Methanol, acetonitrile and isopropanol (HPLC grade) were purchased from Merck (Darmstadt, Germany). Formic acid was obtained from Fluka (Buchs, Switzerland). Ammonium formate and internal standard 1-heptadecenoyl-sn-glycero-3-phosphocholine were from Sigma-Aldrich (St. Louis, Missouri, USA). Chloroform (HPLC grade) was obtained from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Glyoxylic acid(50% in water) was purchased from TCI Development Co., Ltd. (Shanghai, China). Ultrapure water was prepared using a Milli-Q water purification system (Millipore Corp., Billerica, MA, USA).

Dried material of OS was obtained from Anguo (Hebei, China). The material was identified by Prof. Lian-na Sun (School of Pharmacy, Second Military Medical University, Shanghai, China). The extraction method is referred to Yu-Sen Zhong et al. (Yu-Sen et al., 2012). Briefly, the dry material of OS was boiled in 12 volumes of 75% ethanol (v/w) for 1h and the OS ethanol extracts were extracted with 3 volumes of petroleum ether, EtOAc and n-BuOH to obtain 3 fractions. And the component analysis of EEOS fraction was carried out by UPLC-Q-TOF/MS method and the detailed methodology of the experiment was shown in the supplementary materials section. As a intervention, the dry power of EEOS was prepared into a suspension of 24 mg/ml with 0.5% CMC-Na and stored at 4°C.

Animal experiments followed the National Institutes of Health guide for the Care and Use of Laboratory Animals. Eighteen wild-type male C57BL/6 mice (7–8 weeks) were purchased from Shanghai SLAC laboratory Animal Co., Ltd.(Shanghai, China). Mice were housed in groups(control group, model group, EEOS group) of six per standard cage. The animal experiments lasted seven days and the mice were given drug at 8:30 am every day. Model group and EEOS group were injected with glyoxylate at a dose of 120 mg/kg/day and control group was injected with the same volume of saline. Two hours after injection of glyoxylate, the EEOS group was treated EEOS suspension by intragastric administration at a dose of 360 mg/kg/day, while the control group and model group were given 0.5% CMC-Na. In addition, Cystone was selected as the positive reference drug. The selection of the dose of EEOS administered and positive reference drug is described in the supplemental materials section.

On the seventh day of animal experiments, mice were anesthetized with 3% chloral hydrate and blood samples were collected by orbital puncture, and then the kidneys were collected. After clotting at 4°C for 2 h, the blood was centrifuged at 3,500 rpm for 15 min and the serum was saved for biochemical analysis. After removing the capsule and pelvis, three kidneys of each group were fixed in neutral 4% paraformaldehyde purchased from Servicebio company (Wuhan, China) and the others were immediately stored at −80°C.

Kidneys were fixed in 4% paraformaldehyde and embedded in paraffin. Sections (3–4 μm) were prepared and dyed using the Von Kossa stain kit(calcium Stain). The principle of Von Kossa staining is metal replacement. Its main process includes: Section dewaxing to water, Silver nitrate reaction, Hematoxylin- Eosin staining, dehydration and sealing. Calcium deposition was estimated by observing stained plaques.

The contents of serum creatinine and blood urea nitrogen were measured using a Chemray 240 automatic biochemistry analyzer (Shenzhen, Guangdong, China). Six serum samples from each group were tested for serum creatinine and blood urea nitrogen.

Six kidney samples from each group for UPLC-QTOF/MS analysis. Kidney tissue was weighed and homogenized in methanol/chloroform/water (v/v/v,2:1:0.8) followed by the addition of a volume of chloroform and a volume of water to extract the lipids. After vortexing for 30 s, the mixed samples were placed at room temperature for 5 min and then centrifuged at 13,000 rpm for 10 min at 4°C and 200 μl of chloroform were transferred to an Eppendorf tube and evaporated under nitrogen. The dried extracts were reconstituted with 400 μl of an chloroform/methanol (1:1,v/v) solution and transferred to autosampler vials. To monitor system stability, a quality control (QC) sample was prepared by mixing the same volume of each sample.

UPLC-QTOF/MS analysis was performed on a CBM-20A/Alite HPLC system (Shimadzu, Japan) equipped with a TripleTOF 5600 mass spectrometer (AB Sciex, USA). Chromatographic separations were performed on a Waters XBridge™ BEH C18 column (2.1 mm × 100 mm, 2.5 μm, Waters, Milford, MA). The mobile phases consisted of 40:60 water: acetonitrile(A) and 9:10:81 acetonitrile: water: isopropanol (B), both containing 0.1% formic acid and 10mM ammonium formate. The flow rate was held constant at 0.3 ml/min and the injection volume was 1 μl. The gradient elution conditions were: 0min, 40% B; 3 min, 68% B; 5 min, 70% B; 7 min, 70% B; 12 min, 85% B; 15 min, 99% B; 19 min, 99% B; 19.1 min, 40% B. The entire chromatographic gradient time is 24 min and the column temperature was 45°C.

The mass spectrometer was operated in both positive and negative information-dependent acquisition (IDA) modes. The specific instrument parameters were as follow: the source temperature was 550°C; the ion source gas 1 and 2 were 60 psi and the curtain gas was 35 psi; the ion spray voltage floating was 5.5 kV in positive mode and 4.5 kV in negative mode. The accumulation time for full scan was 150 ms and the accumulation time for each IDA experiment was 55 ms. The mass range from m/z 100 to m/z 1,300 and the collision energy was set 45 eV. Eight spectra with an intensity threshold above 100 cps, isotope exclusion were set to 4 Da.

The raw data were converted into “Analysis Base File” (ABF) format files by ABF Converter 4.0.0 software. Then, all data were analyzed using MSDIAL 3.20 software, which can perform deconvolution, streamline criteria for peak identification and identify lipids. Finally, an alignment result, with a list of accurate mass, retention time, metabolite name, and corresponding intensities for all the detected peaks, was exported in.txt format. After the data were internal standard normalized, the resultant data matrices were introduced to the SIMCA-P 11.0 software for unsupervised principal component analysis (PCA) and partial least squares discriminate analysis (PLS-DA).

Statistical analysis was performed using SPSS21.0. The significant differences of three groups were tested by one-way ANOVA followed by a Turkey test for multiple comparisons. P < 0.05 and fold change value (FC) greater than 1.5 or less than 0.67 were considered statistically significant.

UPLC-QTOF/MS approach was applied to identify the compounds of EEOS fraction (Figure S1) and detailed method is listed in supplemental materials section. As shown in Table S1, twenty-two compounds, mainly flavonoids, also contain several phenolic acids and terpenes, were identified from the EEOS fraction. Qualitative analysis of the EtOAc extract was based on the precursor ion and product ions of the standards. Representative qualitative analysis results are shown in Figures S2–S4.

Using the rutin standard as a control, total flavonoids content of EEOS was determined by a colorimetric method, linear relationship between absorbance (A) and rutin concentration (C, μg/ml) by equation A= 0.0128C +0.0073(R² = 0.999) (Table S2). The final results indicated that the total flavonoids content in EEOS was about 88.82% (Table S3). In addition, rosmarinic acid, eupatorin and salvigenin were selected as the indexes to test the reproducibility of the EEOS extraction. The relative standard derivation (RSD) values of the peak areas of each component were less than 5% in six sets of experiments (Table S4), indicating that the EEOS extraction is reproducible.

To investigate whether EEOS could improve the development of stones in mice, histological analysis of renal sections using Von Kossa staining showed calcium spots in three groups mice (Figure S5 and Figure 1A). Calcium spots was significantly reduced in renal sections by EEOS treatment compared with model group mice, which is consistent with the content reduction in calcium of mouse kidney homogenate after EEOS intervention (Figure 1B). Serum creatinine and blood urea nitrogen are routine indicators reflecting the state of renal function. Compared with model group, the serum creatinine and blood urea nitrogen levels are lower in the control group and EEOS group (Figure 1B). In addition, there were significant differences in blood urea nitrogen level between the EEOS group and the control group.

Lipidomics mainly recognizes important lipids in metabolic regulation, and reveals the mechanism of lipids in various life activities by comparing changes in lipid metabolism networks under different physiological conditions. By UPLC-MS/MS analysis, lipids in kidneys of three groups were detected in positive and negative ESI modes. Representative total ion chromatograms (TICs) for kidney samples from pooled QC sample are shown in Figures 2A, B. And, the coefficient of variation(CV) of QC sample was used to test reproducibility of the measured metabolites (Figure S6), and the ions with CV values greater than 20% in the positive and negative modes were approximately 10%, which indicated that the method has a well reproducibility and stability.

For multivariate statistical analysis, the unsupervised PCA model was firstly used to observe the discrete trend in the QC sample and judge the presence or absence of abnormal samples. It can be seen from the PCA score scatter plots (Figures 3A, B) that the QC sample is well polymerized, indicating that the instrument is in good operating condition. The supervised PLS-DA was applied to enhance the classification performance. As illustrated by PLS-DA score scatter plots (Figures 3C, D), model group was obviously separated from control group and the EEOS group had a significant tendency to closer to the control group than model group. In our established model for analysis, the cumulative R²X, R²Y and Q² were above 0.4 (Table S5), indicating that the PCA and PLS-DA models were successfully.

Based on the P value less than 0.05 and the FC value greater than 1.5 or less than 0.67, the differential ions with significant adjustment in the EEOS group were screened. According to the accurate mass-to-charge ratio and their MS/MS product ions (Figure S7), 51 differential lipids shown in Table 1 were identified by database resources(MSDIAL, METLIN, and HMDB). These lipids mainly were glycerophospholipids including glycerophosphocholines(PC), glycerophosphoethanol- amines (PE), glycerophosphoserines (PS), glycerophosphoinositols (PI) and glycerophosphoglycerols (PG) (Figure S8). Cluster analysis with heat map to visualize the distribution of differential lipids in three groups of samples. As shown in Figure 4, the distribution of lipids in the model group was significantly different from that in the control group, but after treatment with EEOS, the lipids distribution of EEOS group was basically consistent with the control group, indicating that EEOS did interfere with the lipid metabolism of mice with stones.

Fatty acids metabolism is closely related to the formation of kidney stones (Takahiro et al., 2008; Hall et al., 2017; Bikulčienė et al., 2018). Esterified fatty acids in glycerophospholipids are potential precursor for lipid mediators, especially ω-3 fatty acids (Kang, 2012; Hishikawa et al., 2017). The concentration of ω-3 fatty acids in glycerophospholipids was obviously elevated in EEOS group mice than in model group, which is consistent with found that ω-3 fatty acids were significantly upper in normal people compared to patients with uronephrolithiasis (Bikulčienė et al., 2018), indicating that EEOS extract may reduce the occurrence of kidney stones by regulating the ω-3 fatty acids metabolism.

Peroxidase proliferator activates the receptor α (PPARα) is a master regulator of lipid metabolism, involved in inflammation and oxidative stress, and can be activated by ω-3 fatty acids. Haruya Takahashi et al. found lysophospholipids are increased on PPARα activation (Haruya et al., 2015). Meanwhile, the lysophospholipid content in the EEOS group was significantly higher than in the stone model group, which shown that the PPARα was more active in EEOS group mice than model group. PPARα antagonizes NF-κB-controlled proinflammatory mediator transcription by binding to NF-κB subunit p65 to form PPARα/NFκBp65 complexes (Jessica et al., 2011). In addition, ω-3 PUFAs can also reduce the expression of pro-inflammatory genes, such as the reduction of TNF-α expression by EPA (Adkins and Kelley, 2010). Although the glycerophospholipids esterified ω-6 fatty acids did not differ in the three groups in this study, there was a significant difference in the ratio of ω-6/ω-3. An unbalanced ω-6/ω-3 ratio is highly proinflammatory in terms of arachidonic acid metabolism and IL-1β production contributes to the prevalence of atherosclerosis and obesity (Simopoulos, 2008), which may be a potential factor for increased risk of stones in atherosclerosis patients and obese people.

Oxidative stress is caused by an imbalance between oxidant and antioxidant. Due to the instability of the double bond, PUFAs in glyperophospholipdis were easily oxidized under oxidative stress. The content of PUFAs in the EEOS group was significantly higher than that in the model group, indicating that the oxidative stress in the stone mice was alleviated after intervention with EEOS. Moreover, the vinyl double bond at the C-1 position of the glycerol backbone of the plasmalogens is preferentially oxidized under oxidative stress, preventing oxidation of polyunsaturated fatty acids and possibly mitigating cellular lipid peroxidation reactions (Lessig and Fuchs, 2009; Dean and Lodhi, 2018). In this study, the plasmalogen containing PUFAs in the model group was significantly higher than the control group, suggesting that the antioxidants in the stone mice were imbalanced and oxidative stress occurred. The plasmalogens in the EEOS group were also lower than in the model group, but the difference was not significant due to the large intra-difference in the EEOS group. Therefore, further research is needed on the effect of EEOS on the metabolism of plasmalogens. In addition, NF-κB upregulates the expression of pro-inflammatory cytokines and adhesion molecules to promote leukocyte adhesion, increase the release of oxygen free radicals and aggravate tissue damage (Holthe et al., 2004). In damaged cells and kidney tissues, crystals preferentially attach and aggregate into the nucleus, and the attached crystals can destroy the cells and form new tissue damage (Devarajan, 2018). From the aforementioned words, this study suggests that a potential mechanism of EEOS extracts in the treatment of stones may be through the regulation of fatty acid composition on glycerophospholipids to mediate oxidative stress and inflammation.

Glycerohospholipids account for about 70% of the total lipids in mammalian cells and are tightly related to the fluidity, flexibility, permeability and electrophysiological properties of biomembrane (Li et al., 2014). In addition to maintaining the integrity of cell membranes, PCs are also involved in the development of inflammatory reactions. PCs with 1-O-alkyl-2-acetyl-sn-glycerol-3-phosphocholine structure are called platelet activating factors(PAF) and are a class of lipid mediators with strong activity. In the kidney, PAF can induce intrarenal infiltration of inflammatory cells, activate mesangial cells and inflammatory cells to release a variety of inflammatory mediators; promote immune deposition in the kidneys and reduce glomerular filtration rate (Reznichenko and Korstanje, 2015). In present study, the acetyl group at sn-2 in PAF is replaced by ω-3 fatty acids, such as PC(O-14:0/22:6), PC(O-16:0/22:6), and PC(O-18:0/22:6), resulting in a decrease in the biological activity of PAF, which is beneficial to anti-inflammatory and prevent tissue damage.

PEs constituting 15%–25% of phospholipids and is very prominent in the energy metabolism of mitochondrion. Depletion of PE in mitochondria leads to dysfunctions in respiration and loss of mitochondrial DNA (Birner et al., 2003), resulting in insufficient energy metabolism in cells. PE in EEOS group was significantly higher than that in model group, indicating that EEOS may be related to mitochondrial productivity. In addition, plasmalogen glycerophosphoethanolamines(PE(P-)) are preferentially oxidized under oxidative stress (Lessig and Fuchs, 2009). In the research, the PE(P-16:0/20:4) and PE(P-16:0/22:6) in EEOS group were lower than that in model group, which suggested that the antioxidant activity of EEOS administration mice was enhanced to resist oxidative damage.

PI, PG and PS are acidic phospholipids and often recognized as second messengers. PI is a type of bioactive water soluble products of PLA2 and lysolipase and is an important precursor of secondary messenger phosphatidylinositol-3,4,5-triphosphate (PIP3) (Corda et al., 2009; Bohdanowicz and Grinstein, 2013). PIP3 can activate Ca²⁺ channels on cells and promote Ca²⁺ influx (Li et al., 2014) to regulate Ca²⁺ concentration intracellular and extracellular. In this study, the PI(16:0/22:6) elevated significantly in EEOS group compared to model group, which indicated that EEOS fraction may regulate the conversion of PI to PIP3, inhibiting the influx of Ca²⁺, thereby preventing cells from irreversible damage caused by Ca²⁺ overload. As an anionic glycerophospholipid, PG has an important role on lipid-protein topology and function (Frimmelová et al., 2016). However, Ca²⁺ can regulate ionic protein-lipid interactions by neutralizing the lipid negative charge, reducing the local concentration of acidic phospholipids (Li et al., 2014), which may be used to explain the decline of the PG in the EEOS group. PS is up to 10% of biological membrane and is the most abundant acid phospholipid (Platre and Jaillais, 2016). Although PS can promote cell proliferation and survival and mediate inflammatory responses by modulating cytokines (Maragno et al., 2015; Oyler-Yaniv et al., 2017), there were no significant differences in PS between the three groups in this study.