Although the suprachiasmatic nucleus (SCN) is widely regarded as the main circadian pacemaker in the mammalian brain, evidence for a rhythm of firing activity based on extensive single unit recording in intact animals has been reported only recently (Saeb-Parsy & Dyball, 2003). There were significant differences in mean spike frequency at different times of the day but the differences were not clearly evident unless activity was averaged over arbitrary time epochs. The use of such averaging techniques represents a substantial of loss of information. We have found that parameters that measure the irregularity of spike activity of cells in the SCN show more obvious changes over the day/night cycle than mean spike frequency.

Using a ventral surgical approach, extracellular single unit activity was recorded from SCN cells in urethane (1.2 g/kg I.P.) anaesthetised rats in vivo. All rats were killed humanely at the end of the experiments. Conventional variance measures were not used to quantify the variability of inter-spike intervals because of the non-Gaussian distribution of the intervals. To measure the irregularity of spike activity, we used the entropy (a widely used measure of disorder) of the log interval distribution. Such a measure quantifies the variability of intervals and thus the coding capacity of each spike. The entropy was plotted against the respective zeitgeber and showed an oscillation that could be described using a second order Fourier expansion with a peak in the mid-dark period.

Neurones within the SCN showed a significant rhythm in entropy (n = 166; P < 0.001) and there was a similar rhythm in the activity of cells recorded from perinuclear zone (PSCN) of the suprachiasmatic nucleus (n = 209; P = 0.0370). SCN and PSCN cells with caudal projections showed a significant rhythm (n = 78; P = 0.006) with a peak in the mid-light period whereas those without a caudal projection showed a rhythm with a peak in the mid-dark period (n = 297; P = 0.001). There was no evidence for a rhythm of activity for hypothalamic more than 1 mm distant from the SCN.

Thus measures of SCN cell activity in addition to mean spike frequency show a daily rhythm that is more obvious. Cells within the SCN show a clearer rhythm than cells in the PSCN and different populations of cells within the SCN show rhythms that appear to be different from each other.

This work was supported by Merck, Sharp, and Dohme and the James Baird fund.