Identifying cancer-causing noncoding RNAs

To circumvent the dependency on prediction models, we developed a microRNA-screen-based assay to establish links between cellular phenotypes and microRNAs (miRNAs). To this end, a miRNA expression library (miR-Lib) was built consisting of 300 annotated miRNAs and around 100 candidate miRNAs. These miRNA 'minigenes' were cloned from genomic DNA of human cells and inserted in a specially engineered retroviral expression vector. This vector, named miR-Vec, is a modified murine stem cell virus (pMSCV) vector, with the expression of the miRNA minigenes under the control of the CMV pol II promoter. The vast majority of naturally transcribed miRNA genes are known to be expressed by pol II promoters. In parallel with the miRNA library, a corresponding microarray was developed (miR-Array) containing DNA spots of the miRNA minigenes from the library. By combining the miRNA library and microarray tools, we designed and performed three distinct functional genetic approaches to identify cancerous microRNAs