Splitting of the circadian rhythm of activity in hamsters: Effects of exposure to constant darkness and subsequent re-exposure to constant light

The circadian rhythm of wheel running behavior was observed to dissociate into two distinct components (i.e. 'split') within 30 to 110 days in 56% of male hamsters exposed to constant light (Figs. 1-2). Splitting was abolished in all 16 animals that were transferred from constant light (LL) to constant darkness (DD) within 1-4 days of DD, and the components of the re-fused activity rhythm assumed a phase relationship that is characteristic of hamsters maintained in DD (Figs. 3-5). Re-fusion of the split activity rhythm was accompanied by a change in period (τ); in 14 animals τ increased while in the other 2 animals τ decreased after transfer to DD. After 10-30 days in DD, the hamsters were transferred back into LL at various time points throughout the circadian cycle. A few of these animals went through two or three LL to DD to LL transitions. The effect of re-exposure to LL was dependent on the phase relationship between the transition into LL and the activity rhythm. A rapid (i.e. 1-4 days) induction of splitting was observed in 7 of 9 cases when hamsters were transferred into LL 4-5 h after the onset of activity (Fig. 5). In the other 2 animals, the activity pattern was ultradian or aperiodic for 20 to 50 days before eventually coalescing into a split activity pattern. In contrast, transfer of animals (n = 13) from DD to LL at other circadian times did not result in the rapid induction of splitting and the activity rhythm continued to free-run with a single bout of activity (Fig. 5). Importantly, a transfer from DD to LL 4-5 h after the onset of activity did not induce splitting if the hamsters had not shown a split activity rhythm during a previous exposure to LL (n=10; Fig. 6). These studies indicate that transfer of split hamsters from LL to DD results in the rapid re-establishment of the normal phase relationship between the two circadian oscillators which underlie the two components of activity during splitting. In addition, there appears to be a history-dependent effect of splitting which renders the circadian system susceptible to becoming split again. The rapid re-initiation of the split condition upon transfer from DD to LL at only a specific circadian time is discussed in terms of the phase response curve for this species.