A hypothetical extension (left to right) of KaiA makes it slim enough to fit inside the KaiC ring.


Jimin Wang (Yale University, New Haven, CT) suggests that an ancient circadian clock rotates much like a hand on an analogue clock. His analysis of cyanobacterial clock protein structures reveals a similarity with the F1-ATPase rotary motor.

The cyanobacterial clock is controlled by the KaiA, KaiB, and KaiC proteins, which form a complex at night that falls apart in the day. Mutations causing more stable complexes correspond to a longer periodicity. But little is known about how the timing of complex formation is controlled. Wang analyzed the recently solved structures of the Kai proteins to suggest a mechanism.

When KaiC is ATP-bound, the Kai complex is stable. But when ATP is bumped off by autophosphorylation near the ATP-binding site, the complex falls apart. Autophosphorylation is stimulated in vitro by KaiA.

Wang realized that the ATP-binding domains of KaiC hexamers resemble the F1-ATPase ring. He proposes that KaiA fits inside the ring much like the γδ subunits fit inside the F1-ATPase ring. Based on previous structures, KaiA dimers are too large to fit inside the ring. But Wang proposes that KaiA dimers must first be activated by the extension of helical domains. This proposed extension makes KaiA dimers resemble γδ subunits. The need for KaiA activation would also explain the two-hour delay between the rise in KaiA and KaiC levels at dusk and complex formation. The stimulus for such KaiA activation is unclear.

In Wang's model, KaiA is expected to contact at most two KaiC monomers at a time and stimulate their autokinase activity. Autophosphorylation would both displace ATP and provide energy for the rotation of KaiA to new KaiC subunits. This cycle would repeat until the KaiC hexamer lacks ATP completely and the complex falls apart.

The rotation of KaiA might be hindered by KaiB, which Wang predicts bridges KaiA and KaiC below the ring. This suggestion fits with previous in vitro data showing that KaiB slows KaiC autophosphorylation. In fact, says Wang, “the autophosphorylation rate of KaiC is very slow, about three to four hours per [dimer].” Thus, one round of KaiA rotation should be approximately equivalent to the span of a night, or one rotation of an average wall clock.


Wang, J.