Improved lipid profiling of the mouse liver by 1H-NMR spectroscopy at 14.1T in vivo

Purpose/Introduction: Whole-body metabolic deregulation features changes in hepatic lipid dynamics. In vivo 1H-MRS allows to noninvasively assess hepatic lipid content (HLC) and composition both in humans [1] and rodents [2]. Transgenic mice provide insight into disease mechanisms but, comparing with bigger subjects, reduced sample size and consequently increased acquisition time to compensate for sensitivity issues, remain limiting factors. We aimed to improve sensitivity and spectral resolution of the 1H-NMR spectrum of the mouse liver employing high magnetic field strength, i.e. 14.1T. Subjects and Methods: Seven C57BL/6J mice (13-20 weeks) under isofluorane anesthesia were scanned in the supine position with a 1H quadrature surface coil over the abdomen. NMR measurements were performed in a horizontal bore 14.1T-26 cm magnet. Multi-slice gradient echo images were acquired in the coronal, sagittal and axial orientations for anatomical identification of the liver. Localized, respiration-gated 1H-NMR spectra were acquired from a 9-15 μl voxel with STEAM (TM, 20 ms; TR, 6.5 s; TE, 8-35 ms; 18-25 scans). Spectra were corrected for B0 drift and phase, summed and analyzed with LCModel. T2 correction was done by mono-exponential fit of peak areas as a function of TE. HLC was estimated as the T2-corrected area of 1.3 ppm-lipid resonance relative to that of the water plus 1.3 ppm-lipid. The quantification method was validated in water-in-oil phantoms (Figure 1). Water suppression was achieved with a Gaussian-shaped pulse during TM period (84 scans). Results: Respiration-gated acquisition yielded well-defined anatomical images of the mouse liver structure (Figure 2) and very stable inter-scans signal intensity (Figure 3). STEAM-determined water T2 was 8.4 ± 0.3 ms, shorter but still comparable with previous reports of ~11 ms with PRESS at 9.4T [2] and 11.7T [3]. HLC was 1.1 ± 0.1% in young adult mice. Fatty acyl resonances were well resolved in water-suppressed spectra (Figure 4) and choline-containing compounds could be identified. Discussion/Conclusion: We report highly stable localized 1H-MRS of the mouse liver at high field. Enhanced sensitivity allowed for accurate determination of HCL from small voxels confined to hepatic tissue, in short experiment series (~10 min). The saturation profile of the fatty-acyl chains can be determined even for healthy mice with low HLC. This approach opens the possibility to study mice models with very low HLC. References: [1] Hamilton G, 2011, NMR Biomed 24, 784–790. [2] Ye Q, 2012, Mang Reson Mater Phy, in press. [3] Tang H, 2007, Proc Intl Soc Mag Reson Med 15.