RNAs as Proximity-Labeling Media for Identifying Nuclear Speckle Positions Relative to the Genome

Abstract

SummaryIt remains challenging to identify all parts of the nuclear genome that are in proximity to nuclear speckles, due to physical separation between the nuclear speckle cores and chromatin. We hypothesized that noncoding RNAs including small nuclear RNA (snRNAs) and Malat1, which accumulate at the periphery of nuclear speckles (nsaRNA [nuclear speckle-associated RNA]), may extend to sufficient proximity to the genome. Leveraging a transcriptome-genome interaction assay (mapping of RNA-genome interactions [MARGI]), we identified clusters of nsaRNA-interacting genomic sequences (nsaPeaks). Posttranscriptional pre-mRNAs, which also accumulate to nuclear speckles, exhibited proximity to nsaPeaks but rarely to other genomic regions. Our combined DNA fluorescence in situ hybridization and immunofluorescence analysis in 182 single cells revealed a 3-fold increase in odds for nuclear speckles to localize near an nsaPeak than its neighboring genomic sequence. These data suggest a model that nsaRNAs are located in sufficient proximity to the nuclear genome and leave identifiable genomic footprints, thus revealing the parts of genome proximal to nuclear speckles.

CPSF6-Dependent Targeting of Speckle-Associated Domains Distinguishes Primate from Nonprimate Lentiviral Integration

“…FIV in particular disfavored SPADs in Jurkat T cells, and MLV, moreover, targeted SPADs at greater frequencies than the nonprimate lentiviruses studied here in both HEK293T and Jurkat T cells. Superenhancers are known to associate with nuclear speckles (46,47), which we suspect accounts for MLV's SPAD preference. In contrast to MLV, superenhancers are not preferred targets of bulk HIV-1 integration (5,48).…”

Mapping RNA–chromatin interactions by sequencing with iMARGI

“…In comparison, iMARGI/ MARGI and ChAR-seq did not provide a one-size-fits-all statistical method for identifying all RNA-DNA interactions. Instead, the authors opted for the statistical methods that best fit each biological question, including calling chromatin enriched RNAs 16 ; assessing enrichments of RNA-DNA interactions at TAD boundaries 16 , TSS 14 , and the chromosomal regions near nuclear speckles 18 ; and comparing global RNA-DNA interactions with genome-wide distributions of histone modifications 14,16 transcription factor binding intensities 18 , and fusion RNAs 15 .…”

“…These genome-wide RNA-chromatin interaction assays include Mapping RNA-Genome Interactions (MARGI) 14 and its improved version called in situ MARGI (iMARGI) 15 , Chromatin-Associated RNA Sequencing (ChAR-seq) 16 , and Mapping Global RNA Interactions With DNA by Deep Sequencing (GRID-seq) 17 . Both GRID-seq and ChAR-seq revealed a range of chromatin-bound RNAs including nascent transcripts, chromosomespecific dosage compensation non-coding RNAs (ncRNAs), and trans-associated RNAs 16,17 GRID-seq also revealed extensive interactions between mRNAs and enhancers 17 MARGI and iMARGI revealed thousands of caRNAs including both coding and noncoding RNAs 14,15,18 These caRNAs are not only associated with the genomic sequences from which they are transcribed (to form proximal interactions), but can also attach to distal genomic sequences (to form distal interactions) on the same chromosomes or to other chromosomes (to form inter-chromosomal interactions). Surprisingly, transcription start sites (TSS) have been identified to be preferred genomic loci targeted by non-coding caRNAs through distal and inter-chromosomal interactions 14 .…”

Stress-induced RNA–chromatin interactions promote endothelial dysfunction

“…Following the rule defined above, we call this Malat1SE. The large number of caRNAs transcribed from Malat1SE is expected because the MALAT1 lncRNA interacts with a large amount of transcription active genomic regions 29 . The number of hubs increased from 1 (Day 0) to 14 (Day 3) and to 25 (Day 7; Fig.…”