You might recall that researchers have been working on the direct conversion of glial cells in the scars produced following ischemic injury to the brain. Overexpression of NeuroD1 via gene therapy appears an effective approach to achieve this goal, at least in the controlled scenario of an animal model. In mice this intervention gives rise to neurons that integrate into existing neural circuits, leading to some degree of functional recovery.
We demonstrated that NeuroD1-mediated in vivo direct reprogramming of astrocytes into neurons promoted their neural circuit integration and led to the visual functional recovery after ischemic injury. Our work bridged the knowledge gap between individual cellular response recovery and animal behavioral recovery, where we characterized the functional synapses formed from specific projections and assessed neuronal response to stimuli in awake mice, which are critical functional characterization at the intermediate neural circuit level. The mouse primary visual cortex is a unique model system providing an opportunity to quantify projection specific functional connectivity and the direct visual responsiveness of the reprogrammed cells. Furthermore, the ability to record responses to different visual features such as orientation and direction provides a unique ability to quantify how well the cells mature and whether the synapses they receive are functional.
In our model system, the visual responses were drastically reduced following ischemic injury, yet they recovered following the NeuroD1 delivery. The putative excitatory neurons started to regain their visual responses 3 weeks after reprogramming, while the putative inhibitory neurons progressively integrated circuit inputs and refined their activity over a longer period of time. This delayed recovery of inhibition after reprogramming is similar to the absence of matured inhibition at an early age. Furthermore, these visual responses became more specific with time, based on our two-photon calcium imaging and extracellular recording results. The NeuroD1 converted cells gradually developed to be selective to the orientations and directions of visual stimuli, which is a typical feature of the mature visual cortical neurons. Interestingly, the reprogrammed cells at 6 weeks post-infection demonstrated higher selectivity compared to the healthy controls, which could be potentially explained by the more functionally developed synaptic inputs received by the reprogrammed cells compared to the healthy controls.
Compared to other studies, the functional recovery achieved by NeuroD1-mediated astrocyte-to-neuron conversion in vivo was similarly efficient. The local functional circuit and visual response recovery were also similar to embryonic neuronal transplantation results. Direct in vivo conversion of astrocytes into neurons removes the possibility of graft rejection and provides a viable solution for this problem, however. Our findings suggest that the NeuroD1-based in vivo direct reprogramming technology may be a promising gene therapy treatment of brain injury by replenishing the lost neurons and successfully integrating them into the existing neural circuit.