TRANSCRIPTOMIC IDENTITY DRIVES SPIKE TIMING

Cortical neurons have diverse and stereotyped spike timing patterns, but it is unknown whether different transcriptomic neuron subtypes have distinct spike timing patterns. Recent technical advances allow identifying transcriptomic subtypes in two-photon calcium imaging (2P) recordings (Bugeon et al., Nature 2022), but 2P observes neural activity on timescales too slow to resolve spike timing differences.
Here, we show that distinct transcriptomic neuron types of mouse visual cortex have distinct temporal responses to visual stimuli. We recorded from the same population of neurons twice, first using 2P, then using Neuropixels probes, and matched neurons across modalities using their responses to “Zebra noise” visual stimuli (Skriabine, Shinn, et al, JVision 2025). We then used CoppaFISH-3D (Prankerd et al., SfN 2023; Zhou et al., SfN 2024) to identify the gene expression and transcriptomic subtype of thousands of neurons from this same population, and 2P to bridge transcriptomic subtypes to neurons recorded with Neuropixels.
Initial analysis of these data suggests marked differences between pyramidal tract (PT) and intratelencephalic (IT) neurons, especially in their temporal trajectory. PT and IT neurons show an inverted hierarchy, with the PT “output neurons” responding earlier than the IT neurons. We also find differences between fine subtypes within these classes, for example, with one PT neuron subtype reliably reducing its firing in response to visual grating stimuli.
Our results indicate that the diversity of neurons’ spiking activity is closely related to their molecular identity, and that transcriptomic subtype represents an important source of heterogeneity at the circuit and population level.