Germline mutation of Brca1 alters the fate of mammary ... - Nature

Oncogene (2013) 32, 2715–2725 & 2013 Macmillan Publishers Limited All rights reserved 0950-9232/13 www.nature.com/onc

ORIGINAL ARTICLE

Germline mutation of Brca1 alters the fate of mammary luminal cells and causes luminal-to-basal mammary tumor transformation F Bai1, MD Smith2, HL Chan1 and X-H Pei1,3 Breast cancer developed in familial BRCA1 mutation carriers bears striking similarities to sporadic basal-like breast tumors. The mechanism underlying the function of BRCA1 in suppressing basal-like breast cancer remains unclear. We previously reported that the deletion of p18Ink4c (p18), an inhibitor of G1 cyclin Ds-dependent CDK4 and CDK6, stimulates mammary luminal progenitor cell proliferation and leads to spontaneous luminal tumor development. We report here that germline mutation of Brca1 in p18-deficient mice blocks the increase of luminal progenitor cells, impairs luminal gene expression and promotes malignant transformation of mammary tumors. Instead of the luminal mammary tumors developed in p18 single-mutant mice, mammary tumors developed in the p18;Brca1 mice, similar to breast cancer developed in familial BRCA1 carriers, exhibited extensive basal-like features and lost the remaining wild-type allele of Brca1. These results reveal distinct functions of the RB and BRCA1 pathways in suppressing luminal and basal-like mammary tumors, respectively. These results also suggest a novel mechanism—causing luminalto-basal transformation—for the development of basal-like breast cancer in familial BRCA1 carriers and establish a unique mouse model for developing therapeutic strategies to target both luminal and basal-like breast cancers. Oncogene (2013) 32, 2715–2725; doi:10.1038/onc.2012.293; published online 9 July 2012 Keywords: Brca1; luminal progenitor; basal-like tumor

INTRODUCTION Mammary epithelia are mainly composed of luminal and basal or myoepithelial cells, both of which are believed to originate from a common stem cell. Accordingly, breast cancer can be divided into two major types: luminal-type and basal-like tumors.1 Approximately 90% of breast cancers occur sporadically. The remaining 10% are inheritable, and half of these are linked with mutations in the BRCA1 gene. BRCA1 mutation carriers bear striking similarities to sporadic basal-like tumors.2–5 The mechanism underlying the function of BRCA1 in suppressing basal-like breast cancer remains unclear. A major hurdle in elucidating this issue has been the lack of suitable mouse models harboring a germline mutation of Brca1 and spontaneously developing basal-like mammary tumors. Basal-like breast tumors developed in BRCA1 mutation carriers are thought to originate from either the mammary stem cell or basal progenitors.6,7 The predominant expression pattern of BRCA1 in mammary luminal epithelial cells in humans and mice, however, suggests that germline mutation of BRCA1 has a role in luminal cell development. Indeed, more recently a few groups observed that aberrant luminal progenitors might be the target for breast basal-like tumor development in BRCA mutation carriers,8,9 and knocking down of BRCA1 impaired luminal cell differentiation.7,10 Owing to the unknown genetic background of the patients and the limitation of the tissue sample collection, however, it remains elusive whether, and how, germline mutation of BRCA1 selectively targets luminal progenitors, leading to basallike tumor development and progression. The functions of BRCA1 have been linked with multiple pathways, but have proven to be difficult to precisely define. Several biochemical functions have been described for the BRCA1

protein including transcriptional regulation, homologous recombination and ubiquitin ligase activity.11,12 At the cellular level, a theme has emerged that links the function of BRCA1 with genome integrity. A reduction of, or deficiency in, Brca1 function causes extensive chromosomal abnormalities, resulting in the activation of DNA-damage checkpoint pathways and inhibition of cell proliferation.13–15 These findings suggest that overcoming growth inhibition resulting from the activation of cell cycle checkpoints may be a necessary step for the development of tumors initiated by Brca1 reduction or loss, as seen in Brca1 mutant mice combined with a mutation in p53 pathway. p18Ink4c (referred to as p18 hereafter) is a member of the mammalian INK4 family that inhibits CDK4 and CDK6 whose activation by mitogen-induced D-type cyclins leads to phosphorylation and functional inactivation of RB, p107 and p130. Functional inactivation of this pathway, consisting of INK4-cyclin D/CDK4/6-Rb-E2F, is a common event in the development of most types of cancer.16,17 p18 is a haploinsufficient tumor suppressor in mice,18 and deletion of p18 results in increased proliferation of stem or progenitor cells and development of pre-neoplastic or tumor phenotypes in multiple tissues or organs, including prostate,19 neuroendocrine organs,20 lung,21 breast22 and brain.23 These studies established the concepts that an INK4 gene, p18, suppresses tumorigenesis by constraining stem or progenitor cells and that reduced RB pathway activity sensitizes cells to more rapid or malignant transformation when combined with heterozygosity of tumor suppressors from other pathways, including Pten, Men1 and p53. Mutation, or reduced expression, of the p18 gene has been observed in several types of human cancers including breast cancer.24–28 p18-deficient mice

1 Molecular Oncology Program, Department of Surgery, University of Miami Miller School of Medicine, Miami, FL, USA; 2Lineberger Comprehensive Cancer Center, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA and 3Sylvester Comprehensive Cancer Center, University of Miami Miller School of Medicine, Miami, FL, USA. Correspondence: Dr X-H Pei, Molecular Oncology Program, Department of Surgery, University of Miami Miller School of Medicine, 1550 NW 10th Avenue, Miami, FL 33136, USA. E-mail: [email protected] Received 9 April 2012; revised 24 May 2012; accepted 5 June 2012; published online 9 July 2012

Brca1 controls luminal cell fate and basal-like tumor F Bai et al

2716 spontaneously develop estrogen receptor (ER)-positive mammary luminal tumors at a high incidence. Deletion of p18 stimulates luminal progenitor cell proliferation at pubertal age and maintains an expanded luminal progenitor cell population throughout life.22 Prompted by the predominant expression of Brca1 in luminal cells and growth defects caused by Brca1 deficiency and reduction, as well as the unique phenotype shown in p18deficient mice—increased luminal progenitor cell proliferation and luminal tumor development—we hypothesized that loss of p18 may overcome Brca1-deficiency-induced growth inhibition, allowing us to determine the effect of germline mutation of Brca1 on mammary gland and tumor development. We generated p18;Brca1 double-mutant mice and discovered the function of Brca1 in maintaining luminal cell fate by preventing luminal progenitor cells from aberrant differentiation and suppressing the differentiation of basal/myoepithelial cell lineage. Germline mutation of Brca1 alters luminal epithelial cell fate and, when combined with the reduced function of the RB pathway, for example, p18 loss, induces luminal-to-basal transformation leading to basal-like tumor formation. RESULTS Deletion of p18 alleviates proliferative and developmental defects caused by heterozygous germline mutation of Brca1 To study the function of Brca1 in mammary gland, we first examined the expression of Brca1 in different epithelial cells. Fluorescence-activated cell sorting (FACS) sorted luminal progenitor-enriched CD24 þ CD29 cells from wild-type mice at a pubertal age expressed 6.7-fold more Brca1 than CD24 CD29 cells that include basal/myoepithelial cells, stromal and other cells (Figure 1a), indicating that Brca1 is predominantly expressed in the luminal epithelium. p18 mRNA level was inversely correlated with that of Brca1, and was significantly lower in luminal progenitor-enriched cells. Immunostaining of pubertal mammary glands further confirmed predominant Brca1 expression in the luminal epithelium (Figure 1b). Haploid loss of Brca1 in primary mammary epithelial cells (MECs, Figure 1c) increased p18 mRNA level, and knockdown of Brca1 in 293 T, T47D and MCF7 cells (Figure 1d) led to a 1.3- to 1.5-fold increase in p18 mRNA and protein levels, suggesting that Brca1 negatively regulates p18. Brca1 / mice die embryonically and Brca1 þ / causes various growth defects in tissues and animals,13,29,30 preventing researchers from determining the effects of germline mutation

of Brca1 in mammary gland and tumor development. Prompted by the observation of expanded luminal progenitor cells and mammary tumor development in p18-deficient mice and the inverse correlation between p18 and Brca1, we determined the genetic interaction between these two genes. We mated p18-/homozygotes with Brca1 þ /– heterozygotes to generate animals heterozygous for both genes. These mice were then intercrossed to generate both p18 / ;Brca1 þ /– and p18 þ /;Brca1 þ / animals. Genotype analysis of more than 600 offspring did not identify any p18 / ; Brca-/– double null mice, nor were any viable p18 / ; Brca / embryos identified beyond embryonic day 14.5 (data not shown). These results indicate that deletion of p18 does not rescue the embryonic lethality caused by Brca1 loss. Haploid loss of Brca1 caused by heterozygous germline mutation of Brca1, as expected, decreased mouse body weight from 12 to 24 weeks of age (Figure 1e), inhibited cellular proliferation in young and old MECs (Ki67-positive cells, 10.2% in wild type and 3.3% in Brca1 þ /– for young, 3.6% in wild type and 0.8% in Brca1 þ /– for old MECs, respectively) (Figures 1f and g), and induced premature senescence (Figures 1g–i). Notably, p18 deficiency rescued body weight decrease and premature senescence (Figures 1e and h) caused by heterozygous germline mutation of Brca1. The proliferation in 30–50% Brca1 þ / mammary glands of each mouse was rescued by p18 loss at 2 months of age (Ki67-positive cells, 3.3% in Brca1 þ /– to 21% in p18 / ;Brca1 þ / glands), although the proliferation of the remaining Brca1 þ / glands (50–70%) was not rescued by p18 deficiency (Figure 1f). When aging mice were examined, 3.6% Ki67-positive MECs were detected in wild-type females at 20 months of age, but only 0.83% Ki67-positive MECs were found in Brca1 þ / littermates (Figure 1g). Loss of p18 rescued the proliferation of most mammary glands and led to tumor development in Brca1 þ / background at this age (Figures 1g–i and data not shown). As it has been shown that overexpression of Brca1 causes cell cycle arrest by transactivating p21 in vitro31,32 and that Brca1 deficiency results in the increase of p21 and premature senescence in vivo,13,30,33 we determined the expression of p21 and p16INK4a, another INK4 member of cell cycle inhibitor, in these mutant mice. We found that p21 mRNA level was not altered in Brca1 þ / mammary glands at 3 months of age, but increased by 2.9- and 3.1-fold in p18 / and p18 / ;Brca1 þ /– glands, respectively (Supplementary Figure S1). p16INK4a mRNA level was increased 1.7fold in Brca1 þ / , and 1.4- and 3.9-fold in p18 / and p18 / ; Brca1 þ /– glands, respectively (Supplementary Figure S1). These results suggest that p18 may not be the only cell cycle inhibitor

Figure 1. Brca1 negatively regulates p18 expression, and loss of p18 rescues growth defects induced by haploid loss of Brca1. (a) Mammary cells from wild-type mice at 7 weeks of age were isolated and sorted by flow cytometry for CD24 and CD29 cell populations. Total RNA from the indicated cell populations (R4, CD24 CD29 , R2, CD24 þ CD29 ) was extracted and analyzed for the expression of Brca1 and p18 by quantitative real-time PCR. Data are expressed relative to the corresponding values for R4 populations as mean±s.d. from triplicates of each of the two independent mice. (b) Eight-week-old littermate mammary glands from wild-type mice were stained with an antibody against Brca1. Note that Brca1-positive staining (red) is highly enriched in luminal epithelial cells. (c) RNA from the primary mammary cells isolated from 3-month-old female mice with the indicated genotype was extracted and analyzed for the expression of p18. Results represent the mean±s.d. of three animals per group. (d) Human MCF7, T47D and 293T cells were infected with lentiviral pGIPZ-empty (sh-Ctrl) or pGIPZshBrca1 (sh-Brca1), and RNA (left) and protein (right) were then extracted and analyzed for the expression of p18 and Brca1 by quantitative real-time PCR and western blot, respectively. Quantitative real-time PCR data (left) are expressed relative to the corresponding values for shCtrl cells as mean±s.d. from triplicates of a representative experiment. Brca1 and p18 protein levels in western blots (right) were quantified relative to the expression in sh-Ctrl cells. (e) Body weight comparison for female mice of different genotypes. s.d. bars and P values are indicated. (f ) Mammary glands of different genotypes at 2 months of age were stained with Ki67. Note the abundant Ki67 þ cells in the rescued and very few Ki67 þ cells in the unrescued p18 / ;Brca1 þ / mammary glands. The percentages of Ki67-positive cells were calculated from cells situated in clear duct/gland structure. Results represent the mean±s.d. of three animals per group. (g) Representative mammary glands from old females were stained with Ki67. Highly proliferative mammary tumors from p18 / and p18 / ;Brca1 þ / are shown. Note the highly heterogeneous tumors derived from double-mutant mice. The percentages of Ki67-positive cells were calculated from cells situated in clear duct/gland structure in normal glands and from all cells in tumors. Results represent the mean±s.d. of three animals per group. (h) Mammary glands from mice of different genotypes at 1 year of age were examined for SA-b-gal staining. (i) Tumor-free mammary glands at 1 year of age were examined for pRb protein phosphorylation at Ser608 by CDK4 and CDK6. Counterstaining is blue and positive staining is brown. The percentages of pRb-S608-positive cells were calculated from cells situated in clear duct/gland structure. Results represent the mean±s.d. of three animals per group. Oncogene (2013) 2715 – 2725

& 2013 Macmillan Publishers Limited


2717 that interacts with Brca1 in vivo. Consistent with our previous finding in other tissues,34 these data also indicate that other cell cycle inhibitors, for example, p21 and p16, compensate for the loss of p18 in mammary glands.


To provide further molecular evidence supporting the rescue of growth inhibition in Brca1 þ / mammary by p18 loss, we directly examined Rb phosphorylation in both normal mammary tissues and tumors of different genotypes by using an antibody

Oncogene (2013) 2715 – 2725


2718

Figure 2. Brca1 heterozygosity inhibits p18-deficiency-enhanced expansion of luminal progenitor cells and expression of genes associated with luminal epithelial differentiation. (a, b) Mammary cells from 6-week-old (a) or 9-month-old (b) virgin mice were analyzed by flow cytometry for CD24 together with CD29 or SCA1. The bar graphs represent the mean±s.d. of three animals per group. (c) Mammary cells from 5-month-old virgin mice were analyzed by flow cytometry for CD24 and CD10 expression. (d) Freshly isolated mammary cells were cultured in Matrigel-coated 24-well plates. Nine days after culture, colonies were stained with CK8 and CK5 and counted as described in the Materials and Methods. The assay was performed in triplicate for each animal. The bar graphs represent the mean±s.d. of three animals per group. (e) RNA from the tumor-free mammary glands of different genotypes at three months of age was extracted and analyzed for the expression of the genes indicated. Results represent the mean±s.d. from triplicates of each of the three independent mice.

specifically recognizing Rb proteins phosphorylated at Ser608 by CDK4 and CDK6,35,36 the functional targets of p18. A visible and consistent increase of pRb-Ser608 phosphorylation, in both intensity and the number of positive cells, was detected in normal p18 / mammary epithelia at 1 year of age (8.9% in wild type to 25.4% in p18 / , Figure 1i), confirming the activation of CDK4 and/or CDK6. Most pRb-Ser608-positive cells are luminal cells (Figure 1i and data not shown), consistent with the finding that p18 deficiency increased luminal progenitor cell proliferation and leads to luminal tumor development.22 In all, 3.3% pRb-Ser608-positive luminal cells were detected in Brca1 þ / females; however, both p18 / ;Brca1 þ / and p18 / mice had a similar and notably higher percentage of positive luminal epithelial cells (22.9% in p18 / ;Brca1 þ / and 25.4% in p18 / , Figure 1i), supporting p18 as a downstream target of Brca1 in controlling luminal cell proliferation. This result also provides an explanation—that reduced RB pathway activity allows Oncogene (2013) 2715 – 2725

Brca1-mutated cells to continue to proliferate—for the development of a higher incidence of mammary tumors in p18 / ;Brca1 þ / mice (87%, see below) than in either germline Brca1 þ / (o10%)14,37 or conditional mammary gland Brca1 / mice (20–25%).15 p18;Brca1 germline double-mutant mice, unlike conditional Brca1 mutant mice in which Brca1 deficiency is directed to a certain cell lineage or developmental stage, offer a unique mouse model to study not only the function of Brca1 in mammary gland and tumor development but also the targeting preference and selectivity for mammary cell lineage and tumor cell type caused by germline mutation of Brca1. Heterozygous germline mutation of Brca1 blocks the expansion of luminal progenitor cells and impairs luminal gene expression in p18 null mice To determine the effect of germline mutation of Brca1 on mammary cell fate determination, we dissected mammary glands & 2013 Macmillan Publishers Limited


2719 Table 1.

Spontaneous mammary tumor development in p18 and Brca1 mutant mice

Tumor

Mammary tumor Metastasis

p18 þ /

Wild type

g

ERa þ tumors % ERa þ cells/tumor Basal marker þ tumorsh %Basal marker þ cells/tumor

p18 /

Brca1 þ /

p18 þ / ;Brca1 þ /

p18 / ;Brca1 þ /

o12 m

12–27 m

o12 m

12–27 m

o12 m

12–22 m

o12 m

12–27 m

o12 m

12–27 m

o12 m

12–22 m

0/5

1/10 (10%)a 0/1

0/3

5/12 (42%)b 0

4/16 (25%) 0

0/3

1/11 (9%)d 0/1

0/4

4/5 (80%) 2–40%

3/4 (75%) 2–40%

10/16 (63%)e 3/10 (30%) 2/10 (20%) o2%

6/16 (38%) 0

1/1 (100%) 5%

19/23 (83%)c 1/19 (5%) 15/19 (79%) 2–40%

1/6 (17%) o2%

13/15 (87%)f 4/13 (31%) 2/13 (15%) o2%

0/1

1/5 (20%) B2%

0/4

7/10 (70%)j 2–95%

4/6 (67%) 2–20%

11/13 (85%)k 2–95%

3/19 (16%)i 2–5%

0/1

1/1 (100%) B2%

Abbreviations: ERa, estrogen receptor alpha; m, months. aA mouse with tumor was 24 months of age. bMice with tumors were 15–27 months of age. cMost mice with tumors were 12–16 months of age, and the oldest one with tumor was 22 months of age. One male developed mammary tumor. dA mouse with tumor was 25.5 months of age. eMice with tumors were 15–27 months of age. fMost mice with tumors were 12–16 months of age and the oldest one with tumor was 20 months of age. One male developed mammary tumor. gMetastasis from mammary tumor was found mostly in lung except one in blood vessel, which was observed in p18 / ;Brca1 þ / mice. hAt least one of the three basal markers (CK5, CK14, and SMA) was positively detected in 42% tumor cells. i One tumor was positively stained with CK5 in B5% tumor cells and the other two were positive in B2% tumor cells. jOne tumor was positively stained with CK5 in B95% tumor cells. kTwo tumors were positively stained with CK5 in B95% tumor cells.

of different genotypes. FACS analysis of primary mammary cell suspensions from virgin females revealed that a CD29loCD24 þ population enriched with luminal stem/progenitors38,39 was significantly increased in p18 / mice compared with wild-type littermates (6 weeks, 45.4±6.1 versus 29±3%; 9 months, 51.7±3.1 versus 40.3±5.1%) as we previously reported.22 Notably, heterozygous germline mutation of Brca1 decreased the expansion of luminal progenitors in p18 / mice, reducing the CD29loCD24 þ population from 45.4±6.1% in p18 / mice to 36.7±5% in p18 / ;Brca1 mice at 6 weeks, or from 51.7±3.1 to 39.3±4.2% at 9 months of age (Figures 2a and b). When the SCA1loCD24high population, also enriched with luminal progenitors,40 was examined, similar results were detected, which was an increase from 46% in wild-type mice to 55% in p18 / mice, and a reduction down to 43% in the p18 / ;Brca1 mice (Figure 2b). Interestingly, the CD24 CD10 þ population enriched for basal/ myoepithelial cells was increased in p18 / ;Brca1 þ / mammary glands (2.1%) compared with either wild type (1.5%) or p18 / counterpart (1.3%, Figure 2c). Colony formation assay further confirmed the inhibitory effect of heterozygous germline mutation of Brca1 on the expansion of luminal progenitor cells caused by p18 loss (230±40.4 versus 377±50.3 per 20 000 cells seeded; Figure 2d). Instead, p18 / ;Brca1 þ / MECs produced more basal colonies than other single-mutant MECs did (215±30.1 for p18 / ;Brca1 þ / versus 104±14.7 for p18 / and 79±29.7 for Brca1 þ / per 20 000 cells seeded), suggesting a function of Brca1 in suppressing basal colony formation. To gain insight into the molecular mechanism underlying the function of Brca1 in maintaining luminal cell fate, we determined the expression of genes associated with luminal epithelial differentiation including Foxa1, Stat5a, Cdh1 and Epcam.41–44 Loss of p18 increased mRNA levels of Foxa1 and CDH1 in pubertal virgin mammary glands (Figure 2e). Heterozygous germline mutation of Brca1, while having inhibitory function on the expression of most genes associated with luminal epithelial differentiation in pubertal glands with wild-type p18 þ / þ , significantly decreased their expression in p18 / glands (Figure 2e). Taken together, these results suggest a function of Brca1 in maintaining proper luminal cell fate by preventing abnormally expanded luminal progenitor cells from aberrant differentiation and suppressing basal/myoepithelial cell differentiation. & 2013 Macmillan Publishers Limited

Heterozygous germline mutation of Brca1 promotes the progression and malignant transformation of luminal tumors Less than 10% of Brca1 þ / mice14,37 and about 20% of mice with both Brca1 alleles conditionally deleted15 developed mammary tumors, suggesting that mutation of Brca1 alone is an insufficient factor to promote mammary tumorigenesis and that it may instead collaborate with mutation of other cancer genes. Taking advantage of the high rate of mammary tumor development in p18 mutant mice and the rescue of Brca1 þ / -induced proliferation defects by p18 loss, we determined how germline mutation of Brca1 affects mammary tumorigenesis in p18-null background. We generated and analyzed a cohort of 51 p18;Brca1 double mutants. p18 / ; Brca1 þ / mice had a similar penetrance of breast cancer (6/16 or 38% before 1 year, 13/15 or 87% by 22 months of age) as p18 / single-mutant mice (4/16 or 25% before one year, 19/23 or 83% by 22 months of age) (Table 1). Similarly, mammary tumor incidence in p18 þ / ;Brca1 þ / double heterozygous mutant mice was also comparable with that in p18 þ / heterozygous mice (Table 1). Importantly, the tumors showed marked morphological and histopathological differences (Figure 3 and below). Mammary tumors developed in p18 / ;Brca1 þ / and p18 þ / ; Brca1 þ / double-mutant mice were more aggressive (increased necrosis, squamous metaplasia, spindle cells, nuclear–cytoplasm ratio, mitotic indices and metastasis) and contained heterogeneous cell types (Table 1). Only 1 out of 19 (5.3%) mammary tumors from p18-/- mice metastasized to the lung after 1 year of age, whereas 3 out of 10 (30%) p18 þ / ;Brca1 þ / and 4 out of 13 (31%) p18 / ;Brca1 þ / mammary tumors metastasized to the lung and blood vessels at a similar age (Table 1). In accordance with the previous finding that Brca1 þ / mice developed mammary tumors during late age at a very low penetrance,14,37 only 1 Brca1 þ / mouse in a cohort of 11 (9%) developed mammary tumor at 25 months of age, similar to that observed in wild-type mice (1 in a cohort of 10, 10%) (Table 1). The mammary tumor developed in the Brca1 þ / single-mutant mouse appeared to be pathologically different (most cells were ER negative and about 2% of cells were positively stained with CK5) from that developed in the wild-type mouse (about 5% cells were ER positive and nearly no CK5-positive cells were detected), but the low incidence prevented us from firmly determining the Oncogene (2013) 2715 – 2725


2720 significance of this finding. Together, these results suggest that p18 loss is an early and initiatory factor in breast cancer development, and that Brca1, although not efficient in suppressing mammary tumor initiation, has a critical role in preventing their progression, invasion and metastasis.

Oncogene (2013) 2715 – 2725

Heterozygous germline mutation of Brca1 converts estrogen receptor alpha (ERa)-positive luminal type tumors into ERanegative basal-like tumors in p18-deficient background In addition to morphological, pathological and cellular differences, the mammary tumors developed in the p18;Brca1 double-mutant



2721 mice also differ in another significant feature from those developed in p18 single-mutant mice. Although 79% (22 out of 28) of p18 / and p18 þ / mammary tumors are positive for ERa, most (24 out of 29, 83%) p18 / ;Brca1 þ / and p18 þ / ; Brca1 þ / mammary tumors were ERa negative (the remaining five double-mutant mammary tumors contained detectable, but less than 2%, ERa-positive cells, Figure 4a, Table 1, and data not shown). We wondered whether transformation of mammary tumors from ERa positive to ERa negative was directly related to Brca1 heterozygosity, or indirectly caused by other changes late in tumor progression in double-mutant mice. Taking advantage of the availability of multiple mammary tumors at different pathologic grades developed in p18 / and p18 / ;Brca1 þ / mice, we determined this issue. We examined and compared the mammary tumors for ERa expression at different pathologic grades. When the tumors were in grade 1 or 2 (tumor cells are well or moderately differentiated and glandular architecture remains distinguishable), nearly all cells from the p18 / ;Brca1 þ / mammary tumors were stained ERa negative, whereas ERapositive cells were frequently detected in the similar grade of p18 / mammary tumors (Figure 4a, left and middle panels). When grade 3 or 4 tumors (tumor cells are poorly differentiated or undifferentiated and glandular architecture is lost) were investigated, similarly, ERa expression could still be detected in p18 / mammary tumors, but not in p18 / ;Brca1 þ / mammary tumors (Figure 4a, right panel and data not shown). More convincingly, the normal glands surrounding the tumors were strongly ERa positive in p18 / mammary, but negative in p18 / ;Brca1 þ / mammary tissue (Figure 4a). These results demonstrate that ERapositive-to-ERa-negative transformation proceeds, yet is not dependent on, malignant transformation and is likely a direct result of Brca1 reduction. p18 / mammary tumors were stained strongly and uniformly for CK8, and only 3 out of 19 (15.8%) tumors were detected CK5 positive in less than 5% tumor cells, suggesting that they predominantly consist of luminal epithelial cells (Figures 4c and d and Table 1). Notably, p18 / ;Brca1 þ / mammary tumors were highly heterogeneous: although the majority of cells (60 - 80%) in most tumors were positive for CK8, a significant portion of highly disorganized clustering cells (20–40%) stained positive by CK5 (in scattering, mixing or chimeric patterns) (Figures 4c and d and Table 1). A similar positive staining pattern for CK14 and SMA was also detected in the p18 / ;Brca1 þ / mammary tumors, but hardly found in p18 / tumors (Figure 4b, Table 1 and Supplementary Figure S1). Three double-mutant mammary tumors, one from p18 / ;Brca1 þ / and the other two from p18 þ / ;Brca1 þ / mice, displayed multiple tumor cell types and were identified as 495% CK5 þ tumors (Figure 4d). Separate immunohistochemical staining and microscopic examination of more and larger sections confirmed CK5 positivity in p18 / ; Brca1 þ / , but not in p18 / , mammary tumors (Figure 4c). Western blot analysis also showed the increased CK5 protein levels in p18 / ;Brca1 þ / tumors over p18 / tumors (Figure 4e). Hence, haploid loss of Brca1 transformed p18 /

mammary tumors from a predominantly luminal subtype into a partially, at least, basal-like subtype. Immunostaining of p18 / ;Brca1 þ / mammary tumors with antibodies specific to Brca1 and CK5 in serial sections revealed a mutually exclusive staining pattern between these two proteins in the tumor cells (Figure 4f), suggesting a function of Brca1 in suppressing the expression of basal-like features. This notion is consistent with a previous finding that the function of Brca1 is important in maintaining luminal cell features. The remaining wild-type Brca1 allele was retained in all 14 p18;Brca1 doublemutant tumors (nine from p18 / ;Brca1 þ / mice and five from p18 þ / ;Brca1 þ / mice) when genomic DNA was prepared from blended tumor cells (data not shown), but was lost in at least 4 out of 10 tumors when genomic DNA was prepared from CK5 þ cells isolated by immunofluorescent-staining-guided laser capture microdissection under microscopy (Figure 4g), confirming an important role of Brca1 in suppressing CK5 þ basal-like tumor development. Together, these five features of the p18 / ;Brca1 þ / mammary tumors—high degree of heterogeneity, increased population of CK5-positive cells, loss of wild-type allele of Brca1 in CK5positive tumor cells, ERa negativity and invasiveness and metastasis—are consistent with an interpretation that haploid loss of Brca1, although not efficiently inducing a high incidence of mammary tumor by itself, significantly changed the properties of mammary tumors induced by p18 loss, resulting in their malignant and luminal-to-basal-like transformation. DISCUSSION In this paper, we reported that Brca1 was predominantly expressed in mammary luminal epithelial cells and that Brca1 negatively regulated p18, whose loss led to the expansion of luminal progenitor cells. Heterozygous germline mutation of Brca1 blocked the expansion of luminal progenitor cells and inhibited luminal gene expression in p18-null mice, suggesting a function of Brca1 in maintaining proper luminal cell lineage by preventing abnormally expanded luminal progenitor cells from aberrant differentiation and suppressing basal/myoepithelial cell differentiation. These data also answered the question as to why conditional deletion of Brca1 in mammary luminal, but not in basal cell lineage, impaired mammary development.45,46 To the best of our knowledge, this is the first genetic evidence directly showing that Brca1 maintains luminal cell fate in normal mammary development. BRCA1 has long been suggested to have an important role in mammary epithelial cell differentiation,10,47–50 which was concluded, however, mostly from either association of BRCA1 expression pattern with mammary cell maturation and differentiation or cell culture system. Knockdown of BRCA1 in human mammosphere-initiating cells blocked differentiation of luminal progenitor cells in vitro and in mouse xenografts.7 Somatic deletion of Brca1 in basal cells by K14- and K6a-cre did not cause any observable changes in the normal mammary gland,45 whereas

Figure 3. Brca1 heterozygosity induces metastatic conversion of p18-deficient mammary tumors. Representative hematoxylin and eosin staining of mammary tumors derived from p18 / ;Brca1 þ / mice are shown (b–j). p18 / mammary tumor (a) is shown as a control. Note the tubular carcinoma (a) in p18 / mice that retained a relatively well-differentiated gland structure; papillary carcinoma in p18 þ / ; Brca1 þ / mice that have finger-like stalks with a central fibrovascular core covered with epithelium (b); poorly differentiated carcinoma with squamous metaplasia that is the major tumor type developed in p18 / ;Brca1 þ / mammary and contains highly pleomorphic tumor cells with numerous mitoses shown in the inset (c); poorly differentiated, infiltrating p18 / ;Brca1 þ / breast cancer with few tubules, highly pleomorphic nuclei and many mitoses surrounded by connective tissues (d); cysts frequently observed in the p18 / ;Brca1 þ / papillary carcinoma that contained abundant fluid and irregular epithelial islands with cribriform pattern and poorly differentiated and high nuclear/ cytoplasmic ratio shown in the inset (e); and mammary carcinoma metastasized to the lung (T) found in a p18 / ;Brca1 þ / mouse (f ). A mammary tumor (g) developed in p18 / ;Brca1 þ / mice metastasized to lung (h) that was stained with antibodies against CK5 (green) and CK8 (red) and shown in inset (h). Note the similar tumor cell type in lung (h) with the original tumor in mammary (g). A mammary tumor from p18 / ;Brca1 þ / mice displayed multiple tumor cell types (i, lower power, j, higher power), which was subsequently identified as 495% CK5 þ basal-like tumor. & 2013 Macmillan Publishers Limited

Oncogene (2013) 2715 – 2725


2722

Figure 4. Brca1 heterozygosity transforms p18-deficient luminal tumors into basal-like tumors. (a) Immunohistochemical staining of ERa in mammary tumors from p18 / and p18 / ;Brca1 þ / mice. Note the highly expressed ERa levels in normal mammary glands (arrows) surrounding p18 / tumors. Only a few cells in normal glands (arrows) surrounding p18 / ;Brca1 þ / tumors were faintly stained by ERa. ERa remained expressed in p18 / tumors, but not in p18 / ;Brca1 þ / tumors. Boxed areas are magnified in the insets. (b, c) Immunostaining of tumors derived from p18 / and p18 / ;Brca1 þ / mice with SMA (b) and CK5 (c). Note the scattering, chimeric and heterogeneous expression patterns of SMA and CK5 in p18 / ;Brca1 þ / tumors. The inset in (b) shows the staining of the normal glands in the same mouse. Boxed areas in (c) are magnified in the insets. (d) Immunostaining of tumors with CK8 (red) and CK5 (green). Note the various degrees of CK5 þ cells in different p18 / ;Brca1 þ / tumors. The tumor in the far right is the same tumor shown in Figures 3i and j, in which 495% tumor cells are CK5 þ , and a normal gland is shown for staining control. (e) Representative mammary tumors from p18 / and p18 / ;Brca1 þ / mice were analyzed for CK8 and CK5 expression by western blot analysis. (f) Immunostaining of p18 / ;Brca1 þ / mammary tumors with Brca1 and CK5 in serial sections. Note the mutually exclusive staining pattern between Brca1 and CK5 in the tumor cells. (g) LOH analysis of CK5 þ patchy tumor cells derived from immunofluorescent-staining-guided laser capture microdissection. N, normal; T/K-, tumor with CK5 staining; T/K þ , tumor with CK5 þ staining.

Oncogene (2013) 2715 – 2725



2723 deletion in luminal cells by MMTV-cre decreased the number of luminal progenitor cells in virgin mice but increased in pregnant mice,46 implying that the mutation of Brca1 preferentially targets luminal cell lineage. Recently, two groups identified luminal progenitors as the target cells for basal-like tumor development in BRCA1 germline mutation carriers.8,9 However, because of the unknown genetic background of the patients in addition to BRCA1 mutation and the limitation of the tissue sample collection, discrepancy in the results—one showing the expanded8 and the other showing the decreased9 luminal progenitors—was observed in disease-free breast tissues of BRCA1 mutation carriers. Our results suggest that germline mutations in Brca1 impair luminal cell lineage and mammary development and, when combined with reduced function of the RB pathway, for example, loss of p18, promote basal-like trans-differentiation. The molecular mechanism underlying the function of Brca1 in maintaining luminal cell fate in normal mammary development and breast cancer formation is largely elusive. Brca1 is a multifunctional protein with roles in DNA-damage repair, cell cycle control, transcriptional regulation and ubiquitination. It is the ability of BRCA1 to regulate transcription that controls normal differentiation and suppresses tumor development.51,52 Brca1 regulates transcription at several levels, from interacting with sequence-specific transcriptional factors, such as p53, Myc, STAT1, CtIP and Oct1, to interactions with the RNA polymerase II complex and enzymes involved in chromatin remodeling, such as HDAC1 and HDAC2, SWI/SNF complex and p300/CBP.53–56 Many genes were found to be regulated by Brca1, including those associated with luminal differentiation, such as Foxa1 and ERa.55–58 We show in this study that heterozygous germline mutation of Brca1 leads to the downregulation of luminal differentiation genes including Foxa1, Stat5a, Cdh1 and Epcam in tumor-free mammary glands, which not only confirms previous findings, most of which were derived from cell culture systems, but also provides the genetic evidence supporting the role of Brca1 in regulation of these genes in cell fate determination and maintenance of normal mammary stem and progenitor cells. Consistent with this finding, we also report that Brca1 deficiency converts ER-positive luminal tumors into ER-negative basal-like tumors, suggesting that Brca1 also functions to regulate luminal differentiation genes and maintain luminal tumor cell fate in breast cancer progression. In accordance with observations derived from Brca1 mutant mice in combination with mutations in the p53 pathway,29,59 we found that Brca1 mutation induced growth inhibition in mice and rarely led to tumor formation, but converted p18-deficient luminal tumors into invasive basal-like tumors, indicating that Brca1 mutation has an important role in tumor progression, rather than in initiation of breast cancer. Human basal-like breast tumors that develop in Brca1 mutation carriers are heterogeneous,60 and some have been found to be mixed epithelial/mesenchymal metaplastic carcinomas with mesenchymal features.61 We observed that p18 / ;Brca1 þ / mammary tumors, unlike p18 / tumors, are highly heterogeneous, with some exhibiting various degrees of whorls and clusters of spindle-shaped tumor cells (Figures 3i and j), a typical mesenchymal characteristic. These observations imply that germline mutation of Brca1 may lead to epithelial-tomesenchymal transition (EMT), which has an important role in intratumoral heterogeneity and breast basal-like tumor progression.62 It has recently been reported that Slug, an EMT-associated transcription factor, is aberrantly expressed in the breast of Brca1 mutant carriers, and that it is necessary for the increase in the basal-like phenotypes of human breast cancers created in mouse by transformation of BRCA1mut/ þ patient-derived breast epithelial cells with four tumorigenic genes (TP53R175H, CCND1, PI3KCA and KRASG12V).9 Brca1 was also found to repress FoxC1, FoxC2 and Twist,63,64 all of which are typical EMT-associated transcription factors, suggesting that Brca1 may function in controlling EMT. Whether Brca1 mutation induces EMT and how it contributes to & 2013 Macmillan Publishers Limited

basal-like tumor development and progression remains to be investigated. Most, if not all, genetic studies used co-mutation of one of the genes in the p53 pathway to overcome the growth defects induced by the mutation of Brca1 in mice.13,29,30,59 Gene expression analysis segregated mammary tumors derived from g-irradiated p53 þ / ;Brca1 þ / mice into basal-like type tumor.65,66 It is, however, unclear in these studies whether germline mutation of Brca1 alone contributed to basal-like tumor development because mammary tumors formed in p53 þ / mice fall into multiple molecular subtypes including the basal-like subtype.67 Conditional deletion of Brca1 in mammary luminal cell lineage by MMTV-Cre or Wap-Cre resulted in basal-like tumor development in p53 / or p53 þ / mice,59 suggesting that the cells of origin of basal-like tumors could be luminal cells. Indeed, conditionally directing Brca1 deficiency to p53 þ / luminal progenitor cells by Blg-Cre led to basal-like tumor development, which was similar to human Brca1 breast cancers, supporting the notion that Brca1 mutant basal-like mammary tumors could originate from luminal progenitors.64 In sum, this study demonstrates that germline mutation of Brca1 is not efficient in initiating mammary tumor formation and that loss of p18 results in the expansion of luminal progenitors, which is an early and initiatory factor in luminal tumor development. Germline mutation of Brca1 alters luminal tumor cell differentiation and activates basal-like features, converting ERpositive luminal tumors into ER-negative basal-like tumors. These results suggest a novel mechanism—causing luminal-to-basal transformation—for the development of basal-like breast cancer in familial BRCA1 carriers and establish a unique mouse model to develop the therapeutic strategies to target both luminal and basal-like breast cancers. MATERIALS AND METHODS Mice The generation and genotyping of p18 and Brca1 mutant mice have been described previously;68,69 p18 and Brca1 mutant mice have been backcrossed for 7 and 410 generations with BALB/c mice, respectively. Cohorts were housed and analyzed in a common setting, and agematched or littermate controls were used for all experiments as indicated. IACUC (Institutional Animal Care and Use Committee) at the University of North Carolina and University of Miami approved all procedures.

Histopathology and immunohistochemistry Histopathology and immunohistochemistry were performed as described previously.18 The following primary antibodies were used: to p-Rb-S608, (Cell Signaling, Danvers, MA, USA), CK5 (Covance, Princeton, NJ, USA), CK8 (American Research Products, Waltham, MA, USA), CK14, SMA (Thermo Scientific, Rockford, IL, USA), ERa, Brca1 (Santa Cruz, Santa Cruz, CA, USA) and Ki67 (Novocastra Laboratories, Newcastle upon Tyne, UK). Immunocomplexes were detected using the Vectastain ABC alkaline phosphatase kit (Vector Laboratories, Burlingame, CA, USA), or using FITC- and rhodamine-conjugated secondary antibodies (Jackson Immunoresearch, West Grove, PA, USA).

Immunofluorescent-staining-guided laser capture microdissection and LOH analysis Ten-micrometer sections were deparaffinized and lightly stained with antibody against CK5. Using a PixCell lle Laser Capture Microdissection system (Arcturus, Mountain View, CA, USA), the lesions that were positively stained with CK5 were isolated from the slides. Particular care was taken to avoid contamination by the surrounding tissue. DNA isolation and LOH analysis were performed as described.18,19

Mammary cell preparation, FACS analysis, cell sorting and colonyformation assay Mammary glands were dissected from female mice at the indicated age. After mechanical dissociation, the tissue was processed as previously described.22,38,70 Oncogene (2013) 2715 – 2725


2724 Cell culture and lentivirus-mediated Brca1 knockdown 293T, MCF-7 and T47D cells were maintained in DMEM with 10% FBS. For knocking down Brca1, pGIPZ-empty and pGIPZ-shBrca1 vectors were purchased from Open Biosystems, Lafayette, CO, USA. Lentivirus supernatants were prepared following the manufacturer’s instructions, and lentiviral infections were carried out accordingly.

Western blot and quantitative real-time PCR Tissue lysates were prepared as previously reported.68 Antibodies to CK5, CK8, p18, actin (NeoMarker, Kalamazoo, MI, USA), Brca1 and tubulin (Santa Cruz) were purchased commercially. Quantification of western blot results was aided by the ImageJ software (http://rsb.info.nih.gov/ij). Total RNA was extracted by RNeasy (Qiagen, Valencia, CA, USA), and cDNA was synthesized with random hexamers by jSuperScript III first-strand synthesis system (Invitrogen, Grand Island, NY, USA). The cDNA was added to a quantitative real-time PCR mixture that contained 1 SYBR Green PCR master mix (Applied Biosystems, Foster City, CA, USA) and 70 nM gene-specific primers. Assays were performed in triplicate on a 7900HT sequence detection system (Applied Biosystems). The expression level of each gene was normalized with GAPDH. The specific PCR primer sequences for genotyping and qPCR are available upon request.

CONFLICT OF INTEREST The authors declare no conflict of interest.

ACKNOWLEDGEMENTS This project was initiated at the Department of Biochemistry and Biophysics, University of North Carolina at Chapel Hill. We thank Dr Yue Xiong for his invaluable support, discussion and critical reading of the manuscript, Dr Beverly Koller for providing Brca1 germline mutant mice and Drs Anthony Capobianco and Xiangxi Xu for discussions. This study was supported in part by a DOD Idea Award (W81XWH-101-0302) and startup funds from the University of Miami Miller School of Medicine to XHP.

REFERENCES 1 Althuis MD, Fergenbaum JH, Garcia-Closas M, Brinton LA, Madigan MP, Sherman ME. Etiology of hormone receptor-defined breast cancer: a systematic review of the literature. Cancer Epidemiol Biomarkers Prev 2004; 13: 1558–1568. 2 Arnes JB, Brunet JS, Stefansson I, Begin LR, Wong N, Chappuis PO et al. Placental cadherin and the basal epithelial phenotype of BRCA1-related breast cancer. Clin Cancer Res 2005; 11: 4003–4011. 3 Foulkes WD, Brunet JS, Stefansson IM, Straume O, Chappuis PO, Begin LR et al. The prognostic implication of the basal-like (cyclin E high/p27 low/p53 þ /glomeruloid-microvascular-proliferation þ ) phenotype of BRCA1-related breast cancer. Cancer Res 2004; 64: 830–835. 4 Lakhani SR, Reis-Filho JS, Fulford L, Penault-Llorca F, van der Vijver M, Parry S et al. Prediction of BRCA1 status in patients with breast cancer using estrogen receptor and basal phenotype. Clin Cancer Res 2005; 11: 5175–5180. 5 Ribeiro-Silva A, Ramalho LN, Garcia SB, Brandao DF, Chahud F, Zucoloto S. p63 correlates with both BRCA1 and cytokeratin 5 in invasive breast carcinomas: further evidence for the pathogenesis of the basal phenotype of breast cancer. Histopathology 2005; 47: 458–466. 6 Foulkes WD. BRCA1 functions as a breast stem cell regulator. J Med Genet 2004; 41: 1–5. 7 Liu S, Ginestier C, Charafe-Jauffret E, Foco H, Kleer CG, Merajver SD et al. BRCA1 regulates human mammary stem/progenitor cell fate. Proc Natl Acad Sci USA 2008; 105: 1680–1685. 8 Lim E, Vaillant F, Wu D, Forrest NC, Pal B, Hart AH et al. Aberrant luminal progenitors as the candidate target population for basal tumor development in BRCA1 mutation carriers. Nat Med 2009; 15: 907–913. 9 Proia TA, Keller PJ, Gupta PB, Klebba I, Jones AD, Sedic M et al. Genetic predisposition directs breast cancer phenotype by dictating progenitor cell fate. Cell Stem Cell 2011; 8: 149–163. 10 Furuta S, Jiang X, Gu B, Cheng E, Chen PL, Lee WH. Depletion of BRCA1 impairs differentiation but enhances proliferation of mammary epithelial cells. Proc Natl Acad Sci USA 2005; 102: 9176–9181. 11 Brzovic PS, Keeffe JR, Nishikawa H, Miyamoto K, Fox 3rd D, Fukuda M et al. Binding and recognition in the assembly of an active BRCA1/BARD1 ubiquitin-ligase complex. Proc Natl Acad Sci USA 2003; 100: 5646–5651.

Oncogene (2013) 2715 – 2725

12 Hashizume R, Fukuda M, Maeda I, Nishikawa H, Oyake D, Yabuki Y et al. The RING heterodimer BRCA1-BARD1 is a ubiquitin ligase inactivated by a breast cancerderived mutation. J Biol Chem 2001; 276: 14537–14540. 13 Cao L, Li W, Kim S, Brodie SG, Deng CX. Senescence, aging, and malignant transformation mediated by p53 in mice lacking the Brca1 full-length isoform. Genes Dev 2003; 17: 201–213. 14 Hakem R, de la Pompa JL, Sirard C, Mo R, Woo M, Hakem A et al. The tumor suppressor gene Brca1 is required for embryonic cellular proliferation in the mouse. Cell 1996; 85: 1009–1023. 15 Xu X, Wagner KU, Larson D, Weaver Z, Li C, Ried T et al. Conditional mutation of Brca1 in mammary epithelial cells results in blunted ductal morphogenesis and tumour formation. Nat Genet 1999; 22: 37–43. 16 Sherr CJ, Roberts JM. CDK inhibitors: positive and negative regulators of G1-phase progression. Genes Dev 1999; 13: 1501–1512. 17 Pei XH, Xiong Y. Biochemical and cellular mechanisms of mammalian CDK inhibitors: a few unresolved issues. Oncogene 2005; 24: 2787–2795. 18 Bai F, Pei XH, Godfrey VL, Xiong Y. Haploinsufficiency of p18(INK4c) sensitizes mice to carcinogen-induced tumorigenesis. Mol Cell Biol 2003; 23: 1269–1277. 19 Bai F, Pei XH, Pandolfi PP, Xiong Y. p18 Ink4c and Pten constrain a positive regulatory loop between cell growth and cell cycle control. Mol Cell Biol 2006; 26: 4564–4576. 20 Bai F, Pei XH, Nishikawa T, Smith MD, Xiong Y. p18Ink4c, but not p27Kip1, collaborates with Men1 to suppress neuroendocrine organ tumors. Mol Cell Biol 2007; 27: 1495–1504. 21 Pei XH, Bai F, Smith MD, Xiong Y. p18Ink4c collaborates with Men1 to constrain lung stem cell expansion and suppress non-small-cell lung cancers. Cancer Res 2007; 67: 3162–3170. 22 Pei XH, Bai F, Smith MD, Usary J, Fan C, Pai SY et al. CDK inhibitor p18(INK4c) is a downstream target of GATA3 and restrains mammary luminal progenitor cell proliferation and tumorigenesis. Cancer Cell 2009; 15: 389–401. 23 Zindy F, Nilsson LM, Nguyen L, Meunier C, Smeyne RJ, Rehg JE et al. Hemangiosarcomas, medulloblastomas, and other tumors in Ink4c/p53-null mice. Cancer Res 2003; 63: 5420–5427. 24 Wiedemeyer R, Brennan C, Heffernan TP, Xiao Y, Mahoney J, Protopopov A et al. Feedback circuit among INK4 tumor suppressors constrains human glioblastoma development. Cancer Cell 2008; 13: 355–364. 25 Leone PE, Walker BA, Jenner MW, Chiecchio L, Dagrada G, Protheroe RK et al. Deletions of CDKN2C in multiple myeloma: biological and clinical implications. Clin Cancer Res 2008; 14: 6033–6041. 26 Bignell GR, Greenman CD, Davies H, Butler AP, Edkins S, Andrews JM et al. Signatures of mutation and selection in the cancer genome. Nature 2010; 463: 893–898. 27 Lapointe J, Lachance Y, Labrie Y, Labrie C. Ap18 mutant defective in CDK6 binding in human breast cancer cells. Cancer Res 1996; 56: 4586–4589. 28 Spirin K, Simpson J, Miller C, Koeffler H. Molecular analysis of INK4 genes in breast carcinomas. Int J Oncol 1997; 11: 737–744. 29 Deng CX, Scott F. Role of the tumor suppressor gene Brca1 in genetic stability and mammary gland tumor formation. Oncogene 2000; 19: 1059–1064. 30 Cao L, Kim S, Xiao C, Wang RH, Coumoul X, Wang X et al. ATM-Chk2-p53 activation prevents tumorigenesis at an expense of organ homeostasis upon Brca1 deficiency. EMBO J 2006; 25: 2167–2177. 31 Somasundaram K, Zhang H, Zeng YX, Houvras Y, Peng Y, Wu GS et al. Arrest of the cell cycle by the tumour-suppressor BRCA1 requires the CDK-inhibitor p21WAF1/ CiP1. Nature 1997; 389: 187–190. 32 Lee YH, Bedford MT, Stallcup MR. Regulated recruitment of tumor suppressor BRCA1 to the p21 gene by coactivator methylation. Genes Dev 2011; 25: 176–188. 33 Xu X, Qiao W, Linke SP, Cao L, Li WM, Furth PA et al. Genetic interactions between tumor suppressors Brca1 and p53 in apoptosis, cell cycle and tumorigenesis. Nat Genet 2001; 28: 266–271. 34 Ramsey MR, Krishnamurthy J, Pei XH, Torrice C, Lin W, Carrasco DR et al. Expression of p16Ink4a compensates for p18Ink4c loss in cyclin-dependent kinase 4/6-dependent tumors and tissues. Cancer Res 2007; 67: 4732–4741. 35 Kitagawa M, Higashi H, Jung HK, Suzuki-Takahashi I, Ikeda M, Tamai K et al. The consensus motif for phosphorylation by cyclin D1-Cdk4 is different from that for phosphorylation by cyclin A/E-Cdk2. EMBO J 1996; 15: 7060–7069. 36 Zarkowska TUS, Harlow E, Mittnacht S. Monoclonal antibodies specific for underphosphorylated retinoblastoma protein identify a cell cycle regulated phosphorylation site targeted by CDKs. Oncogene 1997; 14: 249–254. 37 Bennett LM, McAllister KA, Malphurs J, Ward T, Collins NK, Seely JC et al. Mice heterozygous for a Brca1 or Brca2 mutation display distinct mammary gland and ovarian phenotypes in response to diethylstilbestrol. Cancer Res 2000; 60: 3461–3469. 38 Shackleton M, Vaillant F, Simpson KJ, Stingl J, Smyth GK, Asselin-Labat ML et al. Generation of a functional mammary gland from a single stem cell. Nature 2006; 439: 84–88.



2725 39 Van Keymeulen A, Rocha AS, Ousset M, Beck B, Bouvencourt G, Rock J et al. Distinct stem cells contribute to mammary gland development and maintenance. Nature 2011; 479: 189–193. 40 Sleeman KE, Kendrick H, Robertson D, Isacke CM, Ashworth A, Smalley MJ. Dissociation of estrogen receptor expression and in vivo stem cell activity in the mammary gland. J Cell Biol 2007; 176: 19–26. 41 Bernardo GM, Lozada KL, Miedler JD, Harburg G, Hewitt SC, Mosley JD et al. FOXA1 is an essential determinant of ERalpha expression and mammary ductal morphogenesis. Development 2010; 137: 2045–2054. 42 Yamaji D, Na R, Feuermann Y, Pechhold S, Chen W, Robinson GW et al. Development of mammary luminal progenitor cells is controlled by the transcription factor STAT5A. Genes Dev 2009; 23: 2382–2387. 43 Larue L, Ohsugi M, Hirchenhain J, Kemler R. E-cadherin null mutant embryos fail to form a trophectoderm epithelium. Proc Natl Acad Sci USA 1994; 91: 8263–8267. 44 Stingl J, Raouf A, Emerman JT, Eaves CJ. Epithelial progenitors in the normal human mammary gland. J Mammary Gland Biol Neoplasia 2005; 10: 49–59. 45 Smart CE, Clarke C, Brooks KM, Raghavendra A, Brewster BL, French JD et al. Targeted disruption of Brca1 in restricted compartments of the mouse mammary epithelia. Breast Cancer Res Treat 2008; 112: 237–241. 46 Smart CE, Wronski A, French JD, Edwards SL, Asselin-Labat ML, Waddell N et al. Analysis of Brca1-deficient mouse mammary glands reveals reciprocal regulation of Brca1 and c-kit. Oncogene 2011; 30: 1597–1607. 47 Marquis ST, Rajan JV, Wynshaw-Boris A, Xu J, Yin GY, Abel KJ et al. The developmental pattern of Brca1 expression implies a role in differentiation of the breast and other tissues. Nat Genet 1995; 11: 17–26. 48 Lane TF, Deng C, Elson A, Lyu MS, Kozak CA, Leder P. Expression of Brca1 is associated with terminal differentiation of ectodermally and mesodermally derived tissues in mice. Genes Dev 1995; 9: 2712–2722. 49 Rajan JV, Wang M, Marquis ST, Chodosh LA. Brca2 is coordinately regulated with Brca1 during proliferation and differentiation in mammary epithelial cells. Proc Natl Acad Sci USA 1996; 93: 13078–13083. 50 Kubista M, Rosner M, Kubista E, Bernaschek G, Hengstschlager M. Brca1 regulates in vitro differentiation of mammary epithelial cells. Oncogene 2002; 21: 4747–4756. 51 Shakya R, Reid LJ, Reczek CR, Cole F, Egli D, Lin CS et al. BRCA1 tumor suppression depends on BRCT phosphoprotein binding, but not its E3 ligase activity. Science 2011; 334: 525–528. 52 Buckley NE, Mullan PB. BRCA1 - Conductor of the Breast Stem Cell Orchestra: The Role of BRCA1 in Mammary Gland Development and Identification of Cell of Origin of BRCA1 Mutant Breast Cancer. Stem Cell Rev 2012 (e-pub ahead of print; doi:10.1007/s12015-012-9354-y). 53 Harte MT, O’Brien GJ, Ryan NM, Gorski JJ, Savage KI, Crawford NT et al. BRD7, a subunit of SWI/SNF complexes, binds directly to BRCA1 and regulates BRCA1dependent transcription. Cancer Res 2010; 70: 2538–2547. 54 Gorski JJ, James CR, Quinn JE, Stewart GE, Staunton KC, Buckley NE et al. BRCA1 transcriptionally regulates genes associated with the basal-like phenotype in breast cancer. Breast Cancer Res Treat 2010; 122: 721–731.

55 Mullan PB, Quinn JE, Harkin DP. The role of BRCA1 in transcriptional regulation and cell cycle control. Oncogene 2006; 25: 5854–5863. 56 Rosen EM, Fan S, Ma Y. BRCA1 regulation of transcription. Cancer Lett 2006; 236: 175–185. 57 Williamson EA, Wolf I, O’Kelly J, Bose S, Tanosaki S, Koeffler HP. BRCA1 and FOXA1 proteins coregulate the expression of the cell cycle-dependent kinase inhibitor p27(Kip1). Oncogene 2006; 25: 1391–1399. 58 Hosey AM, Gorski JJ, Murray MM, Quinn JE, Chung WY, Stewart GE et al. Molecular basis for estrogen receptor alpha deficiency in BRCA1-linked breast cancer. J Natl Cancer Inst 2007; 99: 1683–1694. 59 Drost RM, Jonkers J. Preclinical mouse models for BRCA1-associated breast cancer. Br J Cancer 2009; 101: 1651–1657. 60 Honrado E, Benitez J, Palacios J. The molecular pathology of hereditary breast cancer: genetic testing and therapeutic implications. Mod Pathol 2005; 18: 1305–1320. 61 Suspitsin EN, Sokolenko AP, Voskresenskiy DA, Ivantsov AO, Shelehova KV, Klimashevskiy VF et al. Mixed epithelial/mesenchymal metaplastic carcinoma (carcinosarcoma) of the breast in BRCA1 carrier. Breast Cancer 2011; 18: 137–140. 62 Polyak K, Weinberg RA. Transitions between epithelial and mesenchymal states: acquisition of malignant and stem cell traits. Nat Rev Cancer 2009; 9: 265–273. 63 Tkocz D, Crawford NT, Buckley NE, Berry FB, Kennedy RD, Gorski JJ et al. BRCA1 and GATA3 corepress FOXC1 to inhibit the pathogenesis of basal-like breast cancers. Oncogene 2012; 31: 3667–3678. 64 Molyneux G, Geyer FC, Magnay FA, McCarthy A, Kendrick H, Natrajan R et al. BRCA1 basal-like breast cancers originate from luminal epithelial progenitors and not from basal stem cells. Cell Stem Cell 2010; 7: 403–417. 65 Wright MH, Robles AI, Herschkowitz JI, Hollingshead MG, Anver MR, Perou CM et al. Molecular analysis reveals heterogeneity of mouse mammary tumors conditionally mutant for Brca1. Mol Cancer 2008; 7: 29. 66 Herschkowitz JI, Simin K, Weigman VJ, Mikaelian I, Usary J, Hu Z et al. Identification of conserved gene expression features between murine mammary carcinoma models and human breast tumors. Genome Biol 2007; 8: R76. 67 Herschkowitz JI, Zhao W, Zhang M, Usary J, Murrow G, Edwards D et al. Breast Cancer Special Feature: Comparative oncogenomics identifies breast tumors enriched in functional tumor-initiating cells. Proc Natl Acad Sci USA 2011; 109: 2778–2783. 68 Franklin DS, Godfrey VL, Lee H, Kovalev GI, Schoonhoven R, Chen-Kiang S et al. CDK inhibitors p18INK4c and p27KIP1 mediate two separate pathways to collaboratively suppress pituitary tumorigenesis. Genes Dev 1998; 12: 2899–2911. 69 Gowen LC, Johnson BL, Latour AM, Sulik KK, Koller BH. Brca1 deficiency results in early embryonic lethality characterized by neuroepithelial abnormalities. Nat Genet 1996; 12: 191–194. 70 Stingl J, Eirew P, Ricketson I, Shackleton M, Vaillant F, Choi D et al. Purification and unique properties of mammary epithelial stem cells. Nature 2006; 439: 993–997.

Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc)


Oncogene (2013) 2715 – 2725