Introduction
Breast cancer is the most frequent type of cancer among women in the US and ranks as the second leading cause of cancer death [
1]. Reproductive history predicts the potential for breast cancer development. For instance, early menarche or late menopause increases lifetime exposure to estrogen and both are associated with increased breast cancer risk [
2,
3]. However, a full-term pregnancy early in reproductive life reduces breast cancer incidence in women by up to 50% [
4,
5]. Pregnancy also reduces the incidence of carcinogen-induced mammary tumors in rodents [
6]. Treatment with estrogen and progesterone (E+P) to mimic serum levels during pregnancy is sufficient to reduce the incidence of carcinogen-induced mammary tumors in rodents [
7‐
9].
Although E+P can render the mammary epithelium resistant to tumors, the underlying mechanisms mediating the protective effect of parity are unknown. It has been proposed that the bolus of ovarian hormones during pregnancy initiates persistent systemic changes in the parous individual. Reductions in growth hormone and prolactin have been associated with decreased breast cancer risk [
10‐
12]. Serum estrogen is also decreased in parous rodents [
13]. These changes in the hypothalamic-hypophysial axis would limit the proliferative stimulus to preneoplastic cells and, therefore, reduce the promotional environment for cancer.
Parity also induces changes within the mammary gland, which may prevent tumors. Pregnancy causes morphological changes, including differentiation of the epithelial tree in preparation for lactation, which could alter tumor susceptibility [
14]. However, differentiation alone is not sufficient, because prolactin stimulates morphological differentiation but does not render the mammary gland refractory to carcinogen-induced tumors [
15]. In addition to morphological differentiation, pregnancy causes permanent changes in the epithelial cells and stroma of the mammary gland [
16‐
18]. Differential expression of growth-regulatory genes such as amphiregulin, pleiotrophin, insulin-like growth factor-1, and transforming growth factor-beta-3 has been reported in parous mammary tissue [
19,
20] and may limit carcinogenesis in the mammary epithelium.
While alteration in these growth factor signaling pathways may mediate signals from E+P receptors, the p53 pathway appears to be a crucial downstream effector. Both p53 and its downstream transcriptional target,
Cdkn1a (also known as p21/WAF1), are increased in parous and E+P-treated mammary epithelium in response to 7,12-diamethylbenz [a]anthracene [
21]. In the absence of p53, the protection afforded by parity or exogenous E+P is lost [
22,
23]. Exposure to ionizing radiation causes DNA double-strand breaks and is a risk factor for breast cancer in women [
24,
25]. Furthermore, p53-dependent responses to ionizing radiation are modest in mammary epithelium of nulliparous mice [
26] but are enhanced during pregnancy and by treatment with E+P for 3 days [
27]. Breast cancer is the most common tumor among women with Li-Fraumeni syndrome, a syndrome associated with heterozygous mutations in the gene encoding p53 protein [
28,
29], highlighting the important role of p53 in susceptibility to breast cancer. Whether the susceptibility to breast cancer in Li-Fraumeni syndrome is due to increased risk of loss of heterozygosity for
TP53 or due to diminished p53 activity is unclear.
In these experiments, we demonstrate that p53-dependent apoptosis is increased by both parity and E+P treatment and that p53 sensitivity is retained after withdrawal of systemic hormones. Hormone-stimulated p53-dependent apoptosis and hormone-induced protection were evident even when p53 activity was reduced because of haploinsufficiency in BALB/c-Trp53+/- mice. Consistent with the increased responsiveness of p53, the latency of spontaneous mammary tumors in BALB/c-Trp53+/- mice was increased among parous mice compared with nulliparous mice. However, neither enhanced p53 responsiveness nor suppression of tumors was retained in outgrowths from hormone-stimulated BALB/c-Trp53+/- mammary epithelium when transplanted into nulliparous hosts.
Discussion
The significant role of p53 in tumor suppression in the mammary epithelium was demonstrated by the frequent occurrence of spontaneous tumors in transplants of BALB/c-
Trp53-/- mammary epithelium [
22]. As p53 protein promotes cell cycle arrest and apoptosis, both can play significant roles in tumor suppression. However, only the proapoptotic function is required to prevent lymphomas whereas the cell cycle checkpoint activity is dispensable [
33]. Therefore, enhancement of apoptosis in the mammary epithelium appears to be a critical activity for suppression of tumors. Both radiation-induced and spontaneous apoptosis were enhanced in parous and E+P-treated mammary epithelium (Figures
3c and
3d). As radiation-induced apoptosis was 2.5-fold higher in
Trp53+/+ mammary tissues than the
Trp53-/- after hormonal stimulation (Figure
6a), the majority of apoptosis is p53-dependent. Though lower in magnitude, apoptosis in the mammary epithelium of
Trp53-/- mice was also increased by E+P treatment (Figure
6a). Therefore, in addition to enhancing p53 function, hormonal exposures appear to alter the promotional environment of the host to enhance p53-independent apoptosis pathways [
34], both of which may contribute to tumor suppression.
Inducible expression of p53 in mouse models has shown that even short-term expression of p53 can substantially delay the appearance of lymphomas and liver carcinomas [
35,
36]. Therefore, the duration of the enhancement of p53 was of interest. Treatment with E+P for 14 days stimulated changes in the mammary epithelium such that p53-dependent apoptosis in the mammary epithelium persisted even after ovariectomy, indicating that continued exposure to ovarian hormones is not necessary to maintain p53 activity. These results imply that neither the nulliparous nor parous glands are necessarily more vulnerable to genotoxic stress at one phase of the estrus cycle or another. Rather, there is an enhancement of apoptosis that is sustained long after pregnancy and even after cessation of ovarian function.
Breast cancer is the most common tumor among females with Li-Fraumeni syndrome, which is most often associated with heterozygous mutations in the gene encoding p53 [
28], underscoring the prominent role of
TP53 in determining breast cancer susceptibility. However, it is not clear whether the risk of breast cancer among women with Li-Fraumeni syndrome is due to reduced activity of p53 or the increased risk of losing the wild-type allele of
TP53. As levels of radiation-induced apoptosis in
Trp53+/- mouse mammary epithelium were intermediate to the responses in
Trp53+/+ and
Trp53-/- tissues (Figure
6a),
Trp53+/- mammary tissues are haploinsufficient with respect to apoptosis. This intermediate apoptotic response phenotype is consistent with apoptosis observed in
Trp53+/- mice during involution of the prostate following castration [
37]. Stress responses were also attenuated in HCT116 cells that were heterozygous for
TP53 compared with the wild-type cells [
38]. Thus, the activity of p53 may fall below a threshold in
Trp53+/- mammary epithelium, as proposed by Santarosa and Ashworth [
39], to a point that is no longer sufficient to engage appropriate apoptotic responses to genotoxic stress. The appearance of spontaneous mammary tumors was delayed among parous BALB/c-
Trp53+/- mice (Figure
6b). Therefore, the activity of p53 is reduced to critical levels in the mammary epithelium of BALB/c-
Trp53+/- mice, but hormonal stimulation increases p53-dependent apoptosis and renders the tissue more resistant to tumorigenesis. As parity did not prevent tumors in BALB/c-
Trp53-/- mammary epithelial transplants [
22], the decrease in mammary tumors among parous BALB/c-
Trp53+/- (Figure
6b) appears to be largely p53-dependent.
The mechanism by which E+P sensitizes the p53 pathway is initiated by their respective receptors [
27]. In women, the protective effects of parity are limited to the ERα
+/progesterone receptor-positive tumors [
40]. Thus, we investigated whether parity diminished the proportion of ERα
+ tumors in BALB/c-
Trp53+/- mice. In contrast to expectations, the proportion of ERα
+ tumors was increased among parous mice (Table
1). ERα
+ mammary tumors were also observed in C57BL/6J mice rendered p53-deficient by inducible deletion by WAP/Cre [
41]. However, these results should be interpreted with caution because ERα status is confounded with age of mice. Since mammary tumors in BALB/c-
Trp53+/- mice progress from ERα
+ lesions to ERα
- tumors (K.A. Dunphy, A.C. Blackburn, H. Yan, L.R. O'Connell, D.J. Jerry, unpublished data), the ERα expression observed among mammary tumors in parous mice may reflect an earlier stage of progression due to the longer latencies. Additional markers to distinguish basal-like and luminal properties will be needed in order to ascertain whether the cellular origins differ for the mammary tumors in nulliparous and parous BALB/c-
Trp53+/- mice.
Decreased expression of ERα and increased expression of ERβ are persistent changes observed in mammary epithelium of parous rodents [
8]. ERα and ERβ exert antagonistic effects on p53 activity [
42,
43]. This could explain the increased p53-dependent apopotosis in hormone-stimulated mammary tissues. Alternatively, the persistent phenotypic changes in ERs and responsiveness of p53 may reflect alterations in the fates of the progenitor cell pool. Indeed, parity-induced mammary epithelial cells have been identified and serve as progenitors of the lobulo-alveolar structures during subsequent pregnancies [
44,
45]. However, outgrowths of
Trp53+/- mammary epithelium from E+P-treated and parous mice failed to retain the elevated apoptotic response to radiation in nulliparous hosts (Figure
9). The characteristics of the outgrowths were not different from nulliparous transplants or nulliparous glands (Figure
6a). Likewise, the outgrowths of E+P-treated
Trp53+/- mammary epithelium developed tumors with frequency and latency similar to those of the
Trp53+/- transplants from nulliparous mice (Figure
9).
These observations do not preclude the possibility that the fates or number of stem cells may be decreased within the mammary glands of hormone-stimulated mice compared with nulliparous mice. It remains possible that parity decreases the pool of progenitor cells [
46] but does not limit the capacity of the stem cells to repopulate upon transplantation. It is equally possible that hormonal stimulation imparts a stable change in the stromal cells and extracellular matrix which modulate p53 function. The extracellular matrix from involuting glands can promote tumors [
17,
18] but also is essential for differentiation [
47,
48], and parous stroma decreased the incidence of tumors [
49]. Many extracellular matrix proteins change dramatically across developmental phases and can alter integrin-mediated signaling in the mammary gland [
19,
50,
51]. Loss of integrin signaling results in apoptosis in the mammary epithelium [
52]. Therefore, hormonal stimulation may result in stable changes within the stroma which limit either tumorigenic transformation of progenitor cells or their progression to invasive tumors.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
KAD was responsible for designing and carrying out the bulk of the experiments as well as drafting the manuscript. ACB and HY provided the data on tumor phenotypes in Trp53+/- mice. LRO participated in the transplant studies. DJJ was responsible for the design and coordination of the experiments as well as preparation of the manuscript. All authors read and approved the final manuscript.