Circadian entrainment by different daylengths: The roles of dawn and dusk

In the early days of circadian rhythms research there were two competing views on entrainment by light: through parametric action on the velocity of the endogenous cycle (Aschoff) versus non-parametric discrete phase shifts elicited by the lights-on and –off transitions (Pittendrigh). Although the phase shift model became widely accepted, straightforward experimental tests of the alternatives are lacking. In recent years we performed a series of studies in mice to address this issue, using protocols with single and double light pulses of different duration as well as single lights-on and lights-off signals. The results reveal no evidence for a specific role of the transitions per se. All features of the circadian system can be explained by velocity changes in response to light, a response that is reduced under prolonged illumination and restored during prolonged darkness. This arrangement allows robust entrainment when the signal is noisy and behavioural withdrawal from light adds to its variance. It leads to stable timing in species never observing dawn and dusk. It even keeps the clock entrained when mice must work for their food and show diurnal (shiftwork) activity. A model for a neuronal pacer network in the SCN shows exactly these properties and predicts seasonal adaptation to daylength