Examining the roles of DNA2 during mammalian end resection

Abstract

The initiation and extension of end processing of chromosomal breaks to form single-stranded DNA (ssDNA), referred to as end resection, significantly affects the cellular consequences of such breaks.1 Regarding repair, such ssDNA is the optimal substrate for RAD51 nucleoprotein filaments, which catalyze the strand exchange step that is central to homology-directed repair (HDR).1 HDR that uses the identical sister chromatid as the template is relatively non-mutagenic. Alternatively, homologous segments in ssDNA flanking a chromosomal break could be annealed to each other to bridge the break, which is referred to as single-strand annealing (SSA) in the case of relatively extensive homology, or alternative/microhomology-mediated end-joining (Alt-EJ/MMEJ) involving more limited homology.1 Given the preponderance of repetitive elements in the human genome, SSA and Alt-EJ/MMEJ repair events may be prone to causing substantial genome rearrangements. In contrast to facilitating these homology-mediated repair events, end resection likely inhibits non-homologous end joining (NHEJ), since breaks with ssDNA at the terminus are not readily repaired by simple ligation. Such inhibition of NHEJ may be inconsequential to genome stability, but only if the sister chromatid is present to facilitate HDR. Apart from affecting repair outcomes, ssDNA is a potent signal for inducing cell cycle checkpoint pathways.2 Accordingly, examining how cells regulate the initiation and extent of end resection following induction of DNA lesions is critical for understanding the cellular consequences of such DNA lesions. The DNA2 nuclease was initially identified in yeast as an essential factor for DNA replication.3 During Okazaki fragment maturation, the DNA2 endonuclease activity cleaves the 5′ flap DNA in the middle of the ssDNA flap, removing a portion of the flap.4 The resulting short flap structure is then cleaved by the FEN1 nuclease to generate a DNA end that is suitable for ligation. More recently, yeast DNA2 was shown to function during DNA double-strand break (DSB) repair.5-7 Specifically, DNA2 and RecQ helicase Sgs1 were shown to form a complex that processes 5′ DNA ends to generate 3′ ssDNA overhangs. Regarding its role in genome stability, DNA2 was recently shown to be overexpressed in human cancer cells, where it reduces replication stress and supports hyper-DNA replication.7 How overexpression of DNA2 helps to reduce DNA replication stresses and promotes DNA replication in cancer cells remains far from clear. In the study by Karanja et al.,8 DNA2 is demonstrated to play an important role in promoting end resection following exposure to the interstrand crosslinking (ICL) agent cisplatin, as well as the topoisomerase I poison camptothecin. These defects in end resection are shown to affect many of the cellular consequences described above: DNA2-deficient cells show a decrease in SSA, a reduction in cell cycle checkpoint signaling (CHK1 phosphorylation), reduced repair of ICLs and a reduction in NHEJ (DNA-PKcs phosphorylation). Notably, both the resection defects and many of the above consequences are most severe in cells co-depleted of DNA2 and EXO1, indicating that while DNA2 itself is important to promote end resection, there is some degree of functional redundancy with EXO1. Interestingly, DNA2 depletion does not cause a defect in RAD51-dependent HDR and, furthermore, causes a lesser effect on RAD51 recruitment to DNA lesions as compared with its effect on end resection. Accordingly, the role of DNA2 during end resection appears to affect the relative balance of HDR and SSA, and hence the consequences of end resection on chromosomal stability. In addition, Karanja et al. demonstrates that the role of DNA2 on the cellular response to ICLs is dependent on the FA/BRCA pathway, which is critical for recognition and processing of such ICLs.9 DNA2 is shown to interact with the FA/BRCA factor FANCD2, and DNA2 depletion is shown to not clearly affect the cellular response to ICLs in FANCD2-deficient cells. These findings support the model that the FA/BRCA pathway recognizes ICLs to generate breaks that are resected via DNA2 in a manner that is partially redundant with EXO1. In the future, it will be important to elucidate the mechanisms by which DNA2 and EXO1 are regulated during resection, such as the coordination of the FA/BRCA pathway and DNA2.