|Home | About | Journals | Submit | Contact Us | Français|
Breast cancer risk drastically increases in individuals with a heterozygous germline BRCA1 or BRCA2 mutation, while it is estimated to equal the population risk for relatives without the familial mutation (non-carriers). The aim of the present study was to use a G2 phase-specific micronucleus assay to investigate whether lymphocytes of healthy BRCA2 mutation carriers are characterized by increased radiosensitivity compared to controls without a family history of breast/ovarian cancer and how this relates to healthy non-carrier relatives. BRCA2 is active in homologous recombination, a DNA damage repair pathway, specifically active in the late S/G2 phase of the cell cycle. We found a significantly increased radiosensitivity in a cohort of healthy BRCA2 mutation carriers compared to individuals without a familial history of breast cancer (P=0.046; Mann-Whitney U test). At the individual level, 50% of healthy BRCA2 mutation carriers showed a radiosensitive phenotype (radiosensitivity score of 1 or 2), whereas 83% of the controls showed no radiosensitivity (P=0.038; one-tailed Fishers exact test). An odds ratio of 5 (95% CI, 1.07–23.47) indicated an association between the BRCA2 mutation and radiosensitivity in healthy mutation carriers. These results indicate the need for the gentle use of ionizing radiation for either diagnostic or therapeutic use in BRCA2 mutation carriers. We detected no increased radiosensitivity in the non-carrier relatives.
BRCA1 and BRCA2 heterozygous mutation carriers have a strongly increased risk to develop breast cancer (BC) and ovarian cancer (OC). The lifetime risk to develop BC is 70–80% for BRCA1 mutation carriers and 50–60% for BRCA2 mutation carriers (1). For relatives who did not inherit the germline BRCA1/2 mutation segregating in the family (non-carrier relatives), the risk of BC occurrence is generally estimated to be as low as the risk assessed in the general population. This may imply that intensified BC detection screening, using, amongst others, mammography screening and MRI, as applied in individuals at high-risk is unnecessary in non-carriers (2–7). However, one study reported a 2–5-fold increase in BC occurrence in non-carriers of families with either BRCA1 or BRCA2 mutations (8). Another study reported a younger than expected age at diagnosis of BC for non-carriers, that was most evident in BRCA1 families (9). Moreover, one study by Evans et al detected a possible higher relative risk for BC in non-carrier relatives of BRCA2 families, compared to non-carriers in BRCA1 families (10). In summary, these studies suggest that DNA alterations (for example SNPs) in other genes may modify the relative risk for the development of BC in non-carriers, compared to the general population. Moreover, the authors of these studies recommend targeted BC detection screening using for example mammography in non-carriers at a frequency comparable to the intensive BC screening performed in individuals at high-risk.
Both BRCA1 and BRCA2 are caretaker genes playing different roles in the repair of DNA double-strand breaks (DSB), induced by exposure to genotoxic agents such as ionizing radiation (IR). While BRCA1 has a more general function in the detection and signaling of a DSB and in the activation of the G2/M cell cycle checkpoint, BRCA2 exerts a specific function in the recruitment of RAD51 recombinase to the DSB site. This latter event is essential for the activation of the homologous recombination (HR) pathway, that relies on the undamaged sister chromatid as a template for resynthesis of the damaged strand. This occurs in the late S and G2 phase of the cell cycle and leads to error-free repair of DSB (1).
Knowing that both BRCA1 and BRCA2 are important in the repair of DSB, exposure of mutation carriers to IR, a potent inducer of DSB, for either diagnostic or therapeutic purposes appears to be counterintuitive, as mutation carriers may be more prone to develop radiation-induced BC (11).
Radiosensitivity of BRCA1 mutation carriers has previously been reported in the literature and was investigated and confirmed by our research group by means of the G2 micronucleus (MN) assay in combination with an evaluation of the G2/M checkpoint efficiency in peripheral blood lymphocytes of healthy BRCA1 mutation carriers compared to healthy volunteers (12). However, the impact of IR on heterozygous cells of healthy BRCA2 mutation carriers remains to be elucidated.
To date, several cohort studies were able to prove a positive correlation between exposure to diagnostic X-rays and BC risk in BRCA2 mutation carriers (11,13,14). Others however, could not detect a similar correlation (15–18). Furthermore, Bernstein et al detected no increased induction of contralateral BC upon exposure to radiotherapy in BRCA2 mutation carriers (19). Such discrepancies are likely due to differences in inclusion criteria, data acquisition and other issues of the studies. It is however difficult and unethical to design long-term unbiased studies to evaluate the relationship between BRCA2 mutations, the exposure to diagnostic or therapeutic radiation and BC risk. In vitro chromosomal assays are effective tools to investigate radiosensitivity. Chromosomal radiosensitivity testing on lymphocytes from BRCA2 mutation carriers has been performed with techniques such as the G0 MN and the G2 assays for chromatid breaks, occasionally enhanced with a whole-chromosome painting FISH (20–25). However, for several of these studies, it was unclear whether the BRCA2 heterozygotes were healthy individuals or BC patients, which was previously broached by Baeyens et al (20). Furthermore, differences in the experimental setup make comparisons between studies difficult (26). Despite these differences, all but one study was able to detect an elevated chromosomal radiosensitivity in BC patients with a BRCA2 mutation. However, no comparison was made with sporadic BC patients. The study of Baeyens et al previously demonstrated enhanced radiosensitivity in both BC patients with a BRCA1/2 mutation and sporadic BC patients, suggesting that the enhanced sensitivity may not be the result of the mutation (20). No univocal results were achieved for healthy BRCA2 mutation carriers. Radiosensitivity in non-carrier relatives has not been studied extensively, only one study reported no increased radiosensitivity measured with the G0 MN and G2 chromatid break assay in a small cohort (n=10) of relatives of both BRCA1 and BRCA2 families without the familial mutation when compared to a population cohort (20).
In the present study, we aimed to investigate chromosomal radiosensitivity in healthy BRCA2 mutation carriers by means of the G2 MN assay. We previously used this assay and confirmed radiosensitivity in healthy BRCA1 mutation carriers (n=18) compared to healthy controls without a family history of BC or OC (n=20) (12), and in an ataxia-telangiectasia patient and family members (27). In addition, we also included healthy relatives not carrying the familial germline BRCA1 or BRCA2 mutation in the present study. This cohort of non-carriers was included to evaluate radiosensitivity in individuals with a comparable genetic background, but without the familial BRCA1 or BRCA2 mutation.
Blood samples were collected from individuals consulting the Centre for Medical Genetics of the Ghent University Hospital (CMG; Ghent, Belgium), in the context of predictive testing for hereditary BC. Heparin blood samples were collected for the G2 MN assay. In addition, EDTA samples were collected for mutation analysis. We collected blood samples from 18 BRCA2 mutation carriers and 17 subjects from both BRCA1 (n=9) and BRCA2 (n=8) families not showing the familial mutation (non-carriers). None of the individuals selected for the present study had developed cancer at the time of the blood sample collection. We also selected 18 blood samples from a historical cohort of healthy volunteers without a personal or familial history of BC or OC for optimal age and gender match, to determine the normal distribution of MN yields in unaffected individuals from the general population (12).
The present study was approved by the Ethics Committee of Ghent University Hospital (B67020111641 d.d. 20/09/2011) and all participants signed an informed consent.
All healthy individuals selected for the present study had a family history of BC or OC and a mutation in either BRCA1 or BRCA2 was identified in each proband. All BRCA2 mutation carriers are heterozygous for an unequivocal deleterious mutation. This was confirmed by Sanger sequencing of the relevant amplicon. Sanger sequencing was performed on the ABI3730XL instrument using the BigDye® Terminator Cycle Sequencing kit (Life Technologies, Carlsbad, CA, USA) according to the manufacturer's instructions; sequences were analyzed using the SeqPilot software (JSI Medical Systems GmbH, Ettenheim, Germany).
Molecular analyses were not performed in healthy volunteers due to the absence of a personal or familial anamnesis for BC or OC.
The G2 MN assay was performed as previously described (12). In brief, heparinized blood was cultured in the presence of phytohaemagglutinin (PHA; 2% v/v; Gibco, Grand Island, NY, USA) to stimulate T-lymphocyte division. After 3 days, a population of cycling lymphocytes was obtained and the culture was irradiated with a 2 Gy dose of 60Co γ-rays. We opted to use a dose of 2 Gy as this is a well-accepted dose for chromosomal radiosensitivity testing in lymphocytes (20–22,24). Immediately after irradiation, cytochalasin B (cyto B; 6 µg/ml; Sigma-Aldrich, St. Louis, MO, USA) was added to all cultures, including a non-irradiated culture. Cyto B blocks the cytokinesis and allows the identification of first-division cells as a binucleated (BN) cell. After an incubation period of 8 h, all cultures were fixed with the sequential addition of KCl (75 mM), a solution of methanol, acetic acid and Ringer (4:1:5), and a combination of methanol and acetic acid (4:1) to pelleted cells. Finally, the cell suspension was concentrated and spread on slides. Slides were stained with 4′,6-diamidino-2-phenlylindole (DAPI) and scanned with a Metafer 4 platform and MN search software (MetaSystems GmbH, Altlussheim, Germany). The automated image analysis system selects BN cells and determines the number of MN/BN cells. BN cells and MN selection are manually checked for false positives or false negatives. For each condition, 2 cultures were prepared and 2 slides/culture were analyzed. A minimum of 600 BN cells were scored/coded slide. To assess individual radiosensitivity, a radiosensitivity score (RS score) was determined. The mean and SD of the MN yield of the group of healthy volunteers (HV) was set as the cut-off value to determine the RS score of HV, BRCA2 mutation carriers and non-carrier relatives. An MN yield higher than the meanHV + 1SDHV was scored as 1, indicating a milder radiosensitive phenotype, whereas a result higher than the meanHV + 2SDHV was scored as 2, and indicated a more severe radiosensitive phenotype. When the individual value was lower than the meanHV + 1SDHV, a score of 0 was attributed to the tested subject.
Age and gender differences among the 3 groups were judged by means of a one-way ANOVA and Chi-square test, respectively. The median, interquartile range, average and standard deviation of micronuclei yields (number of MN/1,000 BN cells) were assessed in each group of subjects. Intergroup differences of MN yields between HV, BRCA2 mutation carriers and non-carrier relatives of BRCA1 and BRCA2 pedigrees were analyzed by the Mann-Whitney-Wilcoxon test. A one-tailed Fishers exact test was performed to compare the unpaired and independent proportion of patients showing a radiosensitive phenotype, evaluated by RS scoring. For both assays a 5% α error was set as the limit for statistical significance. The odds ratio (OR) was calculated, based on the RS scores in healthy individuals and BRCA2 mutation carriers, to assess the association between the presence of a BRCA2 mutation and radiosensitivity according to the following formula:
The 95% confidence interval (CI) was used as a proxy for significance. The VassarStats platform and the SPSS software (IBM, version 23) were used to perform statistical analysis.
The mean age did not significantly differ for the HV (35.3 years), the BRCA2 mutation carriers (40.9 years) and the non-carrier relatives (40.0 years) (P=0.56; one-way ANOVA). In addition, no significant difference in gender distribution was observed for these 3 groups (68, 61 and 71% of the individuals were female, respectively) (P=0.84; Chi-square test). The number of spontaneously occurring micronuclei (MN yields in non-irradiated samples) was not significantly different among the 3 groups of enrolled subjects (Table I and Fig. 1).
Compared to HV without a family history of BC/OC, BRCA2 mutation carriers showed a significant increase in mean MN yields after exposure to 2 Gy IR (P=0.046; Mann-Whitney). Conversely, the radiation-induced MN yields were similar in relatives who did not inherit the familial BRCA1/2 mutation and HV without a family history of BC/OC. The mean MN yield in BRCA2 mutation carriers was higher compared to the mean yield in non-carriers (86.11 vs. 68.11 MN/1,000 BN cells, respectively). This difference, however, was not significant (P=0.298; Mann-Whitney), probably due to the small cohort and the high SD (Table I and Fig. 1). Furthermore, MN yields did not differ between non-carrier relatives from BRCA1 or BRCA2 pedigrees (Table II).
The individual MN yields after exposure to 2 Gy and the RS score for each BRCA2 mutation carrier, non-carrier relative and healthy volunteer group are listed in Table III. Furthermore, Table III shows mutational data (both nucleotide and protein nomenclature) and individuals with the same family ID are related. Fig. 2 shows the distribution of the 3 groups for the different RS scores. A significantly higher number of BRCA2 mutation carriers (n=9/18; 50%) showed increased RS scores (score 1 or 2) compared to HV (n=3/18; 17%) (P=0.038; one-tailed Fishers exact test). For the relatives who did not inherit the familial germline mutation only 24% (n=4/17) showed an elevated radiosensitivity at the individual level. RS scoring in related individuals (see family ID in Table III) however shows some variation. An OR of 5 (95% CI, 1.07–23.46) for BRCA2 mutation carriers vs. HV, indicates a significant association between the presence of a BRCA2 mutation and radiosensitivity according to our criteria.
All but one of the 18 mutation carriers enrolled in the present study were heterozygous for a mutation predicted to result in a premature termination codon (PTC). The patient with the deleterious missense mutation [BRCA2 c.8167G>C; p.(Asp2723His)] obtained an RS score of 2.
Results of the G2 micronucleus (MN) assay performed after exposure to 2 Gy γ-rays showed a significantly increased radiosensitivity in healthy BRCA2 mutation carriers compared to healthy controls. Previous studies with a large number of different techniques were able to demonstrate enhanced radiosensitivity in BC patients with a BRCA2 mutation, however, no univocal results were achieved for healthy BRCA2 mutation carriers (20–25). Non-carrier relatives of either BRCA1 or BRCA2 families did not show an increased radiosensitive phenotype compared to the cohort of healthy volunteers, which is in agreement with the study of Baeyens et al (20). We previously performed the G2 MN assay in a group of 18 healthy BRCA1 mutation carriers, and found a significantly increased MN yield after exposure to 2 Gy γ-rays (12). These findings are analogous to the results of the present study, performed in healthy carriers of pathogenic BRCA2 mutations. Fig. 3 shows the integration of the data from healthy BRCA1 mutation carriers in the present study. The detection of an increased mean MN yield in both BRCA1 and BRCA2 mutation carriers after exposure to ionizing radiation can be explained by their mutual role in DNA double-strand break repair reviewed by Roy et al (1).
In our previous study we also analyzed the G2/M checkpoint activity by the addition of caffeine, an agent abrogating the G2/M checkpoint, to the irradiated cultures and demonstrated a significantly impaired checkpoint activation in BRCA1 mutation carriers compared to healthy volunteers (12). Analysis of the G2/M checkpoint activation in the current BRCA2 cohort did not reveal a significant difference (data not shown). This result is in agreement with the fact that BRCA2 is not activated in this particular checkpoint pathway as reviewed by Roy et al (1), but does not support the data obtained by Menzel et al (28), suggesting a role for BRCA2 as a regulator of G2 checkpoint maintenance following DNA damage introduced in a human osteosarcoma cell line (U2OS) expressing dominant-negative p53 by a high-dose of ionizing radiation (6 Gy).
The role of BRCA2 in the HR pathway, a DNA repair pathway active in the S and G2 phase of the cell cycle, is extensively reported in literature (1). The present study, focusing on radiosensitivity testing of lymphocytes in these phases of the cell cycle, showed an OR of 5 (95% CI, 1.07–23.47) for healthy individuals with a heterozygous BRCA2 mutation compared to healthy controls. This indicates a positive association between the presence of a BRCA2 mutation and radiosensitivity that could be attributed to deficient HR capacity in heterozygous cells.
Two independent research groups have reported a decreased DSB repair capacity in BRCA2 heterozygous cells. Keimling et al used an enhanced green fluorescent protein (EGFP)-based assay to report impaired HR capacity in lymphoblastoid cells with a BRCA2 monoallelic truncating frameshift mutation. They confirmed this decrease in HR capacity in a BRCA2-knockdown HeLa cell line (29). Arnold et al demonstrated distinct defects in DNA DSB repair in lymphoblastoid cell lines (LCLs) from heterozygous BRCA2 mutation carriers through analysis of γ-H2AX repair kinetics (30). Although, the latter study did not focus on DNA repair by HR, it indicates a malfunction of DSB repair in LCLs from BRCA2 mutation carriers that could be attributed to diminished HR activity.
Most mutation carriers enrolled in the present study (n=17/18, 94%) had a mutation resulting in a premature termination codon (PTC). The presence of a PTC mutation is expected to activate nonsense-mediated decay of the gene transcript. Previous research from various groups including ours, demonstrated a reduction in mutant mRNA to approximately half of the WT mRNA levels in lymphocytes of individuals with a PTC mutation in BRCA1 (12,31,32). Arnold et al (30) detected a similar mutant mRNA reduction for BRCA2 mutations leading to a PTC. Furthermore, Arnold et al (30) and Keimling et al (29) report distinct reduced protein levels in LCLs from heterozygous BRCA2 mutation carriers, although quantitative analysis of this variation was not performed. Previously, haploinsufficiency has been suggested as the mechanism for hereditary BC development in BRCA1 and BRCA2 mutation carriers (33). In the present study, a higher than expected number of radiosensitive individuals in the BRCA2 mutation carriers indicates that haploinsufficiency may also be responsible for the radiosensitive phenotype in carriers of a mutation generating a PTC. In the present study, only one individual with a deleterious missense mutation was included. This substitution results in an amino acid change at position p.2723 and impairs protein functionality as shown by a homology-directed DNA break-repair functional assay (34). For this individual we obtained a high RS score of 2. Further research in larger patient cohorts with different types of mutations is needed to evaluate whether the type of mutation influences the radiosensitive phenotype or whether there are additional parameters determining this phenotype.
Results of the G2 MN assay showed no increased radiosensitivity in the group of non-carrier relatives of both BRCA1 and BRCA2 families compared to a group of healthy volunteers. Furthermore, only 24% of non-carriers showed an elevated radiosensitivity at the individual level (RS score 1 or 2). This was not significantly different from the fraction of healthy volunteers (17%) that was found to have an increased RS score. In addition, no difference was observed between non-carriers from BRCA1 (RS score, 0 in 7/9 investigated relatives) or BRCA2 families (RS score, 0 in 6/8 investigated relatives). However, we observed some variation within the different groups. We hypothesize that modifiers may play a role: indeed, selected SNPs in DNA-damage repair genes and other common variants have been associated with increased radiosensitivity (35–37) and increased BC risk (35,38). Further and larger studies are needed to evaluate the subtle influence of possible modifying factors on BC risk and radiosensitivity.
In conclusion, the present study demonstrated higher radiosensitivity in healthy BRCA2 mutation carriers compared to healthy volunteers by means of the G2 MN assay after exposure of peripheral blood lymphocytes to a dose of 2 Gγ-rays. No increased radiosensitivity was observed in non-carrier relatives of BRCA1 and BRCA2 families. When evaluating radiosensitivity at the individual level, a significantly higher proportion of BRCA2 mutation carriers (50%) showed a mild or more severe radiosensitivity compared to healthy volunteers (17%) and non-carriers (24%). Furthermore, an OR of 5 indicated a positive association between the BRCA2 mutation and an increased radiosensitivity in healthy mutation carriers. These results indicate that care should be taken when applying ionizing radiation for either diagnostic or therapeutic purposes in BRCA2 mutation carriers. However, a study including a larger population of subjects carrying different types of BRCA2 mutations and non-carriers, must be performed to further elucidate the effect of each single mutation on the radiosensitive phenotype and the influence of possible underlying factors.
The authors would like to thank all participants who donated a blood sample for the present study. They thank Céline De Brock, Brecht Crombez, Ilse Coene, Johanna Aernoudt, Toke Thiron, Greet De Smet and Leen Pieters for their technical assistance. Professor Thierens is thanked for the use of the irradiation facility. The present study was funded by the Belgian Foundation against Cancer (project 2012–216).