Integrating bulk RNA-seq, Mendelian randomization and single-cell RNA-seq to elucidate the roles of lactate metabolism related markers-SLC25A4 and keratinocyte in the pathogenesis of psoriasis.

BACKGROUND

Aberrant metabolic reprogramming-characterized by enhanced glycolysis and lactate accumulation-contributes to the pathogenesis of psoriasis (PsO). However, lactate metabolism-related genes that causally link mitochondrial dysfunction, immune dysregulation, and keratinocyte (KC) abnormalities in PsO remain insufficiently defined. This study aimed to identify robust lactate metabolism-associated biomarkers and new therapeutic targets in PsO.

METHODS

Bulk RNA-seq datasets were integrated with a curated lactate-related gene set to identify differentially expressed lactate-related genes (DE-LRGs) in PsO. Weighted Gene Co-expression Network Analysis (WGCNA), differential expression analysis, and two-sample Mendelian randomization (MR) were applied to prioritize candidate genes with causal associations. Machine learning algorithms were employed to identify robust diagnostic biomarkers and construct an artificial neural network (ANN) model. Functional enrichment, immune infiltration analysis, and immune mediator correlation analyzes were performed. Single-cell RNA sequencing (scRNA-seq) data were analyzed to determine cell-type specificity and KC subtype remodeling. Immunohistochemistry and in vitro experiments were conducted to validate the biomarker.

RESULTS

A total of 23 DE-LRGs were identified in PsO. Integrated network analysis and MR prioritized seven DE-LRGs with potential causal relevance. Machine learning approaches identified three robust biomarkers-GOT2, PYGL, and SLC25A4. The ANN model demonstrated excellent diagnostic performance across independent cohorts (AUC > 0.90). Functional enrichment indicated significant involvement in the JAK-STAT signaling pathway and aminoacyl-tRNA biosynthesis. Expression levels of the biomarkers were significantly correlated with immune-cell infiltration, particularly dendritic cells and resting mast cells, as well as key immune mediators. scRNA-seq analysis revealed substantial remodeling of KC subtypes in PsO, with altered differentiation trajectories and disrupted lactate-associated signaling and intercellular communication networks. SLC25A4 exhibited consistent downregulation in PsO at both bulk and single-cell levels, which was confirmed by immunohistochemistry. In vitro, SLC25A4 knockdown in HaCaT cells reduced cell viability, while promoting apoptosis. However, this study has certain limitations, as it is primarily based on public databases and in vitro experiments, without animal model validation. Further validation through multicenter, prospective cohorts and in vivo experiments is required.

CONCLUSION

This study systematically characterized the landscape of lactate metabolism-related genes in PsO and identified SLC25A4 as robust diagnostic biomarker, highlighting the translational potential of targeting lactate metabolism, particularly SLC25A4, as a diagnostic and therapeutic strategy in PsO.