Research efforts investigating the potential of natural compounds in the fight against cancer are growing. Tannic acid (TA) belongs to the class of hydrolysable tannins and is found in numerous plants and foods. TA is a potent collagen cross-linking agent; the purpose of this study was to generate TA-cross-linked beads and assess the effects on breast cancer cell growth. Collagen beads were stable at body temperature following crosslinking. Exposure to collagen beads with higher levels of TA inhibited proliferation and induced apoptosis in normal and cancer cells. TA-induced apoptosis involved activation of caspase 3/7 and caspase 9 but not caspase 8. Breast cancer cells expressing the estrogen receptor were more susceptible to the effects of TA. Taken together the results suggest that TA has the potential to become an anti-ER+ breast cancer treatment or preventative agent.

The immortal strand theory postulates stem cells protect themselves from DNA replication-associated mutations and subsequent cancer risk through selective segregation of template DNA strands. Stem cells self-renew by asymmetric cellular division. During asymmetric division, stem cells maintain their template DNA strands, while the newly synthesized DNA strands segregate to newly formed daughter cells. Previous studies have demonstrated that self-renewing mammary stem cells originate in the expanding mammary ducts during puberty-associated allometric growth. In this study, we labeled newly forming mammary stem cells with the thymidine analog 5-ethynl-2'-deoxyuridine for 2 weeks during allometric ductal expansion. Cells that incorporate and retain the nuclear label following extended chase periods are termed label-retaining cells (LRCs). A second nuclear label, 5-bromodeoxyuridine, was administered before euthanasia to identify cells traversing the cell cycle. Mammary cells collected following euthanasia were sorted based on nuclear label retention. Members of the Notch and Wnt signaling pathways were found differentially expressed by mammary LRCs. These pathways are involved in the regulation of stem cells in the mouse mammary gland. Upon further analysis, we found that in contrast to non-LRCs, Notch1 and Notch2 are expressed and localized in the nuclei of the LRCs. Expression of Notch-inducible genes, Hes1 and Hey2, was elevated in LRCs. Inhibition of Notch1 by shRNA reduced colony forming potential and label retention by mammary epithelial cells in vitro. These results indicate that genes are differentially regulated in the LRC population of mammary glands and Notch1 mediates asymmetric cell division of mammary progenitor cells.

Overexpression of the oncoprotein erbB2/HER2 is present in 20-30% of breast cancer patients and inversely correlates with patient survival. Reports have demonstrated the deterministic power of the mammary microenvironment where the normal mammary microenvironment redirects cells of non-mammary origin or tumor-derived cells to adopt a mammary phenotype in an in vivo model. This phenomenon is termed tumor cell redirection. Tumor-derived cells that overexpress the erbB2 oncoprotein lose their tumor-forming capacity in this model. In this model, phosphorylation of erbB2 is attenuated thus reducing the tumor cell's tumor-forming potential. In this report, we describe our results using an in vitro model based on the in vivo model mentioned previously. Tumor-derived cells are mixed in predetermined ratios with normal mammary epithelial cells prior to seeding in vitro. In this in vitro model, the tumor-derived cells are redirected as determined by attenuated phosphorylation of the receptor and reduced sphere and colony formation. These results match those observed in the in vivo model. This in vitro model will allow expanded experimental options in the future to determine additional aspects of tumor cell redirection that can be translated to other types of cancer.

Based on our results we conclude that AREG is involved in regulating the proliferation and migration of erbB2/HER2-positive breast cancer cells.

Introducing tumor-derived cells into normal mammary stem cell niches at a sufficiently high ratio of normal to tumorous cells causes those tumor cells to undergo a change to normal mammary phenotype and yield normal mammary progeny. This phenomenon has been termed cancer cell redirection. We have developed an in vitro model that mimics in vivo redirection of cancer cells by the normal mammary microenvironment. Using the RNA profiling data from this cellular model, we examined high-level characteristics of the normal, redirected, and tumor transcriptomes and found the global expression profiles clearly distinguish the three expression states. To identify potential redirection biomarkers that cause the redirected state to shift toward the normal expression pattern, we used mutual information relationships between normal, redirected, and tumor cell groups. Mutual information relationship analysis reduced a dataset of over 35,000 gene expression measurements spread over 13,000 curated gene sets to a set of 20 significant molecular signatures totaling 906 unique loci. Several of these molecular signatures are hallmark drivers of the tumor state. Using differential expression as a guide, we further refined the gene set to 120 core redirection biomarker genes. The expression levels of these core biomarkers are sufficient to make the normal and redirected gene expression states indistinguishable from each other but radically different from the tumor state.

The in vitro growth and differentiation of normal mammalian cells is quite different than the growth of cells derived from tumors. Additionally, cells of the same origin (tissue) behave differently depending on the biomaterial matrix in or on which they are grown in vitro. We examined both Matrigel(TM) and a collagen/agarose blend and demonstrated that two murine mammary derived cells lines, 4T1 and NMuMG, derived from a metastatic mammary tumor or a normal mammary gland, respectively, exhibit different growth and differentiation patterns depending on the three-dimensional matrix in which they are grown. The shape and size of the colonies that formed were matrix dependent. The two cell lines produced different levels of growth factors and metalloproteinases, and expressed differentiation markers specific to a matrix. Through the classification of different cell behaviors in different growth matrices, we will be able to intelligently design and tune tissue test systems to ask and answer specific challenging scientific questions.