Development and Validation of a Prognostic Model based on 11 E3-related Genes for Colon Cancer Patients

Abstract

Background: Colon cancer is a common tumor in the gastrointestinal tract with a poor prognosis. According to research reports, ubiquitin-dependent modification systems have been found to play a crucial role in the development and advancement of different types of malignant tumors, including colon cancer. However, further investigation is required to fully understand the mechanism of ubiquitination in colon cancer. Methods: We collected the RNA expression matrix of the E3 ubiquitin ligase-related genes (E3RGs) from the patients with colon adenocarcinoma (COAD) using The Cancer Genome Atlas program (TCGA). The “limma” package was used to obtain differentially expressed E3RGs between COAD and adjacent normal tissues. Then, univariate COX regression and least absolute shrinkage and selection operator (LASSO) analysis were performed to construct the prognostic signature and nomogram model. Afterward, we used the original copy number variation data of COAD to find potential somatic mutation and employed the “pRRophetic” package to investigate the disparity in the effectiveness of chemotherapy drugs between high and low-risk groups. The RT-qPCR was also implied to detect mRNA expression levels in tumor tissues. Results: A total of 137 differentially expressed E3RG3 were screened and 11 genes (CORO2B, KCTD9, RNF32, BACH2, RBCK1, DPH7, WDR78, UCHL1, TRIM58, WDR72, and ZBTB18) were identified for the construction of prognostic signatures. The Kaplan-Meier curve showed a worse prognosis for patients with high risk both in the training and test cohorts (P = 1.037e-05, P = 5.704e-03), and the area under the curve (AUC) was 0.728 and 0.892 in the training and test cohorts, respectively. Based on the stratified analysis, this 11- E3RGs signature was a novel and attractive prognostic model independent of several clinicopathological parameters (age, sex, stage, TNM) in COAD. The DEGs were subjected to GO and KEGG analysis, which identified pathways associated with cancer progression. These pathways included the cAMP signaling pathway, calcium signaling pathway, Wnt signaling pathway, signaling pathways regulating stem cell pluripotency, and proteoglycans in cancer. Additionally, immune infiltration analysis revealed significant differences in the infiltration of macrophages M0, T cells follicular helper, and plasma cells between the two groups. Conclusion: We developed a novel independent risk model consisting of 11 E3RGs and verified the effectiveness of this model in test cohorts, providing important insights into survival prediction in COAD and several promising targets for COAD therapy.