Transcription-dependent R-loop formation at mammalian class switch sequences

Genome-wide mapping reveals conserved and diverged R-loop activities in the unusual genetic landscape of the African trypanosome genome

R-loops are stable RNA-DNA hybrids that have been implicated in transcription initiation and termination, as well as in telomere maintenance, chromatin formation, and genome replication and instability. RNA Polymerase (Pol) II transcription in the protozoan parasite Trypanosoma brucei is highly unusual: virtually all genes are co-transcribed from multigene transcription units, with mRNAs generated by linked trans-splicing and polyadenylation, and transcription initiation sites display no conserved promoter motifs. Here, we describe the genome-wide distribution of R-loops in wild type mammal-infective T. brucei and in mutants lacking RNase H1, revealing both conserved and diverged functions. Conserved localization was found at centromeres, rRNA genes and retrotransposon-associated genes. RNA Pol II transcription initiation sites also displayed R-loops, suggesting a broadly conserved role despite the lack of promoter conservation or transcription initiation regulation. However, the most abundant sites of R-loop enrichment were within the regions between coding sequences of the multigene transcription units, where the hybrids coincide with sites of polyadenylation and nucleosome-depletion. Thus, instead of functioning in transcription termination the most widespread localization of R-loops in T. brucei suggests a novel correlation with pre-mRNA processing. Finally, we find little evidence for correlation between R-loop localization and mapped sites of DNA replication initiation.

Superhelicity Constrains a Localized and R-Loop-Dependent Formation of G-Quadruplexes at the Upstream Region of Transcription

Transcription induces formation of intramolecular G-quadruplex structures at the upstream region of a DNA duplex by an upward transmission of negative supercoiling through the DNA. Currently the regulation of such G-quadruplex formation remains unclear. Using plasmid as a model, we demonstrate that while it is the dynamic negative supercoiling generated by a moving RNA polymerase that triggers a formation of a G-quadruplex, the constitutional superhelicity determines the potential and range of the formation of a G-quadruplex by constraining the propagation of the negative supercoiling. G-quadruplex formation is maximal in negatively supercoiled and nearly abolished in relaxed plasmids while being moderate in nicked and linear ones. The formation of a G-quadruplex strongly correlates with the presence of an R-loop. Preventing R-loop formation virtually abolished G-quadruplex formation even in the negatively supercoiled plasmid. Enzymatic action and protein binding that manipulate supercoiling or its propagation all impact the formation of G-quadruplexes. Because chromosomes and plasmids in cells in their natural form are maintained in a supercoiled state, our findings reveal a physical basis that justifies the formation and regulation of G-quadruplexes in vivo. The structural features involved in G-quadruplex formation may all serve as potential targets in clinical and therapeutic applications.

RNA:DNA hybrids in the human genome have distinctive nucleotide characteristics, chromatin composition, and transcriptional relationships

Background

RNA:DNA hybrids represent a non-canonical nucleic acid structure that has been associated with a range of human diseases and potential transcriptional regulatory functions. Mapping of RNA:DNA hybrids in human cells reveals them to have a number of characteristics that give insights into their functions.

Results

We find RNA:DNA hybrids to occupy millions of base pairs in the human genome. A directional sequencing approach shows the RNA component of the RNA:DNA hybrid to be purine-rich, indicating a thermodynamic contribution to their in vivo stability. The RNA:DNA hybrids are enriched at loci with decreased DNA methylation and increased DNase hypersensitivity, and within larger domains with characteristics of heterochromatin formation, indicating potential transcriptional regulatory properties. Mass spectrometry studies of chromatin at RNA:DNA hybrids shows the presence of the ILF2 and ILF3 transcription factors, supporting a model of certain transcription factors binding preferentially to the RNA:DNA conformation.

Conclusions

Overall, there is little to indicate a dependence for RNA:DNA hybrids forming co-transcriptionally, with results from the ribosomal DNA repeat unit instead supporting the intriguing model of RNA generating these structures intrans. The results of the study indicate heterogeneous functions of these genomic elements and new insights into their formation and stability in vivo.

Electronic supplementary material

The online version of this article (doi:10.1186/s13072-015-0040-6) contains supplementary material, which is available to authorized users.

Human activation-induced cytidine deaminase causes transcription-dependent, strand-biased C to U deaminations

Activation-induced cytidine deaminase (AID) is required for the maturation of antibodies in higher vertebrates, where it promotes somatic hypermutation (SHM), class switch recombination and gene conversion. While it is known that SHM requires high levels of transcription of the target genes, it is unclear whether this is because AID targets transcribed genes. We show here that the human AID promotes C to T mutations in Escherichia coli which are stimulated by transcription. The mutations are strand-biased and occur preferentially in the non-transcribed strand of the target gene. Human AID purified from E.coli is active without prior treatment with a ribonuclease and deaminates cytosines in plasmid DNA in vitro. Further, the action of this enzyme is greatly stimulated by the transcription of the target gene in a strand-dependent fashion. These results confirm the prediction that AID may act directly on DNA and show that it can act on transcribing DNA in the absence of specialized DNA structures such as R-loops. It suggests that AID may be recruited to variable genes through transcription without the assistance of other proteins and that the strand bias in SHM may be caused by the preference of AID for the non-transcribed strand.

R-loops at immunoglobulin class switch regions in the chromosomes of stimulated B cells

The mechanism responsible for immunoglobulin class switch recombination is unknown. Previous work has shown that class switch sequences have the unusual property of forming RNA-DNA hybrids when transcribed in vitro. Here we show that the RNA-DNA hybrid structure that forms in vitro is an R-loop with a displaced guanine (G)-rich strand that is single-stranded. This R-loop structure exists in vivo in B cells that have been stimulated to transcribe the gamma3 or the gamma2b switch region. The length of the R-loops can exceed 1 kilobase. We propose that this distinctive DNA structure is important in the class switch recombination mechanism

Prostaglandin E2 and cAMP promote B lymphocyte class switching to IgG1

Prostaglandins of the E series (PGE) have traditionally been considered as suppressive for immune responses; however, recent data suggest that PGE channels the immune response towards a T helper 2 type response and production of selected immunoglobulin isotypes. Herein, we present data showing that PGE(2) and other agents that induce intracellular rises in cAMP significantly increased B lymphocyte IgG1 production (up to sevenfold). PGE(2) acted on small resting B cells and on uncommitted B cells expressing high levels of surface IgM to increase the number of cells secreting IgG1. PGE(2) even increased IgG1 synthesis by purified B cells in the absence of exogenous IL-4. Finally, PGE(2) synergized with IL-4 to induce germline gamma1 transcripts through the switch region. This transcription is required for isotype switching. These data support the hypothesis that PGE(2) acts on uncommitted resting B cells at the level of germline gamma1 transcription to promote class switching to IgG1. PGE(2) is an important regulator of the immune response, shifting the balance towards a T helper type 2 response, directing selection of the isotypes produced, and promoting memory cell formation.

Preferential cleavage of plasmid-based R-loops and D-loops by Drosophila topoisomerase IIIbeta

The topoisomerase (topo) III enzymes are found in organisms ranging from bacteria to humans, yet the precise cellular function of these enzymes remains to be determined. We previously found that Drosophila topo IIIbeta can relax plasmid DNA only if the DNA is first hypernegatively supercoiled. To investigate the possibility that topo IIIbeta requires a single-stranded region for its relaxation activity, we formed R-loops and D-loops in plasmids. In addition to containing a single-stranded region, these R-loops and D-loops have the advantage of being covalently closed and supercoiled, thus allowing us to assay for supercoil relaxation. We found that topo IIIbeta preferentially cleaves, rather than relaxes, these substrates. The cleavage of the R-loops and D-loops, which is primarily in the form of nicking, occurs to a greater extent at a temperature that is lower than the optimal temperature for relaxation of hypernegatively supercoiled plasmid. In addition, the cleavage can be readily reversed by high salt or high temperature, and the products fail to enter the gel in the absence of proteinase K treatment and are not observed with an active-site Y332F mutant of topo IIIbeta, indicating that the cleavage is mediated by a topoisomerase. We mapped the cleavage to the unpaired strand within the loop region and found that the cleavage occurs along the length of the unpaired strand. These studies suggest that the topo III enzyme behaves as a structure-specific endonuclease in vivo, providing a reversible DNA cleavage activity that is specific for unpaired regions in the DNA.

Transcription-induced cleavage of immunoglobulin switch regions by nucleotide excision repair nucleases in vitro

Immunoglobulin (Ig) heavy chain class switch recombination (CSR) mediates isotype switching during B cell development. CSR occurs between switch (S) regions that precede each Ig heavy chain constant region gene. Various studies have demonstrated that transcription plays an essential role in CSR in vivo. In this study, we show that in vitro transcription of S regions in their physiological orientation induces the formation of stable R loops. Furthermore, we show that the nucleotide excision repair nucleases XPF-ERCC1 and XPG can cleave the R loops formed in the S regions. Based on these findings, we propose that CSR is initiated via a mechanism that involves transcription-dependent S region cleavage by DNA structure-specific endonucleases that function in general DNA repair processes. Such a mechanism also may underlie transcription-dependent mutagenic processes such as somatic hypermutation, and contribute to genomic instability in general.

Stable RNA/DNA hybrids in the mammalian genome: inducible intermediates in immunoglobulin class switch recombination

Although it is well established that mammalian class switch recombination is responsible for altering the class of immunoglobulins, the mechanistic details of the process have remained unclear. Here, we show that stable RNA/DNA hybrids form at class switch sequences in the mouse genome upon cytokine-specific stimulation of class switch in primary splenic B cells. The RNA hybridized to the switch DNA is transcribed in the physiological orientation. Mice that constitutively express an Escherichia coli ribonuclease H transgene show a marked reduction in RNA/DNA hybrid formation, an impaired ability to generate serum immunoglobulin G antibodies, and significant inhibition of class switch recombination in their splenic B cells. These data provide evidence that stable RNA/DNA hybrids exist in the mammalian nuclear genome, can serve as intermediates for physiologic processes, and are mechanistically important for efficient class switching in vivo.