A novel mouse model of human breast cancer stem-like cells with high CD44+CD24-/lower phenotype metastasis to human bone.

CTRC-AACR San Antonio Breast Cancer Symposium: 2008 Abstracts Abstract #4155 Background: At present, our understanding of the development of bone metastasis is limited. Therefore, little progress has been made in preventing skeletal metastasis in the breast cancer patient. A satisfactory animal model that avoids the species-specific factor and simultaneously shares similarities to the clinical pathophysiological progression of breast cancer metastasizing to bone is unavailable. A subpopulation (CD44+/CD24-/lower) of breast cancer cells possesses stem/progenitor cell properties (cancer stem-like cell). CD44 potentiates the adherence of metastatic breast cancer cells to bone marrow endothelial cells. In the present study, we used hepatocyte growth factor to enhance the proportion of CD44+/CD24-/lower subpopulation in the human breast cancer cell line MDA-MB-231. We used these human breast cancer stem-like cells and implantation of human bone to build a novel human-source model of human breast cancer skeletal metastasis. Methods: The human breast cancer cell line MDA-MB-231 was cultured in serum-free DMEM-F12 supplemented with growth factors. Cells in different subpopulations were separated by Percoll gradient centrifugation and incubated in the presence of 50 ng/mL hepatocyte growth factor. The proportion of CD44+/CD24-/lower subpopulation in the human breast cancer cell line MDA-MB-231 was detected with flow cytometry. Before injection with human breast cancer stem-like cells, the experimental animals were implanted with human bone in the right or left dorsal flanks. Animals in Groups A, B, and C were injected with 1 X 105, 1 X 106 human breast cancer stem-like cells, and 1 X 106 parental MDA-MB-231 cells, respectively. A positive control group (D) without implantation of human bone (non-human source model) was also injected with 1 X 106 MDA-MB-231 cells. A group of negative controls (E) with human bone implantation was injected with isotonic sodium chloride. For each group, Micro-SPECT was performed at weeks 4 and 7, and all animals were sacrificed at week 8. Immunohistochemistry was performed for CD34, CD105, SMA, CD44, CD24, CK, CXCR4, and OPN. mRNA levels of CD44, CD24, CXCR4, and OPN in bone metastasis tissues were analyzed by real-time quantitative polymerase chain reaction. Results: New vessels and connective tissue were found on the surface of implanted human bones and cells stained positive for antibodies against human CD105, SMA, and CD34, indicating that implanted human bones were viable and functional. Histologic and immunohistochemical analysis confirmed the metastases as cancer cells. Importantly, the results demonstrated that cells in implanted human bones of group B, which received 1 X 106 cancer stem-like cells, stained strongly positive for CD44, CXCR4, and OPN, whereas that of other groups showed no or minimum staining. Moreover, group B had the highest incidence of human bone metastasis (77.8%, P = 0.0230) and no accompaniment of other tissue metastasis. The real-time polymerase chain reaction (PCR) showed an increase of CD44 mRNA in metastatic bone tissues in group B compared with that of groups C and D (15.2- and 21.1-fold, respectively). The mRNA levels of CXCR4 and OPN (8.4- and 28.4-fold, respectively) in bone metastasis tissues of group B were all higher than that of groups C and D (4.8- and 11.6-fold; respectively). The levels of CD24 mRNA in group B were lowest, measuring only 30 percent of that in groups C and D. Conclusion: This study indicates that in the novel human source model of breast cancer, breast cancer stem-like cells demonstrate a higher human bone-seeking ability, which may contribute to increase metastasis incidence and attenuate species-specific influences. Its mechanism might be related to the higher expressions of CD44, CXCR4 and OPN, and the lower expression of CD24 in breast cancer stem-like cells. The model shares more similarities with clinical pathological features of bone metastatic patients. It will be helpful for further study of the mechanisms and subclinical diagnosis of bone metastasis. Citation Information: Cancer Res 2009;69(2 Suppl):Abstract nr 4155.