Note for: Transcription-Coupled DNA Double-Strand Break Repair: Active Genes Need Special Care

Note: Transcription-Coupled DNA Double-Strand Break Repair: Active Genes Need Special Care

(doi: 10.1371/journal.pgen.1006895)

Specific loci on eukaryotic chromosomes are inherently susceptible to breakage. Transcriptionally active loci are particularly fragile and that a specific DNA damage response is activated and dedicated to their repair.

Review on – crosstalk between transcription and double-strand break repair, from intrinsic fragility of genes to the mechanisms that restore the integrity of damaged transcription units.

DNA double helix is irregular: it can form non-canonical structures such as R-loops (three stranded structures composed of RNA:DNA hybrids and single-stranded DNA), hairpins, G-quadruplex (G4), and underwound or over-twisted DNA helices that are further translated into negative and positive supercoiling. The transcription, replication, and repair machineries must cope with this great variety of secondary and tertiary structures if they are to accurately execute transcriptional programs and maintain genome integrity. Identified active genes as particularly fragile and thus the theater of dedicated repair events. Discuss our current understanding of the crosstalk between transcription and double-strand break (DSB) repair, with a particular focus on recent evidence that (i) points to transcriptional activity as a major threat to the genome and (ii) sheds light on the DNA damage response at active gene.

These “common fragile sites” (CFS), which arise upon mild replication stress, were first defined as chromosomal bands with significantly elevated break/gap frequencies on mitotic chromosomes. CFS have been subjected to intense investigation, as they coincide with translocation break points and are hotspots for gross chromosomal rearrangements in cancer cells.

Moreover, it also became clear that CFS expression (i.e., frequency of breakage) is tissue-specific. CFS instability coincides with the expression of the underlying gene. Since these very long genes need more than one cell cycle to be entirely transcribed, it has been proposed that the collision between the replication and transcription machineries could lead to the slowing or stalling of replication forks. Nevertheless, since not all expressed long genes are prone to breakage, instability might rather result from secondary DNA structures and/or specific chromatin features assembled on a subset of these very long active genes. 'early replicating fragile sites (ERFS)' -- ERFS are replicated early, and like CFS, their instability (detected as gaps on mitotic chromosomes) is cell-type-specific.

this transcriptional activity rather than the timing of replication that promotes their instability. Although not directly linked to the occurrence of DNA damage, it has to be noted that other genome-wide profiles also revealed the accumulation of repair proteins at transcriptionally active loci (P-DNAPK, BRCA1, PALB2). The ability of endogenous DSBs to translocate to a “bait” DSB (introduced by either Cas9 or I-SceI endonucleases) is the basic principle of high-throughput genome-wide translocation sequencing (HTGTS) and translocation capture sequencing (TC‐Seq) techniques.

Combined with global run-on experiments to assess the ongoing transcription at the whole-genome level, these studies showed that:

(i) upon mild replicative stress (and to a lesser extent in unchallenged cycling cells), clusters of DSBs occur in long active genes that replicate late, and

(ii) DSB frequency is generally higher in nucleosome-depleted regions at transcriptional start sites of active genes and is proportional to transcription rate. BLESS, Break-seq, END-seq, and DSBCapture have been further developed as more direct approaches to mapping DSBs at the genomic scale in vivo.

BLESS allowed the clear identification of long genes as preferential sites of DSB following replication stress.

two classes of fragile sites that recurrently experience DSB:

(1) long, active genes, which replicate late (CFS), and are particularly susceptible to replicative stress; and

(2) promoters of transcriptionally active genes that replicate early.

the susceptibility of active genes to breakage, are still not fully understood.

two potentially interlinked structures, R-loops and G4s, contribute to genomic instability.

R-loops are three stranded RNA:DNA hybrids that can form as RNA polymerases progress through duplex DNA, by hybridization of the nascent RNA to the template DNA strand. R-loops form co-transcriptionally and accumulate at promoters that carry transcriptionally active marks. Most R-loops form at sites exhibiting GC-skew (increased GoverCon the non-template strand) due to increased thermodynamic stability of G-rich RNA hybridized with C-rich DNA strands.

This structure termed G-loop has first been proposed to happen on the immunoglobulin locus and to contribute to class switch recombination, and later found to occur in vitro and on plasmid DNA in Escherichia coli.

R-loop accumulation provokes DSB. The persistence of R-loops was also recently shown to impede replication at CFS, suggesting that R-loops may be an important determinant of CFS instability. R-loop and/or G4-driven DSB could arise in a replication-dependent manner through several non-exclusive mechanisms. First, R-loop and G4 formation result in an unannealed DNA strand, which is more sensitive to damaging agents. This could lead to a higher frequency of single-strand breakage (SSB), leading to fork collapse and DSB formation during subsequent replication.

XPG and XPF, known to generate ssDNA gaps during transcription-coupled nucleotide excision repair (TC-NER), are required for DSB induction at R-loop-forming loci. Moreover, prolonged fork stalling can also itself trigger nuclease-induced breaks and fork collapse. Finally, collision between replication and transcription machineries in either head-on (converging) or co-directional orientations can directly generate DSBs in E. coli.

Replication-dependent R-loop/G4-driven DSB induction, these structures can also trigger DSB in non-dividing cells by less well characterized mechanisms in E. coli and mammalian post-mitotic cells.

there is evidence that transcription activation itself can trigger DSB formation. Both estrogen and androgen stimuli trigger the appearance of DSBs at early responsive genes. stimulation of neuronal activity leads to DSB induction not only in a subset of

immediate responsive genes in primary cultured neurons but also in normal mouse brains following fear conditioning or novel environment exploration. Damage induction following stimulation can also occur in a replication-independent manner. Application of technologies such as HTGTS, BLESS, End-Seq, or DSBCapture to mapping DSBs in different cell types in response to various stimuli should soon allow the determination of the frequency of active promoter-associated DSBs and their exact positioning relative to R-loop/G4 structures and to TOP2B and RNA Pol II binding sites.

While further investigation is required to determine whether DSB production is an accidental by-product of transcription activation or a necessity for RNA Pol II release following stimulation, the aforementioned studies clearly show that the DSB landscape is strongly biased toward active genes.

Investigating how the transcriptional status of a damaged locus influences the repair reaction and how transcription is regulated at genes that experience a DSB have proved to be extremely challenging since it requires the induction of DSBs at specific, known loci on the genome. During the past decade, most studies have been performed using irradiation (γ-rays, X-rays, heavy ions) or drugs (topoisomerase poisons, intercalating agents…), which induce (i) uncharacterized damage beyond the DSB, including SSBs and various DNA adducts; (ii) damage at random, unknown positions on the genome; and iii) damage at various stages of the cell cycle.

Consequently, much effort has been made to develop systems where sequence specific and annotated DSBs can be induced on the genome. In higher eukaryotes, the first accurate and controlled method of inducing a single DSB at a specific locus was devised by the laboratory of Maria Jasin in 1994. To accurately quantify homologous recombination (HR) events, they introduced into the mouse genome a transgene that carries the recognition site for the endonuclease I-SceI within a GFP reporter system.

Transcription inhibition occurs at the damaged gene, and can also spread a few kilobases away from the DSB, although not over the entire megabase-wide γH2AX domain. Transcriptional repression is an active process that relies on ATM signaling, ubiquitination of H2A lysine K119, which is a well-known repressive histone mark established by Polycomb group proteins, and histone deacetylation by the NuRD complex. A recent study indicates that transcription of the repaired gene recovers normally in non-dividing cells, indicating that cell cycle progression is not required for the restoration of the epigenetic information and for the resumption of transcription. At present, little is known about these essential steps, probably because investigating transcription recovery at damaged genes has been challenging. The recent development of tools that not only permit DSB induction at controlled loci but also repair completion, owing to the availability of degradable or reversible enzymes, should now enable more rapid advance.

the cell's most important challenge is to accurately recover genetic information at damaged active genes. More recently, it has been proposed that RNA-templated repair also occurs in yeast and possibly in higher eukaryotes. The choice between these pathways at DSBs induced in active genes is therefore critical, as it will clearly determine the quality of the repair event and thus the subsequent functionality of the damaged gene.

Similarly, mammalian cell studies led to the conclusion that transcription enhances recombination rate by a process termed “transcription-associated recombination” (TAR). In yeast and bacteria, an increase of mutation rate was also observed in active genes (also known as TAM or transcription-associated mutagenesis. TAR was attributed to higher damage frequency in genes. However, recent work, described below, also raises the possibility that TAR may be related to repair mechanisms specifically set up at active genes. an active gene exhibits faster repair than an inactive gene, supporting the existence of a “transcription-coupled DSB repair” pathway. AsiSI-induced DSBs, mostly those lying within or close to transcriptionally active loci (e.g., promoters) could recruit the HR factor RAD51. This result clearly demonstrated that damaged genes that are active exhibit a preference for HR repair compared to other sequences located in euchromatin (AsiSI activity being inhibited by DNA methylation, these studies did not allow the investigation of repair at heterochromatin loci).

Transcription itself does not seem to be responsible for HR recruitment at damaged active genes. It is rather the trimethylation of histone H3 on lysine 36 (H3K36me3), well known to be correlated with elongating RNA-Pol II, that acts as a critical determinant for resection and RAD51 loading. H3K36me3 has been reported to be directly recognized by the PWWP domain of LEDGF, a protein that interacts with the CtIP resection factor. Acetylation of histone H4 lysine16, a histone mark correlated with active genomic regions, counteracts the recruitment of 53BP1, an anti-resection factor, to H4 methylated on lysine 20 and favors the binding of BRCA1, a pro-resection factor. Tip60 and hMof, two histone acetyltransferases with known activity on K16, have been shown to favor HR. These studies reveal a complex regulatory network of multiple chromatin marks that fine-tunes repair outcomes by influencing the recruitment and stabilization of repair factors and identified chromatin status as central in HR recruitment at active genes.

HR is strongly suppressed during the G1 phase. HR is strongly suppressed during the G1 phase. This raises the question of how DSBs in active genes are repaired during G1 or in non-dividing cells. NHEJ proteins are clearly recruited at DSBs induced in active genes throughout the cell cycle. This raises the possibility that although recruited, canonical NHEJ might not be entirely efficient at these loci. In agreement, a recent genome-wide analysis by BLESS of repair kinetics revealed that transcriptionally active genes are not fully proficient for repair in G1, in contrast to non-transcribed sequences on the genome.

These data suggest that inG1, a suboptimal efficacy of NHEJ at active genes, combined with the unavailability of HR, could result in persistent DSBs that cluster and await the arrival of S phase to undergo repair by a still uncharacterized mechanism. Thus, while HR seems to be preferentially used to repair active loci in G2, the contribution of other less well-characterized repair pathways at DSBs arising in active genes, especially during G1, needs further investigation. In addition to exhibiting specific chromatin patterns, active genes are distinguished from the rest of the genome by their ability to produce RNA. RNA could serve as a template for HR repair. How RNA-templated DNA repair functions in yeast remains speculative. These findings challenge the traditional view that HR requires a DNA template for repair and highlight the notion that RNA, which is available at transcriptionally active genes, could be used to template DNA repair. RNA-dependent repair might also take place in higher eukaryotes: quiescent human cells use a RNA and RAD52-dependent mechanism for DSB repair. and NHEJ has been proposed to operate at active genes through an RNA-templated mechanism. While direct evidence demonstrating the use of RNA as a template for repair in higher eukaryotes is still lacking, these studies open the possibility that RNA-templated repair is a conserved alternative pathway for repairing regions of the genome in the course of transcription.

DSB induced small RNAs (diRNAs)

it is interesting that under conditions of no DNA damage, double-stranded RNAs have been proposed to function in terminating transcription and assembling repressive chromatin.

active genes are the theater of complex and tightly regulated DSB repair events that together could define a “transcription-coupled DSB repair” pathway. These transcription-coupled DSB repair mechanisms may reveal to be important, given that normal transcription activity is now emerging as a potent DSB-inducing agent. Mutations of some DSB repair factors are

associated with progeria or segmental premature aging phenotypes (such as Werner disease) and neurodegenerative diseases (Ataxia Telangectasia, Nijmegen breakage syndrome…). Whether these disorders impede the correct repair of DSB induced in active regions of the genome is a possibility that deserves attention in the near future.

x