Bioinformatics analysis of shared biomarkers and immune pathways of preeclampsia and periodontitis

Abstract

Abstract Background Epidemiological evidence indicates that preeclampsia (PE) is associated with comorbidities such as periodontitis (PD). However, the underlying mechanism remains unclear. To enhance our understanding of their co-pathogenesis, this research investigated the shared biomarkers and pathological mechanisms. Methods We systematically retrieved transcriptomic datasets from the Gene Expression Omnibus database. These datasets encompass a comparative analysis of the periodontium with and without PD and of the placenta with and without PE. Differentially Expressed Genes Analysis and Weighted Gene Go-expression Network Analysis (WGCNA) were used to identify the key crosstalk genes in patients with PD and PE. The functional characterisation of these genes was performed using enrichment analysis. Protein–protein interaction networks and machine learning methods were leveraged to identify shared hub genes. The XG-Boost algorithm was applied to construct diagnostic models to gain insight into disease aetiology. The identified genes were validated by single-cell RNA sequencing to ensure their robustness and biological relevance. Results A total of 55 key crosstalk genes were identified, which were primarily enriched in immune-related pathways by using limma and WGCNA. Among them, twenty-four shared hub genes were identified using protein–protein interaction analysis and machine learning methods. The diagnostic model constructed using immune-related genes outperformed the other two models (area under the receiver operating characteristic curve [ROC] = 0.7786 and 0.7454 for PE and PD, respectively). Pathways involving these genes were mapped using the Kyoto Encyclopedia of Genes and Genomes analysis. In addition, single-cell RNA sequencing analysis showed that the expression of BIN2, LYN, PIK3AP1, and NEDD9 in neutrophils was significantly downregulated, and LYN in fibroblasts and endothelial cells was consistently upregulated. Conclusions Shared hub genes and immunologic pathways were identified in PE and PD, characterised by crosstalk between BIN2, LYN, NEDD9, and PIK3AP1, suggesting the pathogenesis of PE and PD, which could pave the way for the development of effective diagnostic, treatment, and management strategies.

Keywords

• Preeclampsia
• Periodontitis
• Bioinformatics
• Machine learning
• Single cell
• Immune pathway