Synopsis
Obesity is a pandemic syndrome that leads to reduced
life expectancy, increasing the risk of heart disease, type-2 diabetes and some
type of cancers. Noteworthy, to
understand the mechanisms of obesity onset and development, several animal
models, such as administration high fat diets, have been developed. We used 1H-[13C]
MRS methods in regular and in high fat diet fed mice to investigate the effects
of high caloric diets and obesity in the hypothalamus, its effects in glucose
metabolism and metabolic fluxes in neurons and glia. We found differences that
suggest impaired glucose metabolism in the hypothalamus of obese mice.PURPOSE
Recent investigations have associated consumption of
high fat diets (HFD) with inflammation in the hypothalamus,
the main cerebral regulator
of energy balance in the brain, reporting processes as astroglisosis and
neuronal injury at a local level
1. The implementation of a technique
capable of detecting
in vivo the
changes in metabolites, neurotransmitters and metabolic fluxes in the
hypothalamus, could represent an excellent strategy with which to investigate
the physiologic and pathologic conditions of the mouse brain. Thus, the aim of
this study was to follow the incorporation of
13C-labeled glucose
into its metabolites in the hypothalamus of mice fed regular diet or high fat
diet using
1H-[
13C] MRS methods,
and investigate the HFD induced effects on glucose metabolites and neuroglial
metabolic fluxes.
METHODS
C57BL/6 mice (12weeks old, n=3, 28.5±1.2g) were fed
with a 60% kcal fat diet (http://www.researchdiets.com) during two months or with regular food chow (13 weeks old, n=5, 28.6±1g).
Animals were fasted overnight (12h) before experiments, and body weight and
blood glucose levels were controlled in each experimental session. All MR measurements
were performed in a horizontal 14.1T/26cm Varian Magnet, using a homemade
1H
quadrature surface coil (12mm diameter) combined with a linear
13C
coil (11 mm diameter). The hypothalamus
voxel (1.8x2.7x1.8 mm
3) was selected based on anatomical MR images
and using a mouse brain atlas as a reference
2 and field homogeneity
was improved using FASTMAP methods
3. Indirect
1H-[
13C]
detection was achieved using the full signal intensity NMR sequence BISEP-SPECIAL
4combined
with OVS and VAPOR water supression
5.
Hypothalamic glucose (Glc) metabolism was evaluated during
the infusion of [1,6-
13C
2] glucose solution (20% w/v ,
bolus= 9mL/kg, infusion rate= 15mL/kg/h). To increase the SNR, spectra of all
animals’ from the same group were added together every 13 min of acquisition, and
metabolites were quantified using LCModel. Non-edited 1H MR spectra
acquired with
13C decoupling but without
13C editing
contain
1H resonances coupled to both
12C and
13C
and thus can be quantified with a standard basis set for neurochemical profiles
of mouse hypothalamus. The
13C-editing
spectra were quantified using another simulated basis set as previously4.
The fraction of isotope enrichment (FE) in lactate (Lac), glutamate (Glu),
glutamine (Gln) and g-amino-butyrate
(GABA) were estimated. Time course of Glu C4, Gln C4 and Glx C3 was fitted to a
one compartment model of glucose metabolism using MATLAB nonlinear regression
methods, and TCA rate (Vtca), transmitocondrial flux (Vx), neurotransmission
rate (Vnt), dilution factor (Kdil) and composite flux (Vgt) were estimated. The
errors of all adjusted metabolic fluxes were evaluated by Monte–Carlo
simulation
6.
RESULTS
Body weight of HFD fed and control animals
before the MRS measurements (fasting state) (45.8±2.6g, 25.6±1g) were
significantly different (p<0.005, t-test) as well as their fasting blood
glucose levels (13±1.2mmol, 5.25±1.2 mmol).
Figure 1 shows representative non-edited and edited hypothalamic spectra
in control animals after 160 min of infusion of labeled glucose: note the high
levels of GABA, a typical feature of the mouse hypothalamus
7. SNR of
the edited spectra was >5 for all cases, as measured from LCModel, and
linewidths of total creatine at 3.0ppm <20Hz.
Figure 2 depicts the time course of the
13C label accumulation
in the same animals, and the calculated FE for HFD and control animals is
illustrated in
Figure 3. FE of Lac
C3, which depicts a similar turnover in both groups, was used as an input
function for the metabolic fluxes quantification. The total pool size of the
metabolites was also measured along the glucose infusion, and values are
summarized in
Figure 4. Glc and Asp
show higher values in the HFD group and GABA and Lac higher in the control
group (p<0.001 in all cases, t-test). Finally, derived metabolic fluxes are
illustrated in
Figure 5. Calculated
values show similar Vtca rates for both animal groups, higher Vx and Vgt for
the HFD animals and higher Vnt and Kdil for the control group.
DISCUSSION
Indirect
1H-[
13C] detection
revealed the incorporation of
13C-labeled glucose
in vivo into a number of metabolites of
biological importance in the hypothalamus. We show for the first time, in vivo reliable measurements of the
turnover of Lac C3, Glu C4, Gln C4, Glu C3, Gln C3, GABA C4, GABA C3 in the mouse
hypothalamus. HDF animals, which show prediabetic fasting blood glucose levels,
display increased Glc and Asp levels and decreased Lac and GABA levels in the
hypothalamus, as compared to regular fed animals. Moreover, the increased Vx and
decreased Vnt and Kdil highlight putative alterations in glucose metabolism.
Acknowledgements
Authors want to thank B. Lanz and J. Duarte for his
help in the modeling process and interpretation of results. The work is
supported by the CIBM of the UNIL, UNIGE, HUG; CHUV, EPFL and Leenaards and
Louis-Jeantet Foundations.References
1Thaler
JP, J. Clin. Invest. (2012);2 Paxinos G. (2001) ;3Gruetter
R et al. Magn.Reson.Med (2000);4Xin et al. Magn.Reson.Med.(2010);5Tkac I Magn.Reson.Med.(2009);6
Lanz, B. Front. Endocrinol (2013);
7Lei H. et al. NMR Biomed (2010).