In Vitro Model for Retinal Ganglion Cell Reinnervation of Thalamic Target Structures

Current deep brain stimulation electrodes have a very limited spatial resolution which highly limits their application for targeted single neuron stimulation. However, such highly resolving electrodes would be essential for applications such as vision restoring by direct stimulation of primary visual centers in the brain. To overcome this limitation, we propose a living biohybrid electrode in which we exploit real neurons as relays to transmit information from a multielectrode array (MEA) to a neuronal target structure. We use PDMS microstructures with defined seeding wells and only few µm small axon channels to guide axons across electrodes towards a neuronal target structure. The physical separation into defined PDMS channels electrically isolates the axons and enables us to stimulate individual axons before merging them into a single nerve. The axons of the nerve can then form synapses with individual neurons at the target structure and thereby enable stimulation of only few neurons. Our goal is to create an implantable biocompatible stretchable biohybrid MEA¹in which single retinal ganglion cells (RGCs) grow axons unidirectionally across electrodes into a bioabsorbable axon guidance tube to form an artificial nerve that can innervate and stimulate the brain.<br/>To achieve this goal, we established a PDMS microstructure based <i>in vitro</i> coculture system mounted on conventional glass MEAs in which we can seed embryonic rat RGC and thalamus spheroids at defined locations and systematically optimize the design to achieve axon specific activity patterns in the thalamus target tissue. The basic design of our PDMS microstructure consists of several wells filled with the RGC spheroids. Small axon guidance channels collect the individual axons from each well and guide them into a common 3mm long nerve forming channel. The end of the nerve channel is connected to one or several thalamus wells to emulate innervation.<br/>We could show that within 2-3 weeks we can grow a &gt;3 mm long artificial optic nerve within the PDMS axon guidance structures that innervates the thalamic target spheroids. Time-lapse recordings of migrating RGC axons are used to compare how different structural elements enhance unidirectional axon merging and growth speed. We correlate these structural observations with electrical and optical activity recordings to assess crosstalk between RGC wells and show how individual stimulation of single RGC wells influence the activity pattern in the thalamic target structures. Moreover, we assess how the length and diameter of PDMS axon guidance channels affect the propagation of action potentials. To convert the 2D design into an implantable device we show results on how to guide axons from a PDMS microstructure into a vertically mounted 160 µm diameter and 3 mm long glass axon guidance tube that would guide axons into the brain.