Early searches for mammary-specific stem cell markers were based on the observations of Smith and Medina that mammary epithelial explants contained pale or light-staining cells, and that only these cells entered mitosis when placed into culture (33). Subsequent and painstaking electron microscopic studies confirmed that these pale or undifferentiated cells undergo mitosis occasionally, and that they exist in both small and large forms (34). Other cells, which are darker in appearance because they contain more organelles, are never seen to mitose, and they are assumed to be terminally differentiated.
The small light cells (SLCs) fulfill the presumed criteria for stem cells in that they are division competent and are in the luminal population, but do not contact the lumen. In addition, SLCs can be found either side by side or one above the other (in relation to the basement membrane) in heteroge neous pairs or clustered with large light cells, implying that they have divided asymmetrically. Cell counting at all stages of rodent mammary gland development indicates that the proportion of SLCs remains relatively constant and accounts for approx 3% of the total epithelial population (34). This is higher than the proportion of stem cells calculated from the mammary fat pad repopulation studies (approx 1 in 2500 cells) (27) and has been explained by suggesting that SLCs comprise both multipotent and more lineage-restricted stem cells (35). SLCs have been identified in the mammary epithelium of all mammals examined so far, including humans (35). It is assumed that they are stem cells, although this has yet to be shown definitively.
Studies of other stem cell markers have built on the observations made on the position of SLCs in the rodent epithelium. Specifically, it has been postulated that, because SLCs do not contact the lumen but are situated in a suprabasal position between luminal and myoepithelial cells, they should express a general epithelial marker, but not a luminal-specific, apical membrane marker such as MUC1. Cells isolated using this strategy appear to be multipotent because both luminal and myoepithelial cell types are produced when they are placed into low-density culture. Further analysis of these cells showed that they also express a6 integrin and CK19 (32).
Further characterization of mammary epithelial stem cells has been aided greatly by cross fertilization from studies of other adult stem cell systems. For example, examination of the patterns of deoxyribonucleic acid (DNA) label retention in the study of stem cells in the skin and small intestine has now been applied to the mammary epithelium (36,37). This technique involves administration of a DNA label, usually tritiated thymidine (3H-dT) or a halogenated thymidine analogue such as bromodeoxyuridine (BudR), to the animal and then determination of which cells retain the label at subsequent time points. The label is taken up only by cells actively synthesizing DNA at the time of treatment. In cells that continue to divide, the DNA label will be progressively diluted such that, after a few divisions, levels are reduced to below the level of detection. However, in cells that do not continue to divide, the label will be retained. As quiescence and longevity are considered stem cell properties, these cells would be expected to retain label over long periods of time.
Accordingly, two groups have used label retention as a means of defining the stem cell population in the mouse mammary epithelium (38,39). The first used a bolus of 3H-dT to label the mammary epithelial cells of fully adult mice (10-12 wk old); the mice were then followed for 3 wk and showed that label-retaining cells (LRCs) comprised 0.1 to 1% of the total epithelial population (38). Immunohistochemistry combined with histoautoradio-graphy used to detect the tritium label indicated that a high proportion of the LRCs (approx 95%) also expressed the ERa.
In the second study, BudR was administered continuously for 2 wk to pubertal mice (3-5 wk old), and tissue was sampled until the mice reached 13 wk of age (39). In this study, a greater proportion of LRCs was detected, probably because labeling was carried out over the period of pubertal mammary gland development, when stem cells might be expected to be more active. However, steroid receptor expression was transiently associated with retention of the DNA label because the LRCs detected at the last time point did not contain steroid receptors, although they did at earlier time points. The very long lived LRCs also appeared to be undifferentiated in that they expressed neither CK18 nor CK14, luminal and myoepithelial markers, respectively.
The data from these two studies can be interpreted as evidence for the existence of two stem cell populations in the mammary epithelium: The first is a long-lived, primary, steroid receptor-negative stem cell, and the second is a steroid receptor-positive stem cell that might be more short lived and be the more active during the estrous cycle.
Again, and for obvious reasons, in vivo DNA label retention studies cannot be carried out in women, but a method of implanting small pieces of intact normal breast tissue into immunodeficient mice to track mammary epithelial cells after administration of the label has been used (40). In this study, tissue was labeled intensively with 3H-dT for a period covering two S-phase durations and sampled at various time points afterward. The tissue was taken from adult premenopausal, but nonpregnant and nonlactating, women, so the study was analogous to that of Zeps et al., who used adult mice (38). Accordingly, 2 wk after 3H-dT injection, a population of LRCs was detectable that comprised less than 1% of the total population and expressed steroid receptors in addition to the p27KIP1 cyclin-dependent kinase inhibitor (CDKI), which is consistent with the conception of LRCs as a quiescent population.
The second technique that has crossed over from use in the hematopoietic system to use in the mammary gland is flow cytometric cell sorting based on exclusion of the fluorescent DNA dye Hoechst 33342 (41). Hematopoietic cells that efflux the dye are called the side population (SP) and are able to reconstitute the bone marrow of lethally irradiated mice, suggesting that they are enriched in stem cells. Combining SP sorting with analysis of label retention indicated that, 9 wk after labeling, mammary epithelial LRCs formed 8% of the SP (39).
The mouse mammary SP population is also enriched for cells expressing stem cell antigen (Sca-1, a hematopoietic stem cell marker), a6 integrin (a skin stem cell marker), and telomerase. The technique has now been applied to human breast epithelial cells isolated from reduction mammoplasty specimens, and further characterization of these human SP cells is awaited (42).
It is unclear, at present, how useful SP sorting will be to the study of mammary stem cells as the ability of mouse cells isolated using this method to reconstitute mammary glands in cleared mammary fat pads is no greater than that of non-SP cells (39,42). In this respect, Sca-1 appears to be a better correlate of the ability to reconstitute mammary glands, although immuno-histochemical studies showed that approx 20% of mouse mammary epithelial cells express Sca-1 (39). As it is unlikely that they are all stem cells, more work is needed to define exactly which cell types form the population positive for Sca-1.
Unfortunately, Sca-1 is not detectable in human tissues, which means that other mammary stem cell markers need to be found. In this respect, specific cytokeratins may be useful. Gudjonsson and colleagues (30) reported that the cells that give rise to mixed luminal and myoepithelial cells in culture expressed CK19, whereas Bocker et al. (43) suggested that CK5 positivity defines a stem cell population. Equally, the absence of lineage-specific markers could be used when defining stem cell populations. The CK5-positive stem cell population proposed by Bocker et al. cannot be stained by an anti-CK8/CK18/CK19 antibody or by one that recognizes smooth muscle actin (43). The CK19-positive population identified by Gudjonsson et al. does not express MUC1 (30). These findings illustrate another difficulty in defining human mammary stem cell populations: inconsistency between different research groups. This is probably because of the fact that the study of mammary stem cells in human tissue is in its infancy, and it is hoped that a consensus will be reached as investigations progress.
Was this article helpful?
Are Menopause Symptoms Playing Havoc With Your Health and Relationships? Are you tired of the mood swings, dryness, hair loss and wrinkles that come with the change of life? Do you want to do something about it but are wary of taking the estrogen or antidepressants usually prescribed for menopause symptoms?