www.nature.com/npjbcancer CASE REPORT OPEN Clinical consequences of BRCA2 hypomorphism Laia Castells-Roca1,2,10, Sara Gutiérrez-Enríquez 3,10, Sandra Bonache3, Massimo Bogliolo1,2,4, Estela Carrasco5, Miriam Aza-Carmona1, Gemma Montalban 3,9, Núria Muñoz-Subirana1, Roser Pujol1,2,4, Cristina Cruz 6, Alba Llop-Guevara6, María J. Ramírez1,2,4, Cristina Saura7, Adriana Lasa 2,8, Violeta Serra6, Orland Diez5, Judith Balmaña 5 ✉ and Jordi Surrallés 1,2,4 ✉ The tumor suppressor FANCD1/BRCA2 is crucial for DNA homologous recombination repair (HRR). BRCA2 biallelic pathogenic variants result in a severe form of Fanconi anemia (FA) syndrome, whereas monoallelic pathogenic variants cause mainly hereditary breast and ovarian cancer predisposition. For decades, the co-occurrence in trans with a clearly pathogenic variant led to assume that the other allele was benign. However, here we show a patient with biallelic BRCA2 (c.1813dup and c.7796 A > G) diagnosed at age 33 with FA after a hypertoxic reaction to chemotherapy during breast cancer treatment. After DNA damage, patient cells displayed intermediate chromosome fragility, reduced survival, cell cycle defects, and significantly decreased RAD51 foci formation. With a newly developed cell-based flow cytometric assay, we measured single BRCA2 allele contributions to HRR, and found that expression of the missense allele in a BRCA2 KO cellular background partially recovered HRR activity. Our data suggest that a hypomorphic BRCA2 allele retaining 37–54% of normal HRR function can prevent FA clinical phenotype, but not the early onset of breast cancer and severe hypersensitivity to chemotherapy. npj Breast Cancer (2021)7:117 ; https://doi.org/10.1038/s41523-021-00322-9 1234567890():,; INTRODUCTION HRR activity to prevent FA onset, but it is insufficient to avoid Fanconi anemia (FA) is a rare genetic DNA repair syndrome tumorigenesis during adulthood and results in acute sensitivity to characterized by bone marrow failure (BMF) and cancer predis- chemotherapy. position particularly to acute myeloid leukemia (AML) and/or solid tumors, mainly head and neck squamous cell carcinoma (HNSCC)1. FA cells exhibit chromosomal instability and hypersensitivity to RESULTS DNA damage especially to DNA interstrand crosslinks (ICLs)2. Biallelic BRCA2-FA disorder diagnosis Twenty-two FA-related genes have been identified including Patient FA663 was a thirty-three-year-old woman who developed a tumor suppressors BRCA1 (FANCS) and BRCA2 (FANCD1). malignant breast nodule and, after surgery, showed a severe The breast cancer susceptibility 2 (BRCA2) protein is a key intolerance to chemotherapy. A timeline of patient’s progression is component of the DNA homologous recombination repair (HRR) shown in Fig. 1A. FA663 was an adopted child and her medical pathway3 and a downstream component of the FA/BRCA DNA history only indicated slight abnormalities such as low birth weight, repair pathway1,2. Monoallelic germline pathogenic variants in short stature, a subtle learning disability, skin hyperpigmentation, BRCA2 are a common cause of hereditary breast and ovarian macrocytosis, hiatal hernia, hepatic steatosis, and lumbar disc cancer (HBOC)4, while biallelic pathogenic variants are associated herniation. Due to the severe pancytopenia reaction to che- with FA5. Individuals with biallelic BRCA2 pathogenic variants motherapy, a chromosome fragility test (CFT)11 was performed in usually suffer from a severe form of FA characterized by BMF, blood T-lymphocytes and the levels of chromosome fragility were multiple congenital defects, and severe cancer predisposition as compatible with a diagnosis of FA with somatic mosaicism (FA- patients develop leukemia and/or multiple solid tumors (mainly mosaic) as only one-third of her cells presented multiple Wilms and brain tumors) in the first decade of life. The overall risk chromatid-type breaks and radial figures. (Fig. 1B)11,12, FA663 case of cancer in these patients is about 97% by the age of 5 years5–9. was then included in a cohort of FA patients to be molecularly Here, we describe an unusual clinical case consisting of a female characterized by Exome Sequencing (ES)13, and two variants were patient who reached adulthood with no clear suspicion of a identified in the BRCA2/FANCD1 gene. The first, c.1813dup, p. defined pathology, despite presenting two germline BRCA2 (Ile605Asnfs*11) was a clearly pathogenic variant in BRCA2 exon 10 variants: a null frameshift variant c.1813dup, p.(Ile605Asnfs*11) that produced a frameshift and premature termination of and a missense variant c.7796 A > G, p.(Glu2599Gly). Using a translation (Fig. 1C). This variant had previously been described BRCA2 KO HEK293T cell line stably transfected with an HRR GFP in FA patients8 and in HBOC patients14 and it is reported in ClinVar marker construct10 we demonstrated that the missense allele is as pathogenic class 5 classified by the ENIGMA expert panel partially functional providing 37–54% of HRR activity. We propose (https://www.ncbi.nlm.nih.gov/clinvar/variation/37762/). A second that p.(Glu2599Gly) is a hypomorphic allele that provides sufficient missense variant c.7796 A > G, p.(Glu2599Gly) was identified in 1 Genome Instability and DNA repair Syndromes Group and Join Unit UAB-IR Sant Pau on Genomic Medicine, Biomedical Research Institute IIB-Sant Pau, Hospital de la Santa Creu i Sant Pau, Barcelona, Spain. 2Genetics Department, Hospital de la Santa Creu i Sant Pau, Barcelona, Spain. 3Hereditary Cancer Genetics Group, Vall d’Hebron Institute of Oncology (VHIO), Vall d’Hebron Barcelona Hospital Campus, Barcelona, Spain. 4Center for Biomedical Network Research on Rare Diseases (CIBERER) U-745, Barcelona, Spain. 5Hereditary Cancer Genetics Group, Vall d’Hebron Institute of Oncology (VHIO), Hospital Universitari Vall d’Hebron, Vall d’Hebron Barcelona Hospital Campus, Barcelona, Spain. 6Experimental Therapeutics Group, Vall d’Hebron Institute of Oncology (VHIO), Vall d’Hebron Barcelona Hospital Campus, Barcelona, Spain. 7Breast Cancer and Melanoma Group, Vall d’Hebron Institute of Oncology (VHIO), Hospital Universitari Vall d’Hebron, Vall d’Hebron Barcelona Hospital Campus, Barcelona, Spain. 8Center for Biomedical Network Research on Rare Diseases (CIBERER) U-705, Barcelona, Spain. 9Present address: CHU de Québec – Université Laval Research Center, Oncology division, 9 Rue McMahon, Québec city G1R 3S3 Québec, Canada. 10These authors contributed equally: Laia Castells-Roca, Sara Gutiérrez-Enríquez. ✉email: jbalmana@vhio.net; jsurralles@santpau.cat Published in partnership with the Breast Cancer Research Foundation L. Castells-Roca et al. 2 1234567890():,; Fig. 1 Breast cancer FA663 patient diagnosed with biallelic BRCA2-FA disorder. a Timeline of patient’s cancer history (diagnoses, treatments, studies) and BRCA2 analysis. b Diepoxybutane (DEB) induced chromosome fragility test (CFT) with blood T-lymphocytes. Percentage distribution of aberrant cells and mean the number of DNA breaks per cell, after DEB induction (0.1 µg/ml). A black square-shaped symbol shows the relative position of FA663 in the distribution of historical data from our laboratory. In the “no Fanconi” group only individuals with at least one aberrant cell have been included. c The two BRCA2 pathogenic variants of FA663. BRCA2 protein (P51587) and InterPro domains schematic drawing. Sequenced loss of function variants found in heterozygosis depicted. d the highly conserved amino acid at position 2,599 of the BRCA2 protein. Partial alignment of human BRCA2 with homologous BRCA2 helical domains (IPR015252). BC breast cancer, TNBC triple-negative breast cancer, DCIS ductal carcinoma in situ, IDC invasive ductal carcinoma, ER− estrogen receptors negative, PR − progesterone receptors negative, TN triple-negative, CRC colorectal cancer, M1 metastasis stage 1, MLPA multiplex ligation-dependent probe assay. BRCA2 exon 16 (Fig. 1C) and classified as of “uncertain significance” Evidencing the in trans position of BRCA2 c.1813dup and according to the ACMG guidelines15. In ClinVar three submitters c.7796 A > G variants also interpreted this variant as of uncertain significance (https:// To assess if c.1813dup and c.7796 A > G variants were in trans, www.ncbi.nlm.nih.gov/clinvar/variation/483130/). Human Glu2599 multiplex ligation-dependent probe assay (MLPA) and Sanger is a highly conserved residue among organisms in the BRCA2 DNA sequencing of the BRCA2 gene were undertaken in normal tissue binding domain (Fig. 1D) and the detected variant produced a samples and tumor samples from FA663 (Table 1). The two allelic non-conservative change from a negatively charged amino acid variants of BRCA2 were found in blood, skin tissue, and in three (Glu) to a non-polar amino acid (Gly). The p.(Glu2599Gly) is primary tumor samples: invasive ductal carcinoma (IDC), contral- predicted deleterious for BRASS NN (score 0.912 pathogenic), ateral ductal carcinoma in situ (DCIS), and colorectal cancer (CRC) BRASS MLR (score 1.675 pathogenic), AGVGD (Class C65), identified during a necropsy study (Table 1, Fig. 3A–D and Figure PolyPhen-2 (score 1 probably damaging) and SIFT (score 0.01 S1A). However, four metastatic breast cancer samples (hepatic predict non-tolerated) in silico tools. No other potentially patho- metastasis and supraclavicular, pleural, hilar, and lymph nodes) genic variants were identified in any other known FA gene. infiltrated by the primary breast IDC tumor, and two patient- derived tumor xenograft (PDX) models (derived from supraclavi- cular and hilar metastases) showed loss of heterozygosity (LOH) in Patient fibroblasts retain partial DNA damage response BRCA2 (Table 1, Fig. 3E–F and Figure S1B–E). Specifically, all the infiltrated nodes and PDXs samples presented exclusively the To confirm the diagnosis of FA663 as a FA mosaic11,16, and to study c.7796 G allele, suggesting that the two variants were located in more in-depth patient’s DNA damage response (DDR), FA663 trans (Table 1). In addition, the cDNA obtained from blood RNA primary fibroblasts were analyzed together with fibroblasts from a and cultured fibroblasts was sequenced by Sanger (Figure S2). We wt control and from a patient with a homozygous BRCA2 null observed a lower signal of the allele carrying the duplication in variant c.469 A > T; p.(Lys157*). In the CFT, FA663 fibroblasts exon 10 (as a result of degradation by the nonsense-mediated showed a chromosome fragility level in-between the wt and BRCA2 mRNA decay (NMD) pathway) and a higher signal of the G allele of deficient control and similar to that of blood T-lymphocytes (≈32% the missense variant c.7796 A > G in exon 16 with respect to that aberrant cells; ≈2.88 breaks per aberrant cell) (Fig. 2A). Similarly, of the A wt allele, which is consistent with the two variants being intermediate results were also obtained in ICL induced G2/M phase located in different alleles. Overall, all these results support the cell cycle arrest (Fig. 2B), RAD51 foci formation (Fig. 2C), and position in trans of the two variants. sensitivity to the PARP inhibitors olaparib and veliparib (Fig. 2D, E). npj Breast Cancer (2021) 117 Published in partnership with the Breast Cancer Research Foundation L. Castells-Roca et al. 3 Fig. 2 FA663 primary fibroblasts display partial BRCA2 activity. a Increased DNA breaks per aberrant cell in FA663 fibroblasts, induced by DEB. The graph represents the mean of 25 aberrant metaphases. Error bars show s.e.m. Statistics are done with a non-parametric Mann- Whitney test, *P < 0.05. b MMC induces G2/M cell cycle partial arrest in FA663. Cell cycle distribution was monitored by flow cytometry, 72 h after treatment. The representative experiment shown, from three independent data sets. c Attenuated RAD51 foci formation in FA663 after IR. Immunofluorescence analysis of wt, BRCA2-mutant, and FA663 primary cell lines 6 h after 5 Gy. The graph represents the mean of three independent experiments with duplicates. Error bars show s.e.m. Statistics are done with the One-Way ANOVA test and Bonferroni correction **P < 0.01, ***P < 0.001. d, e PARPi sensitivity of FA663 primary cells. Data represent the mean of three independent experiments. Error bars represent s.d. Statistics with two-tailed t-Test *P < 0.05, ***P < 0.001, comparing BRCA2-mutant cell line. (wt: human primary skin fibroblasts from a healthy individual; BRCA2: primary skin fibroblasts from a homozygous c.469 A > T, p.Lys157* FA patient; FA663: primary skin fibroblasts established from skin biopsy of patient FA663). The hypomorphic missense p.(Glu2599Gly) variant displays KO cell line was generated in our laboratory by CRISPR/Cas9 37–54% of HRR activity technology and it was functionally validated (Figure S3). Plasmids Next, we generated a cellular system to quantify mutation-specific containing wt BRCA2 cDNA, empty vector (EV), and BRCA2 cDNA levels of HRR by genetic complementation of a human-derived bearing either p.(Ile605Asnfs*11) or p.(Glu2599Gly) were transi- BRCA2 deficient cell line, stably transfected with a GFP-based ently transfected in the BRCA2 KO HEK293T-DR-GFP. No expression construct to detect HRR (BRCA2 KO HEK293T DR-GFP)10. The BRCA2 of p.(Ile605Asnfs*11) BRCA2 protein could be observed by Published in partnership with the Breast Cancer Research Foundation npj Breast Cancer (2021) 117 L. Castells-Roca et al. 4 Table 1. BRCA2 biallelic genetic validation and tumor sample analysis from FA663 patient. Sanger sequencing of BRCA2 exons 10 and 16 and allelic copy number analysis by MLPA for all the samples analysed. Tissue Tissue type Sample type c.1813dup c.7796 A > G LOH MLPA BRCA2 Blood N fresh het het 2n Skin biopsy N fresh het het 2n Primary right IDC P FFPE het het 2n Primary left DCIS P FFPE het het 2n Primary colorectal tumour P FFPE het het 2n Supraclavicular node M frozen wt G yes 1n Pleural node M frozen wt G yes 1n Hilar node M frozen wt G yes 1n Hepatic node M frozen wt G yes 1n PDX from supraclavicular node M fresh wt G yes 1n PDX from hilar node M fresh wt G yes 1n N normal tissue, P primary tumour, M metastatic tissue, het heterozygous, IDC invasive ductal carcinoma, DCIS ductal carcinoma in situ. Western blot (WB) analysis, while the missense BRCA2 p. in BRCA2 that survived to adulthood with mild Fanconi Anemia- (Glu2599Gly) was expressed at comparable levels with the associated clinical features, but nonetheless developed young- endogenous wt BRCA2 (Fig. 4A and Figure S4). BRCA2 mRNA onset breast cancer and had fatal evolution. Our functional expression was detected in PDX332 (derived from the supracla- complementation studies showed that p.(Glu2599Gly) missense vicular node) to a similar level of BRCA2-wt PDXs, further allele retains between 37 and 54% of normal HRR activity levels, evidencing the expression of the missense BRCA2 p.(Glu2599Gly) likely sufficient to suppress BMF, leukemia, and solid tumors variant (Figure S5). The HRR capacity of the two BRCA2 variants during childhood, but not breast or colon cancer at adult age. Our was further quantified, using both an ionizing radiation-induced accurate molecular screening of DNA from normal and tumor RAD51 foci formation (IR-IF) assay (Fig. 4B) and the DR-GFP tissues demonstrated that the c.1813dup and c.7796 A > G reporter test by flow cytometry (Fig. 4C and Figure S6). With both variants were positioned in trans. The predicted frameshift caused techniques, cells expressing the missense allele had intermediate by the FA663 variant c.1813dup produces a premature stop levels of HRR compared to the BRCA2 wt reconstituted cells that codon, which implies a degradation of the resulting transcript by exhibited functional HRR, or the cells expressing the p. NMD18. Our results of Sanger sequencing of blood and fibroblasts (Ile605Asnfs*11) that were HRR deficient (Fig. 4B, C). We cDNA show a low signal of one allele supporting the NMD established that the residual HRR function in the missense variant hypothesis (Figure S2). p.(Glu2599Gly) was 54%± 11 (s.e.m) (Fig. 4B) and of 37% ± 7 (s.e. Most of the BRCA-associated tumors undergo the loss of the m) (Fig. 4C). In agreement with these findings, we observed that BRCA-wt allele during tumor evolution, which turns the tumor into the level of baseline RAD51 foci in tumor samples from the an HRR deficient (HRD) state and results in tumor sensitivity to metastatic supraclavicular node and the two PDXs were inter- DNA damaging chemotherapy. Intriguingly, while the primary mediate and significantly different to BRCA2 wt and mutant breast tumors from this patient did not undergo LOH (Table 1; Fig. 3B–D), tumors (Figure S7). the metastatic tumors lose the BRCA2 truncating variant, retaining the missense mutated allele (Table 1; Fig. 3E–F and Figure S1B–E). This could be an evolutionary advantage acquired by the DISCUSSION metastatic tumors, rendering them capable of partly repairing The role of BRCA2 in DDR is relevant not only for tumor the DNA damage induced by chemotherapy and resulting in suppression, but also for organisms’ viability. Preceding studies chemoresistance. This hypothesis underscores an active role of the in mice demonstrated that Brca2 is essential for normal embryonic missense variant in the tumorigenic phenotype. development17 and indirect evidence indicates that most of the Remarkably, at least six FA cases of biallelic BRCA2 patients due time this is also true for humans. Pathogenic variants such as to the combination of one clearly deleterious allele and one c.5946delT (also known as 6174delT) with a carrier frequency of missense allele have been previously reported6. Three of the around 1–1.5% in Ashkenazi Jewish population have not been variants, p.(Lys2729Asn), p.(Asn372His), and p.(Ile2490Thr) showed ever detected in the homozygous state as well as many common functional HRR19,20 and they appear classified in ClinVar as BRCA2 pathogenic variants that are located in exon 119. Never- “Benign” (by expert review), implying that were probably not theless, it has been shown that biallelic deleterious BRCA2 variants the second causal variants in these FA-D1 patient6. The variants, p. in humans are compatible with life, despite being associated to (Leu2510Pro) and p.(Trp2626Cys) were able to partially rescue the severe forms of FA and childhood cancer predisposition6–9. In line lethality of Brca2-null (Brca2KO/KO) ES cells, but with a 70–80% with these formerly reported cases6, FA patients with biallelic reduction in the number of colonies and an HRR severely BRCA2 pathogenic variants previously identified in our laboratory impaired19,20. Cells expressing these variants had residual HRR presented elevated chromosomal fragility, severe congenital activity, below 5%, and a 20-fold increase in ICLs induced aberrations and/or high cancer incidence by age 3 or younger chromosome fragility when compared to the wt BRCA2 expressing (Table S1). On the contrary, patient FA663 had a tumor-free cell lines19,20. Similarly, the p.(Arg2336His) variant, due to the childhood and mild Fanconi Anemia-associated clinical features, c.7007 G > A substitution in DNA, also behaves as a BRCA2 null which remained practically unnoticed to clinicians. Our work allele when expressed in ES cells19,20. The c.7007 G > A change is a focuses on the detailed molecular characterization of a patient splicing alteration that generates two frameshift mRNA transcripts being a carrier of a pathogenic variant and a hypomorphic variant (skipping exon 13 and exons 12–13, respectively), with the npj Breast Cancer (2021) 117 Published in partnership with the Breast Cancer Research Foundation L. Castells-Roca et al. 5 a Blood b Primary right IDC exon 10 exon 10 * * exon 16 exon 16 * * c Primary left DCIS d Primary CRC exon 10 exon 10 * * exon 16 exon 16 * * e Metastatic hilar node f PDX from hilar node exon 10 exon 10 * * exon 16 * exon 16 * Fig. 3 Sanger sequencing and MLPA results from normal and tumor samples indicate that the c.1813dup and c.7796 A > G BRCA2 variants are in trans. Sanger electropherograms for both variants c.1813dup (exon 10) and c.7796 A > G (exon 16) show forward sequences, except for the invasive ductal carcinoma sample (IDC) (b) exon 16. MLPA results show the allelic dosage in a ratio chart format as displayed by Coffalyser.Net software. Probe ratios between 0.7 and 1.3 (blue and red lines) showed the presence of two BRCA2 alleles for blood (a), primary IDC (b), primary ductal carcinoma in situ (DCIS) (c), and primary colorectal cancer (CRC) (d). However, probe ratios for metastatic tumors (e) and their corresponding PDX (f) were around 0.5, suggesting the presence of only one BRCA2 allele. All Sanger sequences are consistent with their corresponding MLPA findings, i.e. presence of heterozygote variants in Sanger matches with the presence of two allele MLPA evidence while the monoallelic presence of a variant matches with one allele from MLPA outcome. MLPA and Sanger sequences for the skin biopsy and the other metastatic and PDXs samples are shown in Figure S1. *indicates the variant location. appearance of an early stop codon in exon 1419,20. Since moderate risks of breast cancer28, being the hypomorphic variants functional studies demonstrated a deep impairment of HRR20, associated with this domain the only missense variants in BRCA2 these last three variants cannot be considered hypomorphic, in increasing cancer risk. Although we show that BRCA2 p. agreement with the acute FA phenotype and early cancer (Glu2599Gly) is hypomorphic in terms of DNA repair function, predisposition shown by these patients6. In contrast, p. this study cannot provide a quantification of its estimated reduced (Glu2599Gly) found in our patient exhibited substantial levels of cancer penetrance, due to the fact that it is presented together HRR in all functional assays performed in our study but well below with another clearly deleterious variant and the lack of a the wt levels (Fig. 2, Fig. 4 and Figure S5), in accordance to recent significant number of patients harboring this missense variant. data21. Moreover, we observed the same intermediate chromo- Our findings together with two family study cases that reported some fragility in both primary lymphocytes and fibroblasts from biallelic BRCA2 patients with a less severe FA phenotype and early FA663 patients. This is indicative of a hypomorphic effect of the p. onset breast and/or colon cancer susceptibility29,30 are relevant for (Glu2599Gly) variant and not of FA somatic mosaicism. Indeed, a the interpretation of variants with unknown significance (VUS) in diagnosis of FA somatic mosaicism is usually consistent with the BRCA2. In the evaluation of the potential pathogenicity of a variant presence of higher induced chromosome fragility in patient’s in BRCA2, it is considered that its co-occurrence in trans with a fibroblasts than patient’s lymphocytes12,22,23. clearly pathogenic variant indicates its benign character31. The p.(Glu2599Gly) variant presented by our patient is located However, our results show that this criterion cannot be applied in the helical domain (2479 to 2668aa) of the DNA binding region without careful exploration of the clinical phenotype. of BRCA2. This domain interacts with DSS124, which promotes the We conclude that the BRCA2 p.(Glu2599Gly) is a truly exchange of BRCA2-RPA1 to BRCA2-RAD51 and thus the ssDNA hypomorphic BRCA2 allele that retains sufficient HRR activity for interaction25. The BRCA2 DNA binding domain is also involved in normal organism development and to minimize FA-related the stabilization of BRCA2 and in the regulation of R-loop- phenotypes, but not enough to avoid early onset of tumorigenesis associated DNA damage, a particular source of damage resolved and severe chemotoxicity. Furthermore, we propose a straightfor- by FA proteins26,27. In addition, it has been shown that ward methodology that produces conclusive results in a reason- hypomorphic missense variants in the DNA binding domain of able turnaround time to classify BRCA2 variants of uncertain BRCA2 with partial effects on protein function may confer significance (VUS) in cancer patients. Published in partnership with the Breast Cancer Research Foundation npj Breast Cancer (2021) 117 L. Castells-Roca et al. 6 Fig. 4 The hypomorphic missense p.(Glu2599Gly) variant has partial HRR activity. a Only the BRCA2 p.(Glu2599Gly) variant is expressed. Immunoblotting of BRCA2 in wt and BRCA2 KO HEK293T transiently transfected with pCDNA3 wt BRCA2, empty vector (EV), BRCA2 p.(Ile605Asnfs*11) and BRCA2 p.(Glu2599Gly), 48 h after transfection. Vinculin antibody was used as a loading control on the same sample run and processed in parallel. b BRCA2 p.Glu2599Gly variant shows partial complementation of RAD51 foci formation. IR- IF analysis of RAD51 in HEK293T BRCA2 KO cell line expressing wt BRCA2, EV, BRCA2 p.(Ile605Asnfs*11) and BRCA2 p.(Glu2599Gly), 6 h after 5 Gy. Mean of three independent experiments represented. Error bars show s.d. Statistics are done with the One-Way ANOVA test and Bonferroni correction **P < 0.01, ***P < 0.001. c Intermediate HR efficiency of p.(Glu2599Gly) BRCA2 expressing cells. GFP expression from the DR-GFP reporter was analysed as a measure of HR activity. The percentage of GFP-positive cells for each variant was shown. Results represent the mean of at least four independent experiments. Error bars represent s.d. Statistics are done with the One-Way ANOVA test and Newman–Keuls multiple comparison tests, *P < 0.05, **P < 0.01, ***P < 0.001. surgery; homozygous BRCA2 c.469 A > T, p.Lys157* and FA663 primary fibroblasts were established from skin biopsies of patients. Cells were grown in DMEM with 20% FBS (Biowest) and antibiotics. Human cell lines HEK293T (Human Embryonic Kidney 293 T cells, ATCC CRL–11268) were cultured in DMEM 10% FBS and antibiotics. Chromosome fragility assays and G2/M cell cycle arrest For the Chromosome fragility test, peripheral blood T-lymphocytes and primary fibroblasts were treated with diepoxybutane (DEB) at the indicated concentrations and processed as in11,32. For the G2/M cell cycle arrest analysis, primary fibroblasts were treated with Mitomycin-C at the indicated concentrations and processed as in32. RAD51 foci formation 75000 fibroblast cells were grown on 4-well Millicell EZ slide units (Merck Millipore) and irradiated with 5 Gy. Cells were fixed with 4% formaldehyde in PBS (Quimigen), incubated in Triton X-100 0.5%, and blocked with 10% BSA (Sigma) TPBS. Slides were incubated with RAD51 antibody (H-92, 8349 Santa Cruz) diluted 1:500 overnight at 4 °C, washed and incubated for 30 min at 37 °C with anti-rabbit Alexa 568 conjugated secondary antibody (Molecular Probes) diluted 1:500. Coverslips were mounted in Vectashield Mounting Medium (H-1000, Vector Laboratories, Inc.), containing DAPI. 300000 HEK293T cells were seeded on coverslips (previously coated with Poly-L-Lysine) in 6 cm Petri dishes and the same procedure was followed. RAD51 on formalin-fixed paraffin-embedded (FFPE) samples The homologous recombination repair activity was assessed on FFPE tumor samples as previously described33,34. In brief, γH2AX and RAD51 foci were determined in cells in the S/G2-phase of the cell cycle (geminin- positive). The following primary antibodies were diluted in DAKO Antibody Diluent (K8006) and incubated at room temperature for 1 h: rabbit anti- RAD51 (Abcam ab133534, 1:1000), mouse anti-geminin (NovoCastra NCL-L, 1:100 in PDX samples, 1:60 in patient samples), rabbit anti-geminin (ProteinTech 10802-1-AP, 1:400) and mouse anti-phospho-H2AX (Millipore #05-636, 1:200). Secondary antibodies were diluted in blocking buffer and incubated for 30 min: goat anti-rabbit IgG Alexa fluor 568 (Invitrogen A11011; 1:500), goat anti-mouse IgG Alexa fluor 488 (Invitrogen A11001; 1:500), donkey anti-mouse IgG Alexa fluor 568 (Invitrogen A10037; 1:500), and goat anti-rabbit IgG Alexa fluor 488 (Invitrogen A11008; 1:500). Finally, sections were dehydrated and mounted with DAPI ProLong Gold antifade reagent (ThermoFisher P36941). Scoring was performed blindly using a 60×-immersion oil objective with a Nikon Eclipse Ti-E microscope. The METHODS amount of dsDNA damage served as a quality check and was quantified by Cell lines and cultures scoring the percentage of geminin-positive cells with γ-H2AX foci. The Blood cultures were established from patient FA663: 10% of blood (in RAD51 score was obtained by quantifying the percentage of geminin- heparin) and 90% RPMI (Gibco) with 15% FBS, 1% antibiotics, 1% L- positive cells with 5 or more nuclear foci. A pre-defined cut-off of 10% was Glutamine, and 1% phytohemaglutinin. Wild type primary fibroblasts were used for RAD51 to differentiate between homologous recombination previously established from a 4-year-old patient undergoing phimosis deficient (HRD) and proficient (HRP) tumors33,34. npj Breast Cancer (2021) 117 Published in partnership with the Breast Cancer Research Foundation L. Castells-Roca et al. 7 PARPi survival assays most stably expressed housekeeping genes (GUSB and ACTB) and 1000 fibroblast cells (wt, BRCA2 and FA663) were seeded on 96-well plates geometric means were calculated to obtain normalized CT values37. and treated with 0, 0.5, 5, 50, 500, and 5000 nM Olaparib (Selleckchem) or 0, 50, 500, 5000, and 50000 nM Veliparib (Deltaclon S.L.). Cells were fixed In silico prediction tools adding 10% trichloroacetic (TCA) (Sigma) solution, stained with 0.4% The in silico prediction of functional impact of the missense BRCA2 Sulforhodamine B (Sigma) acetic acid, and resuspended with 10 mM Tris c.7796 A > G variant was performed using BRASS (a specific tool for BRCA1 (Sigma) solution. OD has measured 510 nm with a microplate reader and BRCA2 genes; https://www.biotoclin.org/BRASS#top)38 and Alamut (Multiskan Sky Touch Drop plate, ref51119600DP). software v2.10 (Interactive Biosoftware) that includes three protein- function prediction tools: Align-GVGD, (http://agvgd.hci.utah.edu/ BRCA2 mutant cell line generation agvgd_input.php), SIFT (https://sift.bii.a-star.edu.sg/), and PolyPhen-2 The selected plasmid to generate CRISPRs constructs was pX330 (pX330 – (http://genetics.bwh.harvard.edu/pph2/). U6 – Chimeric_BB-CBh-HSpCas9, Addgene 42230)35. Selected sequence to target BRCA2 is GTTGATTTCCAGTACCAACT, located in exon 11 and DNA extraction and genetic variant screening previously described35. Primers used for CRISPR pX330 plasmid generation Genomic DNA was extracted from patient peripheral blood and skin biopsy are forward 5’-caccgttgatttccagtaccaact and reverse 5’-aaacagttggtactg- samples using Puregene Genome DNA Purification Kit (Gentra System). gaaatcaac. Three days after, cells were analyzed by flow cytometry using DNA from tumor FFPE samples was extracted either with Maxwell®16 FFPE the FACSAria II (BD Biosciences) and RFP+ GFP+ cells were sorted. Clones Plus LEV DNA Purification Kit (Promega Corp.) or QIAamp DNA FFPE Tissue were analyzed by WB and genotyped with forward 5’-tcagactgcaagtgg- Kit (QIAGEN). DNA from fresh or frozen tissues was isolated using DNeasy gaaaa-3’ and reverse 5’-ggtctttacaggcctctctgt-3’ primers. blood and tissue kit (QIAGEN). Exome sequencing procedure was performed as described in13. BRCA2 plasmids The BRCA2 variants c.1813dup (exon 10; forward 5’-gcctctgaaagtg Human BRCA2 cDNA (NM_000059) was cloned in the pcDNA3 vector gactgga-3’ and reverse 5’-aaaaacacagaaggaatcgtca-3’) and c.7796 A > G (Invitrogen). The two BRCA2 deleterious variants were introduced in the (exon 16; forward 5’-gtgtgatacatgtttactttaaattg-3’ and reverse 5’-gttcgaga pcDNA3-BRCA2 plasmid using the QuikChange II XL Site-Directed gacagttaagagaag-3’) were genotyped by PCR of genomic DNA and Sanger Mutagenesis Kit (Agilent) using for p.(Ile605Asnfs*11) variant forward 5’- sequencing. PCR products were purified using ExoSAP-IT PCR Product ctatacatgatgaaacatcttataaaggaaaaaaaataaccgaaagaccaaaaatca and reverse Cleanup Reagent (ThermoFisher), sequenced using BigDye Terminator v3.1 5’-tgatttttggtctttcggttattttttttcctttataagatgtttcatcatgtatag primers and for Cycle Sequencing kit (Applied Biosystems) and run in a Genetic Analyzer p.(Glu2599Gly) variant forward 5’-acacagagccctataaaatccttcttttccagcctttcc- ABI3130xl (Applied Biosystems). Reference transcript BRCA2 NM_000059.3 and reverse 5’-ggaaaggctggaaaagaaggattttatagggctctgtgt primers. was used for sequence alignment and variant annotation. Plasmid transfection RNA extraction, reverse transcription and cDNA amplification Transfections were carried out using 10 μL Lipofectamine 2000 (Life RNA was isolated from patient’s peripheral blood samples and cultured Technologies) and 250 μL Optimem (Life Technologies) plus 10 μg of each fibroblasts using Trizol Isolation Reagent (ThermoFisher Scientific). RNA pCDNA3 plasmid construct the day after plating 600000 HEK293T cells/well was cleaned up using Rneasy Minikit (QIAGEN) and digested with RNase- in 6-well plates with complete medium. Free DNase Set (QIAGEN) to remove genomic DNA. A total of 500 ng were reverse transcribed with PrimeScript RT reagent kit (Takara) that supplies random and oligo-dT primers. BRCA2 cDNA was amplified in two PCR Western blot antibodies fragments spanning exons 10 to 14 (forward 5’-tatgtccaaatttaattgataat-3’ Primary antibodies used were Rabbit anti-BRCA2 (Ab123491, Abcam) and and reverse 5’-ttggtctgcctgtagtaatc-3’) and 11 to 18 (forward 5’- mouse anti-vinculin (Ab18058, Abcam) were diluted 1:1000 and 1:2500, ttctgatgttcctgtgaaaacaa-3’ and reverse 5’-attctggggcttcaagaggt-3’), respec- respectively; and secondary antibodies, goat anti-rabbit of (Bethyl tively, using Expand™ Long Template PCR System (Roche). Sequencing was Laboratories, Inc.) and goat anti-mouse (Santa Cruz), both diluted 1:2000. performed with BigDye Terminator v3.1 Cycle Sequencing kit (Applied Biosystems) in a Genetic Analyzer ABI 3130xl (Applied Biosystems). Sanger HRR assay electropherograms were visualized using SeqScape v2.6 and Sequencing HRR assay was performed as in36 with some modifications: 200000 cells Analysis v2.6 softwares (Applied Biosystems). HEK293T expressing pDR-GFP construct in a stable manner were seeded in 12-well plates. pCDNA3 plasmids were co-expressed with an I-SceI DNA copy number analysis by multiplex ligation-dependent expressing plasmid by transient transfection. HR-dependent repair of probe assay (MLPA) I-SceI induced DNA double-strand breaks was quantified by fluorescence- Allelic copy number changes in BRCA2 were performed by MLPA kit #P090 activated cell sorting (FACS) of GFP + cells after 48 h. (MRC-Holland), and resulting electrophoresis on an ABI 3130xl capillary sequencer (Applied Biosystems). Analysis was done using Coffalyser.Net Generation of PDX models software (MRC-Holland, Amsterdam, The Netherlands). Blood samples from Biopsy specimens from metastatic lesions were immediately implanted healthy donors were included for sample normalization. into the lower flank of 5-week-old female NOD SCID GAMMA mice (NSG, Charles River) and then expanded in NMRI-Foxn1nu/nu mice (NMRI, Janvier Ethics approval Labs). This procedure was conducted following the European Union’s Patient blood and samples collection was carried out under patient animal care directive (2010/63/EU) and was approved by the Ethical Committee of Animal Experimentation of the Vall d’Hebron Research consent and the positive approval of the Clinical Research Ethics Institute. Committee of Vall d’Hebron Hospital (Barcelona, Spain), project number PR(AG)64/2007. Fresh primary or metastatic tumor samples were collected prospectively for implantation into nude mice at VHIO under an BRCA2 mRNA expression institutional review board (IRB)-approved protocol and the associated RNA was extracted from PDX samples (15–30 mg) by using the PerfectPure informed consent, project number PR(AG)173-2015. The experiments RNA Tissue kit (five Prime). cDNA was obtained using the PrimeScript RT conformed to the principles of the WMA Declaration of Helsinki and the Reagent kit (Takara). Quantitative RT-qPCR was performed in a 7900HT Fast Department of Health and Human Services Belmont Report. Real-Time PCR System (Applied Biosystems) using TaqMan Universal Animal experiments were conducted following the European Union’s Master Mix II (Applied Biosystems) and predesigned human-specific animal care directive (2010/63/EU) and the protocols were approved by primers and TaqMan probes (Hs99999908 for GUSB, Hs99999903 for the Ethical Committee of Animal Experimentation of the Vall d’Hebron ACTB, and Hs01037414 for BRCA2). The comparative CT method was used Research Institute and the appropriate governmental agency and carried for data analysis, in which geNorm algorithms were applied to select the out in accordance with the approved guidelines. Published in partnership with the Breast Cancer Research Foundation npj Breast Cancer (2021) 117 L. Castells-Roca et al. 8 Consent to participate 21. Richardson, M. E. et al. Strong functional data for pathogenicity or neutrality Written informed consent was obtained from the patient for publication of classify BRCA2 DNA-binding-domain variants of uncertain significance. Am. J. this case report. The Clinical Research Ethics Committee of Vall d’Hebron Hum. Genet. (2021) https://doi.org/10.1016/j.ajhg.2021.02.005. Hospital (Barcelona, Spain) approved the study. 22. Oostra, A. B., Nieuwint, A. W. M., Joenje, H. & De Winter, J. P. Diagnosis of fanconi anemia: chromosomal breakage analysis. Anemia 2012, 238731 (2012). 23. Pinto, F. O. et al. Diagnosis of fanconi anemia in patients with bone marrow Reporting summary failure. Haematologica 94, 487–495 (2009). Further information on research design is available in the Nature Research 24. Yang, H. et al. BRCA2 function in DNA binding and recombination from a BRCA2- Reporting Summary linked to this article. DSS1-ssDNA Structure. Science 297, 1837–1848 (2002). 25. Zhao, W. et al. Promotion of BRCA2-Dependent Homologous Recombination by DSS1 via RPA Targeting and DNA Mimicry. Mol. Cell (2015) https://doi.org/ DATA AVAILABILITY 10.1016/j.molcel.2015.05.032. The datasets generated and/or analysed during the current study are available from 26. Bhatia, V. et al. BRCA2 prevents R-loop accumulation and associates with TREX-2 the corresponding author on reasonable request. mRNA export factor PCID2. Nature 511, 362–365 (2014). 27. Schwab, R. A. et al. The fanconi anemia pathway maintains genome stability by coordinating replication and transcription. Mol. Cell 60, 351–361 (2015). Received: 13 August 2020; Accepted: 2 August 2021; 28. Shimelis, H. et al. BRCA2 hypomorphic missense variants confer moderate risks of breast cancer. Cancer Res. 77, 2789–2799 (2017). 29. Degrolard-Courcet, E. et al. Development of primary early-onset colorectal can- cers due to biallelic mutations of the FANCD1/BRCA2 gene. Eur. J. Hum. Genet. 22, 979–987 (2014). REFERENCES 30. Gay-Bellile, M. et al. Is BRCA2 involved in early onset colorectal cancer risk? 1. Nalepa, G. & Clapp, D. W. Fanconi anaemia and cancer: an intricate relationship. Clinical Genetics (2020) https://doi.org/10.1111/cge.13679. Nat. Rev. Cancer 18, 168–185 (2018). 31. Easton, D. F. et al. A systematic genetic assessment of 1,433 sequence variants of 2. Bogliolo, M. & Surrallés, J. Fanconi anemia: a model disease for studies on human unknown clinical significance in the BRCA1 and BRCA2 breast cancer- genetics and advanced therapeutics. Curr. Opin. Genet. Dev. 33, 32–40 (2015). predisposition genes. Am. J. Hum. Genet. (2007) https://doi.org/10.1086/521032. 3. Chen, C. C., Feng, W., Lim, P. X., Kass, E. M. & Jasin, M. Homology-directed repair 32. Bogliolo, M. et al. Mutations in ERCC4, encoding the DNA-repair endonuclease and the role of BRCA1, BRCA2, and related proteins in genome integrity and XPF, cause Fanconi anemia. Am. J. Hum. Genet. 92, 800–806 (2013). cancer. Annu. Rev. Cancer Biol. 2, 313–336 (2018). 33. Castroviejo‐Bermejo, M. et al. A RAD 51 assay feasible in routine tumor samples 4. Prakash, R., Zhang, Y., Feng, W. & Jasin, M. Homologous recombination and calls PARP inhibitor response beyond BRCA mutation. EMBO Mol. Med. (2018) human health: The roles of BRCA1, BRCA2, and associated proteins. Cold Spring https://doi.org/10.15252/emmm.201809172. Harb. Perspect. Biol. (2015) https://doi.org/10.1101/cshperspect.a016600. 34. Cruz, C. et al. RAD51 foci as a functional biomarker of homologous recombination 5. Howlett, N. G. et al. Biallelic inactivation of BRCA2 in Fanconi anemia. Science 297, repair and PARP inhibitor resistance in germline BRCA-mutated breast cancer. 606–609 (2002). Ann. Oncol. (2018) https://doi.org/10.1093/annonc/mdy099. 6. Alter, B. P., Rosenberg, P. S. & Brody, L. C. Clinical and molecular features asso- 35. Cong, L. et al. Multiplex genome engineering using CRISPR/Cas systems. Science ciated with biallelic mutations in FANCD1/BRCA2. J. Med. Genet. 44, 1–9 (2006). 339, 819–823 (2013). 7. Myers, K. et al. The clinical phenotype of children with Fanconi anemia caused by 36. Hernández, G. et al. Decapping protein EDC4 regulates DNA repair and pheno- biallelic FANCD1/BRCA2 mutations. Pediatr. Blood Cancer 58, 462–465 (2012). copies BRCA1. Nat. Commun. (2018) https://doi.org/10.1038/s41467-018-03433-3. 8. Wagner, J. E. et al. Germline mutations in BRCA2: shared genetic susceptibility to 37. Vandesompele, J. et al. Accurate normalization of real-time quantitative RT-PCR breast cancer, early onset leukemia, and Fanconi anemia. Blood 103, 3226–3229 data by geometric averaging of multiple internal control genes. Genome Biol. 3, (2004). RESEARCH0034 (2002). 9. Meyer, S. et al. Fanconi anaemia, BRCA2 mutations and childhood cancer: a 38. Padilla, N. et al. BRCA1- and BRCA2-specific in silico tools for variant interpreta- developmental perspective from clinical and epidemiological observations with tion in the CAGI 5 ENIGMA challenge. Hum. Mutat. 40, 1593–1611 (2019). implications for genetic counselling. J. Med. Genet. 51, 71–75 (2014). 10. Pierce, A. J., Johnson, R. D., Thompson, L. H. & Jasin, M. XRCC3 promotes homology-directed repair of DNA damage in mammalian cells. Genes Dev. 13, ACKNOWLEDGEMENTS 2633–2638 (1999). The authors thank the patient and her family. The authors thank Dr. Detlev Schindler 11. Castella, M. et al. Chromosome fragility in patients with Fanconi anaemia: diag- for providing the primary BRCA2-defective fibroblast cell line (c.469 A > T, p.Lys157*). nostic implications and clinical impact. J. Med. Genet. 48, 242–250 (2011). Surrallés laboratory is supported by ICREA-Academia programme, Spanish Ministry of 12. Nicoletti, E. et al. Mosaicism in Fanconi anemia: concise review and evaluation of Health (projects FANCOSTEM and FANCOLEN), Spanish Ministry of Science, published cases with focus on clinical course of blood count normalization. Ann. Innovation, and Universities (projects CB06/07/0023 MCI, UE and RTI2018-098419- Hematol. 99, 913–924 (2020). B-I00 MCI, UE), European Commission (EUROFANCOLEN project HEALTH-F5-2012- 13. Bogliolo, M. et al. Optimised molecular genetic diagnostics of Fanconi anaemia 305421), Fanconi Anemia Research Fund Inc and the’Fondo Europeo de Desarrollo by whole exome sequencing and functional studies. J. Med. Genet. 57, 258–268 Regional, una manera de hacer Europa’ (FEDER). This work was also funded by (2020). intramural actions (ACCI-U705/U745-2018 to JS and AL) of CIBERER. CIBERER is an 14. Rahman, N. & Stratton, M. R. The genetics of breast cancer susceptibility. Annu. initiative of the Instituto de Salud Carlos III, Spain. The VHIO authors acknowledge the Rev. Genet. (1998) https://doi.org/10.1146/annurev.genet.32.1.95. Cellex Foundation for providing research facilities and thank CERCA Programme/ 15. Richards, S. et al. Standards and guidelines for the interpretation of sequence Generalitat de Catalunya for institutional support. Part of this work was supported by variants: a joint consensus recommendation of the american college of medical the Spanish Instituto de Salud Carlos III (ISCIII) funding (PI16/01218 and PI19/01303 to genetics and genomics and the association for molecular pathology. Genet. Med. SG-E, PI12/02606 to JB and PI17/01080 to VS), an initiative of the Spanish Ministry of 17, 405–424 (2015). Economy and Innovation partially supported by European Regional Development 16. Soulier, J. et al. Detection of somatic mosaicism and classification of Fanconi FEDER Funds. SG-E and VS are supported by the ISCIII Miguel Servet Program (CP16/ anemia patients by analysis of the FA/BRCA pathway. Blood (2005) https://doi. 00034 and CPII19/00033). SB is a recipient of Asociación Española Contra el Cáncer org/10.1182/blood-2004-05-1852. (AECC) fellowship. ALG is a recipient of a PERIS fellowship (Generalitat de Catalunya) 17. Sharan, S. K. et al. Embryonic lethality and radiation hypersensitivity mediated by and innovation funds from the CaixaImpulse program. VS is the recipient of an Rad51 in mice lacking Brca2. Nature 386, 804–810 (1997). 18. Matsuda, D., Sato, H. & Maquat, L. E. Chapter 9 Studying Nonsense-Mediated AGAUR (2017 SGR 540), an AECC (LABAE16020PORTT), and an ERAPERMED2019-215 mRNA Decay in Mammalian Cells. in Methods Enzymology (2008). https://doi.org/ grant. LC-R contract is supported by the European Union’s Horizon 2020 research and 10.1016/S0076-6879(08)02409-9. innovation programme under Marie Skłodowska-Curie grant agreement No 665919. 19. Mesman, R. L. S. et al. The functional impact of variants of uncertain significance in BRCA2. Genet. Med. (2019) https://doi.org/10.1038/s41436-018-0052-2. 20. Biswas, K. et al. A comprehensive functional characterization of BRCA2 variants AUTHOR CONTRIBUTIONS associated with Fanconi anemia using mouse ES cell-based assay. Blood 118, JB and JS conceived this manuscript. LC-R and S G-E contributed equally to this work. 2430–2442 (2011). LC-R, MA-C, RP, NM-S, and MJR performed and analysed PARPi survivals, western blot, immunostainings, DR-GFP, and cell cycle experiments with fibroblasts. MB completed npj Breast Cancer (2021) 117 Published in partnership with the Breast Cancer Research Foundation L. Castells-Roca et al. 9 Exome Sequencing (ES) experiments. MA-C established the BRCA2 KO HEK293T cell Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims line. RP analyzed and interpreted the patient data regarding chromosome fragility in published maps and institutional affiliations. tests (CFT) in blood T-lymphocytes and fibroblasts. SG-E, SB, EC, GM, CC, AL-G, CS, AL, VS, OD contributed with patient samples, experimental data, and animal experiments. LC-R, SG-E, and MB were major contributors in writing the paper. All authors read and approved the final paper. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give COMPETING INTERESTS appropriate credit to the original author(s) and the source, provide a link to the Creative The authors declare no competing interests. Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory ADDITIONAL INFORMATION regulation or exceeds the permitted use, you will need to obtain permission directly Supplementary information The online version contains supplementary material from the copyright holder. To view a copy of this license, visit http://creativecommons. available at https://doi.org/10.1038/s41523-021-00322-9. org/licenses/by/4.0/. Correspondence and requests for materials should be addressed to Judith. Balmaña or Jordi. Surrallés © The Author(s) 2021 Reprints and permission information is available at http://www.nature.com/ reprints Published in partnership with the Breast Cancer Research Foundation npj Breast Cancer (2021) 117