GLIOBLASTOM - Anni Hofmann Stiftung

Summary of the work performed during the 2nd funding period (January 2016 - December 2017) on the project "Tumor metabolism - interactions between glioblastoma cells and brain"

The research project aimed to investigate the relevance of tumor metabolism for the invasive growth of glioblastomas in the brain of patients and to develop therapeutic strategies based on these findings. The project built on previous work, in which we had identified a hypoxia-induced switch between the pentose phosphate pathway (PPP) and glycolysis in glioblastoma cells and had further shown that both metabolic pathways are causatively involved in the dichotomous regulation between tumor cell proliferation and migration, respectively. While the relevance of the PPP for cell division is rather obvious as this pathway produces building bricks for new cells, the functional relevance of glycolysis for tumor cell migration is largely unexplored.
We hypothesized that the glycolysis enzyme glucose-6-phosphate-isomerase (GPI), which is also known as autocrine motility factor (AMF), plays an important role in stimulating glioblastoma cell migration and invasion. Our experimental work showed that both GPI/AMF as well as its receptor AMFR are strongly upregulated by hypoxia and that in particular the extracellular secretion of GPI/AMF is enhanced by hypoxia. Expression analyses demonstrated that both GPI/AMF and AMFR are predominantly expressed in severely hypoxic regions in glioblastomas and that high expression of GPI/AMF correlates with a shorter survival of glioblastoma patients. Functional studies revealed that GPI/AMF strongly stimulates the migration of glioblastoma cells in an autocrine fashion, whereas proliferation tends to be inhibited. These observations were supported by in vivo studies, showing that the knockdown of GPI/AMF expression via shRNA resulted in reduced invasion but increased proliferation. Our findings can explain the association between glycolysis and glioblastoma cell migration, however, the partial anti-proliferative effect of GPI/AMF raises caution as to whether this is a suitable molecular target for interfering with glycolysis and for inhibiting tumor invasion.
In developing therapeutic strategies for metabolic targeting, it has to be taken into account that glioblastomas are highly heterogeneous tumors whose intratumoral and intertumoral molecular heterogeneity is reflected also in metabolic heterogeneity. In order to investigate the causes and mechanisms of metabolic heterogeneity at the clonal level, we developed a fluorescence-based clonal labeling system (optical barcoding). Tracking analyses in vitro and in vivo demonstrated that the metabolic adaptability of the tumor cells and the local microenvironment had a crucial influence on the clonal tumor initiation capacity. In addition, we identified intertumoral metabolic differences in gliomas with a different genetic background in collaboration with Prof. Rainer Glass. RNA sequencing analyses revealed that different metabolic pathways dominate in transgenic mouse glioma cell lines with alterations of the genes TP53, PDGF-B, CDKN2a and EGFR. In particular, the expression of glycolytic enzymes is strikingly differentially regulated, indicating that different subgroups of glioblastomas are likely to respond differently to metabolic targeting. Consequently, clinical trials that attempt to interfere with glycolysis and other metabolic pathways should stratify patients according to the different molecular subgroups that exist.
Our prior analyses further showed that in addition to the PPP several other metabolic pathways that are linked with glycolysis are also regulated by hypoxia. In our project we validated a number of these regulatory changes at the transcript and protein level. Furthermore, we performed expression analyses on human glioblastoma tissue and functional analyses for certain selected enzymes. To conduct the functional studies, we downregulated the expression of individual molecules that are potential candidates for therapeutic targeting by using shRNA and also carried out experiments with specific inhibitors.