Introduction
The vast majority of human cancers arise from epithelial tissues. Understanding cancer development therefore relies on defining the molecules and pathways that support the normal structure and function of epithelia, and how these are disrupted during carcinogenesis. The epithelium of the mammary gland represents a particularly interesting model for epithelial function, as it undergoes recurring dynamic alterations to hew an architecture and physiology specific for different developmental stages. Repeated cycles of mammary epithelial cell proliferation, differentiation, and apoptosis contribute to remodeling of the mammary epithelium, dictating the specific structure and function of the mammary gland throughout morphogenesis. Proper regulation of the processes controlling mammary epithelial form and function is critical, as its dysregulation can impair the normal architecture or behavior of the mammary epithelium, leading to cancer development [
1].
Proper adhesion of cells to one another as well as to the substratum is the key for promoting the polarized organization and integrity of the mammary epithelium [
2]. Cell- in epithelia is facilitated by two major classes of protein complexes: adherens junctions and desmosomes. Adherens junctions promote the cell-cell adhesion required for organizing epithelial cells into sheets and are fundamental for allowing dynamic rearrangements of epithelia necessary for maintaining homeostasis. In the mammary gland, adherens junctions are important for normal tissue development, function, and organization [
3‐
6]. Desmosomes, which are particularly important for reinforcing intercellular adhesion and enabling tissues to resist mechanical stresses, are crucial for promoting adhesion and epithelial integrity in the epidermis. Desmosomes comprise transmembrane cadherin proteins [desmogleins (Dsg) 1 to 4 and desmocollins (Dsc) 1 to 3] that form an adhesive interface at the cell surface, and proteins such as plakoglobin (Pg) and desmoplakin (Dp), which indirectly connect cadherins to the intermediate filament cytoskeleton [
7]. Although desmosomes are present within and between cells in both layers of the mammary epithelium [
8,
9], the contribution of desmosomes to normal mammary epithelial morphogenesis and function has remained unclear. Cell-culture experiments have suggested a role for desmosomal cadherin-mediated adhesion in some aspects of mammary epithelial morphogenesis and organization [
9], but desmosome function in the mammary gland has not been examined
in vivo.
In addition to regulating normal function of tissues, cell-cell adhesion is well established as a tumor-suppression mechanism. Extensive genetic data from mouse models and human cancers have demonstrated a role for adherens junctions in suppressing cancer progression [
10], and reduced expression of adherens junction proteins is often associated with poor clinical outcome for cancer patients [
11‐
15]. In contrast, the role for desmosomes in cancer is less well understood, although more and more evidence supports the idea that desmosomes can inhibit carcinogenesis [
16]. In particular, two recent physiologically relevant studies of mice with conditional knockout of desmosomal component-encoding genes indicate a key role for desmosomes in suppressing cancer [
17,
18].
Perp (
p53 apoptosis
effector
related to
PMP-22) is a tetraspan-membrane protein essential for desmosome function in the skin and oral mucosa [
19,
20].
Perp was originally identified as a transcriptional target of the p53 tumor suppressor induced specifically during apoptosis and was subsequently found to be regulated by another member of the p53 family, p63 [
20], a master regulator of the development of stratified epithelia, such as epidermis [
21‐
23].
Perp-null mice exhibit severe blistering of the skin and oral cavity, which likely contributes to their early postnatal lethality. This blistering phenotype is reminiscent of human diseases or mouse models in which desmosome-mediated intercellular adhesion is compromised [
24,
25]. Immunogold electron microscopy unequivocally demonstrated the localization of Perp to desmosome junctions, and ultrastructurally abnormal desmosomes were observed in the absence of Perp [
20]. Thus, it is clear that Perp plays a crucial role in promoting tissue integrity and homeostasis in the epidermis and oral cavity. This essential role has been underscored by our recent demonstration that conditional deletion of
Perp in mouse skin promotes UVB-induced skin cancer development and progression [
17].
Because the mammary epithelium relies on p63 for proper development [
22,
23] and because adhesion programs are critical for its morphogenesis [
26], we hypothesized that Perp may contribute to mammary gland morphogenesis and function. Here, we tested this hypothesis by characterizing Perp expression in the normal mammary epithelium and examining the consequences of Perp loss in mammary gland development. Because the identified functions of Perp in apoptosis and intercellular adhesion, as well as its identification as a suppressor of skin cancer, suggest a potential role for Perp in breast cancer suppression, we additionally examined this possibility by using a mouse mammary cancer model. Collectively, our findings reveal a role for Perp in mammary gland homeostasis and tumor suppression, thus advancing our understanding of the contribution of desmosomes to these processes.
Discussion
We previously identified Perp as a crucial component of desmosomes in the oral mucosa and epidermis, where it acts to promote tissue integrity [
20], and as a mediator of p53-dependent apoptosis in response to genotoxic stress [
17]. To gain a greater understanding of the function of Perp in tissue homeostasis in other contexts, we investigated a role for Perp in the mammary epithelium, another p63-dependent tissue in which cell adhesion and apoptosis are key for development and homeostasis. Here, we provided the first demonstration that Perp protein is expressed in the mammary epithelium, in a pattern suggesting an adhesive role for Perp in the mammary epithelium, as in the oral epithelium and skin. In addition, our results demonstrate that
Perp deficiency both perturbs mammary epithelial homeostasis and promotes mammary cancer.
In the mammary epithelium, Perp exhibits a punctate pattern of expression at the plasma membrane and colocalizes with other desmosome components. Our results suggest that Perp is expressed in both layers of the mammary epithelium, in luminal and myoepithelial cells. Perp may be an important factor in regulating desmosome assembly and/or stability in this context, as we found that Perp loss in mammary epithelial cells decreased levels of the desmosome proteins Dp, Dsg, and Dsc, compared with those in wild-type cells. These findings suggest that desmosome function in the mammary epithelium is likely compromised in the absence of Perp.
Akin to global Perp deficiency causing impaired desmosome function in the oral mucosa and epidermis, associated with blistering and early lethality [
20], Perp loss in the mammary epithelium also compromises mammary gland homeostasis
in vivo by promoting immune cell recruitment. We demonstrated further that
Perp-null mammary epithelium is able to realize all stages of mammary gland development without the presence of gross structural abnormalities, suggesting that
in vivo mammary gland morphogenesis
per se is not dependent on normal desmosome function. This finding contrasts with a previous study reporting that impairing desmosome-mediated adhesion by simultaneously interfering with the function of both types of desmosomal cadherins, Dsg and Dsc, inhibited mammary epithelial morphogenesis in culture, as seen by the impaired development of alveolar spheres and the inhibition of positional sorting of luminal and myoepithelial cells in aggregates [
9]. This observation could suggest that inactivation of two desmosomal cadherins may be more severe than the loss of Perp in the context of mammary epithelial development. Differences between our findings and those of this report may also be due to the experimental design used in each study: we examined primary mouse mammary epithelium
in vivo, whereas Runswick
et al. [
9] used a mouse mammary cell line in cell-culture experiments.
Because our mammary-transplant assays were performed
in vivo, we were able to examine the mammary epithelium in its native context. Strikingly, in the context of mammary transplantation, we observed abundant immune cell accumulation around the
Perp-deficient mammary epithelium in 8-week-old virgin females, but not that of wild-type counterparts. This observation is reminiscent of our previous findings that ablation of
Perp in the epidermis induced an inflammatory gene signature and that
Perp deficiency in combination with UV irradiation caused the accumulation of immune cells in the skin [
17]. Inflammatory cells are well established to promote tumor progression by producing such factors as cytokines, chemokines, and matrix metalloproteases that can enhance the malignant characteristics of neoplastic cells as well as remodeling the tumor microenvironment to support tumor growth and progression. Indeed, leukocyte accumulation increases during tumor development and progression, and leukocytes contribute to the development of many solid tumors, including breast cancer [
40]. Genetic studies in mouse models have demonstrated the importance of macrophages and T lymphocytes for breast cancer progression [
41,
42], and the presence of macrophages in human breast tumors correlates with poor prognosis [
43]. Therefore, it is possible that the lymphocyte accumulation observed in the
Perp-deficient mammary epithelium could result in microenvironmental changes that potentiate mammary epithelial tumorigenesis.
We showed previously that loss of Perp in mice promotes both the initiation and progression of UVB-induced skin cancer [
17]. To determine whether Perp contributes to tumor suppression in other epithelial cancers, we investigated the effect of
Perp deficiency on mammary carcinogenesis. To date, few studies have examined the specific connection between desmosomes and breast carcinogenesis. A couple of exceptions are studies in which the desmosomal protein Dsc3 was found to be downregulated during human breast cancer development [
44], and in which Dp expression was observed to be inversely correlated with human breast tumor growth and progression [
45]. To more directly assess how desmosomes contribute to carcinogenesis, it is important to evaluate genetic experimental models in which desmosome gene expression can be manipulated in the context of cancer. Here, we found that decreased
Perp dosage in the mammary epithelium reduces tumor-free survival and tumor latency in
K14-Cre/+;p53
fl/fl
mice. Although Perp has roles both in intercellular adhesion and p53-dependent apoptosis, the effects of
Perp deficiency on tumor latency and tumor-free survival in this mouse cancer model likely result from altered desmosome-mediated cell-cell adhesion, because the tumors are null for
p53. Collectively, our results therefore indicate that Perp can display tumor-suppressor activities in more than one type of epithelial cancer and that
Perp deficiency can promote tumor development in the context of different tumor-promoting stimuli. In the future, it would be interesting to investigate how combined targeting of adherens junctions and desmosomes would affect tumorigenesis in this model system.
Our results demonstrate reduced expression of Perp protein in a variety of human breast cancer cell lines, as compared with normal cells, suggesting the possibility that Perp downregulation may contribute to cancer progression. Interestingly, a previous study by Neve
et al. [
46] identified characteristics that led to the segregation of different normal mammary and breast cancer cell lines into functionally distinct subtypes, described as Luminal, Basal A, and Basal B. Some of the cell lines were evaluated for invasive behavior in a Boyden chamber assay, with the most samples being of the Basal B subtype. When we overlap our data describing Perp levels in these same cell lines with the Boyden chamber results of Neve
et al., it suggests that Perp levels may inversely correlate with an invasive phenotype. For example, the nontransformed MCF10A cells exhibited high Perp levels but no invasion, whereas the BT549, MDA-MB-231, and SUM149PT breast cancer cells expressed comparatively low Perp levels and displayed significant invasive activity. Together, the two studies suggest that reduced Perp expression may be one characteristic that contributes to the invasive behavior of breast cancer cells and breast cancer progression
in vivo.
Analyses of human cancers have suggested the importance of Perp as a prognostic marker, as
Perp downregulation is associated with particularly aggressive uveal melanomas [
47] and is predictive for esophageal cancers that will fail to respond efficiently to preoperative combination chemotherapy and radiation treatment [
48]. Moreover, Perp loss correlates with increased rates of local relapse in human head and neck squamous cell carcinoma patients (personal communication, Quynh-Thu Le, M.D.). Although patient outcome has not yet been correlated with Perp expression levels in human breast cancer, human
Perp (also known as
THW) is downregulated in human mammary carcinoma cell lines compared with nonmalignant mammary epithelium [
49]. In addition, the chromosomal region to which human
Perp maps -- 6q24 -- is deleted in human breast cancer, and loss of heterozygosity at this region has been detected both in breast carcinoma cell lines and in human breast tumors [
49,
50]. Future studies will better elaborate the mechanisms by which Perp suppresses epithelial cancers and will evaluate Perp as a prognostic indicator or therapeutic target in breast cancer.
Acknowledgements
We thank Dr. Kathleen Green, Dr. David Garrod, and Dr. W. James Nelson for providing antibodies, Dr. James Ford for human mammary cancer cell lines, and Dr. Steven Artandi for K14-Cre transgenic mice. We also thank Katie Bell at the Center for Comparative Medicine, University of California, Davis, for assistance with the CD45 immunohistochemistry.
This work was supported by the American Cancer Society New England-SpinOdyssey Postdoctoral Fellowship (PF-08-259-01-CSM) to RLD, the Susan G. Komen Foundation (KG080306) to JLB, and by the National Institutes of Health, National Center for Research Resources (K26 RR024037) to ADB. The work from MJB's laboratory is supported by grants from the U.S. Department of Energy, Office of Biological and Environmental Research and Low Dose Radiation Program (contract no. DE-AC02-05CH1123); by the National Cancer Institute (awards R37CA064786, U54CA126552, R01CA057621, U54CA112970, U01CA143233, and U54CA143836; Bay Area Physical Sciences-Oncology Center, University of California, Berkeley, California); and by the U.S. Department of Defense (W81XWH0810736). The National Cancer Institute also provided funding (F32CA130365) to RLD and to LDA (5 R01 CA093665-10). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Cancer Institute or the National Institutes of Health.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
RLD conceived and designed all of the experiments, performed the majority of the experiments, data analysis, and statistical analysis, and drafted the manuscript. JLB contributed substantially to the experimental design, assisted with experiments and provided technical expertise, performed data analysis, and helped to draft the manuscript. HV and ADB contributed to the data analysis of mammary gland and mammary tumor histology and helped with the design and execution of CD45 immunohistochemistry experiments. SB performed some of the Western blots to define levels of desmosome proteins in wild-type and Perp-/- MECs and provided critical feedback on the manuscript. MJB participated in the design of the study and its coordination. LDA conceived of the study and designed experiments, analyzed data, and contributed to writing the manuscript. All authors read and approved the final manuscript.