In this work, we introduce a novel nerve-on-a-chip model designed as a neural interface for deep brain stimulation. Termed as a "biohybrid" approach, it aims to overcome the limitations of standard deep brain implants such as low stimulation resolution. The biohybrid concept leverages on-chip grown retinal neurons to convert electrical signals from a stretchable 2D microelectrode array into synaptic stimulation of a neural target tissue.

The device consists of two primary components: a soft, stretchable multi-electrode array (MEA) and an axon-guiding microstructure [1]. The MEA, fabricated using a transfer stripping method [2], comprises a PDMS substrate and microstructured platinum tracks. The PDMS microfluidic axon guidance structure is aligned and bonded onto the microelectrode array. Spheroids of living retinal neurons labelled with a viral vector (mRuby3) are then seeded into 16 seeding wells and cultured under standard cell culture conditions before implantation.

We describe the fabrication of the biohybrid neural interface and characterize the device electrically and mechanically. We demonstrate how the retinal ganglion cells seeded into the implant form an artificial nerve of up to 3 mm in length. Moreover, we demonstrate how axons transit from the biohybrid implant into a nerve-forming bioresorbable collagen tube that will guide axons from the implant towards a neural target structure for sensory reinnervation and synaptic stimulation of the visual thalamus in vivo. We show that retinal spheroids can be stimulated using the stretchable microelectrode array using functional calcium imaging. To assess stimulation-induced signal transmission in the biohybrid implant we present in vitro data on how spikes propagate within the axon guidance channels using CMOS [3] multielectrode arrays. Lastly, we demonstrate that neurons cultured in the device grow axons in the microstructure, and exhibit spontaneous activity for over 3 weeks when implanted in mice.

With this work, we show that this biohybrid approach has the potential to serve as a new kind of neural interface technology. Although further experiments are necessary for in vivo synapse formation and deep-brain stimulation, previous work has show feasibility of this approach in vitro [4], and the findings presented in this work pave the way for a new kind of neural interfaces.