Credit: DIGITAL VISION

Lower levels of adenosine-to-inosine (A-to-I) editing of mRNAs have been observed in some cancers, particularly in high-grade gliomas. The microRNA (miRNA) cluster miR-376 is subject to A-to-I editing in the brain; however, whether miR-376 editing is disrupted in gliomas, and whether this has relevant functional consequences is not known.

miR-376a* that cannot be edited ... increased in vitro migration and invasion

Shu Wang and colleagues identified reduced editing of miR-376a* (one of five mature miRNAs produced by the cluster) in high-grade human glioma samples, and this correlated with a higher tumour volume on patient magnetic resonance imaging scans. They also showed that miR-376a* editing occurs in glioma cell lines, and used U87 cells to further characterize the biological effects of miR-376a* editing. U87 cells that were selected for metastatic activity by tail-vein injections and reculturing in vitro had higher levels of unedited miR-376a* than parental (less metastatic) U87 cells. Stable expression of miR-376a* that cannot be edited (miR-376a*A) in glioma cell lines increased in vitro migration and invasion, and this was reduced by a stably edited miR-376a* (miR-376a*G; guanosine mimics inosine functionally in miRNAs). Knockdown of miR-376a* in the highly metastatic U87 cells also reduced invasion and migration. Injection of miR-376a*A-expressing U87 cells into the brains of nude mice led to the formation of more invasive tumours and reduced survival time of the animals, compared with tumours formed from miR-376a*G-expressing cells.

The authors used microarray profiling to examine which gene targets are modified by edited and unedited miR-376a*, and identified the RAS family member RAP2A as a potential miR-376a*A target and autocrine motility factor receptor (AMFR) as a potential miR-376a*G target. They confirmed the specificity of each miRNA for its respective target, and showed that unedited miR-376a*A, but not miR-376a*G, reduced the levels of RAP2A mRNA and protein in U87 and SW1783 cells. Conversely, miR-376a*G, but not miR-376a*A, reduced the levels of AMFR mRNA and protein. Additional in vitro experiments showed that RAP2A or AMFR knockdown phenocopied the effects of miR-376a*A or miR-376a*G, respectively, on migration and invasion. Furthermore, RAP2A expression could rescue miR-376a*A-induced invasion, and AMFR expression increased invasion in miR-376a*G-expressing cells.

Unedited miR-376a* levels were inversely correlated with RAP2A mRNA levels and positively correlated with AMFR levels in human and mouse tumour samples. In patient data from publicly available databases, low RAP2A levels were observed in high-grade glioma and significantly correlated with reduced patient survival, suggesting that this miRNA editing and target alteration pathway could be relevant in human cancer.