Mapping of exogenous choline uptake and metabolism in rat glioblastoma using deuterium metabolic imaging (DMI)

Abstract

Introduction: There is a lack of robust metabolic imaging techniques that can be routinely applied to characterize lesions in patients with brain tumors. Here we explore in an animal model of glioblastoma the feasibility to detect uptake and metabolism of deuterated choline and describe the tumor-to-brain image contrast.

Methods: RG2 cells were incubated with choline and the level of intracellular choline and its metabolites measured in cell extracts using high resolution ¹H NMR. In rats with orthotopically implanted RG2 tumors deuterium metabolic imaging (DMI) was applied in vivo during, as well as 1 day after, intravenous infusion of ²H₉-choline. In parallel experiments, RG2-bearing rats were infused with [1,1',2,2'-²H₄]-choline and tissue metabolite extracts analyzed with high resolution ²H NMR to identify molecule-specific ²H-labeling in choline and its metabolites.

Results: In vitro experiments indicated high uptake and fast phosphorylation of exogenous choline in RG2 cells. In vivo DMI studies revealed a high signal from the ²H-labeled pool of choline + metabolites (total choline, ²H-tCho) in the tumor lesion but not in normal brain. Quantitative DMI-based metabolic maps of ²H-tCho showed high tumor-to-brain image contrast in maps acquired both during, and 24 h after deuterated choline infusion. High resolution ²H NMR revealed that DMI data acquired during ²H-choline infusion consists of free choline and phosphocholine, while the data acquired 24 h later represent phosphocholine and glycerophosphocholine.

Discussion: Uptake and metabolism of exogenous choline was high in RG2 tumors compared to normal brain, resulting in high tumor-to-brain image contrast on DMI-based metabolic maps. By varying the timing of DMI data acquisition relative to the start of the deuterated choline infusion, the metabolic maps can be weighted toward detection of choline uptake or choline metabolism. These proof-of-principle experiments highlight the potential of using deuterated choline combined with DMI to metabolically characterize brain tumors.

Keywords: cancer; choline; deuterium; glioblastoma; metabolic imaging.

FIGURE 1

In vitro Cho uptake. (A) ¹H NMR spectra of GL261 cell metabolite extracts at 0, 10, 60, and 120 min, zoomed in on the spectral region of the methyl group of choline-containing molecules. (B) Intracellular levels, normalized to creatine concentration, in RG2 cells (n = 3), and (C) GL261 cells (n = 3), during 120 min incubation with 1 mM unlabeled Cho. Cho, choline; PC, phosphocholine; GPC, glycerophosphocholine. Error bars = SD.

FIGURE 2

Cho infusion. In black: 3-step infusion protocol graphically shown for an animal with body weight of 250 g (left Y-axis). In red: plasma Cho concentration (mM) measured in one animal using the infusion protocol shown in black.

FIGURE 3

In vivo DMI during infusion of ²H₉-Cho. (A) Coronal slice of CE T1W MRI from brain of RG2-bearing rat, showing the tumor lesion and individual voxel positions selected from the ²H MRSI grid. (B) Individual ²H spectra from voxel positions indicated on MRI shown in panel (A). (C) Color-coded ²H MRSI grid overlaid on anatomical MRI. (D) Interpolated color-coded map based on data shown in panel (C). Color scale applies to both C and F. tCho, total choline; HDO, natural abundant ²H-labeled H₂O.

FIGURE 4

In vivo DMI 24 h after infusion of ²H₉-Cho. (A,C) CE T1W MRI of same RG2-bearing rat acquired before 36 min infusion of ²H₉-Cho (A) and 24 h later (C). (B,D) Interpolated color-coded maps of ²H-tCho based on DMI data acquired during (B), and 24 h after (D) infusion of ²H₉-Cho, in the same animal.

FIGURE 5

Tumor ²H-tCho concentration measured in vivo. (A) Average tumor ²H-tCho levels (mM) from DMI data acquired during the 36 min infusion of ²H₉-Cho, and 20–24 h later. Bar graphs indicate mean and standard deviation, with individual data points overlaid. 36 min: n = 8; 24 h: (n = 9). Statistics based on T-test. (B) Subset of data shown in panel (A) of animals (n = 5) for which a pair of 36 in and 24 h data were collected. Statistics based on Wilcoxon Rank test.

FIGURE 6

High resolution ²H NMR. ²H NMR spectra from tumor tissue metabolite extract collected from an animal euthanized immediately after 36 min infusion of ²H₉-Cho (top), and from an animal euthanized 24 h after the infusion (bottom). Cho, choline; PC, phosphocholine; GPC, glycerophosphocholine.