COMMENTARY

Orderly wheels of the cyanobacterial clock

Michael J. Rust1 Department of Molecular Genetics and Cell Biology, Institute for Genomics and Systems Biology, University of Chicago, Chicago, IL 60637

ircadian rhythms, self-sustaining at CII residues near the monomer– workings of the . Using the oscillations in an organism’s monomer interfaces, notably Ser431 and more tractable thermophilic version of the C behavior that repeat roughly Thr432, and does so in a specific ordered system, the authors are able to separate once per day, have fascinated pattern throughout the course of a day the domains of KaiC and introduce biologists for hundreds of years (1, 2). Part (10–12). At times corresponding to the mutations to control the oligomeric state of this fascination comes from the fact subjective morning, KaiA interacts with a of the proteins. This allows them to make that the existence of intrinsic biological sequence at the C terminus of CII and two important discoveries regarding the rhythms implies that an organism carries promotes activity, leading to phos- previously structurally murky KaiB·KaiC within it a model of how it expects the phorylation (13, 14). KaiB, a protein with interaction: first, that KaiB interacts with external world to behave—a plant that a thioredoxin-like fold, is able to interact the CI ATPase domain of KaiC; and moves its leaves in anticipation of the sun with appropriately phosphorylated KaiC second, that the binding of KaiB involves even when grown in constant conditions near subjective dusk, and these KaiB·KaiC some disruption of the ring structure must have some hard-wired knowledge of complexes inhibit KaiA, allowing de- of KaiC. the rotation of the Earth. Fundamentally, to occur and the cycle to Previously the weight of evidence had these predictive biological rhythms must repeat (11, 12). suggested that KaiB made contact with originate mechanistically from carefully This verbal description of the events in the CII domain of KaiC. This ambiguity regulated molecular interactions that the circadian cycle does not itself give probably resulted from two sources: the generate oscillations. In PNAS, Chang a satisfactory account of what the period structural similarity of CI and CII, and et al. (3) present a series of cleverly de- of the clock should be, how the system the requirement for CII phosphorylation, signed structural and biochemical experi- depends on the concentration of the particularly on the key Ser431 residue, for ments that shed light on the inner components, or, indeed, whether stable KaiB interaction (11). The implication of workings of the best-defined biological oscillations should occur at all. Unlike the a structure where KaiB instead interacts clock, the KaiABC protein oscillator. familiar oscillations of a swinging bob in with the CI domain is that that the re- Although it may not be surprising that a pendulum clock, oscillating chemical lationship between CII phosphorylation complex multicellular organisms like fruit reactions are fundamentally nonlinear. and KaiB binding cannot be as simple as flies and humans are capable of creating That is, a cycle of elementary reactions, the direct recognition of a phosphorylated predictive models of their external envi- one leading to the next, that ultimately peptide. Instead, the emerging picture is ronments, autonomous circadian rhythms brings the system back to its starting state, that information about the phosphoryla- can frequently be found in free-living can never itself produce stable oscillations. tion status of CII is transmitted through single cells as well (4). The simplest To drive oscillations, there must also be KaiC to CI via allosteric ring–ring stacking genetic model organism we have today feedback interactions in the reaction net- interactions (17). that exhibits circadian rhythms is the work, strong enough and appropriately The authors find that KaiB cannot in- cyanobacterium elongatus, arranged to enforce synchrony and push teract with a tightly assembled CI ring and, a photosynthetic microbe with a clock that the system away from steady state. in the case of the truncated CI protein, controls pronounced oscillations in A significant amount of mathematical KaiB stably interacts with free CI mono- expression (5). The heart of this cyano- modeling work has been done on the mers that have dissociated from hexamers. bacterial clock is formed by the interaction KaiABC system in an attempt to define In full-length KaiC, ATP binding in the CI of three proteins, KaiA, KaiB, and KaiC. the critical feedback interactions, often domain is the principal force that assem- Following methods pioneered by Takao identifying the phosphorylation-dependent bles the protein into active multimeric Kondo and coworkers (6), a ∼24-h rhythm sequestration of KaiA into protein–pro- particles (9). This information suggests in KaiC phosphorylation can be recon- tein complexes as a crucial event (12, 15, that destabilization of KaiC’s hexameric stituted using only these purified proteins 16). These models have identified core structure is required for KaiB binding, (Fig. 1 A and B). principles of how the clock functions, but helping to explain why exchange of mono- This remarkable finding essentially they leave many of the conspicuous bio- mers between KaiC particles occurs at the reduced the puzzle of how a biological chemical features of the Kai proteins same time in the circadian cycle when rhythm is generated to the problem of unexplained. For example, KaiC is not a KaiB interaction is strongest (12, 18). understanding the chemistry of three traditional protein kinase, but instead is In summary, the work of Chang et al. known components. KaiC is an unusual an ATPase that has expanded its “skill set” promotes an understanding of the struc- relative of the AAA+ superfamily of to include . Is there ture of KaiC, where distinct functions are with very slow enzymatic activity a reason that this phosphorylation-based divided between the CI and CII domains. (7, 8). The protein forms hexameric rings clock was built out of a hexameric AT- In this model, the daily oscillation of the fl in the presence of ATP in which each Pase? Why does KaiC have two catalytic system re ects an oscillation of enzymatic – KaiC monomer consists of two homolo- domains, CI and CII? What is the func- activity and protein protein interaction gous catalytic domains, CI and CII (9). tional role of CI in the ? between CI and CII. During the subjective The KaiC hexamer thus appears as a How are input and output signals trans- “double doughnut,” with the CII ring atop duced into and out of the core oscillator the CI ring. The CI domain is an ATPase, without impairing rhythmicity? Author contributions: M.J.R. wrote the paper. but the CII domain has acquired addi- Chang et al. present a structural and The author declares no conflict of interest. tional phosphotransferase activities: it can biochemical analysis of the physical in- See companion article 10.1073/pnas.1211508109. phosphorylate and dephosphorylate itself teraction of the Kai proteins and the inner 1E-mail: [email protected].

www.pnas.org/cgi/doi/10.1073/pnas.1214901109 PNAS Early Edition | 1of2 Downloaded by guest on September 26, 2021 AB complex with KaiB on CI can be viewed as clock output signaling feeding back on KaiA KaiB KaiC clock input to produce free-running rhythms (Fig. 1C). 80 These data raise many important ques- tions whose answers will likely be critical 60 +ATP to understanding the principles underlying +Mg 40 the biochemical origin of circadian

%P-KaiC rhythms. Because allosteric coupling be- 20 tween CII phosphorylation and CI is now clearly seen as a determinant of KaiB 0 0244872 binding, what is the precise nature of the time (hours) interaction between the two domains? ATP/ADP redox Because disruption of the CI ring is im- C portant for KaiB binding, and the CI ring is held together by ATP, what is the INPUT active functional role of the CI ATPase activity KaiA in the clock, already known to be tem- perature insensitive and linked to clock CII autokinase period? Along these lines, why does the P P assembly of KaiB·KaiC complexes take ring-ring P P hours, even for KaiB kept in a binding- stacking competent dimeric state? The slow kinet- ics of this process are likely fundamental CI ATPase to circadian timing, and our improved SasA KaiB inactive understanding of the nature of the KaiA KaiB·KaiC complex will shed light on this OUTPUT biophysical problem. There are many ways to make a Fig. 1. Roles of two KaiC domains in the cyanobacterial clock. (A) Circadian rhythms can be reconstituted in vitro using a mixture of purified KaiA, KaiB, and KaiC in the presence of ATP. (B) Example in vitro rhythm chemical oscillator, as is clear from non- in KaiC phosphorylation, reaction incubated at 30 °C. (C) Known biochemical interactions mediating input, biological examples, such as the Belousov- output, and oscillations within the KaiABC system. Active KaiA binds to C-terminal tails of KaiC, switching Zhabotinsky reaction (2), and examples CII from a mode to a kinase mode. This process can be modulated by the ATP/ADP ratio from synthetic biology, such as the re- or redox-dependent quinone interactions with KaiA. Multisite phosphorylation of CII on Ser431 and Thr432 pressilator (21). Circadian clocks are – induces ring ring stacking interactions with CI, permitting KaiB to bind to CI. KaiA is sequestered in these a special class of oscillator that have pre- KaiB complexes on CI, constituting a negative feedback loop on KaiC phosphorylation. The kinase SasA also interacts with the CI domain, transducing transcriptional output from the circadian clock. sumably been shaped by natural selection to perform reliably and remain synchro- day, KaiA acts on the CII ring in its role input and output functions. The best- nized with the external environment under as an activator of phosphorylation, understood input mechanisms involve a wide variety of conditions. The existence fi whereas during the subjective night, KaiA modulation of CII activity by the ATP/ of the puri ed KaiABC system, which relocates to inhibitory KaiA·KaiB·KaiC ADP ratio and redox signals that act on shares many of the properties of circadian complexes on CI, permitting the clock to KaiA, the activator of CII phosphorylation clocks in intact organisms, gives us the dephosphorylate. (19, 20). In contrast, the output histidine opportunity to see clearly the structure of This picture also suggests a rough kinase SasA interacts with the CI domain. the reaction networks that living organ- division of KaiC into rings responsible for In this vein, the sequestration of KaiA in isms use to measure time.

1. McClung CR (2006) Plant circadian rhythms. Plant Cell 9. Hayashi F, et al. (2004) Roles of two ATPase-motif-con- 15. Clodong S, et al. (2007) Functioning and robustness of 18(4):792–803. taining domains in cyanobacterial circadian clock pro- a bacterial circadian clock. Mol Syst Biol 3:90. 2.WinfreeAT(2000)The Geometry of Biological Time tein KaiC. J Biol Chem 279(50):52331–52337. 16. van Zon JS, Lubensky DK, Altena PRH, ten Wolde PR (Springer-Verlag, New York), 2nd Ed, p 777. 10. Xu Y, et al. (2004) Identification of key phosphoryla- (2007) An allosteric model of circadian KaiC phosphor- 3. Chang Y-G, Tseng R, Kuo N-W, LiWang A (2012) Rhythmic tion sites in the circadian clock protein KaiC by crystal- ylation. Proc Natl Acad Sci USA 104(18):7420–7425. – ring ring stacking drives the circadian oscillator clockwise. lographic and mutagenetic analyses. Proc Natl Acad Sci 17. Chang YG, Kuo NW, Tseng R, LiWang A (2011) Flexibil- Proc Natl Acad Sci USA, 10.1073/pnas.1211508109. USA 101(38):13933–13938. ity of the C-terminal, or CII, ring of KaiC governs the 4. Mihalcescu I, Hsing W, Leibler S (2004) Resilient 11. Nishiwaki T, et al. (2007) A sequential program of dual rhythm of the circadian clock of . Proc circadian oscillator revealed in individual cyanobac- phosphorylation of KaiC as a basis for circadian rhythm – teria. Nature 430(6995):81–85. Natl Acad Sci USA 108(35):14431 14436. in cyanobacteria. EMBO J 26(17):4029–4037. 5. Kondo T, et al. (1993) Circadian rhythms in prokar- 18. Ito H, et al. (2007) Autonomous synchronization of the 12. Rust MJ, Markson JS, Lane WS, Fisher DS, O’Shea EK yotes: as a reporter of circadian gene ex- circadian KaiC phosphorylation rhythm. Nat Struct Mol (2007) Ordered phosphorylation governs oscillation pression in cyanobacteria. Proc Natl Acad Sci USA 90 Biol 14(11):1084–1088. of a three-protein circadian clock. Science 318(5851): (12):5672–5676. 19. Wood TL, et al. (2010) The KaiA protein of the cyano- 809–812. 6. Nakajima M, et al. (2005) Reconstitution of circadian bacterial circadian oscillator is modulated by a redox-ac- 13. Iwasaki H, Nishiwaki T, Kitayama Y, Nakajima M, oscillation of cyanobacterial KaiC phosphorylation in tive cofactor. Proc Natl Acad Sci USA 107(13):5804–5809. – Kondo T (2002) KaiA-stimulated KaiC phosphorylation vitro. Science 308(5720):414 415. 20. Rust MJ, Golden SS, O’Shea EK (2011) Light-driven changes 7. Leipe DD, Aravind L, Grishin NV, Koonin EV (2000) The in circadian timing loops in cyanobacteria. Proc Natl in energy metabolism directly entrain the cyanobac- bacterial replicative helicase DnaB evolved from a RecA Acad Sci USA 99(24):15788–15793. terial circadian oscillator. Science 331(6014):220–223. duplication. Res 10(1):5–16. 14. Kim YI, Dong G, Carruthers CW, Jr., Golden SS, 21. Elowitz MB, Leibler S (2000) A synthetic oscillatory net- 8. Terauchi K, et al. (2007) ATPase activity of KaiC deter- LiWang A (2008) The day/night switch in KaiC, a central work of transcriptional regulators. Nature 403(6767): mines the basic timing for circadian clock of cyanobac- oscillator component of the circadian clock of cyano- – teria. Proc Natl Acad Sci USA 104(41):16377–16381. bacteria. Proc Natl Acad Sci USA 105(35):12825–12830. 335 338.

2of2 | www.pnas.org/cgi/doi/10.1073/pnas.1214901109 Rust Downloaded by guest on September 26, 2021