DGCR8 is required for microRNA maturation

To determine whether DGCR8 is required for maturation of all miRNAs, we performed miRNA microarray analysis. Using RNA from wild-type ES cells as our reference sample, we observed a global loss of miRNAs in DGCR8 knockout cells, but normal levels of expression in DGCR8 heterozygous cells. The similarity in expression levels between wild-type and heterozygous cells suggests that DGCR8 is not limiting in the maintenance of steady-state levels of miRNAs in ES cells. Of the eighty-nine miRNA array probes that showed significant signals with wild-type RNA, eighty-two were drastically reduced in the DGCR8 knockout cells. The remaining seven were not significantly altered in the DGCR8 knockout cells, but at least four of these seven miRNAs appear to be due to unavoidable contamination with RNA from the mouse embryonic fibroblast (MEF) feeder cells used for ES cell culture prior to the isolation of RNA. These miRNAs were highly expressed in the MEFs and decreased when the ES cells were temporarily passaged on gelatin-coated plates without feeders. The remaining three showed similar levels of expression in the heterozygous, knockout and MEF feeder cells. Therefore, these signals may be small RNAs that are not processed with the help of the microprocessor complex. Alternatively, these signals may result from unavoidable degradation products within the purified small RNA population. In either case, our results show that DGCR8 is broadly required for miRNA processing with little evidence for redundancy or bypass mechanisms. Keywords: cell type comparison, genetic modification