UCSF

The precise pattern of neural activity in the auditory cortex, which governs our perception of sound, is continuously shaped by synaptic inhibition. Cortical inhibition is generated by local GABAergic interneurons, which fall into several subtypes on the basis of their morphology, gene expression, biophysical properties, and innervation patterns. Each subtype is hypothesized to modulate auditory processing in intricate and distinctive ways, yet the links between each subtype and the computations they support have not been definitively made. Using a combination of in vivo extracellular recordings in the primary auditory cortex, optogenetic perturbations, and computational modeling, this thesis explores in what ways, and under what conditions, these inhibitory microcircuits might support different aspects of auditory processing. Firstly, we demonstrate that the two most common types of cortical interneuron – somatostatin-positive (Sst+) and parvalbumin-positive (Pvalb+) interneurons – can differentially alter the gain and tuning of spectral receptive fields, but that observing these differences critically depends on whether one increases or decreases their activity to read out their functions, as well as on the activity state of the network in which they are embedded. Secondly, we show that the ways in which cells process temporal aspects of sound (specifically, stimulus history) are quite diverse, and that this diversity can be explained, in part, by cell-to-cell variability in the overall strength of synaptic inhibition. Finally, we find that inhibition from Sst+ and Pvalb+ interneurons contributes to such temporal processing in distinct ways, potentially stemming from differences in how they are activated by their afferents over time. This thesis demonstrates that cortical circuits can recruit distinct inhibitory microcircuits to differentially fine-tune the spectro-temporal representations of sound, in ways that depend on sensory context and the state of the cortical network.