Significant correlation between ferroptosis-associated genes and ulcerative colitis: implications for diagnosis and treatment

Abstract

Background: Ulcerative colitis (UC) is a significant type of inflammatory bowel disease (IBD). Ferroptosis is a type of procedural death that is associated with different types of diseases to various degrees. This study aimed to explore the key ferroptosis-associated genes in UC and assess their potential implications for biological therapy.

Methods: Bioinformatic prediction combined with experimental verification were applied to this study for the correlation between ferroptosis and UC. The Gene Expression Omnibus database was used to obtain disease-related gene expression data and then the identified genes expression were validated using Dextran Sulfate Sodium Salt (DSS) induced mouse colitis model. Next, the infiltration of immune cells was compared between UC and normal groups, and the correlation between different immune cells and the identified genes was analyzed.

Results: In this study, 91 ferroptosis-associated genes associated with UC were identified. Four genes (TIMP1, LPIN1, SOCS1, and CD44) were captured with diagnostic values among them, whose expression in DSS-induced mouse colitis was higher than control tissues. Besides, the expression levels of these four genes in the biological agents responsive group showed a distinction from those in active UC patients. Immune infiltration analysis revealed that there were positive or negative correlations between four genes with neutrophils, M0 and M1 macrophages, B cells, and NK cells.

Conclusions: This study clearly indicates that immune cell infiltration in UC patients is associated with ferroptosis. The key genes associated with ferroptosis may have diagnostic value and predict therapeutic effects, which also can be helpful for discovering the molecular mechanism of UC.

Keywords: Biological agents; Biomarkers; Ferroptosis; Machine learning; Ulcerative colitis.

Fig. 1

The flowchart for the research

Fig. 2

Screening ulcerative colitis-related genes. A Volcano plot illustrating DEGs between active UC patients and healthy controls. Up-regulated genes are marked in red, down-regulated genes in green, and genes without significant differences in black. B Heatmap showcasing the top 50 DEGs between active UC patients and healthy controls, with a color gradient from red to green indicating high to low expression levels. C Intersection of ferroptosis-related genes with the DEGs in UC reveals 91 overlapping genes. D KEGG pathway enrichment analysis conducted on the 91 overlapping genes. E GO enrichment analysis of the 91 overlapping genes. F Circular visualization of the GO enrichment analysis

Fig. 3

Identification of ferroptosis-associated hub genes and immune-infiltrating landscape of active UC. A, B Identification of key DEFRGs using LASSO regression. A Depicts the LASSO coefficients across varying log lambda values in the metadata cohort. B Illustrates the binomial deviance profiles of DEFRGs within the same cohort). C, D Exploration of key DEFRGs through SVM. E Identification of four intersecting genes via both methodologies. F Illustration of immune infiltration variances between UC patients and control samples, with red indicators highlighting significantly different subpopulations. G Analysis of correlation patterns among immune infiltrates

Fig. 4

The impact of infliximab and vedolizumab therapies on colonic mucosal gene expression. A-D Expression levels of TIMP1, LPIN1, SOCS1, and CD44 in the intestinal mucosa across three groups: healthy controls, patients responding to Infliximab treatment, and active UC patients in dataset GSE73661. E–H. Expression patterns of TIMP1, LPIN1, SOCS1, and CD44 in healthy controls, patients responding to Vedolizumab treatment, and active UC patients from the same dataset

Fig. 5

Experimental verification in the UC mouse model. A, B Changes in body weight of mice in the normal and dextran sulfate sodium (DSS) groups. C Colon length in the control group versus that in the DSS group. D DAI scores increased after acute administration of DSS. E Histological inflammation scores increased after DSS treatment. (*p < 0.05, **p < 0.01, ***p < 0.001, and N = 10 mice per group)

Fig. 6

Experimental verification in the UC mouse model. A The expression levels of CD44, TIMP1, LPIN1 and SOCS1 were detected using real-time quantitative polymerase chain reaction. BThe expression levels of CD44, TIMP1, LPIN1 and SOCS1 were detected by Western blot analysis in the UC mouse model. (*p < 0.05, **p < 0.01, ***p < 0.001, and N = 6 mice per group)

Fig. 7

Experimental verification in the UC mouse model. The expression levels of CD44, TIMP1, LPIN1 and SOCS1 were detected using immunohistochemistry. N = 6 mice per group