Higher eukaryotic genomes are bound by a large number of coding and non-coding RNAs, but approaches to comprehensively map the identity and binding sites of these RNAs are lacking. Here UCSD researchers report a method to capture in situ global RNA interactions with DNA by deep sequencing (GRID-seq), which enables the comprehensive identification of the entire repertoire of chromatin-interacting RNAs and their respective binding sites. In human, mouse, and Drosophila cells, the researchers detected a large set of tissue-specific coding and non-coding RNAs that are bound to active promoters and enhancers, especially super-enhancers. Assuming that most mRNA-chromatin interactions indicate the physical proximity of a promoter and an enhancer, they constructed a three-dimensional global connectivity map of promoters and enhancers, revealing transcription-activity-linked genomic interactions in the nucleus.
(a) Schematic presentation of the GRID-seq technology. Left: steps performed in situ on fixed nuclei. Right: steps performed in solution. The two major bands resolved by native polyacrylamide gel correspond to the products of the linker ligated to both DNA and RNA (upper band) or to either DNA or RNA (lower band). The upper band was excised for adaptor ligation and deep sequencing. (b) Genomic distributions of uniquely mapped RNA/DNA read mates in MDA-MB-231 cells. (c,d) Scatterplots of length-normalized RNA reads from annotated genes detected by GRID-seq in comparison with gene expression detected by RNA-seq (c) or GRO-seq (d) in MDA-MB-231 cells. Highlighted are two representative lncRNAs MATAL1 and NEAT1. R2 values in c and d were determined by linear regression. (e) Comparison of raw MALAT1-chromatin interaction signals captured by RAP-DNA and GRID-seq. RPK: GRID-seq reads per Kb. RPKM: reads per Kb per million mapped reads. (f) MALAT1 GRID-seq signals in a highlighted region of Chr. 17 relative to the background (light blue). (g) Top: scheme for using human MDA-MB-231 cells, Drosophila S2 cells, or their mix for library construction. Bottom: the percentages of human RNAs ligated to human or Drosophila DNA and the percentages of Drosophila RNAs ligated to Drosophila or human DNA. (h) Comparison between the true background based on cross-species RNA–DNA interactions and the deduced background by trans-chromosomal mRNAs, as illustrated on a region of Chr. X in Drosophila S2 cells (top panel) or globally (bottom panel). (i) MALAT1 GRID-seq signals after background correction in comparison with GRO-seq signals in MDA-MB-231.