J Chromatogr B Analyt Technol Biomed Life Sci. 2026 Nov 5;1286:125360. doi: 10.1016/j.jchromb.2026.125360. Online ahead of print.
ABSTRACT
Preterm brain injury (PBI) is a major cause of long-term neurodevelopmental deficits in preterm infants, yet its molecular mechanisms remain poorly defined. Secretory phospholipase A2 (sPLA2) family genes regulate lipid metabolism and inflammation and may contribute to neuronal injury, but their role in PBI is unclear. This study aimed to identify sPLA2-family genes associated with PBI and evaluate their diagnostic potential. Transcriptomic data from 24 PBI patients and 24 healthy controls (GSE112137) were analyzed to identify differentially expressed genes (DEGs) using linear modeling and empirical Bayes statistics. sPLA2-family genes (sPLA2-FGs) were curated from GeneCards, OMIM, PharmGKB, TTD, and DrugBank, and intersected with DEGs to obtain DE-sPLA2-FGs. Protein-protein interaction, GO, and KEGG enrichment analyses were performed, and LASSO regression, SVM-RFE, and Random Forest were applied to select robust feature genes. Single-cell transcriptomic data from mouse brain examined cell type-specific expression. Feature genes were further validated by qPCR and Western blot in peripheral blood from healthy controls, preterm infants without brain injury, and PBI patients, and in brain white matter of a neonatal hypoxic-ischemic mouse model of PBI. A total of 8423 DEGs were identified, including 4056 upregulated and 4367 downregulated genes. Intersection with sPLA2-FGs yielded 13 DE-sPLA2-FGs. PPI and enrichment analyses indicated involvement in vascular smooth muscle contraction, lipid metabolism, and neuroinflammatory pathways. Machine learning identified four robust feature genes (PLA2G3, PLA2G4A, ADCY7, ADCY8), which showed significant differential expression and high diagnostic accuracy (combined AUC = 0.979). Single-cell analysis revealed cell type-specific expression, with PLA2G4A and ADCY8 enriched in microglia, and ADCY7 in microglia and lymphoid populations. Peripheral blood validation confirmed decreased PLA2G3 and PLA2G4A and increased ADCY8 mRNA (protein unchanged) in PBI, with ADCY7 unchanged; in the mouse model, all four genes were robustly and concordantly dysregulated at both mRNA and protein levels in brain white matter. sPLA2-family genes are implicated in the molecular pathogenesis of PBI, potentially through lipid-mediated and neuroinflammatory mechanisms. These feature genes show strong diagnostic potential as novel biomarkers for early PBI detection and intervention.
PMID:43033512 | DOI:10.1016/j.jchromb.2026.125360

