Interrogating Putative Roles for R-Loops in dsRNA Formation and Transcription Regulation

Date

Authors

Abd-Elsattar, Ahmed

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

R-loops are three-stranded nucleic acid structures consisting of an RNA-DNA hybrid and a displaced ssDNA. Their formation is a conserved feature of the genomes of all living organisms, induced naturally by RNA transcription and specific nucleotide-sequence features. Intriguingly, R-loops have been implicated in transcription termination and RNAi-mediated heterochromatin formation through dsRNA formation at specific gene terminators. Yet this role has never been confirmed or tested on a genome-wide scale, and the underlying mechanism is still unclear. Here, I proposed two models for R-loop-mediated dsRNA formation and examined them using S. pombe cells. The first model is based on bidirectional double R-loop formation while the second is based on R-loops coupled with overlapping free asRNA. For a comprehensive and complementary R-loop profiling, I used three different S9.6-based and DRIP-like methods for consistent, high resolution and directional R-loop mapping. Beside the ultra-high-resolution double-stranded DNA-based ChIP-exonuclease (dsChIP-exo) technique, I developed directional and single stranded DNA-based ChIP-exonuclease (ssChIP-exo) to map the DNA strand of the hybrid. Moreover, I adopted the DRIPc method combined with the SMART-seq technology (SMART-DRIPc) to sequence the RNA strand of R-loops. To examine a possible role for R-loops in heterochromatin formation, I mapped the heterochromatic H3K9me2 mark using a high-sensitivity version of ChIP-exo and studied the impact of either RNase H depletion or overexpression on levels of this histone mark. dsChIP-exo revealed bidirectional signals mapping to both DNA strands. Differently, both ssChIP-exo and SMART-DRIPc revealed single R-loops forming in transcription direction. Although my data pointed toward double R-loops formation over tRNAs and rRNAs, it revealed a global anti-correlation between template and non-template strand R-loops even over the double R-loop forming genes. I found that R-loops over 60 to 70% of forming regions were extremely sensitive to physiological RNase H levels, which weakens both models of R-loop-dependent dsRNA formation. However, a category of RNase H-insensitive R-loops were associated with dsRNA-forming sites of sense-antisense transcription. Unexpectedly, RNase H depletion decreased R-loop signals over template strand of protein synthesis and ribosome biogenesis genes. R-loops of these genes are RNase H-insensitive and form on reverse strand. Excitingly, RNase H deletion, globally, increased forward-strand but decreased reverse-strand R-loop signals. Independently, no R-loops were detected over gene terminators or enriched over any of the heterochromatic repeats except tRNA genes and chromosome III telomeres rich in rRNA genes. Strikingly, either RNase H depletion or overexpression, respectively, disrupted H3K9me2 over heterochromatin in fission yeast and depleted global H3K9me2 in mammalian cells. R-loops may have been overrated as contributors for transcription termination and heterochromatin formation as they stand as marks of transcriptionally-active rather than heterochromatic domains. Despite challenges for the currently proposed models, R-loops coupled with overlapping asRNAs seems to be a more plausible model for dsRNA formation compared to double R-loops model. Nevertheless, practical involvement of R-loops in dsRNA formation still needs to be biochemically confirmed. My results suggest that perturbing RNase H levels, independent from R-loops, may alter the transcriptome and proteome, and impact cellular activities. Interestingly, R-loop orientation seems to modulate its response to RNase H. Excitingly, my observations suggest a role for RNase H in regulation of gene expression through suppressing formation of forward-strand R-loops and maintaining expression of sense genes. Transcriptome profiling and quantitative single-cell R-loop mapping using mimic and spike-in R-loop standards are required to confirm these results.

Description

Keywords

Citation

Source

Book Title

Entity type

Access Statement

License Rights

DOI

Restricted until

Downloads

File
Description