EXCITATORY SYNAPTIC INPUT TO PARVALBUMIN INTERNEURONS PREFERENTIALLY SHAPES GAMMA OSCILLATION DYNAMICS AND PLASTICITY IN ACUTE BRAIN SLICES OF HUMAN NEOCORTEX

Gamma oscillations (30-100 Hz, γ-osc) are a type of local rhythmic network activity, generated by fast inhibition via parvalbumin interneurons (PV INs). While being strongly correlated with cognition and neuropsychiatric diseases in in vivo studies, cellular and synaptic characteristics of the involved cell assemblies in the human brain remain elusive.
Here, we present a protocol to reliably induce stable low-γ range (20-40 Hz) oscillations, partially accompanied by a temporally confined epileptiform interval, in in vitro local field potential recordings of human neocortical slices from epilepsy surgical interventions, enabling us to record up to 32 slices parallelly and establish a dataset of ≥ 50 patients.
Within this framework we were able to distinguish between inter-slice and inter-patient variability and resolve cortical area- and layer-dependent differences. Our pharmacological experiments show a strong reliance of γ-osc on excitatory AMPA receptor transmission to maintain peak frequency. Notably, by blocking calcium-permeable AMPA receptors that are preferentially expressed on PV INs, a similar frequency shift is observed. After repeated induction of γ-osc, we observe an NMDA receptor-independent potentiation in peak-power, hinting at PV IN-dependent cell-to-network plasticity. In conclusion, we established a reproducible, high-throughput framework that enables systematic investigation of the network architecture underlying low-γ oscillations on the patient-level, as well as the evaluation of pharmaceutical agents to modulate these dynamics, eventually deciphering the indispensable role of excitatory synaptic input onto PV IN in generating γ-osc.