Identification and validation of paraptosis-related biomarkers in recurrent miscarriage.

Background: Recurrent miscarriage (RM) is a pregnancy complication with growing evidence suggesting a role for paraptosis in its pathogenesis, though the underlying mechanisms remain unclear. This study investigated paraptosis-related genes (PRGs) as potential therapeutic targets. Methods: Transcriptome data for RM were obtained from public databases, while PRGs were sourced from existing literature. Biomarkers were identified through the intersection of differential expression analysis, weighted gene co-expression network analysis, machine learning algorithms and expression validation, followed by the construction and validation of a nomogram. Molecular mechanisms of the biomarkers were further explored through immune infiltration, enrichment analysis, and the construction of regulatory networks. Single-cell RNA sequencing (scRNA-seq) was performed for deeper insights into RM. Results: PCNPP3 and ELOA were selected as biomarkers related to paraptosis. A predictive nomogram was developed with strong accuracy. Enrichment analysis revealed that both PCNPP3 and ELOA were associated with E2F targets and the G2M checkpoint. In immune infiltration analysis, PCNPP3 exhibited a significant positive correlation with smooth muscle cells, while ELOA was notably associated with myocytes. Regulatory network analysis suggested that NEAT1 and NPPA-AS1 might modulate ELOA expression via hsa-miR-49-5p. ScRNA-seq analysis identified decidual natural killer (dNK) cells and macrophages as key cell types, with ELOA expression decreasing in dNK cells as their state changed, while in macrophages, expression followed a pattern of increase, decrease, and increase again. Conclusion: This study identified PCNPP3 and ELOA as biomarkers of RM and provides comprehensive insights into their molecular mechanisms, offering valuable perspectives for future RM research.