Modeling human retinal ganglion cell axonal outgrowth, development, and pathology using pluripotent stem cell-based microfluidic platforms.

Retinal ganglion cells (RGCs) are highly compartmentalized cells, with long axons serving as the sole connection between the eye and the brain. RGC degeneration in injury and/or disease also occurs in a compartmentalized manner, with distinct injury responses in axonal and somatodendritic compartments. Thus, the goal of this study was to establish a novel microfluidic-based platform for the analysis of RGC compartmentalization in health and disease states. Human pluripotent stem cell (hPSC)-derived RGCs were seeded into microfluidics, enabling the recruitment and isolation of axons apart from the somatodendritic compartment. Initial studies explored axonal outgrowth and compartmentalization of axons and dendrites. We then compared the differential response of RGCs differentiated from hPSCs carrying the OPTN(E50K) glaucoma mutation with isogenic control RGCs in their respective axonal and somatodendritic compartments, followed by analysis of axonal transport. Further, we explored the axonal transcriptome via RNA-seq, focusing on disease-related axonal differences. Finally, we established models to uniquely orient astrocytes along the axonal compartment combined with modulation of astrocyte reactivity as a pathological feature of neurodegeneration. Overall, RGC culture within microfluidic chips allowed enhanced cell growth and maturation, including long-distance axonal projections and proper compartmentalization, while patient-specific RGCs exhibited axonal outgrowth deficits as well as decreased rate of axonal transport. Finally, the induction of astrocyte reactivity uniquely along the proximal region of RGC axons led to the onset of neurodegenerative phenotypes in RGCs. These results represent the first study to effectively recapitulate the highly compartmentalized properties of hPSC-derived RGCs in healthy and disease states, providing a more physiologically relevant in vitro model for neuronal development and degeneration.