Heterogeneous Architecture of Neural Networks in Vitro with Precise Unidirectional Synaptic Connectivity

Neuronal networks in vitro today are more often used in a study of basic mechanisms of synaptic plasticity, information coding/decoding and learning. However, such neuronal circuit model lacks ordered connectivity of a network architecture which is crucial to develop functions and behavior in vivo. Here we present a new approach to engineer in vivolike network structure using microfluidic methods, which allows forming a multiple, independent and unidirectional synaptic connection pathways between two populations of neurons, with precise location of synapses in grid like structure. Developed approach can be applied to study information processing and learning experimentally mapping a location of separate synaptic connections between two neuronal populations.