Abstract: The permanent functional deficits resulting from the inability of adult mammalian central nervous system neurons to regenerate after injury present a significant clinical challenge. While traditional stem cell transplantation strategies continue to encounter ethical concerns and the risk of immune rejection, this impasse has shifted regenerative medicine research toward targeting endogenous astrocytes. Due to their intrinsic plasticity, widespread distribution throughout the central nervous system, and affinity for neurodevelopmental lineage, astrocytes are a unique target for in situ neuronal regeneration. This review systematically elucidates the core regulatory network governing astrocyte transdifferentiation, identifying 10 key signaling pathways, such as Wnt signaling pathway, that form a cascade regulatory system. Directed overexpression of transcription factors such as NeuroD1, Ascl1, or Neurog2 can directly initiate neuronal phenotypic conversion. Meanwhile, small molecule compounds such as valproic acid combined with CHIR99021 activate endogenous neurogenic programs by inhibiting the bone morphogenetic protein signaling axis. Notably, polypyrimidine tract binding protein 1 (PTB) gene silencing significantly enhances transdifferentiation efficiency by suppressing the microRNA 124/re1 silencing transcription factor (miR-124/REST) feedback loop. From a translational perspective, a multidimensional evaluation system based on morphological, molecular marker, and electrophysiological properties has demonstrated considerable therapeutic potential. In stroke models, NeuroD1-mediated transdifferentiation replenished approximately 30% of lost cortical neurons and improved motor coordination, evidenced by enhanced performance in food pellet retrieval, grid walking, and cylinder tests compared with controls. In spinal cord injury studies, SOX2-induced glutamatergic neurons moderately reduced glial scar density by about 25%, permitting regenerating axons to pass through while preserving the supportive structure of scar. In neurodegenerative contexts, PTB inhibition yielded functionally mature dopaminergic neurons and reconstructed nigrostriatal pathways in Parkinson’s disease models. In Alzheimer’s disease models, adeno-associated virus-delivered NeuroD1 induced whole-brain neural circuit remodeling, generating 500,000 new neurons widely distributed across the cortex and hippocampus, accompanied by improved cognitive performance. Current technical limitations include off-target effects of adeno-associated virus vectors, which cause nonspecific gene expression and require rigorous validation via Cre-loxP lineage tracing. Transdifferentiation efficiency is also highly influenced by regional microenvironments: gray matter astrocytes show higher conversion rates than those in white matter, and oxidative stress increases apoptosis among newly generated neurons. Clinical translation is further constrained by the safety of delivery systems and the aging tissue microenvironment, where transforming growth factor beta 1 is often elevated. Ferroptosis inhibitors have been shown to nearly double the survival rate of transdifferentiated cells, offering a novel strategy to mitigate oxidative damage. Based on current evidence, astrocyte transdifferentiation enables neural functional recovery across multiple disease models through endogenous repair mechanisms. Future advances should focus on optogenetically inducible vectors for spatiotemporal precision, non-viral delivery systems to mitigate vector-related risks, and integration of long-term safety validation in non-human primates with single-cell multi-omics technologies to facilitate the clinical translation of personalized regenerative therapies.

Key words: Alzheimer’s disease, amyotrophic lateral sclerosis, cell transdifferentiation, Huntington’s disease, ischemic stroke, nervous system diseases, Parkinson’s disease, ptbp1 protein, small molecule, spinal cord injury, transcription factors