Stem cells found within the breast tissue have the ability to self-renew and generate daughter cells, including all of the cell types found in mature breast tissues [4
]. Multi-potent mammary stem cells (MaSCs) produce committed progenitors, which subsequently differentiate into myoepithelial and luminal epithelial cells [4
]. During the differentiation process, the self-renewal capacity of the cells is gradually lost. Of the MaSCs and the committed progenitors, it is believed that at least one is a bi-potent progenitor. These progenitors, which include bi-potent progenitors, all express basal markers, such as CK5, CK6, and CK14 [5
]. In addition to those progenitors previously mentioned, the luminal progenitors also express such markers as CK8 and CK18. Consequently, the existence of a luminal progenitor population has been identified [6
], but the myoepithelial progenitor could not be identified because the progenitors also express basal markers, similar to MaSCs [4
]. Myoepithelial progenitors differentiate into the myoepithelial cells, and luminal progenitors differentiate into cells that are restricted to either ductal or alveolar lineages. The terminal luminal epithelial cells or alveolar cells lose the basal markers, expressing only CK8 and CK18, whereas the terminal myoepithelial cells maintain the expression of basal markers CK5, CK6, CK14 and the new myoepithelial marker P63, SMA (smooth muscle actin), and calponin.
These findings form the basis of the cancer stem cell (CSC) hypothesis, which posits that certain tumors are initiated by one or more self-renewing CSCs that differentiate into large populations of non-self-renewing but rapidly dividing, cells which are responsible for generating the tumor mass [7
]. The cancer stem cells could potentially be derived from bi-potent stem cells or from more differentiated cells that have acquired self-renewal capabilities. In contrast to the normal stem cells, the cancer stem cells lose their capacity for multi-directional differentiation and only produce tumors with features of a particular lineage, including luminal or basal like lineages. Regardings to tumor heterogeneity, there still remains controversy. Current explanations comprise two predominant models: the cancer stem cell hypothesis and the clonal evolution and selection hypothesis. According to the clonal evolution hypothetical model, tumor cell phenotypes are determined by the phenotype of the original cell type of the tumor-initiating cell. In this study, papillary carcinoma cells expressing CK 8/18 should be differentiated from one cancer stem cell/progenitor of luminal cell lineage, and accordingly, the luminal progenitor or terminal cells should express the same markers. This phenomenon would explain the clonal evolution of papillary carcinomas. In our study, the papilloma masses consisted of cells that were positive for CK 5/6 and CK 8/18, clearly indicating that these cells originated from different tumor stem/progenitor cells. Taken together, the two models indicate that the cellular phenotypes are unstable and can change as the tumor evolves. The heterogeneity of the papilloma cells can be considered to characterize a stage of the tumor progression, specifically, a competition among tumor cells with different phenotypes. It is possible that certain papillary carcinomas derived from the papillomas might result from the CK 8/18-positive tumor cells that replace the CK 5/6 positive cells. The two models are complementary in explaining tumor heterogeneity and progression. However, the hypothesis that the papillary carcinoma cells could be derived from the papilloma cells which have the same progenitor of the cancer cells, requires further research. However, the validity of the explanation above regarding tumor progression remains to be confirmed.
In our results, Cyclin D1 staining was predominantly detected in the cells that were also immunoreactive for CK 8/18 in either the papillary carcinomas or papillomas. Previous studies have demonstrated a correlation between Cyclin D1 over-expression and breast cancers. The hypothesis that papillary carcinoma cells could be derived from papilloma cells which have the same progenitors as the cancer cells might be explained by the expression pattern of Cyclin D1 in cells co-expressing CK 8/18.
Cyclin D1, one of the protein mediators of the G1/ S cell-cycle transition, is commonly altered in breast cancer and contributes to tumorigenesis, presumably by increasing proliferation [8
]. It is generally accepted that the initiation of most tumors is triggered by chromosomal instability (CIN), which is characterized by chromosomal abnormalities and an altered gene expression profile. Data from Mathew et al. [9
] suggest that Cyclin D1 contributes to CIN and tumorigenesis by directly regulating a transcriptional program that governs chromosomal stability. During maturation of the breast cell, the differentiation and proliferative behaviors are regulated by a series of signaling pathways, such as the Notch [10
], Hedgehog [11
], and Wnt [12
] pathways. The CCND1 gene encodes a subunit of the Cyclin D1 holoenzyme, which can phosphorylate and inactivate pRB and NRF1 to regulate nuclear synthesis and mitochondrial biogenesis [13
]. The biological effects of Cyclin D1 overexpression in the process of tumorigenesis are exhibited through the pathways mentioned above. Several studies have demonstrated that the disrupted regulation of these pathways can lead to the development of breast cancer in mice [18
] and in humans [22
]. One report indicated that deletion of the CCND1 gene leads to failed mammary gland development in mice [25
]. Another study demonstrated that overexpression of the Cyclin D1 oncogene can induce mammary gland tumors in mice [26
]. These findings further suggest that Cyclin D1 might directly or indirectly trigger the differentiation of mammary glandular cells. Other studies have suggested that Cyclin D1 might inhibit the differentiation of adipocytes [8
]. In addition, many studies have indicated that high Cyclin D1 expression levels correlate with CIN, specifically in the luminal B subtype tumors [9
]. This phenomenon provides a possible explanation for why Cyclin D1 was mainly expressed in the cells that expressed CK 8/18 but not in those that expressed CK 5/6. Cyclin D1 over-expression leads to the sustained proliferation of mammary epithelial cells, which is associated with a delay in acinar development in vitro models [27
] and a failure to undergo terminal differention in mouse models [28
]. The cells in either the papillomas or the papillary carcinomas that expressed CK 8/18 could be derived from the same cancer stem/progenitor cells, which exhibit the capacity of self-renewal and strict luminal differentiation, and which over-express Cyclin D1 proteins because of various genetic or epigenetic events. The distinct expression patterns of Cyclin D1 between the papillomas and papillary carcinomas might be explained by their occurrence during different stages of tumorigenesis.