AND OBJECTIVE: DNA damage-induced chromatin reorganization is emerging to be a key aspect of eukaryotic DNA repair. Much has been learned about the role of histone modifications as landing pads for repair effectors, thereby modulating or directing their recruitment to sites of damage. Recent work suggests that structural changes in the break-surrounding chromatin environment may be equally important in directing the repair process, with implications ranging from repair factor accessibility to break-proximal transcriptional silencing. In light of this complexity, we decided to take an unbiased approach to dissect the role of chromatin in DNA repair, identify its most critical components and explore their functional relevance. RESULTS AND FUTURE DIRECTIONS: To gain insight into the role of chromatin in DSB repair, we performed RNAi-based high-throughput screening of a comprehensive list of 400 Gene Ontology-annotated chromatin modifiers. Repair efficiency was determined using the previously established, U2OS cell-based DR-GFP reporter system, in which DNA repair by homologous recombination (HR) results in restoration of a functional GFP gene and HR efficiency can, thus, be measured as the fraction of GFP+ cells. We found a large number of chromatin-modifying enzymes to be involved in DNA break repair, and in contrast to previous reports suggesting DSB-induced chromatin relaxation, many of these proteins were transcriptional repressors and/or associated with the formation of silent chromatin. Specifically, we identified a DNA repair module consisting of two macro-histone variants and the histone 3 lysine 9 methyltransferase (H3K9MT) PRDM2, which has not previously been implicated in DNA break repair. Both macroH2A1 and PRDM2 promote a biphasic change in the DSB-proximal chromatin microenvironment. In agreement with DSB-induced chromatin decondensation reported previously, we observe initial depletion of macroH2A along with repressive chromatin marks from the break site, and initiation of the DNA damage response is not affected by either PRDM2 or the macro-histone variants. Following this initial expansion, we detect DNA damage signaling-dependent, DSB-proximal enrichment of macro-histone variants as well as the silent chromatin mark H3-dimethyl-K9 and PRDM2, which in turn promotes chromatin condensation. We show that macroH2A mediates PRDM2 recruitment to breaks and are currently investigating the molecular basis for macroH2A1 incorporation at DSB sites as well as its relevance for PRDM2 recruitment. Importantly, macroH2A or PRDM2 depletion causes a significant reduction in H3K9 dimethylation and concomitant break-associated chromatin reorganization. Consistent with a role for macroH2A and PRDM2 during HR, loss of either protein significantly reduced DSB recruitment of the HR mediator and tumor suppressor BRCA1 but not the NHEJ-associated repair protein 53BP1. Moreover, both macroH2A1 and PRDM2 promote the phosphorylation of the single-stranded DNA binding protein RPA, which is a critical step in HR-associated end resection and repair. Moreover, we observed increased sensitivity to inhibition of poly(ADP-ribose) polymerase following knock-down of macroH2A or PRDM2, which is a hallmark of BRCA1-deficient tumor cells. Our data thus implicate the formation of repressive chromatin as a modulator of repair outcome, which is expected to have significant consequences for genomic integrity, specifically in the context of defective BRCA1 function. Further nderlining the functional relevance of structural changes in DSB-associated chromatin, we found that experimentally induced chromatin relaxation via the inhibition of histone deacetylases results in a selective decrease in BRCA1 recruitment following DSB induction. Consistent with this observation, BRCA1 preferably binds to H3K9me2 modified histone tails as opposed to H3K9 acetylated peptides, and future work is directed at identifying the mechanistic basis for this phenomenon. Prompted by the potential of chromatin changes to selectively modulate DSB repair factor choice, we initiated a new aim to investigate the role of other DSB repair-associated chromatin modifiers in the recruitment of BRCA1 and/or 53BP1. Based on recent reports describing ubiquitination of histone H2A as modification that is critical for 53BP1 recruitment to DSBs, we focused on two USP-family deubiquitinating enzymes that were identified as HR promoters in our RNAi screen. We are currently investigating how loss of these enzymes affects the recruitment of 53BP1 and BRCA1 to sites of DSBs and concomitant DSB repair via NHEJ. We further plan to determine possible consequences of USP loss or overexpression on H2A ubiquitination. Together, this aim is expected to shed light on a novel aspect of ubiquitin-dependent DSB repair factor recruitment with the potential to distinguish between 53BP1 and BRCA1-mediated DSB repair pathways. IMPLICATIONS: Defects in BRCA1 function have been linked to tumor initiation and genomic instability. More recently, it has been suggested that a key role for BRCA1 may be to prohibit aberrant 53BP1 recruitment to sites of DNA damage, which accounts for many of the detrimental genomic effects of BRCA1 loss. Determining the factors that control the balance between BRCA1 and 53BP1 at DNA breaks is, thus, critical for our understanding of DSB repair in general and BRCA1-associated tumorigenesis in particular. Both 53BP1 and BRCA1 occupy expansive, DSB-surrounding subnuclear domains, and the identification of selective, chromatin-based modulators of BRCA1/53BP1 recruitment to sites of damage, therefore, has implications for the targeted manipulation of repair outcome and possibly tumor initiation.
| Liu, Jinping; Kruswick, Alex; Dang, Hien et al. (2017) Ubiquitin-specific protease 21 stabilizes BRCA2 to control DNA repair and tumor growth. Nat Commun 8:137 |
| Shi, Lei; Oberdoerffer, Philipp (2012) Chromatin dynamics in DNA double-strand break repair. Biochim Biophys Acta 1819:811-9 |
| Sinclair, David A; Oberdoerffer, Philipp (2009) The ageing epigenome: damaged beyond repair? Ageing Res Rev 8:189-98 |
