Synopsis
This
study demonstrates that 13C MRS of hyperpolarized pyruvate can be
used to detect increased lactate production from pro-inflammatory macrophages,
mechanism mediated by pyruvate dehydrogenase kinase-1 upregulation and pyruvate
dehydrogenase inhibition, in a preclinical model of multiple sclerosis, hence
providing a novel tool for in-vivo
detection of neuroinflammation.Introduction
Activated mononuclear phagocytes (MPs,
macrophages/microglia) are key players in the progression of multiple sclerosis
(MS). Interestingly, upon activation, MPs undergo metabolic reprogramming, and
especially increase their lactate production
1,2.
Importantly, a recent
in vitro study shows that increased
lactate production resulted from inhibitory phosphorylation of the pyruvate
dehydrogenase (PDH) complex by the pyruvate dehydrogenase kinase 1 (PDK1) enzyme
3.
Based on this new knowledge, we questioned if hyperpolarized
13C
Magnetic Resonance Spectroscopic Imaging (HP
13C MRSI) of pyruvate
4
could be used to assess increased pyruvate to lactate conversion linked to MPs
activation in MS (
Fig.1A). To do so,
we longitudinally studied the toxin-induced cuprizone (CPZ) mouse model, a well
established MS model
5. We demonstrated that HP
13C MRSI
of pyruvate could detect MPs activation non-invasively during disease
progression, thus highlighting its potential for monitoring neuroinflammation
levels in MS. We further demonstrated, for the first time
in vivo, that the detected increase HP lactate production was
linked to PDK1-induced PDH inhibition in activated MPs.
Methods
Animals: C57BL/6J mice (n=5) received a 0.2% CPZ diet
for 6 weeks while aged-matched control (CTRL) received a standard rodent chow
(n=8) 6. All CTRL and CPZ fed mice were imaged prior (W0) and after
4 and 6 weeks of CPZ (W4 CPZ, W6 CPZ) (Fig.1B).
MR
acquisitions: T2-weighted
images were acquired for proper assessment of the corpus callosum (CC)
location, region of maximum inflammation in the CPZ model (TE/TR=20/1200ms,
thickness=0.5mm, NA=2, matrix=256x256, FOV=30x30mm²). For 13C MRS,
24μL of [1-13C] pyruvate preparation was hyperpolarized using a
Hypersense DNP polarizer (Oxford Instruments) for one hour3. After
dissolution, HP pyruvate was rapidly dissolved in isotonic buffer (pH~7) and injected
intravenously (iv) over 12sec. From the beginning of iv injection, 2D dynamic
CSI 13C data were acquired on a 14.1T MR system using:
TE/TR=1.2/60ms; SW 2500Hz; 128points; 4sec resolution; FA=10deg; FOV=24x24 mm²;
5mm thickness.
Immunofluorescence
(IF): IF analyses were
performed for myelin (MBP), MPs (Iba-1) and pyruvate dehydrogenase kinase-1
(PDK1), (n=3 mice per group). Pyruvate dehydrogenase (PDH) activity in the CC was
assessed using a PDH activity assay kit (Abcam), (n=4 mice per group).
Data
analysis: HP
lactate and HP pyruvate levels were calculated as the sum of integrals over
time (Fig.1C). For both MR and IF
data, CPZ and CTRL groups were compared using unpaired t-tests and longitudinal
analyses of HP 13C MRSI using paired t-tests (*p<0.05,
**p<0.01, ***p<0.001).
Results
T2-weighted images showed hyperintensity
in the CC region in W4 and W6 CPZ animals (arrows), confirming the expected
effect of the diet (Fig.2A). IF
analysis demonstrated that hyperintensity was linked to modulations in myelin (MBP)
and MPs activation (Iba-1) levels (Fig.2B).
Importantly, even though W4 and W6 lesions are linked to different underlying
cellular events, (W4: high demyelination/high inflammation; W6: partial
remyelination/low inflammation), T2-weighted could not distinguish
between W4 and W6 animals, highlighting the fact that this method is not
specific to inflammation.
In contrast, HP 13C MRS of pyruvate
showed specificity to neuroinflammatory events only. While the pyruvate
delivery remained stable over time in all experimental groups (Fig.3A), the HP lactate/pyruvate ratio was
significantly increased at W4 in the CC of CPZ animals as compared to
age-matched CTRL (72±29% at W4 CPZ, p=0.019,
n=5) (Fig.3B, C, D and E). This
ratio was then significantly decreased by 20±14% at W6 CPZ (p=0.028, n=5, Fig.3F).
Importantly, IF analyses confirmed that
increased HP lactate/pyruvate ratio was associated with a strong upregulation
of PDK1 in activated MPs at W4 CPZ, as demonstrated by the merged image (p<0.0001,
n=3), (Fig.4A-B). Importantly, subsequent
inhibition of PDH activity was also measured in the CC (p<0.0001, n=4), (Fig.4C). Further in line with the HP
results, at W6 CPZ, the number of PDK1-expressing MPs present in the CC significantly
decreased (p=0.001, n=3) and PDH activity subsequently increased as compared to
4W CPZ.
Conclusions
All together, our results demonstrate that
13C
MRSI of HP pyruvate can detect increased lactate production
in vivo in a preclinical MS model, and
that this increase is linked to the presence of activated MPs that upregulate
PDK1 and subsequently inhibit PDH activity, thus pushing the flux towards
lactate production. Importantly, this report is the first to demonstrate the
use of a metabolic imaging method to monitor neuroinflammation. Because HP
13C
MRSI is clinically translatable and expanding rapidly, this study is of high
significance for future clinical trials not only on MS, but also all
neurological diseases presenting an inflammatory component. Such method would
enhance diagnosis and help refine therapeutic regimen, which will ultimately
improve clinical outcome and patient care.
Acknowledgements
Funding sources: IWT-Vlaanderen PhD grant; NMSS_PP3395;
Cal-BRAIN349087; UCSF_RAP7500634; UCSF Department of Radiology seed grants
#14-04 & #14-05; NIH-funded Hyperpolarized MRI Technology Resource Center
#P41EB013598. References
[1] Galvan-Pena et al, Metabolic Reprograming in Macrophage Polarization, Front. Immunol.(2014)
Sep 2;5:420.
[2] Tannahill et al, Metabolic reprograming of mononuclear phagocytes in progressive multiple sclerosis. Front. Immunol.(2015)
Mar 11;6:106.
[3] Tan et al, Pyruvate dehydrogenase kinase 1 participates in macrophage polarization via regulating glucose metabolism J. Immunol. (2015)
Jun 15;194(12):6082-9
[4] Kurhanewicz et al, Analysis of cancer metabolism by imaging hyperpolarized nuclei: prospects for translation to clinical research, Neoplasia (2011) Feb;13(2):81-97.
[5] Kipp et al, The cuprizone animal model: New insights into an old story, Acta Neuropathol. (2009)
118(6):723-36
[6] Praet et al, Cellular and molecular neuropathology of the cuprizone mouse model: Clinical relevance for multiple sclerosis Neurosci. Biobehav. Rev. (2014) Nov;47:485-505.