Circadian Clock Properties and Their Relationships as a Function of Free-Running Period in Drosophila melanogaster

Abstract

The stability of circadian clock mechanisms under cyclic environments contributes to increased Darwinian fitness by accurately timing daily behavior and physiology. Earlier studies on biological clocks speculated that the timing of behavior and its accuracy are determined by the intrinsic period (τ) of the circadian clock under constant conditions, its stability, the period of the external cycle (T), and resetting of the clock by environmental time cues. However, most of these previous studies suffered from certain limitations, the major ones being a narrow range of examined τ values and a non-uniformity in the genetic background across the individuals tested. We present data that rigorously test the following hypotheses by employing Drosophila melanogaster fruit flies with τ ranging from 17 to 30 h in a uniform genetic background. We tested whether 1) precision (day-to-day stability of τ) is greater for clocks with τ close to 24 h; 2) accuracy (i.e., day-to-day stability of the phase relationship (ψ), where ψ is the duration between a phase of the rhythm and a phase of the external cycle) is greater for clocks with τ close to 24 h; 3) Ψ is delayed with an increase in τ; and 4) Ψ becomes more advanced with an increase in length of zeitgeber cycle (T). We show that precision is not always maximum for ~24-h clocks, but that accuracy is greatest when τ approximates T. Further, flies exhibit a delayed phase relationship with increasing τ and an advanced phase relationship under long T-cycles as compared with shorter T-cycles. We also describe relationships between activity and rest durations and how our observations fit predictions from models of circadian entrainment. Overall, we confirm that accuracy and phase of entrained rhythm are governed by both intrinsic clock period and the length of the external cycle; however, we find that the relationship between intrinsic period and precision does not fit previous predictions.