Two-Photon Polymerization 3D-Printing of Micro-scale Neuronal Cell Culture Devices

Neuronal cultures have been a reference experimental model for several decades. However, 3D cell arrangement, spatial constraints on neurite outgrowth, and realistic synaptic connectivity are missing. The latter limits the study of structure and function in the context of compartmentalization and diminishes the significance of cultures in neuroscience. Approximating ex vivo the structured anatomical arrangement of synaptic connectivity is not trivial, despite being key for the emergence of rhythms, synaptic plasticity, and ultimately, brain pathophysiology. Here, two-photon polymerization (2PP) is employed as a 3D printing technique, enabling the rapid fabrication of polymeric cell culture devices using polydimethyl-siloxane (PDMS) at the micrometer scale. Compared to conventional replica molding techniques based on microphotolitography, 2PP micro-scale printing enables rapid and affordable turnaround of prototypes. This protocol illustrates the design and fabrication of PDMS-based microfluidic devices aimed at culturing modular neuronal networks. As a proof-of-principle, a two-chamber device is presented to physically constrain connectivity. Specifically, an asymmetric axonal outgrowth during ex vivo development is favored and allowed to be directed from one chamber to the other. In order to probe the functional consequences of unidirectional synaptic interactions, commercial microelectrode arrays are chosen to monitor the bioelectrical activity of interconnected neuronal modules. Here, methods to 1) fabricate molds with micrometer precision and 2) perform in vitro multisite extracellular recordings in rat cortical neuronal cultures are illustrated. By decreasing costs and future widespread accessibility of 2PP 3D-printing, this method will become more and more relevant across research labs worldwide. Especially in neurotechnology and high-throughput neural data recording, the ease and rapidity of prototyping simplified in vitro models will improve experimental control and theoretical understanding of in vivo large-scale neural systems.