Tamoxifen inhibition of estrogen receptor-alpha-negative mouse mammary tumorigenesis.

Tamoxifen reduces the relative risk of breast cancer developing from specific premalignant lesions. Many breast cancers that arise after tamoxifen treatment are estrogen receptor-alpha (ER-alpha)-negative, although premalignant lesions such as atypical ductal hyperplasia are highly ER-alpha-positive. The p53 null mouse mammary epithelial transplant model is characterized by ER-alpha-positive premalignant lesions that give rise to both ER-alpha-positive and ER-alpha-negative tumors. Given this progression from ER-alpha-positive to ER-alpha-negative lesions, we tested the ability of tamoxifen to block or delay mammary tumorigenesis in several versions of this model. In groups 1 and 2, p53 null normal mammary epithelial transplants were maintained in virgin mice. In groups 3 to 5, the p53 null and mammary transplants were maintained in mice continuously exposed to high levels of progesterone. In groups 6 and 7, transplants of the premalignant outgrowth line PN8a were maintained in virgin mice. Tamoxifen blocked estrogen signaling in these mice as evidenced by decreases in progesterone-induced lateral branching and epithelial proliferation in the mammary epithelium. Tamoxifen did not alter the elevated levels of progesterone in the blood while significantly reducing the circulating level of prolactin. Tamoxifen reduced tumor incidence in p53 null normal mammary epithelial transplants maintained in virgin mice from 55% to 5% and in progesterone-stimulated mice from 81% to 21%. The majority of the resultant tumors were ER-alpha-negative. Tamoxifen also significantly delayed tumorigenesis in the ER-alpha-positive high premalignant line PN8a from 100% to 75%. These results show that tamoxifen delays the emergence of ER-alpha-negative tumors if given early in premalignant progression.

Hormone-induced chromosomal instability in p53-null mammary epithelium.

The absence of p53 function increases risk for spontaneous tumorigenesis in the mammary gland. Hormonal stimulation enhances tumor risk in p53-null mammary epithelial cells as well as the incidence of aneuploidy. Aneuploidy appears in normal p53-null mammary epithelial cells within 5 weeks of hormone stimulation. Experiments reported herein assessed a possible mechanism of hormone-induced aneuploidy. Hormones increased DNA synthesis equally between wild-type (WT) and p53-null mammary epithelial cells. There were two distinct responses in p53-null cells to hormone exposure. First, Western blot analysis demonstrated that the levels of two proteins involved in regulating sister chromatid separation and the spindle checkpoint, Mad2 and separase (ESPL1) were increased in null compared with WT cells. In contrast, the levels of securin and Rad21 proteins were not increased in hormone-stimulated p53-null compared with WT cells. ESPL1 RNA was also increased in p53-null mouse mammary cells in vivo by 18 h of hormone stimulation and in human breast MCF7 cells in monolayer culture by 8 h of hormone stimulation. Furthermore, both promoters contained p53 and steroid hormone response elements. Mad2 protein was increased as a consequence of the absence of p53 function. The increase in Mad2 protein was observed also at the cellular level by immunohistochemistry. Second, hormones increased gene amplication in the distal arm of chromosome 2, as shown by comparative genomic hybridization. These results support the hypothesis that hormone stimulation acts to increase aneuploidy by several mechanisms. First, by increasing mitogenesis in the absence of the p53 checkpoint in G2, hormones allow the accumulation of cells that have experienced chromosome missegregation. Second, the absolute rate of chromosome missegregation may be increased by alterations in the levels of two proteins, separase and Mad2, which are important for maintaining chromosomal segregation and the normal spindle checkpoint during mitosis.

p19ARF determines the balance between normal cell proliferation rate and apoptosis during mammary gland development.

This study demonstrated, for the first time, the following events related to p19(ARF) involvement in mammary gland development: 1) Progesterone appears to regulate p19(ARF) in normal mammary gland during pregnancy. 2) p19(ARF) expression levels increased sixfold during pregnancy, and the protein level plateaus during lactation. 3) During involution, p19(ARF) protein level remained at high levels at 2 and 8 days of involution and then, declined sharply at day 15. Absence of p19(ARF) in mammary epithelial cells leads to two major changes, 1) a delay in the early phase of involution concomitant with downregulation of p21(Cip1) and decrease in apoptosis, and 2) p19(ARF) null cells are immortal in vivo measured by serial transplantion, which is partly attributed to complete absence of p21(Cip1) compared with WT cells. Although, p19(ARF) is dispensable in mammary alveologenesis, as evidenced by normal differentiation in the mammary gland of pregnant p19(ARF) null mice, the upregulation of p19(ARF) by progesterone in the WT cells and the weakness of p21(Cip1) in mammary epithelial cells lacking p19(ARF) strongly suggest that the functional role(s) of p19(ARF) in mammary gland development is critical to sustain normal cell proliferation rate during pregnancy and normal apoptosis in involution possibly through the p53-dependent pathway.

Hormone dependence in premalignant mammary progression.

Human breast cancers that are estrogen receptor (ER) negative convey a poor prognosis for patient survival. A mouse model that mimics essential biological and genetic attributes of a subset of human breast cancer is the BALB/c p53-null mammary epithelium, in which deletion of the tumor suppressor gene p53 results in enhanced tumorigenic risk. The experiments reported herein examine the hormone dependence of premalignant mammary progression in this model. The p53-null normal mammary epithelium exhibits the same dependence as p53 wild-type mammary epithelium on ovarian hormones for growth. However, in contrast to p53 wild-type epithelium, estrogen and progesterone, singly or in combination, strongly enhance tumorigenesis in p53-null mammary epithelium. The removal of progesterone signaling by deletion of the progesterone receptor eliminates progesterone enhancement of tumorigenesis. The immortalized premalignant outgrowth lines, termed PN, possess different tumorigenic capabilities, but the majority of these lines showed a strong dependence on ovarian hormones for growth and tumorigenesis. Although these lines are highly ER positive, a large number of tumors arising from these lines were ER negative and grew when implanted in ovariectomized mice. As was the case for p53-null normal mammary cells, hormonal stimulation was a strong promoter for tumorigenesis in the premalignant outgrowth lines and, surprisingly, was much stronger than the chemical carcinogen 7,12-dimethylbenzanthracene. In summary, these results demonstrate that p53-null mammary cells, which generate a significant percentage of ER-negative tumors, are highly responsive to the absence or presence of ovarian hormones during the normal and premalignant stages. This model would appear an excellent one to test the effects of chemopreventive agents on the development of both ER-negative and ER-positive mammary tumors.

Biological and genetic properties of the p53 null preneoplastic mammary epithelium.

The absence of the tumor suppressor gene p53 confers an increased tumorigenic risk for mammary epithelial cells. In this report, we describe the biological and genetic properties of the p53 null preneoplastic mouse mammary epithelium in a p53 wild-type environment. Mammary epithelium from p53 null mice was transplanted serially into the cleared mammary fat pads of p53 wild-type BALB/c female to develop stable outgrowth lines. The outgrowth lines were transplanted for 10 generations. The outgrowths were ductal in morphology and progressed through ductal hyperplasia and ductal carcinoma in situ before invasive cancer. The preneoplastic outgrowth lines were immortal and exhibited activated telomerase activity. They are estrogen and progesterone receptor-positive, and aneuploid, and had various levels of tumorigenic potential. The biological and genetic properties of these lines are distinct from those found in most hyperplastic alveolar outgrowth lines, the form of mammary preneoplasia occurring in most traditional models of murine mammary tumorigenesis. These results indicate that the preneoplastic cell populations found in this genetically engineered model are similar in biological properties to a subset of precurser lesions found in human breast cancer and provide a unique model to identify secondary events critical for tumorigenicity and invasiveness.