Emergence and long-term maintenance of modularity in plastic networks of spiking neurons
The modular and hierarchical organization of the brain is believed to support the coexistence of segregated (specialization) and integrated (binding) information processes. A relevant question is yet to understand how such architecture naturally emerges and is sustained over time, given the plastic nature of the brains wiring. Following evidences that the sensory cortices organize into assemblies under selective stimuli, it has been shown that stable neuronal assemblies can emerge due to targeted stimulation, embedding various forms of synaptic plasticity in presence of homeostatic and/or control mechanisms. Here, we show that simple spike-timing-dependent plasticity (STDP) rules, based only on pre- and post-synaptic spike times, can also lead to the stable encoding of memories in the absence of any control mechanism. We develop a model of spiking neurons, trained by stimuli targeting different sub-populations. The model satisfies some biologically plausible features: (i) it contains excitatory and inhibitory neurons with Hebbian and anti-Hebbian STDP; (ii) neither the neuronal activity nor the synaptic weights are frozen after the learning phase. Instead, the neurons are allowed to fire spontaneously while synaptic plasticity remains active. We find that only the combination of two inhibitory STDP sub-populations allows for the formation of stable modules in the network, with each sub-population playing a distinctive role. The Hebbian sub-population controls for the firing activity, while the anti-Hebbian neurons promote pattern selectivity. After the learning phase, the network settles into an asynchronous irregular resting-state. This post-learning activity is associated with spontaneous memory recalls which turn out to be fundamental for the long-term consolidation of the learned memories. Due to its simplicity, the introduced model can represent a test-bed for further investigations on the role played by STDP on memory storing and maintenance. Author summaryOne of the most remarkable qualities of the brain is its capacity to learn and adapt. How the learning process imprints and maintains memories, by shaping the architecture of connectivity among neurons in a constantly changing and dynamic environment, is a major question of neuroscience. Here, we explore the idea that the segregation of inputs received by a neural network, with inputs targeting distinct populations, is a key factor for shaping the architecture of the network. We find that the presence of inhibitory neurons is necessary for the emergence and the long-term maintenance of modularity in spiking neural networks with plasticity. In particular, we show that two different inhibitory sub-populations, one subject to Hebbian and the other to anti-Hebbian plasticity, are required to promote the formation of feedback and feed-forward inhibition circuits controlling memory consolidation. On one side, these inhibitory circuits favour long-term memory consolidation by inducing spontaneous memory recalls in the asynchronous irregular resting phase. On another side, the number of inhibitory neurons control the maximal memory capacity of the considered model.