Journal of Cell Science partnership with Dryad

Journal of Cell Science makes data accessibility easy with Dryad


We previously established a rapid three-dimensional assay for discrimination of normal and malignant human breast epithelial cells using a laminin-rich reconstituted basement membrane. In this assay, normal epithelial cells differentiate into well-organized acinar structures whereas tumor cells fail to recapitulate this process and produce large, disordered colonies. The data suggest that breast acinar morphogenesis and differentiation is regulated by cell-extra-cellular matrix (ECM) interactions and that these interactions are altered in malignancy. Here, we investigated the role of ECM receptors (integrins) in these processes and report on the expression and function of potential laminin receptors in normal and tumorigenic breast epithelial cells. Immunocytochemical analysis showed that normal and carcinoma cells in a three-dimensional substratum express profiles of integrins similar to normal and malignant breast tissues in situ. Normal cells express alpha 1, alpha 2, alpha 3, alpha 6, beta 1 and beta 4 integrin subunits, whereas breast carcinoma cells show variable losses, disordered expression, or downregulation of these subunits. Function-blocking experiments using inhibitory anti-integrin subunit antibodies showed a > 5-fold inhibition of the formation of acinar structures by normal cells in the presence of either anti-beta 1 or anti-alpha 3 antibodies, whereas anti-alpha 2 or -alpha 6 had little or no effect. In experiments where collagen type I gels were used instead of basement membrane, acinar morphogenesis was blocked by anti-beta 1 and -alpha 2 antibodies but not by anti-alpha 3. These data suggest a specificity of integrin utilization dependent on the ECM ligands encountered by the cell. The interruption of normal acinar morphogenesis by anti-integrin antibodies was associated with an inhibition of cell growth and induction of apoptosis. Function-blocking antibodies had no inhibitory effect on the rate of tumor cell growth, survival or capacity to form colonies. Thus under our culture conditions breast acinar formation is at least a two-step process involving beta 1-integrin-dependent cellular growth followed by polarization of the cells into organized structures. The regulation of this pathway appears to be impaired or lost in the tumor cells, suggesting that tumor colony formation occurs by independent mechanisms and that loss of proper integrin-mediated cell-ECM interaction may be critical to breast tumor formation.