Dataset Information

Gene expression signature from tumor initiating population predicts clinical outcome in undifferentiated pleomorphic sarcoma (UPS)

ABSTRACT: Undifferentiated pleomorphic sarcomas (UPS) are thought to harbor a small population of cells with unique capability of tumor initiation and maintenance. These tumor initiating cells (TICs) are regarded as the drivers of cancer progression. However, the evidence for this tumor initiating cell (TIC) concept is based on the xenotransplantation assays in mice and the clinical relevance of the TIC concept remains unclear in UPS. We hypothesized that the distinct properties of the TICs would be reflected in the clinical outcome if the TIC concept is relevant in UPS. Here we performed global gene expression profiling of TIC-enriched side population (SP) fractions and non-SP fractions sorted from 15 UPS primary samples. Ninety-three genes were differentially expressed in between SP fractions and non-SP fractions. The UPS TIC gene expression signature score summarizing expression of these genes was calculated for a gene expression data and was correlated with clinical outcome. In the 15 samples used to generate the signature, patients with high scores in the SP fractions were associated with worse survival. We then tested the two independent published datasets with gene expression data on bulk unsorted samples, assuming that the TIC gene expression would persist in the non-TICs. Patients with high scores had significantly worse metastasis-free survival in both datasets. The significance of the score remained significant even considering for the known prognostic factor of UPS. Thus, gene expression signature derived from UPS TICs predicts clinical outcome, suggesting the clinical relevance of the TIC concept in UPS.

ORGANISM(S): Homo sapiens

PROVIDER: GSE55973 | GEO | 2016/02/01

SECONDARY ACCESSION(S): PRJNA241462

REPOSITORIES: GEO

ACCESS DATA

Similar Datasets

Project description:Background: cancer cells rely on glycolysis as main ATP source (Warbürg effect). Tumor-initiating cells (TICs) are the fraction of cells that give raise and repopulate tumors. TICs are exposed to prolonged periods of oxygen and glucose deprivation (OGD), as they live in a hypoxic niche and they withstand prolonged lack of blood vessels during initial tumorigenesis or metastasis formation (avascular phase). Warbürg effect is energetically inefficient; we hypothesize that TICs might have differential metabolic features. Tumor eradication requires killing TICS; finding such features would have therapeutic implications. Methodology/principal findings: 106 MDA-MB-231 breast-cancer cells (hereafter Wt) were exposed for 5 weeks to 0.2% oxygen and 0.1g/l glucose, recovering 9 clones. Both flow-cytometry (50-fold enrichment in the CD24-/CD44+/CD133+ population) and xenografts in NOD/SCID mice using 100 cells (75% vs. 16% engraftment) suggest that OGD-resistant clones are true TICs (hereafter “TIC clones”). TIC clones showed a 30-fold higher replication and viability (BRDU incorporation, colony assay) than Wt. When exposed to OGD, ATP-production dropped 5-fold in WT, but was maintained in TIC clones by increasing 5-fold the fatty acid and oxygen consumption. These properties were explained by lack of upregulation of HIF-1 alpha and PDK1, as well as an increase in ATP-synthase. Analysis with metabolic inhibitors (2-deoxyglucose, antimycin-A) confirmed glycolysis and mitochondrial respiration as main routes of metabolism for Wt and TIC clones respectively. Metabolomics revealed that glutamine catabolism generated the NADPH required to quench reactive oxygen species generated during mitochondrial respiration in TIC clones. Glutamine deprivation or mitochondrial blockers were able to abrogate the viability of TIC clones. Conclusions/significance: the TIC-fraction of a cancer cell population withstands prolonged OGD by switching from the Warbürg effect to mitochondrial respiration. Targeting this metabolic feature abrogates the survival of TICs.

Dataset Information

Gene expression signature from tumor initiating population predicts clinical outcome in undifferentiated pleomorphic sarcoma (UPS)

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure