Production and diffusion model of nitric oxide for bioinspired spiking neural networks

Nitric Oxide (NO) is an intracellular messenger whose diffusive properties enable an unconventional type of communication between neurons in the central nervous system that bypasses their anatomical connectivity. In this work, we modeled NO production and diffusion from a single source and investigated the range of action of the NO signal within a bioinspired spiking neural network. We found that a single active source will produce only a local effect on the individual synapse. While if multiple closely-located sources are active at the same time, NO will act more like a volume transmitter and influence even inactive synapses within that area. We focused our attention on the cerebellum's input layer, where NO is produced by the granule cells. In the granular layer, NO acts as a retrograde second messenger able to enhance presynaptic currents in the mossy fiber - granule cell synapses, thus potentiating them with long-term effects (LTP).