Morphological and behavior studies of kainic acid (KA), an excitotoxic non-NMDA glutamate receptor agonist, in animal models have led to various findings including significant neuronal loss in the hippocampus [1], upregulation of gliosis [2], as well as impaired locomotor activity and sensory function [2]. Magnetic resonance spectroscopy (MRS) provides unique capability of detecting metabolite concentrations in vivo. Previously, MRS of brain has been investigated in KA induced animal models [3]. However, there are still no MRS studies performed on spinal cord injured by KA administration. Thus our current study was to focus on the proton MRS detection of metabolite profile of spinal cord gray matter just caudal to a KA injury in rats 14 days after administration of the excitotoxic agent, and further to correlate the MRS findings to the histopathology of the animal model. Quantitative evaluations of different metabolites were also performed to identify potential MR based biomarkers of neurotoxicity.

Animal Model: Female Sprague Dawley rats (n=9) were separated into three groups (n=3 each). All procedures were performed according to the guidelines of the University of Louisville Institutional Animal Care and Use Committee. Two groups of rats received 0.5 mM and 2.0 mM KA respectively. The third group (sham control) were injected with same amount of saline at the identical locations as the KA groups. For the KA groups, animals received KA injections bilaterally at level L2. Two injections of 0.5 µl each were administered into the intermediate gray matter separated by 1.2 mm rostrocaudally. MR imaging (MRI) and MRS were performed 14 days post-injection.

MRI/S: All MRI and MRS data were acquired using Agilent 9.4 T horizontal bore MRI system equipped with Agilent 205/120 HD gradient coil. Surface coil was carefully positioned on the back of each animal to cover T10 – L4 spinal cord regions, which included the KA injury epicenter as well as lumbar enlargement regions (L3-L4). T2-weighted images were obtained using a standard spin echo multi-slice imaging sequence with the following parameters: TR/TE = 1000/12.9 msec; matrix size = 256 x 256; field of view (FOV) = 40 x 40 mm²; and 15 slices with a slice thickness of 1.2 mm. For localized MRS data acquisition, a voxel with a dimension of 1.7 x 2.2 x 6.0 mm³ was positioned caudal to the KA injury epicenter, aligned with the long axis of the spinal cord (Fig. 1). Localization by Adiabatic SElective Refocusing (LASER) sequence was used with the following parameters: TR/TE = 1500/37 msec; spectrum width = 4006 Hz; complex points = 2048; and number of averages = 512. Variable power and optimized relaxation delays (VAPOR) was applied as water suppression scheme. MRS data were loaded into jMRUI (version 5.0, UCBL, France) for spectroscopy quantification. Full FID was modeled using AMARES plugin, and major metabolite peak areas were calculated (N-acetyl-aspartate, NAA; total creatine, Cr; total choline, Cho, and myo-inositol, mIn).

An overlay of T2-weighted MR images with MRS voxel is shown in Fig. 1 to demonstrate the dimension and the placement of the chosen volume of interest (VOI) for single voxel MRS. A water suppressed MR spectrum obtained from an in vivo acquisition at the selected spinal cord level of a sham rat is shown in Fig. 2 (top). Major metabolites including NAA, Cho, Cr as well as mIn can be readily seen. When comparing MRS obtained from the moderate group(Fig. 2, bottom) with that of the sham control group, the NAA peak was substantially diminished in the moderate group (2 mM KA, 4x injection) and there was an increase for the mIn peak at ~ 3.5-3.6 ppm (both spectra scaled to the same peak height for Cr). Indeed, when comparing metabolites between groups and correlating with percent spared gray matter measured by histology (a gold standard for quantification of gray matter loss), NAA concentration showed a good positive correlation with the histological parameter. The trends, however, were reversed for Cho and mIn (Fig. 3).To the extent of our knowledge this is the first study of longitudinal monitoring of metabolites in lumbar spinal cord using ¹H-MRS in a KA injured rat model. Major findings include elevated total choline and myo-inositol at one and two weeks after KA injury as well as a significant down-regulation of N-acetyl aspartate, a biomarker for neuronal loss. Our preliminary data indicate that ¹H-MRS potentially is a valuable tool for monitoring spinal cord metabolism in a variety of SCI models.