4302

Lactate production but not pool-size is altered in infiltrative mouse model of human GBM: a 1H MRS and hyperpolarized [2H7,13C6]D-glucose study
Mor Mishkovsky1, Olga Gusyatiner2, Bernard Lanz1, Irene Vassallo2, Marie‐France Hamou2, Rolf Gruetter1,3,4,5, and Monika Hegi2

1Laboratory of Functional and Metabolic Imaging, Ecole Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland, 2Department of Clinical Neuroscience, Lausanne University Hospital, Lausanne, Switzerland, 3Department of Radiology, University of Geneva (UNIGE), Geneva, Switzerland, 4Department of Radiology, University of Lausanne (UNIL), Lausanne, Switzerland, 5Centre d'Imagerie Biomédicale (CIBM), École polytechnique fédérale de Lausanne (EPFL), Lausanne, Switzerland

Synopsis

Glioblastoma (GBM) is the most malignant primary brain tumor in adults. Aberrant glucose metabolism is considered a hallmark of cancer, via the so called ‘Warburg Effect’. MRS gives the access to study tumor metabolism. 1H MRS enables to quantify the steady state pool size and 13C MRS of hyperpolarized (HP) endogenous compounds provides real time metabolic information which is related to enzymatic activity. The aim of the present study was to examine whether changes in lactate production through glycolysis can be characterized using HP [2H7,13C6]D-glucose MRS, and if those correspond to changes in lactate pool size.

Introduction

Glioblastoma (GBM) is the most malignant primary brain tumor in adults. It exhibits high metabolic activity and is notorious for its resistance to multi-modal therapy, with a median survival of only 15 months. Hyperactive glucose metabolism is considered a hallmark of cancer, via the ‘Warburg Effect’ manifested by the switch of glucose metabolism and ATP production from oxidative phosphorylation to glycolysis, however recent ex vivo and in vivo studies show evidences for active glucose oxidation in human GBM1-3. Direct detection of tumor glycolysis can provide new evidences on this debate.

MRS gives the access to study tumor metabolism. 1H MRS enables to quantify the steady state pool size of about 20 metabolites4. 13C MRS of hyperpolarized (HP) endogenous compounds, using dissolution dynamic nuclear polarization (dDNP)5, provides real time metabolic information which is related to enzymatic activity6. It was recently reported that de novo synthesis of [1-13C]lactate can be monitored in vivo following the infusion of HP [2H7,13C6]D-glucose in the mouse brain7.

In the present study we addressed the question whether changes in lactate production in GBM can be characterized through glycolysis using HP [2H7,13C6]D-glucose MRS, and whether they correspond to deviations in lactate pool size as determined. The experiments were performed using an invasive patient-derived orthotopic xenograft GBM model, and taking advantage of the fact that metabolite ratios in HP 13C MRS are related to the flux through the up stream biochemical reactions8.

Methods

GBM mouse models: LN-3708GS spheroids (105 cells)9 were stereotactically injected into the left hemisphere of immunodeficient male mice (NSG). The control group was injected with 5 μL of the cell suspension solution solely ( n = 5 per group, 8 weeks of age at implantation, 20 ± 1 weeks at day of dDNP experiment).

Multimodal MR analysis was performed in a 9.4 T/31 cm actively shielded animal scanner (Varian/Magnex) and included the following steps: 1) To define structural changes T2 weighted (T2W) images were acquired. 2) To characterize the neurochemical profile of the tumor, single voxel 1H MRS measurements were carried out in the injected hemisphere. 3) Frozen droplets of a water solution containing [2H7,13C6]D-glucose (3M) and OX63 trityl radical (25 mM) were dynamically polarized in a custom designed 7T/1K DNP polarizer as previously described7,10. To monitor real-time de novo synthesis of [1-13C]lactate, a 540 μL of 56±13 mM HP [2H7,13C6]D-glucose was injected through the femoral vein. A series of pulse-acquire sequence was triggered 5.5s post injection with 20° frequency selective Gaussian pulse (250μs) centered at 182 ppm every 0.5s for 50 s. Lactate-to-glucose ratio (LGR) was calculated from the summed spectra. To minimize variation between individual animals, the ratio was corrected for the dose of 13C glucose at the time of injection, i.e. multiplying by the moles of [2H7,13C6]D-glucose injected, divided by the animal blood volume (cLGR). 4). The integrity of the blood brain barrier (BBB) was assessed by T1 weighted (T1W) images acquired after injection of gadolinium-contrast agent (Gadovist® 5uL/gr body weight),

Results

Implantation of patient-derived LN-3708GS spheroids gave rise to highly diffusive tumors that spread over both brain hemispheres (Fig.1). 1H spectra demonstrated highly distinct metabolic profiles of the tumors as compared to control (Fig.2). In the summed spectra after HP [2H7,13C6]D-glucose injection it was readily observed that the [1-13C]lactate signal in the tumorous brain was smaller than the one in the control (Fig.3). A comparison between lactate production (cLGR) and lactate pool size revealed a significant decrease of the cLGR in the xenografts as compared to the control brains, while the pool size of lactate remained the same in both (Fig.4).

Discussion

We report real time de novo synthesis of [1-13C]lactate following the infusion of HP [2H7,13C6]D-glucose in a diffuse GBM xenograft model. The production of [1-13C]lactate from [2H7,13C6]D-glucose in our experiment is a consequence of 12 enzymatic steps including: glucose transport, 10 enzymatic steps of glycolysis and pyruvate conversion to lactate by lactate-dehydrogenase (LDH). The rate of lactate production is related to the cLGR and is significantly slower in LN-3708GS xenograft bearing brains than in controls, while the pool-size of lactate was similar. This discrepancy between cLGR and lactate pool-size suggests that the HP [2H7,13C6]D-glucose experiment rather detects dynamic lactate production than endogenous lactate pool-size. The similarity of the lactate pool size in both tumor and control suggests that energy production in the tumor is most probably not supplied through ATP production via LDH-mediated lactate production.

Acknowledgements

We thank Drs.Cristina Cudalbu and Arnaud Comment for their assistance putting together the MRS protocols. This work was supported by CIBM of the UNIL, UNIGE, HUG, CHUV, EPFL, the Leenaards and Jeantet Foundations.

References

1. Marin-Valencia, I. et al. Analysis of tumor metabolism reveals mitochondrial glucose oxidation in genetically diverse human glioblastomas in the mouse brain in vivo. Cell metabolism 15, 827-837, doi:10.1016/j.cmet.2012.05.001 (2012).

2 Maher, E. A. et al. Metabolism of [U-13 C]glucose in human brain tumors in vivo. NMR in biomedicine 25, 1234-1244, doi:10.1002/nbm.2794 (2012).

3. Lai, M. et al. In vivo characterization of brain metabolism by (1) H MRS, (13) C MRS and (18) FDG PET reveals significant glucose oxidation of invasively growing glioma cells. International journal of cancer 143, 127-138, doi:10.1002/ijc.31299 (2018).

4. Duarte, J. M., Lei, H., Mlynarik, V. & Gruetter, R. The neurochemical profile quantified by in vivo 1H NMR spectroscopy. NeuroImage 61, 342-362, doi:10.1016/j.neuroimage.2011.12.038 (2012).

5. Ardenkjaer-Larsen, J. H. et al. Increase in signal-to-noise ratio of > 10,000 times in liquid-state NMR. Proceedings of the National Academy of Sciences of the United States of America 100, 10158-10163, doi:10.1073/pnas.1733835100 (2003).

6. Comment, A. & Merritt, M. E. Hyperpolarized magnetic resonance as a sensitive detector of metabolic function. Biochemistry 53, 7333-7357 (2014).

7. Mishkovsky, M. et al. Measuring glucose cerebral metabolism in the healthy mouse using hyperpolarized (13)C magnetic resonance. Scientific reports 7, 11719, doi:10.1038/s41598-017-12086-z (2017).

8. Hill, D. K. et al. Model free approach to kinetic analysis of real-time hyperpolarized 13C magnetic resonance spectroscopy data. PloS one 8, e71996, doi:10.1371/journal.pone.0071996 (2013).

9. Sciuscio, D. et al. Extent and patterns of MGMT promoter methylation in glioblastoma- and respective glioblastoma-derived spheres. Clinical cancer research : an official journal of the American Association for Cancer Research 17, 255-266, doi:10.1158/1078-0432.CCR-10-1931 (2011).

10. Cheng, T. et al. Automated transfer and injection of hyperpolarized molecules with polarization measurement prior to in vivo NMR. NMR in biomedicine 26, 1582-1588, doi:10.1002/nbm.2993 (2013).

11.Mlynarik, V., Gambarota, G., Frenkel, H. & Gruetter, R. Localized short-echo-time proton MR spectroscopy with full signal-intensity acquisition. Magn Reson Med 56, 965-970 (2006).

12. Provencher, S. W. Estimation of metabolite concentrations from localized in vivo proton NMR spectra. Magnetic resonance in medicine 30, 672-679 (1993).

Figures

Figure 1. Typical T2W image (FSEMS, FOV 18x18 mm2, TR = 4000 ms, effective TE = 52 ms, 6 averages)., T1W images post contrast agent (FSEMS, FOV 18x18 mm2, TR = 350 ms, effective TE = 11 ms, 6 averages) and histological sections emphasizing the infiltrative growth of LN-3708GS xenografts and the lack of any striking morphological modifications and focal mass, and absence of enhancement after gadolinium injection. Tumor cells (brown dots in D) display invasion of the contralateral hemisphere.

Figure 2.1H MRS spectra acquired from the voxel designated by the square and the corresponding peak assignment (SPECIAL11, TR/TE = 4000/2.8 ms, VOI = 2x2x2mm3 in 15x16 averages). The spectra detected in control (lower panel, black) and tumor (upper panel, red) show very different metabolic profiles. Metabolite concentrations were quantified using LCModel-based fitting routine12. Significant differences (p < 0.05, 1-way ANOVA OriginPro) in the average metabolite concentrations between LN-3708GS (n = 5) and control (n = 5) groups, and the direction of change is designated by the arrow.

Figure 3. Characteristic summed spectra measured in the tumor (red) and control (black) mouse brains following infusion of HP [2H7,13C6]D-glucose form the area under the RF surface coil (white -1H, red 13C surface coil). Spectra are normalized to their maximal glucose signal respectively. In both acquisitions [1-13C]D-glucose-β (96.8 ppm), [1-13C]D-glucose-α (93 ppm) and [1-13C]lactate (183.5 ppm) can be detected. In the control mouse [1-13C]pyruvate signal is detected at 172 ppm. The broad peak at 175 ppm designated by a star is an impurity in the [2H7,13C6]D-glucose powder.

Figure 4. Comparison of cLGR and lactate total concentration between the tumor and control groups. Statistical analysis was performed using 1-way ANOVA in OriginPro software. Significant differences were found between cLGR ratios while the lactate pool size remains the same in both groups.

Proc. Intl. Soc. Mag. Reson. Med. 27 (2019)
4302