Fast or slow? Clock readout sets internal periods

Abstract

Circadian clocks are essential for the function of a wide array of organisms, from cyanobacteria to humans. Despite decades of productive study of these systems, mysteries remain, including why many biological clocks have non-24-hour internal periods. We take a new approach to circadian clocks by focusing on downstream readout of the clocks state to answer physiologically relevant questions, such as when will the sun rise? Using this framework, we show that systematic errors arising from sunrise and sunset prediction can be compensated by having non-24 hour internal periods. We show that this prediction holds in models of cyanobacterial circadian clocks. Finally, we predict latitude-dependant qualitative changes in circadian clock structure and the performance of different clock phenotypes in common laboratory experimental setups.