Amnesia is one of the essential desirable elements of the anesthetic state along with unconsciousness (hypnosis) and immobility. The post-anesthetic recall of contextually rich (episodic) memories in particular is highly undesirable and a potential cause of morbidity.
Current thinking attributes behavioral anesthetic effects to interactions with specific proteins (as opposed to nonspecific effects on lipid membranes). Indeed, anesthetic interactions with numerous plausible molecular targets have been and continue to be extensively documented.1,2
Clinically used inhalational anesthetics are chemically diverse. Despite different receptor-level activity profiles and potencies, all inhaled anesthetics impair learning and memory at concentrations that are subhypnotic and, typically, are only a fraction of the standard ‘surgical’ concentration that is required for immobility.3–6
For example, both the alkane halothane which markedly enhances γ-aminobutyric acid receptor type A-mediated inhibition and the non-γ-aminobutyric acid receptor-ergic gas nitrous oxide suppress inhibitory avoidance training at comparable lipid solubility-corrected concentrations.7
Similarly, the anesthetic isoflurane and the nonimmobilizer 1,2-dichlorohexafluorocyclobutane (F6 or 2N, an experimental drug) suppress fear conditioning (FC) at similar concentrations.5,8
Notably, both are more potent inhibitors of FC to context (hippocampus-dependent) than FC to tone (hippocampus-independent).5,8
The reason for the preferential sensitivity of hippocampus-dependent learning to suppression by anesthetic(like) compounds is not related to specific molecular targets in any obvious pattern: isoflurane enhances γ-aminobutyratergic and inhibits glutamatergic synaptic transmission9,10
and blocks hippocampal long-term potentiation11
while F6 has no known effect on any of these processes.12,13
These observations beg the question whether all anesthetics similarly affect a single amnesia-promoting molecular target or whether their aggregate actions on different targets converge at some higher level of signal integration that is of particular importance for hippocampal learning and memory. The experiments presented in this manuscript investigate the latter possibility.
The hippocampal θ-rhythm is a prominent network activity of the ‘on-line’ hippocampus that can be separated into atropine-sensitive (type-2) and atropine-resistant (type-1) components. Type-1 theta is suppressed by surgical levels of anesthesia.14
Indeed, based on this observation, it was proposed more than 30 years ago that suppression of this non-cholinergic activation of the cerebrum may mediate behavioral effects of anesthesia.14
Since then, substantial evidence has accumulated demonstrating that the θ-rhythm serves an essential network-level role in hippocampal learning and memory (reviewed in references 15,16
). For example, θ-oscillations facilitate plasticity17
and support mnemonic processes requiring inter-regional signal integration.18–20
Conversely, suppression of the θ-rhythm impairs learning and memory.21–23
We hypothesized that modulation of type-1 θ-oscillations might serve as a common network-level mechanism of anesthetic-induced impairment of hippocampus-dependent learning and memory. If this were correct, some measure of θ-activity should vary with anesthetic concentrations in the amnestic (but subhypnotic) range. We tested this hypothesis by analyzing the effect of three inhaled anesthetics on type-1 θ-oscillations. We found that θ-frequency (but not other parameters of the θ-rhythm) changed systematically with anesthetic dose. We conclude that these results are consistent with the hypothesis that slowing of θ-frequency correlates with suppression of hippocampus-dependent learning and memory.