Identification and validation of paraptosis-related biomarkers in recurrent miscarriage

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.

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.

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.

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.