Neural Activity dynamic in Primate Cortex Across Consciousness Levels: Insights from High-Density Neuropixel Recording

Read the full article See related articles

Discuss this preprint

Start a discussion What are Sciety discussions?

Listed in

This article is not in any list yet, why not save it to one of your lists.
Log in to save this article

Abstract

This study investigates the anesthesia mechanisms induced by sevoflurane and how it modulates neural activity in the posterior parietal cortex (PPC) and prefrontal cortex (PFC) in Non Human Primates (NHPs) using high density Neuropixel probes. Spiking and local field potentials (LFPs) were recorded in two macaque monkeys under going four sevoflurane concentrations (2%, 3%, 4%, and 6%). We aimed to (i) quantify the emergence of anesthesia-induced Up/Down state dynamics, (ii) track changes in oscillatory power and inter-regional synchrony, and (iii) determine whether frontal and parietal areas exhibit differential sensitivity to rising and falling anesthetic depth. Across different anesthetic levels, we observed characteristic slow oscillations in delta range in both PFC and PPC, with neurons transitioning between high-firing "Up" states and near-silent "Down" states. Deeper anesthesia extended Down states, suppressed mean firing rates, and reduced the frequency and duration of Up states. In M1, no single units were detected in PFC, and only a few were recorded in M2. We suspect misalignment of the probe with PFC pyramidal cells and extensive suppression in PFC as the main reasons. Recurrent Neural Networks (RNN) was used to extract Up/Down states from LFP activities, based one the pattern observed in the PPC. The PPC to PFC information flow observed in Transfer Entropy analysis suggests that even under anesthesia, some level of feedforward-like interactions may persist. Up states originate in deep cortical layers and propagate toward superficial layers, following a bottom-up progression, indicating that deep-layer pyramidal neurons, which receive strong thalamic input, may be the primary drivers of Up states. The short Up states under deep anesthesia might represent a failed ignition attempt, where the brain momentarily tries to reactivate but cannot sustain functional activity due to global inhibition. LFP analyses revealed that although the absolute delta power remains high at different anesthetic levels, the relative delta band power is anti correlated with anesthetic depth, due to sporadic short (20ms to 40ms) burst in gamma (30 - 100 Hz) that appeared transient in nature. Lower sensitivity to anesthesia dose changes were observed in PFC as compared to PPC. This could explain why anesthesia first impairs cognitive function before affecting basic sensory responses. These results indicates that the traditional Up/Down state models might oversimplify anesthetic brain dynamics. While anesthesia is often described as a state of simple global slow-wave oscillations, the observed Up/Down state durations are not uniform, they fluctuate, follow non-trivial transition patterns, and differ between PFC and PPC.

Article activity feed