Bioinformatics analysis of immune infiltration and cell cycle hub genes in atopic dermatitis.

BACKGROUND

Atopic dermatitis (AD) is a chronic inflammatory skin disease with complex pathogenesis. This study integrated bioinformatics and experimental validation to elucidate the molecular pathways and discover novel therapeutic targets for AD diagnostic and treatment.

METHODS

Transcriptomic datasets GSE182740 (training cohort, n = 75) and GSE5667 (validation cohort, n = 34) were analyzed. DEGs identified via GEO2R (|logFC|>2, p < 0.01) underwent KEGG/GO enrichment (Metascape; min overlap = 3, p < 0.01). STRING-constructed PPI networks (interaction score>0.4) visualized in Cytoscape (v3.10.1) revealed top 5 hub genes by CytoHubba MCC scoring. Immune cell infiltration patterns were quantified using ssGSEA software, with in vitro validation conducted in TNF-α/IFN-γ-stimulated keratinocytes.

RESULTS

Among 979 significant DEGs, inflammation-related pathways (IL-17, TNF-α, PI3K-Akt, p < 0.05) were enriched. PPI analysis revealed five cell cycle-regulating hub genes (AURKA, BUB1B, CCNB1, MELK, and TTK), all upregulated in AD lesions. In vitro experiments confirmed the upregulation of these hub genes in AD. These genes correlated significantly with activated CD4 T cells, Th2 cells, Treg cells, and central memory CD8 T cells (p < 0.05). In vitro experiments confirmed the overexpression of five hub genes in keratinocytes induced by inflammatory factors.

CONCLUSION

This study identified five hub genes driving AD by disrupting keratinocytes cell cycle regulation. These genes mediated keratinocytes-immune interactions, providing novel targets for AD therapy.