Background:
Cerebral metabolic alterations have been described in MS and are associated with worsened clinical condition1-3. Dysregulation of glucose metabolism may lead to increased glutamate levels, resulting in excitotoxicity and neuronal damage4-7. Importantly, the brain is also capable of utilizing ketones (e.g., β-hydroxybutyrate (βHB)) as alternative fuel when glucose becomes scarce. However, how consumption of glucose and ketone bodies changes in MS is unknown, therefore there is a need for validating novel quantitative biomarkers to improve our understanding of MS pathogenesis and aid the development of novel therapies. Deuterium magnetic resonance spectroscopy (DMRS)8-10 is a new, quantitative approach that enables direct visualization of the uptake and metabolic fate of deuterium (2H)-labeled substrates. DMRS has been shown to map glucose metabolism (i.e., production of lactate and glutamate/glutamine (Glx)) in preclinical models and in humans using [6,6’-2H2]glucose11-13, and ketone metabolism using [3,4,4,4-2H4]βHB14,15, enabling the measurement of βHB conversion into Glx (Fig.1A). To date, neither of these two imaging strategies have been applied to MS models.
Hypothesis/Objective:
We hypothesize that DMRS of [6,6’-2H2]glucose and [3,4,4,4-2H4]βHB can detect metabolic modulations associated with lesion pathogenesis in an MS mouse model.
Methods:
C57Bl/6J mice received regular chow (control, n=12) or chow with 0.2% cuprizone (CPZ, n=14) for 4 weeks to induce demyelination and inflammation. T2-weighted (T2w) MRI data were acquired at 9.4T to confirm brain lesions. For 2H-Glc DMRS studies, localized DMRS scans were acquired from the brain at 11.7T prior (5 scans, 20 min) and following (18 scans, 90 min) a subcutaneous bolus injection of [6,6’-2H2]glucose (2.5g/kg). A separate group of mice was used for 2H-βHB DMRS studies. Similarly, DMRS scans were acquired at 9.4T prior and following [3,4,4,4-2H4]βHB sc bolus injection (3.25g/kg). Experimental overview is shown in Persuasive data. Dynamic DMRS spectra were summed over time and HOD (semi heavy water), Glc, Lac, Glx and βHB concentrations were calculated. Demyelination (MBP) and microglia/macrophages (Iba-1) levels were evaluated using immunofluorescence and glutamate concentration using a colorimetric assay. Statistical analysis is described in figure legends.
Results:
T2w images showed higher normalized T2-w image intensities in the corpus callosum of CPZ mice, confirming demyelination and inflammation (Fig.1B).
To investigate alterations in glycolysis and mitochondrial function we injected 2H-Glc and followed its conversion into Glx and lactate in control and CPZ mice. The summed spectra showed an increase in Glx in CPZ mice (Fig.1C), which was also observable in the time course analysis (Persuasive data). Quantitative analyses revealed no changes in HOD, Glc and lactate levels, but increased Glx concentration (+38%, p=0.0047), increased Glx/Glc (+49%, p=0.0189) and decreased Lactate/Glx (-29%, p=0.0076) in CPZ mice (Fig.1D).
Next, to investigate changes in ketone metabolism and mitochondrial function we injected 2H-βHB and followed its conversion into Glx in control and CPZ mice. The summed spectra showed a clear increase in Glx (Fig.1E). Quantitative analyses confirmed increased Glx concentration (+61%, p=0.0342) and increased Glx/βHB (+47%, p=0.0044) in CPZ mice, while HOD and βHB remained unchanged (Fig.1F).
Ex vivo tissue analyses confirmed reduction in myelin (MBP) and increase in immune cells (microglia/macrophage, Iba-1) levels thorough the brain of CPZ mice (corpus callosum, cortex and hippocampus, Fig.1G,H). Ex vivo photometric glutamate quantification confirmed increased glutamate levels in CPZ mice (Persuasive data).
Conclusion:
This is the first report of DMRS demonstrating brain metabolic alterations in a MS model. Specifically, we showed that both the ketone and glycolytic metabolic pathways feeding into the tricarboxylic acid (TCA) cycle detected increased Glx production, enabling the monitoring of mitochondrial impairment in MS. Because DMRS is clinically available, our findings provide a strong rationale for investigating its potential in MS patients, and other neurological disorders that present metabolic impairment.
Image/Figure:
Click to view full size
Click to view full size
Image/Figure Caption:
Figure 1: (A) [6,6’-2H2]glucose and [3,4,4,4-2H4]β-hydroxybutyrate cellular metabolism pathways. (B) T2w images from a control (left) and CPZ (right) mouse with an arrow indicating the corpus callosum splenium (CCS) and a CPZ demyelinated lesion. T2w image intensities of the CCS normalized to ventricular image intensities. (C) Representative summed [6,6-2H2]glucose spectra of a control and CPZ mouse showing resonances for HOD (4.7 ppm), Glc (3.8 ppm), Glx (2.3 ppm) and Lac (1.3 ppm). Representative spectra were normalized between each other based on their HOD amplitude. (D) Average concentration (mM) of HOD, Glc, Glx, and ratios of Glx/Glc, Lac/Glc, and Lac/Glx. (E) Representative summed [3,4,4,4-2H4]β-hydroxybutyrate spectra of a control and CPZ mouse showing chemical shift peaks for HOD (4.7 ppm), Glx (2.3 ppm), and βHB (1.2 ppm). Representative spectra were normalized between each other based on their HOD amplitude. (F) Average concentration (mM) of HOD, Glx, and βHB and ratio of Glx/βHB. (G) Left: Schematic overview of regions of interest. Right: Representative Iba-1 and MBP immunofluorescence images for a control and CPZ mouse. (H) Top: Graphs of average Iba-1 coverage (% area) in Cortex, Corpus callosum, and Hippocampus. Bottom: Graphs of MBP coverage (%). Error bars = SD, Unpaired t-test, * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.
Author
Nicholas David Vidas-Guscic
Washington University in St. Louis