Divergent lncRNAs that are transcribed in the opposite direction to nearby protein-coding genes comprise a significant proportion (∼20%) of total lncRNAs in mammalian genomes. Through genome-wide analysis, we found that the distribution of this lncRNA class strongly correlates with essential developmental regulatory genes. In pluripotent cells, divergent lncRNAs regulate the transcription of nearby genes. As an example, the divergent lncRNA Evx1as promotes transcription of its neighbor gene, EVX1, and regulates mesendodermal differentiation. At a single-cell level, early broad expression of Evx1as is followed by a rapid, high-level transcription of EVX1, supporting the idea that Evx1as plays an upstream role to facilitate EVX1 transcription. Mechanistically, Evx1as RNA binds to regulatory sites on chromatin, promotes an active chromatin state, and interacts with Mediator. Based on our analyses, we propose that the biological function of thousands of uncharacterized lncRNAs of this class may be inferred from the role of their neighboring adjacent genes.

Thousands of noncoding transcripts exist in mammalian genomes, and they preferentially localize to chromatin. Here, to identify cis-regulatory elements that control RNA-chromatin association, we developed a high-throughput method named RNA element for subcellular localization by sequencing (REL-seq). Coupling REL-seq with random mutagenesis (mutREL-seq), we discovered a key 7-nt U1 recognition motif in chromatin-enriched RNA elements. Reporter assays indicated a direct role for U1 snRNP recognition in regulating RNA-chromatin localization. Globally, U1 motifs and U1 binding are strongly enriched in long noncoding RNA (lncRNA) transcripts. Inhibition of U1 snRNA, and of U2 to a lesser degree, led to global reduction in chromatin association of hundreds of lncRNAs.For promoter-and enhancer-associated noncoding RNAs, U1 binds to their genomic neighborhoods, and their chromatin association depends on both U1 and U2 snRNAs. These findings reveal that U1 snRNP, perhaps together with the splicing machinery, acts widely to promote the chromatin association of noncoding transcripts.

Highlights d SINE, L1, and low-complexity repeats barcode genes with distinct functions d Genomic repeats dictate the time and level of gene expression during development d L1-enriched genes are sequestered in the inactive NAD/LAD domains for silencing d L1 RNA promotes the nuclear localization and repression of L1-enriched genes