Identification of latent core genes and pathways associated with myelodysplastic syndromes based on integrated bioinformatics analysis

Background:Myelodysplastic syndromes (MDS) are relatively common hematological malignancies characterized by dysplastic hematopoiesis in one or more of the lineages of the bone marrow. This study aimed to identify critical pathogenic biomarkers associated with the carcinogenesis and progression of MDS. Methods:To explore the candidate genes, the expression profiles of GSE2779, GSE4619, and GSE19429 were downloaded from the Gene Expression Omnibus (GEO) database, which contained CD34+ cells isolated from MDS patients and normal controls. The three microarray datasets were integrated to obtain differentially expressed genes (DEGs) and were deeply analyzed by bioinformatics methods. The construction of protein-protein interaction (PPI) network together with module analysis was performed based on Cytoscape software and the Search Tool for the Retrieval of Interacting Genes (STRING) database. Results:Our study identified 114 DEGs, which were highly enriched in various key pathways, including forkhead box protein O (FoxO) signaling pathway, the primary immunodeficiency, and hematopoietic cell lineage. Twelve core genes, such as FOXO1, PAX5 and CXCR4 were identified with a high degree of connectivity. It is plausible that FoxO signaling pathway plays an important role in MDS, and the dysregulation of FOXO1 was significantly associated with TGF beta, IL2/STAT5, Notch signaling and apoptosis pathways. Conclusion:The current study for the first time identified twelve latent indicators and their downstream targets, which might become significant biomarkers for worse clinical characteristics in MDS.