EHA Library - The official digital education library of European Hematology Association (EHA)

FLT3 tyrosine kinase (TK) activating mutations (FLT-3^mut) are amongst the most frequent in AML and are associated with a poor outcome. FLT-3^mut promote constitutive activation of survival/proliferation pathways and have also been shown to lead to changes in cellular metabolism, such as increased glycolysis.

The FLT3 TK represents a valid therapeutic target and several FLT3 TK inhibitors (TKI) have been developed. However despite showing activity in the preclinical setting, FLT3 TKI have displayed limited efficacy in clinical trials. Resistance mechanisms to FLT3 TKI include receptor mutations and cell intrinsic adaptive mechanisms. Amongst the latter, metabolic adaptations might play a significant role although the exact mechanisms are still ill-defined.

We hypothesised that metabolic adaptations facilitate FLT3 TKI resistance and aimed to identify early metabolic changes in FLT-3^mut AML, following TKI treatment, in an attempt to unveil novel therapeutic vulnerabilities. 

Liquid chromatography coupled to mass spectrometry (LC/MS), using stable isotope-based carbon flux tracing, and oxygen consumption rate/extracellular acidification rate as measured by an extracellular flux analyser (Seahorse, Agilent Technologies) were used to assess metabolic changes in FLT3^mut cells after FLT3 TKI treatment. Gene expression changes were measured in the same conditions by RNA sequencing. Changes in viability and reactive oxygen species (ROS) in various culture conditions were measured by FACS. Gene silencing was performed using CRISPR-Cas9 gene editing and inducible short hairpin RNA interference.

Analysis of published gene expression datasets demonstrated that glycolytic, citric acid cycle (CAC), and oxidative phosphorylation genes are upregulated in FLT-3^mut compared to FLT3 wild-type (FLT3^wt) patient samples at diagnosis. We then confirmed that both human and murine FLT-3^mut cells display increased glycolytic and respiratory capacity compared to FLT-3^wt cells. Upon treatment with the highly selective FLT-3 TKI AC220 (quizartinib), currently used in a number of clinical trials, these metabolic phenotypes were partially reversed. However, whilst glucose uptake was reduced upon FLT3 TK inhibition, glutamine uptake was not affected. Metabolic flux analysis using [U-¹³C]glutamine demonstrated that glutamine, while providing carbons for the CAC, was primarily used to support production of the major intracellular antioxidant glutathione upon AC220 treatment. This antioxidant function is necessary because, as expected, FLT-3^mut cells displayed a large increase in ROS levels following TKI treatment when grown in the absence of glutamine and these changes correlated with a significant reduction in viability in the same conditions. Glutamine dependency of FLT3^mut cells upon FLT3 TKI treatment was independently validated via a CRISPR-Cas9 drop-out genome-wide screen as glutaminase (GLS), the first enzyme in  glutamine catabolism, and several CAC enzymes were shown to be synthetically lethal with AC220. We went on to show that the combination of AC220 with a specific clinical grade GLS inhibitor (CB-839) or GLS gene silencing resulted in a significant reduction in viability and increase in ROS levels which could be rescued by supplementation of the media with the antioxidant N-acetylcysteine or a cell-permeable form of the CAC intermediate α-ketoglutarate.

Our data suggest that upon AC220 treatment, glutamine metabolism becomes a critical metabolic dependency in FLT3^mut AML. Glutamine metabolism is mostly channelled towards glutathione production, while also supporting the CAC and both these fates contribute to its protective effects following FLT3 TK inhibition by respectively counteracting oxidative damage and sustaining macromolecule biosynthesis and cellular energetics. These data predict that a combined inhibition of glutamine metabolism and FLT3 TK activity may improve the eradication of FLT3^mut AML cells.