Recent studies have identified a subpopulation of highly tumorigenic cells with stem/progenitor cell properties from human breast cancers, and it has been suggested that stem/progenitor cells, which remain after breast cancer therapy, may give rise to recurrent disease. We hypothesized that progenitor cells are resistant to radiation, a component of conventional breast cancer therapy, and that that resistance is mediated at least in part by Wnt signaling, which has been implicated in stem cell survival. To test this hypothesis, we investigated radioresistance by treating primary BALB/c mouse mammary epithelial cells with clinically relevant doses of radiation and found enrichment in normal progenitor cells (stem cell antigen 1-positive and side population progenitors). Radiation selectively enriched for progenitors in mammary epithelial cells isolated from transgenic mice with activated Wnt/β-catenin signaling but not for background-matched controls, and irradiated stem cell antigen 1-positive cells had a selective increase in active β-catenin and survivin expression compared with stem cell antigen 1-negative cells. In clonogenic assays, colony formation in the stem cell antigen 1-positive progenitors was unaffected by clinically relevant doses of radiation. Radiation also induced enrichment of side population progenitors in the human breast cancer cell line MCF-7. These data demonstrate that, compared with differentiated cells, progenitor cells have different cell survival properties that may facilitate the development of targeted antiprogenitor cell therapies.

Using a syngeneic p53-null mouse mammary gland tumor model that closely mimics human breast cancer, we have identified, by limiting dilution transplantation and in vitro mammosphere assay, a Lin⁻CD29^HCD24^H subpopulation of tumor-initiating cells. Upon subsequent transplantation, this subpopulation generated heterogeneous tumors that displayed properties similar to the primary tumor. Analysis of biomarkers suggests the Lin⁻CD29^HCD24^H subpopulation may have arisen from a bipotent mammary progenitor. Differentially expressed genes in the Lin⁻CD29^HCD24^H mouse mammary gland tumor-initiating cell population include those involved in DNA damage response and repair, as well as genes involved in epigenetic regulation previously shown to be critical for stem cell self-renewal. These studies provide in vitro and in vivo data that support the cancer stem cell (CSC) hypothesis. Furthermore, this p53-null mouse mammary tumor model may allow us to identify new CSC markers and to test the functional importance of these markers. [Cancer Res 2008;68(12):4674–82]

Metastatic cancer is almost always terminal, and more than 90% of cancer deaths result from metastatic disease. Combating cancer metastasis and post-therapeutic recurrence successfully requires understanding each step of metastatic progression. This review describes the current state of knowledge of the etiology and mechanism of cancer progression from primary tumor growth to the formation of new tumors in other parts of the body. Open questions, avenues for future research, and therapeutic approaches with the potential to prevent or inhibit metastasis through personalization to each patient's mutation and/or immune profile are also highlighted.

Introduction: Human models of noninvasive breast tumors are limited, and the existing in vivo models do not mimic inter- and intratumoral heterogeneity. Ductal carcinoma in situ (DCIS) is the most common type (80%) of noninvasive breast lesions. The aim of this study was to develop an in vivo model whereby the natural progression of human DCIS might be reproduced and studied. To accomplish this goal, the intraductal human-in-mouse (HIM) transplantation model was developed. The resulting models, which mimicked some of the diversity of human noninvasive breast cancers in vivo, were used to show whether subtypes of human DCIS might contain distinct subpopulations of tumor-initiating cells. Methods: The intraductal models were established by injection of human DCIS cell lines (MCF10DCIS.COM and SUM-225), as well as cells derived from a primary human DCIS (FSK-H7), directly into the primary mouse mammary ducts via cleaved nipple. Six to eight weeks after injections, whole-mount, hematoxylin and eosin, and immunofluorescence staining were performed to evaluate the type and extent of growth of the DCIS-like lesions. To identify tumor-initiating cells, putative human breast stem/progenitor subpopulations were sorted from MCF10DCIS.COM and SUM-225 with flow cytometry, and their in vivo growth fractions were compared with the Fisher's Exact test. Results: Human DCIS cells initially grew within the mammary ducts, followed by progression to invasion in some cases into the stroma. The lesions were histologically almost identical to those of clinical human DCIS. This method was successful for growing DCIS cell lines (MCF10DCIS.COM and SUM-225) as well as a primary human DCIS (FSK-H7). MCF10DCIS.COM represented a basal-like DCIS model, whereas SUM-225 and FSK-H7 cells were models for HER-2⁺ DCIS. With this approach, we showed that various subtypes of human DCIS appeared to contain distinct subpopulations of tumor-initiating cells. Conclusions: The intraductal HIM transplantation model provides an invaluable tool that mimics human breast heterogeneity at the noninvasive stages and allows the study of the distinct molecular and cellular mechanisms of breast cancer progression.

Background: Human breast cancer is a heterogeneous disease consisting of multiple molecular subtypes. Genetically engineered mouse models are a useful resource for studying mammary cancers in vivo under genetically controlled and immune competent conditions. Identifying murine models with conserved human tumor features will facilitate etiology determinations, highlight the effects of mutations on pathway activation, and should improve preclinical drug testing. Results: Transcriptomic profiles of 27 murine models of mammary carcinoma and normal mammary tissue were determined using gene expression microarrays. Hierarchical clustering analysis identified 17 distinct murine subtypes. Cross-species analyses using three independent human breast cancer datasets identified eight murine classes that resemble specific human breast cancer subtypes. Multiple models were associated with human basal-like tumors including TgC3(1)-Tag, TgWAP-Myc and Trp53^-/-. Interestingly, the TgWAPCre-Etv6 model mimicked the HER2-enriched subtype, a group of human tumors without a murine counterpart in previous comparative studies. Gene signature analysis identified hundreds of commonly expressed pathway signatures between linked mouse and human subtypes, highlighting potentially common genetic drivers of tumorigenesis. Conclusions: This study of murine models of breast carcinoma encompasses the largest comprehensive genomic dataset to date to identify human-to-mouse disease subtype counterparts. Our approach illustrates the value of comparisons between species to identify murine models that faithfully mimic the human condition and indicates that multiple genetically engineered mouse models are needed to represent the diversity of human breast cancers. The reported trans-species associations should guide model selection during preclinical study design to ensure appropriate representatives of human disease subtypes are used.