Dynamic [¹⁸F]FSPG PET: impact of glutamate transport and glutaminase inhibition on redox imbalance and cell death in glutamine-dependent triple negative breast cancer

Objectives: Upregulated glutamine metabolism is an important metabolic signature for aggressive and treatment resistant cancers, including many triple negative breast cancers (TNBCs). Glutamate (Glu) produced from glutamine (Gln) via mitochondrial glutaminase (GLS) feeds into the TCA cycle to provide energy and biosynthetic precursors for tumor growth. Recent studies from our lab and others have revealed that Glu also plays an important role in maintaining redox homeostasis which is essential for survival of cancer cells. Glu is a component amino acid of glutathione, the most abundant cellular antioxidant. Furthermore, via the xCT antiporter (SLC7A11), Glu is exported for import of cystine, a rate-limiting step for glutathione synthesis. We assessed the effect of metabolic inhibition of GLS and/or xCT pathways on cellular reactive oxygen species (ROS) level and induction of apoptosis. To track cytosolic Glu transport cross plasma membrane in vivo, we tested (4S)-4-(3-[¹⁸F]fluoropropyl)-L-glutamic acid ([¹⁸F]FSPG), a non-metabolized structural analog of Glu. We hypothesize that mitochondrial Glu supply is necessary for maintaining cellular redox homeostasis and [¹⁸F]FSPG PET can serve as a tool to track Glu transport in glutaminolytic TNBC and guide treatment targeting GLS and redox balance leading to unmitigated oxidative stress and cell death.

Methods: Human glutaminolytic TNBC cell line (HCC1806) and a non-glutaminolytic estrogen-receptor positive cell line (MCF-7) were used to generate xenografts in athymic mice. Cells were exposed to culture media containing GLS inhibitor CB839 (1 µM), xCT inhibitor Erastin (ERA, 3 µM) and/or chemotherapy doxorubicin (DOX, 0.2 µM) for 24 h. ROS were measured by DHE and CellROX Green assays; early apoptosis was estimated by flowcytometry analysis of cells double-stained with Annexin V-FITC and TOR-PR-3. In vivo dynamic [¹⁸F]FSPG PET imaging (1 h) was performed on HCC1806 and MCF-7 xenograft models.

Results: ROS level in HCC1806 cells increased robustly with CB839 (20% over untreated cells), ERA (30%), or ERA plus CB839 (50%), whereas ROS responses in MCF-7 cells were more blunted. Compared to untreated cells, DOX treatment increased ROS level in HCC1806 cells by 40% whereas DOX combined with CB839 plus ERA led to a remarkable increase of cellular ROS by more than 200%. Combined treatment also led to 31% cells undergoing early apoptosis, compared to 3% in DOX only and 2% in untreated cells. Time-activity curves (TACs) from xenograft models (Suppl Figure 1A) were fit by a reversible one-compartment model in keeping with the known biology of the tracer. The tumor-to-blood ratio (T/B) averaged from the last 6 points (30 min) of the TACs reveal a large difference between HCC1806 vs. MCF-7 by a factor > 2 (suppl Figure 1B). Discussions and Conclusions: Our preliminary data support the efficacy of dual targeting of glutaminolysis and Glu transport for sensitizing glutaminolytic TNBC to cytotoxic chemotherapy. T/B of [¹⁸F]FSPG PET appears to be distinct for glutaminolytic versus non-glutamnolytic breast cancers. Future work will focus on developing [¹⁸F]FSPG PET, complementing our prior work with a Gln analog as a measure of Gln metabolism. The two PET tracers may guide treatment targeting both Gln and Glu usage to overcome chemoresistant TNBC. Support: R01CA211337 and Komen SAC130060. We thank Calithera BioScience and Dr. Marina Gelman for providing CB839 and insightful discussions.