Functional repair after ischemic injury through high efficiency in situ astrocyte-to-neuron conversion

Abstract

33Mammalian brains have largely lost internal neural regeneration capability except for a 34 few discrete neurogenic niches. After brain injury, the cerebral cortex is especially 35 difficult to repair due to its extremely low rate of adult neurogenesis. Previous studies 36 have converted glial cells into neurons, but the total number of neurons generated is 37 rather limited, casting doubt about its therapeutic potential. Here, we demonstrate that 38 high-efficiency neuroregeneration can be achieved in adult mammalian brains by 39 making use of an engineered AAV Cre-FLEX system to convert a large number of 40 reactive astrocytes into functional neurons. Specifically, using a combination of 41 GFAP::Cre and FLEX-NeuroD1 AAV system, we were able to regenerate enough new 42 neurons from astrocytes to cover about 40% of the neurons lost from an ischemic injury 43 (400 NeuN+ new neurons/mm2), compared to previously reported an average of <1% of 44 cortical neurons (2-8 NeuN+ neurons/mm2) in an ischemic-injured adult mammalian 45 cortex. Importantly, this in situ astrocyte-to-neuron conversion process also improved 46 survival of injured pre-existing neurons, (additional 400 neurons/mm2), leading to a 47 repaired motor cortex with layered cortical structures. Moreover, NeuroD1-converted 48 neurons not only form functional neural circuits but also rescue motor and memory 49 deficits after ischemic injury. Our results establish the proof-of-principle that a highly 50 efficient in situ astrocyte-to-neuron conversion approach provides a novel treatment for 51 neurological disorders that are in need of new neurons. 52 row). Interestingly, some NeuroD1 AAV-infected astrocytes were clearly captured in a 128 transitional stage toward neurons, showing co-immunostaining of both NeuN and GFAP 129 (Fig. 1g, bottom row). Remarkably, using our AAV Cre-FLEX system, we detected a 130 large number of NeuroD1-converted neurons in the injured mouse cortex (Fig. 1h, i), 131 which was not possible using the other viruses we had tried. 132 133

Regeneration of dorsal spinal cord neurons after injury viain situNeuroD1-mediated astrocyte-to-neuron conversion

“…Importantly, we also observed highly efficient NeuroD1-mediated neuronal conversion in the contusive SCI model making this technique a feasible intervention for regenerating functional new neurons in the grey matter to treat SCI. Consistent with our in vivo cell conversion work in the brain (Chen et al, 2018;Guo et al, 2014), the NeuroD1converted neurons in the dorsal horn of the injured spinal cord were also glutamatergic, but Tlx3+ spinal neurons instead of Tbr1+ cortical neurons. The fact that the same neural transcription factor NeuroD1 can convert cortical astrocytes into cortical neurons and spinal astrocytes into spinal neurons may hold the key to a region-flexible neural repair by using internal glial cells for neuroregeneration.…”

“…3). Therefore, the majority of NeuroD1-converted neurons in the dorsal horn of the spinal cord are glutamatergic neurons, consistent with our finding in the mouse cortex (Chen et al, 2018).…”

“…Following our earlier report of efficient conversion of reactive astrocytes into functional neurons by NeuroD1 through retroviral infection (Guo et al, 2014), intravenous injection of AAV9 expressing NeuroD1 has been reported to infect a small but significant number of resting astrocytes in the striatum and convert them into neurons (Brulet et al, 2017). On the other hand, our intracranial injection of AAV9 expressing NeuroD1 in stab injury model (Zhang et al, 2018) and ischemic injury model (Chen et al, 2018) have both resulted in high conversion efficiency, suggesting that reactive astrocytes are more likely converted into neurons than the resting astrocytes. Consistent with this hypothesis, our results in this study also support that reactive astrocytes in injured spinal cord can be effectively converted into neurons with ~95% efficiency, regardless of retrovirus or AAV delivery system.…”

“…Our group has previously demonstrated that the neurogenic transcription factor NeuroD1 can reprogram reactive astrocytes into functional neurons in the stab-injured brain and in a mouse model for Alzheimer's disease (Guo et al, 2014). Following studies demonstrated that NeuroD1-mediated in vivo astrocyte-to-neuron conversion can reverse glial scar back to neural tissue (Zhang et al, 2018) and repair the damaged motor cortex after ischemic stroke (Chen et al, 2018). The major goal of the current study is to determine whether NeuroD1 can reprogram reactive astrocytes into functional neurons in the injured spinal cord.…”

“…2A). Specifically, we used a Cre-Flex gene expression system, which contains two AAV vectors, with one encoding GFAP-Cre and the other encoding the transgene in reverse form flanked by double LoxP sites (FLEX vector) (Atasoy et al, 2008;Chen et al, 2018;Liu et al, 2015). Thus, when the two AAVs are co-injected into the spinal cord, Cre recombinase will be expressed in the infected reactive astrocytes and turn on the transgene expression in FLEX vector by flipping the transgene sequence into the correct form for transcription ( Fig.…”