Abstract
OBJECTIVE: To explore novel related genes and potential biomarkers of pancreatic neuroendocrine tumors (PanNETs). MATERIALS AND METHODS: Two data sets from ICGC and two from the NCBI GEO database were used to identify the differentially expressed genes (DEGs) in PanNETs. The common DEGs among the four sources were analyzed; furthermore, the relationship of these gene expression patterns with different PanNET grades, their mutation status and corresponding impact on prognosis, the interaction network, and the relationship with three known PanNET genes (ATRX, DAXX, and MEN1) were analyzed by two other GEO data and cBioPortal database. Finally, the expressions of novel DEGs were validated in Chinese PanNET tissues by RT-qPCR. RESULTS: Five new DEGs (ABCC8, PCSK2, IL13RA2, KLKB1, and PART1) and one confirmed DEG-ISL1 were identified. The mutation counts of DEGs increased with the tumor grade increasing from G1 to G3, and PanNET patients present vascular invasion or are deceased. These DEG expression patterns in PanNETs are quite different from that of pancreatic ductal adenocarcinoma and are related to A-D-M (ATRX-DAXX-MEN1) mutation. ABCC8 and KLKB1 are co-occurrence with the A-D-M axis in PanNETs. Importantly, patients with DEG mutations have a lower survival rate. RT-qPCR verification results of KLKB1 (P < 0.01), IL13RA2 (P < 0.01), ABCC8 (P < 0.01), and PART1 (P < 0.0001) expressions in Chinese PanNET tissues are consistent with our database analysis, which were significantly up-regulated. However, the expression of PCSK2 (P < 0.01) was contrary to our bioinformatics analysis, which was significantly down-regulated, suggesting that the expression trend of PCSK2 may be different among different races. These results indicated that these five genes may play an important role in the occurrence and progression of PanNETs. CONCLUSION: Five novel common DEGs identified are related to the development and prognosis of PanNETs and may serve as specific biomarkers and therapeutic targets.