High MAP3K6 expression in spinal cord injury tissues enhances neuronal apoptosis: insights from multi-omics data integrating WGCNA, machine learning and experimental validation.

Abstract

Spinal cord injury (SCI) is a major global health issue with severe complications, yet effective biomarkers remain elusive. We analyzed the GSE226238 dataset from the Gene Expression Omnibus (GEO) database and identified 4621 differentially expressed genes (DEGs) between SCI and controls, comprising 2684 upregulated and 1577 downregulated genes. Functional enrichment analyses revealed these DEGs are predominantly involved in protein degradation pathways, immune-related processes and ubiquitin-mediated proteolysis. Immune infiltration analysis using multiple algorithms showed significant alterations in B cells, T cells, NK cells and neutrophils, indicating a complex immune microenvironment post-SCI. Using WGCNA, we constructed a scale-free co-expression network and identified nine modules; the green module showed the strongest positive correlation with SCI. Intersecting DEGs, WGCNA module genes, and random forest-selected features identified five candidate genes: MAP3K6, ATP5MPL, NDUFB1, RNASE2 and MIR373. Single-cell RNA sequencing further revealed that MAP3K6 exhibited the highest expression among candidates, predominantly in neuroepithelial and neuronal cells. Validation in independent GSE151371 dataset confirmed significantly elevated MAP3K6 expression in SCI, with ROC analysis demonstrating robust diagnostic efficacy (AUC = 0.918). In vitro, MAP3K6 knockdown in mouse spinal cord neuronal cells promoted cell growth and inhibited apoptosis, decreasing pro-apoptotic proteins (BAX, caspase-3, cleaved caspase-3) and increasing anti-apoptotic Bcl-2. Collectively, our multi-omics analysis integrated with experimental validation identifies MAP3K6 as a key regulator in SCI pathogenesis, offering new insights into molecular mechanisms and highlighting its potential as a diagnostic biomarker and therapeutic target.