Abstract
The 1st circadian "clock" gene identified was the X-linked period (per) gene in Drosophila melanogaster. In the pioneering initial report, Konopka and Benzer (1971) characterized 3 alleles of per that shortened (per (S); approximately 19 h), lengthened (per (L); approximately 29 h), or abolished (per (0)) circadian behavioral rhythms. They also showed that transheterozygotes carrying the per (S) and per (L) mutations exhibit robust behavioral rhythms with nearly normal periods of approximately 23 h, highlighting the semidominant nature of many clock mutants. In this study, per (0) flies bearing a doubly mutated per transgene that carries both the per (S) and per (L) alleles (per (0); per (S/L)) were analyzed for behavioral and molecular rhythms. Unlike singly mutated versions, the per (0);per ( S/L) transgenic flies are arrhythmic in constant dark conditions and exhibit little, if any, entrainment to daily light-dark cycles. In a wildtype per (+) background, expression of per ( S/L) abolishes behavioral rhythms, indicating that it functions in a transdominant negative fashion. Biochemical analysis of head extracts revealed that only hyperphosphorylated isoforms of the PERS/L protein are detected throughout a daily cycle, and the levels remain constant. Intriguingly, little if any PERS/L is observed in key pacemaker neurons that control daily activity rhythms, consistent with the notion that hyperphosphorylated isoforms of PER are unstable. Nonetheless, PERS/L is detected in ectopic cells in the brain, in which it exhibits an unusual localization, mainly staining the periphery of the nucleus. These results suggest that posttranslational mechanisms play a key role in limiting the accumulation of PER to specific cells. On a broader scope, our results indicate that the semidominant effects of period-altering alleles observed in trans are not necessarily preserved in the cis-configuration and that novel phenotypes can emerge.
