Abstract

Background and aim: Cancer stem cells (CSCs) are essential challenge in recurrence-free treatment of colorectal cancer (CRC) Patients. Targeting CSCs related genes can provide effective treatment of CRC. However, it may be difficult to identification of appropriate target gene among a complex network of differentially expressed genes (DEGs). For this reason, in present study we analyzed the high throughput gene expression data through network-based approaches to achieve important therapeutic targets. Methods: A Dataset was taken from gene expression omnibus (GEO). Gene expression analysis was done via ExAtlas software. Gene enrichment analysis was done using Database for Annotation, Visualization and Integrated Discovery (DAVID) and g: profiler. Protein-protein interaction network (PPIN) was analyzed by Cytoscape 3.8.0 to find hub genes and significant modules. The clinical importance of candidate genes in predicting CRC patient's outcome was determined using ROC analysis and survival analysis. Results: A total of 501 unique DEGs (318 upregulated and 183 downregulated) were identified that were mainly enriched in molecular functions of RNA Polymerase 2, ribonucleoprotein, DNA and RNA binding and cellular processes of mRNA processing and cell cycle. Main enriched cellular compartments related to DEGs were intracellular regions, nucleoplasm and kinetochores. Also, DEGs were mainly involved in pathways of protein metabolism, post translational protein modification, generic transcription, mitosis, cellular response to external stimuli, interleukins signaling and PI3/Akt signaling. Network analysis revealed 25 hub genes that 17 of them were contributed in three significant modules and considered as key hub genes. However, only 7 key hub genes including RAB3IP, CENPN, NEDD1, CENPU, SPC25, POLR2D and CDC25A were able be to predict chemoradiotherapy response of CRC patients and considered as candidate genes. Among them, CDC25A was negatively correlated with CRC patient's survival. Conclusion: We found seven gene signatures associated with cancer stem cells that predict clinical response of CRC patients to conventional therapies. So, these genes, especially CDC25A, are potentially suitable therapeutic targets for inhibiting CSC proliferation and effective treatment (recurrence free) of CRC patients.