The HsRAD51B-HsRAD51C stabilizes the HsRAD51 nucleoprotein filament

Abstract

There are six human RAD51 related proteins (HsRAD51 paralogs), HsRAD51B, HsRAD51C, HsRAD51D, HsXRCC2, HsXRCC3 and HsDMC1, that appear to enhance HsRAD51 mediated homologous recombinational (HR) repair of DNA double strand breaks (DSBs). Here we model the structures of HsRAD51, HsRAD51B and HsRAD51C and show similar domain orientations within a hypothetical nucleoprotein filament (NPF). We then demonstrate that HsRAD51B-HsRAD51C heterodimer forms stable complex on ssDNA and partially stabilizes the HsRAD51 NPF against the anti-recombinogenic activity of BLM. Moreover, HsRAD51B-HsRAD51C stimulates HsRAD51 mediated D-loop formation in the presence of RPA. However, HsRAD51B-HsRAD51C does not facilitate HsRAD51 nucleation on a RPA coated ssDNA. These results suggest that the HsRAD51B-HsRAD51C complex plays a role in stabilizing the HsRAD51 NPF during the presynaptic phase of HR, which appears downstream of BRCA2-mediated HsRAD51 NPF formation.

Keywords: ATPase, Bloom's syndrome; BLM; Homologous recombination; RAD51 paralogs.

Figure 1. Homology based modeled structures of HsRAD51, HsRAD51B and HsRAD51C indicate similar domain orientations

(A) HsRAD51 structure was modeled using homology based modeling software I-TASSER [49] and the optimized structures were superimposed on the Methanococcus voltae RadA (MvRAD51) crystal structure coordinates (PDB ID: 1XU4) using Pymol. The non-hydrolysable ATP analog AMP-PNP is bound at the interface between MvRAD51 top monomer (cyan) and the bottom monomer (wheat). The Walker A subunit of the bottom monomer is indicated in grey. L2 ssDNA binding regions of MvRAD51 and HsRAD51 are highlighted in salmon and brown colors, respectively. The ATP caps of MvRAD51 and HsRAD51 are indicated in red and magenta colors, respectively. D316 of HsRAD51 is indicated. Two potassium (K⁺) ions at the interface of MvRAD51 monomers are also illustrated. Modeled and superimposed structures of HsRAD51B in Blue (B) and Human HsRAD51C in Green (C). The domain color assignments are similar to A.

Figure 2. HsRAD51B-HsRAD51C heterodimer forms stable complexes on DNA

(A) ssDNA binding and dissociation curves obtained from surface plasmon resonance (SPR). The SPR curves correspond to 800nM concentration of either HsRAD51 or HsRAD51B-HsRAD51C heterodimer. DNA binding corresponds to the timescale 100s-300s and dissociation from 300s-600s. Binding response is given in Response Units. DNA binding and dissociation were analyzed in 25mM Tris-OAc (pH 7.5), 1mM ATP, 1mM Mg(OAc)₂, 2mM DTT and 0.005% Tween-20. Control (without protein) curve is red, HsRAD51B-HsRAD51C binding at 25mM K⁺ (KCl) curve is black, HsRAD51B-HsRAD51C binding at 100mM K⁺ curve is pink, HsRAD51 binding at 25mM K⁺ curve is blue and HsRAD51 binding at 100mM K⁺ curve is green. (B) dsDNA binding and dissociation of HsRAD51B-HsRAD51C heterodimer and HsRAD51 analyzed by SPR. The reaction conditions and the binding curve descriptions are same as in A. (C) ssDNA binding and dissociation curves for HsRAD51 in the presence if indicated concentrations of HsRAD51B-HsRAD51C at 25mM K⁺ (KCl). (D) ssDNA binding and dissociation curves for HsRAD51 in the presence if indicated concentrations of HsRAD51B HsRAD51C at 100mM K⁺ (KCl).

Figure 3. HsRAD51B-HsRAD51C enhances HsRAD51 catalyzed D-loop formation in the presence of HsRPA

(A) Schematic diagram of the experimental approach. Asterisk indicates the [³²P] label on ssDNA. Utilized proteins and DNA substrates are illustrated with labels. B-C, HsRAD51B-HsRAD51C heterodimer. HsRAD51 (1μM) filaments were formed with HsRAD51B-HsRAD51C at the indicated concentrations (if added), on 90mer ssDNA (3μM nt) by incubating at 37°C for 15min in reaction buffer containing 25mM Tris-OAc (pH 7.5), 1mM ATP, 1mM Mg(OAc)₂, 1mM CaCl₂, 2mM DTT and BSA (100μg/mL). After 15min, RPA (200nM) was added (if indicated) and D-loop formation was initiated by the addition of supercoiled dsDNA (50μM bp) and incubation at 37°C for 15min. D-loops were analyzed by electrophoresis on a 0.9% agarose gel after deproteinization. Gel was dried and exposed to a PhosphoImager screen for quantification. (C) Quantification of D-loops formed in B. Error bars indicate S.D.

Figure 4. HsRAD51B-RAD51C protects HsRAD51 nucleoprotein filament against anti-recombinase function of BLM

(A) Experimental scheme. Asterisk denotes the [³²P] label on ssDNA. Utilized proteins and DNA substrates are illustrated with labels. B-C, HsRAD51B-HsRAD51C heterodimer. (B) HsRAD51 (1μM) filaments were formed with HsRAD51B-HsRAD51C (31.2 μM) if indicated, on 90mer ssDNA (3μM nt) by incubating at 37°C for 15min in reaction buffer containing 25mM Tris-OAc (pH 7.5), 1mM ATP, 1mM Mg(OAc)₂, 2mM DTT, BSA (100μg/mL), 20mM phosphocreatine and creatine phosphokinase (30U/mL). After the 10min incubation BLM at indicated concentrations and HsRPA (200nM) were added and incubated further for 10min at 37°C. 1mM CaCl₂ was then added and followed by another 10min incubation at 37°C. D-loop formation was initiated by the addition of supercoiled dsDNA (50μM bp) and incubation at 37°C for 15min. D-loops were analyzed by electrophoresis on a 0.9% agarose gel after deproteinization. Gel was dried and exposed to a PhosphoImager screen for quantification. (C) Quantification of relative amounts of D-loops formed in B. Error bars indicate S.D.

Figure 5. HsRAD51B-HsRAD51C does not suppress the overall ATPase activity of the HsRAD51/HsRAD51B-HsRAD51C nucleoprotein filament

ATPase assays were performed by incubating HsRAD51 (1μM) and HsRAD51B-HsRAD51C at the indicated concentrations with oligo dT₅₀ (3μM nt), ϕX174 ssDNA (3μM nt) or ϕX174 dsDNA (3μM bp) in reaction buffer containing 25mM Tris-OAc (pH 7.5), 1mM Mg(OAc)₂, 2mM DTT, BSA (100μg/mL) and 200μM ATP supplemented with 200nCi of [γ-³²P] ATP for 1hr at 37°C. The reaction was stopped by the addition of 400μL 10% activated charcoal supplemented with 10mM EDTA and incubated on ice for 2 hrs. Following centrifuging for 10 min, 50μL duplicate aliquots were taken for counting [³²P] free phosphate by Cerenkov method. ATP hydrolyzed percentage is indicated.

Figure 6. HsRAD51B-HsRAD51C does not alleviate inhibitory effects of HsRPA to promote HsRAD51 mediated D-loop formation

(A) HsRPA 200nM, 400nM, 800nM and 1600nM (lanes 2-5 and lanes 7-10) was incubated with 90mer ssDNA (3μM nt) at 37°C for 10min in reaction buffer containing 25mM Tris-OAc (pH 7.5), 1mM ATP, 1mM Mg(OAc)₂, 1mM CaCl₂, 2mM DTT and BSA (100μg/mL) for 5min. After 5min, HsRAD51 (1μM) and HsRAD51B-HsRAD51C (32.5 nM) if indicated, was added and incubation was continued for another 15min at 37°C. Reaction in lane 6, RAD51 was incubated with 90mer ssDNA (3μM nt) at 37°C for 10min in reaction buffer. Reaction in lane 11, HsRAD51 (1μM) and HsRAD51B-HsRAD51C (32.5nM) incubated with 90mer ssDNA (3μM nt) at 37°C for 10min in reaction buffer, prior to addition of RPA and further incubation at 37°C for 5min. D-loop formation was initiated by the addition of supercoiled dsDNA (50μM bp) and incubation at 37°C for 15min. D-loops were analyzed by electrophoresis on a 0.9% agarose gel after deproteinization. Gels were dried and exposed to PhosphoImager screens for quantification. (B) Quantification of D-loops formed in A. Error bars indicate S.D.

Figure 7. Proposed model for HsRAD51B-HsRAD51C mediated stabilization of the HsRAD51 nucleoprotein filament

HsRAD51 mediated D-loop formation that can be enhanced by HsRAD51B-HsRAD51C paralogs. The stable nucleoprotein filament that forms could antagonize the action of anti-recombinases, a property that may be enhanced by additional cofactors. Proteins are illustrated with labels. B-C denotes HsRAD51B-HsRAD51C heterodimer.