Network-centered homeostasis through inhibition maintains hippocampal spatial map and cortical circuit function

Area of Science:

• Neuroscience
• Computational Neuroscience
• Systems Neuroscience

Background:

• Neural activity exhibits remarkable stability despite continuous experience, a phenomenon often attributed to homeostatic plasticity.
• The precise conserved aspects of neural activity and mechanisms for maintaining learning flexibility remain incompletely understood.
• Emerging evidence suggests network-centered control complements neuron-centered mechanisms in neural regulation.

Purpose of the Study:

• To computationally investigate input-dependent inhibitory plasticity (IDIP) as a network-centered mechanism for neural stability.
• To explore how IDIP may reconcile neural stability with the flexibility required for learning and memory.
• To assess the functional implications of IDIP across different brain areas and experimental observations.

Main Methods:

• Development of a computational hippocampal model incorporating IDIP.
• Simulation of neural network dynamics under various conditions to observe emergent properties.
• Analysis of place cell activity, network remapping, and persistent activity patterns.

Main Results:

• IDIP successfully models the emergence of active and silent place cells.
• IDIP explains remapping phenomena observed after silencing of active place cells.
• IDIP demonstrates the stabilization of recurrent neural dynamics while preserving firing rate heterogeneity and stimulus representation.
• IDIP supports persistent activity crucial for memory encoding.

Conclusions:

• Input-dependent inhibitory plasticity (IDIP) provides a viable network-centered mechanism for maintaining neural stability.
• IDIP reconciles neural stability with the dynamic flexibility needed for learning and memory.
• IDIP has broad functional implications, potentially explaining diverse experimental findings across brain regions.