crystals could be reproducibly grown in 100 mM Tris (pH 7.5) using 18% (v/v) PEG 400 as a precipitant (Fluka). They belong to space group C2 with cell parameters of a=104.8 Å, b=36.4 Å, c=54.1 Å, and β=98.8°. As wild-type KaiB contains only the N-terminal methionine, a residue often found to be mobile in crystals, mutants that contained an additional methionine residue to increase the probability of success of the MAD phase determination method were generated. Both KaiB mutants (L28 M and L69 M) required the use of higher concentrations of PEG400 (23.5% v/v) (Fluka) as well as the addition of 5% 1,6-hexanediol in order to form crystals. KaiB (L69 M) mutants typically grew as long rod-like crystals that had a strong tendency to split or fracture and therefore were not suitable for data collection. KaiB (L28 M) Se-Met crystals belonged to space group C2 with cell parameters of a=105.0 Å, b=36.7 Å, c=54.2 Å, and β=99.1°.

Structure of KaiA

The crystal structure of KaiA was determined by the bromine-based MAD technique (Dauter and Dauter, 1999). As mentioned earlier, the KaiA protein of S. elongatus is nearly twice as large as the one from Anabaena sp PCC7120 caused by the addition of an N-terminal domain of 180 amino acids (Figure 2), whose structure has recently been reported (Williams et al, 2002). Our analysis deals with the full-length KaiA protein from the closely related species PCC7120, which is highly homologous to the C-terminal domain of the former. The KaiA monomer adopts a four-helix bundle fold with the classical up–down–up–down arrangement of helices and a well-defined hydrophobic core buried between these helices (Carlacci and Chou, 1990) (Figure 1). Within the core, one finds several leucine and isoleucine side chains as well as tyrosine and phenylalanine rings. Residues at the N-terminus (1–7), although partially stabilized for some monomers by crystal contacts, appear to be mostly flexible as the electron density in that region is much less defined. In solution, this region of KaiA is likely free to move, but could adopt specific conformations upon interactions with KaiB, KaiC, or other protein-binding partners.

Figure 1
Structure of KaiA. (A) Ribbon diagram of monomer. Helices are coloured in blue. (B) Ribbon diagram of dimer. The two-fold axis runs vertically. Monomer A is coloured in light blue (for helices) while monomer B is coloured in darker blue. N- and C-termini ...
Figure 2
Alignments of the sequences of KaiA from Anabaene sp PCC7120, Synechococcus sp PCC7942, and S. elongatus. Identical residues across sequences are shaded by dark grey, while conserved residues are shaded by light grey (criteria defined by Clustal X). Hyphens ...

The asymmetric unit in the KaiA crystal lattice contains one monomer with a total of three dimers making up the unit cell. The dimer interface lies along one of the crystal axes of symmetry, with the majority of the interactions occurring along helices 3 and 4 (Figure 2). An inter-subunit disulphide bridge between subunits is formed between symmetry-related Cys 93 residues. In general, the dimer interface consists of predominantly hydrophobic interactions with a few backbone hydrogen bonds surrounding the hydrophobic region. They include the main chain atoms of Val 50a and Glu 90b, Ala 95a and Leu 88b, as well as their symmetry counterparts.

Mutations of KaiA in PCC7942 result in an extended circadian period or completely abolish rhythmicity in cyanobacteria (Ishiura et al, 1998; Nishimura et al, 2002). For PCC7942, the R249 H mutant of KaiA has been described as displaying an extended circadian period of 30 h (Ishiura et al, 1998). In a multiple sequence alignment, this residue is lined up with Arg 69 of the PCC7120 KaiA protein (Figure 2). The crystal structure places Arg 69 in a loop region between helices 3 and 4. One would expect that this mutation should not cause any major change in the internal structure of KaiA. It might, however, affect the way in which the two domains of PCC7942 KaiA interact or it could well be involved in binding to other partner proteins.

Another destructive mutation reported in KaiA of PCC7942 is E274 K (Ishiura et al, 1998). In PCC7120, the corresponding residue is Glu 94 (Figure 2). As this residue links the two subunits via a hydrogen bond, the resulting change in charge would obviously drastically weaken KaiA's ability to form dimers, the oligomerization state observed in vivo (Kageyama et al, 2003). It is easy to imagine how this defect could affect the stability or at least interfere with protein–protein interactions, leading to the observed abolishment of rhythmicity.

Structure of KaiB

The molecular model of KaiB was derived, applying the Se-Met-based MAD technique to the L28 M mutant. The overall fold of the KaiB monomer is a two α-helix/three β-sheet alpha–beta meander (Figure 3), with helical and sheet secondary structure elements alternating along the peptide chain (Figure 4). There is a loop region in each monomer that accounts for the majority of interactions between the monomers. It is flanked by Pro 51 and Pro 63. The first few residues of β-strand 3, Ile 88, and Leu 90 are also involved in the dimer interface, which consists of predominantly hydrophobic interactions coupled with some hydrogen bonding along the edges of the dimer interface between the main chain residues of Leu 53a and Ile 88b. An especially important residue in stabilizing the KaiB dimer seems to be Gln 52. Its side chain is involved in three hydrogen bond interactions linking it to the backbone carbonyls of Ser 81 and Leu 87, as well as the guanadinium group of Arg 85.

Figure 3
Structure of KaiB. (A) Ribbon diagram of monomer. Helices are coloured in purple while β-sheets are coloured in yellow. (B) Ribbon diagram of dimer. The two-fold axis runs vertically. Monomer A is coloured in purple and yellow (for helices and ...
Figure 4
Alignments of the sequences of KaiB from Anabaene sp PCC7120, Synechococcus sp PCC7942, and S. elongatus. Identical residues across sequences are shaded by dark grey, while conserved residues are shaded by light grey (criteria defined by Clustal X). Hyphens ...

Leu 28, which was replaced by methionine in the KaiB (L28 M) mutant used for MAD phasing, is buried within the hydrophobic region between both α-helices and the β-sheet region. In contrast, Leu 69 is located adjacent to the second α-helix, its head group buried within a small hydrophobic pocket, surrounded by Leu 10, Ile 77, and Leu 90. Mutation of this residue into a methionine resulted in the growth of crystals of very poor quality. The mutation may cause α-helix 2 to shift due to the additional bulk of the methionine side chain. This, in turn, would affect the ability for dimers to pack tightly along the corresponding helices, thus introducing multiple defects into the growing crystal lattice.

Circadian period mutants of KaiB in PCC7942 have also been reported (Kondo et al, 1994). In one of them, Leu 11 is replaced with a phenylalanine, resulting in a phenotype with a shortened 21 h circadian cycle (Ishiura et al, 1998). This mutation maps onto Leu 12 of PCC7120 KaiB (Figure 4). The side chain of this amino acid resides directly in the centre of the large hydrophobic core of the KaiB monomer and is surrounded by residues from each of the secondary structure elements (Figure 4). The packing of the leucine and isoleucine side chains within this region is very well defined, suggesting that the packing is tight and well ordered. The space requirements of a flat aromatic ring are quite different from those of a branched side chain. Therefore, in a crowded environment, replacing leucine with a phenylalanine may lead to an overall rearrangement of the side chains, followed by a shift in the overall alignment of secondary structure elements. This, in turn, would certainly affect KaiB's ability to interact with its binding partners.

The second point mutation described for KaiB is R74W. It also results in a slightly shortened circadian rhythm of 22 h (Ishiura et al, 1998). Arg 74 of KaiB from PCC7942 maps onto Lys 75 of the PCC7120 protein (Figure 4). As the side chain of Lys 75 protrudes from the surface of the KaiB dimer, this point mutation should not cause any significant structural changes. Probably, this residue lies within a region involved in binding interactions to other proteins.

Co-precipitation of Kai protein complexes

Tests for physical association between KaiA or KaiB wild-type and mutant proteins with KaiC were performed in vitro using a binding assay analogous to those described in the literature (Iwasaki et al, 1999; Kageyama et al, 2003). We used Ni-NTA agarose (Invitrogen) beads as the specific protein precipitant to immobilize recombinant 6xHis-tagged KaiC, the only one of the Kai proteins modified in this way, and any associated protein complexes.

In the absence of 6xHis-tagged KaiC, neither wild-type KaiA nor KaiB proteins would bind to the Ni beads (Figure 7A, lanes 1, 6, 11, and 16). The tagged KaiC, however, bound readily (Figure 7A, lanes 2, 7, 12, and 17) and was eluted using high concentrations of imidazole. Wild-type KaiB protein could be pulled down by the Ni beads in the presence of KaiC, showing binding between KaiB and KaiC (Figure 7A). Increasing the molar ratio of KaiB:KaiC resulted in a saturation of KaiB bound to KaiC, with increasing amounts of unbound KaiB appearing in the flow through fractions (Figure 7A, lanes 13–15 and 18–20).

Figure 7
Interactions between KaiC and KaiA, KaiB, KaiA69A, and KaiB23A proteins. Ni beads were incubated with the various protein mixtures. Proteins that did not interact with the Ni beads were collected and loaded in the lanes labelled as Flow-through. Proteins ...

Unexpectedly, the presence of KaiA appeared to prevent KaiC from interacting with the Ni beads; consequently, KaiC appears in increasing amounts in the flow-through fractions as the KaiA:KaiC molar ratio is increased (Figure 7A, lanes 3–5 and 8–10). The ability of KaiA to cause KaiC's release from the beads explains why we were unable to see KaiA bound onto the resin even in the presence of KaiC. Loss of KaiC from the resin does indirectly confirm interaction between KaiA and KaiC proteins. Using a fixed concentration of KaiB while adding increasing quantities of KaiA showed a progressive loss of bound KaiC (Figure 7B, lanes 2–5 and 7–10), suggesting that KaiA interacts with KaiC, even in the presence of KaiB. Using increasing amounts of KaiB mixed with a fixed concentration of KaiA progressively rescued tagged KaiC's binding onto the beads taking along KaiB (Figure 7B, lanes 12–15 and 17–20). To provide evidence in a more direct manner, sequential competition-binding assays were also performed that showed that KaiA could be pulled down by pre-immobilized KaiC and could be competed off by subsequently adding KaiB and vice versa (see Supplementary data).

The R69A KaiA and R23A KaiB mutants were also tested for their ability to bind to KaiC and their behaviour was compared to that of the wild-type proteins (Figure 7C). Replacing Arg 69 of KaiA with an alanine produced a mutant protein that was no longer able to cause the release of KaiC from the Ni beads (Figure 7C, lanes 1–12), indicative of a loss of binding. Mutation of Arg 23 of KaiB to an alanine also had a significant effect, substantially reducing its affinity for KaiC (Figure 7C, lanes 13–24).

Discussion

Structural basis of KaiA and KaiB functions

In cyanobacteria, KaiC is a prime candidate for time-keeper of the circadian rhythm (Ishiura et al, 1998). KaiA is believed to be important in regulating KaiC turnover with the common and well-conserved domain of KaiA involved in stimulation of KaiC autophosphorylation (Iwasaki et al, 2002; Williams et al, 2002).

Size and amino-acid sequence of KaiB are also well conserved between cyanobacterial species. However, despite the identification of rhythmicity mutants of KaiB, relatively little is known about the actual mechanistic role played by KaiB in the maintenance of circadian rhythm. Recent data suggest that KaiB is an attenuator of KaiC phosphorylation (Kitayama et al, 2003). The crystal structures of KaiA and KaiB now shed some light on how KaiB mutants might affect its function, thus altering circadian rhythms in cyanobacteria.

Looking at the surfaces of the KaiA and KaiB dimers spanning their two-fold axis of symmetry, one discovers a striking similarity of shape and placement of some charged residues (Figure 5). If the two symmetry-related Arg 69 residues of the KaiA dimer are superimposed with the Arg 23 and Arg 23b residues of the KaiB dimer, the bulk of the molecules matches quite nicely (Figure 5). Especially, the curvature and rectangular footprint of that part of the molecular surface of KaiA and KaiB point to a possible common binding partner. We speculate that the arched surface, stretching between the two Arg 69 residues in the KaiA dimer and between the two Arg 23 residues in the KaiB dimer, may be where these dimers interact with KaiC. KaiA and KaiB appear to exert opposite effects on KaiC: KaiA promotes KaiC phosphorylation while KaiB attentuates it (Iwasaki et al, 2002; Kitayama et al, 2003). One compelling way to fulfil their respective functions would be for KaiA and KaiB to compete for the same binding site on KaiC. The importance of Arg 69 in KaiA is reflected by the loss of rhythmicity PCC7942 R249 H mutant (Nishimura et al, 2002), which would easily be explained by a change in the affinity of KaiA for its docking site on KaiC. The results of our binding assay (Figure 7C) support this interpretation as they show a drastic change in binding affinity between the equivalently located R69A mutant and wild-type forms of PCC7120 KaiA and KaiC.

Figure 5
(A) Rendering of the solvent accessible surface of KaiA. Residues that, when mutated, extend or abolish the circadian rhythm are highlighted in green. Residues that, when mutated, affect KaiBC expression are highlighted in purple. Residues from one monomer ...

Circadian period extending mutations in KaiA have been shown to cluster along residues 239–245 in PCC7942 and typically involve the replacement of charged side chains by uncharged ones (Nishimura et al, 2002). The corresponding residues in the PCC7120 KaiA model are Glu 59, Leu 60, Glu 63, and Asp 62. They flank the sides of the KaiA dimer close to Arg 69 (Figure 6). While the geometry of the residues flanking Arg 23 in KaiB is similar to that found in KaiA, their chemical nature is quite different. This mixture of conserved and deviating features would well explain the common affinity displayed by KaiA and KaiB for KaiC. Whether the phosphorylation state of KaiC is directly related to circadian rhythms remains to be determined.

Figure 6
Surface potential comparisons between KaiA and KaiB dimers using SPOCK with each of the proteins at 50% transparency: brighter regions represent physical overlap. (A) KaiA dimer in the same orientation as in Figure 1B. Arg 69 of the monomers is ...

Mutations of KaiA (F224S and E274 K of PCC7942) that alter KaiBC expression were believed to be located at other regions of KaiA (Nishimura et al, 2002). This assignment has now been confirmed. The mutation sites correspond to Phe 44 and Glu 94 in KaiA from PCC7120, both of which are located at the surface of KaiA opposite from the Arg 69 residues. This finding leads us to believe that the two reported functions of KaiA, interaction with DNA and regulation of autophosphorylation of KaiC, are performed by opposite regions on the KaiA dimer. As overexpression of KaiA has been shown to increase PkaiBC activity (Ishiura et al, 1998), perhaps this region of KaiA may have affinity for the promoter region or some currently unidentified transcription factor specific to the PkaiBC region. While recent studies have shown that PkaiBC is not absolutely necessary for maintaining circadian rhythmicity in cyanobacteria, they do not rule out the importance of PkaiBC upregulation by KaiA under normal conditions in response to external stimuli (Xu et al, 2003).

Given the high degree of sequence identity between corresponding proteins from PCC7942 and PCC7120, we feel fully justified in discussing the findings of mutational analyses obtained with PCC7942 proteins in the context of our structural results. To test whether KaiA and KaiB may be competing for a common binding site on KaiC, we performed in vitro binding assays. Our binding assay using purified KaiA, KaiB, and 6xHis-tagged-KaiC does agree with a competitive nature of wild-type KaiA and KaiB on KaiC in vitro (Figure 7A). Unexpectedly, incubation with increasing amounts of KaiA led to a corresponding reduction of binding of 6xHis-tagged-KaiC to the Ni beads (Figure 7A, lanes 2–5 and 7–10). Potential explanations for this finding are the masking of the His-tag either directly through binding of KaiA or via induction of a conformational change. As pre-incubation of KaiC with Ni beads followed by incubation with KaiA allowed co-elution of KaiA and KaiC, the former seems to be a more probable interpretation (see Supplementary data). KaiA binding to KaiC in this assay could subsequently be competed off using KaiB. Point mutations located at the postulated KaiC interaction sites on both KaiA and KaiB proteins were also generated to test their effects on binding; Arg 69 of KaiA and equivalently placed Arg 23 of KaiB were changed to alanine residues. As expected, the mutations did not affect these proteins' abilities to dimerize (data not shown), but strongly affected their binding affinities to KaiC. Whereas binding of KaiA (R69A) seems to be completely abolished (Figure 7C, lanes 1–3, 7–9), KaiB (R23A) still binds to KaiC albeit with reduced affinity (Figure 7C, lanes 13–15 and 19–21). These results are consistent with our assignment of the concave surface, common to KaiA and KaiB, representing the respective binding sites with KaiC. Despite the competitive nature of KaiA and KaiB binding to KaiC, as seen in our assay, these binding experiments alone cannot prove that KaiA and KaiB specifically bind onto the same site on KaiC. A conformational change of KaiC due to KaiA and KaiB binding as well as partial overlap of their respective binding sites could also be used to explain the current observations without, however, being able to explain the remarkable fit of the size and shape of the corresponding concave molecular surfaces.

The status of KaiC phosphorylation appears to be key to KaiC's ability to bind to the kaiBC promoter (Iwasaki et al, 2002). KaiC shows internal sequence duplication (Iwasaki et al, 1999) and forms hexameric rings in an ATP-dependent manner (Mori et al, 2002). One can imagine KaiA and KaiB dimers competing for binding sites along the outer edges of the KaiC hexamer, affecting KaiC's autophosphorylation activity, thus regulating KaiC function. Recently, Vakonakis et al (2004) proposed a model of a KaiA–KaiC complex consistent with such an interpretation. Autophosphorylation mixing experiments on KaiC in the presence of KaiA and KaiB showed that KaiC's autophosphorylation activity is markedly increased by the presence of KaiA, but that the level of KaiC activity dropped by half if KaiB was present. KaiB alone did not effect KaiC autokinase activity (Kitayama et al, 2003). This evidence, together with molecular structures of KaiA and KaiB, agrees well with a model of competition between KaiA and KaiB for a common binding site on KaiC.

While having no known enzymatic activity on their own, KaiA and KaiB are an intrinsic part of the cyanobacteria circadian system, their function apparently arising from their ability to affect KaiC phosphorylation rates (Williams et al, 2002; Kitayama et al, 2003). In conclusion, we have determined the crystal structures of both KaiA and KaiB from Anabaena sp PCC7120. The atomic models suggest how specific point mutations of KaiA and KaiB could lead to changes in the structures of these molecules, thereby interfering with their roles in establishing proper circadian rhythm. They are also in favour of a structure-based competition model between KaiA and KaiB for a shared binding site on KaiC. However, understanding the exact mechanism through which KaiA and KaiB affect KaiC phosphorylation will require a high-resolution structure of KaiC or, even better, of its complexes with KaiA and KaiB.

Materials and methods

Cloning and mutagenesis

Wild-type KaiA, KaiB, and KaiC genes were cloned by polymerase chain reaction (PCR) from Anabaena sp PCC7120 genomic DNA. The full-length KaiA gene was amplified using a 5′ primer (5′ CGCGCCATGGCTCAAGAAGTTGATCAGCAAATATTATTG) and a 3′ primer (5′ CGCGGGATCCTTATTTAAATATTGCCCCTCTATAGGCTTC) with NcoI and BamH1 restriction enzyme sites. The full-length KaiB gene was amplified using a 5′ primer (5′ CGCGCCATGGATAAAGCCAGAAAGACATACGTCCTCAAGC) and a 3′ primer (5′ CGCGGGATCCTTAAAGATTACTTTGTGCTTCCCAATCTTC), also with NcoI and BamH1 restriction sites. PCR products were gel purified, digested with NcoI and BamH1 (New England Biolabs), and ligated into the pET32 vector (Novagen). Both KaiA and KaiB were expressed as N-terminal 6-His-thioredoxin fusion proteins in BL21/codonplus E. coli cells (Stratagene). The full-length KaiC gene was amplified using a 5′ primer (5′ CGCGCATATGAGTGAGAAAGAGCAGCAGGAACAAC) and a 3′ primer (5′ CGCGGGATCCTTAGGAATCGGAACCCTGTTTGTCTTC) with NdeI and BamH1 restriction sites. The PCR product was gel purified, digested with NheI and BamH1 (New England Biolabs), and ligated into the pET28 vector. KaiC was expressed with an N-terminal 6-His-tag.

Two KaiB mutants were constructed by introducing Leu to Met mutations at either residue 28 or 69 (QuickChange™, Stratagene). These two sites were chosen based on secondary structure predictions placing them within putative helical regions in KaiB. The primers used were (L28 M) 5′ GTCAGGGCGTTAAAAACAATGAAAAATATTTTAGAACAG and (L69 M) 5′ CGACATTATCAAAGATTATGCCCCCACCAGTCCGC (mutated base pairs in bold and underlined). KaiB mutant proteins were expressed in BL21/codonplus Escherichia coli cells.

Protein expression and purification

KaiA-expressing cells were grown at 37°C in Luria–Bertani broth supplemented with 50 μM carbenicillin in a shaker to an OD600 of 0.8. Cells were induced with 1 mM isopropyl-β-D-thiogalactopyranoside (IPTG) and grown for 4 h (h) at room temperature (RT), then placed at 4°C overnight. Cells were pelleted at 4000 revolutions per minute (rpm), resuspended in buffer (50 mM Tris (pH 7.5), 500 mM NaCl, 5% glycerol, 1 mM PMSF), and sonicated using five 1 min pulses on a Branson Sonifier. The lysed cell solution was centrifuged for 1 h at 17 000 rpm, then the supernatant was mixed with 1 ml of Ni-NTA resin (Qiagen) per litre of lysed cells and gently rocked for 1 h at 4°C. The Ni-NTA resin was then transferred into a column and washed using 100 ml of 50 mM Tris (pH 7.5), 500 mM NaCl, 30 mM imidazole, 5% glycerol. Protein was eluted using 50 mM Tris (pH 7.5), 500 mM NaCl, 250 mM imidazole, 5% glycerol and collected in 1 ml fractions. Concentrated samples were pooled and diluted with 50 mM Tris (pH 7.5), 150 mM NaCl to a final volume of 10 ml. CaCl2 was added to a final concentration of 2.5 mM, followed by 0.5 units of thrombin (Calbiochem). The enzymatic digest proceeded for 2 h at RT, then overnight at 4°C. The sample was subjected to three concentration/dilution cycles using 50 mM Tris (pH 7.5), 150 mM NaCl to lower the imidazole concentration. The sample was then batch-bound onto pre-equilibrated Ni-NTA resin for 1 h at 4°C with gentle shaking and transferred into a column. Flow-through fractions were subjected to three concentration/dilution cycles using 10 mM Tris (pH 7.5), 10 mM Bis–Tris (pH 6.5), 30 mM NaCl, and 0.5 mM TCEP. Afterwards, the sample was loaded onto a Sephadex Q-column (Amersham-Pharmacia). Ion-exchange chromatography was performed using a linear gradient of 0–30% buffer B (Buffer A: 20 mM Bis–Tris (pH 6.5), 30 mM NaCl, 0.5 mM TCEP; Buffer B: 20 mM Bis–Tris (pH 6.5), 1 M NaCl, 0.5 mM TCEP). Fractions (1 ml) were collected and peak fractions identified on a Coomassie-Blue-stained 15% SDS–PAGE gel, pooled, and concentrated. This was followed by buffer exchange on a Superdex 200 column (Amersham-Pharmacia) pre-equilibrated with 20 mM Tris (pH 7.5) and 50 mM NaCl. Fractions (1 ml) were collected and their KaiA content and purity were verified on a gel, then fractions were pooled and concentrated.

For KaiB, the growth, induction, cell lysis, and first nickel column purification steps were performed as described for KaiA. In addition, fractions containing concentrated KaiB protein were pooled and diluted to 10 ml total volume using 50 mM Tris (pH 7.5), 150 mM NaCl, 2.5 mM CaCl2, and 5% glycerol, then 0.5 units of thrombin was added and the protein solution was injected into a Slide-A-Lyser (Pierce). Protein was dialysed against 4 l of buffer (50 mM Tris (pH 7.5), 150 mM NaCl, 2.5 mM CaCl2) overnight at 4°C with gentle stirring to remove the 6xHis-Trx fusion tag. The digest was stopped by adding PMSF to a final concentration of 1 mM. The protein solution was mixed with 1 ml of Ni-NTA resin for 1 h at 4°C, packed into a column, and flow-through was collected and concentrated. The protein underwent buffer exchange on a Superdex 200 column (Amersham-Pharmacia) pre-equilibrated with 20 mM Tris (pH 7.5), 50 mM NaCl. Fractions (1 ml) were collected and their KaiB content and purity were checked on a gel, then fractions were pooled and concentrated.

For the expression of the seleno-methionine derivatives (Se-Met), vectors containing the genes for KaiB (L28 M) or KaiB (L69 M) were transformed into B834 (DE3) cells (Novagen). Protein was expressed in M9 media enriched with seleno-methionine and the other 19 amino acids (Ramakrishnan et al, 1993). Purification was the same as for the native protein, except that β-mercaptoethanol was added to bring its concentration to 5 mM in the dialysis buffer and subsequent buffer-exchange steps were performed with 0.5 mM TCEP (BioShop) present.

KaiC-expressing cells were grown at 37°C in Luria–Bertani broth supplemented with kanamycin in a shaker until the OD600 was 0.8. Induction, cell lysis, and nickel column purification steps were performed as described for KaiB protein, except that all buffers were supplemented with 0.5% SDS. Following elution, samples were pooled and buffer was exchanged three times with 50 mM Tris (pH 7.5), 150 mM NaCl using a 10 kDa cutoff protein concentrator (Millipore). Afterwards, the protein sample was placed on ice for 30 min to precipitate the majority of the SDS, then pelleted via centrifugation at 14 000 rpm for 5 min. The supernatant was collected and again placed on ice. Samples were taken to verify the presence and to determine the amount of KaiC remaining in the soluble fraction.

Size exclusion column chromatography

All protein chromatographic steps were performed using an Äkta FPLC system (Amersham-Pharmacia). Protein samples (500 μl) were prepared as above and size-fractionated by gel filtration on a Superdex 200 (Amersham-Pharmacia) column at 4°C. Protein standards used for calibrating the column were ribonuclease A, chymotrypsin, ovalbumin, albumin, and aldolase (Amersham). Peak samples were loaded onto SDS–PAGE gels and visualized using Coomassie Blue staining. For each protein preparation, two or three independent experiments yielded the same results.

Crystallization, data collection, and processing

KaiA was crystallized by vapour diffusion against 100 mM MES (pH 6.0), 1.1 M MgSO4. The crystals were cryoprotected by adding 20% glycerol (Sigma) to their mother liquor. To generate a bromine derivative, a native KaiA crystal was soaked for 15 h in a solution containing 100 mM MES (pH 6.0), 1.1 M MgSO4, 160 mM NaBr, and 8% glycerol. No backsoak was performed on the bromine-soaked crystals. Optimal crystallization conditions for KaiB were 100 mM Tris (pH 7.5), 18% (v/v) PEG 400 (Fluka), and for KaiB (L28 M) 100 mM Tris (pH 7.0), 23.5% PEG400, plus 5% 1,6-hexanediol (Fluka).

For data collection, crystals were flash-frozen in a stream of nitrogen at 100 K. All data sets were collected at the Advanced Photon Source beamline 14D (BioCARS) with a MAR345 area detector. MAD data sets (λremote=0.9407 Å, λpeak=0.9188 Å, λedge=0.9192 Å) were collected on a bromine-soaked KaiA crystal, whereas the native 2.0 Å resolution data set was obtained at λ=0.9407 Å.

A native 2.2 Å resolution data set was collected from a KaiB crystal at λ=0.9407 Å. MAD data sets (λremote=0.9571 Å, λpeak=0.9794 Å, λedge=0.9797 Å) of 2.5 Å resolution were obtained from a KaiB (L28 M) Se-methionine crystal. Data were reduced using DENZO and scaled using SCALEPACK (Otwinowski and Minor, 1997).

Phase determination

The structure for KaiA was determined applying the MAD method of phasing based on the anomalous signals of bromine atoms diffused into the crystal matrix. Using diffraction data of 2.4 Å resolution, the program package SOLVE (Terwilliger and Berendzen, 1999) found a solution to the heavy atom substructure. Three sites with high occupancies (0.445, 0.667, and 0.477) were found and provided phase information (FOM=0.51). The corresponding electron density map was then interactively traced using the program O (Jones et al, 1991), resulting in a preliminary model of KaiA, which was subsequently refined against the native data using the program CNS, version 1.1 (Brünger et al, 1998). Refinement statistics are given in Table I.

Table 1
Data collection and refinement statistics for KaiA

The molecular structure of KaiB was determined by the MAD method using KaiB (L28 M) Se-Met protein crystals. Initial positions of the Se atoms were determined with the help of the program SOLVE. Two high-occupancy (0.779 and 0.820) sites were found, one for each subunit of the dimer, and the overall figure of merit was 0.67 for data to 2.5 Å resolution. The density-modified map was then traced manually using the program O. Subsequent refinements were carvried out against the wild-type data set using the program CNS, version 1.1. Refinement statistics are given in Table II. Figures were prepared using the programs SPOCK (Christopher, 1998), MOLSCRIPT (Kraulis, 1991), and RASTER3D (Merritt and Murphy, 1994). Atomic coordinates and structure factors of KaiA and KaiB have been deposited in the RCSB Protein Databank under the accession codes 1R5Q and 1R5P, respectively.

Table 2
Data collection and refinement statistics for KaiB

Generation of site-specific mutants to probe protein–protein interactions

The R69A KaiA mutant was chosen due to its conservation and location at the proposed site of interaction between KaiC and KaiA. It was generated by using site-specific mutagenesis. The coding strand sequence of the primer used was (R69A) 5′ GAATTTTCTAAGCAGCTAGCACTAGAGGGCAGAGGTGAT G (mutation site is noted in bold and underlined). Similarly for KaiB, Arg 23 was mutated to Ala and the coding strand sequence of the primer used was (R23A) 5′ AACACACCAAATTCAGTCGCGGCGTTAAAAACACTCAAAA AT (mutation site is noted in bold and underlined). Both KaiA and KaiB mutant proteins were expressed and purified from BL21/codonplus E. coli cells as described for wild-type proteins.

6xHis–KaiC interaction assay

The KaiA, KaiB, KaiA (R69A), KaiB (R23A), and 6xHis–KaiC proteins were purified as described above. For each binding assay, varying molar ratios of purified proteins were mixed together in binding solution (50 mM Tris (pH 8.0), 30 mM NaCl, 2.5 mM MgCl2, 5 mM ATP) to a final volume of 400 μl in a 1.5 ml tube. Mixtures were gently rocked on a Nutator (Clay Adams) for 2 h at RT. Ni-NTA agarose resin (Qiagen) was pre-equilibrated using binding solution. In all, 200 μl of a 1:1 slurry of resin and buffer was added to each tube, then the mixture was rocked again for 1 h at RT. Following binding, tubes were centrifuged at 1000 rpm in a microfuge to pellet the resin. The supernatant was carefully removed, resuspended in wash buffer (50 mM Tris (pH 8.0), 30 mM NaCl, 2.5 mM MgCl2, 5 mM ATP, 15 mM imidazole), and transferred into a 0.22 μm spin filtration unit (Millipore). Resin was washed three times using 500 μl of wash buffer and centrifuging the wash through the filter at 1000 rpm for a total wash volume of 1.5 ml. The collection tube was then dried and 500 μl of elution buffer (50 mM Tris (pH 8.0), 30 mM NaCl, 2.5 mM MgCl2, 5 mM ATP, 1 M imidazole) was used to resuspend the resin in the filter unit. The filter unit was centrifuged once more to elute the resin-bound proteins into the collection tubes. Equal volumes of samples from elution and unbound supernatant were loaded onto 10–20% SDS–PAGE gradient gels (NOVEX) and visualized using Coomassie Blue staining.

Sequential 6xHis–KaiC interaction assay

For sequential binding assays, 200 μg of 6xHis–KaiC was mixed with 1000 μl 1:1 slurry of Ni-NTA resin and binding solution in a 2.0 ml tube. Solutions used were identical to those in the previous assay. Mixtures were rocked on a Nutator for 1 h at RT. Following incubation, resin was pelleted at 1000 rpm in a microfuge and supernatants were kept. The resin was washed three times with 1 ml binding solution, then resuspended with 500 μl of the same solution. In all, 200 μl of the resuspended 1:1 slurry was taken as a sample and placed into a 0.22 μm spin filtration unit (Millipore). To the remaining 800 μl of 1:1 slurry, 320 μg of KaiA or KaiB was added and mixtures were gently rocked as before. Again, the resin was pelleted and the supernatant was kept. After the resin was washed as before, it was resuspended with 400 μl of binding solution. Again, a 200 μl sample of the 1:1 slurry was taken. To the remaining 600 μl of slurry, 320 μg of the competitor protein (KaiB if KaiA was already added and vice versa) was added and the mixtures were gently rocked. The resin was then pelleted and the supernatant was kept. After washing, the resin was resuspended with 300 μl of binding solution. A volume of 200 μl of this slurry was taken as the last sample. All resin samples placed in spin filtration units were washed using 500 μl wash buffer and centrifuged to remove excess solution. After that, resin samples were resuspended in 100 μl elution buffer and filter units were centrifuged to elute the resin-bound proteins into collection tubes. Equal volumes of samples from the elution procedure and from unbound supernatants were loaded onto 10–20% SDS–PAGE gradient gels and visualized using Coomassie Blue staining.

Supplementary Material

Supplementary Material

Acknowledgments

We are very grateful to Dr Takakazu Kaneko, Kazusa DNA Research Institute, Kisarazu, Chiba, Japan, for his gift of Anabaena sp PCC7120 genomic DNA. We thank the staff at BioCARS sector 14 beamlines at the Advanced Photon Source, Argonne National Laboratories, for their generous time commitments and support. This work was supported by the Ontario Cancer Institute/Princess Margaret Hospital Foundation, Toronto (RGG & NW), the Canada Research Chairs Program and the National Sciences and Engineering Research Council (EFP), and an Ontario Graduate Studentship (NW). Use of the Advanced Photon Source was supported by the Basic Energy Sciences, Office of Science, United States Department of Energy, under Contract W-31-109-Eng-38. Use of the BioCARS sector 14 was supported by the National Center for Research Resources, National Institutes of Health, under grant RR07707.

References

  • Brünger AT, Adams PD, Clore GM, DeLano WL, Gros P, Grosse-Kunstleve RW, Jiang JS, Kuszewski J, Nilges M, Pannu NS, Read RJ, Rice LM, Simonson T, Warren GL (1998) Crystallography & NMR system: A new software suite for macromolecular structure determination. Acta Crystallogr D Biol Crystallogr 54: 905–921 [PubMed]
  • Carlacci L, Chou KC (1990) Electrostatic interactions between loops and alpha-helices in four-helix bundle proteins. Protein Eng 4: 225–227 [PubMed]
  • Christopher JA (1998) SPOCK., The Center for Macromolecular Design. College Station, TX: Texas A&M University
  • Dauter Z, Dauter M (1999) Anomalous signal of solvent bromides used for phasing of lysozyme. J Mol Biol 289: 93–101 [PubMed]
  • Dunlap JC (1999) Molecular bases for circadian clocks. Cell 96: 271–290 [PubMed]
  • Dunlap JC, Loros JJ, Liu Y, Crosthwaite SK (1999) Eukaryotic circadian systems: cycles in common. Genes Cells 4: 1–10 [PubMed]
  • Ishiura M, Kutsuna S, Aoki S, Iwasaki H, Andersson CR, Tanabe A, Golden SS, Johnson CH, Kondo T (1998) Expression of a gene cluster kaiABC as a circadian feedback process in cyanobacteria. Science 281: 1519–1523 [PubMed]
  • Iwasaki H, Kondo T (2000) The current state and problems of circadian clock studies in cyanobacteria. Plant Cell Physiol 41: 1013–1020 [PubMed]
  • Iwasaki H, Nishiwaki T, Kitayama Y, Nakajima M, Kondo T (2002) KaiA-stimulated KaiC phosphorylation in circadian timing loops in cyanobacteria. Proc Natl Acad Sci USA 99: 15788–15793 [PMC free article] [PubMed]
  • Iwasaki H, Taniguchi Y, Ishiura M, Kondo T (1999) Physical interactions among circadian clock proteins KaiA, KaiB and KaiC in cyanobacteria. EMBO J 18: 1137–1145 [PMC free article] [PubMed]
  • Jones TA, Zou JY, Cowan SW, Kjeldgaard M (1991) Improved methods for building protein models in electron density maps and the location of errors in these models. Acta Crystallogr A 47 (Pt 2): 110–119 [PubMed]
  • Kageyama H, Kondo T, Iwasaki H (2003) Circadian formation of clock protein complexes by KaiA, KaiB, KaiC, and SasA in cyanobacteria. J Biol Chem 278: 2388–2395 [PubMed]
  • Kitayama Y, Iwasaki H, Nishiwaki T, Kondo T (2003) KaiB functions as an attenuator of KaiC phosphorylation in the cyanobacterial circadian clock system. EMBO J 22: 2127–2134 [PMC free article] [PubMed]
  • Kondo T, Strayer CA, Kulkarni RD, Taylor W, Ishiura M, Golden SS, Johnson CH (1993) Circadian rhythms in prokaryotes: luciferase as a reporter of circadian gene expression in cyanobacteria. Proc Natl Acad Sci USA 90: 5672–5676 [PMC free article] [PubMed]
  • Kondo T, Tsinoremas N, Golden S, Johnson C, Kutsuna S, Ishiura M (1994) Circadian clock mutants of cyanobacteria. Science 266: 1233–1236 [PubMed]
  • Kraulis PJ (1991) MOLSCRIPT. J Appl Crystallogr 24: 946–950
  • Leipe DD, Aravind L, Grishin NV, Koonin EV (2000) The bacterial replicative helicase DnaB evolved from a RecA duplication. Genome Res 10: 5–16 [PubMed]
  • Lorne J, Scheffer J, Lee A, Painter M, Miao VP (2000) Genes controlling circadian rhythm are widely distributed in cyanobacteria. FEMS Microbiol Lett 189: 129–133 [PubMed]
  • Merritt EA, Murphy MEP (1994) Raster3D. Acta Crystallogr D 50: 869–873 [PubMed]
  • Mori T, Johnson CH (2001) Circadian programming in cyanobacteria. Semin Cell Dev Biol 12: 271–278 [PubMed]
  • Mori T, Saveliev SV, Xu Y, Stafford WF, Cox MM, Inman RB, Johnson CH (2002) Circadian clock protein KaiC forms ATP-dependent hexameric rings and binds DNA. Proc Natl Acad Sci USA 99: 17203–17208 [PMC free article] [PubMed]
  • Nishimura H, Nakahira Y, Imai K, Tsuruhara A, Kondo H, Hayashi H, Hirai M, Saito H, Kondo T (2002) Mutations in KaiA, a clock protein, extend the period of circadian rhythm in the cyanobacterium Synechococcus elongatus PCC 7942. Microbiology 148: 2903–2909 [PubMed]
  • Nishiwaki T, Iwasaki H, Ishiura M, Kondo T (2000) Nucleotide binding and autophosphorylation of the clock protein KaiC as a circadian timing process of cyanobacteria. Proc Natl Acad Sci USA 97: 495–499 [PMC free article] [PubMed]
  • Otwinowski Z, Minor W (1997) Processing of X-ray diffraction data collected in oscillation mode. Methods Enzymol 276: 307–326
  • Pittendrigh CS (1993) Temporal organization: reflections of a Darwinian clock-watcher. Annu Rev Physiol 55: 16–54 [PubMed]
  • Ramakrishnan V, Finch JT, Graziano V, Lee PL, Sweet RM (1993) Crystal structure of globular domain of histone H5 and its implications for nucleosome binding. Nature 362 (6417): 219–223 [PubMed]
  • Terwilliger TC, Berendzen J (1999) Automated MAD and MIR structure solution. Acta Crystallogr D Biol Crystallogr 55: 849–861 [PMC free article] [PubMed]
  • Vakonakis I, Sun J, Wu T, Holzenburg A, Golden S, LiWang A (2004) NMR structure of the KaiC-interacting C-terminal domain of KaiA, a circadian clock protein: implications for KaiA–KaiC interaction. Proc Natl Acad Sci USA 101: 1479–1484 [PMC free article] [PubMed]
  • Walker AK, See R, Batchelder C, Kophengnavong T, Gronniger JT, Shi Y, Blackwell TK (2000) A conserved transcription motif suggesting functional parallels between Caenorhabditis elegans SKN-1 and Cap‘n'Collar-related basic leucine zipper proteins. J Biol Chem 275: 22166–22171 [PubMed]
  • Williams SB, Vakonakis I, Golden SS, LiWang AC (2002) Structure and function from the circadian clock protein KaiA of Synechococcus elongatus: a potential clock input mechanism. Proc Natl Acad Sci USA 99: 15357–15362 [PMC free article] [PubMed]
  • Xu Y, Mori T, Johnson CH (2000) Circadian clock-protein expression in cyanobacteria: rhythms and phase setting. EMBO J 19: 3349–3357 [PMC free article] [PubMed]
  • Xu Y, Piston DW, Johnson CH (1999) A bioluminescence resonance energy transfer (BRET) system: application to interacting circadian clock proteins. Proc Natl Acad Sci USA 96: 151–156 [PMC free article] [PubMed]
  • Xu Y, Tetsuya M, Johnson C (2003) Cyanobacterial circadian clockwork: roles of KaiA, KaiB and the kaiBC promoter in regulating KaiC. EMBO J 22: 2117–2126 [PMC free article] [PubMed]
  • Young MW, Kay SA (2001) Time zones: a comparative genetics of circadian clocks. Nat Rev Genet 2: 702–715 [PubMed]

Articles from The EMBO Journal are provided here courtesy of The European Molecular Biology Organization
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...