Caroline Guglielmetti1,2, Huihui Li3, Lydia M. Le Page1,2, Lauren Y. Shields3, Jeffrey C. Rathmell4, Ken Nakamura3, and Myriam M. Chaumeil1,2
1Department of Physical Therapy and Rehabilitation Science, University of California San Francisco, San Francisco, CA, United States, 2Department of Radiology and Biomedical Sciences, University of California San Francisco, San Francisco, CA, United States, 3Gladstone Institute of Neurological Disease, San Francisco, CA, United States, 4Vanderbilt Center for Immunobiology, Department of Pathology, Microbiology, and Immunology, Vanderbilt University Medical Center, Nashville, TN, United States
Synopsis
We used hyperpolarized
13C magnetic resonance spectroscopic imaging (HP 13C
MRSI), fluorine-18 fluorodeoxyglucose (18F-FDG) positron emission
tomography (PET) imaging and T2-weighted MRI to detect brain glucose
metabolism in mice harboring deletion of the glucose transporter 3 (GLUT3) in CA1
hippocampal neurons. GLUT3 deletion induced memory impairment in males and
females, highlighting the importance of glucose uptake by neurons. HP [1-13C]lactate-to-pyruvate
ratios and brain volumes were decreased in female GLUT3 deficient mice, but not
in males, indicating sex-specific vulnerability. No changes were detected using
18F-FDG PET imaging, highlighting the potential of HP [1-13C]pyruvate
to detect downstream alterations in brain glucose metabolism.
Introduction
The
brain requires large amount of glucose, but it remains controversial whether
neurons import it and further metabolize it through glycolysis, or if this role
is essentially ensured by glial cells1,2,3. As disruption of glucose metabolism is
an important feature of neurodegenerative disorders, it is of utmost importance
to develop and validate noninvasive biomarkers to improve our knowledge of
brain energetics in the healthy and diseased brain. In this context, animal
models presenting deletion of key transporters and enzymes can provide
mechanistic insights and well-controlled models to evaluate the potential of noninvasive
methods to monitor metabolic changes in vivo. Positron emission
tomography (PET) imaging using fluorine-18 fluorodeoxyglucose (18F-FDG)
is well established to image glucose uptake, however one main limitation lies
in its inability to inform on subsequent glycolytic steps and metabolic fluxes.
Hyperpolarized 13C magnetic resonance spectroscopic imaging (HP 13C
MRSI) has demonstrated great potential to measure metabolic fluxes in real-time
in neurological disorders4-9. Specifically, HP [1-13C]pyruvate detects [1-13C]lactate
production in the brain, thus informing on the metabolic fate of pyruvate, the
end of product of glycolysis. Here, we used mice presenting a deletion of the
glucose transporter 3 (GLUT3) in neurons of the CA1 hippocampal region, a key
structure involved in memory formation. We investigated whether GLUT3 deletion leads
to behavioral deficits and whether PET and MR imaging can detect in vivo
GLUT3 alterations. Methods
Animals:
Floxed GLUT3 mice were
bred with CamKII alpha (CamKCre) mice, which express Cre recombinase in nearly
all CA1 neurons (Figure 1A), to obtain GLUT3 cKO and littermate GLUT3 WT
controls. Mice underwent MRI, PET imaging and behavioral testing as shown in Figure
1B.
Behavioral
analyses: The active
place avoidance test was used to assess hippocampal-dependent spatial learning.
Using visual cues, mice learn to avoid the shock zone of a rotating arena. The
number of entrances into the shock zone was calculated. The open-field test was
used to measure locomotor and exploratory behavior, and the total movement over
a 15-minute period was calculated.
PET/computerized
tomography (CT) acquisitions and analyses: PET/CT was acquired 55 minutes after 18F-FDG intravenous
injection (71±4.5μCi).
We co-registered PET and T2w MR images using VivoQuant software to
delineate the hippocampus and CA1 region based on MR contrast and calculated
the corresponding mean percent-injected dose per grams (%ID/g) values.
MR
acquisitions and analyses: T2-weighted
MRI and HP 2D 13C
CSI were acquired on
a 14.1T MR scanner using the parameters shown in Figure 1C. For 13C MRS, 24μl [1-13C]
pyruvate was polarized for ~1h in a Hypersense polarizer, dissolved in 4.5mL
buffer (80mM NaOH in PBS), and data were acquired 18 seconds after intravenous
injection. HP 13C MRSI data were analyzed
using the SIVIC software and MATLAB. K-space dimensions were zero-filled by two.
The area-under-the-curve of HP [1-13C]pyruvate and [1-13C]lactate
Lorentzian fits were measured and [1-13C]lactate/pyruvate ratios were
calculated. For analyses of T2w MRI data, brain, thalamus, ventricles, hippocampus
and CA1 regions were delineated and the corresponding volumes were calculated.
Statistical
analyses: Two-way ANOVA or unpaired Student’s t-test
were used to compare GLUT3WT and GLUT3cKO groups using GraphPad Prism software. *p<0.05, **p<0.01, ***p<0.001,
****p<0.001.Results
Body
weights were significantly decreased in female GLUT3cKO starting at 7-month-old
mice (p=0.0048),
while no changes were observed in males (Figure 1D-E).
To
determine if GLUT3 is required for normal hippocampal function, we examined how
GLUT3cKO impacts spatial learning and memory assessed by active place avoidance.
Higher number of entrances into the aversive shock zone reflected
memory deficits in female
and male GLUT3cKO (Figure 2A-C p=0.0451
and p=0.0492, respectively). No changes in total movement measured in
the open field indicated no deficits in motor
or sensory function (Figure 2D-F).
Next, in
vivo T2w anatomical MRI revealed smaller brain, hippocampus and
thalamus in female GLUT3cKO (Figure 3A-B, p=0.0009, p=0.0033, p=0.0045,
respectively), while ventricle size remained similar. No changes in
brain regions were detected in males (Figure 3C). Ex vivo T2w
MRI confirmed these results and additionally revealed smaller CA1 volume
(Figure 3D-E, p=0.0458).
Following
injection of HP [1-13C]pyruvate, we observed production of [1-13C]lactate
in the brain (Figure 4A). In the region containing hippocampus and CA1,
we observed lower HP [1-13C]lactate-to-pyruvate ratios in female
GLUT3cKO (p=0.0282), whereas this
difference was not seen in males (Figure 4B-C).
Interestingly,
[18F]FDG PET signal from the hippocampus
and CA1 were not significantly different between female or male GLUT3cKO and WT
(Figure 5A-D). Discussion
We
showed that GLUT3 deletion in CA1 hippocampal neurons led to memory impairment
in male and female mice. Interestingly, anatomical and metabolic MRI, but not [18F]FDG
PET imaging, detected changes linked to GLUT3 deletion in female mice only,
highlighting sex-specific alterations. Furthermore, decreases in body weights
and brain regions volumes in females suggest that GLUT3 deletion can cause
systemic changes, representing an interesting area for future investigation. Altogether,
our findings show that neurons import glucose to maintain normal
function. Moreover, HP 13C-MRSI supports that glucose is metabolized
through glycolysis in vivo, although there are gender differences in the
reliance on GLUT3. Importantly,
these results highlight the potential of HP [1-13C]pyruvate as a
novel method to detect downstream alterations in brain glucose metabolism that
were otherwise undetected using [18F]FDG PET imaging.
Acknowledgements
CG, HL,
KN and MMC contributed equally to this work. CG, HL, LLP, KN and MMC were
supported by NIH 1RF1AG064170-01 to KN and MMC. This work was also supported by
a grant from the California Department of Public Health (KN) and fellowships
from the Larry L. Hillblom Foundation (LS) and the Alzheimer’s Association (HL
and LLP). References
- 1. Patel, A. B. et al. Direct evidence for activity-dependent glucose
phosphorylation in neurons with implications for the astrocyte-to-neuron
lactate shuttle. Proc Natl Acad Sci U S A
111, 5385-5390, doi:10.1073/pnas.1403576111
(2014).
- 2. Lundgaard,
I. et al. Direct neuronal glucose
uptake heralds activity-dependent increases in cerebral metabolism. Nat Commun 6, 6807, doi:10.1038/ncomms7807 (2015).
- 3. Bélanger,
M., Allaman, I. & Magistretti, P. J. Brain energy metabolism: focus on
astrocyte-neuron metabolic cooperation. Cell
Metab. 14, 724-738,
doi:10.1016/j.cmet.2011.08.016 (2011).
- 4. Guglielmetti,
C. et al. Hyperpolarized (13)C MR
metabolic imaging can detect neuroinflammation in vivo in a multiple sclerosis
murine model. Proc Natl Acad Sci U S A
114, E6982-E6991,
doi:10.1073/pnas.1613345114 (2017).
- 5. Le
Page, L. M., Guglielmetti, C., Taglang, C. & Chaumeil, M. M. Imaging Brain
Metabolism Using Hyperpolarized (13)C Magnetic Resonance Spectroscopy. Trends Neurosci 43, 343-354, doi:10.1016/j.tins.2020.03.006 (2020).
- 6. Guglielmetti,
C. et al. In vivo metabolic imaging
of Traumatic Brain Injury. Sci Rep 7, 17525,
doi:10.1038/s41598-017-17758-4 (2017).
- 7. DeVience,
S. J. et al. Metabolic imaging of
energy metabolism in traumatic brain injury using hyperpolarized
[1-(13)C]pyruvate. Sci Rep 7, 1907, doi:10.1038/s41598-017-01736-x
(2017).
- 8. Choi,
Y. S. et al. Hyperpolarized [1-13C]
pyruvate MR spectroscopy detect altered glycolysis in the brain of a
cognitively impaired mouse model fed high-fat diet. Mol Brain 11, 74,
doi:10.1186/s13041-018-0415-2 (2018).
- 9. Choi,
Y. S. et al. Hyperpolarized
[1-13C]lactate flux increased in the hippocampal region in diabetic mice. Mol Brain 12, 88, doi:10.1186/s13041-019-0505-9 (2019).