GABAergic parvalbumin-expressing interneurons underlie distinct local EEG microstructure during different states of unconsciousness in rats

Abstract

Background: The GABAergic mechanism is an important target for the action of anesthetics and the promotion of sleep. We investigated the changes in hippocampal and reticulo-thalamic nucleus (RT) GABAergic parvalbumin (PV)-expressing interneurons as possible underlying mechanisms of the different local cortical and hippocampal EEG microstructures during NREM sleep compared with anesthesia-induced unconsciousness by two anesthetics with different main mechanisms of action. Methods: Twenty adult male Wistar rats were implanted for sleep recordings. After 3 hours of sleep recording, half of the rats were anesthetized with ketamine/diazepam (100 mg/kg, i.p.) and the other half with propofol (100 mg/kg, i.p.). We recorded EEGs of the motor cortex and hippocampus during the one-hour stable surgical phase of both anesthetics. The EEG microstructures of the motor cortex and hippocampus in local NREM sleep were compared with their EEG microstructures during 30 minutes of unconsciousness induced by a given anesthetic. At the end of each recording under stable anesthesia, rats were sacrificed for further PV and postsynaptic density protein 95 (PSD-95) immunohistochemistry. Results: All three states of unconsciousness differed in motor cortical and hippocampal EEG microstructures (χ2≥9.46;p≤0.01). During propofol-induced unconsciousness, attenuated delta and augmented sigma/beta amplitudes (z≥-4.13;p≤0.01) were the globally expressed difference, whereas increased gamma amplitude (z=2.35;p=0.02) was the only difference at the motor-cortical level compared to NREM sleep. During ketamine/diazepam-induced unconsciousness, attenuated theta (z≥-5.53;p≤10-4) and increased beta/gamma amplitudes (z≥-4.82;p≤10-4) were the globally expressed difference from NREM sleep. Both anesthesia-induced unconsciousness expressed globally as increased beta amplitude (z≥-4.13;p≤10-3) and increased motor-cortical gamma amplitude (z≥-4.20;p≤0.02) compared to NREM sleep. In contrast to propofol anesthesia, there was significant suppression of PV expression in the hippocampus (z≤-2.71;p≤ .01) and RT during ketamine/diazepam anesthesia in all rats, but only in the hippocampus was there an inhibitory/excitatory imbalance (increased PSD-95 expression). Conclusions: Although anesthesia and sleep share many neurobiological features, they are distinct states in terms of local EEG microstructure and underlying GABAergic and molecular substrate in local neuronal networks.