During natural sleep and anesthesia, rhythmic hypo- and hyperpolarizations alternate in cortical pyramidal cells and are reflected as slow (<1 Hz) cortical rhythm at the level of the electroencephalogram (EEG). Membrane potential changes in pyramidal neurons were initially attributed to the rhythmic fluctuation of the cholinergic input as the basal forebrain (BF) neurons fire in synchrony with cortical waves, but a more recent proposal suggested that the slow rhythm was of cortical origin. In the present experiments, interaction between the cortex and the BF was examined in urethane-anesthetized rats. BF neuronal activity was inhibited by local infusion of lidocaine into the substantia innominata in one group of rats, while in another group, the slow cortical rhythm was blocked by inducing spreading depression (SD) in the cortex. Slow cortical rhythm persisted after unilateral lidocaine injection, but rhythmic firing in BF neurons disappeared following SD induction. These findings support the view that slow cortical rhythm is generated in the cortex and transmitted to the BF through descending fibers. According to anatomical data, these fibers can excite cholinergic cells only indirectly as they terminate on non-cholinergic neurons. Thus, timing of activity changes in BF neurons during the slow cortical rhythm might give some clue regarding their transmitter specificity.

The influence of serotonergic and benzodiazepine type anxiolytic drugs on the cortical activation and sleep-wakefulness cycle were compared by evaluating the effects of ritanserin and deramciclane (EGIS-3886), two 5-HT2 receptor antagonists, and chlordiazepoxide on the electroencephalogram (EEG) in freely moving rats. Following drug administration (1, 3, and 10 mg/kg, PO for all drugs), EEG was continuously sampled for 6 h and power spectra were calculated for every 5 s to assess changes in slow wave activity and sleep phases. In a separate test, anticonvulsant effects of the drugs were examined in mice. Both deramciclane and ritanserin slightly increased total time spent in deep sleep (DS) and lengthened sleep episodes. In contrast, chlordiazepoxide had a strong inhibitory action on DS, sleep time being shifted to more superficial light sleep (LS). The incidence and length of the high voltage spindle (HVS) episodes characteristic for the motionless, awake rat were increased at the highest dose of both deramciclane and ritanserin, while it was decreased by chlordiazepoxide. In mice, chlordiazepoxide had a marked anticonvulsant effect, while deramciclane was moderately effective and ritanserin ineffective. In conclusion, the 5-HT2 receptor antagonist anxiolytic drugs seem to be superior compared to the benzodiazepine type anxiolytic drug, chlordiazepoxide, as ritanserin and deramciclane improved sleep quality by increasing sleep episode length and time spent in DS, while chlordiazepoxide enhanced sleep fragmentation and decreased DS.