BRCA2 chaperones RAD51 to single molecules of RPA-coated ssDNA

Abstract

Mutations in the breast cancer susceptibility gene, BRCA2, greatly increase an individual's lifetime risk of developing breast and ovarian cancers. BRCA2 suppresses tumor formation by potentiating DNA repair via homologous recombination. Central to recombination is the assembly of a RAD51 nucleoprotein filament, which forms on single-stranded DNA (ssDNA) generated at or near the site of chromosomal damage. However, replication protein-A (RPA) rapidly binds to and continuously sequesters this ssDNA, imposing a kinetic barrier to RAD51 filament assembly that suppresses unregulated recombination. Recombination mediator proteins-of which BRCA2 is the defining member in humans-alleviate this kinetic barrier to catalyze RAD51 filament formation. We combined microfluidics, microscopy, and micromanipulation to directly measure both the binding of full-length BRCA2 to-and the assembly of RAD51 filaments on-a region of RPA-coated ssDNA within individual DNA molecules designed to mimic a resected DNA lesion common in replication-coupled recombinational repair. We demonstrate that a dimer of RAD51 is minimally required for spontaneous nucleation; however, growth self-terminates below the diffraction limit. BRCA2 accelerates nucleation of RAD51 to a rate that approaches the rapid association of RAD51 to naked ssDNA, thereby overcoming the kinetic block imposed by RPA. Furthermore, BRCA2 eliminates the need for the rate-limiting nucleation of RAD51 by chaperoning a short preassembled RAD51 filament onto the ssDNA complexed with RPA. Therefore, BRCA2 regulates recombination by initiating RAD51 filament formation.

Keywords: DNA recombination; DNA repair; RAD51; breast cancer; single-molecule visualization.

Fig. 1.

Direct imaging of BRCA2 binding to RPA-coated ssDNA on single molecules of gapped λ DNA. (A) Schematic of experimental approach combining fluorescence microscopy, a microfluidic flow cell, and optical trapping, as well as the micromanipulation used to capture and image BRCA2 on individual DNA molecules. (B) Illustration of the gapped λ DNA generated through in vitro recombination of circular ssDNA with an engineered λ DNA. (C) Schematic of experimental protocol: Each molecule of gapped λ DNA was captured and micromanipulated between two beads held in separately controllable optical traps. The molecule was moved between solutions in a six-channel flow cell and successively incubated in a solution containing BRCA2. (D) Cartoon and microscopic image of a single molecule of gapped λ DNA (Left, stained with YOYO-1, cyan) that was destained and then successively incubated with BRCA2 (5 nM) plus α-BRCA2 and α-IgG^AF546. Montage shows BRCA2 (magenta) binding to the gapped λ DNA at increasing time intervals. (E) Cartoon representation of the gapped λ DNA between two beads (Top) and histogram (Middle) of binding positions of BRCA2 (number of foci, N = 60). Each data point is also plotted as a single tick (Bottom) where the semi-transparent box represents the SE associated with assigning position owing to the optical resolution of the microscope. Gray bars represent the 10 to 90th percentile range of the 5′- and 3′-termined junctions (N = 98).

Fig. 2.

Direct imaging of RAD51 nucleation and filament formation in the absence and presence of RPA. (A) Schematic of a single molecule of gapped λ DNA attached at each end to a PEG-coated surface via neutravidin. The dsDNA region was initially visualized using SYTOX Orange (red), which was subsequently dissociated upon the addition of binding buffer and fluorescein-RAD51 (green). (B) In the absence of RPA, fluorescent RAD51 rapidly filled the ssDNA region. (C) When RPA was present, RAD51 binding was slower and punctate. (D) Comparison of the lag time in the absence (black filled symbols) or presence (black open symbols) of RPA measured using TIRF microscopy. Lag times in the presence of BRCA2 were measured using optical trapping (see text) and are shown in red symbols for comparison. Lines represent the arithmetic mean and error bars represent SD. No RPA: 5 ± 3 (N = 111); +RPA: 78 ± 40 (N = 20); +RPA/BRCA2: 27 ± 11 (N = 22). (Scale bar in panels B and C is 2 μm.)

Fig. 3.

RAD51-BRCA2 complexes, in contrast to BRCA2 alone, are focused to the ssDNA regions. (A) Schematic of optical-trap experiments designed to visualize nucleation of RAD51 on gapped λ DNA in the absence or presence of BRCA2. (B) A single molecule of gapped λ DNA with bound RPA held between two optical traps and coincubated with 100 nM RAD51 (green) and 5 nM BRCA2 (red, α-MBP^AF546). (C) Cartoon of the gapped λ DNA (Top) and histogram (Middle) showing positions of all RAD51 nucleation events either alone (green) (N = 18) or when coincubated with BRCA2 (red) in the absence of antibody (N = 28). Positions of individual foci are plotted (Bottom) relative to position of the 5′- and 3′-termined junctions. The transparent box around each dash represents the SE owing to the optical resolution of our microscope. Gray bars represent the 10 to 90th percentile range of the 5′- and 3′-termined junctions (N = 98). (D) Bar plot of the number of RAD51 nucleation events in the absence (green) or presence of BRCA2 (red) observed in the regions nearest the 3′-junction, clearly in the middle of the RPA-coated region, or near the 5′-junction. The odds ratios and P-values were calculated using Fisher’s exact test.

Fig. 4.

Nucleation of RAD51 requires a dimer and is accelerated by BRCA2 by delivering a prenucleated complex. (A) The cumulative frequency of RAD51 nucleation measured by optical trapping is plotted as a function of increasing incubation time at varying concentrations of RAD51 in the absence or (B) presence of BRCA2. Inset shows a zoomed-in scale. The solid lines represent nonlinear fitting to a single-exponential rate equation: 50 nM: t_1/2 = 110 ± 10 s, 100 nM: t_1/2 = 47 ± 4 s, 200 nM: t_1/2 = 8 ± 1 s, 300 nM: t_1/2 = 11 ± 4 s (SE). The open symbols and cyan dashed line represent the kinetics of BRCA2 binding in the absence of RAD51 as measured in Fig. 1 and is shown as a comparison to the kinetics of RAD51 nucleation (see text). (C) The arithmetic mean nucleation time plotted as a function of increasing RAD51 concentration in the absence: 50 nM: 179 ± 41 s (N = 10), 100 nM: 76 ± 16 s (N = 14), 200 nM: 38 ± 23 s (N = 4), 300 nM: 23 ± 4 s (N = 4), 400 nM: 30 ± 6 s (N = 6), or presence of BRCA2: 50 nM: 24 ± 10 s (N = 7), 100 nM: 21 ± 2 s (N = 46), 400 nM: 15 ± 6 s (N = 7). The black curve represents a fit to the power law (J=k[RAD51]ⁿ) where n = 1.5 ± 0.3 (SE). The cyan symbol represents the half-time for BRCA2 binding in the absence of RAD51, measured with α-BRCA2 plus α-IgG^AF546: 30 ± 6 s (N = 6). The red line is for visual purposes only; gray symbol is the rate in the absence of RPA: 5.2 ± 0.3 s (N = 111). Error and error bars are SEM, and if not visible, are smaller than the symbol. (D) Model of BRCA2-mediated RAD51 nucleation on RPA-coated ssDNA.