No abstract

MicroRNAs (miRNAs) are a class of noncoding RNAs that post-transcriptionally regulate gene expression in plants and animals. To investigate the influence of miRNAs on transcript levels, we transfected miRNAs into human cells and used microarrays to examine changes in the messenger RNA profile. Here we show that delivering miR-124 causes the expression profile to shift towards that of brain, the organ in which miR-124 is preferentially expressed, whereas delivering miR-1 shifts the profile towards that of muscle, where miR-1 is preferentially expressed. In each case, about 100 messages were downregulated after 12 h. The 3' untranslated regions of these messages had a significant propensity to pair to the 5' region of the miRNA, as expected if many of these messages are the direct targets of the miRNAs. Our results suggest that metazoan miRNAs can reduce the levels of many of their target transcripts, not just the amount of protein deriving from these transcripts. Moreover, miR-1 and miR-124, and presumably other tissue-specific miRNAs, seem to downregulate a far greater number of targets than previously appreciated, thereby helping to define tissue-specific gene expression in humans.

A global decrease in microRNA (miRNA) levels is often observed in human cancers 1,2 , indicating that small RNAs may have an intrinsic function in tumour suppression. To identify miRNA components of tumour suppressor pathways, we compared miRNA expression profiles of wildtype and p53-deficient cells. Here we describe a family of miRNAs, miR-34a-c, whose expression reflected p53 status. Genes encoding miRNAs in the miR-34 family are direct transcriptional targets of p53, whose induction by DNA damage and oncogenic stress depends on p53 both in vitro and in vivo. Ectopic expression of miR-34 induces cell cycle arrest in both primary and tumour-derived cell lines, which is consistent with the observed ability of miR-34 to downregulate a programme of genes promoting cell cycle progression. The p53 network suppresses tumour formation through the coordinated activation of multiple transcriptional targets, and miR-34 may act in concert with other effectors to inhibit inappropriate cell proliferation.The p53 tumour suppressor lies at a nexus of cellular pathways that sense DNA damage, cellular stress and improper mitogenic stimulation 3 . p53 integrates such signals and, in response, induces growth arrest, promotes apoptosis, blocks angiogenesis, or mediates DNA repair in a context-dependent manner 4 . The importance of p53 in preventing tumour formation is indicated by the presence of mutations in the p53 pathway in nearly all cancers 5 . Although p53 is most studied as a transcriptional activator, several reports have suggested that p53 represses the expression of specific genes 6 . Studies of p53-mediated Reprints and permissions information is available at www.nature.com/reprints.

Single-cell transcriptomics reveals gene expression heterogeneity but suffers from stochastic dropout and characteristic bimodal expression distributions in which expression is either strongly non-zero or non-detectable. We propose a two-part, generalized linear model for such bimodal data that parameterizes both of these features. We argue that the cellular detection rate, the fraction of genes expressed in a cell, should be adjusted for as a source of nuisance variation. Our model provides gene set enrichment analysis tailored to single-cell data. It provides insights into how networks of co-expressed genes evolve across an experimental treatment. MAST is available at https://github.com/RGLab/MAST.Electronic supplementary materialThe online version of this article (doi:10.1186/s13059-015-0844-5) contains supplementary material, which is available to authorized users.