Coté G G, Brody S
J Theor Biol. 1986 Aug 21;121(4):487-503. doi: 10.1016/s0022-5193(86)80104-7.
Mutations at the per (period) locus of Drosophila melanogaster affect the period of its circadian rhythms. An analysis of published data on strains having duplications and deletions of the per locus indicates that the period is a logarithmic function of the level of the per gene product. The analysis also indicates that period is relatively insensitive to the level of that gene product; a presumed 300% increase in the gene product level produces only a 4.6% decrease in the period. The period of a strain transformed with per+ DNA conforms to the same logarithmic relationship if the level of mRNA in the transformant, which is one-tenth that in the wild type, is considered equivalent to a gene dosage one-tenth the wild type dose of 2, or 0.2. The periods of strains having various doses of mutant per alleles which shorten (pers) or lengthen (per1) the period can be fitted to the same logarithmic function. The analysis may provide an explanation for the partial dominance of pers over per+ and the dominance of per+ over per1, since it suggests that the per1 gene product is nearly inactive while the pers gene product is more than 34 times as active as the wild type product. Analysis of periods of strains heterozygous at the per locus suggests that the per gene product may be a multimeric protein. Three possible roles for the per gene product in circadian rhythmicity are discussed, including a role in synchronizing rhythm-producing cells.
