Functional Optic Tract Rewiring via Subtype- and Target-specific Axonal Regeneration and Presynaptic Activity Enhancement (bulk RNA-Seq)

Mechanisms underlying functional axonal rewiring after adult mammalian central nervous system (CNS) injuries remain unclear partially due to limited models. Using a mouse intracranial pre–olivary pretectal nucleus (OPN) optic tract injury model, we demonstrate that Pten/Socs3 knockout and CNTF expression in retinal ganglion cells (RGCs) promotes optic tract regeneration and reinnervation into OPN. Revealed by transmission electron microscopy, trans-synaptic labeling and electrophysiology, functional synapses are formed in OPN mainly by intrinsically photosensitive RGCs, thereby partially restoring the pupillary light reflex (PLR). Furthermore, combining with Lipin-1 knockdown accelerates the recovery and achieves functional reconnection in the chronic injury paradigm. PLR can be further boosted by increasing either RGC photosensitivity with melanopsin or optic tract presynaptic release with a voltage-gated calcium channel modulator, the latter of which also improves corticospinal-tract-sprouting-induced functional recovery after stroke. These findings highlight the importance of neuronal types and presynaptic activity for functional reconnection after CNS injuries. To compare the transcriptomic responses of retinal ganglion cells (RGCs) under two distinct injury models, optic nerve crush (ONC) and optic tract injury (OTI), a bulk RNA sequencing approach was employed. Specifically, RGCs were labelled by injecting AAV-DIO-GFP into the vitreous body of vGlut2-cre mice one month prior to any injury was applied to the animals. Subsequently, RGCs were collected for bulk sequencing from various treatment groups, including intact (non-injured) samples, ONC 1D (1 day post-injury), ONC 3D (3 days post-injury), ONC 7D (7 days post-injury), OTI 1D, OTI 3D, and OTI 7D. Fluorescence-activated cell sorting (FACS) was employed for the isolation of RGCs from each treatment group. Principal component analysis (PCA) and differential expression analysis were then performed to elucidate the distinctions between the two injury models.