Distinct roles of the two BRCA2 DNA-binding domains in DNA damage repair and replication fork preservation

SUMMARY

Homologous recombination (HR) removes DNA double-strand breaks (DSBs) and preserves stressed DNA replication forks. Successful HR execution requires the tumor suppressor BRCA2, which harbors distinct DNA-binding domains (DBDs): one that possesses three oligonucleotide/oligosaccharide-binding (OB) folds (OB-DBD) and another residing in the C-terminal recombinase binding domain (CTRB-DBD). Here, we employ multi-faceted approaches to delineate the contributions of these domains toward HR and replication fork maintenance. We show that OB-DBD and CTRB-DBD confer single-strand DNA (ssDNA)- and dsDNA-binding capabilities, respectively, and that BRCA2 variants mutated in either domain are impaired in their ability to load the recombinase RAD51 onto ssDNA pre-occupied by RPA. While the CTRB-DBD mutant is modestly affected by DNA break repair, it exhibits a strong defect in the protection of stressed replication forks. In contrast, the OB-DBD is indispensable for both BRCA2 functions. Our study thus defines the unique contributions of the two BRCA2 DBDs in genome maintenance.

INTRODUCTION

Cells employ several mechanistically distinct tools for the elimination of DNA double-strand breaks (DSBs), lesions that can induce chromosome aberrations and rearrangements. One such tool is homologous recombination (HR), capable of faithfully restoring DNA integrity at DSB sites.^1–3 The HR machinery is also important for the removal of interstrand DNA crosslinks and the protection of stressed and damaged replication forks against nucleolytic attack.^4,5 Given its broad involvement in genome maintenance, HR dysfunction can foster neoplastic transformation of cells and oncogenesis. A plethora of cancer-associated mutations have been identified in HR factors.^3,6,7 In particular, somatic and germline mutations in the BRCA1 and BRCA2 genes, which encode proteins indispensable for HR, crosslink repair, and replication fork preservation, lead to genome destabilization and cancer in the breast, ovary, pancreas, and other organs.^3,6,7

Initiation of HR requires nucleolytic resection of the 5′-terminated DNA strands at DSBs to generate 3′ single-strand DNA (ssDNA) tails.^8,9 These ssDNA tails then serve as the template for the assembly of a nucleoprotein filament comprising multiple units of the recombinase RAD51, termed the presynaptic filament. The presynaptic filament mediates the search for an information donor sequence in a homologous chromatid and initiates the invasion of the latter to form a DNA joint.^1,6 Subsequent steps include repair DNA synthesis and resolution of DNA intermediates to yield repaired products.⁶ Thus, presynaptic filament assembly represents a crucial step in the successful execution of HR. There is considerable evidence that the formation of a RAD51-ssDNA nucleoprotein filament on a stressed or damaged replication fork that has undergone regression is necessary to protect such regressed forks against digestion by cellular nucleases.^4,6

Presynaptic filament assembly is prone to interference by the ssDNA-binding protein RPA.^6,10 Owing to these constraints, cells have evolved mechanistically distinct protein factors, termed HR mediators, to facilitate RPA replacement by RAD51 on ssDNA.^11–13 BRCA2 (3,418 amino acid residues) functions in conjunction with its obligatory partner DSS1 (a 70-residue, highly acidic polypeptide) to mediate RPA-RAD51 exchange on ssDNA to facilitate presynaptic filament assembly.¹¹ BRCA2 provides the DNA-binding and RAD51 interaction attributes of the complex, while DSS1 serves as the RPA recognition module.¹¹ Moreover, DSS1, through its association with a helical domain and oligosaccharide-binding (OB) fold 1 (HD-OB1) in BRCA2, strongly attenuates the dsDNA-binding activity of the BRCA2 OB-fold domain.¹⁴ Importantly, BRCA2 is also important for the protection of stressed replication forks against attack by nucleases such as MRE11.^6,15,16

BRCA2 harbors two distinct classes of RAD51 interaction modules, namely, eight structurally related BRC repeats that can each associate with RAD51^17–20 and the standalone C-terminal recombinase binding domain (CTRB), encoded by exon 27.^1,3,15 Early studies identified the BRCA2 DNA-binding domain that harbors three OB folds,^21,22 which we refer to as the OB-DBD. Structural and biochemical evidence shows that OB folds 2 and 3 (OB2 and OB3) in the OB-DBD are directly involved in ssDNA binding.^21,22 Importantly, our recent studies have uncovered a distinct DNA-binding activity residing within the CTRB, termed the CTRB-DBD.¹⁵ Others have suggested that the CTRB-DBD enables BRCA2 to diffuse along dsDNA in search of ssDNA to initiate RAD51 loading.²³ Mutations that impair either the RAD51 interaction or DNA-binding attribute of the CTRB exert a negative impact on HR mediator activity of BRCA2 in vitro but only modest HR deficiency in cells and cellular resistance to DNA-damaging agents.¹⁵ However, the simultaneous inactivation of both CTRB attributes leads to more pronounced phenotypic consequences in vitro and in cells.¹⁵ Separation-of-function CTRB mutations that specifically impair either RAD51 interaction^15,16 or DNA binding¹⁵ affect the protection of replication forks. These findings help explain why deletion of BRCA2 gene exon 27, which encodes the CTRB, confers defects in HR, DNA damage repair, and cellular sensitivity to DNA damage.^24–27

Here, we conduct biochemical and cellular analyses of OB-DBD and CTRB-DBD mutants to gain insights into the contributions of these two BRCA2 domains to biological functions in HR, DNA damage repair, and replication fork preservation. Our studies reveal that while the OB-DBD is crucially important for optimal DNA damage repair and replication fork protection, the CTRB-DBD is more relevant for preserving the integrity of replication forks but fulfills a lesser role in DSB repair. Thus, our findings provide the first evidence for the distinct roles of the two BRCA2 DNA-binding domains in genome maintenance.

RESULTS

BRCA2-derived polypeptides for biochemical analyses

Previous studies have established that the HR mediator activity of BRCA2 can be interrogated using polypeptides that harbor key functional domains (Figure 1A). We have used BRC4-OB-DBD (consisting of the BRC4 repeat fused to the OB-DBD) and miniBRCA2 (BRC4 fused to the OB-DBD and CTRB-DBD) (Figure 1B) in this study to define contributions of the OB-DBD and CTRB-DBD toward DNA binding affinities and specificities of BRCA2, as well as the HR mediator activity of BRCA2. These BRCA2 polypeptides were expressed in insect cells and purified to near homogeneity using procedures that we have devised (Figure 1B). We note that all BRCA2-derived polypeptides that harbor the OB-DBD were co-expressed with DSS1 and co-purified as stoichiometric complexes. This expression platform is equally suited to obtaining wild-type (WT) and mutant variants of the salient BRCA2 polypeptides (see STAR Methods).

Figure 1. Mutagenesis of OB-DBD and CTRB-DBD for DNA binding analysis.

(A) Schematic of DNA-binding domains in BRCA2 and mutations in OB-DBD and CTRB-DBD. Residues in OB fold 2 (F2841, W2990, R2991, K3017, and K3019), OB fold 3 (F3090, K3104, R3128, and F3139), and CTRB-DBD (K3266, K3267, R3268, and R3269) were mutated to alanine.

(B) Schematic of BRC4-OB-DBD and BRC4-OB-DBD-CTRB (miniBRCA2) and their mutant variants generated for biochemical studies (i). SDS-PAGE showing purified wild-type BRC4-OB-DBD/DSS1 and miniBRCA2/DSS1 (ii). Purity analysis of mutants is presented in Figure S1B.

Specificity of BRCA2 OB-DBD for ssDNA

As determined by X-ray crystallography, OB2 and OB3 of mouse Brca2 directly contact ssDNA.²¹ In isolation, OB2 and OB3 of human BRCA2 bind ssDNA, and compound point mutations that impair the activity of these two OB folds have been described²² (Figures 1A and S1A). Accordingly, we generated alanine substitution variants (OB2-m: F2841A, W2990A, R2991A, K3017A, and K3019A; OB3-m: F3090A, K3104A, R3128A, and F3139A; OB-DBD-9A: combining OB2-m and OB3-m) within the context of the BRC4-OB-DBD polypeptide in complex with DSS1 (Figures 1A, 1B, and S1B). We tested these mutant polypeptides, alongside the WT counterpart, for the binding of an 80-bp ssDNA substrate in an electrophoretic mobility shift assay (EMSA). The results showed that all three mutant BRC4-OB-DBD polypeptides are significantly impaired for ssDNA-binding activity (Figure 2A). Specifically, while the OB3-m variant of BRC4-OB-DBD retains residual ssDNA-binding activity (about 40% at the highest concentration of 800 nM), the OB2-m variant failed to bind the substrate even at the highest concentration (800 nM) tested (Figure 2A). Importantly, we found the OB-DBD-9A mutant to be defective in ssDNA binding (Figure 2A). Our biochemical results thus demonstrate that both OB2 and OB3 contribute to ssDNA binding within the context of the OB-DBD.

Figure 2. Role of OB2 and OB3 in ssDNA binding.

(A) Results from EMSA testing BRC4-OB-DBD WT, OB2-m, OB3-m, and OB-DBD-9A for ssDNA binding (i) and their quantification (ii). Data are represented as mean ± SD (n = 3, error bars = SD).

(B) Results from EMSA testing BRC4-OB-DBD WT, OB2-m, OB3-m, and OB-DBD-9A for dsDNA binding (i) and their quantification (ii). Data are represented as mean ± SD (n = 3, error bars = SD).

We next investigated whether BRC4-OB-DBD in complex with DSS1 also has affinity for an 80-bp dsDNA. The results revealed that it has a much lower affinity for this substrate than for ssDNA (Figure 2B). While ~90% of the ssDNA was bound by 400 nM of BRC4-OB-DBD, ~60% of the dsDNA was bound by as much as 2 μM of this polypeptide (Figure 2B). Even the OB-DBD-9A mutation only engendered a modest defect in dsDNA binding (Figure 2B). These results indicate that the OB2 and OB3 DNA-binding residues under study contribute more prominently to the recognition of ssDNA. Using a pull-down assay, we interrogated the potential impact of these DNA-binding mutations on the ability of BRC4-OB-DBD to interact with the recombinase RAD51 (schematic shown in Figure S2Ai). We find that OB-fold mutant variants (OB2-m, OB3-m, and OB-DBD-9A) of BRC4-OB-DBD exhibit similar levels of interaction with RAD51 compared to the WT (Figure S2A). This finding supports that the residues we identified in the OB-DBD uniquely contribute to DNA, but not RAD51, binding.

We note that even though HD-OB1 within the OB-DBD has significant affinity for dsDNA, a recent study has shown that this activity is strongly attenuated by DSS1.¹⁴ Therefore, the DNA-binding effect of HD-OB1 was not detected in our analyses.¹⁴ Our biochemical results suggest that the OB-DBD utilizes OB2 and OB3 to bind ssDNA. We observe that these OB folds also bind dsDNA but with less propensity than that noted for ssDNA. The mutation of nine select residues (OB-DBD-9A mutant) significantly reduces dsDNA-binding affinity, although not to the same extent as with ssDNA (Figures 2 and 3). We surmise that the 3-helix bundle in the OB2-appended Tower domain²¹ or HD-OB1¹⁴ may contribute to dsDNA binding by the complex of OB-DBD and DSS1.

Figure 3. DNA binding by miniBRCA2 with OB fold and CTRB-DBD mutations.

(A) Quantification of EMSA results for binding of ssDNA (i) and dsDNA (ii) by miniBRCA2 without or with OB2-m, OB3-m, or OB-DBD-9A mutation. Gel images are shown in Figure S3A and show a single, representative run for each indicated experimental comparison. Data are represented as mean ± SD (n = 3, error bars = SD).

(B) Quantification of EMSA results for binding of ssDNA (i) and dsDNA (ii) by miniBRCA2 bearing the OB-DBD-9A, CTRB-DBD-4A, or both mutations, alongside the wild-type protein. Gel images shown in Figure S3B illustrate a single, representative run for each indicated experimental comparison. Data are represented as mean ± SD (n = 3, error bars = SD).

Contributions of OB-DBD and CTRB-DBD to DNA binding affinity and specificity

We have shown previously that the CTRB-DBD alone is a standalone functional module capable of binding ssDNA and dsDNA.¹⁵ The CTRB-DBD-4A mutant, which harbors changes in four DNA-binding amino acid residues (³²⁶⁶KKRR to ³²⁶⁶AAAA), is impaired for ssDNA and dsDNA binding.¹⁵ Here, we strived to determine the contributions of OB-DBD and CTRB-DBD to DNA binding affinity and specificity within the context of miniBRCA2. In particular, we sought to highlight the cooperativity between the two BRCA2 DNA-binding domains. We tested miniBRCA2 variants that harbor salient OB-DBD and CTRB-DBD mutations to examine how they impact ssDNA and dsDNA engagement (Figures 1B and S1B). First, we determined that miniBRCA2 exhibits robust ssDNA-binding activity, with nearly all ssDNA substrates shifted in mobility at 100 nM protein (Figures 3A and S3A). All three OB-fold variants show defective binding of ssDNA (Figures 3A and S3A), but they retain a higher degree of ssDNA binding compared to the corresponding BRC4-OB-DBD mutants (Figures 2A and 3A). We surmise that the presence of intact CTRB-DBD within each OB-fold mutant variant is responsible for the residual affinity for ssDNA in these miniBRCA2 species. The miniBRCA2 OB3-m mutant shows a lesser ssDNA binding defect in comparison to the OB2-m mutant (Figures 3Ai and S3Ai), which recapitulates the trend noted for the BRC4-OB-DBD variants (Figure 2A). Taken together, these results support the premise that, within the context of miniBRCA2, the OB-DBD makes a major contribution toward ssDNA engagement.

Our EMSA results show that miniBRCA2 binds dsDNA with avidity (Figure 3Aii and S3Aii). Given the OB-DBD has only weak affinity for dsDNA (Figure 2B), we posited that the CTRB-DBD represents the major dsDNA-binding entity within miniBRCA2. Indeed, the miniBRCA2 OB2 and OB3 mutants exhibited comparable affinity for dsDNA relative to their WT counterpart, and even the OB-DBD-9A mutant only showed a modest deficiency in dsDNA binding (Figures 3Aii and S3Aii). These observations implicate the CTRB-DBD as the predominant dsDNA-binding entity within miniBRCA2.

We examined the impact of the CTRB-DBD-4A mutation, alone or in combination with the OB-DBD-9A mutation, on ssDNA and dsDNA binding by miniBRCA2 (Figure 1). This analysis confirmed that while the CTRB-DBD-4A mutation has no impact on ssDNA binding, it greatly affects the affinity for dsDNA (Figures 3B and S3B). Notably, the OB-DBD-9A+CTRB-DBD-4A (termed 9A+4A) double mutant is quite defective in binding either ssDNA or dsDNA (Figures 3B and S3B). We also examined miniBRCA2 variants that harbor the OB-DBD-9A mutation, the CTRB-DBD-4A mutation, or both mutations for their affinity for a partially duplex substrate with a 3′ ssDNA overhang (Figure S3C). The results revealed moderate and little impairment of substrate binding by the OB-DBD-9A mutation and CTRB-DBD-4A mutation, respectively, while combining these mutations engendered a strong defect in substrate engagement (Figure S3C).

Contributions of OB-DBD and CTRB-DBD to HR mediator activity

We showed previously that the CTRB-DBD-4A mutation exerts a negative impact on the HR mediator attribute of BRCA2-derived polypeptides that harbor the CTRB.¹⁵ In this present study, we determined the effect of the OB-DBD mutations on the HR mediator attribute of BRC4-OB-DBD (Figure S2B) and miniBRCA2 (Figure 4A) polypeptides in complex with DSS1, which helps target BRCA2 to RPA-coated ssDNA but has no mediator activity on its own.¹¹ In this analysis, RPA was added with RAD51 to a 150-bp ssDNA substrate to impede assembly of the presynaptic filament, followed by the addition of the indicated BRCA2 polypeptide and an incubation, before homologous duplex DNA was incorporated to initiate DNA strand exchange (see Figure 4Ai for schematic). As expected, the co-addition of RPA with RAD51 to ssDNA greatly suppressed the formation of the DNA strand exchange product (Figures S2Bi and S2Bii, lane 3). Importantly, while the addition of 500 nM of BRC4-OB-DBD led to significant restoration of DNA strand exchange, the OB2-m, OB3-m, and OB-DBD-9A variants were all impaired for the HR mediator attribute, with the OB2-m and OB-DBD-9A mutants exhibiting a more severe defect than the OB3-m mutant in this regard (Figures S2Bi and S2Bii). The observed strand exchange defect is attributable to impaired DNA binding and not RAD51 interaction, as, by affinity pulldown, we showed that all three mutant BRC4-OB-DBD polypeptides retain full ability to interact with RAD51 (Figure S3D). We note that BRC4 without the OB-DBD is devoid of HR mediator activity,¹³ and BRC4 can, in fact, disrupt the integrity of RAD51-DNA nucleoprotein filaments.^16,28,29

Figure 4. Roles of the OB-DBD and CTRB-DBD in ssDNA targeting of RAD51.

(A) Strand exchange assay to test recombination mediator activity by miniBRCA2 and variants harboring the OB-DBD-9A, CTRB-DBD-4A, or the B-DBD-9A/CTRB-DBD-4A (9A+4A) double mutation. The reaction schematic is shown in (i), and results are presented in (ii). Results were quantified (iii). Data are represented as mean ± SD (n = 3, error bars = SD).

(B) Schematic of RAD51 ssDNA targeting assay to test ability of miniBRCA2 (wild-type [WT] protein or the OB-DBD-9A, CTRB-DBD-4A, or the OB-DBD-9A/CTRB-DBD-4A (9A+4A) mutant) to nucleate RAD51 onto ssDNA in the presence of an 80-bp dsDNA trap (i). SDS-PAGE to resolve proteins (supernatant containing unbound proteins; bead fraction with ssDNA-bound proteins) (ii). The dsDNA trap was revealed by Cy5 fluorescence (ii). Results were quantified (iii and iv). (iii) shows the percentage of RAD51 trapped on the beads for all samples. (iv) shows the same data but with the Trap sample removed to better visualize the comparison of miniBRCA2 variants with each other and the miniBRCA2-negative sample. For both quantification images, data are represented as mean ± SD (n = 3, error bars = SD).

Next, we tested the impact of the OB-fold double mutant (OB-DBD-9A), alone or in combination with the CTRB-DBD-4A mutation, on the HR mediator attribute of miniBRCA2. Given that miniBRCA2 harbors both the OB-DBD and CTRB-DBD, it is significantly more efficacious than BRC4-OB-DBD in restoring DNA strand exchange upon co-incubation of the ssDNA template with RPA and RAD51. Specifically, 200 nM of miniBRCA2 (Figure 4A), in contrast to 500 nM of BRC4-OB-DBD (Figure S2B), is needed for significant restoration of DNA strand exchange. Importantly, at a 200 nM concentration of miniBRCA2, the OB-DBD-9A mutation imparts as strong a deficiency in HR mediator activity as the CTRB-DBD-4A mutation, and combining these two mutations further compromises mediator activity (Figure 4A). The same general trend in HR mediator efficacy was seen when the miniBRCA2 variants were tested at a 300 nM concentration (Figure 4A).

Impact of OB-DBD and CTRB-DBD mutations in ssDNA targeting of RAD51

We previously showed that the CTRB facilitates the nucleation of RAD51 onto ssDNA, a critical step in generating the presynaptic filament.¹⁵ Here, we sought to elucidate what contributions the two BRCA2 DNA-binding domains make in RAD51 targeting to ssDNA for subsequent nucleoprotein filament assembly. We utilize an in vitro experimental protocol (Figure 4Bi) to interrogate how BRCA2 polypeptides nucleate RAD51 onto ssDNA in the presence of a dsDNA trap. We employed this analytical tool to elucidate how the loss of DNA binding by the OB-DBD or CTRB-DBD would impair the nucleation of RAD51 onto ssDNA (Figure 4B). We found that miniBRCA2 variants that harbor either the OB-DBD-9A or CTRB-DBD-4A mutation are only moderately impaired in RAD51 ssDNA targeting (Figure 4Bii), but the mutant polypeptide that harbors both mutations is quite defective in this regard (Figures 4Bii–4Biv).

Role of the OB-DBD and CTRB-DBD in DNA damage repair and HR

To test the impact of the OB-DBD mutations on the biological functions of BRCA2, we generated DLD1 cells stably expressing the salient mutants in full-length BRCA2 protein (Figures 5A and 5B). We transfected DLD1 cells (BRCA2^−/−) with the salient expression vectors and then isolated stable clones expressing equivalent levels of WT, OB-DBD-9A mutant, and CTRB-DBD-4A mutant versions of BRCA2 (Figures 5B and S4A). We note that all three ectopically expressed BRCA2 species localize to the nucleus properly (Figure S4B). Notably, cells expressing the BRCA2 9A+4A double mutant show severe growth defects and cell morphology changes (Figure S4C). In fact, cells expressing the BRCA2 double mutant are more growth impaired than even BRCA2^−/−cells. These poor growth characteristics of the double mutant render them unsuitable for cell biological studies. Our observations also provide evidence that simultaneous inactivation of the two BRCA2 DNA-binding domains strongly impacts cell health and that the double mutant exerts a dominant-negative effect in this regard.

Figure 5. Mutation of the OB-DBD compromises cellular HR function.

(A) Schematic of the OB-DBD-9A and CTRB-DBD-4A mutant proteins tested in cell-based analyses.

(B) Western blot of DLD1 cells, expressing wild-type BRCA2 or one of the DNA binding mutants, that were used for clonogenic survival and CRISPR-Cas9 gene integration assay studies. Anti-BRCA2 and anti-PALB2 antibodies were used for the blot. Ponceau staining is used as the loading control. EV, empty vector. The asterisk (*) denotes a non-specific band. All three BRCA2 species are present in the nuclear fraction (see Figure S4B).

(C) Clonogenic survival of DLD1 cell lines upon treatment with mitomycin C (MMC) (i), camptothecin (CPT) (ii), and the PARP inhibitors olaparib (iii) and rucaparib (iv). For each experiment, data are represented as mean ± SD (n = 3, error bars = SD).

(D) Schematic of CRISPR-Cas9 gene integration assay used to test HR proficiency of DLD1 cell lines (i). Results are shown in (ii). Data are represented as mean ± SD (n = 3, error bars = SD). ns p > 0.05, **p < 0.01, and ****p < 0.0001.

(E) RAD51 focus formation was accessed with or without exposure to MMC. Shown are representative micrographs of RAD51 foci (i) and quantified results (ii). Representative micrographs of untreated cells are presented in Figure S4D. Scale bar: 10 μm. Data are represented as mean ± SD (n = 3–7 signified by each bar, error bars = SD). ns p > 0.05.

With the requisite cell lines in hand, we first sought to determine the impact of OB-DBD and CTRB-DBD mutations on the sensitivity of cells to several DNA-damaging agents, namely camptothecin (CPT), mitomycin C (MMC), and the PARP inhibitors olaparib and rucaparib (Figures 5C and S4C). This analysis revealed that the OB-DBD-9A mutation engenders a high degree of sensitivity to all four DNA-damaging agents, being equivalent to the hypersensitivity observed for control DLD1 cells (BRCA2^−/−) transfected with the empty vector (Figures 5C and S4C). In comparison, and in concordance with what we have reported recently,¹⁵ the CTRB-DBD-4A mutation engenders only mild DNA damage sensitivity to the same panel of DNA-damaging agents (Figures 5C and S4C). These results suggest that DNA binding by the OB-DBD plays a crucial role in DNA damage repair and ultimately supports cell viability upon the occurrence of a variety of genomic insults.

Next, we used a CRISPR-Cas9-based gene targeting assay³⁰ to interrogate how the OB-DBD and CTRB-DBD mutations affect HR efficiency (Figure 5D). In agreement with our previous studies,¹⁵ we found that DLD1 cells expressing the CTRB-DBD-4A mutant exhibit a modest decrease in HR efficiency (Figure 5D). Importantly, DLD1 cells expressing the OB-DBD-9A mutant are markedly impaired in HR (Figure 5D), which parallels the striking sensitivity to DNA-damaging agents observed for these mutant cells. We also determined the ability of DLD1 cells of different genetic lineages to assemble nuclear RAD51 foci upon MMC treatment (Figures 5E and S4D). In concordance with results from other cellular analyses (Figures 5C and 5D), the OB-DBD-9A mutation eliminates the formation of RAD51 foci in response to treatment with MMC, while the CTRB-DBD-4A mutation mildly affects focus formation in the treated cells (Figures 5E and S4D). Our results thus reveal a prominent role for ssDNA binding via OB-DBD to drive HR-mediated repair of DNA damage, while DNA binding by the CTRB-DBD makes a lesser contribution in this regard.

We note that the CTRB-DBD-4A mutation imparts a modest HR deficiency, even though this same mutation significantly impairs the HR mediator activity of miniBRCA2 (Figure 4A) or CTRB in vitro.¹⁵ We surmise that the mild cellular HR defect engendered by the CTRB-DBD-4A mutation is due to another HR factor, such as the BRCA1-BARD1 complex³¹ or PALB2,^32,33 which both bind dsDNA and together form a higher-order complex with BRCA2,^3,6 providing the same function in parallel.

Involvement of OB-DBD and CTRB-DBD in replication fork preservation

BRCA2 fulfills a prominent role in the protection of stressed replication forks against nucleolytic attrition, predominantly by the MRE11 nuclease.^34–36 Here, we used the DNA fiber assay to determine the role of the two BRCA2 DNA-binding domains in the protection of replication forks upon treatment of cells with hydroxyurea (HU), a potent replication stressor. Cells were sequentially pulsed with CIdU and IdU before being exposed to 2 mM HU for 5 h, and IdU/CldU tract ratios were determined to assess the extent of fork degradation (Figures 6Ai and S4E). As shown in Figure 6Aii, parental DLD1 cells (BRCA2^−/−) or DLD1 cells ectopically expressing WT BRCA2, the OB-DBD-9A mutant, or CTRB-DBD-4A show no significant difference in IdU/CldU tract ratios in the absence of HU. Importantly, cells expressing either the OB-DBD-9A or CTRB-DBD-4A mutant are highly prone to replication fork attrition upon exposure to 2 mM HU (Figure 6Aii). We also verified that replication fork attrition in either OB-DBD-9A or CTRB-DBD-4A mutant cells could be alleviated to a large degree by treatment with the MRE11 nuclease inhibitor Mirin (Figure 6Aii). Collectively, these results support the notion that both DNA-binding domains of BRCA2 are indispensable for the protection of stressed replication forks against nucleolytic attack.

Figure 6. Impact of OB-DBD and CTRB-DBD mutations on replication fork preservation.

(A) Overview of DNA fiber assay to test replication fork protection in DLD1 cell lines and representative micrographs of IdU/CldU-labeled DNA fibers (i). Additional micrographs are presented in Figure S4E. Scale bar: 10 μm. Dot plots of IdU/CldU tract length ratios in DLD1 cells and derivatives left untreated, treated with 2 mM hydroxyurea (HU), or treated with 2 mM HU and 50 μM Mirin. Data are represented as individual IdU/CldU values (points) with means indicated by red bars (n = 3 with 150–500 labeled fibers analyzed per run) (ii). ns p > 0.05, *p < 0.05, **p < 0.01, and ****p < 0.0001.

(B) DNA protection activity of miniBRCA2. Sensitivity of dsDNA to MRN (4 nM), RAD51 (200 nM), BRC4 (1 μM), and miniBRCA2 (nM) was tested, as indicated. The reaction schematic is given in (i).

Denaturing gel electrophoresis was used to resolve reaction products (ii). Results were quantified and plotted (iii). Data are represented as mean ± SD (average of n = 3 [left], average of n = 2 [right], and error bars = SD).

Next, we followed the procedure of Halder et al.³⁴ to test miniBRCA2 and its mutant variants in conjunction with RAD51 for the protection of dsDNA against digestion by the MRE11-RAD50-NBS1 (MRN) nuclease complex. In these experiments, BRC4 was included as a RAD51-DNA destabilizer.^28,34 As shown in Figure 6B, RAD51 prevented MRN-mediated degradation, whereas BRC4 rendered the RAD51-dsDNA complex sensitive to MRN action. Notably, the addition of miniBRCA2, but not the variant harboring the CTRB-DBD-4A mutation, restored the DNA protective capacity (Figure 6B). Intriguingly, miniBRCA2 OB-DBD-9A exhibited WT-like activity in dsDNA protection (Figure 6B). Since miniBRCA2 alone could not prevent MRN-mediated degradation, the protective function requires both miniBRCA2 and RAD51. These results support the idea that BRCA2 relies upon the CTRB-DBD to prevent RAD51-bound DNA from nuclease digestion. We note that in cells, DNA binding by the OB-DBD could play a role in replication fork protection, likely by engaging unique DNA structures that arise during fork stalling, collapse, and regression.

DISCUSSION

The role of BRCA2 in DSB repair and replication fork preservation is well established.^1,3,6 However, how this tumor-suppressor protein utilizes its OB-DBD and CTRB-DBD^6,15,21 to orchestrate repair and replication fork protection remains to be defined. Here, we have addressed not only the contributions of these two BRCA2 DNA-binding domains in ssDNA and dsDNA engagement but also their functional significance in RAD51 presynaptic filament assembly and DNA damage repair, HR, and the protection of stressed replication forks against nucleolytic attrition within the cellular setting. Specifically, by constructing multiple OB-fold mutants within the context of functional BRCA2 polypeptides and testing them in a variety of reconstituted biochemical systems and cell-based assays, we have provided evidence that the OB-DBD is primarily involved in ssDNA engagement. In the cellular context, the OB-DBD appears to have a more critical role than the CTRB-DBD in guiding the completion of HR, though our biochemical studies indicate an importance for both DNA-binding domains in recombination mediator activity. Accordingly, mutations in the OB-DBD that ablate ssDNA binding compromise the DNA damage repair functions of BRCA2 in cells. We also provide evidence that inactivation of the CTRB-DBD DNA-binding attribute leads to only a modest reduction in RAD51 presynaptic filament assembly in vitro and in HR and DNA damage repair efficiency in cells. However, the CTRB-DBD mutation under study engenders a replication fork protection phenotype just as severe as the compound OB-DBD mutation that simultaneously ablates OB2 and OB3 function. Altogether, our results show that the OB-DBD and CTRB-DBD are not redundant entities but, rather, serve distinct biological roles, with the former being indispensable for both HR execution and replication fork protection and the latter being particularly germane for replication fork preservation. A model depicting the distinctive roles of the OB-DBD and CTRB-DBD in the functional engagement of HR and replication fork intermediates is presented in Figure 7.

Figure 7.

Model showing contributions of the BRCA2 DNA interaction via DBD and CTRB toward DNA repair by HR or replication fork protection and the outcome of the mutations disrupting the interactions

Limitations of the study

In our biochemical experiments, we utilized BRCA2 polypeptides containing the OB-DBD and CTRB-DBD. We did not test protein species with the N-terminal DNA-binding domain (NTD) of BRCA2.³⁷ We thus cannot rule out potential contributions of the NTD to DNA-binding and associated repair activity. In HR, BRCA2 functions within a larger complex of DNA repair factors.^3,6 Our biochemical studies focus on the biochemical function of BRCA2 but do not recapitulate the binding partners and other mediators, such as BRCA1-BARD1³¹ and PALB2,^32,33 that assist BRCA2 in its repair function.^3,6 This limitation could explain any discrepancies between our in vitro biochemical findings and our cell biological results. It will be important to understand how additional HR factors within a higher-order repair complex cooperate with the OB-DBD and especially CTRB-DBD, given the observed differences in biochemical and cell biological phenotypes of the CTRB-DBD-4A variant. We observed that the loss of DNA binding by the OB-DBD or CTRB-DBD leads to increased nuclease-mediated digestion of stressed replication forks. Prior studies have shown that a defective BRCA2-RAD51 interaction, notably by the BRCA2 CTRB region, ultimately leads to decreased nucleation of RAD51 onto dsDNA and consequently exposes this DNA to nucleases.³⁴ Further work is needed to clarify if DNA binding by BRCA2 primarily serves to nucleate RAD51 onto nascent DNA, if DNA binding by BRCA2 itself can protect nascent DNA, or if both factors play prominent roles in safeguarding newly replicated DNA. Additionally, evidence has been presented that the BRCA2-NTD is pertinent to replication fork protection by preventing ssDNA gap formation or accumulation.³⁸ The NTD may contribute an additional layer of regulatory control over the cellular functions of BRCA2.

RESOURCE AVAILABILITY

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Patrick Sung (sungp@uthscsa.edu).

Materials availability

All unique reagents generated in this study are available without restrictions, and requests can be made to the lead contact.

Data and code availability

Raw data and uncropped gel images for the figures in this paper are accessible via the Texas Data Repository (https://dataverse.tdl.org/) under the accession number DOI:10.18738/T8/HKOMWM (https://doi.org/10.18738/T8/HKOMWM). Additional information necessary for reanalyzing the reported data is available from the lead contact upon request.

EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS

Bacterial strains

E. coli Rosetta cells were obtained from Sigma-Aldrich, and cells were grown in 2X luria broth at 37°C. E. coli DH10Bac cells were obtained from Thermo Fisher Scientific, and cells were grown in 1X luria broth at 37°C.

Cell lines

Human DLD1 cells lacking endogenous BRCA2 (BRCA2^−/−) were obtained from Horizon Discovery. All DLD1 BRCA2^−/−cell variants were grown in RPMI 1640 medium with 10% FBS at 37°C in a humidified air incubator containing 5% CO₂.

Hi5 insect cells were obtained from Expression systems, and cells were grown in ESF 921 media at 27°C in a humidified air incubator containing 5% CO₂. SF9 insect cells were obtained from Thermo Fisher Scientific. Cells were grown in SF900 media at 27°C in a humidified air incubator.

METHOD DETAILS

Plasmids

We introduced cDNA encoding BRC4-OB-DBD or miniBRCA2 (Figure 1 and Table S1) into the MacroBac vector p438a.⁴² BRC4-OB-DBD and miniBRCA2 were co-expressed with GST-tagged DSS1 in Hi5 insect cells (Tni cells; Expression Systems). For cellular studies, full-length BRCA2, with an N-terminal 2xMBP tag and a C-terminal His6-Flag tag, was introduced into phCMV1 (Table S1). Site-directed mutagenesis (Gene Universal and Twist Bioscience) was used to generate mutations in the OB-DBD, CTRB-DBD, or both (Table S2) into the indicated BRC4-OB-DBD, miniBRCA2, and full-length BRCA2 constructs.

Mammalian cell culture

DLD1 cells (BRCA2^−/−, Horizon Discovery) were maintained in RPMI 1640 media (Gibco, Catalog #11875–093). To express full-length BRCA2, DLD1 cells were transfected with 2 μg of pCMV-2xMBP-BRCA2-His6-Flag (wild-type BRCA2 or the indicated mutant) using lipofectamine 2000 (Invitrogen) according to the manufacturer’s directions. Single colonies stably expressing the indicated full-length BRCA2 constructs were selected and maintained using G418 (0.5 mg/mL). DLD1 cell lines were grown at 37°C and 5% CO₂.

Expression and purification of recombinant proteins

Lysate preparation and protein purification steps were carried out between 0°C and 4°C. Purified proteins were concentrated to a small volume in an Amicon ultracentrifugation device (Millipore), snap frozen in 2–4 μL aliquots, and stored at −80°C. Once thawed, proteins were kept on ice and unused fractions were discarded after 48 h.

BRC4-OB-DBD and miniBRCA2 complexed with DSS1

Recombinant bacmid DNA encoding BRC4-OB-DBD, miniBRCA2, or DSS1 was generated from DH10Bac cells (Thermo Fisher Scientific) using p438a-BRC4-OB-DBD or miniBRCA2, or pDEST20-DSS1. Baculoviruses were amplified in SF9 cells (Thermo Fisher Scientific). Hi5 insect cells were co-transfected with baculovirus encoding BRC4-OB-DBD or miniBRCA2 and baculovirus encoding GST-DSS1 at 27°C for 48 h. Cells were harvested by centrifugation, and pellets were stored at −80°C. Cell lysate was prepared by using a Dounce homogenizer to resuspend the Tni cell pellets in buffer T (25 mM Tris-HCl, pH 7.5, 10% glycerol, 0.5 mM EDTA, 1 mM DTT, 0.01% Igepal CA–630) containing 300 mM KCl, 1 mM benzamidine, protease inhibitors (2 μg/mL each of aprotinin, chymostatin, leupeptin, and pepstatin A). Lysate was clarified by centrifugation at 40,000 rpm × 45 min × 4°C. The supernatant was incubated with 2 mL anti-Flag M2 affinity gel (Sigma-Aldrich) in lysis buffer for 2 h at 4°C on a rocking shaker. Resin was washed with 3 mL each of T buffer containing 300 mM KCl, 1000 mM KCl with 1 mM ATP and 8 mM MgCl₂, and finally 100 mM KCl. Protein was eluted from resin using buffer T containing 100 mM KCl and 0.5 mg/mL Flag peptide. Fractions of eluted protein were pooled and mixed with 2 mL of Glutathione Sepharose 4 Fast Flow (Cytiva) for 1.5 h on a rocking shaker. Glutathione resin was washed in buffer T containing 100 mM KCl, 1000 mM KCl with 1 mM ATP and 8 mM MgCl₂, and finally 100 mM KCl. Protein was eluted off glutathione resin using buffer T containing 100 mM KCl and 50 mM glutathione, pH 8.0. Eluted protein fractions were pooled, diluted 2-fold in buffer T, and loaded into a 1 mL Mono Q column (GE Healthcare). Bound protein was eluted using a salt gradient ranging from 50 mM KCl to 350 mM KCl. Peak protein fractions containing BRC4-OB-DBD/DSS1 or miniBRCA2/DSS1 were pooled and concentrated using a 30K Amicon Ultra centrifugal device (Millipore). Concentrated protein was aliquoted, snap frozen in liquid nitrogen, and stored at 80°C. This purification scheme was used to purify wild-type and all mutant variants of BRC4-OB-DBD/DSS1 and miniBRCA2/DSS1. Throughout each purification, BRC4-OB-DBD/DSS1 and miniBRCA2/DSS1 were verified by western blot with anti-Flag HRP antibody (Thermo Fisher Scientific) for BRCA2 peptides and anti-GST HRP conjugate (GE Healthcare) for DSS1.

RAD51

We expressed His6-Smt3-RAD51 using E coli Rosetta cells (Sigma-Aldrich).⁴³ Bacterial cells were cultured in 2X luria broth (supplemented with 100 μg/mL ampicillin) and induced by incubation with 0.5 mM IPTG for 16 h at 16°C. Bacterial cells were harvested by centrifugation (5000 rpm × 10 min), and pellets were stored at −80°C. For cell lysis, pellets were resuspended in buffer T containing 600 mM KCl, 6 U/μL benzonase, 1 mM MgCl₂, and protease inhibitors (1 mM benzamidine and 2 μg/mL each of aprotinin, chymostatin, leupeptin, and pepstatin A) and then subject to sonication. Lysate was clarified by centrifugation (45,000 rpm × 45 min). Supernatant from centrifugation was combined with (NH4)₂SO₄ (0.277 g per mL lysate) to precipitate protein. The precipitate was harvested by centrifugation 18,000 × g for 30 min. The pellet was dissolved in buffer T (containing 300 mM KCl) and incubated with 2 mL Ni-NTA Superflow resin (QIAGEN) for 2 h on a rocking shaker. The resin was washed with buffer T containing 500 mM KCl, 300 mM KCl, and finally 150 mM KCl. To remove the His6-Smt3 tag, RAD51 bound to resin was digested with Ulp1 (5 μg) in buffer T (150 mM KCl) for 16 h at 4°C RAD51 released from the resin was collected in buffer T containing 150 mM KCl. After adjusting the salt concentration to 150 mM KCl, the protein pool was fractionated in a 1 mL HiTrap Q HP column and eluted over a gradient of 100–500 mM KCl. Peak fractions containing protein were pooled, concentrated, and further fractionated in a 1 mL Heparin HP column, which was eluted with a salt gradient of 100–500 mM KCl. Peak fractions containing protein were pooled, concentrated, and stored.

RPA

We expressed and purified RPA using our published protocol.⁴⁴

MRN complex

The MRE11-RAD50(MBP-tag)-NBS1(FLAG tag) (MRN) complex and NBS1(FLAG tag) were expressed in Tni cells using baculoviruses. Pellets from cells expressing MRN (10 g of pellet) and NBS1 (5 g) were separately lysed in 100 mL of buffer K (20 mM KH₂PO₄, 10% glycerol, 0.5 mM EDTA, 0.05% tween 20, 1mM β-ME, 1 mM phenylmethylsulfonyl fluoride and 5 mg/mL each of aprotinin, chymostatin, leupeptin and pepstatin) containing 200 mM KCl, and the lysates were clarified by ultracentrifugation. For purification, MRN was partially purified using a 4 mL amylose resin for MBP affinity chromatography. We noticed that the complex contained a sub-stoichiometric amount of NBS1. NBS1-Flag was partially purified using anti-Flag agarose resin for Flag affinity chromatography. The eluted proteins were combined followed by further purification via a 1 mL Heparin column (Amersham) with a gradient of 100–600 mM KCl in buffer K over 15 mL. The peak fractions containing the MRN complex were pooled, concentrated, and stored at 80°C.

DNA substrates

HPLC-purified DNA oligonucleotides (Table S3) were purchased from IDT. Blunt-ended dsDNA substrate and partial dsDNA substrate with a 3′ ssDNA overhang were generated by annealing equimolar amounts of the indicated oligonucleotides as described.⁴⁵

Electrophoretic mobility shift assay:

BRCA2-derived polypeptides (BRC4-OB-DBD and miniBRCA2) were incubated with 5 nM of the indicated Cy5-labeled DNA substrate (ssDNA, dsDNA, partial duplex DNA with a 3′-overhang) (Table S3) in reaction buffer (35 mM Tris, pH 7.5, 1 mM DTT, 50 mM KCl, 1 mM MgCl₂, and 100 μg/mL BSA) for 10 min at 25°C. Unbound DNA and nucleoprotein complexes were resolved by electrophoresis on 8% polyacrylamide gels in TBE buffer (50 mM Tris-boric acid, pH 8.4, 0.5 mM EDTA) for 1 h at 40 mA. DNA species were visualized using the ChemiDoc imaging system (Bio-Rad). The proportion of bound versus unbound DNA was quantified using the Image Lab software (Bio-Rad). Mean bound DNA and standard deviation were calculated across 3 independent studies using Microsoft Excel and GraphPad Prism 9. Data were plotted as means on an X-Y scatterplot with a best fit curve, with error bars representing SD for each experimental sample.

DNA strand exchange assay

Unless stated otherwise, reaction steps were carried out at 37°C. To test for recombination mediator activity of BRCA2 polypeptides, strand exchange reactions were assembled (12.5 μL final volume) using 20 nM ssDNA (oligo 863 in Table S3) in buffer B (35 mM Tris, pH 7.5, 1 mM DTT, 50 mM KCl, 1 mM ATP, 1 mM MgCl₂, and 100 μg/mL BSA). The ssDNA was incubated with 1 μM RAD51 and 200 nM RPA for 10 min before BRC4-OB-DBD/DSS1 or miniBRCA2/DSS1 was added to the indicated concentrations. After a 5 min incubation, strand exchange was initiated by adding 20 nM of duplex DNA (5′ Cy5-labeled oligo 1056 with oligo 1057, Table S3) and 4 mM spermidine hydrochloride. Reactions were incubated for 30 min and mixed with 1 μL each of 1% SDS and proteinase K (10 mg/mL). Following a 5 min incubation, reaction mixtures were resolved in 8% polyacrylamide gels in TBE buffer (50 mM Tris-borate, pH 8.4, 0.5 mM EDTA). Cy5-labeled substrate and strand exchange product were visualized in the ChemiDoc imaging system (Bio-Rad). Proportion of product DNA formed was quantified and compared relative to the RAD51-only sample (lane 2). Mean relative strand exchange (SE) activity compared to RAD51-only activity was quantified across 3 separate studies using GraphPad Prism 9 and Microsoft Excel. Mean SE activity was plotted with individual data points and error bars (SD) shown for each experimental sample.

RAD51 ssDNA targeting assay

To assemble the reactions (final volume of 20 μL), 2.5 μM RAD51 was incubated with 0.25 μM miniBRCA2 in buffer C (35 mM Tris-HCl, pH 7.5, 1 mM MgCl₂, 1 mM ATP, 50 mM KCl, and 50 ng/μL BSA) for 10 min at 4°C. Then, 200 nM of 80-bp duplex dsDNA (Cy5-5′-labeled P1:P2, Table S3) and 50 nM of ssDNA (biotin-dT80)-immobilized on 4 μL Dynabeads M270 Streptavadin (Thermo Fisher Scientific) was added to each reaction followed by a 10 min incubation at 25°C. Dynabeads were captured using a Magnetic Particle Separator (Roche Applied Science), which were washed twice in buffer containing 35 mM Tris-HCl, pH 7.5, 1 mM MgCl₂, 0.1 mM ATP, and 50 mM KCl. Bound protein was eluted at 37°C for 3 min with 10 μL of SDS-PAGE loading buffer. The supernatant and eluate fractions were analyzed by SDS-PAGE with Coomassie blue staining. Following electrophoresis, Cy5-labeled dsDNA was visualized in the ChemiDoc imaging system. Mean proportion of RAD51 collected in eluate fractions was calculated for 3 separate experiments. GraphPad Prism 9 was used to tabulate SD and construct graphs depicting mean % RAD51 in eluate samples, with SD represented in accompanying error bars.

Affinity pulldown assay

The indicated BRCA2 polypeptide (2 μg) was incubated with RAD51 (6 μg) in a 25 μL reaction assembled in buffer T (150 mM KCl) and 6 U/μL benzonase for 30 min on ice. Following incubation, reactions were mixed with anti-Flag M2 affinity gel and incubated on a rocking shaker for 1 h at 4°C. Resin was washed three times with 100 μL buffer T (150 mM KCl) and then incubated with 10 μL of SDS-PAGE loading buffer at 37°C for 5 min to elute proteins. The supernatant and eluate fractions were analyzed by SDS-PAGE and Coomassie blue staining.

Cell fractionation

The Rapid, Efficient And Practical (REAP) method for the preparation of cytoplasmic and nuclear extracts was followed.⁴⁶ Briefly, DLD1 cells on 10-cm dishes were washed with ice-cold phosphate buffer saline (PBS) pH 7.4, collected by centrifugation, resuspended in 900 μL of ice-cold PBS with 0.05% NP40 and protease inhibitors, and triturated 5 times using a P1000 micropipette. The lysed cell suspension was centrifuged to separate the supernatant (representing the cytoplasmic fraction), and the pelleted nuclei were washed once with PBS containing 0.05% NP40 and lysed with NETN buffer (20 mM Tris-HCl pH 8, 420 mM NaCl, 1 mM EDTA, 0.5% Igepal CA630, 1 mM DTT) with protease inhibitors to yield the nuclear fraction. The cytoplasmic and nuclear fractions, 20 or 40 μg each, were analyzed by immunoblotting.

Clonogenic survival assay

DLD1 cells expressing full-length BRCA2 (wild-type or the indicated DNA binding mutant), or empty vector (phCMV1–2xMBP), were seeded in 6 well plates at 300 cells per well and incubated for 24 h at 37°C. The next day, cells were treated for 12 days with the indicated concentration of camptothecin (CPT), mitomycin C (MMC), olaparib, or rucaparib. Following drug treatment, cells were fixed in 100% methanol, then stained in 0.5% crystal violet in 20% methanol to visualize DLD1 colonies. To determine clonogenic survival for each drug treatment, the number of surviving colonies in experimental plates was compared to the number of surviving colonies in untreated control plates. The surviving percentage of cells was quantified across 3 independent experiments, and data were plotted, with standard deviation representing error bars, using GraphPad Prism software.

CRISPR-Cas9 gene targeting assay

DLD1 cells expressing 2xMBP-BRCA2-His6-Flag (wild-type or DNA binding mutant) or empty vector were seeded in 6-well plates at 4 × 10⁵ cells per well. Cells were incubated 37°C for 24 h and then were transfected with 4 μL lipofectamine 2000 (Invitrogen), and a 1.5:1 μg ratio of sgRNA plasmid px330-LMNA to pCR2.1-CloverLamin donor template.³⁰ After a 72 h incubation with gene transfer plasmids, cells were collected in 300 μm PBS containing 5% FBS for flow cytometry analysis. Cells positive for GFP were detected by flow cytometry, and the mean proportion of GFP+ cells in each sample was quantified across 3 independent experiments. Data are presented as the mean percent of GFP+ cells normalized to % of GFP+ of BRCA2 WT, with accompanying error bars signifying the standard deviation.

RAD51 foci

Eight-well Permanox chamber slides (LabTek) were seeded with 40,000 cells per chamber 48 h before treatment with 1 μM MMC for 16 h. Following treatment, cells were washed twice with PBS, fixed in 1% paraformaldehyde/2% sucrose in PBS for 15 min at room temperature, washed twice with PBS and permeabilized with methanol for 30 min at −20°C, as described.⁴⁷ Cells were washed twice in PBS before incubation in 0.5% Triton X-100/PBS for 10 min. Samples were blocked in 5% BSA/PBS for 30 min at 25°C before incubation with primary antibody (α-RAD51 from Bio-Academia 70–001, 1:6000) in 5%BSA/0.05% Triton X-100/PBS at 4°C over-night. Cells were then washed 3X in PBS and incubated with Alexa Fluor 594 goat anti-rabbit secondary antibody (Thermo Fisher Scientific; 1:750, Cat #A-11037) in 1%BSA/0.05% Triton X-100/PBS for 45 min at 25°C. After three washes with PBS, chamber slides were mounted in ProLong Gold with DAPI (Thermo Fisher Scientific). Images were taken using a 63× oil objective and a Zeiss Axio-Imager.Z2 microscope equipped with Zen Blue software (Carl Zeiss Microscopy). Images were obtained as z stack sections of 0.2 mm per section containing 18 Z-stacks for each channel. A maximum projection file was generated to identify and count the number of foci and nuclei, and nuclei with >5 RAD51 foci per nucleus were counted as positive. RAD51 foci from 100 nuclei were measured in three to seven independent experiments. Data were processed and plotted using GraphPad Prism 10. Standard deviations were calculated and presented as error bars together with the mean values. p values were calculated using Kolmogorov-Smirnov tests.

DNA fiber assay

Fiber analysis was used to probe replication tract stability as we have done previously.^45,48,49 DLD1 cells (EV or expressing full-length BRCA2) were pulse-labeled with 25 μM CldU (20 min), washed three times with PBS, and then pulse-labeled with 250 μM IdU (20 min) in RPMI growth medium. Following IdU labeling, cells were washed with PBS and treated with 2 mM HU (Sigma-Aldrich) or HU + mirin (50 μM, Sigma-Aldrich) in growth medium. Labeled cells were collected by scraping, resuspended in ice-cold PBS at 4 × 10⁵ cells/ml and 2 μL of this suspension were spotted onto a glass slide and lysed with 7 μL spreading buffer (0.5% SDS, 200 mM Tris-HCl, pH 7.4, 50 mM EDTA). Slides were incubated for 5 min, then tilted to spread labeled DNA fibers. Next, slides were air-dried and fixed in methanol:acetic acid (3:1), rehydrated in PBS for 10 min and denatured in 2.5 M HCl for 1 h at 21°C. Slides were then rinsed in PBS and blocked in PBS +0.1% Triton X-100 (PBS-T) + 5% BSA for 1 h (21°C). Rat anti-BrdU (1:100, Abcam ab6079) and mouse anti-IdU (1:100, Becton Dickinson #347580) were then applied to detect CldU and IdU, respectively. After a 1 h incubation, slides were washed in PBS and stained with Alexa Fluor 488-labeled goat anti-mouse IgG1 antibody and Alexa Fluor 594-labeled goat anti-rat antibody (1:300; Thermo Fisher Scientific). Slides were mounted in Prolong Gold Antifade (Thermo Fisher Scientific) and held at 4°C until image acquisition. Replication tracts were imaged on a Zeiss Axio-Imager.Z2 microscope equipped with ZEN Blue software (Carl Zeiss Microscopy) using a 63× oil objective. CldU and IdU tracts were measured using ImageJ software. Data are from 3 independent experiments with 150–500 CldU fibers measured. Data were plotted in GraphPad Prism 9. Statistical analysis was conducted using the Kruskal-Wallis test.

DNA protection assay

Each DNA protection experiment was performed at a final volume of 12.5 μL and contained Buffer DP (35 mM Tris, pH 7.5; 1 mM DTT; 5 mM MgCl₂; 1 mM MnCl₂; 1 mM ATP; 50 ng/μL BSA). The first stage of the reaction included dsDNA (Cy5-5′-H3:H4 at 5 nM final concentration (Table S3)) and RAD51 (200 nM final concentration) and was incubated at 37°C for 5 min. MiniBRCA2-DSS1 at the indicated final concentrations was then added and the mixture incubated at 37°C for 5 min. Next, BRC4 was added and the mixture incubated at 37°C for 5 min. Finally, MRN was added and the mixture was incubated for 2 h at 37°C. The reaction was stopped with the addition of 1 μL of 2% SDS and 1 μL of 10 mg/mL Proteinase K and incubation for 10 min at 37°C. 6.5 μL of loading buffer (95% formamide, 10 mM Tris, 1 mM EDTA, 0.025% Orange G) was then added to the mixture. The samples were then heated at 95°C for 3 min and then immediately placed in ice. The samples were separated in 15% denaturing polyacrylamide gels in TBE buffer at 150V for 1 h. The gels were imaged with a ChemiDoc MP using the Cy5 application and were analyzed with ImageLab software version 6.1. Mean values and standard deviations were calculated using GraphPad Prism.

QUANTIFICATION AND STATISTICAL ANALYSIS

GraphPad Prism software was used to calculate mean, standard deviation, and statistical significance for all experiments in this manuscript. All gels for in vitro biochemical assays were quantified using Image Lab (Bio-Rad). For clonogenic survival and gene transfer experiments, two-way ANOVA was used to test significance. For RAD51 foci experiments, Kolmogorov-Smimov testing was used to calculate statistical significance of results. For DNA fiber assay, ImageJ software was used to measure CldU and IdU tract length, and Kruskal-Wallis testing was utilized to determine significance. Values of n and p are provided in each figure legend as appropriate.

KEY RESOURCES TABLE

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Anti-Flag M2-HRP	Thermo Fisher Scientific	Cat # MA191878HRP; RRID: AB_2537626
Anti-GST-HRP GE Healthcare	Thermo Fisher Scientific	Cat # MA4004HRP; RRID:AB_2537626
Anti-RAD51	Bio-Academia	Cat# 70-001; RRID: AB_2177110
Anti-RAD51-HRP	Abcam	Cat# ab195548
Anti-rabbit Alexa Fluor 594	Thermo Fisher Scientific	Cat# A-11037; RRID: AB_2534095
Anti-MBP-HRP Novus Biologicals	Thermo Fisher Scientific	Cat# NB10066609H; RRID: AB_3170039
Anti-BRCA2	EMD Millipore	OP95; RRID: AB_2067762
Anti-PALB2	Huo et al.39	N/A
Anti-PARP1	Abcam	Cat# ab6079; RRID: AB_305284
Anti-MSH2	Abcam	Cat# ab52266; RRID: AB_2144800
Anti-H3	Abcam	Cat# ab1791; RRID: AB_302613
Rat Anti-BrdU (for CldU detection)	Bio-Rad	Cat# MCA2060T; RRID: AB_566394
Ant-BrdU(B44) (for IdU detection)	Becton Dickinson	Cat# BD347580; RRID: AB_10015219
Goat anti-mouse Alexa Flour 488	Thermo Fisher Scientific	Cat# A-11001; RRID: AB_2534069
Goat anti-rat Alexa Fluor 594	Thermo Fisher Scientific	Cat# A-11007; RRID: AB_10561522
Bacterial and virus strains
E.coli: DH10Bac Competent Cells MAX Efficiency	Thermo Fisher Scientific	Cat# 10-361-012
E.coli: Rosetta (DE3) Competent Cells	Sigma-Aldrich	Cat# 70954-3
Chemicals, peptides, and recombinant proteins
Ni-NTA Superflow Resin	QIAGEN	Cat# 30410
Anti-FLAG M2 Affinity Gel	Sigma-Aldrich	Cat# A2220-2
Glutathione Sepharose 4 Fast Flow GE Healthcare	Thermo Fisher Scientific	Cat# 45-000-286
G418 disulfate	Sigma-Aldrich	Cat# G8168
Igepal CA-630	Sigma Aldrich	Cat# I3021
Benzamidine	Thermo Fisher Scientific	Cat# 50-488-531
Aprotinin	Gold Biotechnology	Cat# A-655-100
Chymostatin	Sigma-Aldrich	Cat# EI6
Leupeptin	Gold Biotechnology	Cat# L-010-100
Pepstatin A	Gold Biotechnology	Cat# P-020-100
Flag peptide	Alan Scientific	synthesized DYKDDDDK
Spermidine base	Research Products International	Cat# S92150
Proteinase K	Thermo Fisher Scientific	Cat# NC0547027
Dynabeads M270 Streptavadin Invitrogen	Thermo Fisher Scientific	Cat# 65-305
Benzonase (TurboNuclease)	Accelagen	Cat# N0103M
NP40	Thermo Fisher Scientific	Cat# 85124
Protease inhibitor cocktail	Thermo Fisher Scientific	Cat# NC1866304
Camptothecin	Thermo Fisher Scientific	Cat# 50-165-7222
Mitomycin C	Sigma-Aldrich	Cat# M4287
Olaparib	Selleck Chemicals	Cat# S1060
Rucaparib	Selleck Chemicals	Cat# S1098
DAPI Mounting Medium	Thermo Fisher Scientific	Cat# NC1848444
CldU	Sigma-Aldrich	Cat# C6891
IdU	Sigma-Aldrich	Cat# I7125
Mirin	Sigma-Aldrich	Cat# M9948
SF900 SFM Media Gibco	Thermo Fisher Scientific	Cat# 10902088
ESF 921 Media	Expression Systems	Cat# 96-001-01
RPMI 1640 media	Gibco	Cat# 11875-093
Lipofectamine 2000	Invitrogen	Cat# 11-668-019
Experimental models: Cell lines
DLD1 BRCA2−/−	Horizon Discovery	HD 105-007
Hi5 insect cells (Tni cells)	Expression Systems	Cat# 94-002
SF9 cells	Thermo Fisher Scientific	Cat# 11496015
Oligonucleotides
Refer to Table S3	IDT	N/A
Recombinant DNA
pCMV-2xMBP-BRCA2-His6-Flag	This paper	N/A
pCMV-2xMBP-BRCA2OB− DBD− 9A−His6−Flag	This paper	N/A
pCMV-2xMBP-BRCA2CTRB− dbd− 4A-His6-Flag	This paper	N/A
pCMV-2xMBP-BRCA2OB− DBD− 9A+CTRB-DBD-4A-His6-Flag	This paper	N/A
MacroBac p438-A His6 -BRC4-OB-DBD-Flag	This paper	N/A
MacroBac p438-A His6 -BRC4-OB-DBDoB2− 5A-Flag	This paper	N/A
MacroBac p438-A His6 -BRC4-OB-DBDoB3− 4A-Flag	This paper	N/A
MacroBac p438-A His6 -BRC4-OB-DBD9A-Flag	This paper	N/A
MacroBac p438-A His6 -BRC4-OB-DBD-CTRB-Flag	This paper	N/A
MacroBac p438-A His6 -BRC4-OB-DBD°B2− 5A-CTRB-Flag	This paper	N/A
MacroBac p438-A His6 -BRC4-OB-DBD°B3− 4A-CTRB-Flag	This paper	N/A
MacroBac p438-A His6 -BRC4-OB-DBD9A-CTRB-Flag	This paper	N/A
MacroBac p438-A His6 -BRC4-OB-DBD-CTRB4A-Flag	This paper	N/A
MacroBac p438-A His6 -BRC4-OB-DBD9A-CTRB4A-Flag	This paper	N/A
pDEST20-DSS1	Zhao et al.11	N/A
pET-His-Smt3-RAD51	Granéli et al.40	N/A
pET11c-RPA	Henricksen et al.41	N/A
pET32-BRC4	This paper	N/A
MacroBac p438-A Flag NBS1	This paper	N/A
MacroBac strep-MRE11 + RAD50-TEV-MBP + Flag-NBS1	This paper	N/A
sgRNA plasmid px330-LMNA	Pinder et al.30	N/A
pCR2.1-CloverLamin	Pinder et al.30	N/A
Software and algorithms
Prism	GraphPad	https://www.graphpad.com/
Image Lab	Bio-Rad	Cat# 12012931
Microsoft Excel	Microsoft	N/A
ZEN Blue	Carl Zeiss Microscopy	N/A
ImageJ	ImageJ Software	https://imagej.nih.gov/ij/
BioRender	BioRender	https://www.biorender.com