Single-cell transcriptomics identifies keratinocyte-centered lactylation modulation and experimentally validated biomarkers for psoriasis diagnosis

Background Psoriasis (PS) involves lactylation-driven keratinocyte and immune dysfunction. This study sought to identify key lactylation-related biomarkers and networks for improved diagnosis and therapy. Methods PS-related transcriptome and single-cell RNA sequencing data were obtained from Gene Expression Omnibus (GEO). scRNA-seq analysis identified key cell populations and lactylation-related gene (LRG) activity. Weighted Gene Co-expression Network Analysis (WGCNA) identified LRG-correlated modules, intersected with differentially expressed genes (DEGs) to find candidate biomarkers. Machine learning algorithms were integrated for feature selection and model construction, with Receiver operating characteristic (ROC) analysis and a nomogram for diagnostic evaluation. Regulatory networks and drug prediction were established, followed by molecular docking. Biomarker expression was validated by Quantitative reverse transcription polymerase chain reaction (RT-qPCR) and Western blot in peripheral blood from PS patients and healthy controls. Results Single-cell RNA sequencing identified keratinocytes as the key PS-associated cell type with highest LRG activity. WGCNA revealed six LRG-related modules (921 genes) enriched in ribonucleoprotein biogenesis and mitochondrial complexes. Intersection with DEGs highlighted 20 hub genes involved in cell cycle pathways. A four-biomarker model ( UBE2T , FGFBP1 , EIF4EBP1 , GINS3 ) achieved high diagnostic accuracy (Area Under Curve (AUC) > 0.9) and was integrated into nomogram with high clinical utility. RT-qPCR confirmed significant upregulation of all four genes in PS patient samples ( P < 0.05), with FGFBP1 and GINS3 showing particularly pronounced increases ( P < 0.001). Western blot analysis further demonstrated elevated EIF4EBP1 protein expression, consistent with transcriptomic findings. Drug prediction and molecular docking identified resveratrol and calcitriol as potential therapeutic agents targeting GINS3 . Biomarker expression correlated significantly with altered immune infiltration, notably M0 macrophages and resting mast cells. Conclusion This study identified UBE2T , FGFBP1 , EIF4EBP1 and GINS3 as key PS biomarkers with high diagnostic accuracy. These biomarkers show significant immune infiltration correlations and provide crucial insights for PS pathogenesis and clinical management.