Inhibitory cell type heterogeneity in a spatially structured mean-field model of V1

Area of Science:

• Neuroscience
• Computational Neuroscience
• Theoretical Neuroscience

Background:

• Cortical inhibitory interneurons (parvalbumin, somatostatin, vasoactive intestinal polypeptide) modulate excitatory neurons differently.
• The role of heterogeneous spatial connectivity in inhibitory neuron function remains unclear.

Purpose of the Study:

• Investigate how diverse spatial connectivity of inhibitory neurons impacts neural network dynamics and computations.
• Explore the implications of distinct inhibitory cell types on excitation-inhibition balance and stability.

Main Methods:

• Developed a mean-field model of spatially structured neural networks with three inhibitory subtypes.
• Incorporated distinct connectivity probabilities and long-range projections for somatostatin (SST) neurons.
• Utilized stability analysis and linear response theory to examine network properties.

Main Results:

• Heterogeneous inhibitory networks with long-range SST projections maintain stability, unlike homogeneous networks.
• Conductance-based synaptic transmission is essential for reproducing cell-type-specific gain modulations.
• Derived position-dependent firing rate predictions and balanced network solutions.

Conclusions:

• A mix of short- and long-range inhibition is critical for computational diversity and network stability.
• Cell-type-specific and distance-dependent network structures are key to understanding cortical computations.
• Theoretical insights into the computational roles of inhibitory interneuron subtypes.