Yi Li Laboratory
Stem Cells, Wnt signaling, and Breast Cancer
Phone: 713-798-3963
Fax: 7113-798-1673
liyi@breastcenter.tmc.edu
The breast develops from stem cells capable of both self-renewal and differentiation into progenitor cells, which can proliferate many times and progressively differentiate into fully differentiated cells which make up the bulk of the tissue. In mammary ducts, where breast cancers arise, ductal epithelial cells, alveolar epithelial cells, and myoepithelial cells are the principal differentiated cell types. Breast cancers also appear to consist largely of more differentiated but rapidly proliferating cells, but there is evidence that at least some of them also contain small numbers of cancer stem cells or progenitor cells that have the capacity to regenerate “differentiated” tumor cells. Thus, the initial success and eventual failure of many traditional radiation and chemotherapies may well be due to its success in killing the bulk of proliferating “differentiated” tumor cells while leaving the small cadre of cancer stem/progenitor cells able to produce new tumor tissue.
This concept is leading to a new interest in targeting stem cells in therapy, though without a clear understanding of cancer stem cells and where they come from. A natural assumption has been that breast tumors must arise from normal breast stem cells. We have discovered that mice carrying the Wnt-1 transgene develop mammary tumors with transformed ductal epithelial cells, myoepithelial cells, and cells expressing putative stem/progenitor cell markers, suggesting that these tumors originated from multipotential stem/progenitor cells, consistent with this assumption. The Wnt transgenic mouse is now a widely used model for breast cancer having a stem/progenitor cell origin. In contrast, however, mammary tumors arising in mice with Neu (HER2, ErbB2), Ras, or PyMT transgenes did not show multiple cell types. Does this mean that these oncogenes cause tumors only from differentiated epithelial cells? Or could these oncogenes initially transform stem/progenitor cells as the assumption suggests, but force their differentiation? Or conversely, is it possible that certain other oncogenes transform differentiated cells but induce dedifferentiation so that they can produce multiple cell types?
There are distinct subtypes of breast cancer by histopathology and/or gene expression array profile – it may well be that breast cancer subtypes arise from different oncogenic pathways acting at different developmental stages of mammary cells, and these different types require different therapies. Furthermore, since several genetic/ epigenetic alterations are essential to ultimately transform breast cells into malignancy, these secondary changes likely also have a role in defining cancer features. In order to effectively prevent breast cancer initiation and to attack individual breast cancers that do form, we have to understand the interplay of specific oncogenic alterations with specific subsets of breast cells, including stem cells, and to identify oncogenic networks that are crucial in the progression to a transformed state.
This knowledge can only be acquired through prospectively introducing selected oncogenic mutations into different subsets of breast cells (stem cells, progenitor cells, and more differentiated ductal and alveolar epithelial cells), investigating the cellular response in each subset, and discovering what additional oncogenic pathways have to be perturbed in order to achieve the eventual malignant phenotype. And all of these need to occur in the context of normal breast epithelium and stroma (which is also important in cancer development). It is impossible to directly seek answers to these questions in healthy women. Xenografts of human breast tissues in immunodeficient mice are devoid of the effect of the immune system (which is now known to play a critical role in carcinogenesis), and genetic manipulation of isolated cells would destroy the tissue architecture interaction. Cultured cells (even 3-D cultures) from healthy donors lose the cell hierarchy (the myoepithelial layer and stroma contribution are both lost) and the hormonal influences that are present in healthy women.
Transgenic and knockout mice have been valuable in understanding many aspects of breast cancer when they are used for the right questions. For example, they are essential in testing whether a mutation is truly transforming in breast cells, and in identifying “druggable” survival pathways in breast cancers that have already formed. However, conventional models, and even second generation conditional and inducible models, are made by introducing genetic alterations into essentially all mammary epithelial cells without subtype specificity. Consequently, cancers in these models evolve from cells of any differentiation stage in a “field” of already mutated cells. The use of these models in understanding the molecular and cellular picture of the initiation of sporadic human breast cancers is therefore limited. Thus, the initial step of breast carcinogenesis remains largely a mystery, even though it could be a key target for early prevention.
We have very recenty adapted the TVA retroviral gene delivery technology to introduce genetic mutations into selected mammary cells in mice at selected times (BCM News). This method overcomes many of the shortcomings of existing models, and we are using this third generation mouse model to study how specific subsets of breast cells respond to initiating oncogenes and eventually evolve into cancer, whether the stem cells are especially susceptible, and to how oncogenic pathways—especially the Wnt and HER2/Neu signaling pathways—transform breast cells in vivo.
Reproductive history is one of the strongest risk factors for breast cancer — women with a full-term pregnancy at an early age have a 50% reduction in life-time risk of breast cancer compared with women who experience pregnancy late or not at all. However, the mechanism for this protection is not understood, partly because it has been difficult to use animal models to study which oncogenic events pregnancy protects against, and which cell types are protected. The TVA technology now allows us to approach these questions effectively.
Specific Research Topics
Wnt signaling and breast cancer
The origin of ER– and ER+ breast cancers
