Transcriptomic analysis identifies diagnostic genes in polycystic ovary syndrome and periodontitis

Abstract

Purpose: To investigate underlying co-mechanisms of PCOS and periodontitis through transcriptomic approach.

Methods: PCOS and periodontitis gene expression data were downloaded from the GEO database to identify differentially expressed genes. GO and KEGG pathway enrichment analysis and random forest algorithm were used to screen hub genes. GSEA analyzed the functions of hub genes. Correlations between hub genes and immune infiltration in two diseases were examined, constructing a TF-ceRNA regulatory network. Clinical samples were gathered from PCOS and periodontitis patients and RT-qPCR was performed to verify the connection.

Results: There were 1661 DEGs in PCOS and 701 DEGs in periodontitis. 66 intersected genes were involved and were enriched in immune and inflammation-related biological pathways. 40 common genes were selected from the PPI network. RF algorithm demonstrated that ACSL5, NLRP12, CCRL2, and CEACAM3 were hub genes, and GSEA results revealed their close relationship with antigen processing and presentation, and chemokine signaling pathway. RT-qPCR results confirmed the upregulated gene expression in both PCOS and periodontitis.

Conclusion: The 4 hub genes ACSL5, NLRP12, CCRL2, and CEACAM3 may be diagnostic genes for PCOS and periodontitis. The created ceRNA network could provide a molecular basis for future studies on the association between PCOS and periodontitis.

Keywords: Diagnosis; GSEA; Periodontitis; Polycystic ovary syndrome; Transcriptomic analysis.

Fig. 1

Identification of common genes involved in PCOS and periodontitis. A Volcano plot of differentially expressed genes in training set 1; B volcano plot of differentially expressed genes in training set 2; C Venn diagram of co-unregulated genes of training set 1 and training set 2; D Venn diagram of co-downregulated genes of training set 1 and training set 2

Fig. 2

Genes had significant contributions to the RF models. A Boxplot describes the residuals within a cluster in PCOS; B boxplot describes the residuals within a cluster in periodontitis; C feature importance of genes to the RF models in PCOS; D feature importance of genes to the RF models in periodontitis; the horizontal axis of both graphs represents RMSE loss after permutations. The vertical axis is gene names

Fig. 3

Correlation networks and expressions of hub genes in PCOS and periodontitis, and co-expression network of hub genes. A Interactions between hub genes in PCOS training set; B interactions between hub genes in periodontitis training dataset; C expression of hub genes between healthy and PCOS samples. Red and blue represent PCOS and healthy samples, respectively; D expression of hub genes between healthy and periodontitis samples. Red and blue represent periodontitis and healthy samples, respectively; E co-expression network of hub genes

Fig. 4

GSEA enrichment analysis of hub genes. A KEGG pathway analysis of hub gene ACSL5; B KEGG pathway analysis of hub gene NLRP12; C KEGG pathway analysis of hub gene CCRL2; D KEGG pathway analysis of hub gene CEACAM3

Fig. 5

The correlation between the hub gene and immune infiltration. A Immunocyte proportional heatmap of 10 immune cells between healthy and PCOS samples; B immunocyte proportional heatmap of 10 immune cells between healthy and periodontitis samples; C correlations between immune cells in PCOS; D correlations between immune cells in periodontitis; E proportion differences of immune cells between healthy and PCOS samples; F proportion differences of immune cells between healthy and periodontitis samples; G correlations between immune cells and diagnostic genes in PCOS; H correlations between immune cells and diagnostic genes in periodontitis

Fig. 6

The expression levels of four hub genes in PCOS oocytes and periodontitis gingiva samples were detected by RT-qPCR compared to controls. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001