Identification of subclusters and prognostic genes based on GLS-associated molecular signature in ulcerative colitis

Yang Xie,Jun Li,Qing Tao,Yonghui Wu,Zide Liu,Youxiang Chen,Chunyan Zeng
DOI: https://doi.org/10.1038/s41598-024-63891-2
IF: 4.6
2024-06-09
Scientific Reports
Abstract:Ulcerative colitis (UC) is a chronic and recurrent inflammatory disease that affects the colon and rectum. The response to treatment varies among individuals with UC. Therefore, the aim of this study was to identify and explore potential biomarkers for different subtypes of UC and examine their association with immune cell infiltration. We obtained UC RNA sequencing data from the GEO database, which included the training set GSE92415 and the validation set GSE87473 and GSE72514. UC patients were classified based on GLS and its associated genes using consensus clustering analysis. We identified differentially expressed genes (DEGs) in different UC subtypes through a differential expression analysis of the training cohort. Machine learning algorithms, including Weighted Gene Co-Expression Network Analysis (WGCNA), Least Absolute Shrinkage and Selection Operator (LASSO), and Support Vector Machine Recursive Feature Elimination (SVM-RFE), were utilized to identify marker genes for UC. The CIBERSORT algorithm was used to determine the abundance of various immune cells in UC and their correlation with UC signature genes. Finally, we validated the expression of GLS through in vivo and ex vivo experiments. The expression of GLS was found to be elevated in patients with UC compared to normal patients. GLS and its related genes were able to classify UC patients into two subtypes, C1 and C2. The C1 subtype, as compared to the C2 subtype, showed a higher Mayo score and poorer treatment response. A total of 18 DEGs were identified in both subtypes, including 7 up-regulated and 11 down-regulated genes. Four UC signature genes (CWH43, HEPACAM2, IL24, and PCK1) were identified and their diagnostic value was validated in a separate cohort (AUC > 0.85). Furthermore, we found that UC signature biomarkers were linked to the immune cell infiltration. CWH43, HEPACAM2, IL24, and PCK1 may serve as potential biomarkers for diagnosing different subtypes of UC, which could contribute to the development of targeted molecular therapy and immunotherapy for UC.
multidisciplinary sciences
What problem does this paper attempt to address?
The paper aims to address the issue of identifying potential biomarkers for different subtypes in patients with ulcerative colitis (UC) and to explore the association between these biomarkers and immune cell infiltration. Specifically, the main objectives of the study include: 1. **Identifying and exploring different subtypes of UC and their potential biomarkers**: By analyzing gene expression data, particularly genes related to glutamine synthetase (GLS), to identify different subtypes of ulcerative colitis patients. 2. **Evaluating the role of these biomarkers in immune cell infiltration**: By identifying feature genes (such as CWH43, HEPACAM2, IL24, and PCK1) through machine learning algorithms and validating their value in diagnosing different UC subtypes. 3. **Developing more targeted molecular therapies and immunotherapies**: Based on the discovered biomarkers, providing new directions for the treatment of ulcerative colitis. The researchers utilized datasets from multiple gene expression databases (GEO databases) and classified UC patients into different subtypes through consensus clustering analysis. Further, through differential gene expression analysis, weighted gene co-expression network analysis (WGCNA), least absolute shrinkage and selection operator (LASSO), and support vector machine recursive feature elimination (SVM-RFE), they identified biomarkers with diagnostic value. Finally, the CIBERSORT algorithm was used to assess immune cell infiltration in different UC subtypes and to explore the relationship between feature gene expression and immune cell infiltration.