Impaired Anaplerosis Is a Major Contributor to Glycolysis Inhibitor Toxicity in Glioma

Abstract Reprogramming of metabolic pathways is crucial to satisfy the bioenergetic and biosynthetic demands and maintain the redox status of rapidly proliferating cancer cells. In tumors, the tricarboxylic acid (TCA) cycle generates biosynthetic intermediates by oxidation of anaplerotic substrates, such as glucose-derived pyruvate and glutamine20 derived glutamate. We have previously documented that a subset of tumors with 1p36 homozygous deletion exhibit co-deletion of ENO1, in turn becoming extremely dependent on its redundant isoform ENO2 and sensitive to an overall enzymatic deficiency of enolase. Metabolomic profiling of ENO1-deleted glioma cells treated with an enolase inhibitor revealed a profound decrease in TCA cycle metabolites, which correlated with cell-line specific sensitivity to enolase inhibition, highlighting the importance of glycolysis derived pyruvate for anaplerosis. Correspondingly, the toxicity of the enolase inhibitor was significantly attenuated by exogenous supplementation of supraphysiological levels of anaplerotic substrates including pyruvate. These findings led us to hypothesize that cancer cells with ENO1 homozygous deletions treated with an enolase inhibitor might show exceptional sensitivity to inhibition of glutaminolysis because of reduced anaplerotic flow from glycolysis. We found that ENO1-deleted cells indeed exhibited selective sensitivity to the glutaminase inhibitor CB-839, and this sensitivity was also attenuated by exogenous supplementation of anaplerotic substrates including pyruvate. Despite these promising in vitro results, the antineoplastic effects of CB-839 as a single agent in ENO1-deleted xenograft tumors in vivo were modest in both intracranial orthotopic tumors, where the limited efficacy could be attributed to the blood brain barrier (BBB), and subcutaneous xenografts, where BBB penetration is not an issue. This contrasts with the enolase inhibitor HEX, which, despite its negative charge, achieved antineoplastic effects in both intracranial and subcutaneous tumors. Together, these data suggest that at least for 1p36-deleted gliomas, tumors in vivo—unlike cells in culture—show limited dependence on glutaminolysis and instead primarily depend on glycolysis for anaplerosis. Our findings reinforce the previously reported metabolic idiosyncrasies of the in vitro and in vivo environments as the potential reasons for the differential efficacy of metabolism targeted therapies in in vitro and in vivo systems.