Differential Gene Expression Across Species Following Spinal Cord Injury: A Systematic Review and Meta-Analysis

The limited regenerative capacity of human nervous tissue compared to other species has long fascinated researchers. Remarkably, non-mammalian vertebrates such as teleost fish and axolotls possess robust regenerative abilities, capable of regenerating entire limbs and even fully transected spinal cords. Understanding the cellular and molecular mechanisms governing spinal cord regeneration across species may offer insights to accelerate repair processes and improve the quality of life for spinal cord injury (SCI) patients. This systematic review and meta-analysis compiled genome-wide spinal cord gene expression datasets from published SCI studies spanning multiple species. A targeted search was conducted using Gene Expression Omnibus, PubMed, and Science Direct, yielding 167 studies employing microarray and RNA sequencing (RNA-seq) technologies. After screening, 42 studies were selected: 20 used microarray, 17 used bulk RNA-seq, 2 employed single-cell RNA-seq (scRNA-seq), 2 combined bulk and scRNA-seq, and 1 incorporated both single-nucleus RNA-seq (snRNA-seq) and scRNA-seq. Approximately 43% of the selected studies were performed on Rattus norvegicus. Nine studies met the criteria for meta-analysis, allowing for cross-species comparison. Differentially expressed genes (DEGs) varied across models, reflecting species-specific responses. At day 7 post-injury, 214 DEGs were shared between regenerative (REG) and non-regenerative (non-REG) species. Across acute and subacute phases (multiple time points within the first week), 694 shared DEGs were identified. A month post-injury, 51 genes overlapped. Overall, 824 DEGs were common to the REG and non-REG groups at all time points. To explore functional relationships, protein-protein interaction (PPI) analysis was conducted on the oppositely regulated DEGs shared between REG and non-REG groups. Using Cytoscape and the CytoHubba plugin, key hub genes were identified, including CCNA2, CCNB1, CCNB2, CDC20, and FBXO5, all of which are associated with cell cycle processes. These hub genes were upregulated in non-REG (poorly regenerating) mouse models and may influence regenerative divergence across species by modulating glial proliferation or scar formation. Enrichment analysis revealed significant activation of cell cycle and mitotic processes in the non-REG group, which suggests cell cycle progression in response to injury.