From primordial clocks to circadian oscillators

Abstract

Circadian rhythms play an essential part in many biological processes, and only three prokaryotic proteins are required to constitute a true post-translational circadian oscillator¹. The evolutionary history of the three Kai proteins indicates that KaiC is the oldest member and a central component of the clock². Subsequent additions of KaiB and KaiA regulate the phosphorylation state of KaiC for time synchronization. The canonical KaiABC system in cyanobacteria is well understood^3-6, but little is known about more ancient systems that only possess KaiBC. However, there are reports that they might exhibit a basic, hourglass-like timekeeping mechanism^7-9. Here we investigate the primordial circadian clock in Rhodobacter sphaeroides, which contains only KaiBC, to elucidate its inner workings despite missing KaiA. Using a combination of X-ray crystallography and cryogenic electron microscopy, we find a new dodecameric fold for KaiC, in which two hexamers are held together by a coiled-coil bundle of 12 helices. This interaction is formed by the carboxy-terminal extension of KaiC and serves as an ancient regulatory moiety that is later superseded by KaiA. A coiled-coil register shift between daytime and night-time conformations is connected to phosphorylation sites through a long-range allosteric network that spans over 140 Å. Our kinetic data identify the difference in the ATP-to-ADP ratio between day and night as the environmental cue that drives the clock. They also unravel mechanistic details that shed light on the evolution of self-sustained oscillators.

Fig. 1. The extended C-terminal tail of KaiC_RS forms a coiled-coil interaction with an exposed A loop for KaiA-independent phosphorylation of KaiC.

a, Schematic of the phylogenetic tree of kaiC showing the appearance of kaiB and kaiA during evolution. The kaiC clade with an approximately 50-amino-acid C-terminal extension is labelled in red, and a timeline was predicted as previously reported. Ga, billion years ago. b, Phosphorylation rate over time of KaiC_RS (6.5 ± 1.0 h⁻¹) and KaiC_SE in the presence (0.40 ± 0.02 h⁻¹) or absence of KaiA_SE at 30 °C. The s.d. in reported parameters were obtained from the fitting. c, Nucleotide exchange between ATP and mant-ATP in KaiC_RS alone (18.0 ± 1.5 h^–1) compared with KaiC_SE in the presence (4.7 ± 0.3 h^–1) and absence of KaiA_SE (0.08 ± 0.04 h^–1) measured at 30 °C. Representative traces are shown, and the fitted parameters (mean ± s.d.) were obtained from three replicate measurements. d, X-ray structure of dodecameric KaiC_RS (PDB: 8DBA) coloured by hexamer A (light green) and hexamer B (dark green). The CI, CII and coiled-coil domains are indicated, and the A loop is labelled in e. e, Superposition based on an alignment of the CII domain of KaiC_RS (green; PDB: 8DBA, chain B), KaiC_SE (purple; PDB: 1TF7, chain B) and KaiC_SE-S431E/T432A (yellow; PDB: 7S65, chain A) shows that KaiC_RS has an extended A loop orientation that no longer forms the inhibitory interaction with the 422 loop (KaiC_SE numbering). The conformation of the 422 loop in KaiC_RS resembles the one seen in the cryo-EM structure of the phosphomimetic KaiC_SE-S431E/T432A (yellow; PDB: 7S65). No electron density is observed for the C-terminal part of wild-type KaiC_SE and the S431E/T432A mutant owing to flexibility, and the missing 22 residues for wild-type KaiC_SE (46 for S431E/T432A) are represented by a dashed line (not shown for the mutant).

Fig. 2. A coiled-coil partner switch coupled to an allosteric network in the CII domain promotes autodephosphorylation.

a, X-ray structure of KaiC_RS-Δcoil was solved in the C222₁ space group and contained three monomers in the asymmetric unit, with ADP present in all active sites. The obtained electron density map allowed for model building up to Pro463, which indicated that the truncation at position 490 results in enhanced flexibility beyond Pro463. Phosphorylation of Ser414 (pS414) was observed in chain B (cyan) as shown by the electron density mF_o–DF_c polder map (green mesh, 3σ contour level). b, Assembly analysis using the PISA software revealed a hexamer as the most probable quaternary structure (top view). c, Structural comparison of the coiled-coil domain for unphosphorylated KaiC_RS (dark and light green; X-ray structure) and the KaiC_RS-S413E/S414E phosphomimetic mutant (dark and light blue; cryo-EM structure). d, Overlay of interacting dimers of the structures in c using the CII domain of chain A as a reference (dark shades; bottom). Unphosphorylated KaiC_RS (dark green) interacts with the opposite partner on the right (light green), whereas KaiC_RS-S413E/S414E (dark blue) interacts with the partner on the left (light blue). The hydrophobic packing in the coiled-coil domain is mediated by only the Cβ atoms of alanine and arginine residues in unphosphorylated KaiC_RS, but involves the entire side chain of leucine and isoleucine residues in the phosphomimetic structure. e, Allosteric network in the phosphomimetic state (blue) from the coil (light blue) propagating through the KaiC_RS CII domain to the active site (dark blue) compared with the unphosphorylated state (dark green) (Supplementary Video 1). f, Autodephosphorylation of KaiC_RS and KaiC_RS-Δcoil over time in the presence of 4 mM ADP at 30 °C. The phosphorylated (P) and unphosphorylated (U) proteins were separated by Zn²⁺ Phos-tag SDS–PAGE (for gel source data, see Supplementary Fig. 1).

Fig. 3. The regulatory role of KaiB_RS in the phosphorylation–dephosphorylation cycle of KaiC_RS.

a, SDS–PAGE gel of 3.5 μM KaiC_RS and 4 mM ATP in the absence (top) and presence (bottom) of 3.5 μM KaiB_RS at 35 °C, with the percentage of ATP indicated at specific time points. b, Phosphorylation (single and double) of KaiC_RS during the reaction in the absence (grey circles) or presence (red diamonds) of KaiB_RS. c, Phosphorylation–dephosphorylation cycle of 3.5 μM phosphorylated KaiC_RS in the absence and presence of 3.5 μM KaiB_RS in a constant ATP-to-ADP ratio of high ATP (4 mM) to mimic daytime and about 25% ATP to mimic the night time (exact percentage of ATP indicated at specific time points) at 30 °C. U, S and D in a and c represent the unphosphorylated, single phosphorylated (at Ser413 or Ser414) and double phosphorylated state of KaiC_RS, respectively (for gel source data, see Supplementary Fig. 1). d, ATPase activity of wild-type KaiC_RS in the absence and presence of KaiB_RS, KaiC_RS-E62Q/E63Q in the absence and presence of KaiB_RS, and KaiC_RS-E302Q/E303Q in the absence and presence of KaiB_RS at 30 °C. Bar graphs show mean ± s.d. from three replicates. e, Time-dependent autodephosphorylation of ³²P-labelled KaiC_RS bound with ADP in the presence of 20 μM KaiB_RS and 4 mM ADP at 30 °C showing phosphorylated ³²P-KaiC_RS, ³²P-ATP and free ³²Pi. The reaction products were separated by thin layer chromatography. f, The decay of phosphorylated ³²P-KaiC_RS bound with 4 mM ADP in the absence (grey circles) and presence (red diamonds) of KaiB_RS at 30 °C is obtained from autoradiography quantification (Extended Data Fig. 7). g, The nucleotide exchange of 3.5 μM KaiC_RS (grey trace) and 3.5 μM KaiC_RS in complex with 30 μM KaiB_RS (red dotted trace) in the presence of ATP with mant-ATP. Representative traces are shown, and the fitted parameters (mean ± s.d.) were obtained from three replicate measurements.

Fig. 4. KaiB_RS binds to the post-hydrolysis state and accelerates the ATPase activity of KaiC_RS.

a, Cryo-EM structure of KaiC_RS-S413E/S414E (yellow) in complex with KaiB_RS (blue) (PDB: 8FWJ). b, Superposition of KaiC_RS-S413E/S414E (yellow) bound to KaiB_RS (blue) (PDB: 8FWJ) and KaiC_TE-S413E (dark grey) bound to fsKaiB_TE (fold-switch, binding-competent state of KaiB_TE; light grey) (PDB: 5JWQ). c,d, Binding of KaiB_RS (blue) creates a tunnel (grey mesh) that enables water to reach the catalytic position (red sphere) for ATP hydrolysis in the CI domain. e, Binding of wild-type and mutant forms of KaiC_RS to His-tagged KaiB_RS in the presence of ADP or an ATP-recycling system at 25 °C. Bar graphs show mean ± s.d. from three replicates. f, Fluorescence anisotropy of unlabelled KaiB_RS competitively displacing KaiB_RS–6IAF (where 6IAF is the fluorophore) from unphosphorylated KaiC_RS in the presence of ADP (dark green circles) and phosphorylated KaiC_RS in the presence of the ATP-recycling system (light green triangles) at 30 °C. The average anisotropy and standard error were calculated from ten replicate measurements. g, Schematic of the uncovered mechanism of KaiC_RS regulated by coiled-coil interactions and KaiB_RS in the CI and CII domains.

Extended Data Fig. 1. Evolution of kaiC and sequence alignment of kaiC subgroups.

(a) Phylogenetic tree of kaiC homologs, where kaiC genes that have an approximately 50 amino acids C-terminal extension are labeled in red. Rhodobacter sphaeroides strain KD131 studied here and Synechococcus elongatus PCC 7942 (widely studied in the literature) are highlighted in green and pink, respectively. The accession code and organism are shown at the tip of the branches, the numbers at each node represent the aBayes bootstrap values, and the legend for branch length is shown (see also Supplementary Datasets 1 and 2). (b) A sequence alignment of the CII domain of the kaiC subgroups annotated with its sequence similarity. Residue Glu490, the position where the stop codon was introduced in the truncated KaiC_RS-Δcoil construct, is shown in red and marked with an arrow.

Extended Data Fig. 2. Auto-phosphorylation and nucleotide exchange rates of KaiC.

(a) 10% SDS-PAGE gel of 3.5 μM KaiC_RS in the presence of 4 mM ATP and using an ATP-recycling system at 30 °C. U, S, and D represent unphosphorylated, singly, and doubly phosphorylated KaiC_RS, respectively. (b) Densitometric analysis of auto-phosphorylation (single + double phosphorylation) from panel (a) over time yields a rate of 6.5 ± 1.0 h⁻¹. (c) 6.5% SDS-PAGE gel of 3.5 μM KaiC_SE in the presence of 1.2 μM KaiA_SE and 4 mM ATP at 30 °C. U and P represent unphosphorylated and phosphorylated KaiC_SE, respectively. (d) Densitometric analysis of auto-phosphorylation of KaiC_SE activated by KaiA_SE (panel (c)) shows a rate of 0.40 ± 0.02 h⁻¹ and is substantially slower than for KaiC_RS. The standard deviation for parameters in (b) and (d) were obtained from data fitting. (e) 10% SDS-PAGE gels for experiments with 3.5 μM KaiC_RS in the presence of 4 mM ATP and using an ATP-recycling system between 20 and 35 °C show that the level of phosphorylation increases with temperature. U, S, and D represent unphosphorylated, singly, and doubly phosphorylated KaiC_RS, respectively. For gel source data in (a), (c), and (e), see Supplementary Figure 2. (f) Bar graphs indicating the nucleotide exchange rate in the CII domain of KaiC_RS incubated with 50 μM ATP in the presence of an ATP-recycling system, and then mixed with 250 μM mant-ATP. An increase in fluorescence intensity at 440 nm was recorded and the single-exponential time traces were fitted to obtain the exchange rate constants: 3.6 ± 0.8 h⁻¹ (20 °C), 12.2 ± 1.0 h⁻¹ (25 °C), 18.5 ± 1.5 h⁻¹ (30 °C), and 25.2 ± 0.2 h⁻¹ (35 °C). Experiments were performed in triplicate and data are presented as mean values ± s.d.

Extended Data Fig. 3. Oligomeric states of KaiC_RS and effect of coiled-coil domain on rates of nucleotide exchange and auto-phosphorylation.

(a) Oligomerization analysis of KaiC_RS (dodecamer, green line) and truncated KaiC_RS-Δcoil (hexamer, cyan line) by analytical gel-filtration chromatography. The protein size markers are indicated at the top. (b) Comparison of the elution profiles of KaiC_RS-Δcoil (cyan line) and KaiC_SE (purple line) from size-exclusion chromatography shows a hexameric state for both KaiC_RS-Δcoil and KaiC_SE. (c) Oligomeric states of KaiC_RS (dodecamer, green line), KaiC_RS-Δcoil (hexamer, cyan line), and KaiC_SE (hexamer, purple line) were also measured by analytical ultracentrifugation (sedimentation velocity at 30,000 rpm and 20 °C) and the results agree with the data shown in panels (a) and (b). The graph in panel (c) represents the sedimentation coefficient distribution [c(s)]. (d) The change in fluorescence at 440 nm (ΔF_440nm) represents the nucleotide exchange between ATP and mant-ATP at 30 °C for KaiC_RS (green trace, 18.0 ± 1.5 h⁻¹) and KaiC_RS-Δcoil (cyan trace, 19.1 ± 0.8 h⁻¹). Representative traces are shown and the fitted parameters (mean ± s.d.) were obtained from three replicate measurements. (e) Zn²⁺ Phos-tag^TM SDS-PAGE gel shows the level of phosphorylation over time of KaiC_RS (upper gel) and KaiC_RS-Δcoil (lower gel) at 35 °C. P and U represent phosphorylated and unphosphorylated protein, respectively. (f) Phosphorylation level over time of KaiC_RS (green circles, 7.4 ± 0.3 h⁻¹) and KaiC_RS-Δcoil (cyan circles, 5.5 ± 0.4 h⁻¹) analyzed by densitometric analysis of Zn²⁺ Phos-tag^TM SDS-PAGE gel in (e). (g) First derivative of thermal-stability curves measured for unphosphorylated KaiC_RS bound with ADP (brown line) and phosphorylated KaiC_RS bound with ATP (green line). The extracted temperatures of denaturation are 50 °C (unphosphorylated KaiC_RS in the presence of 1 mM ADP) and 58 °C (phosphorylated KaiC_RS in the presence of 1 mM ATP), respectively. (h) Dodecameric state of unphosphorylated KaiC_RS (40 μM) bound with ADP measured by size-exclusion chromatography. (i) SDS-PAGE gel shows dephosphorylation of Ser413 over time at 30 °C in the presence of 4 mM ADP (U and pSer413 represent unphosphorylated and Ser413-phosphorylated KaiC_RS, respectively) with the corresponding kinetics shown in the right panel (confirmed by MS/MS) with a rate constant of 11.5 ± 0.8 h⁻¹. This result suggests that the coiled-coil domain promotes KaiC_RS dephosphorylation. For gel source data in (e) and (i), see Supplementary Figure 2. The standard deviation for parameters in (f) and (i) were obtained from data fitting.

Extended Data Fig. 4. Graphical description of the cryo-EM processing workflow and validation of the final dodecamer structures.

The workflow (see description in Methods section) demonstrates a typical image (scale bar: 60 nm) and representative good class averages. The ab initio and final reconstructions are shown. Shown alongside the final reconstruction is the angular plot demonstrating the distribution of particle views and the Fourier shell correlation curve used for the global resolution estimation (a) KaiC_RS-S413E/S414E alone and (b) KaiC_RS-S413E/S414E:KaiB_RS complex. To validate the final combined dodecamer structures, the data were reprocessed for the full dodecamer. The figure shows representative good class averages, the final reconstruction, angular distribution and Fourier shell correlation curve for the (c) KaiC_RS-S413E/S414E alone and (d) KaiC_RS-S413E/S414E:KaiB_RS complex dodecamers. (e) Comparison of the C1 and D6 reconstructions of KaiC_RS-S413E/S414E alone and KaiC_RS-S413E/S414E:KaiB_RS, and Fourier shell correlation curves for the C1 reconstructions. The C1/D6 comparisons do not reveal discernable differences, suggesting that these complexes have D6 symmetry.

Extended Data Fig. 5. Correlation between the coiled-coil register shift and phosphorylation, and model for consecutive phosphorylation/dephosphorylation events in CII domain of KaiC_RS.

(a) Structural comparison between KaiC_RS (green) and KaiC_RS-S413E/S414E (orange, single-chain for clarity) reveals that the coiled-coil in the phosphomimetic structure points outwards, with an angle of about 20° relative to the KaiC_RS coiled-coil. (b) The conformational change in the coiled-coil domain affects the dimer interface due to partner swaps with the opposite hexamer (see also Fig. 2). From an “outside perspective”: the C-terminal helix in KaiC_RS interacts with the right chain from the opposite hexamer, whereas in KaiC_RS-S413E/S414E the interaction is with the chain on the left. (c) Coiled-coil diagrams describe the heptad register shift that accompanies this structural rearrangement. (d) Based on the overlay of our structures, we propose the following model for the phosphorylation/dephosphorylation events. First, the phosphorylation cycle starts with the transfer of the γ-phosphate of ATP to the hydroxyl group of Ser414 (1; green arrow) in unphosphorylated KaiC_RS (green) or KaiC_RS-Δcoil (cyan). Secondly, pSer414 of KaiC_RS-Δcoil (purple, singly phosphorylated) moves away from the active site placing the hydroxyl group of Ser413 closer to the γ-phosphate of ATP for the second phosphorylation (2; purple arrow). Thirdly, the doubly phosphomimetic state (KaiC_RS-S413E/S414E, orange) reveals that the phosphoryl group of pSer414 moves back towards the active site for dephosphorylation (3; orange arrow). Lastly, we hypothesize that the indole group of Trp419 “pushes” pSer413 into the active site for the second dephosphorylation event (dashed arrow), in agreement with the slower dephosphorylation rate observed in the KaiC_RS-Δcoil construct (cf. Fig. 2d).

Extended Data Fig. 6. Effect of ATP-to-ADP ratio on KaiC_RS auto-dephosphorylation and the dependence of temperature and KaiB_RS binding on the ATPase activity of KaiC_RS.

(a) 10% SDS-PAGE gel of 3.5 μM KaiC_RS and 3.5 μM KaiB_RS in the presence of 4 mM ATP and 10 mM 2-phosphoenolpyruvate at 30 °C shows that the auto-phosphorylation cycle restarts upon regeneration of ATP by the addition of 2 U ml⁻¹ pyruvate kinase at the 24-hour time mark. (b) 10% SDS-PAGE gel of 3.5 μM KaiC_RS without (upper panel) and with (lower panel) 3.5 μM KaiB_RS in the presence of 4 mM ATP with an ATP-recycling system added from the beginning showing that under these conditions KaiB does not accelerate dephosphorylation. For gel source data in (a) and (b), see Supplementary Figure 2. (c) Representative curves for ADP production of KaiC_RS (3.5 μM) alone and (d) in the presence of KaiB_RS (3.5 μM) in 4 mM ATP measured by HPLC. The data were analysed as described in the Methods section and result in ATPase activities of 108 ± 10 day⁻¹ KaiC⁻¹ (with KaiB_RS = 176 ± 29 day⁻¹ KaiC⁻¹) at 20 °C, 163 ± 16 day⁻¹ KaiC⁻¹ (with KaiB_RS = 1052 ± 143 day⁻¹ KaiC⁻¹) at 25 °C, 208 ± 19 day⁻¹ KaiC⁻¹ (with KaiB_RS = 1557 ± 172 day⁻¹ KaiC⁻¹) at 30 °C, and 300 ± 21 day⁻¹ KaiC⁻¹ (with KaiB_RS = 2584 ± 245 day⁻¹ KaiC⁻¹) at 35 °C. The temperature coefficient, Q₁₀, was calculated using the data obtained at 25 °C and 35 °C and yields a value of ~ 1.9. The standard deviations of ATPase activity at each temperature (right panels) were obtained from three replicate measurements and data are presented as mean values ± s.d. (e) The comparison of ADP production of KaiC_RS in the absence (orange circles, data from panel (c)) and presence (orange diamonds, data from panel (d)) of KaiB_RS at 30 °C indicate a 7.5-fold increase in ATPase activity for the complex. The binding of KaiB_RS accelerates the ATPase activity of KaiC_RS in both the CI and CII domains (see also Extended Data Fig. 8b, c). (f) The SDS-PAGE gel of KaiC_RS (10 μg) and KaiB_RS (10 μg) shows that both proteins were purified to homogeneity and the measured ATPase activity is, therefore, not due to impurities. (g) ADP production of KaiB_RS in 4 mM ATP at 30 °C shows, as expected, no ATPase activity for KaiB_RS alone and confirms the increase in ATPase activity shown in panel (d) is due to complex formation. The standard deviation for the representative curves shown in panels (c-d, left), e, and g was set to 6% assuming the largest systematic error originates from the injector.

Extended Data Fig. 7. Dephosphorylation of KaiC_RS occurs via an ATP-synthase mechanism and the phosphoryl-transfer step is unaffected by the binding of KaiB_RS.

(a) Possible mechanisms how KaiB_RS could accelerate KaiC_RS dephosphorylation at nighttime. Binding of KaiB_RS on the CI_RS domain directly accelerates the phosphoryl transfer from pSer to bound ADP to generate transiently bound ATP. The cartoon represents the interface between two monomers in the CII_RS domain. (b) Autoradiograph of separation of ³²P-KaiC_RS at Ser413, transiently formed ³²P-ATP, and free ³²Pi via thin-layer chromatography (TLC) with 4 mM ADP at 30 °C, with the corresponding kinetics shown in (c) where gray circle, purple triangle, and cyan diamonds represent the relative concentrations of phosphorylated ³²P-KaiC_RS, ³²P-ATP, and free ³²Pi, respectively. (d) Comparison of transient ³²P-ATP formation and decay in the absence (open triangle) and presence (solid triangle) of KaiB_RS and free ³²P formation in the absence (open circles) and presence of KaiB_RS (solid circles). Faster decay of transient ³²P-ATP together with higher free ³²P production in the presence of KaiB_RS indicated that KaiB_RS accelerates hydrolysis in KaiC_RS. (e) SDS-PAGE gel (10%) of dephosphorylation of phosphorylated 3.5 μM KaiC_RS at Ser413 without (upper gel) and with (lower gel) 3.5 μM KaiB_RS in the presence of 4 mM ADP at 30 °C (for gel source data, see Supplementary Figure 2). (f) Densitometric analysis of data in panel (e) shows the decay of total KaiC_RS phosphorylation in the absence (gray circles) and presence (red diamonds) of KaiB_RS and yields rates of 11.5 ± 0.8 h⁻¹ and 11.0 ± 0.8 h⁻¹, respectively. This result indicates that binding of KaiB_RS does not accelerate the phosphoryl-transfer step in KaiC_RS.

Extended Data Fig. 8. Effect of KaiB_RS binding on ATPase activity and nucleotide exchange in the CII domain of KaiC_RS.

(a) Second possible mechanism to explain how KaiB_RS accelerates KaiC_RS dephosphorylation at nighttime: binding of KaiB_RS to the CI_RS domain could increase the hydrolysis rate in the CII_RS domain and, thereby, prevent the phosphoryl transfer back from transiently formed or external ATP back to serine residues. (b) Representative curves for ADP production of phosphorylated KaiC_RS with catalytic mutations in the CI domain (KaiC_RS-E62Q/E63Q, 3.5 μM) in the absence (dark green circles) and presence (light green circles) of 3.5 μM KaiB_RS at 30 °C with 4 mM ATP was quantified using HPLC. From these data an ATPase activity in the CII domain of 112 ± 8 day⁻¹ KaiC⁻¹ and 195 ± 16 day⁻¹ KaiC⁻¹ in the presence of KaiB_RS was determined. (c) Representative curves for ADP production measured by HPLC as in panel (b) of KaiC_RS but with catalytic mutations in the CII domain (KaiC_RS-E302Q/E303Q, 3.5 μM) in the absence (dark pink circles) and presence (light pink diamonds) of 3.5 μM KaiB_RS. The corresponding ATPase activities in the CI domain are 110 ± 12 day⁻¹ KaiC⁻¹ in the absence and 320 ± 22 day⁻¹ KaiC⁻¹ in the presence of KaiB_RS. (d) Representative curves for ADP production of KaiCI_RS-E62Q/E63Q (construct of only CI domain with catalytic mutations) in 4 mM ATP at 30 °C shows no ATPase activity indicating that Glu62 and Glu63 are the only two residues that are responsible for ATPase activity in CI domain of KaiC_RS and confirms the ATPase activity shown in panel b is due to ATPase activity in CII domain of KaiC_RS. The standard deviation for the representative curves shown in panels (b-d) was set to 6% assuming the largest systematic error originates from the injector. The experiments in panel (b-d) were performed in triplicate and ATPase rate given in the legend for panels (b) and (c) are presented as mean values ± s.d. (e) Third possible mechanism to explain how KaiB_RS accelerates KaiC_RS dephosphorylation at nighttime: binding of KaiB_RS to the CI_RS domain could promote faster nucleotide exchange in the CII_RS domain to displace transient ATP by ADP. (f) Time course of fluorescence intensity at 440 nm due to binding of mant-ATP to KaiC_RS-S413E bound with ATP in the absence (solid blue trace) and presence (red dotted trace) of KaiB_RS. KaiC_RS-S413E (3.5 μM) was pre-incubated with 3.5 μM KaiB_RS for 16 h at 20, 25, 30, and 35 °C in the presence of 50 μM ATP and an ATP-recycling system and then mixed with 250 μM mant-ATP. The observed exchange rates at each temperature are listed in the table (g). (h) Nucleotide exchange of KaiCI_RS (i.e., only CI_RS domain) cannot be measured since there is no tryptophan residue in close proximity of the nucleotide binding site. In summary, KaiB_RS accelerates KaiC_RS dephosphorylation by increasing the hydrolysis rate in the CI and CII domains and does not affect the nucleotide exchange rate. Representative traces are shown in (f) and (h) and the fitted parameters (g; mean ± s.d.) were obtained from three replicate measurements.

Extended Data Fig. 9. KaiB-KaiB interface in the KaiBC_RS complex affects the solvent accessibility into the active site of KaiC_RS-CI.

(a) Size-exclusion chromatography of KaiB_RS (blue) shows that it is monomeric in solution in contrast to KaiB_SE (gray), which elutes as a tetramer. Molecular-weight standards are shown above the chromatogram. (b) Structural comparison of KaiB_TE (gray, PDB 5jwq) and KaiB_RS (blue) when bound to their corresponding KaiC hexamers. The PISA software package determines that for the KaiBC_TE complex the interface between the KaiB_TE monomers is 255 Ų, whereas the average interface between KaiB_RS monomers is only 45 Ų in the KaiBC_RS complex. (c) To understand how KaiB_RS binding to KaiC_RS-CI domain increases the hydrolysis rate, we investigated whether conformational changes modulated substrate access to the active site. The CAVER software was used to calculate tunnels (gray mesh) leading into the active site of the CI domain of KaiC_RS-S413E/S414E alone (orange) and the KaiC_RS-S413E/S414E:KaiB_RS complex (yellow:blue) with varying probe radii. In both structures, the active site was occupied by ADP:Mg²⁺ (sticks and green sphere, respectively). The crystal structure of bovine F1-ATPase in complex with a transition-state analogue (PDB 1w0j, chain D) was used as a reference to determine the position of the catalytic water molecule in the active site (shown as a red sphere). The calculated tunnels connect bulk solvent to the catalytic water when KaiB_RS is bound for probe radii larger than the default value of 0.9 Å, but never in its absence. These results suggest that KaiB_RS facilitates the access of water into the active site of KaiC_RS-CI via long-range conformational changes and thus enhances ATP hydrolysis.

Extended Data Fig. 10. KaiB_RS binds preferentially to the post-hydrolysis state of KaiC_RS and affects its stability.

(a) Size-exclusion chromatography of 50 μM KaiC_RS-CI (CI_RS domain) in the absence (black line, hexamer) and presence (gray line, monomer) of 50 μM KaiB_RS in 1 mM ATP buffer. (b) Size-exclusion chromatography of 50 μM KaiC_RS-Δcoil in the presence of 50 μM KaiB_RS (purple). The reference sample (50 μM KaiC_RS-Δcoil) is a hexamer in solution (cyan) and after the addition of 50 μM KaiB_RS the mixture was incubated at 30 °C for 3.5 h (purple) before running the samples again on a Superdex-200 10/300 GL column at 4 °C. These data show that binding of KaiB_RS results in (i) disassembly of the hexameric KaiC_RS-Δcoil structure into its monomers and (ii) aggregation as detected by the elution in the void volume of the column (v₀). (c) Thermal denaturation profiles for KaiC_RS-S413E/S414E in the presence of 1 mM ADP are shown from dark to light blue for increasing concentrations of KaiB_RS (between 0 – 4 μM). The black line represents KaiB_RS alone (15 μM), which shows no fluorescence signal as it does not bind to SYPRO Orange due to a lack of a hydrophobic core. The T_m decreases upon the addition of KaiB_RS, indicating that binding of KaiB_RS destabilizes the KaiC_RS dodecamer. Likely due to loosening up of interface and the KaiC_RS structure, thereby allowing for the formation of a tunnel that connects bulk solvent to the position of the hydrolytic water in the active site (see Extended Data Fig. 9). (d) SDS-PAGE analysis showing the control experiment for pull-down assay. The first four lanes after the molecular weight marker are KaiB_RS-Tag samples (red arrow) and show that KaiB_RS-Tag binds tightly to the column. The last four lanes are control pull-down assay experiments for KaiC_RS (green arrow) and show that KaiC_RS alone is unable to bind to the column. The lanes represent the initial sample used in pull-down assay (Before), flow-through after loading sample onto the column (FT), flow-through after washing the column three times with the binding buffer (Wash #3), and sample after elution with imidazole (Elute). (e) SDS-PAGE analysis of pull-down assay to measure the complex formation between KaiB_RS-Tag and wild-type KaiC_RS, KaiC_RS-S413E/S414E, or KaiC_RS-S413A/S414A in the presence of 4 mM ADP or ATP (with an ATP-recycling system). For gel source data, see Supplementary Figure 2. (f) Percentage of wild-type KaiC_RS bound to KaiB_RS-Tag protein for different ATP-to-ADP ratios (4 mM total nucleotide concentration) at 25 °C as measured from pull-down assays. (g) Fluorescence anisotropy at 30 °C of unlabeled KaiB_RS competitively replacing the fluorophore-labeled KaiB_RS (KaiB_RS-6IAF) from KaiC_RS-S413E/S414E in the presence of 4 mM ADP (red circles, K_D value of 0.79 ± 0.06 μM) and 4 mM ATP with an ATP-recycling system (orange triangles). In the latter experiment no change in anisotropy is observed, indicating that only a small fraction of KaiB_RS-6IAF is bound under these conditions. The average anisotropy and standard error were calculated from ten replicate measurements. (h) The mant-ATPγS or mant-ADP release is shown as bar graphs with observed rates of 4.8 ± 0.2 h⁻¹ and 43.6 ± 3.0 h⁻¹ for mant-ATPγS and mant-ADP releasing from KaiC_RS-S413E/S414E, respectively, and 21.0 ± 3.0 h⁻¹ and 65 ± 15 h⁻¹ for mant-ATPγS and mant-ADP releasing from KaiC_RS-S413A/S414A, respectively. The result shows that CII domain of KaiC_RS prefers binding of ATP over ADP. Experiments in panels (f) and (h) were performed in triplicate and data are presented as mean values ± s.d.