Validation of Biomarkers and Immunotherapy With Crohn's Disease Using WGCNA and Two-Sample Mendelian Randomization Study

Abstract

Objective: Crohn's disease (CD) is a chronic systemic inflammatory disease that mainly affects the intestine, accompanied by extraintestinal symptoms and immune problems. The progression of the disease may cause permanent damage to the structure and function of the intestine. Due to unclear early symptoms and lack of precise detection methods, early diagnosis of CD is difficult. Many patients were diagnosis at late stage, which may lead to delayed treatment and increased risk of complications. Identifying hub genes related to CD and using them to predict CD is of great significance. Methods: DEG and WGCNA were employed to identify key genes associated with CD and to detect modules significantly linked to the disease. GO and KEGG analyses were conducted to explore the functions of these identified genes. Additionally, MR method was utilized to assess the causal relationships between the most significant gene and CD. Results: WCGNA identified 3240 differentially expressed genes, with the magenta module being the most significant among the nine clustered modules. The enrichment of GO and KEGG pathways indicates that the hub genes in the magenta module are related to the positive regulation of heme binding, tetrapyrrole binding, carboxylic acid binding, organic acid binding, IL-17 signaling pathway, and amoebiasis pathway. The Top 5 hub genes are CXCL1, LCN2, NOS2, S100A8, and DUOX2. Mendelian randomization analysis found a significant correlation between CXCL1 and CD. Conclusions: The study screened five potential biomarker genes in CD patients using a bioinformatics approach and Mendelian randomization study. Our results provided insights into CXCL1, LCN2, NOS2, S100A8, and DUOX2 in CD and suggested that CXCL1 may potentially be the optimal biomarker that could be a relatively easy path to clinical translation.

Figure 1

Visualization of DEGs in Crohn's disease. (a) Volcanic map and (b) heat map.

Figure 2

WGCNA analyzes important modules. (a) Cluster the dendrogram of genes. (b) The feature heatmap. (c) Positive correlation with CD in the scatter plot of the magenta module. (d) Module membership in magenta module.

Figure 3

Screening of intersecting genes. (a) The Venn diagram of intersecting genes. (b) Pathway analysis of KEGG. (c) GO enrichment analysis.

Figure 4

PPI network. Protein–protein interaction (PPI) network diagram of intersecting hub genes. The core gene map in protein interaction networks, where the depth of colors represents the score of genes.

Figure 5

Nomogram model constructed for predicting the risk of Crohn's disease:(a) Nomogram model constructs based on key central genes (hub genes) to predict disease risk. (b) the predictive accuracy of the model by calibrating the curve to ensure its reliability. (c) CD through receiver operating characteristic (ROC) curves. (d) Lollipop diagram of correlation between CXCL1 and infiltration immune cells in CD. (e) Scatter plots of correlation between CXCL1 and infiltration immune cells in CD.

Figure 6

The investigation of CXCL1's immunological function in Crohn's disease (a) Presented information on the relative proportionate distribution of 22 distinct immune cell types. (b) The variations in immune cell infiltration in Crohn's disease. (c) Examined the relationship between immune cell infiltration and CXCL1.

Figure 7

Results of Mendelian randomization study. (a) Scatter plot. (b) Forest plot. (c) Funnel plot. (d) Leave one plot.