Metabolic and molecular profiling of ovarian cancer stem cells and cancer non-stem counterpart

Epithelial Ovarian Cancer (EOC) is a very malignant neoplasm, accounting for 5% of cancer mortality in women. Although progress has been made in EOC treatments by improved debulking surgery and platinum-taxane regimens, the 5-year survival rate of advanced-stage EOC remains below 30%. This poor prognosis relies on the one hand on the late diagnosis and on the other on the chemo-resistance occurring after only a few months from the completion of treatment. The reasons for recurrence and cancer drug resistance remain uncertain. Recent evidence suggests that EOC, akin most tumors, contains a tiny population of cells, named cancer stem cells (CSC), probably responsible for chemotherapy resistance and tumor recurrence. Ovarian CSC are characterized by the co-expression of two surface markers: CD44 (hyaluronic acid receptor) and CD117 (stem cell factor receptor or c-kit). Recently, our research group demonstrated that ovarian CD44/CD117 co-expressing cells, which represent 1-2% of cancer cells from ascitic effusions of EOC-bearing patients, are endowed with canonical stemness properties and are able to resist in vitro and in vivo glucose starvation. We reported that this glucose deprivation resistance is mostly due to the ability of ovarian CSC to privilege oxidative phosphorylation, rather than the aerobic glycolysis (Warburg effect) exploited by the non-stem tumor bulk. However, independently of CSC fraction, our experiments also highlighted that not all the analyzed EOC samples presented a similar glucose addiction. Thus, investigating this issue and its related metabolic aspects is the first aim of the current project. Concerning this aim, we showed that tumor cells from EOC patients can be categorized, according to their in vitro viability under glucose starvation, into glucose deprivation-sensitive (glucose-addicted, GA) and glucose deprivation-resistant (glucose non-addicted, GNA). Although deregulated glucose metabolism is usually observed in cancer, whether this metabolic trait influences response to or is modulated by cytotoxic drugs is unknown; therefore, we addressed the possible correlation of these glucose addiction profiles with the patient response to platinum (PLT) regimens. In this regard, when EOC cells were cultured in the absence of glucose, all samples from PLT-sensitive patients felt into the GA group; compared to GNA samples, they disclosed higher expression of glucose metabolic enzymes, higher proliferation rates and in vitro sensitivity to PLT, as well as reduced multi-drug resistance pump expression. On the other hand, the samples derived from PLT-resistant patients felt into the GNA category. The close association between PLT sensitivity and glucose metabolic profile was confirmed in a xenograft model, where a stringent parallelism between PLT sensitivity/resistance and glucose metabolism was identified. Finally, in a cohort of naïve EOC patients categorized as GA or GNA at diagnosis, Kaplan Meier curves showed that the GA phenotype was associated with significantly better progression-free survival, compared to GNA patients. Overall, these results suggest that in vitro glucose addiction of EOC cells could be regarded as a reliable marker to predict the patient response to platinum regimens. Investigating the molecular traits leading to the distinctive glucose metabolism of EOC samples is the second aim of this project. In this regard, microRNA (miRNA), that are small non-coding RNA molecules, represent a promising field, in view of their ability to modulate many genes and pathways. Moreover, their involvement in cancer development and progression has already been reported in ovarian cancer, and recently miRNA also emerged to regulate cell metabolism in several normal and cancer tissues. Thus, we performed a miRNA profile on EOC patient-derived samples, comparing both GA versus GNA samples and CSC versus non-CSC, in order to establish whether the miRNA signature behind the metabolic differences of EOC samples is associated with the total tumor bulk or rather with a specific cell fraction. The data did not reveal any miRNA differentially expressed in GA compared to GNA cells, but several miRNA resulted significantly deregulated in CSC versus non-CSC. We focused our attention on mir-602, which was found up-regulated in CSC; indeed, despite little information about this miRNA, its target Casein Kinase 1 Delta (CSNK1D), which displays a key role in cell proliferation and asymmetric division, seemed very interesting in view of its involvement in breast cancer progression. We demonstrated that CSNK1D is down-regulated in CSC, according to mir-602 over-expression. Moreover, the in vitro inhibition of mir-602 reduced the expression of the major stemness-associated genes in CSC, suggesting that mir-602 could regulate some cell stemness pathways. Since none of the analyzed stemness genes is directly targeted by mir-602, it is reasonable to advance that mir-602 could indirectly activate the expression of such genes through its inhibitory function on CSNK1D; thus, we propose that cell stemness signaling is inhibited by casein kinase, whose translation is in turn repressed by mir-602. Although a few further experiments are needed to validate our idea, this project highlights a possible miRNA-mediated regulatory mechanism of cell stemness features in EOC