Application of response surface methodology (RSM) for optimization of the supercritical CO2 extract of oil from Zanthoxylum bungeanum pericarp: Yield, composition and gastric protective effect (2025)

Gastric disorders model establishment

Gastric disorders model was established as previously described method with some minor modification (). Rats were orally given ice-water mixture (0°C, 20mL/kg) twice a day (9:00a.m. and 9:00p.m.), which was prepared with ice and water of 2:1, and then maintained at room temperature for 20min. In addition, a 15min cold-water bath (10±1°C, 10cm) (1:00p.m.) were conducted once a day for 14 consecutive days. Normal rats were given the room temperature bathing and an equal volume of normal temperature water in parallel.

Drug administration

Sixty rats were randomly divided into 6 groups (n=10): normal group (0.1% poloxamer 188 in 0.5% CMC-Na, p.o.), model group (0.1% poloxamer 188 in 0.5% CMC-Na, p.o.), positive group (1000mg/kg Fuzi Lizhong pills, p.o.) and three SZB treatment groups (5, 10 and 20mg/kg SZB, p.o.). All the drugs were ground with poloxamer 188 and then suspended in 0.5% CMC-Na. Each rat was orally administrated with corresponding drugs once daily (4:00p.m.) for 14 consecutive days. The body weight of each rat was measured. The experimental protocol was presented in Fig. 3A.

Sample collection and preparation

All the rats were fasted for 12h after the last dose, which were then anesthetized and sacrificed by intraperitoneal injection of 3% pentobarbital sodium (2mL/kg, i.p.). The blood samples were collected through the rat abdominal aorta and static at 4°C for 2h. After centrifugation at 3000rpm for 10min, the supernatant serum was sub-packed and stored at −80°C for biochemical and metabolomics analysis.

Rat gastric tissue was washed with normal saline and preserved in 4% paraformaldehyde for pathological examination. The thymus and spleen were removed and weighted, and the viscera indices of which was measured as the weight of the viscera to the body weight (mg/g) of rat.

Extraction of serum metabolite was primarily performed according to previously reported method (Sarafian et al., 2014). Shortly, 100μL aliquot of each serum sample was mixed with 300μL precooled methanol and acetonitrile (2:1, v/v), and 10μL internal standards were added for quality control. After vortexed for 1min and placed at −20°C for 2h, samples were centrifuged at 4000rpm for 20min. The 300μL of supernatant were transferred for vacuum freeze drying, and the dried residue was resuspended with 150μL methanol and centrifuged for 30min at 4000rpm. The aliquot of 5μL supernatant was injected for metabolomics analysis.

Pathological observation of gastric tissue damage

For pathological observation, the gastric tissue of each group (n=5) was preserved, dehydrated, and then embedded in paraffin wax. Each tissue block was sectioned at 5-μm thickness using a microtome, and hematoxylin and eosin (HE) staining were performed. After sealing, the sections were photographed using an optical microscope at 200 to 400 magnifications, which were then evaluated by an experienced pathologist who was blinded to the treatment protocols.

Biochemical index detection

The SP and CGRP level of rat gastric tissue were determined by ELISA kits according to the manufacturer’s instructions.

Metabolomics analysis

Serum metabolic profiling data were acquired by a UPLC (Waters, USA) coupled to a Q-Extractive mass spectrometer (Thermo Fisher Scientific, USA) and a heated electrospray ionization (HESI) source. Chromatographic separation was achieved using a Waters ACQUITY UPLC BEH C₁₈ column (2.1mm×100mm, i.d. 1.7μm) maintained at 45°C of column temperature. The flow rate was set as 0.35mL/min and the injection volume was 5μL. The mobile phase consisted of 0.1% formic acid in water (A) and 0.1% formic acid in methanol (B) in the positive ion mode. In the negative mode, 10mM ammonium formate in water (A) and 10mM ammonium formate in 95% methanol (B) were used as the mobile phase. The gradient elution program was as follows: 0–1min, 2% B; 1–9min, 2–98% B; 9–12min, 98% B; 12–12.1min, 98–2% B; 12.1–15min, 2% B. The ionization source was operated in the positive and negative modes with the flow rates of sheath gas (N₂) and aux gas (N₂) at 40 arb and 10 arb, respectively. The spray voltage was set at 3.8kV (in positive ion mode) and 3.2kV (in negative ion mode). The full scan range was from 70 to 1050m/z with a resolution of 70,000 (MS) and 17,500 (MS²).

A quality control (QC) sample was prepared by mixing equal volume of each serum sample for monitoring the reliability and reproducibility of the analytes found in the samples. QC sample was analyzed consecutively prior to the analytical run and was injected at regular intervals of every 8 samples.

Raw data was firstly analyzed using the Compound Discoverer 3.1 software (Thermo Fisher Scientific, USA). The retention time (RT), m/z values and peak area intensity of each analyst were extracted and were then transferred into R software metaX () to perform the principal component analysis (PCA) and partial least squares-discriminate analysis (PLS-DA). Differential analysts for group discrimination were obtained from the variable importance in the projection (VIP) and independent t test analysis. Those analysts with VIP>1 and p<0.05 were considered as the significant metabolites and were identified by comparing with MS/MS analysis, databases of BGI Library, HDMB (https://www.hmdb.ca/), and KEGG (https://www.kegg.jp/). Pathway analysis was performed with online platform of MetaboAnalyst 5.0, a web-based tool for metabolomics enrichment.

Identification of volatile compounds of SZB

As shown in supplement Table S3, in total 22 volatile compounds were determined with retention time from 9.3min to 23.2min, and the ionic chromatogram was exhibited in Fig. 2A. These identified compounds accounted for 82.71% of the total peak area and mainly belonged to the monoterpenoids hydrocarbons (40.26%), oxygenated monoterpenes (21.0%), and ester compounds (16.95%). The predominant compounds were detected as limonene (34.76%), linalool (20.45%) and linalyl acetate (15.65%), which have been commonly discovered in Zanthoxylum species with their representative aroma flavors and documented bioactivity (Sun et al., 2020a, Sun et al., 2020b, Yang, 2008). In addition, some smaller amounts such as β-ocimene (2.62%), germacrene D (1.49%), and (4E,6Z)-2,6-dimethyl-2,4,6-octatriene (1.46%) were also found to be the important component of SZB.

Identification of non-volatile compounds of SZB

As shown in Fig. 2B, SZB obtained excellent separation effect and ionization efficiency, and produce [M+H]⁺ ions in positive mode. Nineteen unsaturated alkylamides including hydroxy-sanshools, dihydroxy-sanshools and bungeanool, were tentatively identified by comparing with the characteristic diagnostic fragment ions of standards and reported literature (; Ji et al., 2019; Ke et al., 2021, Sun et al., 2020a, Sun et al., 2020b, Zhang et al., 2019, Zhao et al., 2013). The non-volatile compounds of SZB including the name, formula, RT and main secondary fragment ions were summarized in supplement Table S4.

The relative abundance of these identified compounds was relatively higher, and No. 6, No. 7 with the largest peak area, were eluted at 19.29 and 19.86min; the positive ionization mode revealed the same first-order ion [M+H]⁺ at m/z 264.1958, and similar secondary fragment ions of m/z 246.1850, 147.1168, 107.0858 and 79.0548, which were identified as hydroxy-α-sanshool and hydroxy-β-sanshool (Li et al., 2020), respectively. Although the sanshool derivatives were almost identified in positive mode, it is worthy of attention that there remain other components of SZB in negative ion mode.

Effect of SZB on rat body weight and immune organ indexes

In the present study, an experimental model of gastric disorders induced by oral administration of ice-water mixture and cold-water bath was established. This was reported to be similar to human symptom of gastrointestinal disorders, and can increase the intragastric pressure and elevated the visceral sensitivity in functional dyspepsia-epigastric pain syndrome (Mustafa and Thulesius, 2001, Wang et al., 2013, Yang et al., 2015). While rats treated with SZB obviously reversed this disorder, at the macro level, as indicated by the increased body weights and organ indexes.

As displayed in Fig. 3B, compared to the normal group, the body weights of rats in the model group were gradually decreased with the prolong of modeling time, especially after 11days intervention. Whereas, after treatment with 20mg/kg of SZB, rats body weight was significantly higher than that in the model group (p<0.05) after continuous oral treatment for 11days, which was similar to the positive group.

Thymus and spleen are the important immune organs which can directly reflect the organism immunity (Wang et al., 2020a, Wang et al., 2020b). In this study, the thymus indexes of model group showed a noticeable decrease (p<0.01) compared with the normal group (Fig. 3D). Interestingly, SZB treatment exhibited similar improvement to positive drug of Fuzi Lizhong pills and significant increases of SZB in thymus indexes were observed in a dose-dependent manner, especially at the high dose of 20mg/kg (p<0.01). There was no obvious difference in spleen index between the normal and model group, as well as the model and SZB treatment group (Fig. 3C).

Effect of SZB on histopathology of gastric tissue

Gastric tissues from normal rats revealed well-preserved intact gastric mucosa, regularly arranged gastric glands and normal round nuclei (Fig. 4A). Compared to the normal group, obvious cell degeneration, cell swelling at the superficial mucosal layer, as well as a few lymphocytes or eosinophils and small focal necrosis of the gastric mucosa in model group were observed (the yellow arrow points to). However, the rats treated with 10 and 20mg/kg of SZB and especially those treated with positive drug exhibited slight gastric mucosal lesion with mild edema and cell hyperplasia. These results demonstrated that SZB administration could ameliorated cold-stimulation induced rats gastric tissue damage.

Biochemical analysis

It has been shown that SP and CGRP are involved in sensory signal modulation and play an important role in protecting the stomach against the harmful effect of acrylamide (). As shown in Fig. 4B & C, cold stimulation evoked an obvious gastric mucosal lesion as indicated by decreased SP and CGRP levels in rat gastric tissue (by 37.5% and 44.2%) compared to the normal control group (p<0.01). However, the SP level was fully reversed after SZB treatments at doses of 5, 10 and 20mg/kg and were almost restored to normal levels (p<0.05 & p<0.01), while SZB at 5mg/kg showed an improved tendency on CGRP level (p<0.05). Meanwhile, positive drug displayed a remarkable improvement in model-induced reduction of SP and CGRP level ( p<0.01 & p<0.05).

Metabolomics analysis

Metabolites such as lipids, amino acids, short peptides and nucleic acids are routinely produced by endogenous catabolism (Wishart, 2019). These molecules called “primary” metabolites, play an important role in human health and many key physiological functions (Wang et al., 2019). In this study, the untargeted metabolic profiling was acquired by representative total ion chromatograms (TICs) of each serum sample (supplement Fig. S2). A total of 2643 compounds in HESI (+) and 780 compounds in HESI (−) were obtained after data pre-processing, and were then subjected to R software for PCA and PLS-DA analysis.

As shown in Fig. 5A, scores scatter plot of QC and test samples in positive and negative mode were mostly within the scope of 3 std except for S1 and S5 sample, which meant a good reproducibility of the detecting system and the high quality of sample data. A clear separation trend among three groups in the 3-D PLS-DA was obtained (Fig. 5B). Obviously, the rat gastric disorder model was successfully replicated from the metabolic perspective, as indicated by the good separation effect between normal and model group in both positive and negative ion modes. Whereas the metabolic profiling in rats treated with SZB was much closer to the normal group, suggesting that SZB played a positive effect in rats with gastric disorder model.

Subsequently, the supervised PLS-DA displayed that the serum samples from the same group clustered together and score plots clearly separated the samples into two blocks between normal and model group, as well as model and treatment group, in both positive and negative ion modes (Fig. 5C & D). The explanation and prediction of PLS-DA model were obtained, of which R²Y and Q² were higher than 0.93 and 0.44 in positive ion mode, and higher than 0.86 and 0.63 in negative ion mode, respectively. These results indicated that the PLS-DA model with satisfactory explanatory and predictive capability is capable of discriminating the different intervention groups.

Based on the parameters of VIP>1 and p<0.05, a total of 464 differential metabolites were expressed between normal and model group, while 182 were observed between model and SZB treatment group. After intersection and determination, 19 serum metabolites were identified as the potential biomarkers of SZB for protecting rats against cold-stimulated gastric disorder. Compared with the normal group, 11 metabolites were remarkedly increased in the model group and 8 metabolites were decreased, while these metabolites were reversely regulated after oral administration of SZB. The detailed information of the potential biomarkers was exhibited in Table 1. To visualize the patterns, a heatmap of the potential metabolites in three groups was constructed by using Bioinformatics (https://www.bioinformatics.com.cn/), as shown in Fig. 5E. The higher abundant metabolites were shown in red color, while lower abundant metabolites in blue.

Furthermore, the metabolic pathway analyses were performed with MetaboAnalyst 5.0 according to the 19 potential metabolites. As shown in Fig. 5F, 9 metabolic pathways seem contribute to the protective effect of SZB against gastric disorder, including beta-alanine metabolism, glutathione metabolism, pantothenate and CoA biosynthesis, pyrimidine metabolism, biosynthesis of unsaturated fatty acids, primary bile acid biosynthesis.

Nineteen potential metabolites and nine pathways were primarily related to amino acid metabolism and lipid metabolism. Amino acids are essential for various cellular functions and energy source, and participate in metabolic pathways, such as the tricarboxylic acid cycle (TCA) (). Spermine is one of the polyamine metabolites and is synthesized from ornithine, and the acetylation of spermine could be driven by spermidine. Furthermore, β-alanine could be produced by spermidine through a series of reactions, which was then converted into pyruvate and enters the TCA cycle to regulate tissue energy metabolism (Shimizu et al., 2021). In this study, spermine concentration of serum in model group was significantly increased, whereas it was decreased after SZB treatment for 14days. It is speculated that spermine was decomposed and involved in the production of β-alanine, subsequently up-regulated β-alanine metabolic pathways, and then promoted energy release.

Lipid metabolism pathway mainly includes unsaturated fatty acid biosynthesis, fatty acid elongation, fatty acid degradation, and primary bile acid synthesis (Lv et al., 2021). Fatty acids are one of the sources of mammalian, and can be decomposed into H₂O and CO₂ in the body and release a lot of energy. Palmitic acid and stearic acid, the typical saturated fatty acids, involved in body energy metabolism and food digestive absorption (Gangl and Ockner, 1975, Ito et al., 2010). It is reported that acetyl co-enzyme A produced by palmitic acid, could generate CO₂ and ATP (Ramprasad et al., 2001). After SZB intervention, the relative abundance of palmitic acid and stearic acid in serum were elevated, subsequently, which maybe up-regulate the biosynthesis pathway of fatty acids and unsaturated fatty acids, and then decomposes to CO₂ and H₂O under aerobic conditions to releases a lot of energy. Consistent with the research results, Wang et al. (2015) reported that phosphoenolpyruvate hydroxyl kinase could be increased by Z. bungeanum treatment, and promote TCA cycle and thereby increasing the internal energy generation (Wang et al., 2015). These findings proved that the protective effect of SZB against gastric disorder not only related to a single metabolic pathway but also a complex network of metabolic pathways. In the network, accelerate energy metabolism of the body is the core mechanism. The main metabolic network of gastric protective effect of SZB was shown in supplement Fig. S3.

Supplementary data 1