Towards understanding the microcircuit in monkey primary visual cortex in-vivo

The primary visual cortex (V1) is fundamental for visual processing. Transcriptomics suggests that the microcircuit in V1 consists of many inhibitory and excitatory cell types1. However, exactly how those diverse cell populations participate in visual cognition in vivo is unknown.
To address this question, we recorded from V1 of two anesthetized monkeys using high-density Neuropixel probes across all layers of cortex2. We identified single neuron waveforms using Kilosort2, followed by rigorous manual curation. Our final dataset was composed of 801 negative spiking neurons (from 2529 units, 5 sessions). We used histology and current source density to assign neurons to layers and then applied a recently developed nonlinear dimensionality reduction technique (WaveMAP3) to delineate candidate cell types.
WaveMAP revealed 9 candidate cell types ranging from narrow and triphasic, to broad biphasic waveforms. These candidate cell types were localized to distinct laminae, demonstrated heterogeneous firing rate responses, functional properties, and spatial profiles. Four of the narrow spiking cell classes were largely localized to Layer 4 of V1 and were likely stellate cells. Of these narrow spiking classes, one cell class demonstrated backpropagating action potentials (recently shown in mouse V14), whereas another had the highest and most sustained firing rates. Narrow, triphasic waveforms were in all layers except layer 2/3 and had the slowest and lowest visual response suggesting these are perhaps outputs from V1. In contrast, broad spiking waveforms were found in all layers. Of note, there was a cell class localized to Layer 5/6, with the broadest waveforms, robust firing rates, and the shortest response latency, which resembles Layer 6 pyramidal neurons that receive direct geniculocortical input.
Together, our results show that high density electrophysiology and machine learning analysis help delineate cell types within a cortical column, furthering our understanding of microcircuits in the visual cortex.