Transcriptome Analysis, Machine Learning, and Experimental Identification of CDK7 Affecting the Progression of...

Pregnancy-induced hypertension (PIH) is a severe pregnancy complication characterized by placental insufficiency, abnormal vascular remodeling, and immune dysregulation, but personalized therapeutic markers remain unclear. This study aimed to identify key genes and explore immune mechanisms in PIH using transcriptome analysis, machine learning, and experimental validation.

We analyzed the GSE204835 transcriptomic dataset to screen differentially expressed genes (DEGs) and performed Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG), Reactome, and Gene Set Enrichment Analysis (GSEA) for functional annotation. Immune infiltration analysis was also performed to examine the immune landscape in PIH. Least Absolute Shrinkage and Selection Operator (LASSO) regression identified key genes, which were validated in a PIH cell model. Flow cytometry and immunofluorescence assays assessed the effect of CDK7 knockdown on macrophage polarization.

A total of 1,598 DEGs (1,123 upregulated, 475 downregulated) were identified. Enrichment analyses highlighted associations with embryonic organ development, oxidative phosphorylation, angiogenesis, and oxidative stress. Immune infiltration analysis revealed altered eosinophil and macrophage polarization in PIH. LASSO regression selected 12 key genes, with CDK7 showing the most significant upregulation in the PIH model. CDK7 knockdown promoted macrophage polarization toward the anti-inflammatory M2 phenotype.

These findings link CDK7 to immune dysregulation in PIH by modulating macrophage polarization, expanding our understanding of PIH's molecular mechanisms. The study's limitations include reliance on public datasets and in vitro models, warranting in vivo validation.

CDK7 emerges as a potential therapeutic target for PIH, offering new insights into immunoregulatory interventions for this complication.