Circular RNA (circRNA), a type of non-coding RNA, has received a great deal of attention with regard to the initiation and progression of tumors. However, the molecular mechanism and function of circRNAs in breast cancer (BC) remain unclear. In the current study, we discovered that hsa_circ_0028899 (also called circRNF10) was significantly reduced in BC tissues, and a higher level of circRNF10 was markedly related to a favorable prognosis. The results of CCK8, colony formation, Transwell, ELISA, and NK cell-mediated cytotoxicity assays indicated that increased circRNF10 expression could significantly repress the proliferation, invasion, and migration of BC cells and enhance the killing efficiency of NK cells against BC cells. According to these biological functions, the possible role and molecular mechanism of circRNF10 in BC cells were further investigated. We used bioinformatics prediction tools to predict circRNF10-bound miRNAs, which were verified by many experimental studies, including FISH, luciferase reporter assays, RIP, and Western blots. These data suggest that circRNF10 serves as a molecular sponge for miR-934 to further regulate PTEN expression and PI3k/Akt/MICA signaling in vitro and tumor growth in vivo. Altogether, these findings reveal that circRNF10 functions as a novel anti-oncogene in BC via sponging miR-934 and suppressing the PI3K/Akt/MICA pathway.

Background: Circular RNAs (circRNAs) have drawn lots of attention in tumorigenesis and progression. However, circRNAs as crucial regulators in multitudinous biological processes have not been systematically identified in breast cancer (BC). Our research aims to explore novel circRNAs in BC and their mechanisms of action.Methods: The circRNA expression profile data, as well as RNA-sequencing data of BC, were downloaded from public database, respectively. The differentially expressed circRNAs, miRNA, and mRNA were determined via fold change filtering. The competing endogenous RNAs (ceRNAs) network were established on the foundation of the relationship between circular RNAs, miRNAs and mRNAs. GO and KEGG analysis of the overlapped genes were performed to predict the potential functions and mechanisms of circRNAs in BC. The CytoHubba was used to determine the hub genes from the PPI regulatory network. Morever, we further used Kaplan–Meier plotter to perform survival analysis of these hub genes. Real-time PCR was used to validate the expression of the circRNAs in BC tissues.Results: A total of seven differential expressed circRNAs were screened. After the predicted target miRNA and DEmiRNA were intersected, four circRNA-miRNA interactions including three circRNAs and four miRNAs were determined. Furthermore, the Venn diagram was used to intersect the predicted target genes and the downregulated differentially expressed genes, and screened 149 overlapped genes. Moreover, we constructed a PPI network, and selecting six hub genes, including DGAT2, ACSL1, ADIPOQ, LPL, LEP, PCK1. Moreover, the survival analysis results revealed that low expression of ADIPOQ, LPL, LEP were obviously correlated with poor prognosis of BC patients. The real-time PCR results demonstrated that, the levels of circ_0028899, circ_0000375, and circ_0000376 were significantly down-regulated in breast cancer tissues.Conclusions: Our study constructed and analyzed a circRNA-associated ceRNA regulatory network and discovered that circ_0028899, circ_0000375, and circ_0000376 may function as ceRNAs to serve key roles in BC.