Understanding the mechanisms regulating human mammary epithelium requires knowledge of the cellular constituents of this tissue. Different and partially contradictory definitions and concepts describing the cellular hierarchy of mammary epithelium have been proposed, including our studies of keratins K5 and/or K14 as markers of progenitor cells. Furthermore, we and others have suggested that the p53 homolog p63 is a marker of human breast epithelial stem cells. In this investigation, we expand our previous studies by testing whether immunohistochemical staining with monospecific anti-keratin antibodies in combination with an antibody against the stem cell marker p63 might help refine the different morphologic phenotypes in normal breast epithelium. We used in situ multilabel staining for p63, different keratins, the myoepithelial marker smooth muscle actin (SMA), the estrogen receptor (ER), and Ki67 to dissect and quantify the cellular components of 16 normal pre- and postmenopausal human breast epithelial tissue samples at the single-cell level. Importantly, we confirm the existence of K5+ only cells and suggest that they, in contrast to the current view, are key luminal precursor cells from which K8/18+ progeny cells evolve. These cells are further modified by the expression of ER and Ki67. We have also identified a population of p63+K5+ cells that are only found in nipple ducts. Based on our findings, we propose a new concept of the cellular hierarchy of human breast epithelium, including K5 luminal lineage progenitors throughout the ductal-lobular axis and p63+K5+ progenitors confined to the nipple ducts.

Ovarian cancer disseminates primarily intraperitoneally. Detached tumor cell aggregates (spheroids) from the primary tumor are regarded as ‘metastatic units’ that exhibit a low sensitivity to classical chemotherapy, probably due to their unique molecular characteristics. We have analyzed the cellular composition of ascites from OvCa patients, using flow cytometry, and studied their behavior in vitro and in vivo . We conclude that ascites‐derived cultured cells from OvCa patients give rise to two subpopulations: adherent cells and non‐adherent cells. Here, we found that the AD population includes mainly CD90 + cells with highly proliferative rates in vitro but no tumorigenic potential in vivo , whereas the NAD population contains principally tumor cell spheroids (EpCAM + /CD24 + ) with low proliferative potential in vitro . Enriched tumor cell spheroids from the ascites of high‐grade serous OvCA patients, obtained using cell strainers, were highly tumorigenic in vivo and their metastatic spread pattern precisely resembled the tumor dissemination pattern found in the corresponding patients. Comparative transcriptome analyses from ascites‐derived tumor cell spheroids ( n = 10) versus tumor samples from different metastatic sites ( n = 30) revealed upregulation of genes involved in chemoresistance ( TGM1 , HSPAs, MT1s), cell adhesion and cell‐barrier integrity ( PKP3 , CLDNs, PPL ), and the oxidative phosphorylation process. Mitochondrial markers (mass and membrane potential) showed a reduced mitochondrial function in tumoroids from tumor tissue compared with ascites‐derived tumor spheroids in flow cytometry analysis. Interestingly, response to OXPHOS inhibition by metformin and IACS010759 in tumor spheroids correlated with the extent of mitochondrial membrane potential measured by fluorescence‐activated cell sorting. Our data contribute to a better understanding of the biology of ovarian cancer spheroids and identify the OXPHOS pathway as new potential treatment option in advanced ovarian cancer.