4.1 Gene Set Enrichment Analysis for All Expressed Genes
Using the GSEA software, we explored the distribution of pathway gene sets across PCOS cases and controls. In the PCOS patients, 96/156 gene sets were upregulated, and 48 were significantly enriched with FDR < 25%. In enrichment results, the T cell receptor pathway and B cell receptor pathway indicated that the gene expression in PCOS may associate with immune (Figure 1a, b).
4.2 DEGs and IRDEGs Identification in PCOS
After data were preprocessed, an expression matrix of 15,913 genes was obtained from the dataset. Among the 2131 DEGs identified between PCOS samples and normal controls, 1195 upregulated genes and 936 down-regulated genes were selected for subsequent analysis (Supplementary Figure S1a). The expression profile of the top 50 DEGs was intuitively visualized through a heatmap (Supplementary Figure S1b). After overlapping genes were removed from the above database, 1793 immune-related genes in Homo sapiens were obtained (Supplementary Table1). 2131 DEGs were intersected with 1793 immune-related genes, and 125 IRDEGs in PCOS were identified (Figure 2).
4.3 GO Enrichment Analysis of IRDEGs
An analysis of the IRDEGs in PCOS using g: Profiler revealed that there was significant dysregulation across all categories, which included 36 significant pathways for MF, 425 important pathways for BP, 26 effective pathways for CC, 36 significant pathways for KEGG, 21 significant pathways for REAC, and 16 significant pathways for WP (Figure 3a). The number of genes that were enriched in the different pathways of BP, MF, and CC was annotated by WebGestalt (Figure 3b). According to g: Profiler analysis, MF GO terms highlight "signaling receptor binding", “receptor ligand activity”, “signaling receptor activator activity”, “signaling receptor regulator activity”, and “cytokine activity” (Figure 4a). In GO BP analysis, immune processes were involved in most pathways, such as “response to stimulus”, “immune system process” and “immune response” (Figure 4b). As for CC, "extracellular region", " extracellular space" and "cell periphery" were the most significant terms (Figure 4c). A separate analysis was conducted on the biological functions of genes up-regulated and down-regulated. A study of GO BP data showed that up-regulated IRDEGs were significantly enriched in immune response, cell proliferation, and innate immunity (Figure 5a), while down-regulated IRDEGs primarily involved immune response, inflammation, and signal transduction (Figure 5d). MF also showed that up-regulated genes involved protein binding, calcium ion binding, and growth factor activity (Figure 5b), and down-regulated were mainly enriched in protein binding, receptor binding, and cytokine activity (Figure 5e). Additionally, CC analysis revealed that upregulated IRDEGs increased in the extracellular space, the extracellular region, and the plasma membrane (Figure 5c). Protein binding, receptor binding, and growth factor activity were down-regulated genes (Figure 5f). The multiple biological pathways could be involved in PCOS pathogenesis associated with these genes.
4.4 Pathway Enrichment Analysis of IRDEGs
Profiler analyses were performed on KEGG pathways, REAC pathways, and WikiPathways for all IRDEGs. (Figure 6a). It appears that inflammatory response and immune regulation were the most frequently used terms in all three datasets, such as TNF signaling, cytokine-cytokine receptor interaction, and cytokine signaling in the immune system. To assess their distinctive functions, we analyzed up-regulated and down-regulated genes using WebGestalt and DAVID analysis. Based on WebGestalt overrepresentation analysis showed that up-regulated genes tended to be enriched in the T cell receptor signaling pathway and the estrogen signaling pathway, while the down-regulated genes were mainly in TNF signaling pathway and cytokine-cytokine receptor interaction (Figure 6b, c). According to DAVID analysis, up-regulated genes were enriched in T cell receptors and axon guidance. The down-regulated pathways were also enriched in TNF signaling and the interaction between cytokines and cytokine receptors (Figure 6d, e). The TCR signaling pathway, TNF signaling pathway, and cytokine-cytokine receptor interaction were significantly enriched.
4.5 PPI Network Construction and Module Analysis Identified
After eliminating unconnected nodes, the PPI network was constructed by combining pairs of IRDEG with a combined score > 0.4 (Figure 7a). The PPI network of IRDEGs, comprising 101 nodes (DEGs) and 413 edges, was visualized using Cytoscape (Figure 7b). As a result of MCODE's analysis of the PPI network, three significant modules were identified (Figure 7c). From the PPI network complex, the most highly connected subnetwork is cluster rank 1, with a score of 9.000 (Figure 7d). There were 21 (FOS, NFKBIA, HLA-B, BST2, PTGS2, HLA-A, CXCR4, TNFAIP3, CXCL6, ISG15, CCL20, HLA-C, IL18, CXCL3, NOS2, OAS1, RSAD2, PSMB8, IRF5, IL15, IL1RN) nodes and 90 interactions in this functional module, which was considered critical. The expression of these genes were shown in Figure 7e. These genes were described in detail in Supplementary Table2, along with their molecular functions. Additionally, Metascape results showed that these candidate hub genes are associated primarily with immune responses (Figure 7f). An analysis of correlations was conducted to investigate the whole interrelationship of 21 genes (Figure 7g).
4.6 TFEA of Hub IRDEGs Involved in Regulation of PCOS
To screen for genes that may regulate PCOS, TFEA by ChEA3 identified 1632 transcription factors (Supplementary Table3), and the top 10 were shown in Table 1. Among the ten genes, RELB, CSRNP1, and NFKB2 in these ten genes were expressed in our data (Table 2). Venn graph showed common genes regulated by these three TFs. These five genes, HLA-A, HLA-B, FOS, PTGS2, and CXCR4, were identified as hub genes (Figure 8a). PPI networks were revealed by co-expression analysis using the GeneMANIA database. The interactions were 77.64% physical, 8.01% co-expression, 3.63% colocalization, 2.87% genetic, 1.88% pathway, and 0.60% shared protein domains, prostaglandin biosynthesis, and antigen presentation were the primary functions of these enzymes (Figure 8b).
4.7 Identification of Differential Immune Cell Types in PCOS
As outlined above, the hub genes showed high immunological enrichment. Therefore, we used ImmuCellAI to predict the abundance of 24 different immune cell types in samples. And the variation in immune cell infiltration in various groups was examined while the immune cell abundance in groups was checked (Figure 9a, b). Monocyte cells, nTreg cells, iTreg cells, and Tcm cells were all significantly altered in this study.
4.8 Analysis of m6A RNA Methylation Regulators in PCOS
RNA metabolisms, including stability, translation, degradation, transport, and splicing, have all been impacted by m6A alteration.(37,38). Thus, we analyzed whether PCOS m6A regulators were related to the expression of these hub genes. As shown in Figure 10a, 19 m6A regulators expressed in PCOS are based on the GSE155489 dataset. Between PCOS and control samples, three m6A regulators (YTHDF2, HNRNPA2B1, and ALKBH5) were significantly altered (Figure 10b). Afterward, we conducted a correlation analysis between 19 genes associated with m6A and hub gene expression (Figure 10c). A significant correlation was found between HLA-B, FOS, and CXCR4 expression and three genes associated with m6A. Readers YTHDF2 and HNRNPA2B1 were present, and a demethylase, ALKBH5, was present. In regulating immune cells, m6A is increasingly thought to play a crucial role through several mechanisms.