Clinical consequences of BRCA2 hypomorphism

Abstract

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.

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.

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).

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.

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.