Oncogene (2000) 19, 6159 ± 6175 ã 2000 Macmillan Publishers Ltd All rights reserved 0950 ± 9232/00 $15.00 www.nature.com/onc Lessons learned from BRCA1 and BRCA2 Lei Zheng1, Shang Li1, Thomas G Boyer1 and Wen-Hwa Lee*,1 1Department of Molecular Medicine, Institute of Biotechnology, University of Texas Health Science Center at San Antonio, 15355 Lambda Drive, San Antonio, Texas, TX 78245, USA BRCA1 and BRCA2 are breast cancer susceptibility susceptibility genes has provided two human genetic genes. Mutations within BRCA1 and BRCA1 are models for studies of breast cancer. responsible for most familial breast cancer cases. To understand how the loss of BRCA1 or BRCA2 Targeted deletion of Brca1 or Brca2 in mice has function leads to breast cancer formation, mouse revealed an essential function for their encoded products, genetic models for BRCA1 or BRCA2 mutations have BRCA1 and BRCA2, in cell proliferation during been established. This work has revealed that Brca1 embryogenesis. Mouse models established from condi- homozygous deletions are lethal at early embryonic tional expression of mutant Brca1 alleles develop days (E)5.5 ± 13.5 (Gowen et al., 1996; Hakem et al., mammary gland tumors, providing compelling evidence 1996; Liu et al., 1996; Ludwig et al., 1997). Three that BRCA1 functions as a breast cancer suppressor. independent groups generated distinct mutations within Human cancer cells and mouse cells de®cient in BRCA1 Brca1, yet nonetheless observed similar embryonic or BRCA2 exhibit radiation hypersensitivity and phenotypes, including defects in both gastrulation and chromosomal abnormalities, thus revealing a potential cellular proliferation, and death at E6.5 (Hakem et al., role for both BRCA1 and BRCA2 in the maintenance of 1996; Liu et al., 1996; Ludwig et al., 1997). A fourth genetic stability through participation in the cellular group that generated a distinct Brca1 mutation response to DNA damage. Functional analyses of the observed embryos that survived until E13.5 and BRCA1 and BRCA2 gene products have established exhibited defects in neural development, including their dual participation in transcription regulation and anencephaly and spina bi®da to varying degrees DNA damage repair. Potential insight into the molecular (Gowen et al., 1996). A ®fth group generated a mouse basis for these functions of BRCA1 and BRCA2 has model with a targeted deletion of Brca1 exon 11. The been provided by studies that implicate these two tumor resultant mutant embryos expressed an exon 11- suppressors in both the maintenance of genetic stability deletion variant of Brca1 and died at E12 ± 18.5 (Xu and the regulation of cell growth and dierentiation. et al., 1999b). Collectively, these ®ndings imply a role Oncogene (2000) 19, 6159 ± 6175. for the Brca1 gene product in growth and/or dierentiation during mouse embryogenesis. Keywords: BRCA1; BRCA2; breast cancer; tumor Similarly, three separate groups have demonstrated suppressor; transcription regulation; DNA damage that mice homozygous for a Brca2 truncation mutation repair at the 5' end of exon 11 die at E8.5 ± 9.5 of gestation (Ludwig et al., 1997; Sharan et al., 1997; Suzuki et al., 1997). Prior to growth arrest, the mutant embryos BRCA1 and BRCA2, two genetic models of appear to have been dierentiating and to be forming breast cancer mesoderm, suggesting that the in¯uence of Brca2 during mouse embryogenesis is manifest more on Breast cancer is one of the most frequent malignancies proliferation than dierentiation. Mice homozygous aecting women. The cumulative lifetime risk of a for a Brca2 truncation mutation at the 3' end of exon female for the development of this disease is about 11 also exhibit progressive proliferative impairment 10% (Claus et al., 1991). For this reason, breast cancer and die prenatally or perinatally; those animals that do has been the subject of intense study; however, the survive to adulthood, however, develop lethal thymic mechanism underlying breast cancer formation is still lymphomas (Connor et al., 1997; Friedman et al., largely unknown. In the last decade of the 20th 1998). century, two breast cancer susceptibility genes, BRCA1 Although these mouse models have revealed a and BRCA2, which together are responsible for most fundamental role for BRCA1 and BRCA2 in embry- of the hereditary breast cancer cases, were identi®ed ogenesis, mice carrying heterozygous Brca1 and Brca2 (Hall et al., 1990; Miki et al., 1994; Wooster et al., mutations develop normally and are no more suscep- 1994, 1995). Mutations in BRCA1 account for almost tible to tumors than their normal littermates. The lack all of the hereditary breast and ovarian cancer cases of tumor formation in mice heterozygous for Brca1 or and up to 40 ± 50% of families with hereditary breast Brca2 has rendered it dicult to study the pathogenesis cancer only (Easton et al., 1993). Mutations in BRCA2 of breast cancer. However, the recent establishment of are linked to the other half of inherited breast cancer a conditional knockout animal model has provided a families and also to male breast cancer (Wooster et al., useful system with which to study early events in breast 1994, 1995). The identi®cation of familial breast cancer tumor formation (Deng and Scott, 2000). By exploiting the mammary epithelium-speci®c expression of a MMTV-Cre or WAP-Cre transgene to induce a Cre-LoxP mediated deletion of brca1 exon *Correspondence: W-H Lee 11 in mammary epithelium, it was reported that ®ve Lessons learned from BRCA1 and BRCA2 L Zheng et al 6160 out of 23 MMTV-Cre or WAP-Cre female mice Analysis of BRCA1- and BRCA2-de®cient cells developed diverse mammary tumors by 10 ± 13 months of age (Xu et al., 1999a). Most of the tumors analysed The establishment of culture cell lines de®cient in either were found to carry p53 mutations, an observation BRCA1 or BRCA2 has facilitated studies designed to consistent with previous reports from studies of human de®ne and characterize their corresponding biological BRCA1 familial breast tumors (Crook et al., 1997; activities. Cell lines established from clinical tumor Eisinger et al., 1997). These observations imply a link specimens include the BRCA1-de®cient human breast between p53 mutation and Brca1-associated mammary adenocarcinoma HCC1937 cell line and the BRCA2- tumor development, a notion further supported by a de®cient human pancreatic carcinoma CAPAN-1 cell documented acceleration in both the frequency and age line. Brca1 and Brca2-de®cient mouse embryos derived of onset of breast tumor formation accompanying from gene targeting events have also served as an inactivation of one germline copy of p53 in these invaluable source of Brca1 and Brca2-de®cient stem conditional Brca1 knockout mice (Xu et al., 1999a). (ES) and ®broblast (MEF) cell lines for fundamental Collectively, these observations support a role for research purposes. BRCA1 as a breast cancer suppressor gene. BRCA1-de®cient HCC1937 cells, BRCA1-null ES Evidence to support a role for BRCA2 as a tumor cells, and Brca1-exon 11 deletion MEF cells are all suppressor includes the observation of tumorigenesis in characterized by radiation hypersensitivity. Increased mice homozygous for a Brca2 truncation mutation at sensitivity to the radiomimetic agent methyl methane- the 3' end of exon 11 (Connor et al., 1997; Friedman et sulfonate (MMS) and ionizing radiation (IR), but not al., 1998). Inactivating mutations in mitotic checkpoint to ultraviolet (UV) radiation, has also been observed in genes such as Bub1, Mad3L and p53, whose products BRCA1-de®cient cells (Gowen et al., 1998; Scully et are pressed into action as a consequence of chromo- al., 1999; Xu et al., 1999b; Zhong et al., 1999). somal damage, are believed to relieve growth arrest Reintroduction of a wild-type BRCA1 allele, but not caused by BRCA2 de®ciency and precipitate neoplastic clinically validated BRCA1 missense mutant alleles, transformation (Lee et al., 1999). Nevertheless, a can complement the MMS and IR sensitivity of suitable breast cancer model for studying the role of BRCA1-de®cient cells, suggesting that the cellular BRCA2 in breast cancer development remains to be response to DNA damage is compromised in breast established. cancer patients carrying BRCA1 mutations (Scully et Both BRCA1 and BRCA2-de®cient cells are al., 1999; Zhong et al., 1999). BRCA1-null ES cells are characterized by cumulative chromosome abnormal- defective in the repair of both oxidative DNA damage ities, including chromosomal breaks, aberrant mitotic by transcription-coupled processes (Gowen et al., 1998) exchanges and aneuploidy (Lee et al., 1999; Patel et and chromosomal double-strand breaks by homolo- al., 1998; Xu et al., 1999b). Chromosomal instability gous recombination (Moynahan et al., 1999). Defective has been proposed as the pathogenic basis for control of the DNA-damage induced G2/M checkpoint mammary tumor formation caused by BRCA1 and has also been observed in BRCA1-exon 11 deletion BRCA2 de®ciency. Paradoxically, chromosomal in- MEFs, thereby implicating BRCA1 in cell cycle stability is invariably accompanied by growth arrest checkpoint control (Xu et al., 1999b). Improper or increased cell death, and the early embryonic centrosome duplication is another prominent charac- lethality associated with BRCA1 or BRCA2 de®ciency teristic of these cells leading to multipolar spindle has been attributed to these cellular responses. How formation and consequent unequal chromosomal then might BRCA1 or BRCA2 mutation lead segregation and micronuclei formation (Xu et al., ultimately to uncontrolled cell growth and tumor 1999b). Although a function in centrosome duplication formation? One answer to this question may lie in the is consistent with the reported localization of the observation that tumors found in Brca1 and Brca2- BRCA1 protein to centrosomes (Hsu and White, 1998), de®cient mice harbor additional inactivating mutations multiple centrosomes could be formed as a conse- in p53 and mitotic checkpoint genes (Lee et al., 1999; quence of accumulated DNA damage in Brca1- Xu et al., 1999b).