Abstract: BACKGROUND: The origin of chronic myeloid leukemia lies in clonal hematopoietic stem cells, and it is marked by uncontrollable growth of myeloid cells, mostly caused by the BCR-ABL1 fusion gene. Although imatinib has significantly improved patient survival, drug resistance remains a major obstacle to treatment.
OBJECTIVE: To identify genes associated with imatinib resistance in chronic myeloid leukemia using bioinformatics analysis within the Gene Expression Omnibus database and to explore the mechanisms of resistance.
METHODS: The present study utilized the Gene Expression Omnibus database, which was created and is currently maintained by the National Center for Biotechnology Information. Two datasets, GSE267522 and GSE174800, were retrieved from the Gene Expression Omnibus database, and each of the two datasets comprised three imatinib-resistant samples and three imatinib-sensitive samples. First, the GEO2R tool was used to screen the differential genes, and the DAVID tool was used to analyze them for Gene Ontology and Kyoto Encyclopedia of Genes and Genomes enrichment. A protein interaction network framework was constructed using the STRING database, and the top 10 hub genes with the highest connectivity values were identified from the network using Cytoscape software. Furthermore, a weighted gene co-expression network analysis algorithm was used to identify key module feature genes. Venn diagram analysis was performed with these genes and the aforementioned 10 hub genes, and the intersection genes were selected as core genes. Finally, a K562 imatinib resistance model was constructed and validated using real-time fluorescence quantitative PCR and western blotting.
RESULTS AND CONCLUSION: (1) A total of 273 differential genes were screened between the two datasets, of which 81 were downregulated and 192 were upregulated. (2) Gene ontology enrichment analysis revealed that the differential genes were involved in immune response and T-cell receptor signaling. Focusing on the cellular component level, cell components like the outer and inner plasma membranes and cellular exosome showed significant enrichment. Molecular function analysis indicated that the differential genes were involved in interactions between transmembrane receptor proteins and actin. (3) Kyoto Encyclopedia of Genes and Genomes enrichment analysis showed that the differential genes were enriched in the hematopoietic cell lineage, phosphatidylinositol 3-kinase/protein kinase B signaling pathway and cancer pathway. (4) Cytoscape software screened out the top 10 differentially expressed genes by connectivity and intersected them with key module signature genes obtained using a weighted gene co-expression network analysis algorithm. The resulting intersection genes obtained included IRS1, CD52, CD53, CORO1A, KIT, LAPTM5, and PECAM1. (5) The K562 Imatinib resistance beads were successfully constructed. The real-time fluorescence quantitative PCR results showed that compared with the K562 group, the mRNA expression of CD52, CD53, CORO1A, and PECAM1 was significantly increased in the imatinib-resistant K562 group (P < 0.05), while the mRNA expression of IRS1 was significantly decreased (P < 0.05). Furthermore, western blotting revealed increased expression of CD52, CD53, CORO1A, and PECAM1 proteins (P < 0.05), and decreased expression of IRS1 protein (P < 0.05) in the imatinib-resistant K562 strain (consistent with the real-time fluorescence quantitative PCR results). (6) Differences in the expression of K562 imatinib resistance core genes may provide new insights into the mechanisms of imatinib resistance in chronic myeloid leukemia in the future.

Key words: chronic myeloid leukemia, tyrosine kinase inhibitor, imatinib, drug resistance gene, gene expression, bioinformatics, weighted gene co-expression network analysis, protein interaction network