Introduction

Elevated levels of phosphocholine (PC) and phosphoethanolamine (PE) have emerged as MR-detectable biomarkers of malignant transformation^1,2,3. Interestingly, however, PC and PE levels are reduced in low-grade gliomas carrying the isocitrate dehydrogenase 1 (IDH1) mutation^4,5,6. The goal of the current study was, therefore, to understand the mechanism underlying reduced PC and PE levels in mutant IDH1 gliomas. Using a combination of ¹³C- and ³¹P-MRS and biological assays, we measured the activities of choline kinase (CK) and ethanolamine kinase (EK) as well as CTP:PC cytidylyltransferase (CCT) and CTP:PE cytidylyltransferase (ECT), the enzymes directly involved in PC and PE production and consumption respectively⁷. Our results indicate that the activities of CK, EK, CCT and ECT are down-regulated in mutant IDH1 cells. However, the down-regulation of CK/EK is greater than that of CCT/ECT resulting in an overall reduction in steady-state levels of PC and PE. Our study thus delineates the mechanism underlying the unusual down-regulation of PC and PE levels in mutant IDH1 gliomas.

Methods

U87 and NHA cells expressing IDH1 wild-type (U87IDHwt/NHAIDHwt) or mutant (U87IDHmut/NHAIDHmut) enzyme and orthotopic tumor xenografts were generated as described^6,8. ¹³C-MRS studies to quantify flux from [1,2-¹³C]-choline and [1,2-¹³C]-ethanolamine to PC and PE were performed on a 600-MHz Bruker spectrometer using a 10 mm broadband probe and an MR-compatible bioreactor system⁹. ³¹P-MR-based assays to measure CK, EK, CCT and ECT activity were performed on a 500-MHz Bruker spectrometer with a triple resonance cryoprobe^10,11. Protein expression was determined by immunoblotting. Experiments were performed in triplicate (n=3) and statistical significance assessed using an unpaired t-test (p<0.05 was considered significant).

Results and Discussion

Using ³¹P-MRS (Fig.1A), we first confirmed that PC (Fig.1B&1C) and PE (Fig.1D&1E) levels were significantly reduced in mutant IDH1 cells relative to wild-type. We then used ¹³C-MRS to monitor ¹³C-PC and -PE synthesis in live cells perfused with [1,2-¹³C]-choline and [1,2-¹³C]-ethanolamine over 48h (Fig.2A). The kinetic fit for PC (Fig.2B&2C) and PE (Fig.2D&2E) build-up indicated that the pseudo-first-order rate constant for CK was significantly reduced by 39.6% from 0.1±0.008h-1 in U87IDHwt to 0.06±0.007h-1 in U87IDHmut and by 40.4% from 0.08±0.007h-1 in NHAIDHwt to 0.05±0.004h-1 in NHAIDHmut. Similarly, the pseudo-first-order rate constant for EK was significantly reduced by 39.8% from 0.08±0.005h-1 in U87IDHwt to 0.05±0.004h-1 in U87IDHmut and by 50.2% from 0.06±0.005h-1 in NHAIDHwt to 0.03±0.004h-1 in NHAIDHmut. Next, we found a significant reduction in CK expression in mutant IDH1 cells (43.2±5.6% in the U87 model and 49.6±3.4% in the NHA model; Fig.3A&3B). Measurement of CK and EK activity using ³¹P-MR-based assays (Fig.3C&3D) revealed a significant reduction in CK activity in mutant IDH1 cells (59.8% from 9.24±0.79fmol/cell/h in U87IDHwt to 3.71±0.62fmol/cell/h in U87IDHmut and by 61.4% from 5.63±0.54fmol/cell/h in NHAIDHwt to 2.17±0.51fmol/cell/h in NHAIDHmut). EK activity was similarly reduced in mutant IDH1 cells (59.8% from 5.29±0.42fmol/cell/h in U87IDHwt to 2.13±0.41fmol/cell/h in U87IDHmut and 69.3% from 4.03±0.29fmol/cell/h in NHAIDHwt to 1.23±0.24fmol/cell/h in NHAIDHmut). We then measured the expression and activity of CCT and ECT, the enzymes that convert PC and PE to CDP-choline and CDP-ethanolamine respectively⁷. We found a significant reduction in CCT expression (30.1±5.4% in the U87 model and 28.9±6.3% in the NHA model; Fig.4A&4B) and ECT expression (31.3±7.9% in the U87 model and 25.3±4.1% in the NHA model; Fig.4C&4D) in mutant IDH1 cells. CCT activity (Fig.4E) was significantly reduced in mutant IDH1 cells (39.3% from 4.24±0.39fmol/cell/h in U87IDHwt to 2.57±0.21fmol/cell/h in U87IDHmut and by 40% from 2.74±0.31fmol/cell/h in NHAIDHwt to 1.65±0.29fmol/cell/h in NHAIDHmut). ECT activity (Fig.4F) was likewise reduced in mutant IDH1 cells (40% from 2.62±0.22fmol/cell/h in U87IDHwt to 1.57±0.23fmol/cell/h in U87IDHmut and by 37.2% from 1.52±0.16fmol/cell/h in NHAIDHwt to 0.96±0.14fmol/cell/h in NHAIDHmut). Taken together, these results indicated that the reduction in CK/EK activity was higher (~60%) than the reduction in CCT/ECT activity (~40%) leading to an overall reduction in PC and PE levels in mutant IDH1 cells. Finally, we confirmed our findings in U87IDHwt and U87IDHmut orthotopic tumor xenografts. Ex vivo ³¹P-MR analysis of tumor extracts indicated that PC and PE levels were reduced in U87IDHmut tumors relative to U87IDHwt (Fig.5A&5B). CK (Fig.5C), EK (Fig.5D), CCT (Fig.5E) and ECT (Fig.5F) activities were concomitantly reduced. These results confirmed our findings in genetically-engineered cells in vivo.

Conclusions

Collectively, our results mechanistically linked the reduction in PC and PE levels in mutant IDH1 gliomas to reduced CK and EK activity. Importantly, although CCT and ECT activity were reduced as well, the reduction in CK and EK activity was greater leading, on balance, to reduced PC and PE levels. Reduced PC and PE levels, therefore, constitute unique metabolic biomarkers and potential therapeutic opportunities in mutant IDH1 gliomas.

Acknowledgements

Grant Acknowledgements: NIH R01CA172845, NIH R01CA197254, NIH R01CA154915, NIH P41EB013598, UCSF Brain Tumor Loglio Collective and NICO.

References

1) Glunde et al., Nature Reviews Cancer, 11:835-48, 2011. 2) Zhu et al., Biochemistry & Cell Biology, 90(2): 188-199, 2012. 3) Shah et al. ISMRM Toronto, 0523, 2015. 4) Yang et al., Clinical Cancer Reseach, 18:5562-5571, 2012. 5) Esmaeili et al., Cancer Research, 74:4898-4907, 2014. 6) Izquierdo-Garcia et al., PLOS ONE, 10:e0118781, 2015. 7) Pellech & Vance, Biochimica et Biophysica Acta, 779, 217-251, 1984. 8) Chaumeil et al. Nature Communications, 4: 2429-2441, 2013. 9) Ronen et al., Biochimica et Biophysica Acta, 1095:5-16, 1991. 10) Gabellieri et al., NMR in Biomedicine, 22: 456-461, 2009. 11) Ward et al., PLOS ONE, 8:e62610, 2013.