Upregulation of ECT2 Predicts Adverse Clinical Outcomes and Increases 5-Fluorouracil Resistance in Gastric Cancer Patients

An integrated analysis of prognostic mRNA signature in early- and progressive-stage gastric adenocarcinoma

The pathogenesis and vital factors of early and progressive stages of stomach adenocarcinoma (STAD) have not been fully elucidated. In order to discover novel and potential targets to guide effective treatment strategies, a comprehensive bioinformatics study was performed, and the representative results were then validated by quantitative polymerase chain reaction (qPCR) and immunohistochemical (IMC) staining in clinical samples. A total of 4,627, 4,715, and 3,465 differentially expressed genes (DEGs) from overall-, early-, and progressive-stage STAD were identified, respectively. Prognostic models of 5-year OS were established for overall-, early-, and progressive-stage STAD, and ROC curves demonstrated AUC values for each model were 0.73, 0.87, and 0.92, respectively. Function analysis revealed that mRNAs of early-stage STAD were enriched in chemical stimulus-related pathways, whereas remarkable enrichment of mRNAs in progressive-stage STAD mainly lay in immune-related pathways. Both qPCR and IHC data confirmed the up-regulation of IGFBP1 in the early-stage and CHAF1A in progressive-stage STAD compared with their matched normal tissues, indicating that these two representative targets could be used to predict the prognostic status of the patients in these two distinct STAD stages, respectively. In addition, seven mRNAs (F2, GRID2, TF, APOB, KIF18B, INCENP, and GCG) could be potential novel biomarkers for STAD at different stages from this study. These results contributed to identifying STAD patients at high-risk, thus guiding targeted treatment with efficacy in these patients.

ECT2 promotes malignant phenotypes through the activation of the AKT/mTOR pathway and cisplatin resistance in cervical cancer

Epithelial cell transforming sequence 2 (ECT2) is expressed at high levels in various malignancies and contributes to malignant phenotypes in cancers. However, ECT2 is still not fully understood regarding its function and carcinogenic mechanism in cervical cancer. This research indicated that ECT2 expression was elevated in cervical cancer based on bioinformatics analysis and clinical specimens. Experiments in vitro and in vivo confirmed that ECT2 knockdown could suppress the proliferation and metastasis of cervical carcinoma cells. In addition, we found that silencing ECT2 could enhance the sensitivity to cisplatin and promote cell apoptosis. Mechanistically, we observed that ECT2 knockdown could inhibit the AKT/mTOR pathway and activate apoptosis, while ECT2 overexpression induced the opposite effect. The relationship between ECT2 and AKT was further confirmed by immunoprecipitation and rescue experiments. We found that the ECT2 and AKT could interact to form a complex, and knockdown AKT could offset all of the effects induced by ECT2. Our study emphasized the key point of ECT2 in the reversal of cisplatin resistance, and ECT2 could become a potential therapeutic target in cervical cancer.

Identification of key biomarkers for STAD using filter feature selection approaches

Gastric cancer (GC) is the fifth most common cancer and the third leading cause of cancer death worldwide. Discovery of diagnostic biomarkers prompts the early detection of GC. In this study, we used limma method combined with joint mutual information (JMI), a machine learning algorithm, to identify a signature of 11 genes that performed well in distinguishing tumor and normal samples in a stomach adenocarcinoma cohort. Other two GC datasets were used to validate the classifying performances. Several of the candidate genes were correlated with GC tumor progression and survival. Overall, we highlight the application of feature selection approaches in the analysis of high-dimensional biological data, which will improve study accuracies and reduce workloads for the researchers when identifying potential tumor biomarkers.