Eukaryotic resectosomes: A single-molecule perspective

Abstract

DNA double-strand breaks (DSBs) disrupt the physical and genetic continuity of the genome. If unrepaired, DSBs can lead to cellular dysfunction and malignant transformation. Homologous recombination (HR) is a universally conserved DSB repair mechanism that employs the information in a sister chromatid to catalyze error-free DSB repair. To initiate HR, cells assemble the resectosome: a multi-protein complex composed of helicases, nucleases, and regulatory proteins. The resectosome nucleolytically degrades (resects) the free DNA ends for downstream homologous recombination. Several decades of intense research have identified the core resectosome components in eukaryotes, archaea, and bacteria. More recently, these proteins have been characterized via single-molecule approaches. Here, we focus on recent single-molecule studies that have begun to unravel how nucleases, helicases, processivity factors, and other regulatory proteins dictate the extent and efficiency of DNA resection in eukaryotic cells. We conclude with a discussion of outstanding questions that can be addressed via single-molecule approaches.

Figure 1. DNA Double Strand Break Resection for Homologous Recombination

Resection is initiated by the MRN complex, which recognizes the free DNA ends. MRN and CtIP initially process the DNA to promote the loading and assembly of a resectome containing either the nuclease Exo1 or the nuclease/helicase DNA2. The helicase BLM participates in both pathways. These two resectosomes catalyze long-range DNA resection. The resulting single-stranded DNA is bound by RPA and other SSBs. Finally, resection is terminated by an unknown mechanism and RPA is exchanged for Rad51, which facilitates the homology search for HR.

Figure 2. Single-molecule studies of MRN

(A) Domain map (top) and structural map (bottom) of the Mre11-Rad50-Nbs1 complex. (B) Atomic force microscopy studies reveal that MRN undergoes mesoscale conformational changes upon binding to DNA. Reproduced with permission from (Moreno-Herrero et al., 2005). Additionally, MRN complexes can bridge two DNA molecules via the long coiled-coil arms of Rad50. (C) A single-molecule FRET assay reveals that MRN harbors a limited ATP-dependent end-opening activity. Reproduced with permission from (Cannon et al., 2013). Briefly, MRN complexes were immobilized on the surface of a flowcell. DNA oligonucleotides containing a Cy3 label on one strand and a Cy5 label on the other are then bound by MRN, showing a high-FRET state. However, in the presence of ATP, MRN opens the ends, separating the Cy dyes and shifting the population to a low-FRET state.

Figure 3. Single-molecule studies of Exo1

(A) Domain map (top) and graphical map (bottom) of Exo1 and known interacting partners. Exo1 contains at least two distinct domains: a structured N-terminal nuclease domain (NTD) and an unstructured C-terminal domain (CTD). Previous studies have hypothesized an auto-inhibitory interaction between the NTD and CTD (Orans et al., 2011). (B) DNA Curtains assay for studying Exo1 activity and regulation. Reproduced with permission from (Myler et al., 2016). (C) Kymograph and particle tracking (top) of a single Exo1 molecule resecting from a DNA end. Rate and processivity measurements are indicated. Boxplots (bottom) show the velocity and processivity of Exo1 molecules from the lambda DNA end (magenta), nicked DNA (orange), or with the nuclease-dead Exo1(D78A/D173A) (black). (D) Kymographs of Exo1 lifetime upon injection of RPA (left) or SOSS1 (right). White arrow indicates dissociation of Exo1. Lifetimes (bottom) of Exo1 upon no injection (left), RPA (middle), or SOSS1 (right). (E) Paramagnetic bead assay for monitoring Exo1 resection. Reproduced with permission from (Jeon et al., 2016). Briefly, a paramagnetic bead attached to nicked DNA is tethered to a microscope slide surface via a biotin-streptavidin interaction. In the presence of flow, the DNA is extended and Exo1 is added. As Exo1 resects the DNA, it generates single-stranded DNA, which extends differently than dsDNA, resulting in a change in bead location. This change is monitored over time to determine the rate and processivity of Exo1 (right). In the presence of SSB, the bead does not move. However, MSH2-6 is able to overcome SSB inhibition of Exo1 activity. (F) Model for Exo1 resection in the presence of RPA. In the absence of a processivity factor (left), Exo1 is physically removed by RPA, which binds the single-stranded DNA generated by resection. However, processivity factors (green ring) may promote Exo1 degradation in the presence of RPA (right).

Figure 4. Single-molecule studies of DNA2/BLM

(A) Domain map of DNA2 (top) and BLM (bottom). Functional domains and interacting partners are labeled. (B) A smFRET-based assay for BLM unwinding. Reproduced with permission from (Yodh et al., 2009). A dsDNA containing both Cy3 and Cy5 labels is tethered to a flowcell surface. A high FRET state is observed in the absence of BLM. In the presence of BLM, the dyes are physically separated, resulting in a low-FRET state. Interestingly, BLM can repetitively unwind and re-anneal the DNA, suggesting a strand-switching mechanism. Right panels show the wait time between strand switching events (top) and FRET trace (bottom). For full length BLM, strand switching wait time is unaffected by the presence of RPA. (C) A single-molecule assay for DNA2 nuclease-dead (E675A) helicase activity. Reproduced with permission from (Levikova et al., 2013). The position of a magnetic bead is monitored as DNA2 unwinds DNA in the presence of RPA. Right panel shows multiple processive DNA2 helicase trajectories. (D) Models for the functions of the DNA2-BLM-RPA complex. DNA2, is a potent, bidirectional nuclease in the absence of other factors, but cannot processively degrade DNA. In the presence of BLM, DNA2 can processively degrade both strands of DNA. Finally, in the DNA2-BLM-RPA complex, RPA directs DNA2 to only degrade the 5′ strand, creating the necessary 3′ overhangs for recombination.

Figure 5. Speculative models for termination of Exo1-dependent resection

In the first model (left), a processivity factor (green ring) promotes degradation in the presence of RPA. Then, at a later step of DNA resection, the processivity factor dissociates or is removed, allowing RPA to displace Exo1. For the second model (middle), the resectosome encounters a chromatin block, which physically stalls resection and leads to dissociation of the resectosome. In the third model (right), a termination factor (red oval) is recruited to physically disassemble the resectosome. We propose that one or several of these concepts will play a role in both Exo1 and DNA2-dependent resection.