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

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.

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 are labelled.

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 indicate gaps. Secondary structural elements determined in the crystal structure are aligned with the corresponding sequence.

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 sheets), while monomer B is coloured in darker purple and gold. N- and C-termini are labelled. (C) Stereo view of the hydrophobic core of KaiB. Residues are labelled to the right and the electron density map is shown in light grey.

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 indicate gaps.

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 are given the suffix a, while residues from the other monomer are given the suffix b. (B) Rendering of the solvent accessible surface of KaiB. Residues that, when mutated, decrease or abolish circadian rhythm are highlighted in green. Residues from one monomer are given the suffix a, while residues from the other monomer are given the suffix b. Both molecules are oriented with the two-fold rotation axis relating the two subunits located in the plane of the paper.

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 located in the blue region at the bottom of the surface plot. (B) KaiB dimer in the same orientation as in Figure 3B. Arg 23 of the KaiB monomers is located in the blue region at the ‘bottom' of the surface plot. (C) Overlay of (A) and (B). (D) KaiA dimer ‘bottom' surface. (E) KaiB dimer ‘bottom' surface. (F) Overlay of (D) and (E). (G) KaiA dimer side surface. (H) KaiB dimer side surface. (I) Overlay of (G) and (H).

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 that bound to the Ni beads were then eluted using 1 M imidazole and loaded in the lanes labelled ‘Bound'. Known protein bands are indicated by arrows on the right side of the gels; an intersubunit disulphide bond of KaiA is quite resistant to reduction by β-mercaptoethanol, resulting in the persistence of dimers. (A) Binding profiles of wild-type KaiA or KaiB to a fixed amount of 6x-His-tagged KaiC bound to Ni beads. The presence of a particular protein in a given solution is indicated by ‘+' signs, and its absence by ‘−' signs. ‘+': 40 μg of protein was added; ‘++': 80 μg of protein was added; and ‘+++': 160 μg of a particular protein was added. Equal volumes of sample were loaded in each lane. Lanes are numbered below. (B) Competition binding profiles of increasing amounts of KaiA competing with a fixed amount of KaiB, or vice versa, on a fixed amount of 6x-His-tagged KaiC bound to Ni-beads. (C) Comparison of binding profiles of mutant and wild-type KaiA and KaiB proteins to 6x-His-tagged KaiC bound to Ni beads.