BRIP1, a G-quadruplex DNA helicase, protects against AID-induced genomic instability by limiting R-loop accumulation

Abstract

Background

BRIP1 (BRCA1-interacting protein 1) or FANCJ is a 5′–3′ DNA helicase that plays critical roles in the maintenance of genome stability through its involvement in DNA repair, replication stress responses, and the resolution of DNA secondary structures or non-B DNAs. Defects in the BRIP1 helicase gene are associated with an increased risk of various cancers. It has been shown that BRIP1-deficient mice are predisposed to B-cell lymphoma. Because activation-induced cytidine deaminase (AID)-dependent genomic instability resulting from its mistargeting to non-immunoglobulin genes is a major hallmark of B-cell lymphoma, and non-B DNA structures, including G-quadruplexes (G4) and RNA–DNA hybrids (R-loops), are implicated in AID-induced genomic instability, we asked whether BRIP1 helicase limits these secondary structures to suppress AID-mediated oncogenesis.

Methods

We knocked down (KD) BRIP1 in the CH12F3-2A mouse B-cell, a model cell line for studying AID-induced physiological and non-physiological events and their underlying mechanisms. We also used biophysical assays and in silico tools to study non-B DNAs at human AID off-target sites.

Results

Here, we report that BRIP1 KD increased the frequency of AID-induced events, including IgM-to-IgA antibody switching, somatic hypermutation (SHM), DNA double-strand breaks (DSBs), and IgH/c-Myc chromosomal translocations. However, BRIP1 KD does not affect either the end-joining of AID-induced DNA breaks or the synapsis of switch regions. Importantly, BRIP1 KD resulted in the accumulation of R-loops at both AID targets and off-targets. Furthermore, in silico analysis revealed G4s as the most prevalent form of non-B DNA conformation at AID off-target sites in human. We propose that BRIP1 suppresses AID-induced genomic instability by resolving non-B DNA, thereby potentially limiting the accumulation of genomic lesions that contribute to B-cell lymphoma.

Conclusion

These findings revealed that BRIP1 prevents the excessive accumulation of non-B DNAs at AID target sites and suppresses AID-mediated genomic instability. These data also provide a possible explanation for why BRIP1-deficient mice are predisposed to B-cell lymphoma.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s10020-026-01627-w.

Background

Preserving genomic integrity is essential for normal cellular function and survival; thus, cells have evolved intricate DNA damage surveillance and repair pathways that continuously detect and resolve DNA lesions to maintain genome stability. Interestingly, the sites of genomic instability are not randomly distributed throughout the genome but are frequently concentrated within sequences capable of adopting non-B DNA conformations (Wang and Vasquez 2006). The accumulation of single-stranded DNA (ssDNA) and negative supercoils during transcription and replication favors the formation of non-B DNA structures, which include G-quadruplexes (G4s), RNA/DNA hybrid (R-loop), Z-DNA, hairpins, cruciforms, and H-DNA (Zhao et al. 2010). Among these, G4 arises when four guanine bases interact to form a planar structure, G-quartet (or G-tetrad), which subsequently rearranges into a quadruplex configuration through the stacking of two or more such tetrads (Spiegel et al. 2020; Sen and Gilbert 1988). Another prevalent non-B DNA structure, the R-loop, forms when RNA molecules anneal to their DNA template, generating a distinctive three-stranded structure. R-loops display a remarkable propensity to form at highly transcribed GC-rich sequences (García-Muse and Aguilera 2019). These non-canonical DNA structures are inherently susceptible to aberrant processing, thereby promoting genomic instability and contributing to the development of a wide range of human diseases.

Non-B DNA structures also serve important physiological functions in B-lymphocytes. Specifically, non-B-DNA conformations formed at the immunoglobulin (Ig) switch (S) region, which are highly enriched in G-rich repetitive sequences, provide a favorable substrate for the recruitment and activity of activation-induced cytidine deaminase (AID), the enzyme responsible for initiating secondary antibody diversification. AID expression in antigen-activated B-cells is necessary for Ig gene diversity via somatic hypermutation (SHM) and class switch recombination (CSR) (Muramatsu et al. 2000; Revy et al. 2000; Honjo et al. 2005). The introduction of random mutations into the recombined Ig variable (V) region through SHM is essential for generating high-affinity antigen-binding sites in the encoded antibodies (Chi et al. 2020). In contrast, CSR, which involves DNA cleavage at the switch (S) region, alters the Ig heavy-chain (IgH) constant region, thereby altering antibody effector functions and allowing it to interact with different effector molecules while retaining specificity for the same antigens (Stavnezer and Schrader 2014). AID normally targets the V and S regions of the Ig gene; however, its mistargeting and aberrant expression are linked with oncogenic mutations and chromosomal translocations in various cancers (Honjo et al. 2012).

Transcription at physiological AID target sites (S-regions), as well as frequent off-targets (such as c-MYC) generates negative supercoils and ssDNA. These conditions, particularly on repeat-containing DNA sequences, favor the formation of aberrant or non-B DNA structures, including G4s and R-loops (Duquette et al. 2005; Zhang et al. 2014a; Yu et al. 2003; Qiao et al. 2017; Feng et al. 2020). It has also been shown that the co-transcriptional formation of G4 DNA on the non-template ssDNA stabilizes R-loops on the template strand (Lee et al. 2020; Wulfridge et al. 2023). It was previously shown that AID preferentially binds and deaminates G4-containing DNA substrates, and that G4 binding promotes AID oligomerization in vitro. This oligomerization may enhance AID activity at G4-associated sites in vivo, leading to higher mutation density and breaks (Qiao et al. 2017). In addition to c-MYC, various other AID off-targets, in both humans and mice, are GC-rich and have the potential to form G4s, promoting genomic instability associated with B-cell malignancies (Pasqualucci et al. 2001; Lu et al. 2013; Meng et al. 2014; Chiarle et al. 2011; Kato et al. 2012). G4s have also recently been reported to be involved in repairing AID-induced DNA breaks, suggesting a multifaceted role of G4s in AID-induced genomic instability (Refaat et al. 2023). Given the dual role of non-B DNA structures in supporting physiological processes while simultaneously threatening genomic integrity, their formation and persistence must be tightly regulated by genome maintenance factors. DNA topoisomerase 1 (TOP1) is one such factor that eliminates transcription-associated helical stress, preventing the formation of non-B DNA. Consistently, the knockdown (KD) of TOP1 in B-cell line enhances the formation of negative supercoils and facilitates DNA cleavage in the V and S regions, resulting in increased frequencies of SHM, CSR, and IgH/c-Myc chromosomal translocations (Kobayashi et al. 2009, 2011; Husain et al. 2016).

Helicases are another class of genome-maintenance proteins that unwind DNA or RNA by disrupting hydrogen bonds between complementary base pairs. Beyond their essential roles in replication, transcription, and DNA repair, they resolve non-B DNA structures that arise during these processes, preventing genomic instability (Liu et al. 2021). Notably, DDX1, an RNA helicase, facilitates S-region R-loop formation and promotes CSR by resolving G4s in intronic switch transcripts (Ribeiro de Almeida et al. 2018). BLM and WRN DNA helicases contribute to DNA end-resection at AID-induced breaks in the switch regions during CSR. Inhibiting BLM or WRN partially rescues CSR defects in 53BP1- or H2AX-deficient cells by reducing end-resection (Keijzers et al. 2014; Babbe et al. 2009). Given the emerging role of G4-resolving helicases in genome maintenance, BRIP1, an ATP-dependent helicase, is of particular interest because, in addition to its established role in DNA crosslink repair, it can resolve G4 DNA structures, thereby preventing DNA damage (Cantor et al. 2004; London et al. 2008; Youds et al. 2008; Sarkies et al. 2012; Castillo Bosch et al. 2014). Furthermore, germline mutations in BRIP1 are associated with increased susceptibility to several cancers, including ovarian (Castillo Bosch et al. 2014) and breast cancers (Rafnar et al. 2011), while BRIP1-deficient mice are predisposed to epithelial cancers and lymphoma (Matsuzaki et al. 2015).

Here, we asked whether BRIP1 suppresses oncogenic AID-induced genomic instability in B-cells by preventing excessive accumulation of non-B DNA at AID target and off-target sites. We revealed that KD of the BRIP1 helicase led to a significant increase in AID-induced mutations and DNA breaks, resulting in elevated levels of both CSR and IgH/c-Myc chromosomal translocation. These effects were associated with the accumulation of non-B DNA, leading to increased DNA break formation without affecting end joining or S-region synapsis. Additionally, in silico analysis revealed that, among various non-B DNA structures, G4s were preferentially enriched in human AID off-target regions. Together, the data point to a role of BRIP1 helicase as a suppressor of AID-induced genomic instability and suggest that it may help prevent B-cell lymphoma.

Material and methods

Materials

The plasmid expressing the human BRIP1 protein, pcDNA3-myc-his-BRIP1, was purchased from Addgene (#17642) and cloned into the pEF1a-3xFLAG-IRES-ZsGreen vector. Endotoxin-free plasmid DNA was purified using a plasmid purification kit (NAP Life Science). The source and specifications of other materials used in this study are either described in the methods or given in Supplementary Tables S1-S3.

Analysis of class switch recombination

The mouse B-cell lymphoma line CH12F3-2A, expressing Bcl2, was cultured and maintained in RPMI 1640 supplemented with glutamine, 5% NCTC-109, fetal bovine serum (10%), β-mercaptoethanol, and penicillin/streptomycin, as previously described (Stanlie et al. 2010; Doi et al. 2003). Chemically modified Stealth siRNA oligonucleotides (Supplementary Table S1) were introduced into the cells by electroporation to KD the expression of DNA helicase genes. Following electroporation, the cells were cultured for 24 h, stimulated with CIT (CD40L, IL4, and TGFβ) (Matsuzaki et al. 2015) to induce IgM-to-IgA isotype switching, and cultured for an additional 24–48 h before collection and analysis. The surface expression of IgM and IgA was analyzed by staining cells with FITC-conjugated anti-mouse IgM and PE-conjugated anti-mouse IgA antibodies (Supplementary Table S1), respectively, as previously described (Nakamura et al. 1996). For the rescue experiment, CH12F3-2A cells were co-transfected with either empty vector (pEF1a-3xFLAG-IRES-ZsGreen) alongside siCONT or siBRIP1, or 2 μg of human BRIP1 expression construct (pEF1a-BRIP1-3xFLAG-IRES-ZsGreen) alongside siBRIP1.The cells were cultured for 24 h, stimulated with CIT, and cultured for an additional 24 h. The whole cell lysates were subjected to western blotting with anti-FLAG antibody (Sigma). Flow cytometry analysis was performed using a BD FACS Calibur instrument, and the data were analyzed using CellQuest software (BD Biosciences) and the Floreada.io tool.

Expression of Aid and germline transcripts

RNA from control and BRIP1 siRNA-transfected CH12F3-2A cells was isolated using TRIzol (Ambion), followed by DNase I treatment. One microgram of purified RNA was used as a template for cDNA synthesis with SuperScript IV and an oligo (dT) primer. Real-time PCR (Applied Biosystems, 7900HT Fast Real-Time PCR System) was performed using PowerUp SYBR Green Master Mix (Thermo Fisher Scientific) and gene-specific RT-qPCR primers, as previously described (Stanlie et al. 2010, 2012). Primer pairs used for transcript analysis are described in Supplementary Table S2. The 2^−∆∆Ct method was used to quantify the data, and individual genes were normalized to Hprt expression.

AID-induced IgH/c-Myc translocations

IgH/c-Myc chromosomal translocations (derivative chromosome 15) were detected by nested PCR using genomic DNA isolated from CIT-stimulated CH12F3-2A cells transfected with BRIP1 or control siRNAs. A total of 24 to 48 genomic DNA aliquots were PCR-amplified in separate reactions using the Expand Long Template PCR System (Sigma), as previously described (Boboila et al. 2012a; Ramiro et al. 2004). The conditions for both the first and second rounds of PCR were as follows: 94 ℃ for 3 min, followed by 25 cycles at 94 ℃ for 15 s, 62 ℃ for 15 s, 68 ℃ for 7.5 min, and a final extension of 5 min at 68 ℃. The PCR products were electrophoresed on 1% agarose gels containing ethidium bromide and subjected to Southern blotting with a Myc-specific probe. The primer and probe sequences are shown in Supplementary Table S2.

Analysis of AID-induced somatic hypermutations

To analyze SHM in the 5’ Sµ region, genomic DNA was isolated from CH12F3-2A cells transfected with control or BRIP1 siRNA and stimulated with CIT for 48 h. A 550-base pair region from the 5’ Sµ region was PCR amplified using homemade Pfu-sso7d high-fidelity polymerase (Farooqui et al. 2023; Mahboob et al. 2024). The purified PCR fragments were A-tailed using Taq DNA polymerase, and TA-cloned into the pGEM-T Easy vector, as previously described (Nagaoka et al. 2002; Denépoux et al. 1997). Plasmids from each of the BRIP1 or control siRNAs plates were sequenced for Sµ-SHM analysis using T7 forward and SP6 reverse primers. Clones with unique modifications were counted, and the mutation frequency was calculated as the number of mutations per total bases sequenced. The primers used for 5’ Sµ region PCR and sequencing are shown in Supplementary Table S2.

Analysis of the Sµ-Sα repair junctions

Genomic DNA was prepared from the siCONT- or siBRIP1-transfected CH12F3-2A cells stimulated with CIT for 48 h. To analyze DNA repair and end-joining, the recombined Sµ-Sα junctions were amplified by nested PCR using the forward and reverse primers within the 5’Sµ and 3’Sα regions. PCR reactions were performed using homemade Pfu-Sso7d DNA Polymerase (Farooqui et al. 2023; Mahboob et al. 2024). For the first PCR round, the conditions were 98 °C for 2 min, followed by 22 cycles of 98 °C for 10 s, 55 °C for 10 s, and 72 °C for 2 min. The PCR products from the first round were purified and subjected to a second round of PCR under the same conditions, except that the elongation time at 72 °C was 1 min 20 s. The PCR products, ranging from ~300 bp to 1.2 kb, were gel-extracted, A-tailed by Taq DNA polymerase, and cloned into the pGEM-T Easy vector (Supplementary Table S1). Plasmids from the BRIP1 or control siRNAs plates were individually sequenced using T7 forward and SP6 reverse primers. Sequence recombination was analyzed by aligning the S-region sequences with the reference sequences (Sμ: AH005309.2 and Sα: D11468.1) using pairwise nBLAST at NCBI, as previously described (Stavnezer et al. 2010; Gothwal et al. 2024). The primers used for junction amplification and sequencing are shown in Supplementary Table S2.

Chromatin immunoprecipitation assay

DNA break formation was analyzed by histone γH2AX chromatin immunoprecipitation (ChIP) assay, as previously described (Stanlie et al. 2010). Briefly, CIT-stimulated control or BRIP1 KD CH12F3-2A cells (5 x 10⁶) were fixed using 1% formaldehyde for 10 min, and the crosslinking reaction was quenched by adding glycine to a final concentration of 0.125 M. The chromatin fraction from the formaldehyde-fixed cells was extracted as previously described, with minor modifications (Husain et al. 2020). Chromatin was sheared to obtain a soluble chromatin fraction containing fragmented DNA of 200–2,000 bp by sonication (Bioruptor, Diagenode), followed by centrifugation to clarify the chromatin lysate. The clarified chromatin lysate was immunoprecipitated using 3 µg of γH2AX antibody (Supplementary Table S1) and protein A/G-coupled magnetic beads at 4 °C with rotation. The bead-bound chromatin was sequentially washed with a standard series of wash buffers, and the bound chromatin was eluted, proteinase K-digested, reverse crosslinked, and purified using a DNA purification kit. Locus-specific enrichment was detected by real-time PCR and normalized to the input DNA. The primer sequences used for ChIP are shown in Supplementary Table S2.

Chromosomal conformation capture assay

The chromosomal conformation capture (3C) assay was performed as described previously (Wuerffel et al. 2007; Stanlie et al. 2014). A total of 7 × 10⁶ cells was harvested, washed, and cross-linked with 1% formaldehyde for 5 min at room temperature. After cross-linking, nuclear lysis was performed, and the chromatin was digested with Hind III overnight. The digested chromatin was then ligated using T4 DNA ligase (Takara). Following ligation, the samples were treated with proteinase K and reverse-cross-linked, and the DNA was subsequently purified by phenol/chloroform extraction. PCR amplification was done using the primers listed in Supplementary Table S2.

DNA-RNA Immunoprecipitation

The DNA-RNA immunoprecipitation (DRIP) assay for the IgH and c-Myc locus was adapted from previously established protocols (Refaat et al. 2023; Zhang et al. 2014b, 2015). Genomic DNA from CH12F3-2A cells transfected with control or BRIP1 siRNAs was digested at 37 °C using a cocktail of restriction enzymes (HindIII, EcoRI, XbaI, SspI, BsrGI) in digestion buffer (10 mM Tris-HCl, pH 8; 10 mM MgCl2; 50 mM NaCl; 1 mM DTT), followed by RNase A treatment and genomic DNA purification through phenol-chloroform extraction. Digested DNA (6 µg) was either untreated or treated with RNase H overnight at 37 °C. The RNase H-treated and untreated genomic DNA samples were incubated overnight with the 3 μg of S9.6 antibody in DRIP buffer (10 mM NaPO4, pH 7.0; 140 mM NaCl; 0.1% Triton X-100) at 4 °C, followed by further incubation for 2 h with 30 µL of Protein G magnetic beads at 4 °C. The antibody-DNA complexes bound to Protein G beads were washed three times with DRIP buffer and once with TE buffer (20 mM Tris-HCl, pH 8; 1 mM EDTA), then eluted twice with 60 µL elution buffer (100 mM NaHCO3; 1% SDS), followed by Proteinase K digestion and DNA purification. The DRIP-qPCR % IP/input ratio was used to calculate the enrichment. The primers used in PCR are given in Supplementary Table S2.

Fluorescence spectrometry

The G4 RNA screener predicted G4 forming oligonucleotides from the Sµ and c-Myc regions were purchased from Integrated DNA Technologies (IDT) and prepared at concentrations of 0.5–1 µM in folding buffer (10 mM Tris-HCl, pH 7.4; 100 mM KCl), heated at 95 ℃ for 10 min, followed by 2–3 h incubation at room temperature, then overnight incubation at 4 ℃. Folded G4 or control oligos were incubated with either 1.4 µM N-methyl mesoporphyrin IX (NMM) for 10 min or 0.5 µM of Thioflavin T (ThT) for 30 min, followed by fluorescence measurements at Shimadzu RF-5301PC spectrofluorometer (Kyoto, Japan). The excitation and emission wavelengths for NMM were 393 nm and 610 nm, respectively, whereas those for ThT were 425 nm and 490 nm, respectively, as previously described (Doi et al. 2003; Arthanari et al. 1998). Fluorescence data were normalized using tubes folded in the presence of LiCl, a G4 destabilizing agent, and C-rich oligonucleotides that serve as a negative control. The list of G4-forming oligonucleotides and C-rich oligonucleotides is given in Supplementary Tables S2 and S3.

Native-polyacrylamide gel electrophoresis

For the analysis of control or G4-forming oligonucleotides on native-polyacrylamide gel electrophoresis (native-PAGE), 15% (w/v) polyacrylamide (acrylamide: bisacrylamide, 29:1) gel was prepared in 0.5xTBE (0.89 M Tris-borate, 1 mM EDTA, pH 8.3) containing 50 mM KCl, as described previously (Sharma et al. 2022; Zaug et al. 2005). Oligonucleotides were prepared at 35 µM, using the same procedure as for fluorescence spectrometry, in a total volume of 15 µL. 5 µL of 4X sample loading buffer containing 50 mM KCl, 60% glycerol, and 0.15 mg/mL Orange G, prepared in 1xTBE, was added to each sample, and the samples were loaded onto the gel. The running buffer consisted of 0.5xTBE with 50 mM KCl, and the electrophoresis was performed at a constant voltage of 100 V at room temperature with a cooling pack to prevent overheating. The gel was stained using a silver staining protocol to visualize G4 formation, as previously described (Bassam and Gresshoff 2007).

In silico G4 analysis

A literature survey was conducted to compile AID-targeted human non-Ig loci along with their genomic coordinates. Using the University of California, Santa Cruz (UCSC) genome browser, the genomic sequences of these AID non-Ig regions for the human GRCh37.p10 assembly were retrieved. The Eukaryotic Promoter Database (https://epd.expasy.org/epd) was used to identify the promoter regions and transcription start sites of the respective genes. For genes whose AID targeting genomic coordinates were not available in the literature, we used a window of −1 kb to +4kb relative to the transcription start site. To identify potential G4-forming sequences within the selected genomic coordinates, the NBBC (Non-B DNA Burden explorer in Cancer) https://kowalski-labapps.dellmed.utexas.edu/NBBC/, a web based non-B DNA visualisation platform (https://kowalski-labapps.dellmed.utexas.edu/NBBC/) was queried using the genomic coordinates for each locus (Xu and Kowalski 2023). Each gene length query was normalized by default non-B burden unit: counts per kilobase per million, or CPKM. Normalization enables comparison of non-B loads across different non-B categories and gene locations.

Results

BRIP1 deficiency enhances Ig class-switch recombination

Given that AID-induced genomic instability is a key feature of B-cell lymphomas, and that BRIP1 deficiency predisposes mice to B-cell lymphoma, we asked whether BRIP1 prevents AID-induced genomic instability by suppressing the accumulation of non-B DNA structures at both on- and off-target loci. To investigate this, we first studied the effect of BRIP1 deficiency on AID-induced physiological recombination, namely CSR. We used either control siRNA (siCONT) or three independent siRNAs targeting BRIP1 to KD BRIP1 expression in CH12F3-2A cells, a mouse B-cell line that undergoes IgM-to-IgA switching upon stimulation with CIT. We examined the frequency of CSR at 24 and 48 h after CIT stimulation (Fig. 1A). As shown by the flow cytometry profiles, BRIP1 KD increased CSR frequencies at both 24 and 48 h, with siRNA-41 showing the largest increase (Fig. 1B and C). To validate the efficiency of BRIP1 KD in CH12F3-2A cells, we performed an RT-qPCR assay, which revealed that siRNA-41 caused the most robust BRIP1 KD, followed by siRNA-43 and siRNA-42. We also confirmed that BRIP1 KD did not alter the expression of Aid or switch germline transcripts (µGLT and αGLT), the essential prerequisites for CSR initiation at the IgH locus (Fig. 1D), indicating that BRIP1 is directly involved in CSR. Since CSR frequency can be indirectly influenced by cell proliferation and death, we also assessed the effect of BRIP1 KD on cell proliferation and cell death of CH12F3-2A cells by analyzing cell density and propidium iodide (PI) exclusion, respectively. While none of the siRNAs caused significant changes in PI exclusion, siRNA-43 strongly reduced cell proliferation (Fig. 1E and Supplementary Fig. 1 A). This aligned well with the poor enhancement of CSR by siRNA-43 despite good KD efficiency. On the other hand, siRNA-41 resulted in the most substantial KD of BRIP1 without adversely affecting cell proliferation or cell death, making it the preferred choice for subsequent analysis (Fig. 1E).

Fig. 1.

BRIP1 loss increases the efficiency of Ig class-switch recombination. A Scheme of the IgA class switching assay in CH12F3-2A cells. After electroporation with siRNAs, cells were cultured for 24 h, stimulated with CIT for an additional 24 h and 48 h, and analyzed by flow cytometry. B Representative profiles of CH12F3-2A cells undergoing IgA switching after transfection of the indicated siRNAs. Cells were cultured for 24 h and 48 h in the absence (-) or presence (+) of CIT. The dot plot shows the percentage of cells expressing IgA on their surfaces. C Summary of flow cytometry data obtained from two independent experiments at 24 h and 48 h, respectively. Data show % IgA switching in siBRIP1 relative to control siRNA (n=2; mean +SD; two-tailed unpaired Student’s t-test). D Quantitative RT-PCR analysis of the Aid, Brip1, αGLT, and µGLT expression. Data were normalized to Hprt (n=2; mean +SD; two-tailed unpaired Student’s t-test). E Cell proliferation of CH12F3-2A cells as revealed by cell density analysis. F Top: Schematic representation of IgA class switching of CH12F3-2A cells co-transfected with either siCONT and empty vector (EV) or siBRIP1 together with either EV or the human BRIP1 expression vector (pEF1a-BRIP1-3xFLAG-IRES-zsGreen). After electroporation, cells were cultured for 24 h, stimulated with CIT for an additional 24 h, and analyzed by flow cytometry. Bottom: Representative profiles of IgA switching of indicated cells. The numbers in the density plot represent the percentages of the GFP-gated cells expressing IgA on their surface. G Summary of the flow cytometry data from the rescue experiment at 24 h. Data show % IgA switching in the indicated samples (n = 2; mean + SD; two-tailed unpaired Student’s t-test). H Western blot analysis of BRIP1 expression 24 h after CIT stimulation. I Summary of the effect of Brip1 or Top1 KD alone, or their combination in CH12F3-2A cells on CSR (n=2; mean +SD; two-tailed unpaired Student’s t-test). CSR, class switch recombination; ns, not significant; CIT, CD40L-IL4-TGFb, *p < 0.05; **p < 0.01; ***p < 0.001; ns, not significant

To further validate the specific role of BRIP1 in suppressing CSR and rule out off-target effects of siRNAs, we performed a rescue experiment by transient ectopic expression of human BRIP1 in siBRIP1 transfected CH12F3-2A cells (Fig. 1F). While BRIP1 KD resulted in elevated IgM-to-IgA switching relative to siCONT, a decreased percentage of IgA⁺ cells was observed upon ectopic expression of human BRIP1 when compared to those transfected with siBRIP1 alone, indicating that BRIP1 indeed suppresses CSR, and that the enhancement in CSR efficiency by BRIP1 siRNA was solely due to BRIP1 deficiency (Fig. 1G and H).

It has been shown that the KD of TOP1 in the CH12F3-2A cell line increases negative supercoiling and CSR frequency (Kobayashi et al. 2009; Husain et al. 2016). It is also well established that the accumulation of negative supercoils promotes aberrant DNA structures, such as G4s and R-loops (Hegazy et al. 2020). We therefore speculated that BRIP1 depletion might lead to further accumulation of non-B DNA structures in TOP1-depleted cells. To test this, we examined whether the combined KD of TOP1 and BRIP1 in CH12F3-2A cells would exert an additive effect on CSR frequency. Indeed, the combined KD of BRIP1 and TOP1 in CH12F3-2A cells led to a greater increase in CSR frequency than the individual KD of either BRIP1 or TOP1 (Fig. 1I), suggesting that BRIP1 and TOP1 suppress CSR through independent pathways. Next, we asked whether, like BRIP1, other G4-unwinding DNA helicases suppress CSR in CH12F3-2A cells. We specifically selected PIF1 (Paeschke et al. 2013), WRN (Mohaghegh et al. 2001), RTEL1 (Wu et al. 2020), RECQL5 (Budhathoki et al. 2016), DHX9 (Tang et al. 2022), and DHX36 (Chen et al. 2018), based on their established roles in unwinding G4 structures, while excluding BLM, another known DNA helicase whose role in CSR has already been described (Babbe et al. 2009). Individual KD of these candidate helicases in CH12F3-2A cells was performed, and their impact on CSR frequency was evaluated. As revealed by the lack of correlation between KD efficiency and changes in CSR, KD of PIF1, RTEL1, WRN, RECQL5, and DHX9 did not significantly affect the frequency of IgM-to-IgA switching (Supplementary Fig. 2B and 2 C). However, we observed a modest increase in CSR with one of the siDHX36 oligos, but the effects were likely independent of its role in G4 resolution, as DHX36 does not bind to the Sμ region in a G4-dependent manner (Refaat et al. 2023). Hence, we concluded that an increase in CSR upon BRIP1 KD is not a general consequence of helicase deficiency, but rather a phenotype specific to BRIP1.

BRIP1 does not affect end-joining and S-region synapsis

AID deaminates cytosine residues, triggering downstream repair pathways that resolve these lesions into chromosomal breaks, which are essential for CSR. Most AID-induced DSBs have staggered ends that require extensive end-processing for their repair through classical non-homologous end-joining (c-NHEJ) or alternative end-joining (alt-EJ) pathways (Boboila et al. 2012b; Chang et al. 2017). While c-NHEJ does not require homology between DNA ends, alt-EJ utilizes microhomology (MH) to join the cleaved DNA. There is growing interest in understanding whether non-B DNA structures, such as G4s and R-loops at DSBs, affect the function of NHEJ pathway proteins. With respect to CSR, studies have shown that the RNA-binding protein HNRNPU forms a ribonucleoprotein complex with c-NHEJ repair factors at S-region G4s (Refaat et al. 2023), suppressing the excessive accumulation of R-loops and ssDNA, thereby preventing incorrect DNA repair and persistent DNA breaks. Given that BRIP1 maintains genomic stability by resolving non-B DNA structures and facilitating DNA damage repair (Panday et al. 2021; Brosh and Matson 2020), we next investigated whether BRIP1 influences DNA repair and the choice of DSB repair pathways during CSR. We analyzed the characteristics of the recombined Sµ and Sα CSR junction in CH12F3-2A cells (Fig. 2A). Having first confirmed BRIP1 KD efficiency (Fig. 2B), we amplified Sµ-Sα junctions through nested PCR of genomic DNA from CH12F3-2A cells transfected with either control or BRIP1 siRNA, followed by sequencing and analysis of the junctional features (Fig. 2C). Junction analysis revealed that BRIP1 KD did not significantly affect the proportion of junctions repaired by c-NHEJ, as junctions with 0–3 nucleotides (ntds) of microhomology, which accounted for nearly two-thirds of all junctions, remained largely unchanged between control and BRIP1 KD cells. However, the proportion of junctions containing ≥4 ntds of microhomology increased modestly from 6.1% to 11.5%, while those with insertions decreased from 21.2 % to 11.5 % (Fig. 2D). Since CSR predominantly relies on c-NHEJ, these modest alterations in junction characteristics are unlikely to explain the enhanced CSR observed upon BRIP1 KD. Overall, these data suggest that BRIP1 KD does not substantially alter the DNA end joining or the choice of DSB repair pathways during CSR.

Fig. 2.

BRIP1 loss does not affect end-joining or S-region synapsis. A Schematic representation of the Sµ-Sα junction analysis. B Quantitative RT-PCR analysis of the BRIP1 expression in the indicated sample. The data was normalized to Hprt (n=2; mean +SD, two-tailed unpaired Student's t-test, **p < 0.01). C Agarose gel electrophoresis of PCR amplified Sµ-Sα junctions from genomic DNA obtained from CH12F3-2A cells transfected with siCONT or siBRIP1. D Pie charts showing the distribution of Sµ-Sα junction recombination derived from control and BRIP1 KD cells. Statistical significance was evaluated by Fisher’s exact test (0–3 nt vs ≥4 nt, p = 0.655; 0–3 nt vs insertions, p = 0.488; ≥4 nt vs insertions, p = 0.328). E Left: Scheme of long-range interactions among Eµ-Eα, Eµ-Sα, and Sµ-Sα elements of the IgH locus before and after CIT stimulation in CH12F3-2A cells. Right: Representative gel images of 3 C assay using siCONT-transfected CH12F3-2A cells with or without CIT stimulation, and siBRIP1-transfected CH12F3-2A cells with CIT stimulation for 24 h. Each PCR panel illustrates paired long-range interactions at the IgH locus, as indicated on the right. Gapdh was amplified as a loading control. CIT, CD40L-IL4-TGFb

In addition to DSBs, CSR also relies on three-dimensional chromatin architecture mediated by intra-chromosomal interactions between distantly located transcriptional regulatory elements: the Eµ intronic enhancer situated at the 5’ end, and the Eα at the 3’ regulatory region, with S regions looped between these enhancer regions. Therefore, we next used a chromosomal conformation capture (3C) assay in CH12F3-2A cells transfected with control or BRIP1 siRNA to examine whether BRIP1 influences the establishment of synapsis between the Sµ and downstream Sα region. While the constitutive interaction between the Eµ and Eα regions of the IgH locus remained unchanged in both unstimulated and CIT-stimulated cells, the interactions between Sµ and Sα, as well as Eµ and Sα, significantly increased upon CIT stimulation. However, no significant differences were observed in Sµ–Sα and Eµ–Sα interactions between CIT-stimulated CH12F3-2A cells transfected with control or BRIP1 siRNA (Fig. 2E). Taken together, these findings indicate that the increase in CSR resulting from BRIP1 KD was not attributed to either the altered end-joining of DSBs in Sµ and Sα or the defects in the synapsis of the Sµ and Sα regions.

BRIP1 suppresses SHM, DNA breaks, and IgH/c-Myc translocations

Since the synapsis and end joining of AID-induced DNA breaks remained unperturbed upon BRIP1 KD, we next examined the frequency of AID-induced mutations, breaks, and chromosomal translocations. We first examined the effect of BRIP1 KD on mutation in the upstream 5' Sμ region, an efficient target for SHM, in CH12F3-2A cells (Fig. 3A). We PCR-amplified a 550 bp DNA segment from the 5' Sμ region using genomic DNA from control or BRIP1-KD cells, followed by cloning and sequencing to detect mutations (Fig. 3B and C). SHM analysis revealed that similar to CSR, the BRIP1 KD in CH12F3-2A cells resulted in 1.5-fold increase in mutation frequency at the 5' Sμ region when compared with cells transfected with control siRNA (Fig. 3D). However, the mutation base profile remained unchanged, suggesting that the type of mutations induced by AID was not affected by BRIP1 KD, which is also in line with unperturbed repair pathways choice during DSB end-joining (Fig. 3E). Since changes in SHM also reflect changes in the DSBs, therefore, we assessed whether increased mutations due to BRIP1 KD were accompanied by the changes in DSB levels at 5' Sμ. Sα, and c-Myc regions, through ChIP assay for Ser139 phosphorylated histone H2AX (γH2AX), a marker for DSBs and DNA damage response. We found increased levels of γH2AX at 5' Sμ, Sα, and c-Myc regions in BRIP1 KD cells, indicating an increased frequency of AID-induced DNA breaks (Fig. 3F and G).

Fig. 3.

BRIP1 loss elevates SHM, DNA breaks, and IgH/c-Myc chromosomal translocations. A Top: Schematic illustration of the experimental design for the SHM assay in CH12F3-2A cells. Bottom: Schematic representation of the 5’ Sμ region of CH12F3-2A cells analyzed for SHM. B Agarose gel electrophoresis of PCR amplified 5’ Sμ region of CH12F3-2A cells transfected with siCONT or siBRIP1. C Details of the mutation analysis of the 5’ Sμ region. Ins, insertion; Del, deletion. D Mutation frequency in the 5’ Sμ region of BRIP1 KD CH12F3-2A cells. Statistical significance, as evaluated by Fisher’s exact test, is shown. E Base substitution patterns in the Sμ region of CH12F3-2A cells transfected with siCONT or siBRIP1. F Top: Positions of the ChIP-qPCR amplicons (black bars) on the IgH and c-Myc loci. Bottom: Western blot analysis of histone γH2AX expression in siCONT and siBRIP1-transfected CH12F3-2A cells. G Histone γH2AX-ChIP assay to assess the level of DSBs at Sµ and c-Myc region. Data are shown as the fraction of immunoprecipitated DNA (% IP) normalized to the total amount of DNA used for the immunoprecipitation. The γH2AX-ChIP signals at β2m and β-Actin genes were used as a negative control (n = 2; mean + SD; two-tailed unpaired Student’s t-test). H Top: Schematic illustration of the experimental design for the chromosomal translocation assay in CH12F3-2A cells transfected with siCONT or siBRIP1. Bottom: PCR amplification scheme to detect IgH/c-Myc chromosomal translocations. Triangles represent the positions of nested PCR primers used to amplify the recombined regions. The position of the Myc probe used in the Southern blot hybridization is shown as a horizontal black bar. I Southern blot analysis of PCR-amplified fragments using a Myc-specific probe from two independent experiments. First lane in each blot represents a Myc-positive control. J Summary of the frequency of IgH/c-Myc chromosomal translocations. Statistical significance was evaluated by Fisher’s exact test. SHM, somatic hypermutation; CIT, CD40L-IL4-TGFb, *p < 0.05; ns, not significant

B-cell malignancies are characterized by recurrent chromosomal translocations involving Ig loci and proto-oncogenes, including c-MYC, BCL-2, BCL-6, and FGFR, which are linked to Burkitt's lymphoma, follicular lymphoma, diffuse large B-cell lymphoma, and multiple myeloma, respectively. Among these, the chromosomal translocations between the IgH locus and the c-Myc oncogene are extensively studied due to their clinical significance. Since chromosomal translocations arise from DSBs, and BRIP1 KD elevated DSB formation, we investigated whether these DSBs led to increased IgH/c-Myc chromosomal translocations. In light of this, we investigated the effect of BRIP1 KD in CH12F3-2A cells on the frequency of IgH/c-Myc chromosomal translocations using nested PCR followed by Southern blotting with a c-Myc-specific probe (Fig. 3H). Similar to CSR, BRIP1 KD resulted in a significant increase in the frequency of IgH/c-Myc translocations (Fig. 3I and J). Together, these findings suggest that BRIP1 suppresses AID-induced DNA breaks, preventing aberrant chromosomal translocations, and imply that BRIP1-inactivating mutations might contribute to the development of B-cell lymphoma.

BRIP1 prevents the accumulation of R-loops at AID target sites

Since non-B DNA structures such as G4s and R-loops are implicated in AID-induced genomic instability, and BRIP1 is known to resolve G4 structures, we next examined the formation of G4s by sequences derived from the Ig Sµ and c-Myc regions using biophysical assays. G4s in the Sµ region and the c-Myc promoter are well-documented; however, the formation of G4s in exon 1 and intron 1 of the mouse c-Myc region, where translocation breakpoints are preferentially enriched (Husain et al. 2016), has not been thoroughly investigated. We first predicted G4 structures in the mouse c-Myc gene using in silico tools (Supplementary Table S3) and then analyzed the formation of these predicted G4s along with Sµ G4-forming sequences reported in the literature and a well-characterized G4-forming sequence, MycPu28, as a positive control (Wu et al. 2019) using fluorescence spectrometry and native-PAGE. First, we performed a fluorescent spectrometry assay using NMM, a G4-selective fluorescent light-up probe. In the presence of potassium chloride (KCl), we observed a dramatic increase in the NMM fluorescence for all the oligonucleotides from the Sµ (Sµ and Sµ4G) and predicted c-Myc (Myc1-Myc4) regions, indicating that these sequences form G4s (Fig. 4A). However, the increment in NMM fluorescence in the presence of lithium chloride (LiCl), a weak stabilizer of G4, was significantly lower when compared to KCl. We further validated G4 formation of the Sµ and c-Myc oligonucleotides along with their C-rich counterparts as a negative control, utilizing another G4 fluorescent probe, ThT. The Sµ and c-Myc oligonucleotides showed very high ThT fluorescence compared to the negligible fluorescence of the C-rich oligonucleotide controls (Fig. 4B). These findings suggest the G4-forming potential of both Sµ and c-Myc sequences in vitro.

Fig. 4.

AID target and off-targets form G4s, and BRIP1 loss increases R-loop formation. A Top: Schematic representation of the experimental design to detect G4 formation using NMM. Bottom: G4-forming sequences from IgH and c-Myc loci were folded in the presence of LiCl or KCl and incubated with NMM. Fluorescence of the G4-NMM complex was recorded at 610 nm (n = 2; mean + SD; two-tailed unpaired Student’s t-test). B Top: Schematic representation of G4 formation detection using ThT. Bottom: Predicted G4-forming sequences from IgH and c-Myc loci, and their C-rich control sequences were folded in the presence of KCl and incubated with ThT. Fluorescence of the G4-ThT complex was recorded at 490 nm (n = 2; mean + SD; two-tailed unpaired Student’s t-test). C Top: A schematic representation of G4 conformation analysis using native-PAGE. Middle: Position of the c-Myc primer sequences (black triangles) used in the experimental analysis. Bottom: G4-forming DNA oligonucleotides and their control C-rich oligonucleotides were incubated with or without 50 mM KCl, respectively, and electrophoresed on native TBE gel in 0.5X TBE running buffer, both of which contained 50 mM KCl. C-rich oligonucleotides served as negative controls. Asterisk (*) indicates the position of a linear C-rich oligonucleotide band, hash symbol (#) indicates intramolecular G4s, and the vertical bar (|) indicates intermolecular G4s. D Top: A schematic illustration of an R-loop immunoprecipitation with an S9.6 antibody from genomic DNA treated or untreated with RNaseH. Bottom: Positions of the DRIP-qPCR amplicons on the IgH and c-Myc loci are shown by the horizontal black bars. E R-loop enrichment at the indicated sites within the mouse IgH and c-Myc loci, expressed as percentage of input DNA, in S9.6-immunoprecipitated DNA from CH12F3-2A cells transfected by siCONT or siBRIP1 (n = 2; mean + SD; two-tailed unpaired Student’s t-test). RT is room temperature, ON is overnight incubation; *p < 0.05, **p < 0.01; ns, not significant

To determine the molecularity and strandedness of G4s formed at the Sµ and c-Myc region, we performed native PAGE followed by silver staining to assess their structural states. As C-rich oligonucleotide does not form G4s, oligonucleotide containing G to C mutations in the G-tetrad of the G-rich oligonucleotides was used as a control. The mutant oligonucleotides served as linear ssDNA molecular weight markers for each Sµ or c-Myc region sequence, and helped determine the position on the native PAGE where a folded G4 DNA oligonucleotide of the same length would be expected to migrate. The G4 DNA oligonucleotides with faster mobility than C-rich markers suggest the presence of intramolecular G4s, while those with slower mobility indicate the formation of intermolecular G4 structures. The MycPu28 oligonucleotide, which was used as a control, showed discrete bands corresponding to intermolecular and intramolecular G4 structures. Similarly, the Ig Sµ and c-Myc oligonucleotide sequences exhibited varying degrees of structural heterogeneity corresponding to both intramolecular and intermolecular G4 structures, including bimolecular and/or tetramolecular structures. Sµ, Sµ4G, Myc2, and Myc4 exhibited the co-existence of intramolecular and intermolecular G4 structures, whereas Myc1 and Myc3 predominantly formed slower-migrating intermolecular G4s with fewer intramolecular species. These biophysical data revealed that predicted G4 DNA sequences in the Sµ and c-Myc regions form G4s in vitro (Fig. 4C).

Transcription through the G-rich S-region on the non-template strand facilitates the formation of R-loops, exposing ssDNA on the non-template strand that can adopt non-B DNA conformations, including G4s, and serve as substrates for AID (Yu et al. 2003; Chaudhuri et al. 2007). Since G4 formation on the displaced strand is known to promote R-loop formation (Wulfridge et al. 2023; Duquette et al. 2004), we next examined R-loop levels at the Sµ, Sα, and c-Myc regions upon BRIP1 KD. R-loops were assessed by immunoprecipitation with the S9.6 R-loop antibody, followed by PCR with primers specific to these regions. We found that in BRIP1 KD cells, R-loop levels at the Sµ, Sα, and c-Myc regions were significantly higher than in control cells, suggesting that BRIP1 suppresses R-loop formation (Fig. 4D and E). The specificity of the S9.6 immunoprecipitation of R-loops was verified using genomic DNA pre-treated with RNase H, which degrades RNA in RNA–DNA hybrids. Together, our findings indicate that BRIP1-mediated resolution of G4s limits excessive accumulation of R-loops, thereby reducing AID-induced mutations and DSBs at AID target and off-target sites, mitigating genomic instability.

G4 structures are preferentially enriched at human AID off-target sites

BRIP1 deficiency increased AID-induced DNA breaks, IgH/c-Myc chromosomal translocations, and R-loop formation, without interfering with DNA end-joining and S-region synapsis, suggesting a possible role for BRIP1 in suppressing AID-induced genomic instability through the resolution of G4s. Hence, we analyzed the presence of G4s and other non-B DNA forms at human AID off-target sites previously identified in various studies. Through the literature survey, we compiled a dataset of 107 human non-Ig genes reported to undergo AID-dependent mutation, DNA breaks, chromosomal translocations, or kataegis (Meng et al. 2014; Kato et al. 2012; Xu et al. 2020; Qian et al. 2014; Álvarez-Prado et al. 2018). To assess the prevalence of different non-B DNA structures at these loci, genomic regions corresponding to the reported AID off-target sites were retrieved and analyzed using the NBBC bioinformatics tool, revealing the non-B DNA burden associated with each site (Fig 5A). Among the different non-B DNA motifs, G4s were the most abundant, followed by Z-DNA, short tandem repeats, inverted, mirror, and direct repeats, as well as triplex and cruciform structures (Fig. 5B and C). To determine whether any of these non-B structures are preferentially enriched at AID-target sites relative to non-targeted regions, we compared the non-B DNA distribution at targeted and non-targeted regions. AID-induced mutations occur within the promoter region of Ig genes, typically located around 100 bp upstream of the transcription start site (TSS), peaking at 200 bp downstream, and gradually tapering off to undetectable levels over the following 1.5 to 2 kb (Chi et al. 2020; Jiao et al. 2023). Therefore, as a control for the non-AID-targeted regions, the non-B burden was evaluated in a region approximately 30 kb downstream of the target sites. We observed significantly higher G4 burden in AID-targeted regions than in non-targeted control regions; however, all other non-B DNA types showed no significant difference between targeted and non-targeted regions, suggesting a higher prevalence of G4s at AID off-target sites compared to other non-B DNA types (Fig. 5D). These observations are in line with previous reports suggesting the presence of G-rich sequences at AID hotspots (Xu et al. 2020) and indicate that factors influencing G4 structures could potentially affect AID-induced genomic instability.

Fig. 5.

G4s are preferentially enriched at human AID off-target sites. A Summary of the workflow for in silico analysis of non-B DNA structure at human AID off-target genes. B Pie-chart showing the overall distribution of different non-B DNA structures at human AID off-target genes. C Extent of different non-B DNA structures at individual AID off-target sites. D Left: Schematic illustration of the experimental design to identify non-B DNA structures at the AID-targeted site versus the non-targeted control site located 30 kb downstream of the same target gene. Right: Box plot representation of non-B DNA burden of different non-B DNA motifs in both AID-targeted and non-targeted regions. Statistical significance was evaluated by using a paired two-tailed t-test; ***p < 0.001; ns, not significant. E A proposed model illustrating the accumulation of non-B DNA structures at AID-targeted sites upon BRIP1 deficiency, leading to genomic instability. Under normal conditions, BRIP1 resolves secondary DNA structures, whereas BRIP1 deficiency promotes the accumulation of G4s and R-loops, exposing both non-template and template strand deoxycytidines to AID and promoting DNA breaks and genomic instability

Discussion

Several studies have demonstrated that G4 structures contribute to genomic instability, as G4-forming sequences are frequently enriched at DNA breakpoints in cancer (Waisertreiger et al. 2025; Bacolla et al. 2016). DNA helicases maintain genomic stability by resolving these non-canonical DNA structures and preventing their excessive accumulation. Extending from the previous findings that mice deficient in the BRIP1 G4 helicase are prone to lymphoma, and that non-B DNAs, such as G4s and R-loops, are enriched at both Ig and non-Ig loci, we showed that BRIP1 prevents the accumulation of non-B DNAs. By doing so, BRIP1 suppresses AID-induced physiological recombination during CSR as well as aberrant recombination involving the IgH locus and AID off-target sites.

CSR is a locus-specific genomic rearrangement driven by the transcription of S regions, followed by AID-mediated DNA break, and their subsequent repair through the c-NHEJ or Alt-EJ pathways (Boboila et al. 2012b; Saha et al. 2021). In CH12F3-2A cells stimulated for CSR, the IgM-to-IgA switching efficiency increased upon BRIP1 KD, and these effects were rescued upon ectopic expression of BRIP1 to a normal level relative to siCONT, confirming the specific role of BRIP1 in CSR. Given that non-coding transcription forming GLTs and Aid expression are essential for CSR, we verified that BRIP1 KD did not affect the expression of these genes. Similarly, we also confirmed that BRIP1 KD did not affect cell proliferation or cell death, both of which can affect the rate of CSR. Furthermore, the 3C assay revealed that the synapsis of S-regions undergoing recombination remains unaffected in BRIP1 KD CH12F3-2A cells. These observations suggest that BRIP1 plays a direct role in regulating CSR.

Previous studies have demonstrated diverse roles of BRIP1, including aiding the repair of DNA interstrand crosslinks (ICLs) (Odermatt et al. 2020), DNA-protein crosslinks (DPCs) (Yaneva et al. 2023), and resolving G4s (London et al. 2008; Wu et al. 2008), highlighting BRIP1’s role in preserving genome stability. For instance, BRIP1 suppresses frequent genomic deletions at or near the G4 signature in Caenorhabditis elegans mutants lacking Dog-1 helicase, the homolog of BRIP1 (Cheung et al. 2002); and BRIP1 deficiency in Xenopus laevis egg extracts causes persistent replication stalling at G4 sequences (Castillo Bosch et al. 2014). Similarly, BRIP1-deficient cells or cells expressing helicase-defective BRIP1 variants are sensitive to G4-stabilizing ligands (Odermatt et al. 2020). Beyond G4 processing, BRIP1 helicase participates in DSB repair via homologous recombination and suppresses long-tract gene conversions (Nath et al. 2017). In addition, as G4 and R-loops have recently been implicated in DSB repair pathway choice decision, BRIP1 may also indirectly affect DSB repair by regulating these structures at the DSB ends (Castillo Bosch et al. 2014; Wu and Spies 2016). In fact, HNRNPU has been shown to promote the repair of AID-induced DSBs during CSR by favoring c-NHEJ through binding to R-loops and/or G4s at the S regions, thereby limiting DNA end-resection and ssDNA accumulation (Refaat et al. 2023).

While the AID-induced DSBs are primarily repaired by the c-NHEJ pathway, involving no or short MH (1–3 ntds), in the absence of the c-NHEJ pathway, they can also be repaired via alt-EJ, utilizing longer MHs (≥4 ntds) (Chang et al. 2017; Lee-Theilen et al. 2011). Analysis of the recombined Sµ and Sα junction characteristics in CH12F3-2A cells revealed that BRIP1 KD does not affect the use of MH during DSB repair. Consistent with our findings, Sµ and Sα junction sequences from a child with homozygous BRIP1 deficiency revealed no significant effect on direct/blunt end-joining but showed a slight decrease in small insertions at the recombined CSR junction (Björkman et al. 2015). Conversely, CSR junctions from an adult with heterozygous BRIP1 deficiency showed a reduction in direct end joining and small insertions, along with a concomitant increase in the use of longer MHs (Björkman et al. 2015). Since only heterozygous but not homozygous deficiency in BRIP1 affected the use of MHs at CSR junctions in these patients, it is difficult to conclude, based on these results, that BRIP1 plays a direct role in the repair of AID-induced DNA breaks. Alternatively, the differences in end-joining between heterozygous and homozygous patients could result from age-related variations in the utilization of MH-dependent end-joining pathways. However, consistent with these reports, we observed a modest decrease in the frequency of insertions at recombined Sμ–Sα junctions and an increase in the frequency of junctions containing ≥4 nucleotides of microhomology.

While AID is essential for immune function, aberrant expression or mistargeting of AID can predispose B-cells to DSBs, resulting in chromosomal translocations that can deregulate oncogenes and contribute to the development of lymphoma (Robbiani and Nussenzweig 2013; Küppers and Dalla-Favera 2001; Leder 1985). An increase in SHM, DSBs, and IgH/c-Myc translocations upon BRIP1 KD demonstrates that BRIP1 suppresses AID-induced genomic instability, suggesting that BRIP1-inactivating mutations may increase the occurrence of translocations and the development of B-cell malignancies. Together, these observations suggest that BRIP1 does not actively participate in the DSB repair pathway choice or synapsis during CSR, but rather functions upstream to suppress break formation, thereby preventing genomic instability.

The S and c-Myc regions of the mouse and human IgH loci are rich in G-clusters and form non-canonical DNA structures (Chaudhuri et al. 2003; Dézé et al. 2023; Ramiro et al. 2003). While TOP1 unwinds the DNA helix to relieve negative supercoiling and prevent the accumulation of non-B DNA, DNA helicases such as BRIP1 resolve these abnormal structures. Biophysical analysis of mouse Sµ and c-Myc regions revealed G4 formation, indicating the presence of distinct G4 DNA substrates for BRIP1 within these regions. Furthermore, we observed increased levels of R-loops at both S regions and the c-Myc locus in BRIP1 KD CH12F3-2A cells, suggesting that the unresolved G4 resulting from BRIP1 KD promotes R-loop formation. Consistent with our results, BRIP1-deficient neuronal (HT22) and neuroblastoma (SH-SY5Y) cell lines showed increased R-loop levels (Mani et al. 2022). Similarly, PIF1, another G4 helicase, has been shown to suppress R-loop formation in yeast (Tran et al. 2017). Furthermore, treatment of U2OS cells with G4-stabilizing ligands increases R-loop levels (De Magis et al. 2019). Taken together, these results are consistent with the observation that persistent G4 structures stabilize and promote R-loop formation.

Extending these locus-specific insights to a genome-wide scale, computational analysis of previously identified human AID off-target sites revealed a preferential enrichment of G4-forming motifs, compared with other non-B DNA forms. Consistent with these observations, previous studies have shown enrichment of G4-forming motifs within AID-targeted genomic regions in tissues from B-cell malignancies (Xu et al. 2020), suggesting that G4 persistence may influence AID-induced genomic instability. However, it is important to note that these in silico analyses complement, rather than provide direct evidence for, G4 formation in vivo, as G4 formation and stability are also influenced by chromatin organization, transcriptional dynamics, and helicase-mediated G4 and R-loop resolution. Furthermore, because G4 structures can adopt different conformations, AID may not recognize all G4 conformations with equal affinity. Therefore, future studies are needed to determine the G4 topologies preferentially targeted by AID and to validate the physiological relevance of predicted human AID off-target sites. Given that reduced BRIP1 function increases non-B DNA structure and enhances the persistence of AID-induced lesions, we propose that this increased genomic instability stems from the accumulation of G4s on unresolved non-template strand and R-loops on the template strand. As previously reported, G4s serve as substrate for AID, which may also expose template-strand deoxycytidines to AID through hybrid G4 structure formed by the displaced non-template strand and the RNA component of the R-loop (Pucella and Chaudhuri 2017), resulting in DNA breaks and ultimately genomic instability (Fig. 5E).

While our study in the CH12F3-2A cell line provides insights about the role of BRIP1 in AID-induced events, extending these findings to primary B-cells, in vivo models, or patient-derived samples with BRIP1 aberrations would further elucidate its role in genome instability. Ultimately, future studies examining the response of BRIP1-deficient B-cells to G4-stabilizing ligands, DNA cross-linking agents, or DNA repair inhibitors will be required to determine whether BRIP1 can serve as a biomarker of genomic instability or therapeutic vulnerability. In summary, our findings reveal that BRIP1-mediated resolution of G4s prevents the accumulation of transcription-associated R-loops, thereby limiting AID-induced DNA damage and chromosomal rearrangements, safeguarding B-cell genome integrity. These findings suggest that BRIP1 dysfunction may contribute to B-cell malignancies.

Abbreviations

AID

Activation-induced cytidine deaminase

BRIP1

BRCA1-interacting protein 1

KD

Knockdown

CSR

Class switch recombination

SHM

Somatic hypermutation

G4

G-quadruplex

DSB

Double strand DNA break

ssDNA

Single stranded DNA

EJ

End-joining

Authors’ contributions

AM: Conceptualization, Methodology, Formal analysis, Investigation, Writing - Original Draft, Writing - Review & Editing, Visualization; HA: Methodology, Investigation; NF: Methodology, Investigation; NB: Conceptualization, Methodology, Investigation, Resources; AH: Conceptualization, Methodology, Formal analysis, Writing - Original Draft, Writing - Review & Editing, Visualization, Resources, Supervision, Project administration, Funding acquisition.

Funding

This work was financially supported by the Start-Up Research Grants from the Department of Science and Technology (SRG-2020-000819) and University Grant Commission (F.30-564/2021 (BSR), the Government of India.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Consent for publication

All co-authors have approved the manuscript and agreed to its publication.

Competing interests

The authors declare no competing interests.