Pseudotargeted lipidomics analysis of scoparone on glycerophospholipid metabolism in non-alcoholic steatohepatitis mice by LC-MRM-MS

Drug and chemicals

The 42 kinds of glycerophospholipids isotope-labeled internal mix standards were obtained at Avanti (Birmingham, AL, USA) (Table S1). All reagents were HPLC grade and used for sample preparative and analytical UPLC analysis (Thermo Fisher Scientific, Waltham, MA, USA). Scoparone (≥98%, HPLC, CAS number 120-08-1) was supplied by purechem-standard Corp. (Chengdu, China). High-fat diet (HFD: 45% energy from fat) and normal chow diet (NCD: 15% energy from fat) were purchased from Keao Xieli (Beijing, China).

Experimental animals and design

The experiments were approved by the animal management regulations of the Ethical and Animal Welfare Committee of Hebei University (No. IACUC-2019016XG). 4-week-old male C57BL/6 mice, weighing 18–22 g, were purchased from Qinda (Qingdao, China). Animals were acclimatized to their housing environment for 1 week prior to experimentation. The temperature of the housing facility was maintained at 24 ± 1 °C with a relative humidity of 60 ± 5% and a 12-h day-night cycle. They were subsequently randomized into three groups, with 8 mice per group in the control group, NASH group and scoparone group. After 1 week of acclimation, the control group was fed the NCD while other groups were fed the HFD for 20 weeks. At week 21, olive oil (control and NASH groups) or scoparone (scoparone group) was administered by gavage (50 mg/kg) for seven consecutive days.

After the end of 7-day treatment, all mice were humanely sacrificed. Liver samples were collected, and blood was immediately centrifuged (579 g for 15 min at 4 °C). The liver tissue samples were stored at −80 °C for subsequent lipidomic analysis.

Biochemical assays

The serum levels (n = 8 per group) of triglyceride (TG), aspartate aminotransferase (AST) and alanine aminotransferase (ALT) were detected by commercially available kits (Millipore, Burlington, MA, USA), according to the manufacturer’s instructions. These assay kits were purchased from Jiancheng Bioengineering (Nanjing, China).

Hematoxylin and eosin and oil red O staining analysis

For hematoxylin and eosin (H&E), liver tissues were fixed within 4% paraformaldehyde for 24 h, followed by dehydration, paraffin embedding, serially sectioning (4µm), and staining with H&E for evaluation of histopathological changes. Liver sections for Oil Red O staining were cut 8 µm-thick, fixed with 4% paraformaldehyde and stained with hematoxylin and Oil Red O. These reagents were obtained from Solarbio Life Science (Beijing, China).

Lipid extraction and sample preparation

Hepatic lipids (n = 8 per group) were extracted from 30 mg of liver tissue by homogenizing in 300 µL methanol: water solution (v/v =1:1) (Song et al., 2022). Each homogenate was added to 300µl chloroform, vortexed and then sonicated for 10 min. After the lower layer (chloroform fraction) was taken by centrifugation, 300 µL chloroform/methanol mixture (v/v =2:1) was added for extraction followed by vortexing and then sonicating for 10 min. After centrifugation and recovery of the lower layer of chloroform, both chloroform extracts were combined and evaporated completely. The lipid residue was dissolved in 200 µL isopropanol-methanol solution (v/v =1:1), and 20 µL mixed isotope internal standard was added. After vortex and centrifugation, the supernatant was collected for LC-MRM-MS analysis.

Quantitative phospholipid profiling by LC-MRM-MS

Chromatography. The LC system was carried out using an ExionLC System that consisted of a binary high-pressure mixing gradient pump with a degasser, a column oven, and a thermostated autosampler. Finally, the conditions were optimized as follows:

The injection volume was 5 µL, and the autosampler temperature was held at 10 °C. Eluents used were 4:6 water/acetonitrile (eluent A, v/v) and 9:1 methanol/acetonitrile (eluent B, v/v), both added with 10 mM ammonium formate and 0.1% formic acid. The column temperature was 55 °C, and the flow rate was 0.35 mL/min. Chromatographic separation was performed using a UPLC HSS T3 (1.7 um, 2.1 × 100 mm, Waters). The elution gradient began at 0% B for 1.5 min, then a linear increase to 55% at 5 min, 60% at 10min, 70% at 13 min, 90% at 15 min and 100% B at 16 min, and then held at 100% for 2 min. Thereafter, the system returned to the initial conditions and was held constant for up to 2 min to allow column equilibration.

Mass spectrometry. The MS was QTRAP^® 6500+ (Sciex, Framingham, MA, USA) fitted with an IonDrive™ Turbo V source. MS data were performed using the scheduled multiple-reaction monitoring (MRM) mode operating in the positive- and negative-ion ESI mode. Source condition was as follows: ion spray voltage was −4.5KV/+5.5KV, collision gas was medium, curtain gas was 40psi, ion source temperature was 400 °C, gas 1 was 50 psi, and gas 2 was 55 psi.

Data processing

All data for samples were uploaded to MRMPROBS software (RIKEN Center for Sustainable Resource Science, Saitama, Japan) for automated data acquisition, processing, and report generation. Integrated peak area of metabolites was used to calculate the relative concentration of each detected lipid. The formula is as follows: CA = A1 / A2 * C * V / N; where, CA, analyte concentration (ng/mL); A1, analyte intensity; A2, internal standard intensity; C, internal standard concentration (mg/mL); V, constant volume (0.2mL); N, sample volume. The data were output to Microsoft Excel and the concentrations were reported as ng/mL in the reconstituted solution. The multivariate data analysis was performed using Ezinfo3.0 software (Waters Corp., Milford, MA, USA) for the unsupervised principal components analysis (PCA) and orthogonal projection to latent structure-discriminant analysis (OPLS-DA). OPLS-DA was utilized to validate the PCA model. The furthest metabolites from the origin showing a higher value of VIP-plots generated from OPLS-DA were potential biomarkers.

Statistical analysis

All statistical analyses were performed with SPSS 19.0 statistical analysis software (IBM, Armonk, NY, USA). A value of p < 0.05 was considered statistically significant.

Scoparone alleviates HFD-induced hepatic steatosis and injury

To examine the effects of scoparone on NASH, the mice were fed with an HFD for up to 20 weeks. As expected, HFD-fed animals accelerated weight gain and increased serum TG, AST, and ALT levels, reflecting hepatocyte damage. With the daily bodyweight of the mice measured over the course of the scoparone treatment, a significant weight change was observed in the 5-day treatment ( p < 0.05) (Fig. 1A). After 7 days of treatment with scoparone, the levels of serum TG, AST and ALT had been dramatically reduced (p < 0.01) (Figs. 1B, 1C and 1D).

H&E staining of liver tissues (Fig. 1E) exhibited numerous large lipid droplets (hepatic steatosis), local infiltration of inflammatory cells (hepatic inflammation) and diffuse vacuolization of hepatocellular cytoplasm (hepatocellular ballooning) in the NASH group compared with the control group. Strikingly, scoparone treatment effectively improved hepatocellular ballooning, hepatic steatosis, and inflammation in HFD-fed mice. The above histological outcomes indicating the effects of scoparone on hepatic steatosis were confirmed by staining with Oil Red O (Fig. 1F). Results showed that hepatic fat accumulation was apparent in NASH mice, while scoparone strongly suppressed the effect of HFD on hepatic steatosis. Collectively, these results suggested that scoparone robustly inhibited hepatic inflammation, hepatic steatosis, and hepatocellular ballooning.

The glycerophospholipid profile analysis by LC-MRM-MS

The representative LC-MRM-MS total ion flow chromatogram of each group showed good peak shape, separation, and intensity, indicating that the chromatographic and MS conditions optimized were suitable for the liver analysis in this study (Fig. 2). A total of 340 glycerophospholipids were detected with quantitative thresholds in liver. The stability and reproducibility of the data were assessed by the quality control (QC) samples during the experiment (Fig. S1). After peak alignment, filtering and normalization, the RSD of intensity of the above glycerophospholipids in repeated QC sample measurement was less than 40%. Unsupervised PCA analysis was employed to visualize the separation between the control and NASH group (Figs. 3A and 3B). Obvious separation was observed in the OPLS-DA analysis, indicating the unique differential glycerophospholipids expression of each group (Figs. 3C and 3D). In order to discover potentially valuable disease-related biomarkers, glycerophospholipids that differed with statistical significance between the control and NASH groups after correcting for multiple comparisons were screened. Variables that made significant contributions to the classification were determined by a threshold of VIP values, which could be obtained from OPLS-DA (Figs. 3E and 3F). A total of 14 glycerophospholipids associated with HFD-induced mouse model were screened following the above-mentioned criteria and performed clustering heat map analysis (Fig. 4). In addition, the levels of these glycerophospholipids were significantly changed in the NASH group compared with the control group (Fig. S2).

Effect of NASH treated with scoparone based on metabolic profile

The PCA analysis of liver samples revealed distinct separation among the control, NASH and scoparone groups (Figs. 5A and 5B). The OPLS-DA score plot revealed the metabolic trajectory of the scoparone group was similar to the control group and in the direction away from the NASH group (Figs. 5C and 5D). The apparent trend suggested that the pathological process of NASH might be reversed by scoparone.

Scoparone ameliorates metabolic dysregulation of glycerophospholipids in NASH mice

Lipidomic analysis showed that the amount of phosphatidylethanolamine (PE) (18:1/20:4), lysophosphatidylcholine (LPC) (18:0), phosphatidylcholine (PC) (16:0/20:4), PE(16:0/20:4), PE(18:1/20:3), PE(18:0/22:4), PC(18:1/20:4), and PE(18:0/22:5) was higher in the NASH group compared with the control group (Fig. 6A). In parallel, HFD-fed mice had a notable decrease of PC(18:0/18:2), PC(16:0/20:5), and PC(18:2/20:2). It was noteworthy that these glycerophospholipids exhibited clear differences in mice after taking scoparone. Scoparone had striking and consistent increase of PC(18:0/18:2), PC(16:0/20:5), and PC(18:2/20:2) compared to HFD-fed mice, significantly reduced PE(18:1/20:4), LPC(18:0), PE(16:0/20:4), PC(16:0/20:4), PE(18:1/20:3), PE(18:0/22:4), PC(18:1/20:4), and PE(18:0/22:5) as anticipated. Correlation analyses showed significant relationships between the above glycerophospholipids (Fig. 6B). In summary, HFD-fed induced dysregulations of glycerophospholipids, which were found to be regulated with scoparone treatment.