Growth factor control of myoepithelial-cell differentiation in cultures of human mammary gland

Abstract

Abstract The relationship between growth and cytodifferentiation was studied in cultured human mammary myoepithelial cells under serum-free culture conditions. Myoepithelial-cell differentiation was monitored by quantifying cells showing immunoreactivity to the muscle isoform of actin; to the membrane glycoprotein common acute lymphoblastic leukemia antigen (CALLA); and to type IV collagen. Growth was quantified either by measuring the actual increase in cell number, or in a more-sensitive assay using immunoreactivity to the cell-proliferation-associated nuclear antigen Ki-67 as a measurement of the number of cells leaving the G 0-phase of the cell cycle. The results showed that: (a) Primary cultures of myoepithelial cells on DMEF12 supplemented with cholera toxin (CT) alone resulted in the formation of quiescent cell islets (in the G 0-phase of the cell cycle) showing phenotypic traits preserved from the in vivo situation (actin- and CALLA-positive cells with little or no type-IV-collagen immunoreactivity). (b) After addition of epidermal growth factor (EGF), with an ED 50 of 1–10 ng/ml, in the presence of CT, the cells entered the G 1-phase of the cell cycle, without further increase in cell number. At the same ED 50 of EGF, the frequency of CALLA-positive cells decreased, while the number of cells immunoreactive for type IV collagen increased with a maximal effect of EGF seen after 7–11 days. During the same period, the cells remained fully differentiated with respect to actin immunoreactivity. (c) Further addition of insulin (I) to the medium in the presence of EGF and CT resulted in the cells entering an exponential growth phase associated with simultaneous decrease in actin immunoreactivity with a maximal effect of I after 11 days of exposure. The dose-response curve to I was virtually identical for stimulating cell proliferation and for reducing the frequency of actin-immunoreactive cells (ED 50 in the range of 30 ng/ml), suggesting that the two processes were controlled by the same initial I-receptor interaction. (d) Some reduction in the number of actin-positive cells was exerted by I-EGF-CT independently of the mitogenic response, but this reduction was further augmented if the cells were allowed to proliferate. (e) Time-course studies of quiescent (G 0-phase) cells stimulated to exponential growth revealed that entrance of cells into the G 1-phase of the cell cycle preceded the loss of muscle actin filaments. (f) Exponentially growing actin-negative epithelial cells did not resume a myoepithelial phenotype in density-arrested postconfluent cultures. Rather, they entered a state of terminal squamous differentiation, showing immunoreactivity to cytokeratin 10. However, if the exponentially growing cells were growth-arrested in mitogen-deprived medium such as basal medium, a time-course study revealed a rapid (approx. 3 days) but limited squamous differentiation of some cells, while others showed myoepithelial differentiation after a latency period of more than 8 days. In conclusion, we have developed a model system that may answer several general questions concerning growth-factor control of cytodifferentiation as well as more-specific questions related to normal and abnormal myoepithelial differentiation with particular reference to the human mammary gland and the derived tumors.