A Sorting Strategy For Bovine Mammary Progenitors

The mammary gland is a bilayered system organized into a series of branching ducts that terminate in secretory alveoli. Their epithelium is composed of an inner layer of cytokeratin(CK)18+ luminal cells, which secrete milk, and an outer layer of contractile CK14+ myoepithelial cells, which contribute to milk ejection during lactation. The existence of mammary renewable stem cells (MaSCs) has been deduced from the profound regenerative ability of the mammary epithelium. It undergoes different morphogenetic changes during puberty and during the subsequent pregnancies. The knowledge of MaSCs properties and functions should represent an interesting opportunity in the field of animal production as the appropriate regulation of these cells should benefit milk yeld. The aim of our study is to characterize the MaSCs population of bovine species and set up a sorting strategy to purify different primitive populations. The Colony-Forming Cell (CFC) assay performed on bovine mammary samples obtained from virgin, two years old cows revealed the presence in the mammary gland of both myoepithelial and luminal-restricted progenitor cells. In order to select and isolate the different types of progenitors we have performed a flow cytometric analysis of cells dissociated from bovine mammary tissue. Cells were then stained with antibodies directed to CD49f, whose expression at high levels is associated with progenitor and stem cell activity, and p-cadherin, that is normally expressed in the myoepithelial/basal layer of the mammary epithelium. After depletion of hematopoietic cells (using an anti-CD45 antibody) we identified four different subpopulations. They were characterized by different levels of CD49f/P-cad expression: CD49f/P-cad- (15,6%), CD49f++/P-cad- (25,4%), CD49f+/P-cad+ (27,1%), CD49f+/P-cad++ (10,2%). After sorting, the CFC assay performed with the cells of the different subpopulations showed that CD49f++/P-cad- population was able to generate colonies with an higher frequency compared to the others, revealing an enrichment in stem cells. In particular, 35,7% of colonies showed a myoepithelial phenotype and 64,3% a luminal one. In order to analyze whether these cells also had an in vivo tissue regenerative activity, the four subpopulation were implanted in collagen gels under the kidney capsule of NOD/SCID immunodeficient mice in numbers proportional to their frequency in the CD45- population. Immunohistochemical analysis of gels recovered 4 weeks later revealed that only CD49f++/P-cad- subpopulation was able to regenerate in vivo mammary alveolar structures characterized by a lumen surrounded by two layers of cells: an inner K18+ layer and an outer K14+ layer. The CFC assay performed with single cell suspension obtained from recovered gels showed a higher colony-forming ability of CD49f++/P-cad- population compared to the others. Therefore the structures produced in these gels contained clonogenic progenitors, that were able to form the same types of colonies observed from freshly isolated cells (36,2% of myoepithelial colonies and 63,8% of luminal ones). By using a novel sorting strategy we show that the mammary myoepithelial cell layer can be divided in different subpopulations. Each one has a different progenitor content and only in one fraction (CD49f++/P-Cad-) we were able to detect a subset of adult stem cells able to regenerate a mammary epithelium. This sorting strategy might prove useful for understanding the stem cells population dynamics during the different phases that the mammary gland undergoes. Besides, given the similar organization of the mammary tissue in the bovine and the human species, the cow might represent a much better animal model than the mouse.