HER2 induced EMT and tumorigenicity in breast epithelial progenitor cells is inhibited by coexpression of EGFR.

The members of the epidermal growth factor receptor (EGFR) kinase family are important players in breast morphogenesis and cancer. EGFR2/HER2 and EGFR expression have a prognostic value in certain subtypes of breast cancer such as HER2-amplified, basal-like and luminal type B. Many clinically approved small molecular inhibitors and monoclonal antibodies have been designed to target HER2, EGFR or both. There is, however, still limited knowledge on how the two receptors are expressed in normal breast epithelium, what effects they have on cellular differentiation and how they participate in neoplastic transformation. D492 is a breast epithelial cell line with stem cell properties that can undergo epithelial to mesenchyme transition (EMT), generate luminal- and myoepithelial cells and form complex branching structures in three-dimensional (3D) culture. Here, we show that overexpression of HER2 in D492 (D492(HER2)) resulted in EMT, loss of contact growth inhibition and increased oncogenic potential in vivo. HER2 overexpression, furthermore, inhibited endogenous EGFR expression. Re-introducing EGFR in D492(HER2) (D492(HER2/EGFR)) partially reversed the mesenchymal state of the cells, as an epithelial phenotype reappeared both in 3D cultures and in vivo. The D492(HER2/EGFR) xenografts grow slower than the D492(HER2) tumors, while overexpression of EGFR alone (D492(EGFR)) was not oncogenic in vivo. Consistent with the EGFR-mediated epithelial phenotype, overexpression of EGFR drove the cells toward a myoepithelial phenotype in 3D culture. The effect of two clinically approved anti-HER2 and EGFR therapies, trastuzumab and cetuximab, was tested alone and in combination on D492(HER2) xenografts. While trastuzumab had a growth inhibitory effect compared with untreated control, the effect of cetuximab was limited. When administered in combination, the growth inhibitory effect of trastuzumab was less pronounced. Collectively, our data indicate that in HER2-overexpressing D492 cells, EGFR can behave as a tumor suppressor, by pushing the cells towards epithelial differentiation.

Immortalization protocols used in cell culture models of human breast morphogenesis.

Defining the key players in normal breast differentiation is instrumental to understanding how morphogenesis becomes defective during breast cancer progression. During the past 2 decades much effort has been devoted to the development of technologies for purification and expansion of primary human breast cells in culture and optimizing a relevant microenvironment, which may help to define the niche that regulates breast differentiation and morphogenesis. In contrast to the general property of cancer, normal human cells have a finite lifespan. After a defined number of population doublings, normal cells enter an irreversible proliferation-arrested state referred to as replicative senescence. To overcome this obstacle for continuous long-term studies, replicative senescence can be bypassed by treatment of cells with chemical agents such as benzopyrene, by radiation or by transfection with viral oncogenes or the gene for human telomerase (human telomerase reverse transcriptase, hTERT). A drawback of some of these protocols is a concurrent introduction of chromosomal changes, which sometimes leads to a transformed phenotype and selection of a subpopulation, which may not be representative of the tissue of origin. In recent years, we have sought to establish immortalized primary breast cells, which retain crucial characteristics of their original in situ tissue pattern. This review discusses various approaches to immortalization of breast-derived epithelial and stromal cells and the application of such cell lines for studies on human breast morphogenesis.

The plasticity of human breast carcinoma cells is more than epithelial to mesenchymal conversion.

The human breast comprises three lineages: the luminal epithelial lineage, the myoepithelial lineage, and the mesenchymal lineage. It has been widely accepted that human breast neoplasia pertains only to the luminal epithelial lineage. In recent years, however, evidence has accumulated that neoplastic breast epithelial cells may be substantially more plastic in their differentiation repertoire than previously anticipated. Thus, along with an increasing availability of markers for the myoepithelial lineage, at least a partial differentiation towards this lineage is being revealed frequently. It has also become clear that conversions towards the mesenchymal lineage actually occur, referred to as epithelial to mesenchymal transitions. Indeed, some of the so-called myofibroblasts surrounding the tumor may have an epithelial origin rather than a mesenchymal origin. Because myoepithelial cells, epithelial to mesenchymal transition-derived cells, genuine stromal cells and myofibroblasts share common markers, we now need to define a more ambitious set of markers to distinguish these cell types in the microenvironment of the tumors. This is necessary because the different microenvironments may confer different clinical outcomes. The aim of this commentary is to describe some of the inherent complexities in defining cellular phenotypes in the microenvironment of breast cancer and to expand wherever possible on the implications for tumor suppression and progression.

Human mammary luminal epithelial cells contain progenitors to myoepithelial cells.

The origin of the epithelial and myoepithelial cells in the human breast has not been delineated. In this study we have addressed whether luminal epithelial cells and myoepithelial cells are vertically connected, i.e., whether one is the precursor for the other. We used a primary culture assay allowing preservation of basic phenotypic traits of luminal epithelial and myoepithelial cells in culture. The two cell types were then separated immunomagnetically using antibodies directed against lineage-specific cell surface antigens into at best 100% purity. The cellular identity was ascertained by cytochemistry, immunoblotting, and 2-D gel electrophoresis. Luminal epithelial cells were identified by strong expression of cytokeratins 18 and 19 while myoepithelial cells were recognized by expression of vimentin and alpha-smooth muscle actin. We used a previously devised culture medium (CDM4) that allows vigorous expansion of proliferative myoepithelial cells and also devised a medium (CDM6) that allowed sufficient expansion of differentiated luminal epithelial cells based on addition of hepatocyte growth factor/scatter factor. The two different culture media supported each lineage for at least five passages without signs of interconversion. We used parallel cultures where we switched culture media, thus testing the ability of each lineage to convert to the other. Whereas the myoepithelial lineage showed no signs of interconversion, a subset of luminal epithelial cells, gradually, but distinctly, converted to myoepithelial cells. We propose that in the mature human breast, it is the luminal epithelial cell compartment that gives rise to myoepithelial cells rather than the other way around.