Since the discovery of the tumour suppressor BRCA2 (encoded by breast-cancer susceptibility gene 2), cells lacking the fully functional protein have consistently been found to show increased sensitivity to a variety of DNA-damaging agents, particularly those that cross-link DNA. In this short review, we will bring together these findings and discuss them in the light of our recent in vivo data in the mouse small intestine, which suggests that deletion of cells lacking Brca2 is necessary to avoid the development of potentially tumorigenic clones in this tissue, a system that may be less effective in the mammary glands of humans with germline mutations in BRCA2.
- cross-linking agents
- DNA damage
- small intestine
Inheritance of a mutated copy of the BRCA2 (breast-cancer susceptibility 2) gene predisposes carriers to a variety of cancers in humans, most notably breast and ovarian tumours . BRCA2 is a large protein whose primary role is thought to be involvement in homologous recombination, a process that repairs DNA double-strand breaks in an error-free manner, through interaction with the RAD51 protein . This occurs via a number of repeated motifs (BRC repeats) in BRCA2, the majority of which are encoded within exon 11 of the gene . Cells lacking BRCA2 repair DNA by the error-prone mechanism of non-homologous end-joining (NHEJ), leading to chromosomal rearrangements and genomic instability . Many types of BRCA2-deficient cells exhibit increased sensitivity to DNA-damaging agents, a feature that we will now discuss in some detail, before concluding with our recent in vivo data in Brca2-deficient-mouse small intestine (SI).
Homozygosity for Brca2
In one of the original papers describing the embryonic death caused by homozygosity for a Brca2 allele in which exon 11 had been mutated, the inner-cell mass of 3.5-day-old embryos grown in culture was almost completely destroyed, and the number of trophoblast cells significantly reduced, following exposure to γ-irradiation in homozygous mutants compared with wild-type and heterozygous controls . At around the same time, another strain of mice was created in which the Brca2 gene was mutated within exon 11, such that the resulting protein was truncated at approx. 45% of its full length . Owing to the partial activity of this truncated protein, approx. 10% of mice survived to adulthood . MEFs (mouse embryonic fibroblasts) derived from these mice also showed increased sensitivity to γ-irradiation, as well as to the alkylating agent MMS (methyl methanesulphonate) and UV light, by a cell death assay . This suggested that, in addition to a role in homologous recombination, Brca2 may also be involved in the process of BER (base excision repair), a notion that has since been suggested elsewhere . Interestingly, these cells did not exhibit increased sensitivity to MMC (mitomycin C) at a range of concentrations, although it was noted that the same homozygous truncation of Brca2 in lymphoid cells did lead to hypersensitivity to MMC, suggesting cell-type-specific variation in sensitivity to this drug .
A number of C-terminal truncations have been designed in the Brca2 protein. Exon 27 is the last in the Brca2 gene and encodes the region containing the most C-terminal RAD51-interacting domain . ES (embryonic stem) cells were created in which exon 27 was deleted, and these were found to be hypersensitive to γ-irradiation but not to UV light, in a colony-forming assay . Immortalized MEFs from mice derived from these ES cells were also shown to be hypersensitive to γ-irradiation, as well as to MMC, again by colony-forming assay . The apparent contradiction in the response to UV light between this study  and that of Patel et al.  may be explained by one of several reasons, including differences in experimental design (survival compared with proliferation) or cell type (MEFs compared with ES cells), or, more likely, the possibility that the region deleted in the larger truncation is involved in the response to UV, by repairing pyrimidine dimers in BER .
Capan-1 is a human pancreatic carcinoma cell line which lacks one copy of the BRCA2 gene and carries the 6174delT mutation in the other allele . This allele is found in the Ashkenazi Jewish population and is the most common cancer-causing BRCA2 allele, once again because of a truncation which ablates the C-terminal half of the protein, thereby disrupting binding to RAD51 . Capan-1 cells are unable to repair double-strand breaks induced by treatment with ionizing radiation and, using a colony-forming assay, were shown to exhibit hypersensitivity to ionizing radiation as well as to three drugs which induce DNA strand breaks (etoposide, amsacrine and mitoxantrone), but not to two which do not (hydroxyurea and paclitaxel) . Capan-1 cells have also been shown to exhibit increased sensitivity to MMS , again presumably because there is a larger truncation of the BRCA2 protein in these cells, which interferes with the process of BER . This hypersensitivity to MMS was rescued in Capan-1 cells when a full-length wild-type BRCA2 cDNA was exogenously expressed via lipofection .
Another Brca2-deficient cell type that has been extensively studied is the Chinese-hamster mutant V-C8. These cells have a complex phenotype, characterized by extreme sensitivity to cross-linking agents (such as MMC), UV light, MMS and ionizing radiation . Once again, ectopic expression of Brca2 [either via an intact human chromosome 13 or a BAC (bacterial artificial chromosome) containing the murine Brca2 gene] complemented the hypersensitivity phenotype exhibited in these cells . Although the nature of the mutation was not mentioned in this study, it is likely that any Brca2 protein present in these cells is missing at least the C-terminal half, thereby allowing cross-sensitivity to a variety of agents, as suggested by previous studies [4,9]. A follow-up study revealed that V-C8 cells were also hypersensitive to camptothecin, an inhibitor of topoisomerase-I, and that once again this phenotype could be complemented by exogenous Brca2 .
Heterozygosity for Brca2
A hypersensitivity phenotype has also been seen in the context of heterozygosity for Brca2. Using a clonogenic survival assay, Foray et al.  showed that immortalized lymphoblastoid cell lines from BRCA2 germline-mutation carriers exhibited a slightly increased sensitivity to γ-irradiation. In addition, disruption of the Brca2 gene within exon 11, thereby disrupting RAD51 interaction, in a chicken B-cell line, resulted in hypersensitivity to MMC and cisplatin in heterozygous cells, as demonstrated by a clonogenic survival assay . Interestingly, sensitivity to MMS, UV and X-rays in these cells was unaffected, suggesting that only hypersensitivity to cross-linking agents such as MMC may be dose-dependent .
Role of BRCA2 in the FA (Fanconi anaemia) pathway
BRCA2 has recently been shown to be involved in the FA pathway, although exactly how it fits into the pathway is still a little unclear . BRCA2 is thought to be the FANCD1 protein, which is known to function downstream of, or parallel to, the main pathway, owing to the ability of cells in this complementation group to mono-ubiquitinate FANCD2 . One of the strongest pieces of evidence that BRCA2 is indeed FANCD1, is that the full BRCA2 cDNA can complement hypersensitivity to agents such as MMC and cisplatin in a number of FANCD1 patient cell lines . Interestingly, hypersensitivity to cross-linking agents, but not other types of DNA damage, such as UV and ionizing radiation, is a hallmark of FA cell lines .
In vivo hypersensitivity in the mouse SI
In the light of this wealth of data, we set out to examine the sensitivity of Brca2-deficient cells in an in vivo system, namely the mouse SI. Using a CYP-1A1-driven Cre-LoxP approach, we deleted exons 9 and 10 of the Brca2 gene within SI crypt cells, thereby leaving a truncated protein which is unable to bind RAD51 . Using a direct method of counting histological apoptosis, we demonstrated increased rates of spontaneous apoptosis in the crypt region in these mice, as well as hypersensitivity to a low dose of MMC (0.1 mg/kg), but not a higher dose (15 mg/kg), compared with heterozygous mice  (Figure 1). We saw no significant difference following treatment with cisplatin, which also cross-links DNA, or a number of other drugs which do not cross-link DNA, such as etoposide and 5-fluorouracil (5-FU) . Additionally, we used a clonogenic survival assay, which measures the ability of whole crypts to recover from DNA damage, to show that Brca2-deficiency resulted in a reduced capacity for crypt survival following treatment with either 15 mg/kg MMC, 10 mg/kg cisplatin or 15 Gy ionizing radiation  (Figure 2). Our data enabled us to make a number of statements: first, that the ability to engage an apoptotic pathway in crypt cells does not directly influence crypt survival; secondly, that we had successfully been able to corroborate published in vitro data regarding the sensitivity of Brca2-deficient cells, in our in vivo system and finally, we suggest that the reason for the lack of tumours caused by somatic BRCA2 mutation may be at least partly due to the rapid removal of cells lacking BRCA2, the number of which will be very low to start with. Individuals with germline mutations in the gene would need to delete many more cells that have lost both functional copies of BRCA2, and may fail to do so efficiently, particularly in the mammary gland and, to a lesser extent, the ovary.
The overwhelming message from the data discussed above is that deficiency in BRCA2 leads to increased sensitivity to DNA-damaging agents. However, the nature of the sensitivity depends on a number of things, including cell type, agent used, the mutation in BRCA2 and the method of analysis. In our system, loss of Brca2 causes increased sensitivity to DNA damage, particularly cross-linking agents such as MMC. Cells which lack Brca2 are removed from the tissue, thereby preventing a build up of cells with the potential to acquire further mutations that may lead to malignancy. It seems likely that, in the mammary gland of individuals with a germline mutation in BRCA2, this process is less effective, allowing BRCA2-deficient tumours to develop specifically in this tissue.
The Molecular Biology of Colorectal Cancer: Focused Meeting held at the UBHT Education Centre, Bristol, U.K., 10–11 March 2005. Organized and edited by T. Corfield (Bristol, U.K.), C. Paraskeva (Bristol, U.K.) and H. Wallace (Aberdeen, U.K.).
Abbreviations: BER, base excision repair; BRCA2, breast-cancer susceptibility 2; ES, embryonic stem; FA, Fanconi anaemia; MEF, mouse embryonic fibroblast; MMC, mitomycin C; MMS, methylmethanesulphonate; SI, small intestine
- © 2005 The Biochemical Society