More than 70% of human low grade gliomas (WHO grades II and III) and secondary glioblastomas (WHO grade IV) carry the heterozygous R132H-mutation that causes a neomorphic change in the isocitrate dehydrogenase1 (IDH1) enzyme¹, resulting in modulated metabolic pathways. Wildtype IDH1 is responsible for the conversion of isocitrate into α-ketoglutarate (αKG), whereas IDH1^R132H has lost its isocitrate binding capacity, and converts αKG into the oncometabolite D-2-hydroxyglutarate (2HG) instead². This results in depletion of αKG, a metabolite which is essential for synthesis of fatty acids in peroxisomes³. To obtain a more comprehensive understanding of IDH1-mutant glioma metabolism it is important to study the anaplerotic pathways that rescue αKG shortage in more detail. Since glutamate (Glu) and glutamine (Gln) are ubiquitously present in the brain as part of the neurotransmitter cycle, the role of the glutamatergic anaplerotic pathway is of special interest.

The goal of this study was to investigate the role of glutamate in 2HG production next to those of glutamine and glucose (Glc). Using ¹³C NMR we show that the glutamatergic pathway is the major carbon source for the production of 2HG in IDH1^wt/R132H HCT116 cells and that glutamate can replace glutamine as a carbon source.

HCT116 cells with a heterozygous IDH1^wt/R132H mutation (Horizon Discovery, Cambridge UK) were grown in standard DMEM (+Glc, +Gln, +10% FCS). 17 hours prior to cell extraction the medium was replaced by fresh glucose- and glutamine-free medium supplemented with 4 mM [1-¹³C]-Gln and 5.5 mM [1,6-¹³C₂]-Glc to determine fractional enrichments; and with 0 mM Gln, 5.5 mM Glc and 4 mM [1-¹³C]-Glu to examine the incorporation of extracellular Glu into 2HG. Cells were placed on ice and washed with cold PBS twice, immediately followed by incubation with 2.5 ml ice-cold MeOH, containing 280 mM formic acid per 150 cm² flask. After 10 minutes of incubation the cell extracts were removed from the flask with a rubber policeman. Subsequently the extract was centrifuged 5 minutes at 1200 x g. The protein content of the pellet, as determined via BCA assays, was used to normalize data for cell number. The supernatant was dried in a Savant SpeedVac evaporator and redissolved in 400 µl D₂O for analysis on a Bruker Avance III 500MHz spectrometer. ¹H and ¹³C NMR spectra were acquired with pulse-acquire experiments (respectively TR = 18s, 90° flip angle, NS = 16; and TR = 5.1s, 30° flip angle, NS = 7000, proton-decoupled) and analyzed with Bruker Topspin software. Integrated peak intensities were corrected for T₁ saturation, number of contributing spins, cell number, and reference compound concentration. Total 2HG pools were determined with LC-MS³. To exclude that culture conditions caused cell death, a five-day cell viability SRB assay was performed.

We could discriminate Glc- from Gln-derived 2HG due to the fact that ¹³C from Glc first flows to αKG through the tricarboxylic acid cycle and ultimately ends up at positions 4, 3 and 2, while [1-¹³C]-Gln yields [1-¹³C]-2HG (Fig.1). Based on the ratios of [2-,3-,4- ¹³C]-2HG:[1-¹³C]-2HG we calculated that 69% of the total pool of 2HG was Gln-derived (2.4 fmol/cell) compared to 16% that was Glc-derived. Glu fractional enrichments were 25% Gln-derived and 8% Glc-derived (Fig.2).

Cell growth in medium without Gln and Glu was significantly compromised and was proportionally restored by incrementing Glu in the medium. Normal protein levels were achieved by adding 4 mM Glu (Fig.3).

Substitution of Gln by [1-¹³C]-Glu resulted in dramatically reduced Glu and Gln (Glx) pools. Interestingly, the amount of Glu-derived [1-¹³C]-2HG did not drop significantly (Fig.4).

Results of our study show that carbons in the 2HG backbone mainly originate from the glutamatergic pathway. 2HG concentrations were comparable to reported values in literature⁴. Remarkably, the fractional enrichment of 2HG is much higher than the upstream metabolites Glu and Gln (~25%), suggesting a possible intracellular compartmentalization.

The SRB viability assay pointed out that there was no immediate cell death due to simultaneous Glu administration and Gln deprivation within the duration of our second experiment.

Despite the minimal Glu uptake from the medium, as reflected by decreased intracellular pools of Glx, Glu-derived 2HG production was not significantly decreased if compared to the experiments with labeled glutamine. This suggests that under standard culture conditions the total Glx levels are not rate limiting in the 2HG production pathway. Other mechanisms, like IDH1-inhibiting effects by 2HG itself, might play a more important regulating role than the availability of glutamatergic precursors⁵. The capacity of Glx uptake for αKG replenishment might be a reason for the typical diffusive growth in environments with high Glx levels.