FIG. 4.

Overexpression of DBT^S and DBT^L proteins in otherwise wild-type flies produces the same circadian period for locomotor activity as in the original mutant flies, while overexpression of DBT^K/R produces high levels of arrhythmicity and some long periods. Shown are representative actograms and periodograms of tim-GAL4>UAS-DBT-MYC flies (the DBT genotype is indicated to the left of the periodogram) as well as those of a sibling control fly that carried the balancer rather than the responder gene (tim-GAL4>TM3; denoted “DBT^TM3”). The actogram plots the magnitudes of successive 0.5-h activity bouts. There are two 24-h intervals shown on each line, and each day plotted on the right is plotted again on the left side of the line underneath (a double plot). The periodogram to the left of each actogram plots the amplitude of the oscillation for each test period applied to the activity record. The period which produces the highest amplitude is taken as the best period for the data set. Periods for which the amplitude lies above the sloping line are statistically significant with a P value of <0.01. The responder inserts were tested in the hemizygous condition with the tim-GAL4 driver (tim-GAL4/+; UAS-DBT/+), as were all the flies listed in Table 1. Representative actograms and periodograms are shown for both an arrhythmic and a long-period tim-GAL4>UAS-DBT^K/R fly.