Patterned Neuronal Networks for Robotics, Neurocomputing, Toxin Detection and Rehabilitation

摘要

Biological systems can, in many aspects, be superior to existing control, computing, and warfare-agent detection systems. On the other hand, silicon-based electronics have many benefits over biological systems. Thus, an ideal system could use concepts and solutions from both biological and engineering principles. In this study we have developed basic methods to create hybrid neuronal systems consisting of single neurons or a simple neuronal network integrated with silicon-based electronics. We used the Self Assembled Monolayers (SAMs), DETA (trimethoxysilylpropy ldiethylenetriamine) and 13F (tridecafluoro- 1,1,2,2-tetrahydroctyl- 1-trichlorosilane) in combination with deep UV photolithography to create surface patterns to determine cell attachment and dendritic/axonal growth. We have utilized the necessary methods to characterize these patterns and evaluated the survival and physiology of the patterned neurons. Elemental composition analysis using XPS, contact angle measurements and electroless metallization proved the formation of the surface patterns. The 'two cell networks', 'polarity determination', and 'axon- outgrowth' patterns successfully directed the adhesion, growth, and differentiation of embryonic rat hippocampal cells and motoneurons in serum- free culture media. The cultured cells displayed a compliance of more than 50% to the cell-adhesive patterns of the SAMs at 4-6 days in vitro. The neurons survived up to 35 days on the patterns. Immunostaining with MAP2 and Neurofilament, neuronal markers indicated that surface patterns alone can determine polarity of the neurons. Single and dual patch clamp electrophysiological recordings proved that the patterned neurons exhibited normal physiological properties and formed functional synaptic connections. In this study we successfully integrated living cells with silicon structures using standard tools which are compatible with industrial microchip manufacturing methods.