Cytometric and biochemical characterization of human breast cancer cells reveals heterogeneous myoepithelial phenotypes.

To determine whether cell cultures maintain the cellular heterogeneity of primary tissues and may therefore be used for in vitro modeling of breast cancer subtypes, we evaluated the expression of a cell surface marker panel in breast cancer cell cultures derived from various subtypes of human breast carcinoma. We used a four-color flow cytometry strategy to immunophenotype seven human breast cancer cell cultures and four reference breast cancer cell lines. We analyzed 28 surface markers selected based on their potential to distinguish epithelial or mesenchymal lineage, to identify stem cell populations, and to mediate cell adhesion and migration. We determined their ability to form mammospheres and analyzed luminal cytokeratins CK18, CK19, and myoepithelial/basal CK5, SMA (alpha-smooth muscle actin), and vimentin expression by western blot. All cell surface markers showed a unimodal profile. Ten/28 markers were homogenously expressed. Four (CD66b, CD66c, CD165, CD324) displayed negative/low expression. Six (CD29, CD55, CD59, CD81, CD151, CD166) displayed homogenous high expression. Eighteen (CD9, CD10, CD24, CD26, CD44, CD47, CD49b, CD49f, CD54, CD61, CD90, CD105, CD133, CD164, CD184, CD200, CD227, CD326) were heterogeneously expressed. Spearman's rank test demonstrated a significant correlation (p< 0.001) between mesenchymal phenotype and breast cancer cell cultures. Breast cancer cell cultures, all CD44+, displayed concomitant high expression of only three antigens (CD10, CD54, CD90), and low expression of CD326; cell cultures formed mammospheres and expressed CK5, SMA and vimentin, and were weakly CK19-positive. We demonstrate that breast cancer cell cultures preserve inter-tumor heterogeneity and express stem/progenitor markers that can be identified, quantified and categorized by flow cytometry. Therefore, cell cultures can be used for in vitro modeling of breast cancer subtypes; immunophenotyping may mirror breast cancer heterogeneity and reveal molecular characteristics of individual tumors useful for testing target therapy.