Neural Dynamics in Primate Cortex During Exposure to Subanesthetic Concentrations of Nitrous Oxide

Abstract

Nitrous oxide (N₂O) is a hypnotic gas with anti-depressant and psychedelic properties at subanesthetic concentrations. Despite longstanding clinical use, there is insufficient understanding of its effect on neural dynamics and cortical processing, which is important for mechanistic understanding of its therapeutic effects. We administered subanesthetic (70%), inhaled N₂O and studied dynamic changes of spiking rate, spectral content, and somatosensory information representation in primary motor cortex (M1) in two male Rhesus macaques implanted with Utah microelectrode arrays in the hand area of M1. The average, sorted multi-unit spiking rate in M1 increased from 8.1±0.99 to 10.6±1.3 Hz in Monkey W (p < 0.001) and from 5.6±0.87 to 7.0 ± 1.1 Hz in Monkey N (p = 0.003). Power spectral densities increased in beta and gamma band power. To evaluate somatosensory content in M1 as a surrogate of information transfer, fingers were lightly brushed and classified using a naïve Bayes classifier. In both monkeys, the proportion of correctly classified fingers dropped from 0.50±0.06 before N₂O to 0.34±0.03 during N₂O (p = 0.018), although some fingers continued to be correctly classified (p = 0.005). The decrease in correct classifications corresponded to decreased modulation depth for the population (p = 0.005) and fewer modulated units (p = 0.046). However, the increased single-unit firing rate was not correlated with its modulation depth (R² < 0.001, p = 0.93). These data suggest that N₂O degrades information transfer, although no clear relationship was found between neuronal tuning and N₂O-induced changes in firing rate. Significance Statement There are few intra-cortical studies characterizing the influence of nitrous oxide (N₂O) on neuronal behavior in the primate brain. Herein we demonstrate increased spiking rate in primary motor cortex (M1) as well as increased beta/gamma power during administration of subanesthetic N₂O. In a previously validated model of primary somatosensory to M1 information transfer, we also show a degradation of somatosensory representation. The degraded representation, as assessed by modulation depth, was not correlated with neuronal firing rate changes.