@prefix vivo: . @prefix edm: . @prefix ns0: . @prefix dcterms: . @prefix skos: . vivo:departmentOrSchool "Medicine, Faculty of"@en, "Medical Genetics, Department of"@en ; edm:dataProvider "DSpace"@en ; ns0:degreeCampus "UBCV"@en ; dcterms:creator "Makarem, Maisam"@en ; dcterms:issued "2013-06-25T09:15:42Z"@en, "2013"@en ; vivo:relatedDegree "Doctor of Philosophy - PhD"@en ; ns0:degreeGrantor "University of British Columbia"@en ; dcterms:description "Recent findings have suggested that the normal breast is a hierarchically organized tissue in which differentiated cells incapable of further division are continuously produced from a common, self-sustaining stem cell population. However, little is known about the origin, timing or properties of mammary stem cells during development. In some tissues, cells with tissue-specific stem cell properties are known to develop after the first recognizable tissue elements are in place, and these early stem cells also have qualitatively different properties than their adult counterparts. I hypothesized that this would also be true of the mammary gland. I therefore designed a series of experiments to identify, isolate and characterize mammary stem cells when they first appear during development. I also established robust in vitro and in vivo systems to enable their growth and regenerative potential to be compared with that of similarly assessed adult mammary stem cells. I found that a reproducibly detectable mammary stem cell population with transplantable regenerative activity is first detected late in gestation after the early mammary bud is formed. This population then expands in parallel with a more prevalent population detectable as cells with clonogenic activity in vitro. I also discovered that single EpCAM⁺ fetal mammary epithelial cells in semi-solid Matrigel cultures supplemented with added irradiated fibroblasts have a direct and inducible but highly variable growth and regenerative potential that is, nevertheless greater on average than that of their adult counterparts. Analysis of the 4-week regenerative activity of fetal as compared to adult mammary epithelial cells in vivo also indicated an increased output of total cells and clonogenic progenitors from the fetal cells, again with marked variation in this assay. Analysis of published gene expression profiles of fetal and adult mammary enriched stem cell subsets have suggested a number of changes in the external cues that govern their responses as well as the intrinsic molecular network they use to execute these responses. Further elucidation of developmentally-determined changes in the properties of mammary stem cells are likely to be important in understanding perturbations of adult mammary cells that endow them with malignant properties."@en ; edm:aggregatedCHO "https://circle.library.ubc.ca/rest/handle/2429/44592?expand=metadata"@en ; skos:note " i DEVELOPMENTAL CHANGES IN THE PROPERTIES OF MOUSE MAMMARY STEM CELLS by MAISAM MAKAREM B. Sc., University of Toronto, 2007 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Genetics) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) June, 2013 © Maisam Makarem, 2013 ii Abstract Recent findings have suggested that the normal breast is a hierarchically organized tissue in which differentiated cells incapable of further division are continuously produced from a common, self-sustaining stem cell population. However, little is known about the origin, timing or properties of mammary stem cells during development. In some tissues, cells with tissue- specific stem cell properties are known to develop after the first recognizable tissue elements are in place, and these early stem cells also have qualitatively different properties than their adult counterparts. I hypothesized that this would also be true of the mammary gland. I therefore designed a series of experiments to identify, isolate and characterize mammary stem cells when they first appear during development. I also established robust in vitro and in vivo systems to enable their growth and regenerative potential to be compared with that of similarly assessed adult mammary stem cells. I found that a reproducibly detectable mammary stem cell population with transplantable regenerative activity is first detected late in gestation after the early mammary bud is formed. This population then expands in parallel with a more prevalent population detectable as cells with clonogenic activity in vitro. I also discovered that single EpCAM+ fetal mammary epithelial cells in semi-solid Matrigel cultures supplemented with added irradiated fibroblasts have a direct and inducible but highly variable growth and regenerative potential that is, nevertheless greater on average than that of their adult counterparts. Analysis of the 4-week regenerative activity of fetal as compared to adult mammary epithelial cells in vivo also indicated an increased output of total cells and clonogenic progenitors from the fetal cells, again with marked variation in this assay. Analysis of published gene expression profiles of fetal and adult mammary enriched stem cell subsets have suggested a number of changes in the external cues that govern their responses as well as the intrinsic iii molecular network they use to execute these responses. Further elucidation of developmentally- determined changes in the properties of mammary stem cells are likely to be important in understanding perturbations of adult mammary cells that endow them with malignant properties. iv Preface Under the supervision and conceptual guidance of my supervisor, Dr. Connie Eaves, I designed and performed all experiments that contributed to the results Chapters 2 and 3. Nagarajan Kannan, Long Nguyen and David Knapp gave advice on data analysis and helped with data interpretation. Afshin Raouf and Peter Eirew provided preliminary experiments for the growth of human mammary cells in a similar 3D Matrigel culture system. John Stingl and Michael Prater identified the superior ability of basal colonies to form in vitro in the medium I adopted for this purpose. Darcy Wilkinson and Glenn Edin helped with immunohistochemistry staining and experimental set up for the work in Chapter 2. For all chapters, I analyzed and interpreted the data and wrote the associated manuscripts with the assistance of Dr. Eaves. Sections of Chapter 1 have been accepted for publication in the Journal of Mammary Gland Biology and Neoplasia as a review entitled “Stem cells and the developing mammary gland”1. Chapter 2 has been submitted as a manuscript entitled, “Developmental changes in the in vitro activated regenerative activity of primitive mammary cells” for publication in a peer- reviewed journal. Chapter 3 is being prepared for submission as a manuscript entitled, “Comparative in vivo analysis of fetal and adult primitive mammary cell potentials” for publication in a peer- reviewed journal. All animal experiments were carried out in accordance with the policies and guidelines presented by the University of British Columbia Animal Care Committee. Canadian Council on Animal Care Approval was granted under the certificate number: #A11-0037. v Table of Contents Abstract ......................................................................................................................................... ii Preface ........................................................................................................................................... iv Table of Contents ........................................................................................................................... v List of Tables ............................................................................................................................... vii List of Figures ............................................................................................................................ viii List of Abbreviations ..................................................................................................................... x Acknowledgements ..................................................................................................................... xii Dedication .................................................................................................................................... xiv Chapter 1: Introduction ................................................................................................................ 1 1.1 Key stages of mammary gland development ........................................................................ 1 1.2 Structure and cellular composition of the mouse mammary gland during development. ............................................................................................................................................ 4 1.3 Historical evidence for cells with regenerative potential in the mammary gland ............. 7 1.4 Detection and quantification of MRU, assay requirements, and modifications ............... 8 1.5 Detection and quantification of in vitro mammary colony-forming cells ........................ 11 1.6 Post-natal changes in MRU numbers ................................................................................. 14 1.7 Origin and appearance of MRU during embryogenesis ................................................... 14 1.8 Phenotypic properties of fetal mammary stem cells .......................................................... 15 1.9 Lineage fate of embryonic mammary cells as demonstrated by lineage-tracing ............ 16 1.10 Molecular regulators of fetal and adult stem cell properties ............................................ 16 1.11 Self-renewal pathways involved in cancer .......................................................................... 19 1.12 Thesis objectives ................................................................................................................... 20 Chapter 2: Developmental changes in the in vitro activated regenerative activity of primitive mammary epithelial cells ............................................................................................ 27 2.1 Introduction .......................................................................................................................... 27 2.2 Methods ................................................................................................................................. 28 2.3 Results .................................................................................................................................... 31 2.3.1 First appearance of MRUs in the developing mouse embryo ................................................. 31 vi 2.3.2 Changes in the frequency, content and phenotype of total cells, CFCs and MRUs in the mammary epithelium between E18.5 and adulthood .............................................................. 31 2.3.3 Fibroblast-containing Matrigel cultures support the production of large numbers of CFCs and MRUs ............................................................................................................................... 33 2.3.4 3T3 cells enhance CFC and MRU production in Matrigel cultures by novel factors ............. 34 2.3.5 Immunophenotypic characterization of the cells present in mammary structures produced from single purified fetal, adult basal and luminal cells ......................................................... 35 2.3.6 Functional characterization of the cells present in mammary structures produced from single purified fetal, adult basal and luminal cells ............................................................................ 35 2.4 Discussion .............................................................................................................................. 37 Chapter 3: Comparative in vivo analysis of fetal and adult primitive mammary cell potentials ....................................................................................................................................... 60 3.1 Introduction .......................................................................................................................... 60 3.2 Methods ................................................................................................................................. 62 3.3 Results .................................................................................................................................... 63 3.3.1 Establishment of a quantitative measure of MRU activity reveals an increased growth potential of fetal versus adult mammary cells ........................................................................ 63 3.3.2 Production of CFCs in 8-week regenerated glands correlates with the output of total EpCAM+ cells ......................................................................................................................... 64 3.3.3 MRU outputs at 8 weeks post-transplantation are highly heterogeneous ............................... 66 3.3.4 CFC and EpCAM+ cell outputs at 4 weeks correlate with input cell numbers and show fetal cells have regenerated more CFC and EpCAM+ cells at that time ......................................... 67 3.4 Discussion .............................................................................................................................. 68 Chapter 4: Discussion and future directions ............................................................................. 83 4.1 Late detection of fetal MRUs ............................................................................................... 84 4.2 Clonal analysis of fetal and adult mammary cell outputs ................................................. 85 4.3 In vitro activation of stem cell properties ........................................................................... 86 4.4 Increased potency of fetal mammary cells ......................................................................... 88 4.5 Heterogeneity in cell outputs from fetal and adult mammary cells ................................. 91 4.6 Implications for breast cancer ............................................................................................. 93 4.7 Concluding comments .......................................................................................................... 95 References ..................................................................................................................................... 97 vii List of Tables Table 1.1 Reported MRU frequencies in suspensions of unfractionated adult breast tissue from different strains of mice and the effect of including Matrigel in the innoculum ......................... 25 Table 2.1 Number of positive fat pads from transplantations of mammary fragments ............... 42 Table 2.2 Number of positive fat pads from dissociated mammary buds at various stages of embryonic development............................................................................................................... 42 Table 2.3 Frequency of total cells, CFCs and MRUs in unseparated suspensions and different subsets of fetal and adult mammary cells. ................................................................................... 45 Table 2.4 LDA of the enhanced MRU frequency in unseparated adult mammary cells tested in recipients given an E/P pellet ...................................................................................................... 47 Table 2.5 LDA of the MRU frequency in different adult mammary basal and CD61+ luminal subsets .......................................................................................................................................... 47 Table 2.6 LDA of the MRU frequency in unseparated E18.5 fetal mammary cells ................... 48 Table 2.7 LDA of the MRU frequency in purified fetal mammary subsets ................................ 48 Table 2.8 LDA of the MRU frequency in 7-day Matrigel cultures initiated with fetal mammary cells .............................................................................................................................................. 53 Table 2.9 LDA of the MRU frequency 7-day Matrigel cultures initiated with adult mammary cells .............................................................................................................................................. 53 Table 2.10 Frequency of single cells that form structures and generate CFCs and MRUs from different subsets of fetal and adult mammary cells. .................................................................... 56 Table 2.11 LDA of the MRU frequency in 7-day Matrigel cultures initiated with single mammary epithelial cells from different sources ........................................................................ 57 Table 3.1 Fetal and adult MRU frequency calculated based on different criteria for a positive regenerated gland. ........................................................................................................................ 74 Table 3.2 In vivo expansion of MRUs in primary recipients of unseparated fetal mammary cells ..................................................................................................................................................... 80 viii List of Figures Figure 1.1 Schematic representation of early mammary gland development. ............................ 22 Figure 1.2 Schematic representation and immunohistochemical images of mammary duct cross- sections. ....................................................................................................................................... 23 Figure 1.3 Schematic representation of the MRU assay protocol ............................................... 24 Figure 1.4 Assay protocol for detecting mammary CFCs ........................................................... 26 Figure 2.1 Increasing numbers of EpCAM+ cells, MRU and CFC in developing mammary glands ........................................................................................................................................... 43 Figure 2.2 Representative FACS profiles of fetal and adult mammary epithelial cells. ............. 44 Figure 2.3 Enhancing effects of low O2 and ROCKi on colony formation by fetal and adult cells. ..................................................................................................................................................... 46 Figure 2.4 Quantification of the CFC and MRU content of fetal compared with adult mammary cells. ............................................................................................................................................. 49 Figure 2.5 Effects of 2D and 3D culture on fetal and adult mammary CFC production ............. 49 Figure 2.6 The effect of added ROCK inhibitor and oxygen tension on fetal and adult CFC production after culture................................................................................................................ 50 Figure 2.7 Production of MRUs and CFCs in 7-day Matrigel cultures of adult and fetal mammary cells. ............................................................................................................................ 51 Figure 2.8 A transwell system reveals a role for secreted factors in fetal CFC production ........ 52 Figure 2.9 The effect of 3T3 cell CM, Wnt, FGF, CSF-1 and HGF on adult basal and luminal, and fetal mammary cell production of CFCs. .............................................................................. 54 Figure 2.10 Morphology and cellular composition of structures generated in 7-day Matrigel cultures of single fetal and adult mammary cells. ....................................................................... 55 Figure 2.11 Comparison of CFC and MRU outputs in 7-day Matrigel cultures initiated with single fetal or adult mammary cells. ............................................................................................ 58 Figure 2.12 Analysis of regenerated glands from basal and luminal cell-derived structures ...... 59 Figure 3.1 Classification of regenerated trees based on GFP/EpCAM expression ..................... 73 Figure 3.2 Schematic showing the ideal window during which CFC should be measured......... 75 Figure 3.3 CFC are produced in regenerated glands and correlate with EpCAM expression ..... 76 Figure 3.4 Low numbers of MRU saturate the fat pad at 8 weeks and reveal possible inhibition at high cell doses .......................................................................................................................... 77 ix Figure 3.5 CFC outputs display similar outcomes of saturation and inhibition .......................... 78 Figure 3.6 In vivo self-renewal activity of fetal MRUs reveals inhibition of large input MRU . 79 Figure 3.7 Fetal mammary cells produce heterogeneous but higher CFC outputs per starting EpCAM+ cell after 4 weeks in vivo ............................................................................................. 81 Figure 3.8 Differences between fetal and adult cell in vivo regeneration depends on the stringency of the criteria used to calculate MRU frequency ....................................................... 83 x List of Abbreviations 3H-Tdr tritiated thymidine APC allophycocyanin AREG amphiregulin B6 C57Bl/6 bFGF basic fibroblast growth factor BrDU 5-bromo-2'-deoxyuridine CALLA common acute lymphocytic leukemia antigen CD cluster of differentiation CFC colony-forming-cell CK cytokeratin DAPI 4',6-diamidino-2-phenylindole E (number) embryonic day E estrogen ECM extracellular matrix EGF epidermal growth factor EGFR epidermal growth ER estrogen receptor FGFR fibroblast growth factor receptor FITC fluorescein isothiocyanate GFP green fluorescent protein GH growth hormone HGF hepatocyte growth factor HSC hematopoietic stem cell IGF-1 Insulin-like growth factor gene 1 K27 lysine 27 KO knock-out LDA limiting dilution analysis MRU mammary repopulating unit NSC neural stem cell P progesterone xi PE phycoerythrin PI propidium iodide PR progesterone receptor Prl prolactin PrlR prolactin receptor PTHR1 type 1 parathyroid hormone-related protein receptor PTHrP parathyroid hormone-related protein ROCK rho-associated kinase ROCKi rho-associated kinase inhibitor RT reverse transcriptase SEM standard error of the mean SF scatter factor SMA smooth muscle actin TGFβ transforming growth factor β THY-1 thymocyte differentiation antigen xii Acknowledgements First and foremost, I am deeply indebted to my supervisor, Connie Eaves, for her exceptional commitment, dedication and her support and mentorship in every aspect of my work and scientific efforts. I thank her for providing the perfect environment to discuss and engage in new scientific concepts and ideas; for instilling in me sound scientific principles, for teaching me how to ask important biological questions and how to critique my own and others’ work. I thank her for her guidance and attention to detail, for staying up late to finalize many manuscripts, and finally for her full support of my future career goals. I also want to acknowledge the support and criticisms of my supervisory committee, Dr. Aly Karsan, Dr. Sam Aparicio and Dr. Keith Humphries. I thank Dr. John Stingl and Michael Prater for sharing unpublished data to identify and detect basal colonies from the adult mammary gland. As well as, Dr. Afshin Raouf, for sharing his results from investigating the use of irradiated feeders in 3D Matrigel cultures with human mammary cells. I want to acknowledge the support of Darcy Wilkinson for her help with preparing and staining immunohistochemistry slides and Glenn Edin for help with culture preparations for experiments. I want to thank the mammary team, past and present; especially, Raj, Long, Peter and Sneha for helping to make my time here very enjoyable and productive. I thank other members of the Eaves’ Lab for their input, feedback and late night discussions throughout the years; especially; Alice, Claudia, Mel, Mike, Yun, Paul, David and Philip. I would like to acknowledge current and past members of the Terry Fox Labs for being my friends and family throughout the years. I want to thank the Aboulhosn’s and the Modad’s for being my second family in Vancouver. xiii And lastly, I want to thank my parents, my brother and my sister, who are always my continuous support in every aspect of my life. xiv Dedication I dedicate this thesis to my parents 1 1 Chapter: Introduction Chapter 1: Introduction 1.1 Key stages of mammary gland development The adult mammary gland is a continuous system of ductal structures and end lobules that are responsible for providing milk and nutrients to the newborn. Throughout development, the mammary gland undergoes many changes in its cellular make-up. This begins with the formation of a placode in the ectoderm of the mid-gestation embryo and is followed after birth by a sequence of hormonally regulated changes during puberty and adulthood2. As is the case for many other organs derived from the ectoderm, the development of the mammary gland also involves reciprocal interactions between the mammary epithelial cells and the underlying mesenchyme3. The first evidence of future mammary tissue can be visualized histologically in the mouse on embryonic day 10.5 (E10.5) as cellular thickening at 10 distinct regions along the ventral side of the embryo. In the rabbit, these appear as elevated “mammary lines” by scanning electron microscopy4. This line extends from the anterior to the posterior limbs along symmetrical sides of the embryo. Some of this epithelial thickening is hypothesized to arise from migrating cells in the epidermal layer of the skin and forms distinct lens-shaped structures known as mammary placodes4,5. These placodes consist of epithelial cells, that are thought to form by recruitment of cells from the surrounding epidermal layer rather than by in situ proliferation due to their lack of mitotic activity from E11.0 – E14.56,7. Markers used to distinguish the embryonic mouse mammary line and the early mammary placodes from surrounding epidermal tissue include Wnt10b transcripts at E10.5 and a Src homology 2 2 domain-containing inositol-5-phsophatase s-SHIP driven green fluorescent protein (GFP) reporter at E11.55,8. Subsequent mammary cell growth between E10/11.5 and E13.5/14.5 involves the generation of spherical protrusions, or “bulbs” of epithelial cells, just prior to E14.5 that later penetrate into the underlying dermis as mammary buds, as seen both histologically and by scanning electron microscopy6,7. An important role of the mesenchyme underlying the forming epidermis is evident in male embryos as early as E13. At this time, the mammary cells that have formed in male embryos respond to activated androgen receptors on the adjacent mesenchymal cells by condensing and starting to undergo necrosis resulting in later detachment of the epithelial bud from the surface9,10. In female embryos, mammary cells are thought to remain in a “resting” state until E16. Then epithelial cells begin to proliferate and form an elongated mammary sprout that invades the mammary mesenchyme and comes in close proximity to the underlying fat pad precursor tissue (Figure 1.1). Differentiating adipocytes become evident at that stage and by birth form fully mature adipose tissue11-13. Two main types of mesenchyme then surround the mammary rudiment, one is the mammary mesenchyme that is thought to be capable of inducing non-mammary epidermis to form mammary epithelium and supports complete mammary cell differentiation as exemplified by the production of α-lactalbumin14,15. The second mesenchyme is the fat pad precursor tissue which has less of a mammary-specific inductive role, based on the finding that tissue taken from earlier embryonic stages, such as E14, is unable to direct salivary epithelial cells towards a mammary cell fate16,17. The fat pad precursor mesenchyme becomes the mature adipose tissue that forms the fat pad at birth, and this tissue is important for the early formation of the rudimentary branched structure, as mice that lack white adipose tissue form less branched structures at E1817. 3 At E18.5, 10-20 branched ducts are present and lumen formation begins. This branched structure then grows isometrically until puberty18. The embryonic mammary gland grows independently of hormonal regulation as estrogen receptor (ER) knockout mice (KO) and progesterone receptor (PR) KO mice show no defects in the early development of the branched E17.5/18.5 structure19,20. However, it is interesting to note that receptors for these hormones are already being expressed at these early developmental stages. Moreover, certain endocrine- disrupting chemicals, such as diethylstilbestrol, a synthetic nonsteroidal estrogen, and a structurally related compound, bisphenol A, can cause mammary gland lesions and hyperplasias indicating that functional receptors are also present at these early times21-23. The majority of mammary gland growth takes place after puberty starting in the 3rd week after birth when the secretion of estrogen (E) and progesterone (P) by the ovaries induces mammary cell proliferation and the formation of terminal end buds, which are regions of high proliferative activity13,24-26. Mammary cells continue to invade through the fat pad and proliferate until they reach its outer limits in the 10 to 12 week-old adult mouse. Some of the hormone receptors that are important for the postnatal development of the mammary gland (branching and alveogenesis) include ERα, PR, growth hormone receptor (GHR, required in the stroma), and the prolactin receptor (PrlR)19,27-31. Notably, epithelial cells show a particular dependence on the expression of ERα, independent of the ER status of the stromal cells19. During pregnancy and lactation, PrlR, and downstream Janus-Kinase/Signal Transducer and Activator of transcription (JAK/STAT) signaling become important for the formation of alveolar structures and subsequent milk production32-34. Upon weaning of newborn litters, the gland undergoes massive remodeling through apoptosis mechanisms and regresses to a state that is similar to the virgin adult gland35. 4 1.2 Structure and cellular composition of the mouse mammary gland during development Histological studies of the bi-layered structure of the mouse mammary gland have led to the description of two main cell types (Figure 1.2). One of these makes up the inner layer of polarized cuboidal and columnar-like “luminal” cells that line the ducts and alveolar structures of the gland and are responsible for milk secretion during lactation. The other makes up the outer (basal) layer of elongated, flat myoepithelial cells which enclose the luminal cells. Myoepithelial cells possess contractile properties important for milk secretion. They are also important for maintaining tissue integrity and play important roles in controlling ductal growth and differentiation36-39. This bi-layered cellular structure is fully encased by a basement membrane, an extracellular matrix (ECM) sheath that is composed primarily of laminin and collagen, and plays an important role in regulating the overall branching and morphogenesis of the gland40,41. Cells in the luminal and basal layers of the adult mammary gland express different cytokeratins (CKs) and various other proteins. In both mouse and human mammary glands, CK8, 18 and 19, Mucin-1 (MUC1) and Prominin-1 (CD133) are distinguishing markers of luminal cells, although the expression of some of these CKs (such as CK8/18) has proven to be less prominent during lactation. Expression of CK14 may be seen in luminal cells of pre- pubertal glands from birth until puberty42-45. Markers of basal mammary cells include CK5, CK14, Common Acute Lymphoblastic Leukemia Antigen (CALLA, CD10), Thymocyte differentiation antigen 1 (THY1, CD90) and p63. Notably, a number of these markers are shared by cells in the basal layer of other epithelial tissues, including the prostate, skin and salivary gland46-51. 5 In the embryo, many of the CKs are expressed simultaneously in the same cells in the developing ducts52, suggesting that many mammary epithelial cells may be in a more “uncommitted” state. However, little is known about the lineage potential of cells that express CKs associated with adult basal or luminal cells. The only CK marker that has been used to track the lineage output of cells during embryonic development is K14, the expression of which marks precursors of both basal and luminal cells at E1753. Several integrins, such as α6, β1 and β3 have also been recently reported to be useful markers for the prospective isolation of adult basal and luminal cells, as well as for enriching for or selectively identifying primitive cells within various phenotypically defined compartments54- 56. Some integrins are specific to one of the two cellular subsets. For example, α1β1 (collagen receptor) and α5β1 (fibronectin receptor) are expressed exclusively in myoepithelial cells, whereas α2β1 (collagen receptor), α3β1 (laminin) and α6β1 (laminin) are expressed by both myoepithelial and luminal cells. Some integrins have also been found to be expressed on the lateral surface of luminal cells and thought to be more involved in cell-cell versus cell-ECM interactions57. Flow cytometric analysis has shown that myoepithelial cells express higher levels of α6 and β1 than most luminal cells54,55. The expression of these integrins is also variable during development which may have as yet unknown functional relevance. Many deletion studies of particular subunits (α6, α3, β4) have not been found to cause mammary defects58, but this may be due to their ability to form heterodimers with other integrins. However, deletion of α2 causes some reduction in branching59. β1 integrin deletion in basal cells using a K5-Cre mouse showed that mammary development is disrupted and results in abnormal and reduced branching56,57,60. Other cellular components of the fat pad in which the mammary gland is contained include fibroblasts, hematopoietic and endothelial cells as well as adipocytes and possibly other 6 stromal cells. The fat pad has been found to be an important stromal environment for regulating mammary epithelial cell branching, although other adipose tissue has been shown to induce growth and ductal branching of mammary cells, suggesting some non-specific induction by the mammary fat pad40. Some key growth factors shown to be important mediators of signaling between mammary epithelial cells and their stromal neighbors include Hepatocyte Growth Factor/Scatter Factor (HGF/SF), Transforming Growth Factor B (TGFβ), and Epidermal Growth Factor (EGF). All of these are produced by surrounding fibroblasts in mammary tissue. HGF/SF is a potent stimulator of ductal branching and proliferation of mammary cells in vitro and in vivo61,62. In contrast, stromal TGFβ is a potent inhibitor of mammary cell proliferation, and plays a role in branching patterns when this inhibitory morphogen’s concentration is at its lowest63-66. EGF receptor (EGFR) ligands have also been shown to be important. The most dramatic effects on mammary development have been observed in mice lacking Amphiregulin (AREG), one of the EGFR ligands expressed in mammary epithelium67, although other members of the ErbB signaling family affect early gland development68,69. Tissue recombination experiments with transplanted tissue fragments from either EGFR KO epithelium or stroma showed that the expression of EGFR (which is normally expressed by the stroma and epithelial cells) is most important in the stroma30,70. Similar tissue recombination experiments revealed the important role of Parathyroid Hormone-related Protein Receptor (PthR1) expression in stromal cells71. GHR expression in the stroma is also essential for mammary gland growth, and this has been demonstrated in studies where GHR KO stroma impedes ductal branching and mammary gland growth29. GH mediates some of its effects by upregulating insulin-like growth factor 1 (IGF-1), which is expressed in the stroma, (and its receptor expressed in epithelial cells)72. 7 Overexpressing IGF-1 in basal cells causes increased ductal proliferation before puberty73. Additional regulators of this GH/IGF-signaling axis which may be important in mammary gland development, include the IGF-binding proteins, some of which are expressed in the stroma. Interestingly, these genes are localized in the same chromosomal regions as the homeobox (HOX) gene clusters, which include several HOX genes also shown to be relevant in mammary gland development74-76. Overall, it is clear that complex interactions between the epithelium and stroma are necessary for the growth and development of the mammary gland. Many of these are hormonally-regulated and mediate their effects through several described growth factors and signaling pathways. 1.3 Historical evidence for cells with regenerative potential in the mammary gland There have been major advances in recent years in characterizing the cells that produce and sustain a supply of differentiated mammary cells in the adult gland. These have identified primitive mammary cells that can be detected and quantified by their ability as isolated single cells to regenerate an entire, normally structured and functional mammary gland as well as give rise to daughter cells that retain this same regenerative potential54,55,77,78 Additional experiments have also begun to investigate their first appearance during development as well as examine their in vivo lineage restriction through the use of lineage-tracing experiments53-55,79-81. In fact the existence of such cells derives from much older experiments. The earliest evidence of cells with long-term potential to repopulate a mammary gland in the mouse was documented in transplantation studies first established by Deome et al. in 1959. He introduced the use of surgery to cut out the small part of the young fourth inguinal mammary fat pad that contains all of the nascent mammary epithelial tissue, thereby creating a “cleared” fat pad in which test mammary tissue could then be placed and its subsequent growth assessed82. Early 8 work using this technique demonstrated that tissue from various developmental stages of mice from E13 to very old (2 years) can regenerate the mammary fat pad and undergo up to 9 serial transplantations83-85. In addition, it was shown that fragments of tissue from any part of the mammary tree demonstrated this potential, suggesting the presence of stem cells throughout the mammary gland83,85,86. Later experiments showed that regenerated mammary trees are composed of clonal outgrowths of mammary cells. Using mouse mammary tumor virus (MMTV) mouse models, unique insertion sites were tracked in primary and secondary transplants; and specific patterns of clones could be detected in the primary transplant, with some of those clones also present in secondary recipients. These experiments indicated that the primary and secondary regenerated glands were composed of clonal populations of repopulating cells87. In the human system, the evidence for clonal populations of cells contributing to the mammary gland was provided by studies of X-chromosome inactivation in different regions of the human gland. These showed that distinct regions containing ducts and lobules of human breast epithelial tissue contained the same inactivated X chromosome88,89. These results indicate that the human mammary gland develops from cells with extensive proliferative and multi-lineage differentiation potential, but they do not establish how long the latter property is active. 1.4 Detection and quantification of MRU, assay requirements, and modifications Quantification and characterization of primitive mammary cells has been facilitated by the development of robust in vivo and in vitro assays to measure individual cells based on the proliferative activity and regenerative activity they display under defined conditions. Assays for stem cells have made use of the cleared fat pad transplantation assay90 detect dissociated, single cells able to regenerate a complete mammary gland54,55. Advances in the development of 9 reagents and methods to obtain viable single cell suspensions of adult mouse mammary epithelial cells in high yields enabled cleared fat pad transplants to be performed with limiting numbers of mammary cells. These allowed a phenotypically distinct and rare subset with “mammary repopulating unit” (MRU) activity that can generate a complete tree-like structure within 6-7 weeks, to be quantified by employing limiting dilution assay (LDA) principles. In addition, serial transplantation experiments showed that these MRUs could self-renew, thus making MRUs meet the most stringent criteria for defining mammary stem cells54,55. The mammary transplant protocol includes inducing pregnancy in the recipients during the last 3 weeks to produce enlarged structures for easier visualization (Figure 1.3). This assay has now been used by many laboratories but has resulted in the reporting of widely different mouse MRU frequencies, particularly for dissociated, but “unseparated” suspensions of mouse mammary glands from young, virgin adult females (Table 1.1). There are a number of likely explanations for this variation. An important one is the composition of the non-MRU content of the test cells being assayed which will influence the frequency of MRU measured. Because the mouse mammary gland is embedded in the fat pad which is the entity removed, a large and variable number of stromal elements, hematopoietic cells and endothelial cells are always present in the suspension ultimately obtained. Normalizing according to the epithelial cell content of the test cell suspension (e.g., by measuring its content of EpCAM+ cells, which robustly and uniquely identifies the mammary epithelial cells91,92), or by citing absolute numbers of MRU per gland can help to minimize this extraneous source of variability. It is also possible that assessments of MRU frequencies may be affected by less readily controlled exogenous factors, exemplified by the failure of mouse MRUs to regenerate glands in the absence of colony-stimulating factor 1 (CSF-1)-producing macrophages93 or of human MRUs to grow in the fat pad of immunodeficient mice unless the local density of fibroblasts is artificially 10 increased94,95. It can thus be envisaged that the properties of co-transplanted stromal cells from different sources, including changes in those that occur during the early development of the gland, could affect the efficiency of MRU detection. Recently, it has been shown that higher MRU frequencies are also measured if the cells being injected into the fat pad are suspended in Matrigel (a tumor extract that contains a high concentration of laminin, and other extracellular matrix components96,97). However, the effects reported are again variable, ranging from 3-100- fold80,94 (Table 1.1). Interestingly, these effects of Matrigel are not restricted to the ability of mammary cells to engraft, but also extend to tumor-initiating cells98,99. Whether Matrigel activates cells to acquire a more primitive cell state, or provides better survival signals or conditions for cells that already have these properties (i.e., an increase in detection, compared to activation of latent property) remains to be investigated100,101. Defining the optimal time frame for readout and setting criteria for serial transplantation is also a challenge considering the heterogeneity now being revealed by modifications to the assay. Whether or not the host mice are made pregnant is another critical variable. Some protocols include induced pregnancy in the recipients during the last 3 weeks to produce enlarged structures for more facile visualization (Figure 1.3). Estrogen and progesterone implants may partially replace the enhancing effects of host pregnancy and also overcome the inherent stochastic failure rate associated with pregnancy induction77. However, the effect of these hormones may also be mouse strain-dependent, as some strains efficiently accept and grow transplants without hormonal supplementation80. By coupling various cell separation strategies with a subsequent MRU assay, it has been shown that adult female mouse MRUs have features of basal cells; i.e., high expression of α6 integrin (CD49f) and β1 integrin (CD29), and lack markers characteristic of luminal cells; i.e., MUC1, CD24, high expression of EpCAM, CD14 and, in certain mouse strains, CD117/c-kit. 11 MRUs also do not express CD45, the pan-hematopoietic cell markers nor CD31, a marker of endothelial cells91,102-104. The exploitation of these differentially expressed markers has allowed enrichments of adult female mouse MRUs of up to ~1% to be achieved91,101,105. A LDA approach similar to that developed for quantifying mouse MRUs has been devised to detect and quantify human “MRUs” in transplanted mice. This assay relies on the use of highly immunodeficient recipients as hosts and the suspension of the cells in fibroblast- containing collagen gels that are then placed under the kidney capsule, or the cells are co- transplanted with associated feeders into the mammary fat pad91,94,106. The human MRUs are then identified either by their ability to produce mammary CFCs 4 weeks later (kidney capsule assay), or the production of identifiable structures 6-7 weeks later (fat pad assay). Like mouse MRUs, the human MRUs thus detected, express markers characteristic of basal cells (low levels of EpCAM and high levels of CD49f) and have been obtained in such enriched populations at frequencies of ~0.01-0.1%94,95,106. 1.5 Detection and quantification of in vitro mammary colony-forming cells Some of the earliest in vitro methods to study the proliferative potential of single mammary cells came from studies in the human mammary gland. These showed that human mammary epithelial cells can be seeded at low densities (<5000 cells/cm2) under conditions where the cells grow adherent to plastic, in the presence of EGF. These studies also confirmed that the efficiency of colony formation is enhanced in the presence of irradiated fibroblast feeders107,108. Later studies confirmed these properties are also exhibited by dissociated mouse and rat mammary cells54,55,109,110. The mouse mammary colony-forming cells (CFCs) also rely on EGF and irradiated feeders for their optimal detection54,55,110 (Figure 1.4). Although human CFCs that are myoepithelial-restricted and bipotent as well as luminal-restricted have been identified 12 and well characterized for over a decade51,107,111, only luminal-restricted mouse CFCs have been detected54,55 until recently. As discussed in Chapter 2, this is now attributable to the unique dependence of basal mouse CFCs on low O2 conditions for their growth, a finding that builds on an older observation that low oxygen conditions enhance overall colony yields from unseparated cells110. In the adult mouse, mammary epithelial CFCs are at least 100-fold more numerous than cells currently detectable as MRUs and thus far, their characterization has been largely restricted to analyses of the CD24/EpCAM++CD49f+ (luminal) fraction including the recent addition of CD61, CD49b (Integrin α2), Sca-1, and c-kit to allow their further enrichment to purities of 40- 50%91,103,104. However, their progeny generated in vitro express CKs associated with both luminal and basal cells generated in vivo104,110. In contrast, the distinctions between luminal and basal CFCs in the human mammary gland are well-characterized, and their progeny generated in vitro more closely resemble those produced in vivo45,112. Additional methods to study mammary cells with proliferative potential have employed 3D Matrigel cultures54,80,94,113. Historically, such cultures have been used primarily to study lumen formation and branching mechanisms in cell lines114, but Matrigel has now been recognized as a useful matrix to support the growth of a number of epithelial tissues96,97,115,116. For example, single intestinal stem cells identified by their expression of Lgr5 can proliferate in Matrigel in the absence of any mesenchymal cells116. Although Matrigel is thought to provide many important extracellular matrix components such as laminins and collagen type IV, it is also rich in many growth factors, including TGFβ, EGF, basic FGF (bFGF), and platelet-derived growth factor96. In mammary cell cultures, both mouse and human luminal cells produce single- layered sphere-like structures, whereas mouse basal cells produce solid, sometimes branched, structures and human basal cells produce dense duct-like structures54,55,94. 13 The frequency of adult mammary cells that generate colonies in 3D assays appears to be similar to the frequency of mammary CFCs detected in 2D assays (~5-10% for purified mouse basal and luminal progenitors)94,113 and ~1-10% for purified human luminal progenitors94. Both basal and luminal mouse mammary cells produce 3D structures containing CK14+ and CK18+ cells, but basal cells produce more structures containing CK14+ cells and the luminal cells produce more structures containing CK18+ cells55. In 3D Matrigel cultures of human cells, the structures formed from basal cells contain cells expressing p63, CK5/6 and CK14, and those generated from luminal cells contain mostly cells expressing CK8/18 and CK5/6 but not CK14 or p6394,117. Additional in vitro methods to culture mammary epithelial cells have included the use of another non-adherent cell culture method, termed mammosphere culture118. This assay was developed based on the neurosphere assay established as a primary tool for the identification and propagation of mouse neural stem cells (NSCs)119. The neurosphere assay uses EGF plus Fibroblast Growth Factor (FGF)-2 to support the formation of free-floating spheres of proliferating neural cells whose progeny can be passaged to generate secondary spheres as well as differentiating neurons and glia upon plating on an adherent substrate119. In the mammary field, this assay was first developed using human mammary epithelial cells where it was shown that ~0.4% of cells cultured in this non-adherent manner can generate sphere-like structures that retain multi-lineage differentiation potential (generate luminal, bipotent and myoepithelial colonies), and can regenerate spheres in secondary and tertiary passage118,120. Although such work has demonstrated the sphere-forming potential of mammary cells, it has been shown that CFCs are depleted after the first passage even under “optimal” conditions, indicating a lack of support of stem cell populations in this system (Raouf et al. unpublished). In the mouse system, it has been shown that mammospheres can be derived from single cells, but when more than a 14 few cells are present will also represent aggregates of input cells121, thus confounding the utility of this sytem to examine self-renewal events. 1.6 Post-natal changes in MRU numbers Recently, changes in MRU numbers have been measured during the mouse estrus cycle and during pregnancy 77,78. It was demonstrated that the total number of cells at the diestrus phase of the mouse estrus cycle increases by ~2-fold, and is accompanied by a ~10-fold increase in the absolute number of MRUs78. Since MRUs do not express ER, PR or ErbB2 receptors78,122, the increase in numbers appears to be mediated by paracrine effects from luminal cells, which do express these receptors and were shown to respond to P. These paracrine mechanisms act through receptor activator of nuclear factor kappa-B ligand (RANKL) and subsequently through amphiregulin (AREG) and Wnt-4 signaling to basal cells77,78. Such effects on MRU by steroid hormones are also observed in ovariectomized mice, where a ~4-fold decrease in MRUs is detected77. An opposite effect is observed when mice are treated with a combination of both E and P. At mid-pregnancy, MRUs increase by ~10-fold but these regenerated glands have reduced potential to repopulate a fat pad in a secondary recipient, suggesting the expansion of “short-term” MRU77. 1.7 Origin and appearance of MRU during embryogenesis Changes in the properties of primitive cells during early development may reveal important clues about the mechanisms that regulate their properties. Early evidence indicated that whole rudiments from E13 mouse embryos will regenerate a mammary tree when transplanted into the cleared pubertal fat pad123-125. However, once the cells have been dissociated, this activity is not apparent until later during mid-gestation. In 2012, Spike et al.80 showed that dissociated 15 mammary cells from very early mouse embryos regenerate a mammary tree at very low frequencies and only with the addition of Matrigel to the cell innoculum. The earliest time point at which mammary stem cells could be detected reproducibly was at E17.5/18.5 and the frequency at E18.5 was ~1 in 1000 cells, but could be increased to 1 in 60 if injected with 50% Matrigel. It is not clear whether embryonic mammary stem cells lose their potential to read out in the MRU assay after dissociation from their neighboring mammary cells, or if the early cells contributing to the development of the fetal mammary gland are not bona fide stem cells which then develop later. This latter possibility is evident in the hematopoietic system where the progenitors with short-lived proliferative ability appear much earlier than definitive stem cells with durable self-renewal potential126,127. 1.8 Phenotypic properties of fetal mammary stem cells The importance of establishing a set of surface markers to prospectively isolate cells at high purities provides a powerful tool to examine and compare gene expression profiles of different mammary subpopulations and study cells at a single cell level. Fetal MRUs have been shown to be present at high purities in the CD24++CD49f++ mammary population80. Within that fraction, MRUs are found at ~10% frequency (with injections containing 50% Matrigel). Earlier during development at E14.5 and during puberty, there are cells that express high levels of CD24 and CD49f, but distinguishing these cells from nearby epidermal cells is a challenge earlier during development80. This data revealed that fetal cells share a similar phenotypic profile to that of the adult where basal cells containing MRU and CFC are found in the CD24+CD49f++ fraction. The difference appears to be the absence of a compartment that resembles the CD24++/EpCAM++CD49f+/CD29+ luminal cells present in the adult. 16 1.9 Lineage fate of embryonic mammary cells as demonstrated by lineage-tracing Growth properties of fetal mammary cells have also been recently interrogated using different lineage-tracing techniques. These allow the genetic tracing, upon activation of a Cre recombinase, of all the progeny of cells that expressed a particular gene at the time the Cre recombinase was activated by insertion of a floxed reporter gene downstream of the promoter of the gene of interest. Experiments using K14 as the gene in which a fluorescent reporter was activated at E17 showed these gave rise to both basal and luminal progeny seen in the adult mammary gland. However, when Cre was activated in K14+ cells one day after birth, or later, the majority of the marked progeny were myoepithelial53. In contrast, experiments using a different reporter gene; i.e., Axin2, a downstream target of Wnt, revealed different results81. Lineage-tracing of the progeny of Axin2-positive cells marked at either E12.5, after placode formation, or at E17.5 showed these were mostly luminal in the adult, with <1% of cells contributing to the basal layer. However, when Axin2-positive cells were marked 14 days after birth (just before the onset of puberty), their progeny were mostly basal. Additional studies will clearly be required to discern the cause of these confusing results and the order and types of fetal cells that give rise to the basal and luminal cells present in the adult. 1.10 Molecular regulators of fetal and adult stem cell properties Several molecular regulators have been proposed to be important for the pre- and post-natal development of the mammary gland. Some of these regulators have overlapping functions with those important for maintaining the self-renewal and differentiation of primitive mammary cells. One of these signaling pathways that seem to be important for the early formation mammary 17 rudiments is the canonical Wnt pathway. Its role during development has been indicated by the genetic deletion of Lef-1128,129, a nuclear target of Wnt signaling, the loss of Lrp6130, a Wnt ligand co-receptor, and the deletion of Pygopus 2 (Pygo2)131, which is part of a conserved family of plant homeodomain proteins recently shown to bind lysine 4-methyl on histone 3, and facilitate its trimethylation at Wnt target genes, thereby activating their transcription. In addition, the expression of a Wnt pathway inhibitor (Dikkopf-1, DKK1) driven by a K14 promoter disrupts mammary bud formation5,132,133. Some of the crucial roles for Lef-1 in mammary epithelial cells (between E11 – E15) are mediated by paracrine signaling through Parathyroid Hormone-related Protein (PTHrP), which is secreted by mammary epithelial cells, and binds to its receptor (PTHR1) expressed in the mesenchyme underlying the mammary placode. In mice that have both of these genes deleted, mammary buds disintegrate and fail to form a sprout due to defects in the formation of this mesenchyme. However, this effect can be rescued by addition of exogenous Bone morphogenetic protein (Bmp)-4 to cultured explants134,135. Some of these regulators play a role not only in development but also in the maintenance of the regenerative properties of mammary cells. In particular, Lrp5, another co- receptor of Wnt, is important for maintaining the basal lineage of mammary cells, since Lrp5 deletion reduces the frequency of cells detected as MRUs by ~100-fold105. In addition Pygo2- deficient animals have reduced CFC and MRU frequencies, with less extensive fat pad filling131. Wnt-1 activation using an MMTV mouse model reveals an increased number of mammary stem cells and even shows an increased repopulation potential of luminal progenitor cells136. It is still unclear whether fetal MRUs respond similarly to changes in Wnt signaling, although there is some evidence that this pathway is not over-represented in gene expression signatures of fetal MRUs80. 18 There is also evidence of an important role of FGF signaling in the developing mammary gland. Mice lacking FGF10 (secreted by the somites) or FGFR2b, which is expressed in epithelial cells at E10-11, fail to form mammary buds (with the exception of mammary bud 4)137. However, FGF on its own is not a requirement for the formation of fetal mammary cell- derived spheres in vitro80. On the other hand, there does seem to be an important role for ErbB signaling in the maintenance of fetal mammary stem cell-containing spheres in vitro as the addition of Neuregulin or EGF ligands to cultures greatly enhances sphere formation80. ErbB signaling through ErbB4 has also been shown to be important in the development of mammary placodes138,139. Additional molecular regulators of mammary cell differentiation include Notch, which plays an important role in directing luminal cell fate. Knockdown of Cbf-1, a downstream Notch effector gene, reduces luminal cells and expands MRU numbers consistent with a blockade of luminal cell differentiation and the overexpression of Notch in basal cells increases the number of cells with a luminal phenotype, but results in hyperplastic nodules140. A similar dependence of human bipotent cells on Notch3 for their ability to commit to the luminal pathway (but not to execute it) has also been demonstrated50. Gata-3 is another transcription factor that appears to regulate mammary stem cell properties. In the embryo, Gata-3 is highly expressed in the fetal MRU/CFC-enriched populations and is required for the formation of mammary structures56,80. In the adult, Gata-3, along with its upstream negative regulator FoxM1, are well-characterized for playing a role in luminal cell differentiation80,141. Loss of Gata-3 leads to an expansion of luminal progenitor cells and ectopic expression of Gata-3 causes an expansion of mature luminal cells, even inducing luminal features (expression of milk proteins) in basal cells56,142. Gata-3 is also thought to bind to the p18 promoter and negatively regulate its transcription, and mice deficient for 19 P18Ink4c on a Balb/C background develop tumors of a luminal phenotype. P18Ink4c is important for HSC self-renewal and may also have implications in understanding mammary stem cell regulation56,142-144. Members of the polycomb-group (PcG) proteins, such as Bmi-1, Ezh2 and Eed, which are involved in global epigenetic regulation of chromatin, are essential for maintaining self- renewal of adult populations enriched for mammary, prostate, and hematopoietic stem cells145- 147. Bmi-1-deficient mice have reduced MRU numbers and fail to regenerate a cleared fat pad, an effect rescued by co-deletion with the Ink4a/Arf locus147. Additional roles for Ezh2, a member of the PcG proteins, and a component of the polycomb repressive complex 2 (PRC2), which catalyzes lysine 27 trimethylation on histone 3, has been recently shown to play an important role in the proliferative and regenerative properties of mammary stem cells, with Ezh2 deletion in MMTV-positive cells affecting the colony-forming activity of basal and luminal cells, as well as reducing MRU numbers by ~14-fold148. These studies provide initial evidence for molecular pathways that are both similar and different in fetal and adult mammary stem cells. 1.11 Self-renewal pathways involved in cancer There is now increasing evidence that self-renewal and differentiation, the defining features of normal stem cells, are also relevant to malignancies that arise from mammary cells. Cells with stem cell properties in mammary tumors are known as cancer stem cells149. These cells are identified functionally by their ability to initiate and maintain tumors when transplanted into mice and in general, have been found to represent a rare subset that can be prospectively isolated and phenotypically distinguished from the bulk of the tumour cells. Interestingly, many of the molecular regulators and signaling pathways important for the self-renewal of normal 20 stem cells also appear to be important in initiating or maintaining malignancies, e.g. Bmi-1147,150- 152, Ezh2145,148,153, Wnt113,154, FGF80,155,156, Met65,157, ErbB68,80,158, and others159. The expression of many of these has also now been shown to be increased in fetal mammary cells, and a role for FGF and ErbB signaling has been demonstrated in non-adherent conditions, to be important for fetal mammary cell growth in vitro80. Although many of these self-renewal mechanisms likely play a role in tumor initiation and progression, a major recent interest has focused on determining the specific primitive cell states within the mammary epithelial hierarchy that launch a first malignant clone160-162. For example, it has been suggested that BRCA1-associated breast cancer cells, which share many features of normal mammary stem cells (ER-, PR-, and ErbB2-), actually arise from luminal progenitors. Pre-neoplastic BRCA1 mutation-carriers have expanded luminal progenitor populations, and these cells share similar expression profiles to basal-like breast cancers94,163. In addition, p53 heterozygous mice generate basal-like tumors following deletion of Brca1 in luminal cells164. Others have demonstrated a role for BRCA1 in the basal compartment165,166, and a collaborative effort with Dr. Scott Bultman suggests that the basal compartment may be the cell-of-origin for mammary tumors heterozygous for Brahma-related gene 1 Brg1, which forms a catalytic subunit of the SWI/SNF-related complex167. 1.12 Thesis objectives The development of robust in vitro and in vivo assays and the identification of key molecular regulators important for the maintenance of self-renewal properties of mammary stem cells in the adult has allowed some of the mechanisms that control their functional properties to be investigated. We hypothesized that the developing mammary gland would serve as a model to 21 understand whether and how normal mammary stem cell properties may be altered and hence be useful for elucidating how this is controlled. This hypothesis was based on evidence in the hematopoietic and neural stem cell systems, which has uncovered unique molecular mechanisms that regulate different self-renewal and lineage determination potentials in developing stem cells in early embryonic stages compared with the adult168,169. These findings have raised the question as to whether the self- renewal programs in embryonic, fetal, young and aged mice may be intrinsically controlled by changes in key developmental regulators. In order to begin to address these questions in the mammary stem cell system, it was necessary to first identify the earliest stage at which self- renewing mammary stem cells are detected, and subsequently characterize their properties. At the time this work was initiated, nothing was known about the timing of CFC or MRU appearance in the developing mammary gland. Therefore I sought first to address these questions and validate the assay conditions and endpoints developed originally for adult cells. The methods developed allowed me to compare the regenerative and growth properties of fetal and adult cells using in vitro and in vivo systems. I then sought to determine whether cells with latent MRU activity were also present. To detect such cells, I developed a sensitive in vitro system that allowed the latent MRU output of single cells to be measured. These experiments are described in Chapter 2. I also designed experiments to compare the regenerative ability of fetal and adult MRUs in vivo. These results are presented in Chapter 3. Chapter 4 summarizes the key findings and discusses their significance and implications for future studies. 22 E11.5 E14.5 E16.5 E18.5 Mammary mesenchyme Placode Epithelial bud Mammary sprout Fat pad precursor Dermis Branching mammary structure Figure 1.1 Schematic representation of early mammary gland development. The mouse mammary gland develops through reciprocal interactions between epithelial cells (blue) and mesenchymal cells (red and green). Between E11.5 and E14.5, the mammary buds invade the underlying dermis and interact with the mammary mesenchyme (red). At E16.5, mammary epithelial cells form a sprout that invades the fat pad precursor tissue (green) and starts to form a rudimentary branched structure with initiation of lumen formation late by E18.5. Histological images are of an E14.5 mammary bud (left) and E18.5 branching structure (right). Scale bars show 50µm. 23 Basement  membrane Fibroblasts Adipocytes Luminal cells K8, K18, Muc1 Basal cells K5, K14, p63,SMA K8 K5 A B Figure 1.2 Schematic representation and immunohistochemical images of mammary duct cross- sections. The mammary gland is a bi-layered system composed of basal (blue) and luminal (red) cells. (A) Basal cells express a combination of cytokeratins; K5 (B), K14 and p63 as well as smooth muscle actin (SMA). Luminal cells express K8 (B), 18, Muc1, CD133, CD14 and in some mouse strains c-kit. Scale bars show 30 µm. 24 positive negative 21-24 day-old cleared fat pad 7-10 wks Adult Fetal (E/P pellet/ pregnancy) Figure 1.3 Schematic representation of the MRU assay protocol Cells are dissociated into a single cell suspension and then transplanted into the cleared fat pad of a pubertal female mouse. 7- 10 weeks later, glands are removed and scored for the presence or absence of a large positive tree. Photomicrographs show carmine-stained examples of positive and negative glands. MRU detection can often be increased by inducing pregnancy or implanting an E/P pellet in the host 3-4 weeks after the transplant is performed. 25 Table 1.1 Reported MRU frequencies in suspensions of unfractionated adult breast tissue from different strains of mice and the effect of including Matrigel in the innoculum Mouse strain Frequency of MRU (95% CI) -Matrigel Frequency of MRU (95% CI) + Matrigel Innoculum solution References FVB 1/1,400 (1/600 – 1/3,000) HF 54 FVB 1/110 (1/72 – 1/160) PBS: Matrigel (1:1) 170 FVB 1/600 * (1/300– 1/1,100) HF: Matrigel (1:1) 78 BALB/c 1/7,800 (1/1,900 – 1/31,000) 5µg/ml Matrigel 105 BALB/c 1/1,600 ** (1/940 – 1/2,700) 5µg/ml Matrigel 171 BL/6J 1/2,000 (1/900 – 1/4,500) HF 54 BL/6J x 129Sv 1/4,900 (1/3,200 – 1/7,500) 50% FCS in PBS 55 BL/6J x B6SJLF1/J 1/700 (1/500 – 1/1,000) PBS: Matrigel (1:1) 131 BL6/J x CD-1 1/30,000 (1/11,000 – 1/80,000) 1/300 ( 1/150 – 1/500) HF or HF: Matrigel (1:1) 80 HF: 2% FBS in Hanks’ Balanced salt solution * Frequency at diestrus ** Injected with synthetic triacylglyceride (vehicle control arm frequency) 26 Adult Clonal density 5% O2 +irradiated fibroblasts 5-7 days Figure 1.4 Assay protocol for detecting mammary CFCs Cells are dissociated into a single cell suspension and then plated at a sufficiently low density for individual colonies to be scored as non-overlapping entities on irradiated fibroblasts (3T3 cells) and cultured in the presence of EGF, insulin, FBS, cholera toxin, hydrocortisone and for 5-7 days at 37oC in a humidified atmosphere of 5% O2. At the end of this time, the assay dishes were stained with Giemsa and colonies scored using an inverted microscope. Images on the right show 2 typical colonies generated from adult epithelial cells. Scale bars show 50 µm. 27 2 Chapter: Developmental changes in the in vitro activated regenerative activity of primitive mammary epithelial cells Chapter 2: Developmental changes in the in vitro activated regenerative activity of primitive mammary epithelial cells 2.1 Introduction The presence of cells in the adult mammary gland with extensive regenerative ability was suggested initially by the outgrowths obtained when mouse mammary tissue fragments were serially transplanted into cleared mammary fat pads82. Subsequent experiments exploiting endogenous retroviral insert tracking established that this regenerative activity could be attributed to single cells87. More recently, it was shown that this activity could be obtained by single cells transplanted in isolation and that these cells represent a rare subset with features of basal cells (CD24+EpCAM+CD49f++)54,55. The regenerated mammary structures contain the same spectrum of cell types that are present in the adult mammary gland, including CFCs that have a luminal (CD24++EpCAM++CD49flow/-) phenotype, as well as cells with the same in vivo regenerative activity as the original parental input cell. Cells with this mammary regenerative activity are referred to operationally as MRUs. MRUs can be quantified by LDA of their ability to regenerate large branched glandular structures when transplanted into the cleared fat pad of pre-pubertal mice54,55. The MRU assay has now been widely used to investigate changes in their numbers that occur during and after pregnancy, and some of the mechanisms that regulate their growth and differentiation172,173. Previous studies of the development of the mouse mammary gland have provided an extensive histological description of this process and some of the molecular features of the cells present at different early stages44,52. However, at the time this thesis work was initiated, there was no information as to the timing of appearance or properties of the earliest cells with MRU 28 or CFC activity. The experiments described in this Chapter were therefore initiated to address these questions. Simultaneously, Spike et al. 201280 obtained similar results showing that these cells could first be detected at reproducible frequencies at E18.5 when an advanced rudiment is already formed. These findings raised the possibility that MRUs might arise in the embryo from pre-existing mammary epithelial cells that had not yet acquired all of the properties required for their operational recognition as MRUs. I therefore next sought to devise an in vitro system to address this possibility and then compare the results with adult mammary cells cultured under the same conditions. The results provide new evidence that both fetal and adult glands contain cells that do not display MRU activity when assayed directly, but can generate such cells under certain conditions in vitro. Moreover, the frequency and potency of these fetal and adult cells appears to differ. 2.2 Methods Mice. C57Bl6/J mice were used for all experiments according to procedures that had been approved by the Animal Care Committee of the University of British Columbia. Mice were considered E0.5 on the day of observed plug. Cell isolation. Mammary glands from adult and E18.5 female C57Bl6/J mice were digested overnight (adult) or for 1.5 hours (fetal) at 37°C in DMEM/F12 medium containing 1 mg/ml collagenase A (Roche Diagnostics) and 100 U/ml hyaluronidase (Sigma) and single cell suspensions obtained as described54. Flow cytometry. Blocking of non-specific antibody binding was performed by incubating cells for 10 minutes on ice in rat serum (Sigma) and anti-mouse CD16/32 Fc-gamma III/II Receptor 29 antibody (Clone 2.4G2, STEMCELL Technologies). Mammary cells were depleted of hematopoietic, endothelial and stromal cells using biotinylated antibodies to CD45 (clone 30- F11, Biolegend), erythroid cells (clone TER-119, Biolegend), CD31 (clone MEC 13.3, BD Pharmingen), and for adult cells only, also to BP-1 (clone 6C3, eBioscience), followed by streptavidin-eFluor780 (eBioscience) or streptavidin-phycoerythrin (PE, BD Pharmingen). Anti- CD49f-fluorescein isothiocyanate (FITC, clone GoH3, BD Pharmingen) and anti-CD326 (EpCAM)-AlexaFluor 647 (clone G8.8, Biolegend), and anti-CD61-PE (integrin β3) (clone 2C9.G2, BD Pharmingen) were used to isolate the fractions described. Cells were then exposed to 4', 6-diamidino-2-phenylindole (DAPI) or propidium iodide (PI) to eliminate dead (DAPI+ or PI+) cells. The CD61+ fraction was isolated using fluorescence-minus-one controls, sorted from the adult luminal (CD45-CD31-Ter119-BP1-EpCAM++CD49f+) fraction. Cell sorting was performed using a FACSAria II or Influx II cell sorter (BD Biosciences). 2D CFC assays. Mammary cells and irradiated 3T3 fibroblasts were cultured for 6-7 days in media consisting of DMEM/F12 (3:1, STEMCELL Technologies), 10% fetal bovine serum, 10ng/ml EGF (Sigma), 1.8 x 10-4 M adenine (Sigma), 5 µg/ml insulin, 0.5 µg/ml hydrocortisone, 10-10 M cholera toxin (Sigma), and 10 µM Y-27632 (Reagents Direct). Cultures were incubated at 37oC in an atmosphere of 5% O2 in N2 unless indicated otherwise. MRU assays. These were performed as described (Stingl et al., 2006) with the following modifications. 25% Matrigel (BD Biosciences) was added to the cell innoculum and, unless indicated otherwise, a silicone elastomer E/P pellet containing 2 mg 17β-estradiol and 4 mg progesterone (both from Sigma) was implanted subcutaneously 3-4 weeks post-transplant. Another 4 weeks later, glands were fixed and stained. Outgrowths that occupied >25% of the fat 30 pad were scored as positive. All MRU frequencies were calculated using ELDA software (http://bioinf.wehi.edu.au/software/elda/)174. Matrigel cultures. Dissociated mammary cells in 200 µl of the same medium used for CFC assays were added on top of 50 µl of solidified (100%) Matrigel (Cat no. 354234, BD Biosciences) in each well of a 96-well plate and then incubated at 37oC in an atmosphere of 5% CO2 in air (20% O2) for 7 days with or without irradiated 3T3 cells or other additives as indicated. At the end of the 7 days, a cell suspension was obtained by first adding 5 mg/ml dispase for 1-1.5 hrs at 37°C and subsequently 0.25% trypsin/EDTA (both from STEMCELL Technologies) for 3-4 minutes. For histological analyses, 10% buffered formalin (Fisher) was added to the culture followed by a 70% ethanol wash. Structures were then removed and pooled for embedding in paraffin. 4 µm sections were prepared using Target Retrieval solution (DAKO), blocked using Cleanvision solution (Immunologic) and stained with an anti-CK 5 antibody (Clone AF138, Covance), anti-CK 8 antibody (ab 59400, Abcam), anti-CK 18 antibody (Clone E431-1, Millipore), anti-CK 14 antibody (LL002, Novocastra), and an anti-p63 antibody (Clone 4A4, BioCare Medical), and developed using the UltraVision ONE detection system (Fisher Scientific). Transwell experiments were performed with scaled cultures in 24- well tissue culture plates with 1.0 µm pore size inserts. Additional constituents tested (where indicated) included 200 ng/ml Wnt3a (R&D Systems), 500 ng/ml R-Spondin 1 (R&D Systems), 1 µM/5 µM XAV939 (Cellagen), 200 ng/ml mouse DKK1 (R&D Systems), 20 ng/ml bFGF (STEMCELL Technologies), 20 ng/ml mouse CSF-1 (STEMCELL Technologies), 50 ng/ml HGF (gift from Dr. Pamela Hoodless) and 3T3 cell CM (80%) obtained by incubating the cells in plain medium for 48 hrs. 31 2.3 Results 2.3.1 First appearance of MRUs in the developing mouse embryo Preliminary experiments showed that mammary fragments from E14.5 and E15.5 embryos could regenerate mammary trees, and also hair follicles after transfer to cleared pre-pubertal mammary fat pads (Table 2.1 and data not shown). However, results with similar transplants of dissociated cells from E13.5 – E16.5 mammary buds were consistently negative (Table 2.2) and the first MRU activity was not observed until E17.5. 2.3.2 Changes in the frequency, content and phenotype of total cells, CFCs and MRUs in the mammary epithelium between E18.5 and adulthood I then asked how the size of the MRU, CFC and total epithelial (EpCAM+) cell populations change from E18.5 to adulthood. The results show that all 3 parameters increase approximately in parallel (Figure 2.1). Thus the ratio of MRU to CFC in the E18.5 mammary gland does not change significantly, although the frequency of both relative to all EpCAM+ cells decreases slightly after E18.5. I then analyzed the phenotypic subsets of EpCAM+ cells present in the E18.5 fetal gland. As illustrated in Figure 2.2-A, after removal of the endothelial (CD31+) and blood (CD45+ and Ter119+) cells, 3 major subsets could be defined according to their expression of EpCAM and CD49f: EpCAM- cells (which include CD49f+ and CD49- cells), and EpCAM+ cells and EpCAM++ cells (both of which are CD49f+). The fetal gland thus appears to lack the large EpCAM++CD49f-/low population of luminal cells present in the adult gland (Figure 2.2-B). In addition, the majority of the CD49f+ cells in the fetal gland express very high levels of EpCAM, 32 whereas the majority of the cells in the adult gland express lower levels of this surface marker. The relative prevalence of EpCAM+/++ cells in fetal and adult suspensions is shown in Table 2.3. I then measured the frequency and content of CFCs and MRUs in the different subsets of EpCAM+ cells isolated E18.5 mammary tissue and compared these with the corresponding frequencies in subsets of adult mammary epithelial cells. For these comparisons, I introduced several modifications of both the originally described CFC and MRU assays (Figure 2.3-A-C). For the CFC assay, these modifications included the use of low (5%) oxygen conditions110 and the addition of a ROCKi (Y-27632) to the culture medium. Together, I found these 2 modifications enhanced the detection of fetal mammary epithelial CFCs and the CFCs in the basal fraction of the adult mammary epithelial population ~1.5 (fetal) to 5-fold (adult), but had no effect on the detection of CFCs in the adult luminal (EpCAM++CD49f+) population (Figure 2.3-A-C). For the MRU assays, I incorporated Matrigel into the solution in which the test cells were injected into the cleared fat pad101 and used a subcutaneously implanted E/P pellet to replace inducing pregnancy in the transplant host (as the latter is only variably successful). These modifications together increased the detection of MRUs from both the fetal and adult mammary gland ~15-fold, Figure 2.3-D). Application of these improved assay protocols to the different subsets of fetal and adult mammary cells showed that all fetal CFCs (Table 2.3) and MRUs (Table 2.3 and Table 2.7) are EpCAM+CD49f+. They also show that both CFCs and MRUs are largely concentrated in the EpCAM++ fraction of the fetal gland and few if any are present in the corresponding EpCAM- fraction (Figure 2.4-A and Table 2.3). Finally, it is evident that the total numbers of MRU, CFC as well as the total number of EpCAM+ cells expand ~100-fold from E18.5 to adulthood (Figure 2.1 and Table 2.3). 33 2.3.3 Fibroblast-containing Matrigel cultures support the production of large numbers of CFCs and MRUs I then undertook a preliminary survey of culture conditions that might support the production of CFCs (and MRUs) through either self-renewal and/or differentiation divisions from adult cells with “latent” MRU or CFC properties. These showed superior yields of CFCs were obtained in 3D Matrigel cultures of fetal and adult luminal cells as compared to 2D adherent culture conditions used for CFC assays, although this difference did not extend to adult basal cells (Figure 2.5). However, the use of reduced O2 conditions or added ROCKi did not affect CFC production in the Matrigel cultures (Figure 2.6). Nevertheless, ROCKi was added to Matrigel cultures in all subsequent experiments because of the positive effect these conditions have on colony formation in 2D assays of basal cells. Initial experiments with varying inputs of unseparated adult cells showed that large numbers of CFCs are produced in 7-day 3D Matrigel cultures in a cell dose-dependent fashion until the cell concentrations becomes inhibitory (above input cell numbers of ~5 x 103 EpCAM+ cells per 250 µl, Figure 2.7-A and B). Interestingly, in cultures initiated with low numbers of adult EpCAM+ cells, I found CFC outputs to be greatly enhanced by the addition of irradiated 3T3 cells. Under these conditions, ~3x104 CFCs per 100 initial adult EpCAM+ cells are produced - representing a ~1,000-fold increase over the number of CFC seeded into the culture 7 days earlier (Figure 2.7-B and Table 2.3). Fetal mammary cells showed a similar cell dose- dependent increase in CFC outputs in the same type of culture with much higher CFC outputs per initial EpCAM+ cells seeded by comparison to the results obtained with adult mammary cells (Figure 2.7-C). 34 I then assessed the ability of this culture system to support the production of MRUs (Figure 2.7-D). LDA of in vivo transplants of cells harvested from 7-day Matrigel cultures (containing 3T3 cells) showed that the number of MRUs also increases in these cultures (Figure 2.7-D and Table 2.8 and Table 2.9). Interestingly, the output of MRUs from the same initial numbers of EpCAM+ fetal and adult cells again differed, with ~4-fold higher MRU outputs coming from the fetal EpCAM+ cells. 2.3.4 3T3 cells enhance CFC and MRU production in Matrigel cultures by novel factors Experiments using a transwell system suggested that the enhancing effect of the added 3T3 cells is mediated, at least in part, by soluble factors (Figure 2.8). Subsequent experiments using 3T3- cell conditioned media (CM, 80%) confirmed that the effects of the irradiated 3T3 cells on adult basal and luminal cells, as well as fetal cells, could be at least partially replaced by factors secreted by the 3T3 cells (Figure 2.9-A). However, the activity in the CM could not be attributed to basic fibroblast growth factor (bFGF) and also appears to be different from the Wnt3a effect recently reported113, as neither Wnt3a alone nor Wnt3a plus R-spondin 1 could replace the 3T3 cells in our cultures (Figure 2.9-A). In addition, the effect of the added 3T3 cells could only be minimally inhibited, and only on basal cells (P=0.04), by the addition of one of 2 Wnt pathway inhibitors tested (XAV939 and mDKK1, Figure 2.9-B). I also observed no effect with added colony-stimulating factor 1 (CSF-1) and HGF on purified adult EpCAM+ cells (Figure 2.9-C). 35 2.3.5 Immunophenotypic characterization of the cells present in mammary structures produced from single purified fetal, adult basal and luminal cells In cultures initiated with low numbers of input cells, discrete structures were readily visualized within 7 days, and by that time, the structures derived from fetal cells were consistently larger (Figure 2.10-A). Immunostaining showed that the structures derived from single adult basal cells (EpCAM+CD49f+, Figure 2.2-B, blue gate) consisted primarily of CK5+, CK14+ cells and some p63+ cells (basal markers) with very few CK8+ or CK18+ cells (luminal markers) detectable (Figure 2.10-B, left panels). In contrast, structures derived from single CD61+ adult luminal progenitor cells (EpCAM++CD49f+CD61+, Figure 2.2-B, solid red gate) contained cells expressing markers associated with both types of progeny (Figure 2.10-B, middle panels). Flow cytometric analysis showed that structures derived from both adult cell types were composed of homogeneous populations of EpCAM+CD49f+ cells with slightly higher levels of EpCAM expression on the cells derived from the luminal cells (Figure 2.10-C). Structures produced from fetal “basal” cells (EpCAM++CD49f+, Figure 2.2-A) consisted of the same homogeneous EpCAM+CD49f+ populations (Figure 2.10-C) and most of these expressed basal markers (CK5, CK14, and p63), although a few expressing luminal markers (CK8 and CK18) were also seen (Figure 2.10-B, right panels). 2.3.6 Functional characterization of the cells present in mammary structures produced from single purified fetal, adult basal and luminal cells To characterize the cellular origin of the CFCs and MRUs produced in the 7-day Matrigel cultures, I examined these in structures generated from single input cells. In total, 163 such cultures were initiated with single fetal “basal” (EpCAM++CD49f+) cells, 124 with single adult 36 basal (EpCAM+CD49f+) cells, and 100 with single adult luminal progenitor (EpCAM++CD49f+CD61+) cells (Table 2.10). For these experiments, I focused on the CD61+ adult luminal cells because this subset had been reported to be more enriched in progenitor potential than the rest of the luminal cells56. Similarly, I focused on the EpCAM++ subset of CD49f+ fetal cells because I had found these contained >90% of the CFCs and MRUs present at the E18.5 stage of development (Table 2.3 and Figure 2.4). Visible structures were seen in 43% of cultures initiated with a single fetal cell (70/163), 30% of the cultures initiated with a single adult basal cell (37/124), and 18% of those initiated with a single adult CD61+ luminal cell (18/100). I also found that structure size (cells per culture) and CFC content were highly correlated (Figure 2.11). The proportion of cultures initiated with a single fetal or adult basal cell that contained CFCs was very similar to the proportion that contained a visible structure (Table 2.10-A). This finding is consistent with 2 outputs representing readouts of the same starting cells. However, CFCs were detected in approximately twice as many cultures initiated with a single CD61+ luminal cell as contained a visible structure (41% vs 18%), consistent with the generally smaller size of the structures produced from these input cells (Figure 2.10-A). Interestingly, the content of CFCs in individual cultures generated from all 3 sources of single input cells was highly variable, (ranging up to >104 per well). Again the highest CFC outputs were present in clones derived from fetal cells (median value ~10-fold higher than for the basal clones and almost 100-fold higher than for the luminal clones, Figure 2.11-B). In addition, these clonal CFC output measurements showed that the frequency of cells in the input populations that generated CFCs in vitro was always higher than the frequency of CFCs that were initially detectable in all 3 sources of input cells (1.5 to 3-fold, Table 2.3 and Table 2.10). 37 I also identified MRUs in the cultures initiated with single cells. The glands produced by these culture-generated MRUs were indistinguishable in morphology to those generated from freshly isolated cells (Figure 2.12). Most of the visible structures derived from the adult basal cells and some of those derived the luminal progenitors also contained MRUs (86% and 33%, respectively, Table 2.10). If it is assumed that all MRUs produced in these cultures are associated with the formation of a visible structure, the frequency of adult mammary cells that can produce MRUs is at least 50-fold higher for the basal population (26% vs 0.3) and >500- fold higher for the CD61+ luminal progenitor cells (6% vs 0.01%) than the frequency of MRUs in the respective input populations (as shown by comparing the above values multiplied by the frequency of structures present, with the corresponding input frequencies given in Table 2.3, Table 2.5 and Table 2.10). The proportion of clonal structures produced from fetal basal mammary cells that contained MRUs was even higher (95%, Table 2.10). This corresponds to a minimal frequency of fetal basal cells that can generate MRUs in vitro that is 20-fold higher than the frequency of freshly isolated fetal cells that can be directly detected as MRUs (41% vs 2%, Table 2.3 and Table 2.7). Importantly, a comparison of the Matrigel cultures initiated with single fetal and adult cells also shows that average clonal outputs of MRUs were significantly higher in the fetal cell cultures (~4-fold, P<0.01, Figure 2.11-C and Table 2.11). 2.4 Discussion There are 4 major findings emanating from these studies. First, I identified profound differences in the stage-specific regulation of the proliferation and differentiation of isolated primitive mammary cells in vitro. Second, I show that the in vitro stimulation of latent regenerative 38 activity is dependent on as yet undefined factors produced by fibroblasts that do not signal through the Wnt pathway. Third, I demonstrate a strong association of this cellular response with the formation of a visible structure in cultures initiated with single cells although the magnitude of the response exhibited by individual cells varies over several orders of magnitude consistent with the operation of a self-renewal process that includes a stochastic component. Finally, I present the first and strong evidence of a marked superiority of the regenerative activity in vitro of primitive fetal as compared to adult mammary cells. Current methods widely used to detect mammary progenitors that form colonies within 7 days in low-density 2D cultures and include irradiated 3T3 cells to optimize their detection support primarily luminal colony formation54. Here I show that colony formation by adult basal CFCs requires a low O2 atmosphere and the presence of ROCKi, whereas the clonal growth of adult luminal progenitors appears to be indifferent to these conditions. The selective sensitivity of the basal CFCs to ambient O2 explains the specificity inherent in an earlier report of this effect on adult mammary cells110. The extension of both of these requirements to EpCAM++CD49f+ fetal mammary cells that contain the stem cells (MRUs) reinforces the interpretation that these may be more “basal-like”. Thus restriction of the differentiation potential of mammary cells to the luminal lineage appears to release them from a sensitivity possessed by their upstream precursors to an ambient O2 microenvironment and pathways blocked by ROCKi. These differences add to a growing list of programmatic changes that accompany this differentiation step which is known to be governed by transcription factors like Notch350 and Gata-356, and its transcriptional repressor FoxM1141. However, the specific changes identified here may involve other transcription factors or epigenetic regulators of this step that have not yet been identified. 39 Several recent studies have indicated that the frequency of mammary cells able to regenerate a full branched tree-like structure in the cleared mammary fat pad over a 6-8-week period in vivo can be enhanced if the cells are co-injected with Matrigel80,94,101,175. I have confirmed this finding for both fetal and adult MRUs from C57Bl/6 mice. I have also confirmed that the enhancing effect of pregnancy of the host can be reliably replaced by implanting an E/P pellet in the host for the last 3-4 weeks of the assay77. More importantly, however, is the clear demonstration that, in spite of these improved direct assay conditions, both CFC- and MRU- generating activity is readily demonstrable in all subsets examined and includes cells that do not exhibit either of these capabilities when assessed for these functional properties directly. Particularly intriguing is the observation that MRUs could be derived from adult CD61+ luminal progenitors, previously thought to have irreversibly lost the bipotent, gland-generating activity of MRUs. However, it is interesting to note that evidence of persisting self-renewal of luminal cells has recently been reported53 as well as reactivation of bipotent features when “passaged” in collagen gels under the kidney capsule91. In addition, both basal and luminal cells have been shown to display increased potential to repopulate a mammary gland following constitutive Met activation65 or after transient expression of Slug and Sox9175. Thus it is tempting to speculate that the acquisition of a cell surface luminal phenotype may sometimes precede the irreversible molecular “shut-down” of bipotency and self-renewal mechanisms that operate in MRUs that have a basal phenotype, and which can then be reactivated by factors produced by fibroblasts. Indeed evidence for such a model of alternative/latent stem cell populations in the skin176 and crypt of the small intestine177,178 that can be reactivated under defined conditions has recently been reported. The finding that the addition of fibroblasts to Matrigel cultures strongly promotes the regenerative activity of cells in all subsets of adult and fetal mammary cells with MRU and CFC 40 activity raises the important question as to the molecular mediators of this response and its effectuation by the target cells. Recent reports have shown that induced Wnt signalling stimulates or is involved in qualitatively similar responses by developing81 or adult113 mammary or adult intestinal epithelial stem cells115. The CM experiments described here showed that some of the effects of the added 3T3 fibroblasts in our cultures could be elicited by soluble factors that they release, but this could not be mimicked by the addition of Wnt3a, bFGF, CSF-1 or HGF. On the other hand, a contribution of Wnt signaling cannot be entirely ruled out since a minor effect was obtained using one of 2 Wnt inhibitors, although this latter experiment could also reflect non-specific effects of the inhibitor or a role of Wnt signaling in the production or release of different factors that may mediate the effects obtained on the mammary cells. Additional experiments to elucidate all of these possibilities and define the numbers and type of factors involved as well as their mode of action will clearly be of great interest and of potential relevance to understanding mammary cell oncogenesis. The availability of a rapid and robust clonal assay to discriminate active agents as now described should greatly accelerate such future investigations. The present findings also provide the first evidence of a greater mammary stem cell- producing activity of fetal as compared to adult mammary epithelial cells demonstrable at the single-cell level. A higher self-renewal activity of fetal stem cells has been well-documented in the hematopoietic system179,180 and in the neural system181,182. Several transcription factors and chromatin regulators that have been implicated in maintaining the unique properties of these fetal stem cells and include Sox17, Ezh2, Lin28 and Hmga2145,181-184. We predict that similar intrinsic programs may be operative in fetal mammary cells, given the increased potency in the regenerative behaviour of fetal cells compared with adult mammary cells as assessed under identical conditions. Reported evidence of differences in gene expression of fetal and adult 41 mouse mammary cells that are enriched in MRU content52,80 and further analyses of these should help to elucidate the mechanistic basis of the heightened regenerative properties of fetal mammary cells. Taken together, our observations also appear relevant to growing evidence that some breast cancers may originate in luminal cells65,164, or expand from cells that have or may acquire luminal features94. The ease and rapidity with which this enormous proliferative potential of many normal mammary epithelial cells can be activated in vitro suggests that the mechanisms involved may also be targets of transforming events and act as covert contributors to the process of oncogenesis and clonal evolution in nascent breast cancers. 42 Table 2.1 Number of positive fat pads from transplantations of mammary fragments Embryonic day Fragments of mammary buds transplanted No of positive fat pads/total fat pads 14.5 2-4 2/2 15.5 2-4 2/2 Table 2.2 Number of positive fat pads from dissociated mammary buds at various stages of embryonic development Embryonic day Addition of 25% Matrigel No of dissociated buds transplanted No of positive fat pads/ total fat pads 13.5 - 5 0/2 13.5 + 4 0/4 14.5 - 9.5 0/10 14.5 + 10 0/2 15.5 - 14 0/2 15.5 + 5 0/2 16.5 - 11 0/2 17.5 - 4 2/2 Transplants performed with (+) and without (-) 25% Matrigel in the innoculum 43 Figure 2.1 Increasing numbers of EpCAM+ cells, MRU and CFC in developing mammary glands The MRU, CFC and EpCAM+ cell content of dissociated mammary glands are measured at E18.5, 3-4 weeks, and 8-12 weeks. N o pe r g la nd 44 Ep C AM CD49f C D 61 CD49f EpCAM+49f+ EpCAM++49f+ Adult Ep C AM ++ + _ CD49f Fetal SS C -A FSC-A D AP I FSC-A SS C -W SSC-A EpCAM Li n SS C -A FSC-A D AP I FSC-A SS C -W SSC-A FSC-A Li n Figure 2.2 Representative FACS profiles of fetal and adult mammary epithelial cells. Mammary cells were dissociated from E18.5 fetal mammary glands (A) or adult (B) and gated to eliminate non-viable (DAPI+) cells and doublets. Cells were depleted of hematopoietic (CD45 and Ter119), endothelial (CD31) and (in the case of adult mammary cells only) stromal (BP-1) cells (shown as Lin). Fetal EpCAM++ cells and EpCAM+ cells are shown with a green gate, and adult basal and luminal CD61+ cells are isolated using the blue and red gates. A fluorescence-minus-one control was used to set the CD61+ gate. Embedded images are negative controls. 45 Table 2.3 Frequency of total cells, CFCs and MRUs in unseparated suspensions and different subsets of fetal and adult mammary cells. Cell fractions Source Unseparated EpCAM+ 49f+ EpCAM++ 49f+ EpCAM++ 49f+61+ EpCAM- Fetal % of total cells ― 2.5 ± 0.3 10 ± 0.6 N.D. 87 ± 0.7 CFC/100cells 55 ± 6* 13 ± 2 26 ± 3 N.D. 0.09 ± 0.03 MRU/100cells N.D. 0.3 1.6 N.D. 0.001 (95%CI) (0.1 – 0.7) (0.8-3.2) (0.0002 - 0.009) Adult % of total cells ― 20 ± 1 45 ± 2 30 ± 4 30 ± 2 CFC/100cells 27 ± 2* 22 ± 2 9 ± 1 13 ± 3 N.D. MRU/100cells 0.8* 0.3 N.D. 0.01 N.D. (95%CI) (0.3 - 2) (0.1 - 1) (0.004- 0.04) * Per 100 EpCAM+ cells Total cells and CFCs are the mean ± SEM. MRU values and their 95% CI were derived by LDA (Table 2.4 -Table 2.7). 46 M RU p er 1 00 E pC AM ++ c el ls 10-2 10-1 100 101 Fetal 0 2 4 6 8 C ol on ie s pe r 1 00 ce lls Fetal O2 ROCKi 5 + 5 - 20 + 20 - C A 0 2 4 6 8 10 C ol on ie s pe r 1 00 c el ls Adult O2 ROCKi 5 + 5 - 20 + 20 - 0 10 20 30 C ol on ie s pe r 1 00 c el ls Adult Basal Adult Luminal O2 ROCKi 5 + 5 - 20 + 20 - 5 + 5 - 20 + 20 - E/P - + D B M RU p er 1 00 E pC AM + c el ls 10-2 10-1 100 101 Adult E/P - + Figure 2.3 Enhancing effects of low O2 and ROCKi on colony formation by fetal and adult cells. CFC assays of unseparated E18.5 fetal cells (A, data pooled from 2-11 experiments) and unseparated adult cells (B, data pooled from 3 experiments). (C) CFC assays of purified basal and CD61+ luminal cells (data pooled from 3-18 experiments). All values shown are the mean ± SEM. Differences between results for 5% O2 + ROCKi in all assays are significantly higher than corresponding results using 20% O2 - ROCKi (P<0.05, one-way ANOVA with Bonferroni’s multiple comparison test). (D) MRU assays of fetal and adult mammary epithelial cells in the absence and presence of an E/P pellet (results expressed per 100 EpCAM++ (fetal) or EpCAM+ (adult) cells). Unseparated cells were used to estimate MRU numbers for adult (± E/P pellet) and purified EpCAM++ cells used to estimate MRU numbers for fetal (± E/P pellet), Table S3. 47 Table 2.4 LDA of the enhanced MRU frequency in unseparated adult mammary cells tested in recipients given an E/P pellet Addition of E/P pellet Cell dose Positive fat pads/total MRU frequency (95% CI) + 1,000 7/8 1/480 (1/200 – 1/1,200) - 2,500 1/4 1/5,000 2,000 0/3 1/2,400 – 1/10,600 1,000 6/20 500 0/7 Table 2.5 LDA of the MRU frequency in different adult mammary basal and CD61+ luminal subsets Addition of E/P pellet Fraction Cell dose Positive fat pads/total MRU frequency (95% CI) + EpCAM+CD49f+ 250 3/5 1/300 (Basal) 50 1/6 1/100 – 1/780 10 0/5 + EpCAM++CD49f+CD61+ 5,000 3/7 1/8,900 (Luminal) 1/2,800 – 1/28,000 48 Table 2.6 LDA of the MRU frequency in unseparated E18.5 fetal mammary cells Addition of E/P pellet Cell dose Positive fat pads/total MRU frequency (95% CI) - 25,000 2/2 1 /1,000 5,000 2/3 (1/500 - 1/2,500) 1,000 5/7 500 3/4 100 1/2 Table 2.7 LDA of the MRU frequency in purified fetal mammary subsets Addition of E/P pellet Fraction Cell dose Positive fat pads/total MRU frequency (95% CI) + EpCAM- 10,000 0/8 <1/75,000 + EpCAM+ 500 6/8 1/360 (1/150 - 1/860) + EpCAM++ 250 4/4 1/50 50 8/15 (1/30 - 1/100) 10 3/11 - EpCAM++ 2,000 4/4 1/130 500 4/4 (1/60 - 1/280) 200 0/2 100 4/4 50 1/4 10 1/6 49 BA EpCAM levels ++ + - EpCAM levels ++ + - Figure 2.4 Quantification of the CFC and MRU content of fetal compared with adult mammary cells. (A) Distribution of fetal MRUs and CFCs according to their expression of EpCAM. CFC values are the mean ± SEM for ≥4 experiments. MRU and 95% CI values were determined by LDA. The arrow indicates that MRUs, if present in the EpCAM- fraction, were below the limit of detection (indicated by the line, for details see Table 2.7). (B) Comparison of CFC and MRU content of cells from >15 and 2 adult no. 4 glands, respectively, and >50 fetal glands. CFC values are geometric means. Error bars for MRU and CFC values show 95% CI. A B 3D 2D 3D 2D Fetal Adult Figure 2.5 Effects of 2D and 3D culture on fetal and adult mammary CFC production (A) Unseparated fetal cells (Bulk) or cells enriched for EpCAM (Ephigh), or adult basal and luminal cells (B) were plated in 2D culture conditions on Matrigel-coated (1:60) dishes (filled circles or squares) or in a 3D Matrigel matrix (100% Matrigel) (open circles or squares) and the production of CFC measured relative to the starting number of CFC before culture. 50 5 20O2 5 20 5 20 ROCKi -+ -+ -+ A B Figure 2.6 The effect of added ROCK inhibitor and oxygen tension on fetal and adult CFC production after culture Values shown are the fold-changes in CFCs detected in 7-day cultures initiated with fetal, adult basal or adult luminal cells incubated at 5% or 20% O2 (A) and assayed for CFC activity under optimal 5% O2 conditions as compared to input CFC numbers. (B) Similarly calculated changes in CFC numbers in cultures initiated with fetal, adult basal or adult luminal cells maintained at 20% O2 in the presence or absence of ROCKi. 51 B D C A Adult Fetal Input MRU 7 days Examine for structures Dissociate cells Output CFC Output MRU Input cells Input CFC Figure 2.7 Production of MRUs and CFCs in 7-day Matrigel cultures of adult and fetal mammary cells. (A) General experimental design. Mammary cells were added to 50 µl of solidified Matrigel (± 2.5x104 irradiated 3T3 fibroblasts) in 200 µl of medium and incubated for 7 days. Each well was then examined for the presence of one or more visible structures and the contents then fixed and stained or dissociated into a suspension of viable single cells to perform FACS, CFC or MRU assays for comparison with corresponding starting (input) values. (B) CFC outputs from 7-day Matrigel cultures of adult mammary cells as a function of the input cell number (expressed as the number of input EpCAM+ cells) and the addition of irradiated fibroblasts. (C) CFC outputs from 7-day cultures of fetal cells as a function of the input cell number (expressed as the number of input EpCAM+ cells). Comparison of the corresponding relationship for adult cells (solid line redrawn from (B)) shows a ~5-fold higher CFC output by the fetal cells. (D) Comparison of increased numbers of MRUs obtained from 7-day Matrigel cultures of fetal and adult mammary cells (values determined by LDA as described in Tables 2.4, 2.6, 2.8 – 2.9). 52 C FC o ut pu t +3T3 -3T3transwell Figure 2.8 A transwell system reveals a role for secreted factors in fetal CFC production CFC outputs from 500 unseparated fetal mammary cells cultured with or without irradiated 3T3 fibroblasts added directly into the same culture or in a transwell system (with the feeders in the transwell, 1.0 µm pore size). Results are pooled from 2 experiments. Differences for cultures containing no feeders and cultures containing feeders or with feeders in a transwell are significantly different (P<0.05, one-way ANOVA with Bonferroni’s multiple comparison test). 53 Table 2.8 LDA of the MRU frequency in 7-day Matrigel cultures initiated with fetal mammary cells Exp no EpCAM+ cells/well Input MRU (95% CI) Fraction of well/ fat pad Positive fat pads/total Output MRU/well (95% CI) MRU/ 100 input EpCAM+cells 1 30 0.5 (0.2 - 1) 1/6th 1/12th 4/4 2/4 16 (8 - 30) 53 (27 - 100) 2 30 1/10th 6/7 Cultures were initiated with 325 (Exp 1) or 300 (Exp 2) unseparated fetal mammary cells (containing a calculated number of EpCAM+ cells) and co-cultured with irradiated 3T3 fibroblasts for 7 days. The contents of each well were then individually dissociated and assayed as described in the Methods. The output MRU values are derived from the data pooled from both experiments. Table 2.9 LDA of the MRU frequency 7-day Matrigel cultures initiated with adult mammary cells Exp no EpCAM+ cells/well Input MRU (95% CI) Fraction of well/fat pad Positive fat pads/total Output MRU/well (95% CI) MRU/ 100 input EpCAM+ cells 1 80 0.6 (0.2 - 1.6) 1/6th 1/20th 4/4 1/4 11 (5 - 23) 14 (6 - 29) 2 80 0.6 (0.2 - 1.6) 1/20th 1/3 3 80 0.6 (0.2 -1.6) 1/20th 1/3 Cultures were initiated with 300 unseparated adult mammary cells (containing a calculated number of EpCAM+ cells) and co-cultured with irradiated 3T3 fibroblasts for 7 days. The contents of each well were then individually dissociated and assayed as described in the Methods. The output MRU values are derived from the data pooled from all 3 experiments. 54 3T3 mDKK1 XAV939 + + ++ - + -- - - + A B Basal Luminal Fetal 3T3 Wnt3a R-Spo bFGF CM - + - - - - - - - - - - - - + - - - - + + - - - - - + - C 3T3 CSF-1 HGF + - - - + - - - + - - - Figure 2.9 The effect of 3T3 cell CM, Wnt, FGF, CSF-1 and HGF on adult basal and luminal, and fetal mammary cell production of CFCs. CFC assays were performed on 7-day cultures of adult basal (60 EpCAM+CD49f+ cells), adult luminal (100 EpCAM++CD49f+CD61+ cells), and fetal (30 EpCAM++ cells) under various conditions. (A) Effect of putative stimulators compared to added 3T3 cells (set =100%): 3T3 cell CM (80%), Wnt3a (200 ng/ml) ± R-Spondin 1 (R-Spo, 500 ng/ml), or bFGF (20 ng/ml). Results pooled from 3-6 experiments. The difference in CFC output between CM and no added 3T3 cells was significant for cultures initiated with all types of cells (P<0.05, one-way ANOVA with Bonferroni’s multiple comparison test). (B) Effect of Wnt pathway inhibitors: XAV939 (1 µM for adult, 5 µM for fetal) or mDKK1 (200 ng/ml) in cultures with irradiated 3T3 cells (set =100%). Results are pooled from 3-9 experiments. Only the effect of added XAV939 was significant, and only for basal cells (P=0.04, one-way ANOVA with Bonferroni’s multiple comparison test). (C) Effect of CSF-1 (20 ng/ml) and HGF (50 ng/ml) on adult EpCAM+ cells. Data from 3 biological replicates. All cultures with added factors are significantly different from cultures with 3T3 cells (P<0.05, one-way ANOVA with Bonferroni’s multiple comparison test). 55 FetalBasal Luminal K5 K14 p63 K8 K18 A B C 4716 37 0.1 6916 15 0.1 Ep C AM CD49f CD49f CD49f 91 0.7 0.3 8 Figure 2.10 Morphology and cellular composition of structures generated in 7-day Matrigel cultures of single fetal and adult mammary cells. (A) Gross appearance of representative structures from each of the 3 types of initial cells analyzed. The scale shown indicates 30 µm. (B) Representative photomicrographs of sectioned structures stained with antibodies against the markers shown. Scale indicates 50 µm. (C) Flow cytometric analysis of dissociated PI- cells obtained from pooled harvests of cultures initiated with the same types of cells. 56 Table 2.10 Frequency of single cells that form structures and generate CFCs and MRUs from different subsets of fetal and adult mammary cells. Endpoint analyzed Fetal Basal Luminal Visible structures 43% (70/163) 30% (37/124) 18% (18/100) Wells with CFCs 44% (72/163) 35% (44/124) 41% (41/100) Structures with MRUs* 95% (20/21) 86% (12/14) 33% (2/9) Wells with MRUs† 41% (95% x 43%) 26% (86% x 30%) 6% (33% x 18%) *Based on assessment of wells containing a visible structure. †Calculated assuming MRUs would be found exclusively in wells containing visible structures. Values shown in brackets are the number of positive wells/number of wells assayed. 57 Table 2.11 LDA of the MRU frequency in 7-day Matrigel cultures initiated with single mammary epithelial cells from different sources Cell source Fraction of well/fat pad Positive fat pads/total Output MRU/well (95% CI) Fetal 1% 1/3 8 10% 15/18 (5 - 15) 60% 7/7 100% 6/7 Basal 10% 2/4 2 100% 12/14 (1- 4) Luminal 10% 0/1 0.4 100% 3/9 (0.1 – 1) Cells in cultures containing a visible structure derived from single adult basal or CD61+ luminal cells, or EpCAM++ fetal cells were dissociated and assayed for MRUs as described in the Methods. 58 B A Total CFC per clone Basal Luminal Fetal C % o f t ot al c lo ne s % o f t ot al c lo ne s % o f t ot al c lo ne s Figure 2.11 Comparison of CFC and MRU outputs in 7-day Matrigel cultures initiated with single fetal or adult mammary cells. (A) Correlation between the total number of cells in each 7-day culture with the corresponding number of CFCs present. (B) Distribution of clonal CFC outputs in 7-day cultures initiated with different types of input cells. The dotted line indicates the median values of positive clones: 1,000 for clones derived from basal adult cells (n=44), 140 for clones from CD61+ adult luminal cells (n=35) and 13,000 for clones from fetal cells (n=55). (C) Comparison of MRU outputs determined by LDA in 7-day Matrigel cultures initiated with single adult and fetal mammary cells. The value for fetal cells is significantly higher than either of the adult values (P<0.01, Chi-square P-value, pair-wise comparison of stem cell frequencies (ELDA, Table 2.11). 59 CD49f Fetal Basal Luminal E pC A M Figure 2.12 Analysis of regenerated glands from basal and luminal cell-derived structures Whole mounts (upper panels) of mammary glands produced in fat pads injected with cells from cultures initiated with single fetal cells (left panels), single basal cells (middle panels) or single or 6 CD61+ luminal cells (right panels). Scale bars show 100 µm (whole mounts). Bottom panels show representative FACS profiles of regenerated mammary glands from single fetal or basal cell-derived structures or from 6 CD61+ luminal cells after depletion of CD31, CD45, and BP-1. 60 3 Chapter: Comparative in vivo analysis of fetal and adult primitive mammary cell potentials Chapter 3: Comparative in vivo analysis of fetal and adult primitive mammary cell potentials 3.1 Introduction As described in Chapter 1, many adult tissue stem cells have been defined using retrospective functional assays that measure the potential of cells to maintain and regenerate a tissue long- term54,55,185-189. In the adult mouse mammary gland, individually isolated cells from a rare phenotypically distinct subpopulation of mammary epithelial cells with basal features can regenerate an entire functional mammary gland in the cleared fat pad of a transplanted pubertal mouse54,55. This unique property is maintained through multiple divisions in vivo as demonstrated in serial transplantation experiments. The MRU assay enables cells with this ability to be quantified by LDA of cleared fat pads that have been reconstituted or not and has thus provided a powerful method to characterize cells with this potential54,55. However, the use of the MRU assay rests on a number of assumptions. One is that the formation of a “complete mammary tree” in a fixed and stained fat pad will be a readily discernible “all-or-nothing” outcome. In addition, it assumes that there are no cell concentration limitations on the activation of MRU activity that may become relevant in various experimental situations. The studies presented in Chapter 2 showed excessive cell concentrations are strongly inhibitory to the ability of adult mammary cells to proliferate and generate structures containing CFCs (Figure 2.7). Thus it is possible that similar constraints may affect MRU growth and hence their detection in the cleared fat pad assay. Conversely, the addition of Matrigel to the innoculum in which cells are injected into the cleared fat pad has been reported to increase the 61 frequency of cells detectable as MRUs122 (Table 1.1) and I confirmed this finding. Also of note is the fact that the original endpoint of MRU activity, which relies on detecting a regenerated tree in a fixed, cleared and stained fat pad, cannot be paired with measurements of functionally defined progeny that must be viable to be detected (e.g., derived MRUs and CFCs). In Chapter 2, I showed evidence of differences in the regenerative properties of fetal and adult mammary cells stimulated to proliferate in vitro. Therefore it was of interest to investigate whether similar differences might be seen when these different sources of mammary cells are stimulated to grow in vivo. Serial transplantation experiments of clonally regenerated adult mammary glands provided an early indication that they do self-renew but that this function is also highly variable (with the number of daughter MRUs produced from a single MRUs ranging from ~10 to >100054). Similar heterogeneity in stem cell self-renewal behaviour has been documented in the hematopoietic system since analyses of their clonally regenerated progeny have been possible190,191. It has also been shown that fetal hematopoietic stem cells have the potential to produce greater numbers of daughter hematopoietic stem cells in transplanted recipients than adult hematopoietic stem cells, and that this enhanced self-renewal capacity persists until 3-4 weeks of age, after which, fetal stem cells become more adult-like in their repopulation kinetics179,180. There is also growing evidence that this developmental change in hematopoietic stem cell self-renewal potential is intrinsically regulated184,192(Copley and Eaves, unpublished data) and may also extend to neural stem cells182. All of these findings fuelled interest in examining how development might affect mechanisms that control MRU growth and self-renewal responses to signals they receive in vivo. To address this question, a comparison of the cellular outputs of fetal and adult mammary MRUs following their transplantation into cleared pubertal fat pads would be an obvious experimental strategy (Figure 3.1). However, the potential limitations of the system used to 62 detect MRUs and the endpoints used to define them were unknown. I therefore designed a first set of experiments, to examine how the assessment of MRUs (in vivo) might be made more precise and to define the limitations of their growth in time and as a function of the number of cells transplanted. The results provide evidence of a close correlation in mammary tree size, total EpCAM+ cells and total CFCs produced, enabling the latter to be developed as rapid quantitative endpoints of MRU activity. In addition, the use of these endpoints showed that fetal mammary cells have, on average a higher proliferative potential than adult mammary cells as assessed at both 4 and 8 weeks post-transplant, in spite of a wide variability in their numbers. Finally, evidence of inhibition of these 2 parameters in glands transplanted with very high numbers of cells was revealed. 3.2 Methods Mice and cell preparation. These were as described in Chapter 2 with the additional use of C56Bl6/J mice engineered to express GFP ubiquitously as donors of mammary gland tissue for some experiments as indicated193. In vitro CFC assays and in vivo MRU assays. These were performed as described in Chapter 2 with the following modifications. Regenerated mammary glands were dissociated into a single cell suspension and a fraction of cells assayed for in vitro CFC activity and EpCAM+ cell output. Flow cytometry. All flow cytometry was conducted as described in Chapter 2 with the additional use of the following antibodies. A streptavidin-perinidin-chlorophyll protein (PerCP)- 63 Cy5.5 (Biolegend) conjugate was used to deplete CD31+, CD45+, Ter-119+ and BP-1+ cells stained with the relevant biotinylated antibodies (described in methods in Chapter 2). In addition, I used a PE-cojugated, anti-CD49f antibody (clone GoH3, eBioscience) and in some cases a biotinylated EpCAM (clone G8.8, Biolegend) followed by staining with streptavidin- allophycocyanin (APC, BD Pharmingen). 3.3 Results 3.3.1 Establishment of a quantitative measure of MRU activity reveals an increased growth potential of fetal versus adult mammary cells To determine whether the total number of epithelial cells present could be used as a more objective and quantitative alternative to visualizing a complete tree, I transplanted cleared fat pads of B6 mice with mammary cells from syngeneic GFP+ mice and measured their content of GFP+ cells after 4 and 8 weeks in vivo to capture an expansion phase of growth and their final output. All regenerated glands were then visualized under a fluorescent microscope and subsequently dissociated to measure their total content of GFP+ (donor-derived) cells and the proportion of these that also expressed EpCAM to determine its utility as a candidate specific marker of all mouse mammary epithelial cells (Figure 3.1-A). Glands were then assigned different criteria based on the size of GFP+ trees observed. The size of the GFP+ mammary trees produced were scored according to the following criteria: fat pads that contained no GFP+ cells or GFP+ small sphere-like structures were scored as negative, structures that contained few ducts of GFP+ cells or a small branched GFP+ structure were scored as “+”, and fat pads that contained a large GFP+ tree that occupied the majority of the fat pad were scored as “++” (Figure 3.1-B). I then calculated the absolute number of GFP+ and EpCAM+ cells per fat pad (most GFP+ cells 64 were EpCAM+ and vice versa (Figure 3.1). All fat pads rated as “++” contained >1.3 x 104 EpCAM+ cells and those rated as “+” contained between 0.3 and 13 x 104 GFP+/EpCAM+ cells (Figure 3.1-C). I then used these 2 threshold values to measure MRU frequencies in assays of fetal and adult MRUs and to compare their EpCAM+ cell outputs when assessed at limiting dilutions. For fetal MRUs, the stringency of criteria for measuring their activity in terms of total EpCAM+ cell output at 8 weeks post-transplant had no effect on the frequency of MRUs detected, which remained at ~1 in 1,000 – 1 in 2,000 unseparated fetal cells (Table 3.1). This suggests that fetal mammary cells produce larger trees exclusively (>1.3 x 104 EpCAM+ cells). In contrast, adult mammary cells showed a dramatic change in the MRU frequency when the stringent criteria were used (increasing from 1 in 8,600 using the more stringent threshold to 1 in 180 using the reduced threshold, Table 3.1) This result indicates that the adult gland contains many cells that mostly produce smaller outgrowths in the 8 week period of the assay. 3.3.2 Production of CFCs in 8-week regenerated glands correlates with the output of total EpCAM+ cells In order to compare the growth and regenerative activity of fetal and adult mammary stem cells, I also assessed their output of a related primitive progenitor subset; i.e. the cells detectable in vitro as CFCs. The production of CFCs has been used to measure the activity of primitive human and bovine mammary stem cells proliferating and regenerating mammary tissue in collagen gels placed under the kidney capsule and by the detection of CFC106,194. This strategy was developed based on its success in detecting very primitive human precursors of hematopoietic CFCs produced after 6-8 weeks in cultures containing supportive stromal cells195,196. The assumption is that more committed progenitors/CFCs, with limited proliferative 65 activity, will be depleted from the starting cell population in the initial phase of the assay, and hence their presence at later times can be used to infer the proliferative and differentiation activity of a more primitive precursor (schematic, Figure 3.2). To test the assumption that CFCs in the starting population would be rapidly depleted, thus ensuring that the CFC output measured after 8 weeks would reflect the activity of more primitive cells, I isolated EpCAM++CD49f+ mammary cells (luminal progenitors) from an adult virgin GFP+ mouse after removing contaminating hematopoietic and endothelial cells (CD45-, Ter119-, CD31-) and transplanted cleared fat pads with ~4,000 cells (~200 CFCs) each. Luminal CFCs were chosen for this experiment because they are known to be devoid of MRUs, whereas the relationship between basal CFCs and MRUs in the mouse is not clear (basal CFCs are much more numerous than MRUs throughout development (Table 2.3), but at no stage are they phenotypically separable from MRUs). Then at weekly intervals, I used FACS to re-isolate GFP+ cells from the fat pads that had been injected with the GFP+ luminal CFCs and assayed the number of CFCs they contained. The results showed that within a week, the number of input CFCs had decreased to ~10% of the input value and remained low for another 2 weeks (Figure 3.3-A). This suggests that CFC outputs, if greater than 10% of the number transplanted at later times is indicative of regenerated CFCs. Additional transplants were then performed with both adult and fetal mammary cells and at the end of 4 and 8 weeks the number of CFCs and total EpCAM+ cells present in the same cell suspensions harvested from the fat pads were measured and compared (Figure 3.3-B). This showed that the sizes of these 2 populations produced in individual fat pads were significantly correlated over a >100-fold range. The frequencies of CFCs within the EpCAM+ fraction (of which it is a subset) was also similar at both time points (~30%) which is similar to the frequency of CFC in the total EpCAM+ population in the normal adult mammary gland (~20% - 66 30%) (Figure 3.3-C). These findings suggest a close relationship between the regulation of CFC numbers and the total size of the mammary gland. 3.3.3 MRU outputs at 8 weeks post-transplantation are highly heterogeneous I then undertook a series of experiments to examine the relationship between the 8-week output of CFCs post-transplant and the number of mammary cells transplanted, when either unseparated or purified EpCAM+ cells were used as the input population from fetal or adult mammary tissue. In order for these to be compared, the EpCAM+ (Figure 3.4) and CFC (Figure 3.5) outputs from unseparated cells were normalized according to the content of EpCAM+ cells or MRU equivalents in them based on my finding that ~10% of viable unseparated fetal cells express EpCAM and ~30% of unseparated adult mammary cells express EpCAM (Table 2.3). The number of MRUs injected was calculated based on the frequency of MRUs measured in unseparated cells using the more stringent criteria (Table 3.1). Eight weeks post-transplant, fat pads were individually dissociated and then assayed for their content of CFCs and EpCAM+ cells. The results revealed a very broad range (>100-fold) of both cell types had been produced in individual fat pads from both fetal and adult cells over a very wide range of cells transplanted, even within a single experiment (Figure 3.4-A and Figure 3.5-A). This was expected when limiting numbers of mammary epithelial cells from either fetal or adult glands were injected but thus revealed that even a single MRU could regenerate as many CFCs or EpCAM+ cells at 8 weeks post-transplant as are found in a normal adult mammary gland (up to 105 CFCs and 1.5 – 2 x 105 EpCAM+ cells from <100 EpCAM+ cells, Figure 3.4 and Figure 3.5, dotted line). Thus it appeared that the outputs of CFCs and EpCAM+ cells from MRUs may already reach saturation by 8 weeks even when low numbers of MRUs are transplanted. 67 More surprising was the finding in some experiments in which high numbers of unseparated cells were transplanted, that the number of CFCs or EpCAM+ cells produced were very low (>0.25 x 104 EpCAM+ fetal cells, and >6.5 x 104 EpCAM+ adult cells, Table 3.1, Figure 3.4 and Figure 3.5). These observations were also made when an attempt to measure the self-renewal activity of fetal MRUs was made using a double LDA experimental design that involved assessing the EpCAM+ cell content of each regenerated gland at both stages (Figure 3.6-A). In these experiments, I injected cleared fat pads with fetal cells calculated to contain either ~3 or ~15 MRUs and then 8 weeks later, injected the cells obtained from these fat pads into secondary recipients in limiting dilutions (Figure 3.6-C and Table 3.2). The results of this experiment showed that only 4 MRUs were produced on average in the one fat pad that had been injected with an estimated 15 fetal MRUs (i.e. a 4-fold decrease). In contrast, the cells harvested from the 3 fat pads initially transplanted with only ~3 MRUs each contained ~130 MRUs (i.e., a 10-fold increase, Table 3.2). These results suggest that MRU self-renewal might also be highly sensitive to the number (and/or types) of non-MRUs that are co-injected. 3.3.4 CFC and EpCAM+ cell outputs at 4 weeks correlate with input cell numbers and show fetal cells have regenerated more CFC and EpCAM+ cells at that time To determine a time point when the growth of MRUs might not yet have reached saturation and hence CFC and EpCAM+ cell outputs might be more correlated with the number of MRUs injected, I repeated the previous experiment using an earlier time point for the analysis. Assessment of CFCs and EpCAM+ cells 4 weeks after the transplantation of increasing numbers of EpCAM+ cells (ranging from 5 to 103 fetal cells and 50 to 3 x 104 adult cells) in either EpCAM purified or unseparated cell suspensions) revealed a similarly heterogeneous spread of CFC and EpCAM+ cell outputs (Figure 3.7). As shown, both EpCAM+ cell and CFC outputs 68 from any given cell dose extended over more than a 10- to 100-fold range, although evidence of saturating outputs at low cell doses and inhibited outputs at high cell doses appeared less than what had been seen at 8 weeks. For the fetal cells, there was a statistically acceptable linear relationship between CFC and EpCAM+ cell outputs and the number of cells injected (slopes of 0.83, 0.90 not significantly different from 1 for CFC and EpCAM+ cells recovered, P = 0.19, 0.47, Figure 3.7-C to F). For the adult cells, a linear input-output relationship was also obtained for transplant doses of between 50 and 2,600 EpCAM+ cells (slopes of 0.85, 0.77 not significantly different from 1 for CFC and EpCAM+ cells, P = 0.3, 0.13, Figure 3.7-E and F). However, inclusion of values from higher input doses caused the slope to become reduced (blue line, slope unconstrained, Figure 3.7-C and D), suggesting these represented doses where inhibition was occurring. Comparison of the fetal and adult mammary CFC and EpCAM+ outputs over the range of input cell doses where these increase linearly with the input showed that fetal mammary cells produce ~5-fold more epithelial cells and CFCs per starting input EpCAM+ cell compared to the adult (Figure 3.7-G). When we estimated the number of MRUs that would have been present in the cells transplanted, the increased CFC/EpCAM+ cell output of the fetal cells held true when the less stringent criteria MRU detection were used (Table 3.1 and Figure 3.8-B and D) but not when the more stringent criteria were applied (Figure 3.8-A and C) consistent with the observation that adult mammary cells include a subset not present in the fetal gland that regenerate smaller trees. 3.4 Discussion There are several findings that can be drawn from the work described in this Chapter. One is that the epithelial cell content of a primary regenerated mammary gland can provide a robust 69 and quantitative measure of a positive gland, and that it correlates with the production of CFCs and total epithelial cells produced. Second, this system demonstrates a higher growth potential of fetal as compared to adult mammary cells in vivo both in terms of total epithelial cell and CFC production. Third, is the evidence provided of both saturation and inhibition of CFC and EpCAM+ cell outputs that is pronounced by 8 weeks after transplantation, but reduced when these parameters are assessed 4 weeks earlier. As a result, extensive heterogeneity in these measures of MRU growth is seen at 8 weeks which is reduced at 4 weeks post-transplant sufficient to reveal the greater output potential of fetal as compared to adult MRUs. The results show that the criterion used to define MRU activity can indeed affect the frequency of MRU obtained. Here, I show that the total number of EpCAM+ cells present serves as a specific and objective measure of the size of the regenerated epithelial population and thus circumvents the need for a GFP+ donor. The minimal content of epithelial cells in large regenerated trees is 1.3 x 104 cells and imposition of this threshold eliminates >90% of the MRUs in the adult mammary gland that produce smaller trees. In contrast, fetal cells produce consistently larger trees than adult cells and the frequency of fetal MRUs is unaffected by reducing the tree size requirement. The frequency of adult MRU reported by others is highly variable (e.g., ranging from 1 in 100 to 1 in 30,000, Table 1.1) and this may well be explained, at least in part by variability in the threshold adopted by different investigators. The present studies now establish a method to standardize this parameter so that it does not compound investigation of other factors that may affect MRU detection. Examination of the CFC and EpCAM+ cell content of regenerated mammary trees in fat pads injected with <1 MRU provides insight into the heterogeneity in proliferative ability of individual MRUs. After 8 weeks post-transplant, trees regenerated from fetal cells spanned a range of sizes from ~ 104 to 4 x 105 EpCAM+ cells and contained from 200 to 105 CFCs. Due to 70 the predominant production of smaller trees from adult cells, their CFC outputs also spanned an even wider range than that characteristic of fetal MRUs. Notably, some trees regenerated in fat pads injected with a single MRU produced as many cells and CFCs as are present in an entire normal adult gland. This suggests that these outputs can reach saturating levels within the fat pad within 8 weeks. It is not known what determines the ultimate size that the normal mammary gland achieves, although it is clear that this is highly constrained since individual cells can be removed and stimulated either in vitro (Chapter 2) or in vivo54,55 to produce entire new glands of equivalent size and CFC content. Interestingly, with higher doses of injected cells in unseparated cell suspensions, occasional evidence of very low EpCAM+ cell and CFC outputs were observed. Since the innocula would have contained many MRUs, the simplest interpretation of these findings is that MRU activity can be inhibited by other co-injected cells. Variability in the size of regenerated glands obtained could be determined by the symmetry of the initial self-renewal divisions undertaken, by factors affecting their proliferative status and factors affecting their survival. Moreover, all of these may be regulated by intrinsic as well as extrinsic mechanisms, as exemplified by the causes of similar clonal variability identified to date in the hematopoietic system197. At 4 weeks, saturation of outputs by single MRUs and their inhibition by co-injected cells appeared to be diminished as expected for a system examined when it has not reached its full growth potential. Nevertheless, considerable heterogeneity was still observed suggesting an important role of early events contributing to this heterogeneity. Investigations of the hematopoietic system indicate that clone longevity but not initial size is unrelated to the self-renewal activity of the stem cell from which the clone originated198- 200. This is attributed to the large number of divisions that hematopoietic cells may undergo after they have left the stem cell compartment and during which time the final outputs of 71 differentiated cells may be modulated by factors unrelated to the self-renewal activity of the stem cell that initiates the formation of a clone. In the mammary system, it is not known whether there are short-lived/short-term MRUs that do not produce daughter MRU detectable in secondary fat pads assays. This might be anticipated in light of the observation that the mammary gland of pregnant mice, has increased numbers of MRUs (~10-fold more than in virgin adults), but their regenerative activity in secondary recipients appears reduced102. Careful serial transplantation experiments of large and small primary glands containing high and low CFC numbers may offer an avenue to explore this possibility. Another important variable to also keep in mind is the extrinsic hormonal environment of recipient animals. As none of these transplants involved pregnancy induction, or the transplantation of a hormonal pellet containing E and P, both of which show an increase in MRU detection77 (Figure 2.3), it is possible that some of the heterogeneity seen here may have been diminished if the recipient environment were more optimized for MRU activity. It would therefore be of interest to know whether inducing pregnancy or implanting a hormone pellet would alter the overall heterogeneity observed. In summary, the results presented in this Chapter provide 2 lines of evidence that primitive fetal mammary cells have a greater growth potential than their adult counterparts. The first is the finding that fetal cells consistently regenerate mammary trees 8 weeks after their transplantation into a cleared fat pad that contain a minimum of 1.3 x 104 EpCAM+ cells, whereas this size of structure is achieved by <10% of all adult mammary cells that can, nevertheless regenerate a detectable tree-like structure. Secondly, if regenerating trees are examined at an earlier time point (4 weeks post-transplant), the average output of EpCAM+ cells (and CFCs) per starting EpCAM+ cell (defined using the same stringent 8-week endpoint) is ~5- fold higher for fetal as compared to adult cells. Whether this is due to differences in self-renewal 72 probabilities, cell cycle length, proliferative activity or apoptosis control in these 2 stages of mammary cell development are all possible mechanisms that now need to be investigated. 73 100 102 104 106 - + ++ +++- 100 102 104 106 - + ++ G FP + ce lls /g la nd C B E pC A M GFP 3% A Adult or fetal Unseparated GFP+ cells 4/8 weeks Adult/4 week Fetal/8 week 100 102 104 106 - + ++ Adult/8 week Adult/fetal 8 week Figure 3.1 Classification of regenerated trees based on GFP/EpCAM expression (A) Experimental design. Unseparated adult or fetal GFP+ mammary cells are transplanted into pre- cleared fat pads and 4 or 8 weeks later, glands are visualized under a fluorescent microscope and then dissociated to calculate the numbers of GFP and/or EpCAM+ cells in each gland. Representative FACS plot shown on the right. (B) Representative images of what categories regenerated glands are assigned (-, no GFP detected/spheres or 1-2 ducts; +, small tree, with few ducts; ++, large branched structure that occupies most of the fat pad). (C) GFP+ cell content of 4 and 8 week regenerated mammary glands from adult and fetal cell sources. The dotted lines show the cutoffs for a positive gland based on more (>13,000 EpCAM+ cells upper dotted line) or less (>4,000 EpCAM+ cells lower dotted line) stringent criteria. 74 Table 3.1 Fetal and adult MRU frequency calculated based on different criteria for a positive regenerated gland. Source Cell dose Positive/total fat pads (more stringent) More stringent MRU frequency (95% CI) Positive/total fat pads (less stringent) Less stringent MRU frequency (95% CI) Fetal 200 0/8 1/8 250 3/3 3/3 1,000 10/14 1/1,700 13/14 1/900 3,000 1/2 (1,000 – 2,700) 2/2 (540 – 1,500) 4,000 4/7 5/7 5,000 6/6 6/6 10,000 3/3 3/3 25,000 4/6 4/6 50,000 2/2 2/2 80,000 2/2 2/2 Adult 200 2/6 4/6 1,000 2/3 3/3 5,000 2/3 1/8,600 3/3 1/180 10,000 4/5 (3,900 – 19,000) 5/5 (70 – 500) 30,000 3/3 3/3 50,000 1/2 2/2 90,000 3/3 3/3 250,000 0/2 0/2 270,000 2/2 2/2 Cell doses in italics represent doses at and above which inhibition was observed with negative fat pads 75 MRU-derived CFC Time post-transplant C FC o ut pu t Input Figure 3.2 Schematic showing the ideal window during which CFC should be measured In order to measure the production of CFC from the fat pad, CFC injected in the starting population have to be depleted (blue line) and CFC that are produced in the fat pad need to be expanding (linear part of the curve, red line, within the red shaded window) avoiding the saturation phase on the far right of the red curve. 76 C FC o ut pu t/g la nd EpCAM+ cells/gland 8 weeks4 weeks A B C C FC /E pC A M + ce ll 4wk 8wk Lu m in al C FC ( % o f i np ut ) Time post-transplant (weeks) Figure 3.3 CFC are produced in regenerated glands and correlate with EpCAM expression (A) 4,400 luminal (GFP+CD31-CD45-Ter119-CD49flowEpCAMhigh) cells containing ~200 CFCs were injected into cleared fat pads and glands then dissociated at weekly intervals to measure the CFC content in each gland. (B) A correlation between CFC and EpCAM+ cells at 4 and 8 weeks post-transplantation. Every dot represents one fat pad. Lines are the linear regression with the slope and R squared values stated.(C) Frequency of CFC output (per EpCAM+ cell) measured at 4 weeks and 8 weeks after transplantation. Error bars are SEM. No significant differences (P=0.39, two-tailed unpaired t-test). 77 E pC A M + ce lls /g la nd FetalAdult B E pC A M + ce lls /g la nd EpCAM+ cells MRU equivalents E pC A M + ce lls /g la nd C 42 43 44 45 46 47 48 42 43 44 45 46 47 A Figure 3.4 Low numbers of MRU saturate the fat pad at 8 weeks and reveal possible inhibition at high cell doses Unseparated or EpCAM+ mammary cells from adult and fetal glands are transplanted in varying input cell doses (expressed as EpCAM+ (A and B) or as MRU equivalent (C) doses) into cleared fat pads. 8 weeks later, regenerated glands are dissociated and a fraction analyzed for EpCAM expression to measure epithelial content in each gland. Each open or closed circle represents one dissociated gland. Each set of similar colored open circles represents dissociated glands from one experiment. Regenerated fat pads shown in panel (A) are binned according to the range of EpCAM+ cells transplanted. The dotted line represents the mean of total EpCAM+ cells from a normal adult virgin gland. Triangles represent glands that contained undetectable CFC. Shaded blue areas show the distribution of EpCAM+ cell output from adult glands. 78 C FC o ut pu t/g la nd FetalAdult A 42 43 44 45 46 4742 43 44 45 46 47 48 C FC o ut pu t/g la nd C FC o ut pu t/g la nd EpCAM+ cells MRU equivalents B C Figure 3.5 CFC outputs display similar outcomes of saturation and inhibition Unseparated or EpCAM+ mammary cells from adult or fetal glands are transplanted in various cell doses (expressed as EpCAM+ (A and B) or MRU equivalent doses (C)) into cleared fat pads. 8 weeks later, regenerated glands are dissociated and their CFC content is measured. Each open or closed circle represents one dissociated gland. Each set of similar colored open circles represents isolated glands from one experiment. Regenerated fat pads shown in panel (A) are binned according to the range of EpCAM+ cells transplanted. The dotted line represents the mean of total CFC from a normal adult virgin gland. Triangles represent undetected CFC measured at the limit of detection. Shaded blue areas show the distribution of CFC output from adult glands. 79 23.4 E pC A M FSC B C A 8 weeks 8 weeks 1o MRU assay 2o MRU assay Adult or fetal Unseparated or EpCAM+ cells FACS for EpCAM+ cells E pC A M + ce lls /g la nd 1° 2° EpCAM+ cells FACS for EpCAM+ cells Figure 3.6 In vivo self-renewal activity of fetal MRUs reveals inhibition of large input MRU Experimental design. Unseparated cells containing known numbers of EpCAM+ adult or fetal mammary cells (or FACS-purified EpCAM+ adult or fetal mammary cells) are transplanted into pre-cleared fat pads and 8 weeks later, all glands are dissociated and the numbers of MRUs they contain measured by performing limiting dilution transplants in secondary mice. (B) Representative FACS profile of a regenerated gland from a secondary recipient of cells regenerated from primary fetal cells. (C) No. of EpCAM+ cells in all primary (1°) and secondary (2°) fat pads obtained in 2 independent experiments with fetal mammary cells. The dotted line represents the EpCAM+ cell cut off of positive glands based on the more stringent criteria for an MRU. The green filled circles and blue filled circle show the primary fat pads used for the secondary transplants. 80 Table 3.2 In vivo expansion of MRUs in primary recipients of unseparated fetal mammary cells Cell Source No. of MRU/ 1o fat pad No. of 1o fat pads assayed Cells per 2o fat pad Positive/total fat pads Frequency of MRUs in cells injected into 2o fat pads Output MRU MRU expansion in 1o fat pads Fetal 15 1 1,000 1/2 1/180,000 4 0.2 10,000 0/3 (1/23,000 – 1/145,000) 50,000 0/3 3 3 5,000 1/2 1/5,800 130 10 20,000 2/2 (1/1400 – 1/24,000) Secondary limiting dilution analysis of MRU expansion in 1o mice transplanted with unseparated fetal mammary cells containing the indicated number of input MRUs calculated from the frequencies shown in Table 1. In Experiment 1, a primary regenerated gland containing 15 MRU (based on primary LDA) was transplanted in a second LDA shown in table. In Experiment 2, a total of 9 input MRU pooled from 3 regenerated pads containing ~3 MRU equivalents each. Output MRU numbers are based on the MRU frequency values measured and the total cell content of the primary glands. 81 EpCAM+ cells C FC o ut pu t/g la nd 102 103 104 105 41 42 43 44 4542 43 44 45 46 47 48 EpCAM+ cells C FC o ut pu t/g la nd Fetal and Adult EpCAM+ cells E pC A M + ce lls /g la nd C EpCAM+ cells C FC o ut pu t/g la nd Fetal and Adult D A B FetalAdult FetalAdult C FC /i np ut E pC A M + ce ll G E pC A M + ce lls /g la nd E C FC o ut pu t/g la nd F E pC A M + ce lls /in pu t E pC A M + ce ll * FetalAdult H Figure 3.7 Fetal mammary cells produce heterogeneous but higher CFC outputs per starting EpCAM+ cell after 4 weeks in vivo Unseparated or EpCAM+ mammary cells from fetal and adult glands are transplanted in varying input cell doses (expressed as EpCAM+ cells per cell dose) into cleared fat pads for 4 weeks, and the CFC output (A, B and C, E) or EpCAM+ cell output (D and F) is measured per gland. Each open or closed circle represents a dissociated fat pad. Similar colored symbols represent glands isolated from an individual experiment. C- F, Correlation between pooled fetal (green dots) and pooled adult (blue dots) CFC and EpCAM+ cell output from regenerated glands with a linear regression line. Slope not significantly different from 1 for green lines in C and D (P = 0.19, 0.47) and for green and blue lines in E and F (Blue lines P = 0.3, 0.13). (G) The CFC output per starting EpCAM+ input cell for EpCAM+ cell doses up to 2,600 cells where the slope is not significantly different from 1, similarly H shows the EpCAM+ cell output per input EpCAM+ cell. Unpaired two-tailed t-test *P<0.05. Error bars represent SEM. Triangles in C and E represent glands with undetected CFC measured at the limit of detection. 82 E pC A M + ce lls /g la nd C FC o ut pu t/g la nd E pC A M + ce lls /g la nd MRU equivalents BA C FC o ut pu t/g la nd DC More stringent criteria Less stringent criteria Figure 3.8 Differences between fetal and adult cell in vivo regeneration depends on the stringency of the criteria used to calculate MRU frequency CFC or EpCAM+ cell content of regenerated glands from fetal (green symbols) and adult (blue symbols) expressed relative to the estimated number of MRU in the starting cell population based on 2 readouts for a positive regenerated gland (Table 3.1). 83 4 Chapter: Discussion and future directions Chapter 4: Discussion and future directions The overall objective of the work described in this thesis was to compare the growth potential of very primitive cells in the fetal and adult mammary gland to determine whether these might be qualitatively different, as has been documented in other tissues. For this, it became rapidly apparent that existing methods to define and quantify mammary cells with stem cell functionality were inadequate for investigating this important developmental question. Accordingly, a first task was to establish the earliest time during the development of the mammary gland in the embryo when cells detectable as MRUs could be identified robustly and reproducibly. The next task was to set up an in vitro system that would allow both fetal and adult MRU activity to be examined and measured. Addressing these 2 tasks formed the foundation of the experiments described in Chapter 2. A third task was to develop a more objective and precise endpoint of MRU activity in the cleared fat pad assay so that the functionality of fetal and adult MRU in vivo could also be compared. Addressing this task formed the foundation of the experiments described in Chapter 3. The earliest time point at which robust MRU activity was detected in the embryo, was found to be at E17.5/18.5. This is in agreement with simultaneous studies80, which also showed that MRU can occasionally be detected at E15.5 when a higher concentration of Matrigel is added to the injection innoculum. In further time course studies, I found that the numbers of MRUs, CFCs and total epithelial cells (EpCAM+ cells) in the developing mammary gland increase progressively and in concert with one another. Subsequent characterization of their regenerative and growth properties at a clonal level in vitro (Chapter 2) and in a transplantation system in vivo (Chapter 3) shows that the outputs of individual fetal and adult cells are highly variable. Whether this heterogeneity is associated with cell intrinsic or extrinsic mechanisms is 84 not clear. Further investigation of this heterogeneity may reveal different types of repopulating cells in fetal and adult mammary glands. Nevertheless the outputs from the fetal cells appear to be greater. I also discovered that fetal and both basal and luminal cells in the adult have latent MRU activity that is only revealed after the cells have been stimulated to proliferate in vitro Table 2.10. These findings set the stage for examining the cellular and molecular mechanisms responsible for this heightened output of fetal cells, the timing of its switch to an “adult” phenotype and the explanation of the “latent” mammary stem cell activity. 4.1 Late detection of fetal MRUs The finding that MRU are not detected in dissociated cell suspensions of mammary tissue from the early embryo raises several questions about the acquisition of mammary stem cell activity in early mammary buds (prior to E15.5). My own and historical findings show that intact mammary fragments can engraft in the mammary fat pad and generate mammary structures in vivo as early as E13123. In Spike et al. 201280, some MRU activity could be detected at E15.5 in cell suspensions in the presence of 50% Matrigel, but most robustly when a rudimentary mammary tree structure is seen at E18.5. Overall, these findings suggest three possibilities; one is that a specific microenvironment provided by surrounding mesenchymal tissue of the developing mammary buds plays an important role in allowing dissociated mammary stem cells to display repopulating potential in the fat pad assay, and that this potential is not acquired in individually dissociated cells until later in gestation. Another is that the fat pad environment is not supportive of early mammary stem cell proliferation/survival or self-renewal divisions. As described in Chatper 1, the mammary buds of E11.5-15.5 embryos are not in contact with fat pad precursor tissue until ~E16, suggesting that exposure to differentiated adipocytes in the pubertal cleared fat pad upon transplantation may not be a supportive environment. Lastly, it is 85 possible that the lack of activity prior to E18.5 is due to an inability of mammary cells before that time to survive a dissociation procedure. This would again emphasize the importance of the surrounding microenvironment. Interestingly, in the developing hematopoietic system, it is well established that definitive HSCs are detected after the appearance of less primitive progenitors and differentiated red cells126,127. Whether or not mammary CFCs are present before E18.5 in the developing mammary gland is unclear. Mammary CFCs from early embryos are more difficult to quantify specifically due to the many contaminating EpCAM+ epidermal cells that form morphologically similar colonies in vitro. 4.2 Clonal analysis of fetal and adult mammary cell outputs Preliminary experiments in our group comparing different in vitro conditions indicated that human mammary CFCs and MRUs were optimally maintained and produced in the presence of Matrigel (Raouf et al., unpublished). I then tested this system to investigate the growth of mouse mammary cells, using a modified culture medium (Methods in Chapter 2) and the addition of irradiated 3T3 fibroblasts. This system was shown to support the extensive growth of single input mammary epithelial cells resulting in the output of large numbers of CFCs and some MRUs. Simultaneous with these observations was a report that fetal mammary epithelial cells can grow as single cells in liquid cultures at 10% efficiency80. However, as described in Chapter 2, I found that ~40% of single fetal cells expressing high levels of EpCAM generate progeny that possess MRU activity. Personal communication from Dr. John Stingl is that the activation of adult MRU properties in vitro under certain conditions is possible at a single cell level, and can happen within 24 hours (Prater M and Stingl J, unpublished). Recent work has shown that the addition of Wnt3a to 3D Matrigel cultures of basal cells, cultured at low density, 86 increases the absolute number of MRU in culture113. By coupling a method that supports the clonal growth of primitive mammary cells with a quantitative functional readout, it will now be possible to examine the mechanisms that activate regenerative potential at a single cell level and measure the effects of specific growth factors and other external stimuli on the output of single cells. Future experiments to examine such effects would need to involve the use of CM to prevent effects associated with unintended feeder manipulation. This raises the important question about the content of CM and the identity of the factors that are produced that maintain the remarkable regenerative potential observed. Continuing surveys of possible candidates (HGF, CSF-1, FGF and Wnt) have not yet proven fruitful. An analysis of the transcriptome of irradiated 3T3 cells, now underway, may be more informative. 4.3 In vitro activation of stem cell properties Another important finding described in Chapter 2 is the high frequency of cells after culture that possess self-renewal activity. Interestingly, cells within the luminal progenitor fraction, which have a very low frequency of stem cells (<1 in 15,000) are shown to generate structures that contain cells with MRU activity in vivo (Figure 2.12). Similar findings are now being reported in mouse mammary luminal progenitors which seem to acquire some repopulating activity after being “passaged” in vivo under the kidney capsule in collagen gels or in collagen/Matrigel gels91. In addition, basal and luminal cells have an increased potential to repopulate a mammary gland following the transduction of HGF, leading to constitutive Met activation65, as well after ectopic expression of Slug and Sox9175. The same properties seem to be present in single-cell cultures of adult basal and fetal cells, where the predicted MRU frequency post-culture is as high as 1 in 3 cells (Table 2.10). These findings suggest a >200-fold increase in the number of cells detected as MRU after culture. This raises several issues, one of which is whether our 87 current MRU assay is optimal for the detection of all stem cell activity. It is possible that conditions in which the assay is currently run are sub-optimal conditions for them to express their nascent full potential. The MRU properties measured may also have been activated or “acquired” due to the interaction of the cells with the irradiated feeders, or their liberated products, and/or growth factors present in the Matrigel. It is also not clear whether the MRU generated in culture are able to regenerate daughter MRUs detectable in secondary transplants. Preliminary experiments to address this question suggest that both basal and luminal structures generated in vitro contain MRUs able to produce daughter MRUs with secondary repopulating ability. The present examples of activation of latently “suppressed” MRU potential are in keeping with emerging observation of similar phenomena in other systems demonstrating that cellular states are less “fixed” than suggested by normal physiology and, in fact, are reversible under certain conditions. For example, skin wounding experiments have shown that cells that normally display a restricted potential to contribute to cells of the hair follicle can be activated to become long-lived epidermal stem cells176,201. In the intestine, progenitors of the secretory lineage that express high levels of Delta-like 1, a Notch ligand, are induced to express Lgr5 and generate small organoids upon addition of Wnt3a in vitro. Notably, these cells can also display stem cell activity in vivo following sublethal irradiation of the animal202. Recent work by Barroca and colleagues demonstrated that spermatgonoial progenitors can revert back to a stem cell state in vitro, although at low efficiency, by the addition of GDNF and FGF2, and subsequently repopulate germ cell-depleted testes upon transplantation203,204. The most striking examples of reversible changes in cellular hierarchies are given by the multitude of reprogramming experiments both back to an ES-like state from many types of relatively mature starting cells205 and those that enable direct reprogramming of cells across 88 tissue types as well as across cell lineages206,207. These observations also apply to malignant phenotypes, depending on the micro-environmental context of cells. Even cells with stem cell properties in tumors may not be fixed in their potential, which can be “lost” and “re-acquired” depending on extrinsic factors98,208,209. 4.4 Increased potency of fetal mammary cells The results in Chapter 2 demonstrate that fetal cells generate ~10-100-fold larger structures in the 7-day cultures described and these contain ~1000-fold more CFCs and ~3-fold more MRU as compared to purified basal and CD61+ luminal cells. In Chapter 3, an increased growth potential of fetal cells is further supported by the finding that fetal cells produce larger mammary trees in vivo and more CFCs per EpCAM+. In this section, I discuss the possible mechanisms that may be responsible for this enhanced output. The first is that fetal mammary cells overall undergo more symmetric self-renewal divisions than adult mammary cells. In order to expand the numbers of stem cells, fetal cells would be predicted to divide symmetrically more often and produce daughter progeny with the same regenerative potential. This may be possible considering the increased number of MRU (~3-fold) produced from single fetal cells compared to the adult, and the overall higher number of MRUs produced from 100 EpCAM+ cells (Table 2.8). Evidence for the enhanced self-renewal activity of fetal stem cells is well-documented in the hematopoietic and neural stem cell systems. In the hematopoietic system, fetal liver HSCs regenerate more progeny HSCs more rapidly than adult HSC, and also ultimately produce many more HSCs per input HSC in irradiated recipients179,180,210,211. In the brain, the mitotic activity and self-renewal activity of neural stem and progenitor cells also decreases during development181,212,213. Changing properties during development have included the differential self-renewal responses of fetal 89 cells to growth factors214 and a variation in response to factors directing lineage fate in neural crest progenitors181. The investigation of the self-renewal properties at different developmental stages of mammary stem cells is still at an early stage, but the recent documentation of this changing property during development will open up new avenues to explore the underlying mechanisms. Another property that could explain the increased MRU and CFC outputs of fetal cells would be possible differences in their cycling control, either affecting their cell cycle transit time and/or the proportions of these that are cycling as opposed to being in a quiescent state. To date, this has been poorly studied. In the adult, the very large increase in MRU numbers during pregnancy and the estrus cycle, suggests that mammary stem cells can be actively recruited into cycle77,78. However, attempts to functionally measure the cycling properties of mammary cells are limited to analyses Hoechst/Pyronin-stained cells from the adult gland where the majority (>95%) of mouse MRU and CFC have been found in the G1/ S/G2/M fractions54, and measurements of tritiated thymidine (3H-Tdr) and 5-bromo-2'-deoxyuridine (BrdU) labeling. These latter experiments have revealed that >50% of mammary cells incorporate 3H-Tdr but most lose this label after 2 weeks, with a few remaining label-retaining cells being eventually lost after 5 weeks215. Increased frequency of 3H-Tdr injections also suggests that a high proportion of mammary cells are in cycle, and can incorporate additional dyes such as BrdU216- 218. Unfortunately, because of the lack functional assays performed on the cells of interest, it is difficult to make direct inferences about the cycling properties of adult mammary MRU and CFCs. Little is known about the cycling properties of fetal mammary cells, although Ki-67 staining of mammary buds suggests that cells in the 14.5 day embryo are not mitotically active7. In addition fetal mammary stem cell signatures and fetal stromal signatures do not have 90 significant representation of proliferation-associated genes80. Experiments to address some of these questions include performing 3H-Tdr suicide assays to measure the numbers of CFC and MRU that are in S-phase in fetal compared to adult mammary cells. This could also be done using a transgenic mouse that expresses a fused histone 2B-GFP under the control of a tetracycline-responsive regulatory element, in which cycling cells incorporate GFP or lose GFP, depending on the inducible system219. In our system, it may be more relevant to measure the cell cycle status of fetal and adult cells in culture. In order to investigate the time at which cells of fetal and adult sources enter the cell cycle as well as measure their cell cycle transit time, one could envisage employing single-cell tracking methods to monitor division kinetics in a microfluidics system, as shown for HSCs220. The observed higher output of fetal cells could also be attributed to a decreased rate of cell death as compared to adult cells. Future analysis of their gene expression profile after culture may reveal insight into the representation of apoptosis-related genes. In the hematopoietic system, overexpression of B-cell lymphoma -2 (Bcl-2), a pro-survival gene, leads to an increase in HSC numbers221,222. The role of apoptosis in the mammary gland has historically been associated with the involution phase223, but little is known about its possible role in regulating mammary stem cell outputs. There may also be intrinsic molecular mechanisms that regulate fetal and adult stem cell self-renewal differently. In the hematopoietic system, these have included transcription factors, such as Sox17, specifically expressed in fetal HSCs and its deletion in the fetus or neonate eliminates reconstituting activity or induces cell death, but this factor seems to be dispensable for adult HSC maintenance183. Additional factors that appear to be important for the maintenance of self-renewal programs in the fetus include chromatin regulators such as Ezh2, AML1 and the Let7-Hmga2 family145,181-184. Interestingly, some of these genes are now being 91 implicated as also important for normal mammary stem cell self-renewal. For example, deletion of Ezh2 using an MMTV mouse model reduces MRU frequency by ~14-fold148. How these regulators influence fetal mammary stem cell self-renewal properties remains to be investigated, but there is evidence that Hmga2, which plays a role in neural stem cell self-renewal182, and HSC self-renewal (Copley M et al., submitted), is highly expressed in fetal (E15.5) mammary buds compared to adult but its expression declines by 18.580. Although some of these regulators are important for maintaining fetal stem cell properties, there are others that are important for maintaining adult stem cell self-renewal. These include Bmi-1, which also plays a role in regulating the self-renewal properties of hematopoietic, neural, and prostate stem cells146,147,224,225. The Wnt signalling pathway in adult mammary tissue113, in the intestine226, and skin227 is another such example. However, we demonstrate that Wnt signalling is not essential for the production of CFCs from fetal or adult cells in our culture system (Figure 2.9). 4.5 Heterogeneity in cell outputs from fetal and adult mammary cells There is very little known about the heterogeneity of normal mammary cell growth properties. Chapter 2 of this thesis documents the profound heterogeneity in primitive mammary cell activity in vitro and Chapter 3 shows it also occurs in vivo. At any given cell dose, the CFC and EpCAM+ cell output of primitive mammary cells from fetal or adult cells is distributed over 2 orders of magnitude even after 4 weeks when the regenerating glands have not yet finished growing. This suggests there may be intrinsic and/or extrinsic components that influence the in vivo growth displayed by individual MRUs. These findings also raise many questions; namely, are there different subsets of MRU (short-term or long-term, high CFC-producing or low CFC-producing) that are intrinsically 92 programmed and contribute differentially to the observed output? Another possibility is whether this is the result of stochastic events that may or may not be significantly influenced extrinsically by the stimulatory or inhibitory effect of other mammary or non-epithelial cells or paracrine and/or hormone factors in the surrounding environment. One approach to further investigate this heterogeneity is through the identification of novel surface markers to prospectively isolate different MRU subsets at high purities to allow studies on the repopulation kinetics of single cells transplanted in isolation. Thus far, this approach has not been successful, as MRUs and CFCs are both found in the basal fractions and no markers have yet been identified to separate the two functional subsets54,106,228. Alternate methods that we are currently using to explore the heterogeneity in the lineage output of basal cells in primary and secondary regenerated glands, in place of large limiting dilution experiments, include a bar-coding strategy that aims to track the progeny of single cells in vivo with a unique 27-nucleotide non-coding barcode sequence downstream of a fluorescent reporter. Coupled with high-throughput sequencing of barcodes from a large population of regenerated cells, we can identify sequences at high depth and resolution to yield accurate quantitative measures of clonal expansion, and content of cell types generated. The bar-coding approach builds upon previous low resolution gene-tracking strategies that have been employed in the mammary system87. The bar-coding strategy can now be used with functional assays and thus allows us to make inferences about the proliferative and regenerative potentials of the progeny of clonally tracked cells. Several studies in the hematopoietic system have also used such a method to track the output of individual HSC clones229,230. We have recently employed this methodology to uniquely label mouse and human basal mammary cells in vivo (Nguyen, Makarem et al. unpublished). Our findings confirm that mouse basal cells display extensive heterogeneity in the 93 size and lineage contributions they make to individual clones generated simultaneously in vivo. Strikingly, we find that some clones that contribute to both basal and luminal lineages (so called ‘bipotent’ clones) can be detected in both primary and secondary recipients, and conversely, some are only detected in secondary recipients, or are lineage-restricted in primary recipients and later give rise to large bipotent clones (Nguyen, Makarem et al. unpublished). But this approach also comes with some caveats. These include the fact that cells have to be in culture for a period of time before being transplanted in order to select for the transduced cells based on their expression of the co-transduced GFP DNA. Transplantation of exclusively transduced cells helps to ensure that rare cells of interest will be tracked but they must be initially transduced at a relatively low efficiency in order to ensure that each cell takes up only one barcode. In addition, there are caveats associated with analyzing large sequencing data sets where a large number of clones are small and may fall below the threshold of detection. Defining the threshold thus becomes critical for interpreting what are real clones. It is likely that a combination of stochastic, intrinsic and extrinsic factors will be found to contribute to the heterogeneous growth behaviour of individual mammary stem cells as shown in other tissues197,231. In the mammary system, extrinsic factors such as TGFβ and CSF-1 play an important role in mammary stem cell proliferation and self-renewal93,232 and additional extrinsic and intrinsic regulators may be identified in future gene expression analyses when stem cell purities are higher. 4.6 Implications for breast cancer Many of these findings will inform on our current understanding of breast cancer and may provide us with new methods to study malignant cells and understand the mechanisms important for their maintenance, proliferation and activation. The majority of breast cancer research has 94 focused on the use of breast cancer cell lines to study the mechanisms that are perturbed in these cells, and many of these may not be true representations of the tumors from which they arise233,234. It would be of interest to study the effects of using the culture conditions developed in this thesis on the growth of primary human breast tumor tissue. Taking into account recent findings of tumor-initiating cells in primary breast tumors, one could envision the use of 3D culture conditions to explore ways of studying the heterogeneity of breast tumors in vitro, and even genetically manipulate normal and malignant cells. The finding of higher regenerative activity of fetal compared with adult mammary stem cells will shed light on the mechanisms that are important for this temporal “switch” in activity. Are the mechanisms activated in breast cancer similar to those important for maintaining the high regenerative activity of fetal mammary cells? The fetal mammary stem cell signature has been found to be over-represented in poorly differentiated breast tumors (basal-like)80, and additional molecular pathways involved in fetal self-renewal properties may inform on similar pathways activated during malignancy. Some of these questions are starting to be explored in the mammary field80, although the activation of “fetal programs” alone may not be sufficient to confer malignancy, as shown by the overexpression of the fetal-specific self-renewal transcription factor Sox17192 in adult HSC which leads to fetal HSC properties but only results in nonlymphoid leukemic transformation after secondary transplantation, suggesting that some of these key factors can act in concert with other mutations leading to malignant transformation. 4.7 Concluding comments The findings presented here open up many avenues to further understand the regulation of stem cell self-renewal and differentiation properties. A comparison of the gene expression profiles of 95 fetal and adult mammary stem cells can reveal novel mechanisms that regulate changes in regenerative potentials. Another interesting direction is to identify the molecular pathways important for activating the latent regenerative potential common to fetal and adult stem cells in culture. Such investigations are important considering the relevance of these properties to mammary tumorigenesis. Additional questions pertain to the role of the microenvironment in determining mammary stem cell competency in the developing embryo. Can the manipulation of dissociation procedures and the mammary transplantation assay affect MRU detection? This has already been touched upon through the use of Matrigel and the stimulation with ovarian hormones, but there may be additional ways that the transplantation assay can be manipulated to optimize the detection of MRU such as the addition of stromal feeders or other factors to the cell innoculum as is the case in the human MRU assay95,106. The question of whether such manipulations activate or “reprogram” mammary cells will always be a more challenging one to address. Additionally, the heterogeneity revealed in MRU potential both in vitro, in individual clones, and in vivo, at low MRU inputs, raises important questions about whether smaller or large clones both possess long-term self-renewal ability. What is the mechanistic basis for the observed heterogeneity in potential? The difference in MRU potential between fetal and adult MRU also raises the question of when, at what developmental time point, do such differences arise? In the hematopoietic system, a remarkable switch in the cycling properties, expansion potential and differentiation output occurs between 3 and 4 weeks after birth180,214, which is the time at which the secretion of ovarian hormones begins. Further investigation of when such a switch occurs in MRU potency will be very meaningful for further investigating mammary stem cell self-renewal properties. 96 In summary, the isolation and characterization of mammary stem cells from the developing embryo, and the demonstration of their enhanced regenerative and proliferative properties has set the stage for investigating the mechanisms that are important for establishing and maintaining the expression of mammary stem cell properties and how these may change during development. 97 References 1 Makarem, M. et al. Stem Cells and the Developing Mammary Gland. J Mammary Gland Biol Neoplasia (2013). 2 Richert, M. M., Schwertfeger, K. L., Ryder, J. W. & Anderson, S. M. An atlas of mouse mammary gland development. J Mammary Gland Biol Neoplasia 5, 227-241 (2000). 3 Pispa, J. & Thesleff, I. Mechanisms of ectodermal organogenesis. Dev Biol 262, 195-205 (2003). 4 Propper, A. Y. Wandering epithelial cells in the rabbit embryo milk line. A preliminary scanning electron microscope study. Dev Biol 67, 225-231 (1978). 5 Veltmaat, J. M., Van Veelen, W., Thiery, J. P. & Bellusci, S. Identification of the mammary line in mouse by Wnt10b expression. Dev Dyn 229, 349-356 (2004). 6 Balinsky, B. I. On the prenatal growth of the mammary gland rudiment in the mouse. J Anat 84, 227-235 (1950). 7 Howard, B. A. In the beginning: the establishment of the mammary lineage during embryogenesis. Semin Cell Dev Biol 23, 574-582 (2012). 8 Rohrschneider, L. R., Custodio, J. M., Anderson, T. A., Miller, C. P. & Gu, H. The intron 5/6 promoter region of the ship1 gene regulates expression in stem/progenitor cells of the mouse embryo. Dev Biol 283, 503-521 (2005). 9 Kratochwil, K. & Schwartz, P. Tissue interaction in androgen response of embryonic mammary rudiment of mouse: identification of target tissue for testosterone. Proc Natl Acad Sci U S A 73, 4041-4044 (1976). 10 Durnberger, H. & Kratochwil, K. Specificity of tissue interaction and origin of mesenchymal cells in the androgen response of the embryonic mammary gland. Cell 19, 465-471 (1980). 11 Slavin, B. G. Fine structural studies on white adipocyte differentiation. Anat Rec 195, 63-72 (1979). 12 Sakakura, T., Sakagami, Y. & Nishizuka, Y. Dual origin of mesenchymal tissues participating in mouse mammary gland embryogenesis. Dev Biol 91, 202-207 (1982). 13 Neville, M. & Daniel, C. The Mammary gland: Development, regulation, and function. 37-66 (1987). 14 Propper, A. & Gomot, L. [Tissue interactions during organogenesis of the mammary gland in the rabbit embryo]. C R Acad Sci Hebd Seances Acad Sci D 264, 2573-2575 (1967). 15 Cunha, G. R. et al. Mammary phenotypic expression induced in epidermal cells by embryonic mammary mesenchyme. Acta Anat (Basel) 152, 195-204 (1995). 16 Sakakura, T., Kusano, I., Kusakabe, M., Inaguma, Y. & Nishizuka, Y. Biology of mammary fat pad in fetal mouse: capacity to support development of various fetal epithelia in vivo. Development 100, 421-430 (1987). 17 Couldrey, C. et al. Adipose tissue: a vital in vivo role in mammary gland development but not differentiation. Dev Dyn 223, 459-468 (2002). 18 Hogg, N. A., Harrison, C. J. & Tickle, C. Lumen formation in the developing mouse mammary gland. J Embryol Exp Morphol 73, 39-57 (1983). 19 Mallepell, S., Krust, A., Chambon, P. & Brisken, C. Paracrine signaling through the epithelial estrogen receptor alpha is required for proliferation and morphogenesis in the mammary gland. Proc Natl Acad Sci U S A 103, 2196-2201 (2006). 98 20 Hovey, R. C., Trott, J. F. & Vonderhaar, B. K. Establishing a framework for the functional mammary gland: from endocrinology to morphology. J Mammary Gland Biol Neoplasia 7, 17-38 (2002). 21 Bern, H. A. et al. Long-term alterations in histology and steroid receptor levels of the genital tract and mammary gland following neonatal exposure of female BALB/cCrgl mice to various doses of diethylstilbestrol. Cancer Res 47, 4165-4172 (1987). 22 Durando, M. et al. Prenatal bisphenol A exposure induces preneoplastic lesions in the mammary gland in Wistar rats. Environ Health Perspect 115, 80-86 (2007). 23 Murray, T. J., Maffini, M. V., Ucci, A. A., Sonnenschein, C. & Soto, A. M. Induction of mammary gland ductal hyperplasias and carcinoma in situ following fetal bisphenol A exposure. Reprod Toxicol 23, 383-390 (2007). 24 Humphreys, R. C. et al. Apoptosis in the terminal endbud of the murine mammary gland: a mechanism of ductal morphogenesis. Development 122, 4013-4022 (1996). 25 Zinser, G., Packman, K. & Welsh, J. Vitamin D(3) receptor ablation alters mammary gland morphogenesis. Development 129, 3067-3076 (2002). 26 Daniel, C. W., Silberstein, G. B. & Strickland, P. Direct action of 17 beta-estradiol on mouse mammary ducts analyzed by sustained release implants and steroid autoradiography. Cancer Res 47, 6052-6057 (1987). 27 Bocchinfuso, W. P. et al. Induction of mammary gland development in estrogen receptor-alpha knockout mice. Endocrinology 141, 2982-2994 (2000). 28 Lydon, J. P. et al. Mice lacking progesterone receptor exhibit pleiotropic reproductive abnormalities. Genes Dev 9, 2266-2278 (1995). 29 Gallego, M. I. et al. Prolactin, growth hormone, and epidermal growth factor activate Stat5 in different compartments of mammary tissue and exert different and overlapping developmental effects. Dev Biol 229, 163-175 (2001). 30 Wiesen, J. F., Young, P., Werb, Z. & Cunha, G. R. Signaling through the stromal epidermal growth factor receptor is necessary for mammary ductal development. Development 126, 335-344 (1999). 31 Brisken, C. et al. Prolactin controls mammary gland development via direct and indirect mechanisms. Dev Biol 210, 96-106 (1999). 32 Liu, X. et al. Stat5a is mandatory for adult mammary gland development and lactogenesis. Genes Dev 11, 179-186 (1997). 33 Shillingford, J. M. et al. Jak2 is an essential tyrosine kinase involved in pregnancy- mediated development of mammary secretory epithelium. Mol Endocrinol 16, 563-570 (2002). 34 Ormandy, C. J. et al. Null mutation of the prolactin receptor gene produces multiple reproductive defects in the mouse. Genes Dev 11, 167-178 (1997). 35 Gouon-Evans, V., Rothenberg, M. E. & Pollard, J. W. Postnatal mammary gland development requires macrophages and eosinophils. Development 127, 2269-2282 (2000). 36 Dulbecco, R., Allen, W. R., Bologna, M. & Bowman, M. Marker evolution during the development of the rat mammary gland: stem cells identified by markers and the role of myoepithelial cells. Cancer Res 46, 2449-2456 (1986). 37 Rudland, P. S. & Hughes, C. M. Immunocytochemical identification of cell types in human mammary gland: variations in cellular markers are dependent on glandular topography and differentiation. J Histochem Cytochem 37, 1087-1100 (1989). 38 Deugnier, M. A., Teuliere, J., Faraldo, M. M., Thiery, J. P. & Glukhova, M. A. The importance of being a myoepithelial cell. Breast Cancer Res 4, 224-230 (2002). 99 39 Gudjonsson, T., Adriance, M. C., Sternlicht, M. D., Petersen, O. W. & Bissell, M. J. Myoepithelial cells: their origin and function in breast morphogenesis and neoplasia. J Mammary Gland Biol Neoplasia 10, 261-272 (2005). 40 Parmar, H. & Cunha, G. R. Epithelial-stromal interactions in the mouse and human mammary gland in vivo. Endocr Relat Cancer 11, 437-458 (2004). 41 Fata, J. E., Werb, Z. & Bissell, M. J. Regulation of mammary gland branching morphogenesis by the extracellular matrix and its remodeling enzymes. Breast Cancer Res 6, 1-11 (2004). 42 Dairkee, S. & Heid, H. W. Cytokeratin profile of immunomagnetically separated epithelial subsets of the human mammary gland. In Vitro Cell Dev Biol Anim 29A, 427- 432 (1993). 43 Mikaelian, I. et al. Expression of terminal differentiation proteins defines stages of mouse mammary gland development. Vet Pathol 43, 36-49 (2006). 44 Sun, P., Yuan, Y., Li, A., Li, B. & Dai, X. Cytokeratin expression during mouse embryonic and early postnatal mammary gland development. Histochem Cell Biol 133, 213-221 (2010). 45 Stingl, J., Raouf, A., Emerman, J. T. & Eaves, C. J. Epithelial progenitors in the normal human mammary gland. J Mammary Gland Biol Neoplasia 10, 49-59 (2005). 46 Signoretti, S. et al. p63 is a prostate basal cell marker and is required for prostate development. Am J Pathol 157, 1769-1775 (2000). 47 Barbareschi, M. et al. p63, a p53 homologue, is a selective nuclear marker of myoepithelial cells of the human breast. Am J Surg Pathol 25, 1054-1060 (2001). 48 Mills, A. A. et al. p63 is a p53 homologue required for limb and epidermal morphogenesis. Nature 398, 708-713 (1999). 49 McKeon, F. p63 and the epithelial stem cell: more than status quo? Genes Dev 18, 465- 469 (2004). 50 Raouf, A. et al. Transcriptome analysis of the normal human mammary cell commitment and differentiation process. Cell Stem Cell 3, 109-118 (2008). 51 Stingl, J., Eaves, C. J., Kuusk, U. & Emerman, J. T. Phenotypic and functional characterization in vitro of a multipotent epithelial cell present in the normal adult human breast. Differentiation 63, 201-213 (1998). 52 Wansbury, O. et al. Transcriptome analysis of embryonic mammary cells reveals insights into mammary lineage establishment. Breast Cancer Res 13, R79 (2011). 53 Van Keymeulen, A. et al. Distinct stem cells contribute to mammary gland development and maintenance. Nature 479, 189-193 (2011). 54 Stingl, J. et al. Purification and unique properties of mammary epithelial stem cells. Nature 439, 993-997 (2006). 55 Shackleton, M. et al. Generation of a functional mammary gland from a single stem cell. Nature 439, 84-88 (2006). 56 Asselin-Labat, M. L. et al. Gata-3 is an essential regulator of mammary-gland morphogenesis and luminal-cell differentiation. Nat Cell Biol 9, 201-209 (2007). 57 Taddei, I. et al. Integrins in mammary gland development and differentiation of mammary epithelium. J Mammary Gland Biol Neoplasia 8, 383-394 (2003). 58 Klinowska, T. C. et al. Epithelial development and differentiation in the mammary gland is not dependent on alpha 3 or alpha 6 integrin subunits. Dev Biol 233, 449-467 (2001). 59 Chen, J., Diacovo, T. G., Grenache, D. G., Santoro, S. A. & Zutter, M. M. The alpha(2) integrin subunit-deficient mouse: a multifaceted phenotype including defects of branching morphogenesis and hemostasis. Am J Pathol 161, 337-344 (2002). 100 60 Taddei, I. et al. Beta1 integrin deletion from the basal compartment of the mammary epithelium affects stem cells. Nat Cell Biol 10, 716-722 (2008). 61 Kamalati, T., Niranjan, B., Yant, J. & Buluwela, L. HGF/SF in mammary epithelial growth and morphogenesis: in vitro and in vivo models. J Mammary Gland Biol Neoplasia 4, 69-77 (1999). 62 Yant, J., Buluwela, L., Niranjan, B., Gusterson, B. & Kamalati, T. In vivo effects of hepatocyte growth factor/scatter factor on mouse mammary gland development. Exp Cell Res 241, 476-481 (1998). 63 Daniel, C. W., Silberstein, G. B., Van Horn, K., Strickland, P. & Robinson, S. TGF-beta 1-induced inhibition of mouse mammary ductal growth: developmental specificity and characterization. Dev Biol 135, 20-30 (1989). 64 Yang, Y. et al. Sequential requirement of hepatocyte growth factor and neuregulin in the morphogenesis and differentiation of the mammary gland. J Cell Biol 131, 215-226 (1995). 65 Gastaldi, S. et al. Met signaling regulates growth, repopulating potential and basal cell- fate commitment of mammary luminal progenitors: implications for basal-like breast cancer. Oncogene (2012). 66 Nelson, C. M., Vanduijn, M. M., Inman, J. L., Fletcher, D. A. & Bissell, M. J. Tissue geometry determines sites of mammary branching morphogenesis in organotypic cultures. Science 314, 298-300 (2006). 67 Luetteke, N. C. et al. Targeted inactivation of the EGF and amphiregulin genes reveals distinct roles for EGF receptor ligands in mouse mammary gland development. Development 126, 2739-2750 (1999). 68 Wansbury, O. et al. Dynamic expression of Erbb pathway members during early mammary gland morphogenesis. J Invest Dermatol 128, 1009-1021 (2008). 69 Howard, B. A. The role of NRG3 in mammary development. J Mammary Gland Biol Neoplasia 13, 195-203 (2008). 70 Daniel, C. W. & Robinson, S. D. Regulation of mammary growth and function by TGF- beta. Mol Reprod Dev 32, 145-151 (1992). 71 Dunbar, M. E. et al. Stromal cells are critical targets in the regulation of mammary ductal morphogenesis by parathyroid hormone-related protein. Dev Biol 203, 75-89 (1998). 72 Wood, T. L., Richert, M. M., Stull, M. A. & Allar, M. A. The insulin-like growth factors (IGFs) and IGF binding proteins in postnatal development of murine mammary glands. J Mammary Gland Biol Neoplasia 5, 31-42 (2000). 73 de Ostrovich, K. K. et al. Paracrine overexpression of insulin-like growth factor-1 enhances mammary tumorigenesis in vivo. Am J Pathol 173, 824-834 (2008). 74 Flint, D. J., Tonner, E., Beattie, J. & Allan, G. J. Role of insulin-like growth factor binding proteins in mammary gland development. J Mammary Gland Biol Neoplasia 13, 443-453 (2008). 75 Daza, D. O., Sundstrom, G., Bergqvist, C. A., Duan, C. & Larhammar, D. Evolution of the insulin-like growth factor binding protein (IGFBP) family. Endocrinology 152, 2278-2289 (2011). 76 Lewis, M. T. Homeobox genes in mammary gland development and neoplasia. Breast Cancer Res 2, 158-169 (2000). 77 Asselin-Labat, M. L. et al. Control of mammary stem cell function by steroid hormone signalling. Nature 465, 798-802 (2010). 101 78 Joshi, P. A. et al. Progesterone induces adult mammary stem cell expansion. Nature 465, 803-807 (2010). 79 Van Keymeulen, A. & Blanpain, C. Tracing epithelial stem cells during development, homeostasis, and repair. J Cell Biol 197, 575-584 (2012). 80 Spike, B. T. et al. A mammary stem cell population identified and characterized in late embryogenesis reveals similarities to human breast cancer. Cell Stem Cell 10, 183-197 (2012). 81 van Amerongen, R., Bowman, A. N. & Nusse, R. Developmental Stage and Time Dictate the Fate of Wnt/beta-Catenin-Responsive Stem Cells in the Mammary Gland. Cell Stem Cell 11, 387-400 (2012). 82 Deome, K. B., Faulkin, L. J., Jr., Bern, H. A. & Blair, P. B. Development of mammary tumors from hyperplastic alveolar nodules transplanted into gland-free mammary fat pads of female C3H mice. Cancer Res 19, 515-520 (1959). 83 Hoshino, K. & Gardner, W. U. Transplantability and life span of mammary gland during serial transplantation in mice. Nature 213, 193-194 (1967). 84 Hoshino, K. Regeneration and Growth of Quantitatively Transplanted Mammary Glands of Normal Female Mice. Anat Rec 150, 221-235 (1964). 85 Young, L. J., Medina, D., DeOme, K. B. & Daniel, C. W. The influence of host and tissue age on life span and growth rate of serially transplanted mouse mammary gland. Exp Gerontol 6, 49-56 (1971). 86 Daniel, C. W., De Ome, K. B., Young, J. T., Blair, P. B. & Faulkin, L. J., Jr. The in vivo life span of normal and preneoplastic mouse mammary glands: a serial transplantation study. Proc Natl Acad Sci U S A 61, 53-60 (1968). 87 Kordon, E. C. & Smith, G. H. An entire functional mammary gland may comprise the progeny from a single cell. Development 125, 1921-1930 (1998). 88 Diallo, R. et al. Monoclonality in normal epithelium and in hyperplastic and neoplastic lesions of the breast. J Pathol 193, 27-32 (2001). 89 Tsai, Y. C. et al. Contiguous patches of normal human mammary epithelium derived from a single stem cell: implications for breast carcinogenesis. Cancer Res 56, 402-404 (1996). 90 Daniel, C. W. & Deome, K. B. Growth of Mouse Mammary Glands in Vivo after Monolayer Culture. Science 149, 634-636 (1965). 91 Shehata, M. et al. Phenotypic and functional characterization of the luminal cell hierarchy of the mammary gland. Breast Cancer Res 14, R134 (2012). 92 Smalley, M. J. et al. Isolation of Mouse Mammary Epithelial Subpopulations: A Comparison of Leading Methods. J Mammary Gland Biol Neoplasia (2012). 93 Gyorki, D. E., Asselin-Labat, M. L., van Rooijen, N., Lindeman, G. J. & Visvader, J. E. Resident macrophages influence stem cell activity in the mammary gland. Breast Cancer Res 11, R62 (2009). 94 Lim, E. et al. Aberrant luminal progenitors as the candidate target population for basal tumor development in BRCA1 mutation carriers. Nat Med 15, 907-913 (2009). 95 Kuperwasser, C. et al. Reconstruction of functionally normal and malignant human breast tissues in mice. Proc Natl Acad Sci U S A 101, 4966-4971 (2004). 96 Vukicevic, S. et al. Identification of multiple active growth factors in basement membrane Matrigel suggests caution in interpretation of cellular activity related to extracellular matrix components. Exp Cell Res 202, 1-8 (1992). 97 Kleinman, H. K. & Martin, G. R. Matrigel: basement membrane matrix with biological activity. Semin Cancer Biol 15, 378-386 (2005). 102 98 Quintana, E. et al. Phenotypic heterogeneity among tumorigenic melanoma cells from patients that is reversible and not hierarchically organized. Cancer Cell 18, 510-523 (2010). 99 Quintana, E. et al. Efficient tumour formation by single human melanoma cells. Nature 456, 593-598 (2008). 100 Lewis, M. T., Landua, J. D., Adams, H. C., 3rd & Medina, D. A mystery wrapped in an enigma: matrigel enhancement of mammary cell growth and morphogenesis. J Mammary Gland Biol Neoplasia 17, 99-101 (2012). 101 Vaillant, F., Lindeman, G. J. & Visvader, J. E. Jekyll or Hyde: does Matrigel provide a more or less physiological environment in mammary repopulating assays? Breast Cancer Res 13, 108 (2011). 102 Asselin-Labat, M. L. et al. Gata-3 negatively regulates the tumor-initiating capacity of mammary luminal progenitor cells and targets the putative tumor suppressor caspase-14. Mol Cell Biol 31, 4609-4622 (2011). 103 Regan, J. L. et al. c-Kit is required for growth and survival of the cells of origin of Brca1-mutation-associated breast cancer. Oncogene 31, 869-883 (2012). 104 Sleeman, K. E. et al. Dissociation of estrogen receptor expression and in vivo stem cell activity in the mammary gland. J Cell Biol 176, 19-26 (2007). 105 Badders, N. M. et al. The Wnt receptor, Lrp5, is expressed by mouse mammary stem cells and is required to maintain the basal lineage. PLoS One 4, e6594 (2009). 106 Eirew, P. et al. A method for quantifying normal human mammary epithelial stem cells with in vivo regenerative ability. Nat Med 14, 1384-1389 (2008). 107 O'Hare, M. J., Ormerod, M. G., Monaghan, P., Lane, E. B. & Gusterson, B. A. Characterization in vitro of luminal and myoepithelial cells isolated from the human mammary gland by cell sorting. Differentiation 46, 209-221 (1991). 108 Emerman, J. T., Stingl, J., Petersen, A., Shpall, E. J. & Eaves, C. J. Selective growth of freshly isolated human breast epithelial cells cultured at low concentrations in the presence or absence of bone marrow cells. Breast Cancer Res Treat 41, 147-159 (1996). 109 Dundas, S. R., Ormerod, M. G., Gusterson, B. A. & O'Hare, M. J. Characterization of luminal and basal cells flow-sorted from the adult rat mammary parenchyma. J Cell Sci 100 ( Pt 3), 459-471 (1991). 110 Smalley, M. J., Titley, J. & O'Hare, M. J. Clonal characterization of mouse mammary luminal epithelial and myoepithelial cells separated by fluorescence-activated cell sorting. In Vitro Cell Dev Biol Anim 34, 711-721 (1998). 111 Stingl, J., Eaves, C. J., Zandieh, I. & Emerman, J. T. Characterization of bipotent mammary epithelial progenitor cells in normal adult human breast tissue. Breast Cancer Res Treat 67, 93-109 (2001). 112 Bruna, A. et al. TGFbeta induces the formation of tumour-initiating cells in claudinlow breast cancer. Nat Commun 3, 1055 (2012). 113 Zeng, Y. A. & Nusse, R. Wnt proteins are self-renewal factors for mammary stem cells and promote their long-term expansion in culture. Cell Stem Cell 6, 568-577 (2010). 114 Debnath, J., Muthuswamy, S. K. & Brugge, J. S. Morphogenesis and oncogenesis of MCF-10A mammary epithelial acini grown in three-dimensional basement membrane cultures. Methods 30, 256-268 (2003). 115 Sato, T. et al. Single Lgr5 stem cells build crypt-villus structures in vitro without a mesenchymal niche. Nature 459, 262-265 (2009). 116 Barker, N. et al. Lgr5(+ve) stem cells drive self-renewal in the stomach and build long- lived gastric units in vitro. Cell Stem Cell 6, 25-36 (2010). 103 117 Villadsen, R. et al. Evidence for a stem cell hierarchy in the adult human breast. J Cell Biol 177, 87-101 (2007). 118 Dontu, G. et al. In vitro propagation and transcriptional profiling of human mammary stem/progenitor cells. Genes Dev 17, 1253-1270 (2003). 119 Reynolds, B. A. & Weiss, S. Generation of neurons and astrocytes from isolated cells of the adult mammalian central nervous system. Science 255, 1707-1710 (1992). 120 Ginestier, C. et al. ALDH1 is a marker of normal and malignant human mammary stem cells and a predictor of poor clinical outcome. Cell Stem Cell 1, 555-567 (2007). 121 Booth, B. W., Boulanger, C. A. & Smith, G. H. Alveolar progenitor cells develop in mouse mammary glands independent of pregnancy and lactation. J Cell Physiol 212, 729-736 (2007). 122 Asselin-Labat, M. L. et al. Steroid hormone receptor status of mouse mammary stem cells. J Natl Cancer Inst 98, 1011-1014 (2006). 123 Sakakura, T., Nishizuka, Y. & Dawe, C. J. Capacity of mammary fat pads of adult C3H/HeMs mice to interact morphogenetically with fetal mammary epithelium. J Natl Cancer Inst 63, 733-736 (1979). 124 Klinowska, T. C. & Streuli, C. H. Analyzing cell-ECM interactions in adult mammary gland by transplantation of embryonic mammary tissue from knockout mice. Methods Mol Biol 139, 345-358 (2000). 125 Sakakura, T., Nishizuka, Y. & Dawe, C. J. Mesenchyme-dependent morphogenesis and epithelium-specific cytodifferentiation in mouse mammary gland. Science 194, 1439- 1441 (1976). 126 Medvinsky, A. & Dzierzak, E. Definitive hematopoiesis is autonomously initiated by the AGM region. Cell 86, 897-906 (1996). 127 Muller, A. M., Medvinsky, A., Strouboulis, J., Grosveld, F. & Dzierzak, E. Development of hematopoietic stem cell activity in the mouse embryo. Immunity 1, 291-301 (1994). 128 van Genderen, C. et al. Development of several organs that require inductive epithelial- mesenchymal interactions is impaired in LEF-1-deficient mice. Genes Dev 8, 2691-2703 (1994). 129 Boras-Granic, K., Chang, H., Grosschedl, R. & Hamel, P. A. Lef1 is required for the transition of Wnt signaling from mesenchymal to epithelial cells in the mouse embryonic mammary gland. Dev Biol 295, 219-231 (2006). 130 Lindvall, C. et al. The Wnt co-receptor Lrp6 is required for normal mouse mammary gland development. PLoS One 4, e5813 (2009). 131 Gu, B. et al. Pygo2 expands mammary progenitor cells by facilitating histone H3 K4 methylation. J Cell Biol 185, 811-826 (2009). 132 Eblaghie, M. C. et al. Interactions between FGF and Wnt signals and Tbx3 gene expression in mammary gland initiation in mouse embryos. J Anat 205, 1-13 (2004). 133 Chu, E. Y. et al. Canonical WNT signaling promotes mammary placode development and is essential for initiation of mammary gland morphogenesis. Development 131, 4819-4829 (2004). 134 Hens, J. R. et al. BMP4 and PTHrP interact to stimulate ductal outgrowth during embryonic mammary development and to inhibit hair follicle induction. Development 134, 1221-1230 (2007). 135 Foley, J. et al. Parathyroid hormone-related protein maintains mammary epithelial fate and triggers nipple skin differentiation during embryonic breast development. Development 128, 513-525 (2001). 104 136 Vaillant, F. et al. The mammary progenitor marker CD61/beta3 integrin identifies cancer stem cells in mouse models of mammary tumorigenesis. Cancer Res 68, 7711-7717 (2008). 137 Mailleux, A. A. et al. Role of FGF10/FGFR2b signaling during mammary gland development in the mouse embryo. Development 129, 53-60 (2002). 138 Howard, B., Panchal, H., McCarthy, A. & Ashworth, A. Identification of the scaramanga gene implicates Neuregulin3 in mammary gland specification. Genes Dev 19, 2078-2090 (2005). 139 Panchal, H., Wansbury, O., Parry, S., Ashworth, A. & Howard, B. Neuregulin3 alters cell fate in the epidermis and mammary gland. BMC Dev Biol 7, 105 (2007). 140 Bouras, T. et al. Notch signaling regulates mammary stem cell function and luminal cell- fate commitment. Cell Stem Cell 3, 429-441 (2008). 141 Carr, J. R. et al. FoxM1 regulates mammary luminal cell fate. Cell Rep 1, 715-729 (2012). 142 Kouros-Mehr, H., Slorach, E. M., Sternlicht, M. D. & Werb, Z. GATA-3 maintains the differentiation of the luminal cell fate in the mammary gland. Cell 127, 1041-1055 (2006). 143 Pei, X. H. et al. CDK inhibitor p18(INK4c) is a downstream target of GATA3 and restrains mammary luminal progenitor cell proliferation and tumorigenesis. Cancer Cell 15, 389-401 (2009). 144 Yuan, Y., Shen, H., Franklin, D. S., Scadden, D. T. & Cheng, T. In vivo self-renewing divisions of haematopoietic stem cells are increased in the absence of the early G1-phase inhibitor, p18INK4C. Nat Cell Biol 6, 436-442 (2004). 145 Mochizuki-Kashio, M. et al. Dependency on the polycomb gene Ezh2 distinguishes fetal from adult hematopoietic stem cells. Blood 118, 6553-6561 (2011). 146 Lukacs, R. U., Memarzadeh, S., Wu, H. & Witte, O. N. Bmi-1 is a crucial regulator of prostate stem cell self-renewal and malignant transformation. Cell Stem Cell 7, 682-693 (2010). 147 Pietersen, A. M. et al. Bmi1 regulates stem cells and proliferation and differentiation of committed cells in mammary epithelium. Curr Biol 18, 1094-1099 (2008). 148 Pal, B. et al. Global Changes in the Mammary Epigenome Are Induced by Hormonal Cues and Coordinated by Ezh2. Cell Rep (2013). 149 Nguyen, L. V., Vanner, R., Dirks, P. & Eaves, C. J. Cancer stem cells: an evolving concept. Nat Rev Cancer 12, 133-143 (2012). 150 Lessard, J. & Sauvageau, G. Bmi-1 determines the proliferative capacity of normal and leukaemic stem cells. Nature 423, 255-260 (2003). 151 Bruggeman, S. W. et al. Bmi1 controls tumor development in an Ink4a/Arf-independent manner in a mouse model for glioma. Cancer Cell 12, 328-341 (2007). 152 Liu, S. et al. Hedgehog signaling and Bmi-1 regulate self-renewal of normal and malignant human mammary stem cells. Cancer Res 66, 6063-6071 (2006). 153 Doherty, L. F., Bromer, J. G., Zhou, Y., Aldad, T. S. & Taylor, H. S. In utero exposure to diethylstilbestrol (DES) or bisphenol-A (BPA) increases EZH2 expression in the mammary gland: an epigenetic mechanism linking endocrine disruptors to breast cancer. Horm Cancer 1, 146-155 (2010). 154 Teuliere, J. et al. Targeted activation of beta-catenin signaling in basal mammary epithelial cells affects mammary development and leads to hyperplasia. Development 132, 267-277 (2005). 105 155 Pond, A. C. et al. Fibroblast Growth Factor Receptor Signaling is Essential for Normal Mammary Gland Development and Stem Cell Function. Stem Cells (2012). 156 Penault-Llorca, F. et al. Expression of FGF and FGF receptor genes in human breast cancer. Int J Cancer 61, 170-176 (1995). 157 Edakuni, G., Sasatomi, E., Satoh, T., Tokunaga, O. & Miyazaki, K. Expression of the hepatocyte growth factor/c-Met pathway is increased at the cancer front in breast carcinoma. Pathol Int 51, 172-178 (2001). 158 Korkaya, H. & Wicha, M. S. HER-2, notch, and breast cancer stem cells: targeting an axis of evil. Clin Cancer Res 15, 1845-1847 (2009). 159 Lindeman, G. J. & Visvader, J. E. Insights into the cell of origin in breast cancer and breast cancer stem cells. Asia Pac J Clin Oncol 6, 89-97 (2010). 160 Sorlie, T. et al. Repeated observation of breast tumor subtypes in independent gene expression data sets. Proc Natl Acad Sci U S A 100, 8418-8423 (2003). 161 Perou, C. M. et al. Molecular portraits of human breast tumours. Nature 406, 747-752 (2000). 162 Curtis, C. et al. The genomic and transcriptomic architecture of 2,000 breast tumours reveals novel subgroups. Nature 486, 346-352 (2012). 163 Proia, T. A. et al. Genetic predisposition directs breast cancer phenotype by dictating progenitor cell fate. Cell Stem Cell 8, 149-163 (2011). 164 Molyneux, G. et al. BRCA1 basal-like breast cancers originate from luminal epithelial progenitors and not from basal stem cells. Cell Stem Cell 7, 403-417 (2010). 165 Liu, S. et al. BRCA1 regulates human mammary stem/progenitor cell fate. Proc Natl Acad Sci U S A 105, 1680-1685 (2008). 166 Liu, X. et al. Somatic loss of BRCA1 and p53 in mice induces mammary tumors with features of human BRCA1-mutated basal-like breast cancer. Proc Natl Acad Sci U S A 104, 12111-12116 (2007). 167 Serber, D. W. et al. The BRG1 chromatin remodeler protects against ovarian cysts, uterine tumors, and mammary tumors in a lineage-specific manner. PLoS One 7, e31346 (2012). 168 Molofsky, A. V., Pardal, R. & Morrison, S. J. Diverse mechanisms regulate stem cell self-renewal. Curr Opin Cell Biol 16, 700-707 (2004). 169 Levi, B. P. & Morrison, S. J. Stem cells use distinct self-renewal programs at different ages. Cold Spring Harb Symp Quant Biol 73, 539-553 (2008). 170 Moraes, R. C. et al. Constitutive activation of smoothened (SMO) in mammary glands of transgenic mice leads to increased proliferation, altered differentiation and ductal dysplasia. Development 134, 1231-1242 (2007). 171 Klos, K. S., Kim, S. & Alexander, C. M. Genotoxic Exposure during Juvenile Growth of Mammary Gland Depletes Stem Cell Activity and Inhibits Wnt Signaling. PLoS One 7, e49902 (2012). 172 Visvader, J. E. Keeping abreast of the mammary epithelial hierarchy and breast tumorigenesis. Genes Dev 23, 2563-2577 (2009). 173 Lindeman, G. J. & Visvader, J. E. Cell fate takes a slug in BRCA1-associated breast cancer. Breast Cancer Research 13 (2011). 174 Hu, Y. & Smyth, G. K. ELDA: extreme limiting dilution analysis for comparing depleted and enriched populations in stem cell and other assays. J Immunol Methods 347, 70-78 (2009). 175 Guo, W. et al. Slug and Sox9 cooperatively determine the mammary stem cell state. Cell 148, 1015-1028 (2012). 106 176 Levy, V., Lindon, C., Zheng, Y., Harfe, B. D. & Morgan, B. A. Epidermal stem cells arise from the hair follicle after wounding. FASEB J 21, 1358-1366 (2007). 177 Tian, H. et al. A reserve stem cell population in small intestine renders Lgr5-positive cells dispensable. Nature 478, 255-259 (2011). 178 van Es, J. H. et al. Dll1(+) secretory progenitor cells revert to stem cells upon crypt damage. Nat Cell Biol 14, 1099-1104 (2012). 179 Pawliuk, R., Eaves, C. & Humphries, R. K. Evidence of both ontogeny and transplant dose-regulated expansion of hematopoietic stem cells in vivo. Blood 88, 2852-2858 (1996). 180 Bowie, M. B. et al. Identification of a new intrinsically timed developmental checkpoint that reprograms key hematopoietic stem cell properties. Proc Natl Acad Sci U S A 104, 5878-5882 (2007). 181 Kruger, G. M. et al. Neural crest stem cells persist in the adult gut but undergo changes in self-renewal, neuronal subtype potential, and factor responsiveness. Neuron 35, 657- 669 (2002). 182 Nishino, J., Kim, I., Chada, K. & Morrison, S. J. Hmga2 promotes neural stem cell self- renewal in young but not old mice by reducing p16Ink4a and p19Arf Expression. Cell 135, 227-239 (2008). 183 Kim, I., Saunders, T. L. & Morrison, S. J. Sox17 dependence distinguishes the transcriptional regulation of fetal from adult hematopoietic stem cells. Cell 130, 470-483 (2007). 184 Yuan, J., Nguyen, C. K., Liu, X., Kanellopoulou, C. & Muljo, S. A. Lin28b reprograms adult bone marrow hematopoietic progenitors to mediate fetal-like lymphopoiesis. Science 335, 1195-1200 (2012). 185 Yang, L. et al. Identification of Lin(-)Sca1(+)kit(+)CD34(+)Flt3- short-term hematopoietic stem cells capable of rapidly reconstituting and rescuing myeloablated transplant recipients. Blood 105, 2717-2723 (2005). 186 Jones, R. J. et al. Characterization of mouse lymphohematopoietic stem cells lacking spleen colony-forming activity. Blood 88, 487-491 (1996). 187 Spangrude, G. J., Heimfeld, S. & Weissman, I. L. Purification and characterization of mouse hematopoietic stem cells. Science 241, 58-62 (1988). 188 Morris, R. J. et al. Capturing and profiling adult hair follicle stem cells. Nat Biotechnol 22, 411-417 (2004). 189 Lawson, D. A., Xin, L., Lukacs, R. U., Cheng, D. & Witte, O. N. Isolation and functional characterization of murine prostate stem cells. Proc Natl Acad Sci U S A 104, 181-186 (2007). 190 Benz, C. et al. Hematopoietic stem cell subtypes expand differentially during development and display distinct lymphopoietic programs. Cell Stem Cell 10, 273-283 (2012). 191 Jordan, C. T. & Lemischka, I. R. Clonal and systemic analysis of long-term hematopoiesis in the mouse. Genes Dev 4, 220-232 (1990). 192 He, S., Kim, I., Lim, M. S. & Morrison, S. J. Sox17 expression confers self-renewal potential and fetal stem cell characteristics upon adult hematopoietic progenitors. Genes Dev 25, 1613-1627 (2011). 193 Wright, D. E. et al. Cyclophosphamide/granulocyte colony-stimulating factor causes selective mobilization of bone marrow hematopoietic stem cells into the blood after M phase of the cell cycle. Blood 97, 2278-2285 (2001). 107 194 Martignani, E., Eirew, P., Eaves, C. & Baratta, M. Functional identification of bovine mammary epithelial stem/progenitor cells. Vet Res Commun 33 Suppl 1, 101-103 (2009). 195 Sutherland, H. J., Eaves, C. J., Eaves, A. C., Dragowska, W. & Lansdorp, P. M. Characterization and partial purification of human marrow cells capable of initiating long-term hematopoiesis in vitro. Blood 74, 1563-1570 (1989). 196 Miller, C. L. & Eaves, C. J. Long-term culture-initiating cell assays for human and murine cells. Methods Mol Med 63, 123-141 (2002). 197 Copley, M. R., Beer, P. A. & Eaves, C. J. Hematopoietic stem cell heterogeneity takes center stage. Cell Stem Cell 10, 690-697 (2012). 198 Dykstra, B. et al. High-resolution video monitoring of hematopoietic stem cells cultured in single-cell arrays identifies new features of self-renewal. Proc Natl Acad Sci U S A 103, 8185-8190 (2006). 199 Audet, J., Miller, C. L., Eaves, C. J. & Piret, J. M. Common and distinct features of cytokine effects on hematopoietic stem and progenitor cells revealed by dose-response surface analysis. Biotechnol Bioeng 80, 393-404 (2002). 200 Suda, T., Suda, J. & Ogawa, M. Proliferative kinetics and differentiation of murine blast cell colonies in culture: evidence for variable G0 periods and constant doubling rates of early pluripotent hemopoietic progenitors. J Cell Physiol 117, 308-318 (1983). 201 Ito, M. et al. Wnt-dependent de novo hair follicle regeneration in adult mouse skin after wounding. Nature 447, 316-320 (2007). 202 van Es, J. H. et al. Dll1+ secretory progenitor cells revert to stem cells upon crypt damage. Nat Cell Biol 14, 1099-1104 (2012). 203 Barroca, V. et al. Mouse differentiating spermatogonia can generate germinal stem cells in vivo. Nat Cell Biol 11, 190-196 (2009). 204 Nakagawa, T., Sharma, M., Nabeshima, Y., Braun, R. E. & Yoshida, S. Functional hierarchy and reversibility within the murine spermatogenic stem cell compartment. Science 328, 62-67 (2010). 205 Takahashi, K. & Yamanaka, S. Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell 126, 663-676 (2006). 206 Sancho-Martinez, I., Baek, S. H. & Izpisua Belmonte, J. C. Lineage conversion methodologies meet the reprogramming toolbox. Nat Cell Biol 14, 892-899 (2012). 207 Graf, T. Historical origins of transdifferentiation and reprogramming. Cell Stem Cell 9, 504-516 (2011). 208 Vermeulen, L. et al. Wnt activity defines colon cancer stem cells and is regulated by the microenvironment. Nat Cell Biol 12, 468-476 (2010). 209 Mani, S. A. et al. The epithelial-mesenchymal transition generates cells with properties of stem cells. Cell 133, 704-715 (2008). 210 Rebel, V. I., Miller, C. L., Eaves, C. J. & Lansdorp, P. M. The repopulation potential of fetal liver hematopoietic stem cells in mice exceeds that of their liver adult bone marrow counterparts. Blood 87, 3500-3507 (1996). 211 Morrison, S. J., Hemmati, H. D., Wandycz, A. M. & Weissman, I. L. The purification and characterization of fetal liver hematopoietic stem cells. Proc Natl Acad Sci U S A 92, 10302-10306 (1995). 212 Maslov, A. Y., Barone, T. A., Plunkett, R. J. & Pruitt, S. C. Neural stem cell detection, characterization, and age-related changes in the subventricular zone of mice. J Neurosci 24, 1726-1733 (2004). 108 213 Molofsky, A. V. et al. Increasing p16INK4a expression decreases forebrain progenitors and neurogenesis during ageing. Nature 443, 448-452 (2006). 214 Bowie, M. B., Kent, D. G., Copley, M. R. & Eaves, C. J. Steel factor responsiveness regulates the high self-renewal phenotype of fetal hematopoietic stem cells. Blood 109, 5043-5048 (2007). 215 Zeps, N., Dawkins, H. J., Papadimitriou, J. M., Redmond, S. L. & Walters, M. I. Detection of a population of long-lived cells in mammary epithelium of the mouse. Cell Tissue Res 286, 525-536 (1996). 216 Welm, B. E. et al. Sca-1(pos) cells in the mouse mammary gland represent an enriched progenitor cell population. Dev Biol 245, 42-56 (2002). 217 Booth, B. W., Boulanger, C. A. & Smith, G. H. Selective segregation of DNA strands persists in long-label-retaining mammary cells during pregnancy. Breast Cancer Res 10, R90 (2008). 218 Smith, G. H. Label-retaining epithelial cells in mouse mammary gland divide asymmetrically and retain their template DNA strands. Development 132, 681-687 (2005). 219 Tumbar, T. et al. Defining the epithelial stem cell niche in skin. Science 303, 359-363 (2004). 220 Lecault, V. et al. High-throughput analysis of single hematopoietic stem cell proliferation in microfluidic cell culture arrays. Nat Methods 8, 581-586 (2011). 221 Domen, J. & Weissman, I. L. Hematopoietic stem cells need two signals to prevent apoptosis; BCL-2 can provide one of these, Kitl/c-Kit signaling the other. J Exp Med 192, 1707-1718 (2000). 222 Domen, J., Cheshier, S. H. & Weissman, I. L. The role of apoptosis in the regulation of hematopoietic stem cells: Overexpression of Bcl-2 increases both their number and repopulation potential. J Exp Med 191, 253-264 (2000). 223 Strange, R., Metcalfe, T., Thackray, L. & Dang, M. Apoptosis in normal and neoplastic mammary gland development. Microsc Res Tech 52, 171-181 (2001). 224 Molofsky, A. V., He, S., Bydon, M., Morrison, S. J. & Pardal, R. Bmi-1 promotes neural stem cell self-renewal and neural development but not mouse growth and survival by repressing the p16Ink4a and p19Arf senescence pathways. Genes Dev 19, 1432-1437 (2005). 225 Park, I. K. et al. Bmi-1 is required for maintenance of adult self-renewing haematopoietic stem cells. Nature 423, 302-305 (2003). 226 Barker, N. et al. Identification of stem cells in small intestine and colon by marker gene Lgr5. Nature 449, 1003-1007 (2007). 227 Huelsken, J., Vogel, R., Erdmann, B., Cotsarelis, G. & Birchmeier, W. beta-Catenin controls hair follicle morphogenesis and stem cell differentiation in the skin. Cell 105, 533-545 (2001). 228 Dykstra, B. et al. Long-term propagation of distinct hematopoietic differentiation programs in vivo. Cell Stem Cell 1, 218-229 (2007). 229 Gerrits, A. et al. Cellular barcoding tool for clonal analysis in the hematopoietic system. Blood 115, 2610-2618 (2010). 230 Lu, R., Neff, N. F., Quake, S. R. & Weissman, I. L. Tracking single hematopoietic stem cells in vivo using high-throughput sequencing in conjunction with viral genetic barcoding. Nat Biotechnol 29, 928-933 (2011). 231 Metcalf, D. Lineage commitment and maturation in hematopoietic cells: the case for extrinsic regulation. Blood 92, 345-347; discussion 352 (1998). 109 232 Boulanger, C. A., Wagner, K. U. & Smith, G. H. Parity-induced mouse mammary epithelial cells are pluripotent, self-renewing and sensitive to TGF-beta1 expression. Oncogene 24, 552-560 (2005). 233 Lacroix, M. & Leclercq, G. Relevance of breast cancer cell lines as models for breast tumours: an update. Breast Cancer Res Treat 83, 249-289 (2004). 234 Neve, R. M. et al. A collection of breast cancer cell lines for the study of functionally distinct cancer subtypes. Cancer Cell 10, 515-527 (2006). "@en ; edm:hasType "Thesis/Dissertation"@en ; vivo:dateIssued "2013-11"@en ; edm:isShownAt "10.14288/1.0073898"@en ; dcterms:language "eng"@en ; ns0:degreeDiscipline "Genetics"@en ; edm:provider "Vancouver : University of British Columbia Library"@en ; dcterms:publisher "University of British Columbia"@en ; dcterms:rights "Attribution-NonCommercial-NoDerivatives 4.0 International"@en ; ns0:rightsURI "http://creativecommons.org/licenses/by-nc-nd/4.0/"@en ; ns0:scholarLevel "Graduate"@en ; dcterms:title "Developmental changes in the properties of mouse mammary stem cells"@en ; dcterms:type "Text"@en ; ns0:identifierURI "http://hdl.handle.net/2429/44592"@en .