Glycolysis-related gene signatures in spinal cord injury pathophysiology identification through integrative gene expression analysis

Ai Xin^1*Xiaoqin Liu²Zhuang Wang¹Jiating Hu²Guodong Shi²

• ¹Yanan University Affiliated Hospital, Yanan, China
• ²Medical College of Yan'an University, Yan'an, China

Background: Spinal cord injury (SCI) is a debilitating condition that significantly impacts patients' mobility and quality of life, posing a substantial economic burden on healthcare systems. Increasing evidence suggests that metabolic reprogramming, particularly glycolysis, is involved in inflammatory responses following SCI. This study aims to systematically investigate the association between the role of glycolysis-related genes (GRGs) and SCI, and to identify potential candidate biomarkers. Methods: The GSE151371 dataset was retrieved from the Gene Expression Omnibus (GEO) to identify differentially expressed genes (DEGs), which were subsequently subjected to Weighted Gene Co-expression Network Analysis (WGCNA) to pinpoint glycolysis-associated modules. Hub genes associated with SCI were initially identified through machine learning algorithms and subsequently evaluated using the independent GSE45006 dataset. Immune infiltration in SCI was profiled by single-sample gene set enrichment analysis (ssGSEA) and correlated with hub gene expression. After establishing TF-miRNA-mRNA and protein-chemical networks, hub gene expression patterns were characterized by scRNA-seq and further validated experimentally in vivo by qRT-PCR and Western blotting. Results: From 1138 differentially expressed genes (DEGs), WGCNA identified 704 in glycolysis-associated modules. Intersecting these with glycolysis-related genes (GRGs) yielded 13 candidates. Subsequent machine learning pinpointed six hub genes (ALK, GGH, IRS1, PPARG, SLC1A3, UBTD1), of which only IRS1 and UBTD1 showed consistent expression patterns in an external dataset. ssGSEA identified 20 differentially abundant immune cell types in SCI. Subsequently, IRS1 expression was associated with activated T cells and natural killer (NK) cells, while UBTD1 expression correlated with activated dendritic cells, monocytes, and neutrophils. scRNA-seq revealed that Irs1 was mainly expressed in endothelial and epithelial cells, while Ubtd1 was broadly expressed, with higher levels in endothelial cells and microglia. qRT-PCR revealed significant upregulation of Ubtd1 in the SCI group, whereas Irs1 expression did not differ significantly. Western blot further confirmed elevated UBTD1 protein levels in SCI compared with the Sham group. Conclusion: Our integrative transcriptomic and experimental analyses suggest that UBTD1, a glycolysis-related gene, as significantly associated with SCI and immune cell infiltration, highlighting its potential as a biomarker and suggesting its role in metabolic–immune interactions post-SCI.