Figures and data
Theoretical background and metabolic basis of POPE13C-MRS
a Following administration of uniformly labelled 13C-glucose ([U-13C₆]Glc), 13C label is incorporated into lactate, glutamate, glutamine, and GABA through glycolysis, mitochondrial metabolism, and neurotransmitter cycling. Exchange between metabolic pools and compartments shapes the time course of 13C labelling. b POPE13C-MRS detects this incorporation indirectly via changes in 1H signals bound to 12C and the appearance of 1H-13C satellite resonances, enabling measurement of neurotransmitter-specific labelling. This approach provides a targeted readout of excitatory (glutamate) and inhibitory (GABA) metabolic pathway in vivo. [U-13C₆]Glc labels lactate via exchange with pyruvate generated by glycolysis, while glutamate and glutamine are labelled through exchange with tricarboxylic acid (TCA) cycle intermediates. GABA becomes labelled through the decarboxylation of glutamate. POPE13C-MRS detects protons (¹H) bound to 12C (or 13C) in specific positions in these metabolites, including glutamate and glutamine H2 (GlxH2, purple), glutamate H4 (GluH4, red), GABA H4 (GABAH4, blue), and lactate H3 (LacH3, green). As protons become bound to 13C, their 1H signal therefore decreases, producing a measurable signal loss. For some resonances (e.g., GlxH2 and LacH3 here), this loss is accompanied by the appearance of characteristic 13C-coupled satellite doublets flanking the parent resonance (see purple and green coloured boxes). Due to their lower signal and for the sake of simplicity, satellites of GABAH4 and GluH4 are not represented here. Carbon numbers in grey are not detectable with MRS due to the small metabolite pool size. IN, inhibitory neuron; EN, excitatory neuron; TCA, tricarboxylic acid cycle; Glc, glucose, Glu, glutamate; Gln, glutamine; Lac, lactate; Pyr, pyruvate; OA, oxaloacetate; aKG, alpha-ketoglutarate; AcCoA, acetyl-coenzyme-A; NAA, N-acetyl-aspartate; bHB, beta-hydroxybutyrate
Time-resolved detection of neurotransmitter labelling following 13C-glucose administration
a Acquisition of POPE13C-MRS in mouse brain (in purple, voxel size: 6.5 x 3.4 x 3.2 mm3) following a ∼2h30 min s.c. administration of uniformly 13C-labelled glucose ([U-13C6]-glucose). b Progressive attenuation of 1H-12C resonances reflects incorporation of 13C label into downstream metabolites, consistent with metabolic flux through glutamatergic and GABAergic pathways. GABA- and lactate-POPE13C-MRS acquired from baseline scan (t0 = start of the infusion) until the end of the experiment (tf = 2h30). c Fitting of POPE13C-MRS signals tracks neurotransmitter-specific metabolic labelling in vivo (averaged spectra acquired between 70- and 140-min post infusion). Peaks of interest include glutamate+glutamine-H2 (Glx) in purple and glutamate-H4 (Glu) in red, gamma-aminobutyric acid-H4 (GABA) in blue and lactate-H3 (Lac) in green. Spectra are all shown with a 5Hz Lorentzian apodization.
Validation of indirect detection of 13C-labelled metabolites
a Both attenuation of 1H-12C resonances and emergence of 1H-13C satellite signals provide complementary and internally consistent measures of metabolite labelling. b GABA-POPE13C-MRS can indirectly detect 13C labelling of GluC4 (via a drop in the 12C-GluH4 signal), GABAC4 (via a drop in the12C-GABAH4 signal) and of the total pool of glutamate and glutamine (via both the drop in 12C-GlxH2 signal and increase in satellite 13C-GlxH2 signals). Red arrows indicate the directionality of that change. c Lactate-POPE13C-MRS can detect indirect 13C labelling of lactate-C3 via both the drop in 12C-LacH3 signal and increase in satellite 13C-LacH3 signals (not detected here, red crosses). Experiments were performed in mice under medetomidine anaesthesia, making 13C-LacH3 signals undetectable. Spectra are averages of 5 adult mice after 1h [U-13C6]-glucose infusion and shown with 1 Hz Lorentzian apodization.
Feasibility of POPE13C-MRS in the human brain
a Timeline of the intravenous [U-13C6]-glucose infusion and POPE13C-MRS baseline (t0) and post-infusion (tf) acquisitions. b Voxel positioning in dorsal anterior cingulate cortex (dACC) for POPE13C-MRS acquisition. c Overlaid POPE13C-MRS before (black) and after (blue) the [U-13C6]-glucose infusion. d,e Detectable 13C-labelled GlxH2 satellite resonances and reduction in 1H-12C signals for GABA, Glu and Glx were observed following 13C-glucose administration, consistent with incorporation into neurotransmitter metabolic pools. Average GABA-POPE13C-MRS (d) and lactate-POPE13C-MRS (e) spectra before (baseline scan, t0) and after (post infusion scan, tf) the [U-13C6]-glucose administration. Glx = glutamate + glutamine; NAA, N-acetyl-aspartate; βHB, hydroxybutyrate; Mac/Lipid, macromolecules or lipids.
Consistent detection of POPE13C-MRS readouts across mice
Quantification results from POPE13C-MRS in adult male mice (N=5) from Study 2. Metabolite quantifications are reported for baseline scans (t0) or after the [U-13C6]-glucose infusion (tf) using either water reference (given in millimolar, mM) or using the isotopic fractional enrichment of GlxH2 (GlxFE) as internal reference (given as arbitrary units, a.u.). Coefficient of variation (CV) represent the inter-individual variability of the quantification.
Assessment of metabolite labelling dynamics and ratio stability in human a,b
Simulation of arterial input function (AIF) based on our 13C-glucose infusion protocol (g) and estimation of the resulting 13C fractional enrichment (FE) of brain metabolites (h) in human POPE13C-MRS. c Effective 13C FE of metabolites measured with POPE13C in human dACC for different number of averages and SNR, i.e. (left) five individual acquisitions (SNRaverage=9.7), (middle) two averages (SNRaverage=9.8), and (right) a single average (SNR=11.6). GlxH2 FE was measured based on either the drop of basal GlxH2 peak from t0 (pink) or based on the GlxH2 satellite resonances (purple).
POPE13C-MRS quantifications in the human
Quantification results from POPE13C-MRS in human in Study 3 (N=1). Metabolite quantifications are reported for baseline scans (t0) or after the [U-13C6]-glucose infusion (tf) using either water reference (given in millimolar, mM) or using the isotopic fractional enrichment of GlxH2 (GlxFE) as internal reference (given as arbitrary units, a.u.). Coefficient of variation (CV) represent the variability of the quantification within the ∼45min acquisition.
