The emerging role of RNAs in DNA damage repair

Abstract

Many surveillance and repair mechanisms exist to maintain the integrity of our genome. All of the pathways described to date are controlled exclusively by proteins, which through their enzymatic activities identify breaks, propagate the damage signal, recruit further protein factors and ultimately resolve the break with little to no loss of genetic information. RNA is known to have an integral role in many cellular pathways, but, until very recently, was not considered to take part in the DNA repair process. Several reports demonstrated a conserved critical role for RNA-processing enzymes and RNA molecules in DNA repair, but the biogenesis of these damage-related RNAs and their mechanisms of action remain unknown. We will explore how these new findings challenge the idea of proteins being the sole participants in the response to DNA damage and reveal a new and exciting aspect of both DNA repair and RNA biology.

Figure 1

A schematic of the DNA repair pathway. The formation of a DSB induces the phosphorylation of ATM, which contributes to the activation of the DNA repair pathway and cell cycle arrest. A series of molecular signalling events lead to the deployment of ubiquitylation (Ub) marks on the histones (red cylinders) in the proximity of DNA breaks, facilitated by RNF8 and RNF168. The recruitment of 53BP1 marks the key crossroad of DSB repair (DSBR) pathway, which branches out into error-free HR or relatively error-prone NHEJ. Small RNAs have been proposed to function at two distinct steps in DSBR. Francia et al. suggested that it affects early signal propagation through ATM phosphorylation (blue arrow), while Gao et al. proposed that it only affects the HR sub-pathway via modulation of Rad51 binding (red arrow)

Figure 2

Outline of the microRNA biogenesis pathway in humans, and how plants utilise RdRPs to amplify these. The miRNA gene is transcribed by RNA polymerase II and typically capped and polyadenylated. This primary miRNA (pri-miRNA) contains the hairpin structure that is recognised and cleaved by Drosha, as part of the Microprocessor complex. The stem loop is then further trimmed by Dicer forming the pre-miRNA. In the canonical miRNA pathway, a single strand of the small RNA duplex is loaded into an Argonaute protein (Ago), which leads to repression of target transcripts. In plants, a dsRNA precursor is cleaved by Dicer into small dsRNA (green box). There exist multiple amplification pathways; broadly, an RdRP can synthesise a complementary strand by elongating a small RNA bound to its target RNA. Plant Dicer proteins can then cleave this newly generated dsRNA. to produce many secondary siRNAs that can repress target transcripts via Ago, or begin another cycle of small RNA amplification

Figure 3

Schematics of reporter systems used for small RNA sequencing experiments. (a) Wei et al. DR-GFP genomically integrated reporter. Transfection of I-SceI results in cleavage near the transcriptional start site (TSS) of GFP. Small RNA was sequenced and mapped back to the reporter sequence. (b) Francia et al. genomically integrated Lac-/Tet-operator-flanked I-SceI site. This reporter lacks transcriptional activity but is highly repetitive. Small RNA was detected after I-SceI transfection but at low levels (47 total reads after transfection versus 20 reads without). No information on where these RNAs mapped to was provided. (c) Michalik et al. Drosophila expression plasmids, either circularised or linearised. Here only the BamHI linearised vector is shown as it produced the highest number of small RNA reads. Grey dashed lines represent the approximate distribution of small RNA mapping back to the locus, where positional data was supplied in the manuscript. Right-angled arrows represent TSS, whereas vertical wavy line denotes integration within the genome

Figure 4

Detailed schematic of the DR-GFP HR repair reporter as in Figure 3a, used for RNA rescue experiments. A copy of the reporter is integrated into the genome to provide appropriate chromatin context. Insertion of an I-SceI restriction site within the GFP ORF results in a nonsense product that will produce no green fluorescence. After the induction of I-SceI cleavage, the cell can repair the resulting DSB via HR using the downstream internal GFP sequence (iGFP), producing a full-length GFP product. If HR is impaired, the break will instead be repaired via an alternate pathway, such as NHEJ, resulting in a sequence lacking full GFP coding region. The extent of deletion is dependent upon the non-HR mechanism chosen by the cell, but any loss of sequence within the I-SceI restriction site will prevent any further cutting. In the experiments by Wei et al. and Wang and Goldstein, the loss of Drosha and Dicer resulted in a lack of GFP indicating a deficiency in HR repair; however, when small RNAs extracted from control cells were incubated with these deficient cells for 1 h, GFP was found to be expressed (denoted by dashed arrow)

Figure 5

Postulated biogenesis of diRNAs and their function in DDR. Directionality of transcription, transcripts, and resection is 5' to 3', denoted by arrows. (a) A new transcript is formed from the break site. Any secondary structure formed by this transcript that can be recognised by Drosha and Dicer is processed into small RNAs, mirroring miRNA generation. (b) Where a cut occurs within an actively transcribing gene, a new end-dependent transcription event could take place resulting in formation of an antisense RNA. These RNA species can anneal to form dsRNA, and are then diced into small RNA substrates by Dicer. (c) The resulting small RNAs could then be incorporated into an Argonaute or another RBP, to carry out diRNA functions, which are hypothesised to be: (d) recruitment of repair and chromatin remodelling factors to the site of damage, (e) degradation of potentially aberrant transcripts or other as-of-yet unexplored functions