Molecular and therapeutic aspects of breast cancer stem cells

Abstract: Breast cancer stem cells (BCSCs) are defined as a minor cellular component in the solid tumor, displaying mammary stem cell properties by expressing embryonic stem cell genes. They are usually undergoing self-.‐renewal process, but highly potent to give rise to other types of tumor cells and create a hierarchically arranged tumor under certain stimuli. Accumulating evidences suggest that BCSCs are more resistant to conventional therapies and play an important role in mediating tumor metastasis and relapse. Therefore more targeted therapy on the scope of BCSCs is required to improve prognosis of breast cancer patients. Two estrogen receptors (ER) exist in breast cancer: ERα and ERβ. Around 75% of all primary breast tumors can be identified as ERα positive. In study I, we confirm BCSCs derived from human tumors as well as cell lines are absent of ERα expression but dominant in ERβ expression. We show that ERβ is correlating with phenotypic cancer stem cell markers and responsible for the proliferative role of estrogens and thereby essential for tumor growth. We also observe ERβ activated gene signature in BCSCs associated with poor outcome in three independent breast cancer cohorts. By in vivo xenograft experiments we show that agonizing or antagonizing ERβ enhances or inhibits tumor growth respectively. Hence, we suggest that ERβ is a central mediator of estrogen action in BCSCs. Although BCSCs are considered the apex of hierarchy in breast cancer progression, the origin of this unique cell subpopulation is not well established. Two alternative and contradictive models have been proposed to describe the tumor growth governed by BCSCs. In study II, we observe a large degree of shared mutations as well as similar spread of allele frequencies between BCSCs and the differentiated tumor cells by performing whole exome sequencing. We thereby conclude the existence of BCSCs as a population of cells dynamically and reversibly switching from differentiated tumor cells. In paper I, we observed that tamoxifen appears to be insufficient to reduce mammosphere formation and proliferation even from BCSCs derived from ERα+ tumors. Therefore, in study III we conducted genome-.‐wide transcriptional analysis to explore potential mechanism posed by tamoxifen. Within the tamoxifen-.‐induced transcriptional pathways, ribosomal biogenesis and mRNA translation are the most significantly and physiologically relevant pathways. We also identified induction of the key mTOR downstream targets S6K1, S6RP and 4EBP1 in BCSCs by tamoxifen on protein level. Using mTOR inhibitors along or in combination with tamoxifen, we observed significant reduction in mammosphere formation. We conclude that tamoxifen can activate mTOR-.‐signaling pathways in BCSCs associated with endocrine resistance. The rapid evolving genomic technologies and the advantages of using ex vivo modeling widens the possibilities to perform improved diagnostics for precision oncology. In study IV, we report a simple but robust superficial scraping method to collect viable primary cells from small tumors. Validation of scraping material has proven the method to be sufficient for analayses at DNA and mRNA level. By using this method for routine biobanking, the inclusion of resected tumors for fresh frozen sample storage could increase significantly from 60% up to 85% of the total breast cancers. In conclusion, studying BCSCs as a model system and as a target will be important for development of cancer therapies in advanced breast cancer.