Abstract
Circadian rhythms are observed in most organisms on earth and are known to play a major role in successful adaptation to the 24-h cycling environment. Circadian phenotypes are characterized by a free-running period that is observed in constant conditions and an entrained phase that is observed in cyclic conditions. Thus, the relationship between the free-running period and phase of entrainment is of interest. A popular simple rule has been that the entrained phase is the expression of the period in a cycling environment (i.e., that a short period causes an advanced phase and a long period causes a delayed phase). However, there are experimental data that are not explained by this simple relationship, and no systematic study has been done to explore all possible period-phase relationships. Here, we show the existence of stable period-phase relationships that are exceptions to this rule. First, we analyzed period-phase relationships using populations with different degrees of genome complexity. Second, we generated isogenic F1 populations by crossing 14 classical period mutants to the same female and analyzed 2 populations with a short period/delayed phase and a long period/advanced phase. Third, we generated a mathematical model to account for such variable relationships between period and phase. Our analyses support the view that the circadian period of an organism is not the only predictor of the entrained phase.