In vivo studies reported that cholinergic agents affect mammalian circadian rhythmicity. To study phase resetting properties of cholinergic compounds more directly, we carried out experiments in rat suprachiasmatic nuclei slices. Compounds were added to the perfusate for 1 h at specific phases of the circadian cycle. On the following day, the time of peak neuronal activity, a measure of the phase of the endogenous circadian pacemaker, was assessed by means of extracellular recording in the suprachiasmatic nuclei. The peak of neuronal activity occurred at circadian time 5.8 +/- 0.7 (mean +/- 95% confidence limits) in the control slice (circadian time 0: lights-on). Ten-micromolar carbachol had no effect on the phase of the circadian rhythm when given at circadian times 6 and 15, while at circadian time 21 a phase advance of one hour was observed. By contrast, 10 microM nicotine significantly phase advanced (> 1 h) the neuronal circadian rhythm at all but one experimental circadian phase. The circadian times of maximal nicotinic phase advances were 15 (+2.6 h) and 21 (+2.8 h). A concentration response curve for nicotine was generated and pharmacological blocking experiments were performed at circadian time 15. The estimated maximum response of nicotine was 3.4 h, and the estimated concentration for half maximal response was 5 microM. The Hill coefficient (= 1.08) indicated that the effects of nicotine may be explained by a single receptor occupancy model. Mecamylamine (20 microM) almost completely antagonized the nicotinic phase-advances, whereas tetrodotoxin (1 microM) or high Mg2+ (10 mM) did not significantly attenuate the nicotinic phase-advances.(ABSTRACT TRUNCATED AT 250 WORDS)