Cell type- and layer-specific inhibitory synaptic plasticity duringassociative learning

Neocortical GABAergic inhibitory neurons dynamically regulate cortical activity across various contexts. Modulation of inhibition has been proposed as a key mechanism enabling circuit rewiring during learning. By using in vivo Ca++ imaging and electrophysiology, prior studies have shown changes in the activity of different inhibitory neuron types during learning. However, these modulations vary depending on cell types, brain regions, layers, and tasks. Furthermore, inhibitory neurons form canonical microcircuits with excitatory and other inhibitory neurons, but the target specificity of these changes remains unclear.In this study, we investigated whether changes in inhibitory synaptic output onto excitatory pyramidal (Pyr) neurons during associative learning are cell type- and layer- specific. Using in vitro electrophysiology combined with transgenic mouse lines, we examined how somatostatin (SST)- and parvalbumin (PV)-expressing inhibitory neurons change their synaptic output onto Pyr neurons in primary somatosensory (barrel) cortex following whisker-dependent sensory association training. We found that SST output onto layer 2/3 (L2/3) Pyr neurons, but not L5 Pyr neurons, was significantly reduced after one day of training. This reduction was specific to the training in which whisker stimulus predicts a reward and was observed only at the onset of the training.Given evidence of potential sex differences in PV neurons, we first investigated potential sex-dependent variations in PV neurons within barrel cortex. We found that PV synaptic output onto L5 Pyr neurons was higher in males than females. Intrinsic properties of PV neurons did not differ between sexes. We further examined whether PV synaptic output was differentially modulated between sexes during associative learning. Despite similar learning rates in males and females, PV synaptic output in L2/3 was reduced only in females. No training-dependent change was observed in L5 in either sex.Taken together, our findings indicate that reducing inhibitory synaptic output is a circuit modification unique to superficial layers during the early phase of learning. Furthermore, our study highlights sex as an important biological variable influencing circuit adaptations during learning.<br>