Characterizing Synchronous Activity Propagation in Modular Neuronal Networks using Microscale Unidirectional Culture Devices

The brain is a complex system with a highly organized modular structure. The astronomically large amount of nonlinearly interacting elements in the brain makes it challenging to map the relation between modular structures and their functions, even with the state of the art methodologies. Unfortunately, traditional reduced in vitro systems cannot serve as a model, since they have a homogeneous and random structure, lacking the essential modularity that is required to probe the relation of the structure and function. It follows logically, that the required structure must be synthesized in vitro. During the last decade, several attempts have been made to reorganize the structure of in vitro systems and introduce modularity, for the purpose of studying communication among these synthetic modules. In these studies, to achieve micro-scale control over the neuronal populations and to realize the synthetic modules, a common method for microprocessing has been used which is borrowed from the field of microelectronics micro-manufacture. This thesis aims to expand the previous works in three directions: Micro-manufacturing methodology, design of the cross-module connectivity that permits the anatomical connections between neuronal populations only in one direction, and study of the developmental change in signal propagation across synthetic modules of neuronal networks ex vivo. It starts with the exposition of the proposed microprocessing method, in the format of a detailed reproducible protocol. Similarly, the new design for the cross-module connectivity pattern is introduced, and validated by visual and functional assays, that confirm the unidirectional propagation of the signal across modules. The micro-scale neuronal culture device with unidirectional connections, coupled to micro-electrode array extracellular recording systems, served as an optimal experimental platform to study the propagation of the signal across the modules of neuronal networks. Multiple methods for investigation of signal propagation across the modular neuronal networks are proposed and implicated in this work, e.g., cross correlation of population activity, cross correlogram of single unit activity and algorithmic search. These methods, in the form of scripting language, along with the blueprints of the micro-scale devices have been made publicly available. Investigation into the propagation of signals across modules revealed a notable delay of up to 200 milliseconds in the transmission of synchronous activity. A series of hypotheses were tested, in particular, by changing the number of anatomical connections between the modules, manipulation of the dynamics of fast synaptic transmission, and spiking network simulation, to elucidate the nature of this effect.