Model for circadian oscillations in the Drosophila PER protein

Post-translational modification of PER is also involved in the mechanism of its circadian oscillations (Zwiebel et al., 1991). Thus, a recent study (Edery et al., 1994) showed that PER becomes phosphorylated. 

Another important study (Zeng et al., 1994) demonstrated that the overexpression of PER in the Drosophila eyes represses per transcription and suppresses circadian rhythmicity in these cells, without affecting circadian PER oscillations in other per-expressing cells in the brain, or the circadian rhythm in locomotor activity. This work also shows that the action of PER on transcription is intracellular, and suggests that each per-expressing cell contains an autonomous oscillator of which the per feedback loop is a component (Zeng et al., 1994). 

An alternative, more detailed model for circadian PER oscillations in Drosophila has been developed independently be Abbot et al. (1995) (M. Rosbasch, personal communication results of that study were first presented at the 4th conference of the Society for Research on Biological Rhythms held in May 1994 in Amelia Island, Florida). That model, which is also based on the negative feedback exerted by PER on per transcription, takes into account a larger number of phosphorylated residues and focuses on the role of PER phosphorylation in delaying the entry of the protein into the nucleus (Curtin et al. 1995). 

Based on the experimental observations summarized above, the model for the circadian variation in PER is schematized in fig. 11.6. We assume that per mRNA, whose cytosolic concentration is denoted by M, is synthesized in the nucleus and rapidly transfers to the cytosol, where it accumulates at a maximum rate v there it is degraded enzymi-cally, in a Michaelian manner, at a maximum rate v . 

The rate of synthesis of PER, proportional to M, is characterized by an apparent first-order rate eonstant k. To take into account the fact that PER is multiply phosphorylated (Edery et al, 1994), and to keep the model as simple as possible (the precise number of phosphorylated residues is still unknown), we eonsider only three states of the protein unphosphorylated (Pq), monophosphorylated (Pi) and bisphosphorylat-ed (P2). The model eould readily be extended to include a larger number of phosphorylated residues such an extension would, however, unnecessarily complicate the model without altering significantly its dynamic behaviour. 

---