Abstract
BACKGROUND: Stromal-epithelial interactions mediate breast development, and the initiation and progression of breast cancer. In the present study, we developed 3-dimensional (3D) in vitro models to study breast cancer tissue organization and the role of the microenvironment in phenotypic determination. METHODS: The human breast cancer MCF7 cells were grown alone or co-cultured with primary human breast fibroblasts. Cells were embedded in matrices containing either type I collagen or a combination of reconstituted basement membrane proteins and type I collagen. The cultures were carried out for up to 6 weeks. For every time point (1-6 weeks), the gels were fixed and processed for histology, and whole-mounted for confocal microscopy evaluation. The epithelial structures were characterized utilizing immunohistochemical techniques; their area and proliferation index were measured using computerized morphometric analysis. Statistical differences between groups were analyzed by ANOVA, Dunnett's T3 post-hoc test and chi-square. RESULTS: Most of the MCF7 cells grown alone within a collagen matrix died during the first two weeks; those that survived organized into large, round and solid clusters. The presence of fibroblasts in collagen gels reduced MCF7 cell death, induced cell polarity, and the formation of round and elongated epithelial structures containing a lumen. The addition of reconstituted basement membrane to collagen gels by itself had also survival and organizational effects on the MCF7 cells. Regardless of the presence of fibroblasts, the MCF7 cells both polarized and formed a lumen. The addition of fibroblasts to the gel containing reconstituted basement membrane and collagen induced the formation of elongated structures. CONCLUSIONS: Our results indicate that a matrix containing both type I collagen and reconstituted basement membrane, and the presence of normal breast fibroblasts constitute the minimal permissive microenvironment to induce near-complete tumor phenotype reversion. These human breast 3D tissue morphogenesis models promise to become reliable tools for studying tissue interactions, therapeutic screening and drug target validation.