Experimental drug solutions preparation

SLP was purchased from Guangdong Yifang Pharmaceutical Co. (Guangdong, CN). It consists of 12 herbs including Ren Shen (panax ginseng), Bai Zhu (atractylodes macrocephala), Fu Ling (poria cocos), Yiyi Ren (coix lacryma), Sha Ren (amomum villosum lour), Lian Zi (nelumbo nucifera gaertn), Jie Geng (platycodon grandifloras), Baibian Dou (dolicho lablabl), Chen Pi (citri reticulatae pericarpium), Shan Yao (dioscoreae rhizoma), Da Zao (ziziphus jujuba mill) and Gan Cao (glycyrrhiza uralensis). SLP was dissolved in distilled water at 40 ℃ and configured into 1.89 g drug product/ml of SLP solution, which was stored at 4 ℃.

Establishment of MAFLD rat model

Sprague–Dawley male rats (6 weeks old, weight 250 ± 20 g) were purchased from Changsha Tianqin Biotechnology Co., td (Changsha, CN), license no. SCXK (Xiang) 2019-0013. All rats were fed in the laboratory animal room of Hainan Medical University, with the room temperature maintained at 22 °C and humidity at 50%. Rats were adaptively fed and watered ad libitum for 1 week, before being randomly divided into control group (n = 6) and MAFLD group (n = 12). The MAFLD rat was established according to a previous study [16]. Normal diet was administered to control group rats and MAFLD group rats were fed a high-fat diet (HFD, 20% lard, 4% sugar, 2% milk powder, 1% cholesterol, 73% standard feed, Suzhou Shuangli Animal Food Company, Suzhou, CN) for 4 weeks. The MAFLD rat model was successfully established by testing liver lipid levels and pathological status.

Treatment of MAFLD rats

Diets of all the rats continue to be kept the same. Rats in the MAFLD group were randomly divided into the HFD group (n = 6) and the SLP group (n = 6). The human clinical dose (Drug1) was converted to the equivalent dose administered to rats (Drug2) according to the formula: Drug2 (g/kg) = Drug1g/70 kg × 6.3 [17]. The adult clinical dose of SLP was 121 g/d, which was converted to the rat equivalent dose of Drug2 = 121 g/70 kg × 6.3 = 10.89 g/kg, SLP was dissolved in distilled water. Rats in the SLP group were given SLP solution(10 ml/kg/d) by gavage for 3 weeks, and rats in the control and HFD groups were given saline(10 ml/kg/d) by gavage for 3 weeks, after which the blood and liver tissues were collected from the rats for subsequent testing.

The whole experimental procedure is shown in the Fig. 1.

Fig. 1

The Experimental flow chart

Q-Orbitrap LC–MS/MS determination of the incoming blood components of SLP

Identification and separation of blood components of SLP were performed by Q-Orbitrap LC–MS/MS system (Thermo Fisher Scientific, MA, USA). The chromatographic mass spectrometric setup was consistent with the method according to a previous report [18]. Briefly, the Q-Orbitrap LC–MS/MS was combined with an electrospray ion source in positive and negative ion mode. Chromatographic separations were performed on an UltiMate 3000 RS column (AQ-C18, 150 × 2.1 mm, 1.8 µm, Welch) (Thermo Fisher Scientific, MA, USA) with an oven temperature of 35 °C and a flow rate of 0.3 ml/min; 0.1% formic acid aqueous solution (A)-methanol; (B) as mobile phase, using a gradient elution. The Q-Orbitrap LC–MS/MS acquisition data were initially classified using CD2.1 (Thermo Fisher Scientific, MA, USA) followed by mzCloud database (https://www.mzcloud.org/) search for comparison.

Biochemical analysis

Liver was mixed with anhydrous ethanoland homogenized at low temperature, centrifuged for 10 min (2500 rpm/min) and then the supernatant was extracted for e determination of total cholesterol (TC) and triglyceride (TG) (Nanjing Jiancheng Institute of Biological Engineering, Nanjing, CN).

Oil red O staining

Experiments were performed according to the instructions of the oil red O staining kit (BioCloud Biotechnology, Shanghai, CN). Briefly, frozen liver sections were dried, fixed in 10% formaldehyde solution for 10 min, washed with washing solution and then immersed in diluted oil red O staining solution for 20 min, washed in washing solution and stain with hematoxylin (Meyer) for 2 min, followed by tap water rinsed for 5 min, ddH2O washed for 10 s and sealed with glycerol gel. Finally, observed and photographed under microscope.

Data-independent acquisition (DIA) proteomicsProtein extraction and digestion

Liver tissue was ground to powder in liquid nitrogen, added to protein lysis solution (7 M urea, 2% SDS, 0.1% PMSF, 65 mM DTT) and lysed by sonication. After centrifuging at 4 °C for 30 min (14,000 rpm/min), the precipitate was discarded, and the supernatant was measured for protein content. 50 μg protein was mixed with 1 μL 1 M DTT and incubated at 55 °C for 1 h; 5 μL 1 M iodoacetamide was added and left for 1 h in a dark place; mixed with 300 μL pre-cooled acetone and precipitated at − 20 °C for 2 h. Finally, the precipitate and Trypsin (Promega, Madison, WI) were mixed in a 50:1 ratio and left to digest overnight at 37 °C.

High PH reverse phase separation

All sample peptide mixtures were redissolved in buffer A (buffer A: 20 mM ammonium formate in water, adjusted to pH 10.0 with ammonia) and an Ultimate 3000 system (Thermo Fisher Scientific, MA, USA) was connected to a reverse column (XBridge C18 column, 4.6 mm × 250 mm, 5 μm, (Waters Corporation, MA, USA)) for high pH separation. High pH separation conditions were set to a linear gradient from 5 to 45% B in 40 min (B: 80% ACN with 20 mM ammonium formate, adjusted to pH 10.0 with ammonia). The column was equilibrated for 15 min under initial conditions, with the flow rate maintained at 1 mL/min and the temperature at 30 °C. Six fractions were collected, and each fraction was dried in a vacuum concentrator for use.

Data-dependent acquisition (DDA) analysis and DIA analysis

The desalted lyophilized peptide was re-dissolved into a suspension by adding 30 μL buffer A (0.1% formic acid solution). 9 μL suspension was mixed with 1 μL 10 × iRT and analyzed by on-line nanospray LC-MS/MS on an Orbitrap Fusion Lumos coupled to EASY-nLC 1200 system (Thermo Fisher Scientific, MA, USA). The 3 μLpeptide sample was transferred to an analytical column (Acclaim PepMap C18, 75 μm × 25 cm) and separated with a 120 min gradient separation, from 5 to 35% B (B: 0.1% formic acid ACN solution). The column flow rate was maintained at 200 nL/min and the column temperature at 40 °C. The electrospray voltage was maintained at 2.0 kV throughout the process. The peptides were separated by chromatography and then detected by an Orbitrap Lumos mass spectrometer (Thermo Fisher Scientific, MA, USA) in DDA mode. The specific parameters were as follows: (1) MS: scan range = 350–1500(m/z); resolution = 120,000; AGC target = 4e5; Maximum injection time (MIT) = 50 ms; Dynamic exclusion time (DET) = 30 s; (2) HCD-MS/MS: Resolution = 15,000; AGC target = 5e4; MIT = 35 ms; Collision energy (CE) = 32.

The liquid chromatography conditions for DIA were the same as those for DDA. The Orbitrap Lumos mass spectrometer takes a DIA approach, and the DIA MS parameters were set as follows: (1) MS: Scan range = 350–1500 (m/z); Resolution = 120,000; AGC target = 4e6; MIT = 50 ms; (2) HCD-MS/MS: Resolution = 30,000; AGC target = 1e6; CE = 32; stepped CE = 5%. (3) Variable window acquisition with 60 windows set up and overlapping serial ports set up with 1 m/z overlap per window.

Database search and analysis

The raw DDA data was combined and analyzed by Spectronaut Pulsar X and a database of spectra was created. Trypsin was set up for enzymatic digestion, while allowing for two amino acid miss-sites. Library search parameters: the fixed modification: carbamidomethylation; the variable modification: methionine oxidation. Library standards are 1% precursor false discovery rate(FDR), 1% protein FDR, 1% peptide FDR (Additional file 1: File S1).

The raw data from DIA were processed and analyzed by Spectronaut X (Biognosys AG, Switzerland). The qualitative criteria for protein were a precursor threshold of 1.0% FDR and a protein threshold of 1.0% FDR. The peak areas of the first three MS1 peptides with FDR ≤ 0.01 were chosen as the criteria for protein quantification.

Bioinformatics analysis

Proteins were annotated against the gene ontology (GO, http://www.geneontology.org) and kyoto encyclopedia of genes and genomes (KEGG, https://www.genome.jp/kegg/) databases to obtain their function. Any absolute fold change > 1.5 and Q value < 0.05 was considered differentially expressed proteins (DEPs). DEPs with overlapping but opposite expression trends between the two comparison groups (control group vs. HFD group and HFD group vs. SLP group) were screened and then enrichment analyses were performed on those by the OmicShare tool (https://www.omicshare.com/tools).Whenp < 0.05 for a KEGG pathway or a GO entry, which indicates these DEPs are significantly enriched in the KEGG pathway or GO entry.

Building protein–protein interaction (PPI) networks

These DEPs, with overlap and opposite expression trends, were uploaded to the STRING database (http://www.string-db.org/) for PPI network analysis. The nodes of PPI network need to interact with each other in order to function biochemically, and the interaction score represents the degree of association between nodes and nodes. Having too low an interaction score between nodes means that there are more interactions and more false positives. To avoid more false positives, the minimum required interaction score was set to a medium confidence level (0.400) and the unconnected nodes in the network were hidden. The TSV files was downloaded and uploaded to Cytoscape 3.7.2. CytoNCA is a plugin for Cytoscape, evaluation and visualization analysis for multiple centrality measures [19]. CytoNCA was used to obtain the topological parameters of each node in the network: including “degree centrality (DC)”, “betweenness centrality (BC)”, “closeness centrality (CC)”. The definitions and formulae for the calculation of these parameters have been defined previously [20]. Briefly, the higher the scores of “DC”, “BC” and “CC”, the more central and important the node is in the network. In order to obtain core DEPs, nodes with scores above the median of DC, BC and CC were generally considered to be important.

Western blotting analysis

Added 0.4 ml RIPA buffer (P0013B, Beyotime Biotechnology, CN) containing protease inhibitor (P1050, Beyotime Biotechnology, CN) to 40 mg rat liver tissue, homogenized the liver by tissue grinder (KZ-III-FP, Servicebio, CN), placed on ice for 1 h, centrifuged at 4 °C for 10 min (12000 rpm/min), and collected the supernatant. Protein concentration was determined using the BCA kit (P10012S, Beyotime Biotechnology, CN), samples were mixed with SDS-PAGE protein loading buffer (5 ×) (P0015, Beyotime Biotechnology, CN) in a 4:1 ratio and placed in dry bath incubator at 100 °C for 5 min to denature the proteins. The proteins were separated on 8% or 12% SDS-PAGE gel, containing 20–45 ug of protein per lane. Afterwards, the separated proteins were transferred to PVDF membranes (FFP24, FFP32, Beyotime Biotechnology, CN), which were then blocked in 5% skimmed milk for 1.5 h. After blocking, the PVDF membranes were washed three times with TBST, then rabbit anti-Fabp1 (1:2000, 32,764–1, Signalway Antibody, USA), rabbit anti-Hmgcs2 (1:2000, 29,188–1, Signalway Antibody, USA), rabbit anti-Pck1 (1:2000, 41,644–1, Signalway Antibody, USA) and rabbit anti-β-actin (1:2000, GB11001, Servicebio, CN) were added and incubated overnight at 4 °C. After incubation, PVDF membranes were washed three times with TBST and incubated with HRP-conjugated secondary antibodies (1:10,000, G1213, Servicebio, CN) for 1 h at room temperature. The membranes were washed against with TBST and then immunoreactivity was determined using enhanced BeyoECL chemiluminescence reagent (P0018AS, Beyotime Biotechnology, CN) and analyzed with Image J.

Statistical analysis

All data were obtained from at least three independent experiments and data were statistically analyzed using GraphPad Prism 8.0, with results expressed as mean ± standard deviation (\(\overline\) ± SD). Data between groups were analyzed using Student's t-test and one-way ANOVA. p < 0.05 was considered statistically significant.