Archives
We generated an interactive fractal model
We generated an interactive fractal model of mammary lobules based on parameters measured in histological sections. The fractal tree can be virtually sectioned at arbitrary depth and angle and then compared to breast tissue sections. This model can quantitate the number of branching generations, the total number of Etonogestrel in a lobule, and identify different generations of ductules present in a section. We utilized this model to identify the location of stem cell markers that were present in small clusters of cells scattered throughout the mammary epithelium, in apparently random fashion. Several of the stem cell markers used were present in such small clusters, representing a minority of the total cells. An important finding of this study is that the majority of stem cell markers investigated, namely ALDH1A1+, SSEA4+, CK14+CK19+, and CD44high, co-localize. This provides consensus among various published studies (Choudhury et al., 2013; Ginestier et al., 2007; Shipitsin et al., 2007; Villadsen et al., 2007). Importantly, the cells in which these markers co-localize have both luminal and basal characteristics and are situated at very distinct positions in the 3D structure of the lobule at the growing ends of ductules, which represent future points of branching. This is particularly clear when comparing the virtual and histological sections. Our study underscores the necessity
of moving beyond the limited definitions of basal and luminal localization. As Gusterson and colleagues convincingly discuss, it is important to understand the distinction between basal and luminal phenotypes versus basal and luminal localization of mammary epithelial cells, and the distinction between basal cells and myoepithelial cells (Gusterson and Stein, 2012).
Interestingly, ALDH1A1+ cells were predominantly quiescent, the majority of them expressing the cell-cycle arrest marker p27 and only very rarely expressing the proliferation marker MCM2. Consistent with these findings, enhanced expression of p27 recently has been demonstrated in CD44+ progenitor population in the normal breast, and the p27+ cell population was found to be expanded in nulliparous breast and correlated with breast cancer risk (Choudhury et al., 2013).
Another finding was that the candidate stem cell phenotypes that never co-localized with the set enumerated above, ALDH1A3+ and CD49fhighEpCAMlow, are present exclusively in the extralobular ducts. It is possible that the human breast epithelium, unlike the mouse mammary epithelium, is sustained by more than one distinct population of stem or progenitor cells. Ductal progenitors may be active primarily before and around puberty while lobular progenitors may sustain the growth of lobules in adult life, including pregnancy. The stem cell population described by Villadsen and colleagues (Villadsen et al., 2007), present at the emergence of secondary ducts and lobules from big ducts, may be the cell pool that generates the lobular stem cells, which will subsequently move away from the parental duct. Our findings indicate that these lobular stem cells are ALDH1A1+. The change in expression of cytokeratins and ALDH isoforms between ducts and lobules is consistent with that seen during developmental patterning in other tissues. We focused on stem cells present in the lobules, given the implications for cancer origin. Association of ALDH activity or CD44 with functional or molecular cancer stem cell properties is also one of the most consistently reported in the literature. Several studies established a link between expansion of the ALDH1A1 cell population and cancer risk or cancer initiation (Ali et al., 2011; Ginestier et al., 2007; Khoury et al., 2012).
Numerous studies demonstrate that pregnancy before 35 years of age has a protective effect against breast cancer development (Clavel-Chapelon and Gerber, 2002). Russo and colleagues showed that the nulliparous breast contains predominantly L1 and L2, whereas the parous breast contains mostly L3. One of the explanations proposed for this correlation between parity and breast cancer risk was the persistence and/or expansion of the stem cell population in nulliparous women, as opposed to a decrease of this cell population with each pregnancy in parous women (Russo et al., 1999). Our findings that ALDH1A1+ cells are considerably enriched in immature L1 compared to L2 and L3 support such a correlation between long persistence of undifferentiated cells and risk of malignant transformation. Moreover, a significant difference was seen in the cellularity of immature lobules in nulliparous versus parous women and in patients that carry BRCA1 or BRCA2 mutations compared to non-carriers. These findings are consistent with previous reports from others and us, pointing to abnormalities in cellular differentiation associated with BRCA1/2 mutations, in addition to defects in DNA repair (Lim et al., 2009; Liu et al., 2008). Similar conclusions were reached by Kuperwasser and colleagues in a study using methods of analysis different from the ones we utilized (Arendt et al., 2014).