Mélissa Vincent1,2, Yohann Mathieu-Daudé1,2, Julien Flament1,2, and Julien Valette1,2
1Molecular Imaging Research Center (MIRCen), Commissariat à l'Energie Atomique et aux Energies Alternatives (CEA), Fontenay-aux-Roses, France, 2UMR 9199, Neurodegenerative Diseases Laboratory, Centre National de la Recherche Scientifique (CNRS), Université Paris-Sud, Université Paris-Saclay, Fontenay-aux-Roses, France
Synopsis
It is still unclear how diffusion properties of water differ from one
compartment to another (neurons, glial cells, extracellular space…). Here we
propose the idea that Chemical Exchange Saturation Transfer of Glutamate
(gluCEST) may be used to specifically reduce the contribution of intraneural
water to the overall signal attenuation, thus providing enhanced sensitivity to
non-neuronal compartments. Acquisitions performed in two rats yields water ADC
slightly but significantly higher when gluCEST is performed, supporting the
idea that water diffusion is slower inside neurons.
Introduction
Diffusion MRI
is a powerful tool to decipher the intricacies of tissue microstructure. However, water
ubiquitously diffuses in all compartments (neurons, glial cells, extracellular
space…), and it is unclear how diffusion properties differ in these
compartments, which severally limits the robustness of biophysical models needed
to extract quantitative microstructural parameters (e.g.1). Experimental
approaches using intracerebral injection of exogenous probes such as gadolinium
or 19F-labeled molecules have been proposed to gain some specificity
to the extracellular space2-4.
Nevertheless, such approaches are not ideal since: i) they might alter the osmotic
balance and induce microstructural alterations; ii) entry inside cells cannot
be ruled out; iii) they cannot be applied to Humans.
Here we propose
that Chemical Exchange Saturation Transfer (CEST) of endogenous metabolites may
be used to specifically modulate the relative contribution of water in the
different compartments, hence providing some compartment-specific insights into
water diffusion properties. We implement that idea for glutamate, a widely acknowledged
neuronal marker which can generate a strong CEST effect (gluCEST)5
thanks to fast exchange between protons of its amine group and water. By
incorporating a gluCEST module within a diffusion-weighted sequence, we aim at reducing
the contribution from intraneural water to the overall signal (Figure 1). Acquisitions performed in
two rats reveal that water ADC is slightly but significantly lower when gluCEST
is performed, i.e. water diffusion is slower inside neurons. Methods
Sequence:
For this initial work, we wanted to favor detection sensitivity over spatial
resolution, to be able to capture subtle differences in diffusion properties.
We also wanted high acquisition speed to avoid bias due to potential drifts.
Therefore, we decided to measure water signal using localized
diffusion-weighted magnetic resonance spectroscopy (DW-MRS) instead of DW-MRI. Briefly,
a gluCEST module was positioned in front of a fully adiabatic LASER
localization sequence (Figure 2). Isotropic
diffusion-weighting was achieved thanks to QMAS gradients6 inserted around
the first refocusing pulse7,8. The stability
of water ADC with respect to varying saturation offsets was assessed in a phantom.
Animal
experiments: Two female Sprague-Dawley rats were anesthetized
using isoflurane and scanned in an 11.7 T Bruker scanner. A volume coil was
used to ensure homogenous transmit-B1 for gluCEST saturation, while a quadrature
surface coil was used for reception to maximize SNR. A voxel was positioned in a
glutamate-rich cortical area above the corpus
callosum. For the gluCEST saturation time, we wanted to use a “short” tsat
to limit potential water exchange between the different compartments, and hence
avoid distribution of saturated water in all compartments that would obscure
any compartment-specificity. We decided to use tsat=300 ms, which is
much shorter than tsat=1 second as widely used for gluCEST. The
optimal saturation parameters were found to be fsat=+3 ppm and B1=9
µT. In the end, three saturation offsets were used: +3 ppm corresponding to
glutamate resonance, -3 ppm as a symmetric control, and -20 ppm as off-resonance
control (the water resonance is here at 0 ppm according to the usual CEST convention).
The ADC (two-point measurement at b=0.2 ms/µm² and 1.2 ms/µm²) was evaluated for
each of these saturation frequencies, and the measurement was repeated ten times
over an hour in each animal.Results
Figure 3-A shows the voxel localization together with a
water-suppressed metabolite spectrum, which confirms high glutamate
concentration in our voxel. Figure 3-B displays a few water peaks as measured
for different fsat, and Figure 3-C exhibits the whole z-spectrum as
acquired between -5 and 5 ppm by
integrating water spectra acquired at different fsat in
one animal. The strong gluCEST contrast can be appreciated, yielding an
asymmetrical magnetization transfer ratio (MTRasym) at 3 ppm of ~15% in both
animals. Figure 4 displays the mean water spectra averaged over ten
acquisitions and obtained at the three aforementioned saturation offsets, with and
without diffusion weighting, for the two animals. Figure 5 shows the boxplots obtained for the ten ADC values repeatedly
measured in each animal for the three saturation offsets. In both cases, the
ADC for fsat=3 ppm is slightly (~1%) but very significantly (p<0.005)
higher than for fsat=-3 and -20 ppm, whereas there is no significant
difference between the latter two.Discussion/ Conclusion
The approach presented
here opens up many exciting opportunities. However, it is fundamentally limited
by the necessity to generate a strong CEST contrast while keeping tsat
short enough to minimize water exchange between compartments. Shorter tsat
can be partly, but not fully, compensated by higher B1. Further investigations
are hence required to determine optimal tsat. Interestingly, we also
performed experiments at tsat=1 second (not shown), but could not
observe ADC variations when glutamate was saturated, suggesting that too much
inter-compartment exchange has occurred over 1 second.
Here we used a
spectroscopic approach, as we thought that high sensitivity would be required
to detect subtle variations in ADC. However, partial volume effect from
ventricles cannot be entirely ruled out when using large voxels. Future works will
investigate the possibility to reduce voxel size, and to adapt the approach to
DW-MRI. Besides alleviating partial volume effects, higher spatial resolution
would enable measurements in highly-ordered structures, such as the corpus callosum, parallel and perpendicular
to fibers.
Acknowledgements
This project has received funding from the
European Research Council (ERC) under the European Union’s FP7 and Horizon 2020
research and innovation programmes (grant agreements No 336331 and 818266). The 11.7 T MRI scanner was funded by a grant
from “Investissements d'Avenir - ANR-11-INBS-0011 - NeurATRIS: A Translational
Research Infrastructure for Biotherapies in Neurosciences".References
1 Jelescu,
I. O., Veraart, J., Fieremans, E. & Novikov, D. S. Degeneracy in model
parameter estimation for multi-compartmental diffusion in neuronal tissue. NMR in biomedicine 29, 33-47,
doi:10.1002/nbm.3450 (2016).
2 Silva,
M. D. et al. Separating changes in
the intra- and extracellular water apparent diffusion coefficient following
focal cerebral ischemia in the rat brain. Magnetic
resonance in medicine 48, 826-837, doi:10.1002/mrm.10296 (2002).
3 Kunz,
N., da Silva, A. R. & Jelescu, I. O. Intra- and extra-axonal axial
diffusivities in the white matter: Which one is faster? NeuroImage 181, 314-322, doi:10.1016/j.neuroimage.2018.07.020
(2018).
4 Duong,
T. Q. et al. Extracellular apparent
diffusion in rat brain. Magnetic
resonance in medicine 45, 801-810, doi:10.1002/mrm.1108 (2001).
5 Cai,
K. et al. Magnetic resonance imaging
of glutamate. Nature medicine 18,
302-306, doi:10.1038/nm.2615 (2012).
6 Topgaard,
D. Director orientations in lyotropic liquid crystals: diffusion MRI mapping of
the Saupe order tensor. Physical
chemistry chemical physics : PCCP 18, 8545-8553, doi:10.1039/c5cp07251d
(2016).
7 Marchadour,
C., Brouillet, E., Hantraye, P., Lebon, V. & Valette, J. Anomalous
diffusion of brain metabolites evidenced by diffusion-weighted magnetic
resonance spectroscopy in vivo. Journal
of cerebral blood flow and metabolism : official journal of the International
Society of Cerebral Blood Flow and Metabolism 32, 2153-2160,
doi:10.1038/jcbfm.2012.119 (2012).
8 Ligneul,
C. & Valette, J. Probing metabolite diffusion at ultra-short time scales in
the mouse brain using optimized oscillating gradients and
"short"-echo-time diffusion-weighted MRS. NMR in biomedicine 30, doi:10.1002/nbm.3671 (2017).