The skeleton is a unique structure capable of providing support for the body. Bone resorption and deposition are controlled in a tightly regulated balance between osteoblasts and osteoclasts with no net bone gain or loss. However, under conditions of disease, the balance between bone resorption and deposition is upset. Osteoblasts play an important role in bone homeostasis by depositing new bone osteoid into resorption pits. It is becoming increasingly evident that osteoblasts additionally play key roles in cancer cell dissemination to bone and subsequent metastasis. Our laboratory has evidence that when osteoblasts come into contact with disseminated breast cancer cells, the osteoblasts produce factors that initially reduce breast cancer cell proliferation, yet promote cancer cell survival in bone. Other laboratories have demonstrated that osteoblasts both directly and indirectly contribute to dormant cancer cell reactivation in bone. Moreover, we have demonstrated that osteoblasts undergo an inflammatory stress response in late stages of breast cancer, and produce inflammatory cytokines that are maintenance and survival factors for breast cancer cells and osteoclasts. Advances in understanding interactions between osteoblasts, osteoclasts, and bone metastatic cancer cells will aid in controlling and ultimately preventing cancer cell metastasis to bone.

Understanding how cancer cells resist ferroptosis is a significant problem that impacts ongoing efforts to stimulate ferroptosis as a therapeutic strategy. We reported that prominin2 is induced by ferroptotic stimuli and functions to resist ferroptotic death. Although this finding has significant implications for therapy, specific prominin2 inhibitors are not available. We rationalized that the mechanism by which prominin2 expression is induced by ferroptotic stress could be targeted, expanding the range of options to overcome ferroptosis resistance. Here, we show that that 4‐hydroxynonenal (4HNE), a specific lipid metabolite formed from the products of lipid peroxidation stimulates PROM2 transcription by a mechanism that involves p38 MAP kinase‐mediated activation of HSF1 and HSF1‐dependent transcription of PROM2 . HSF1 inhibitors sensitize a wide variety of resistant cancer cells to drugs that induce ferroptosis. Importantly, the combination of a ferroptosis‐inducing drug and an HSF1 inhibitor causes the cytostasis of established tumors in mice, although neither treatment alone is effective. These data reveal a novel approach for the therapeutic induction of ferroptosis in cancer.

Introduction: In a cancer-free environment in the adult, the skeleton continuously undergoes remodeling. Bone-resorbing osteoclasts excavate erosion cavities, and bone-depositing osteoblasts synthesize osteoid matrix that forms new bone, with no net bone gain or loss. When metastatic breast cancer cells invade the bone, this balance is disrupted. Patients with bone metastatic breast cancer frequently suffer from osteolytic bone lesions that elicit severe bone pain and fractures. Bisphosphonate treatments are not curative. Under ideal circumstances, osteoblasts would synthesize new matrix to fill in erosion cavities caused by osteoclasts, but this is not what occurs. Our prior evidence demonstrated that osteoblasts are diverted from laying down bone matrix to producing cytokines that facilitate breast cancer cell maintenance in late-stage disease. Here, we have new evidence to suggest that there are subpopulations of osteoblasts in the tumor niche as evidenced by their protein marker expression that have distinct roles in tumor progression in the bone.Methods: Tumor-bearing tibia of mice was interrogated by immunofluorescent staining for the presence of osteoblasts and alterations in niche protein expression. De-identified tissue from patients with bone metastatic breast cancer was analyzed for osteoblast subpopulations via multi-plex immunofluorescent staining. Effects of breast cancer cells on osteoblasts were recapitulated in vitro by osteoblast exposure to breast cancer-conditioned medium. Triple-negative and estrogen receptor-positive breast cancer proliferation, cell cycle, and p21 expression were assessed upon contact with “educated” osteoblasts.Results: A subpopulation of osteoblasts was identified in the bone tumor microenvironment in vivo of both humans and mice with bone metastatic breast cancer that express RUNX2/OCN/OPN but is negative for IL-6 and alpha-smooth muscle actin. These tumor “educated” osteoblasts (EOs) have altered properties compared to “uneducated” osteoblasts and suppress both triple-negative and estrogen receptor-positive breast cancer cell proliferation and increase cancer cell p21 expression. EO effects on breast cancer proliferation were mediated by NOV and decorin. Importantly, the presence of EO cells in the tibia of mice bearing tumors led to increased amounts of alkaline phosphatase and suppressed the expression of inflammatory cytokines in vivo.Conclusions: Our work reveals that there is a subpopulation of osteoblasts in the bone tumor microenvironment that demonstrate a functional role in retarding breast cancer cell growth.

Breast tumors are heterogeneous and composed of different subpopulation of cells, each with dynamic roles that can change with stage, site, and microenvironment. Cellular heterogeneity is, in part, due to cancer stem–like cells (CSC) that share properties with stem cells and are associated with treatment resistance. CSCs rewire metabolism to meet energy demands of increased growth and biosynthesis. O-GlcNAc transferase enzyme (OGT) uses UDP-GlcNAc as a substrate for adding O-GlcNAc moieties to nuclear and cytoplasmic proteins. OGT/O-GlcNAc levels are elevated in multiple cancers and reducing OGT in cancer cells blocks tumor growth. Here, we report that breast CSCs enriched in mammosphere cultures contain elevated OGT/O-GlcNAcylation. Inhibition of OGT genetically or pharmacologically reduced mammosphere forming efficiency, the CD44^H/CD24^L, NANOG+, and ALDH+ CSC population in breast cancer cells. Conversely, breast cancer cells overexpressing OGT increased mammosphere formation, CSC populations in vitro, and also increased tumor initiation and CSC frequency in vivo. Furthermore, OGT regulates expression of a number of epithelial-to-mesenchymal transition and CSC markers including CD44, NANOG, and c-Myc. In addition, we identify Krüppel-like factor 8 (KLF8) as a novel regulator of breast cancer mammosphere formation and a critical target of OGT in regulating CSCs. Implications: These findings demonstrate that OGT plays a key role in the regulation of breast CSCs in vitro and tumor initiation in vivo, in part, via regulation of KLF8, and thus inhibition of OGT may serve as a therapeutic strategy to regulate tumor-initiating activity.

A specific cell surface glycan enables the identification of breast cancer stem cells and contributes to their function.