Elevated phosphocholine (PC) levels and increased expression of choline kinase, the enzyme catalyzing PC synthesis from choline, have been reported in most types of cancer and are often considered hallmarks of the disease¹. Interestingly however, in cells that express the isocitrate dehydrogenase 1 (IDH1) mutation – a recently discovered driver mutation in low-grade gliomas² – we and others have found that PC levels are reduced when compared to IDH1 wild-type cells^3,4,5. The goal of the current study was therefore to understand the underlying mechanism for the drop in PC in IDH1 mutant cells. We used ¹³C-magnetic resonance spectroscopy (MRS) to monitor the metabolism of [1,2-¹³C]-choline and quantify choline flux to PC in two genetically engineered glioma models expressing wild-type or mutant IDH1. We also investigated the expression of choline kinase and the ubiquitous choline transporter SLC44A1. Our results indicate that PC synthesis is reduced in IDH1 mutant cells in both of our models, and that this drop is associated with reduced expression of choline kinase α, pointing to a unique pattern of metabolic reprogramming that is associated with the IDH1 mutation. U87 and NHA cell lines expressing IDH1 wild-type (U87IDHwt/NHAIDHwt) or mutant (U87IDHmut/NHAIDHmut) enzyme were generated as described³. Cells were cultured in DMEM with 10% fetal bovine serum. MRS studies were performed on a 600-MHz Bruker Avance spectrometer using a broadband probe and an MR-compatible cell perfusion (bioreactor) system⁶. Cells were perfused for 48h with DMEM in which choline was replaced with [1,2-¹³C]-choline (56μM). Proton-decoupled ³¹P spectra were acquired using a 30° pulse, 3s relaxation delay and 1024 scans to confirm cell viability and quantify PC levels. Proton-decoupled ¹³C spectra were acquired in 2h blocks using a 60° pulse, 6s relaxation delay and 2400 scans. Peak integrals were quantified using Mnova, corrected for saturation and NOE and normalized to cell number and to a metabolite of known concentration (1.59mM inorganic phosphate for ³¹P spectra and 5mM 1-¹³C-glucose for 13C spectra). Assuming that PC is generated primarily by choline phosphorylation, the kinetics of PC build-up were expected to follow the equation ^13CPC(t)=A (1-e^-kt) where ^13CPC represents ¹³C-labeled PC at time point t, A represents the final ¹³C-labeled pool of PC and k is the pseudo-first-order rate constant for choline kinase⁷. GraphPad Prism was used to fit the build-up of the ¹³C-labeled PC peaks as a function of time. Enzyme expression was determined using western blotting for SLC44A1 (Abcam) and for choline kinase α and β (Sigma-Aldrich) with GAPDH (Cell Signaling) as loading control. All experiments were performed in triplicate (n=3) unless otherwise mentioned and significance assessed using an unpaired Student’s t-test with p<0.05 considered significant. ³¹P-MRS data (Fig. 1) was used to quantify total PC levels (1.4±0.3 fmol/cell in NHAIDHwt, 0.5±0.2 fmol/cell in NHAIDHmut, 1.9±0.3 fmol/cell in U87IDHwt and 1.1±0.2 fmol/cell in U87IDHmut) confirming our previous findings that PC levels were reduced in IDH1 mutant cells. ¹³C-MRS of live cells perfused with [1,2-¹³C]-choline showed simultaneous build-up of two ¹³C-labeled PC doublets, corresponding to [1-¹³C]-PC and [2-¹³C]-PC, over the course of 48h (Fig. 2A&2B). As expected, the concentration of [1,2-¹³C]-choline (56μM) was below detection. Fig. 3A illustrates the kinetic fit to ¹³C-labeled PC build-up in the NHA model, and the goodness of fit (R²=0.94 for NHAIDHwt and R²=0.87 for NHAIDHmut) confirmed that PC was generated primarily from exogenous ¹³C-labeled choline. This was further confirmed by comparison of the steady-state ¹³C-labeled PC pool (determined from the kinetic fit) to the total PC pool (determined from the ³¹P spectra), which indicated that the two were within experimental error (Fig. 3B&3C). Furthermore, the kinetic data indicated that the pseudo-first-order rate constant for choline kinase was significantly reduced by 49±11% (n=3, p<0.05) in NHAIDHmut cells relative to NHAIDHwt (Fig. 4A) and by 39±9% (n=2) for U87IDHmut cells relative to U87IDHwt (Fig. 4B). Western blotting showed a significant decrease in choline kinase α expression in IDH1 mutant cells compared to IDH1 wild-type (43±5%, p < 0.005 in the NHA model and 29±10%, p<0.05 in the U87 model, Fig. 5A&5B), whereas choline kinase β (Fig. 5C&5D) and choline transporter SLC44A1 (Fig. 5E&5F) were comparable. Collectively, these results mechanistically link the MRS-detectable reduction in PC levels to reduced choline-derived PC synthesis and decreased expression of choline kinase α in IDH1 mutant cells. Our MR findings and associated studies point to an unexpected metabolic reprogramming in IDH1 mutant cells. Reduced PC synthesis mediated by down-regulated choline kinase α expression provides a unique metabolic biomarker and potential therapeutic target in IDH1 mutant glioma cells.