Improved Quantification of Hepatic Fatty Acid Metabolism in Nonalcoholic Steatohepatitis: Serum Biochemistry and In vivo Proton MRS Study with Spin-Spin Relaxation Time Correction at 9.4 T
Kyu-Ho Song1, Min-Young Lee1, Song-I Lim1, Chi-Hyeon Yoo1, and Bo-Young Choe1

1Department of Biomedical Engineering and Research Institute of Biomedical Engineering, College of Medicine, The Catholic University of Korea, Seoul, Republic of Korea


Changes in saturated and unsaturated fatty acids with hepatic triglycerides following the formation of abnormal metabolites, which play an important pathogenic role, can be measured by magnetic resonance (MR) spectroscopy. High-field-strength MR imaging scanners, which have an improved signal-to-noise ratio and high resolution for multiple lipid resonance components, are used to detect each component of lipid resonance. The aims of this study were to quantify hepatic lipid content and triglyceride composition in a preclinical nonalcoholic steatohepatitis model during the progression of steatohepatitis by assessing potential biomarkers, including spin-spin relaxation time, and applying in vivo proton MR spectroscopy with serum biochemistry.


Currently, liver biopsies are not routinely performed in patients with suspected non-alcoholic fatty liver disease (NAFLD) or non-alcoholic steatohepatitis (NASH) because of its invasiveness and perceived risks.1 Therefore, this method does not accurately assess the degree of NASH. Studies of NASH have suggested the need for a method to accurately diagnose and assess the progressive severity of NASH independent of NAFLD.2 The aims of this study were to establish an animal model for steatohepatitis with serum biochemistry and quantify hepatic lipid contents and composition of triglycerides with potential biomarkers, including spin-spin relaxation time, during the progression of steatohepatitis in a preclinical NASH model. Additionally, the relationships of the lipid components were investigated by in vivo 9.4 T proton MR spectroscopy to overcome limitations such as low resolution of the stimulated echo acquisition mode (STEAM).

Materials and Methods

All animals (eight male C57BL/6J mice weighting 22-24 g) were housed in standard plastic cages with ad libitum access to water. Their weight was monitored for 10 weeks. All animals were fed a methionine- and choline-deficient (MCD) diet of pellets composed of 10% cornstarch, 10% dextrin, and 41% sucrose. Acquisition of MR imaging and spectroscopy were performed on a 9.4 T magnetic resonance animal scanner (Biospec 94/20 USR, Bruker BioSpin MRI GmbH, Ettlingen, Germany) equipped with a 20-cm bore magnet with a 400 mT/m gradient. We used the localized STEAM [repetition time (TR) = 5000 ms; mixing time (TM) = 3.5 ms; echo time (TE) = 20 ms; number of signal averages (NSA) = 128; acquisition data points = 2,048; acquisition bandwidth = 5,000 Hz; voxel size = 3×3×3 mm3 (2.7 mL)] and multiple-TE STEAM [TR = 5000 ms; TM = 3.5 ms; multiple-TE = 20, 25, 30, 35, 40, 50, 60, and 70 ms; NSA = 16; acquisition data points = 2,048; acquisition bandwidth = 5,000 Hz; voxel size = 3×3×3 mm3 (2.7 mL)]. The water resonance signal (~4.75 ppm) of the volume of interest in the MCD-fed mice liver was suppressed by variable pulse power and optimized relaxation delays with outer volume suppression.3 All MR spectroscopy data were quantified with Linear Combination of Model spectra (LCModel, version 6.3.1-1K) software, which is useful for fitting data acquired by STEAM sequences.4 The mono-exponential equation (M(TE)=M0e(-TE/T2)) and enhanced-curve equation (M(TE)=M0e(-TE/T2)[cos(πJETE)+b]) were used to measure the spin-spin relaxation time (T2) of the lipid resonances with the J-coupling evolution value (JE) and zero TE value (M0).5 The saturated component [SC; 3(CH2)n/2(CH3)] was used as an estimate for saturated fatty acids. The unsaturated components [fraction of unsaturation (FU): CH2-CH=CH-CH2-CH2/2(CH2‑CH2‑CO); total unsaturated fatty acid index (TUFA): 3(CH2-CH=CH-CH2)/4(CH3); total unsaturated bond index (TUBI): 3(CH=CH)/2(CH3); and polyunsaturated bond index (PUBI): 3(CH=CH-CH2-CH=CH)/2(CH3)] were scaled using the terminal methyl group (CH3) resonance as an internal reference.


As shown in Fig. 1, early metabolic changes and body weight loss were observed at three weeks. As shown in Fig. 2, we obtained strong signals for the lipid resonances with significant increases. Figure 3 shows that serum albumin and glucose levels was significantly lower in MCD-fed mice (ALB, 2.4 ± 0.51 g/dL, p < 0.001; GLU, 86.8 ± 9.32 mg/dL, p < 0.001 compared to in control mice). Total bilirubin (TBIL), aspartate aminotransferase (AST), alanine aminotransferase (ALT), and alkaline phosphatase (ALP) levels were significantly higher in MCD-fed mice (TBIL, 0.8 ± 0.24 mg/dL, p < 0.001; AST, 316.8 ± 123.34 U/L, p < 0.001; ALT, 560.8 ± 141.90 U/L, p < 0.001; ALP, 1131.0 ± 209.13 U/L, p < 0.001 compared to in control mice). As shown in Fig. 4, the differences in the T2 relaxation times were > 1 ms, and these resonances differed significantly between 0 and 10 weeks (~2.03 ppm, p < 0.05; ~2.25 ppm, p < 0.001; ~2.78 ppm, p < 0.01). Figure 5 shows that the lipid contents (TL, p < 0.001; SC, p < 0.05; TUFA, p < 0.05; TUBI, p < 0.01) were significantly upregulated in MCD-fed mice compared to the values obtained at 0 weeks, while FU (p = 0.096) and PUBI (p = 0.188) were not significantly different.

Discussion and Conclusion

The results of this study demonstrated that hepatic unsaturated fatty acid levels increased during the development of NASH. In addition, we investigated the relaxation behaviors of the resonances of lipid components (methyl, methylene, β-methylene to the carboxyl group, allylic group, α-methylene to the carboxyl group, and diallylic group). Our findings suggest that fatty acid metabolism in NASH induced by a MCD diet was distinguishable from progressive NAFLD by in vivo spectra quantification and relaxation measurements.


This study was supported by grants (2012-007883) from the Mid-career Researcher Program through the National Research Foundation (NRF) funded by the Ministry of Science, ICT & Future Planning (MSIP) of Korea.


1. Cheung JS, Fan SJ, Gao DS, et al. In vivo lipid profiling using proton magnetic resonance spectroscopy in an experimental liver fibrosis model. Acad. Radiol. 2011;18(3):377-383.

2. Hamilton G, Middleton MS, Bydder M, et al. Effect of PRESS and STEAM sequences on magnetic resonance spectroscopic liver fat quantification. J. Magn. Reson. Imaging. 2009;30(1):145-152.

3. Tkáč I, Starčuk Z, Choi IY, et al. In vivo 1H NMR spectroscopy of rat brain at 1 ms echo time. Magn. Reson. Med. 1999;41(EPFL-ARTICLE-177519):649-656.

4. Provencher SW, 2001. Automatic quantitation of localized in vivo 1H spectra with LCModel. NMR Biomed. 2001;14(4):260-264.

5. Yahya A, Tessier AG, Fallone BG. Effect of J-coupling on lipid composition determination with localized proton magnetic resonance spectroscopy at 9.4 T. J. Magn. Reson. Imaging. 2011;34(6):1388-1396.


Fig 1. Intake of the MCD diet (A) and body weight (B) of the male mice were measured over 10 weeks. The percentage deviation (delta) from baseline (0 week) of intake of the MCD diet and the body weight is presented.

Fig 2. (A) Spectra were obtained in the axial plane with the voxel of interest (red box) in the livers of the MCD-fed mice. Methylene protons were increased at the affected region compared with the baseline spectrum. (B) The spectra acquired at 10 weeks ranging from 20 to 80 ms. Representative in vivo proton magnetic resonance spectra obtained with STEAM sequences with different areas from (a, methyl protons; b, methylene protons; c, β-methylene to the carboxyl group; d, allylic group; e, α-methylene to the carboxyl group; f, diallylic group; g, olefinic group) in the liver region of MCD-fed mice.

Fig 3. (A) Necropsy and images (x 100) of liver sections stained with H & E of MCD diet group for 10 weeks: mononuclear cell infiltration (blue arrow), hyaline change (green arrow), fatty change (black arrow), fibrosis (yellow arrow). (B) – (G) Serum biochemistry parameter for individual animals in each group: (B) ALB, (C) GLU, (D) TBIL, (E) AST, (F) ALT, and (G) ALP. Normal male mice (n=8; normal chow-diet group) were used as controls. The asterisks (*, **, and ***) indicate statistical significance at p < 0.05, p < 0.01, and p < 0.001.

Fig 4. The mean areas of the resonances ([A], CH3, ~0.90 ppm; [B], (CH2)n, ~1.30 ppm; [C], CH2CH2CO, ~1.60 ppm; [D], CH2CH=CHCH2CH2, ~2.03 ppm; [E], CH2CH2CO, ~2.25 ppm; [F], CH=CHCH2CH=CH, ~2.78 ppm) fit on mono-exponential curve (baseline [0 week] and 10 weeks; black dash line) and enhanced-exponential curve (baseline [0 week] and 10 weeks; blue solid line).

Fig 5. Time courses of the lipid content ([A], total lipid [TL; (-CH2-)n/H2O]; [B], saturated component [SC; 3(CH2)n/2(CH3)]; [C], fraction of unsaturation [FU; CH2-CH=CH-CH2-CH2/2(CH2‑CH2‑CO)]; [D], total unsaturated fatty acid index [TUFA; 3(CH2-CH=CH-CH2)/4(CH3)]; [E], total unsaturated bond index [TUBI; 3(CH=CH)/2(CH3)]; [F], polyunsaturated bond index [PUBI; 3(CH=CH-CH2-CH=CH)/2(CH3)]) yields while the mice are on the MCD diet. The asterisks (*, **, and ***) indicate the statistical significance at p < 0.05, p < 0.01, and p < 0.001, respectively.

Proc. Intl. Soc. Mag. Reson. Med. 26 (2018)