LEO (purity>99.9%, the chemical structure of LEO shows in Figure 1) was purchased from the Chengdu MUST Biotechnology Co., Ltd. (Chengdu, China). A stock solution of LEO was prepared in dimethyl sulfoxide (DMSO) at 10 mM and stored at 4°C until use. Adrenalin Hydrochloride injection was purchased from Dikang-Changjiang Pharmaceutical Co., Ltd. BCA Protein Assay Kit and PMSF were obtained from Multi Sciences Biotech Co., Ltd. (Hangzhou, China). Total RNA Isolation kit, RT Easy™ Ⅱ, and Real-Time PCR Easy™-SYBR Green Ⅰ were purchased from FOREGENE Biotechnology Co. Ltd. (Chengdu, China). Protein phosphatase inhibitor and protease inhibitor mixture were purchased from Solarbio Sciences and Technology Co., Ltd. (Beijing, China). RIPA lysis buffer was obtained from Beyotime Biotechnology (Shanghai, China). 2,7-Dichlorofluorescein diacetate (DCFH-DA) was purchased from Yeasen Biotechnology Co., Ltd. (Shanghai, China). PI3K, phospho-PI3K (Tyr607), Akt1/2/3, phospho-Akt1/2/3(Ser473), and GAPDH were purchased from Affinity Biosciences, Erk1/2 mAb, phospho-Erk1/2 (Thr202/Tyr204) were purchased from Abmart Shanghai Co., Ltd. FIB was obtained from ABclonal Technology Co., Ltd. (Wuhan, China). DCF-DA was obtained from Yeasen Bio-technology Co., Ltd. (Shanghai, China). Other chemicals and reagents used in this study were obtained from Kelong Chemical Reagent Factory (Chengdu, China).

AB line zebrafish obtained from China Zebrafish Resource Center (Wuhan, China) were kept in a breeding system (Shanghai Haisheng Biotech Co., Ltd.), with a constant temperature of 28 ± 0.5°C and 14 h/10 h light/dark cycle. Adult male and female zebrafish in a 2:1 ratio were placed in a breeding tank to natural mating and laid eggs. Embryos were collected within 2 h after spawning and rinsing three times with fresh medium. The clean embryos were moved to tanks with fresh medium (5.00 mM NaCl, 0.44 mM CaCl₂, 0.33 mM MgSO4, 0.17 mM KCl, 0.1%methylene blue, equilibrated to pH 7.0) and bred at 28.5°C for subsequent experiments.

The experiment was carried out on 5 dpf (days post-fertilization) zebrafish larvae. All zebrafish were assigned to 5 groups and placed in a 6-well microplate (30 larvae per well) to study the effect of anti-thrombosis. In the control group, larvae were incubated in the embryonic medium with 0.1% DMSO. Larvae of the other groups were exposed to AH (45 μM), ASA (20 μg/ml), and LEO solutions (2.5, 5, 10 μM) for 16 h. All drugs were diluted in fresh medium. After incubating in an incubator at 28°C for 16 h, zebrafish larvae were washed three times with fresh medium and dyed with O-Dianisidine in the dark for 30 min. Next, the O-Dianisidine solution was discarded, the larvae were washed with DMSO for three times. Then the heart erythrocyte staining intensity and caudal vein thrombus of zebrafish larvae were photographed under a microscope (Leica Microsystems, Germany) and finally analyzed by Image-Pro Plus 6.0 software (Media Cybernetics, United States), as shown in Figure 2. The antithrombotic effect of LEO was evaluated based on the following formula (Zhu et al., 2016):

Therapeutic efficacy (%) = [S(drug) − S(model)]/[S(control) − S(model)] × 100%.

Zebrafish larvae were grouped and treated as described in Antithrombotic Activity of Leonurine on Adrenalin Hydrochloride-Induced Thrombosis in AB Strain Zebrafish Larvae. After treatment, zebrafish were washed and incubated with 25 ng/ml DCFH-DA solution for 30 min in the dark at 28.5°C. Afterward, the embryos were rinsed with fresh medium and photographed under a fluorescence microscope (Leica Microsystems, Germany). The fluorescence intensity in individual zebrafish larvae was analyzed by Image-Pro Plus 6.0 software to quantify the ROS accumulation.

Zebrafish larvae were grouped and treated as described in Antithrombotic Activity of Leonurine on Adrenalin Hydrochloride-Induced Thrombosis in AB Strain Zebrafish Larvae. Afterward, zebrafish were collected, and tissue homogenates were obtained by ultrasonic crushing in PBS (0.01M, PH 7.4). The concentration of lactate dehydrogenase (LDH), malondialdehyde (MDA), superoxide dismutase (SOD), glutathione (GSH), nitric oxide (NO), and endothelin-1 (ET-1) in tissues were detected with specific commercial assay kits according to the manufacturer’s instructions. MDA test kit (E-BC-K025-M), LDH test kit (E-BC-K048-M), SOD test kit (E-BC-K019-M), GSH test kit (E-BC-K030-M), NO test kit (E-BC-K035-M), and ET-1 test kit (E-EL-H0064c) were purchased from Elabscience Biotechnology Co., Ltd. (Wuhan, China). The absorbance of LDH, MDA, SOD, GSH, ET-1, and NO were measured by Tecan Infinite M1000 PRO microplate reader (Tecan, Switzerland) at wavelengths of 450 nm, 532 nm, 550 nm, 550 nm, 450 nm, and 55 0 nm, following the manufacturer’s instructions respectively. Each biological sample has 30 zebrafish larvae, and each experiment was performed in triplicate.

Zebrafish in each group were fixed with 4% paraformaldehyde for 24 h, dehydrated and embedded in paraffin at 65°C. All other processes were performed following standard procedures. Then 4 μm slide was stained with hematoxylin and eosin (H&E) for pathological analysis, all the staining were observed and photographed under the microscope.

Zebrafish larvae were grouped and treated as described in Antithrombotic Activity of Leonurine on Adrenalin Hydrochloride-Induced Thrombosis in AB Strain Zebrafish Larvae. The total RNA was extracted using a Total Animal RNA Isolation kit, and the RNA purity was measured at 260/280 nm by detecting optical density (OD). Then the RNA was converted to single-strand cDNA with a cDNA Synthesis System for RT-qPCR (FOREGENE). After that, RT-qPCR was performed using an Applied Biosystems 7500 HT Sequence Detection System. The RT-qPCR reaction conditions were set as follows: 95°C for 10 min, 40 cycles of 95°C for 15 s, and 60°C for 30 s. For RT-qPCR reactions, three independent biological samples were used for each experiment. The expression of pkcα, pkcβ, f2, fga, fgb, fgg and vwf mRNA were analyzed according to the quantification method described by the manufacturer. Then the relative mRNA expression levels were calculated by the 2-^ΔΔCT method. GAPDH served as the internal control. All primers sequences (TsingKe) used in RT-qPCR were listed in Table 1.

Previous studies have shown that activation of the PI3K/Akt pathway can activate platelet and α_IIbβ₃ to promote platelet adhesion and aggregation, then inducing thrombus formation (Guidetti et al., 2015). To explore the potential mechanisms of LEO in thrombosis zebrafish induced by AH, the expression of several proteins including PI3K, phospho-PI3K, Akt, phospho-Akt, ERK, phospho-ERK, and FIB were determined by Western blot.

Zebrafish were grouped and bred according to Antithrombotic Activity of Leonurine on Adrenalin Hydrochloride-Induced Thrombosis in AB Strain Zebrafish Larvae After 16 h of administration, zebrafish were washed twice with PBS, then added lysis buffer (RIPA lysis buffer: protein phosphatase inhibitor: PMSF: protein mixer inhibitor = 100:1:1:1) and steel balls. Zebrafish tissues were broken in a low-temperature crusher for 3 min and then centrifuged at 10,000 ×g, 4°C. The protein concentration was adjusted to the same by lysate buffer after the concentration was determined by the BCA method. After that, protein loading buffer was added in the ratio (protein sample: loading buffer = 4:1) and then denatured at 100°C for 5 min. Equal amounts of proteins from each group were loaded onto 10% SDS-PAGE for electrophoresis and then transferred to PVDF membranes in an ice bath. Subsequently, the membranes were blocked at room temperature for 2 hours in TBST containing 5% skimmed milk and incubated with primary antibody (PI3K, phospho-PI3K, Akt, phospho-Akt, ERK, phospho-ERK, and GAPDH) overnight at 4°C. The next day, the membranes were washed three times with TBST and incubated with the appropriate HRP-conjugated Goat Anti-Rabbit IgG (1:10,000) at 37°C for 1 hour. The protein bands were detected by the ECL kit and quantified by Image J. GAPDH was used as a standard reference. The relative density of each protein band was normalized to GAPDH.

100 mg of each zebrafish sample from the control group, AH group, and LEO group were put in 2 ml centrifuge tubes. Then 1 ml of tissue extract (75% methanol: chloroform: H₂O = 67.5: 7.5: 25) and three steel balls were added to each centrifuge tube. After that, the samples were ground three times (−20°C, 50 Hz, the 60 s/time) in a Tissue Grinding Device. The tissue homogenates were ultrasonicated at room temperature for 30 min and then placed on ice for 30 min. Subsequently, the tissue homogenates were centrifuged for 10 min (4°C, 12,000 rpm). Finally, 850 μL supernatant was transferred into a 2 ml centrifuge tube and lyophilized in a vacuum concentrator. Before analysis, the samples were resolved with 200 μL 2-chlorophenylalanine acetonitrile solution (50% acetonitrile: 2-chlorophenylalanine = 1: 1, 4°C) and filtered through a 0.22 μm microfilter membrane. Then 20 ul of each sample were collected and mixed into a quality control sample for correcting the analysis results, the remaining samples were tested by LC-MS.

Chromatographic separation was performed on a High-performance liquid chromatography (HPLC, Ultimate 3000 system, Thermo), which equipped with a Chromatographic column of ACQUITY UPLC^® HSS T3 1.8 μm (2.1 × 150 mm), column temperature was 40°C. Gradient elution of analytes was carried out at a flow rate of 0.25 ml/min with 5 mM ammonia formate water (A) and acetonitrile (B) in the negative ion mode or with 0.1% formic acid water (C) and 0.1% formic acid acetonitrile (D) in the positive ion mode. Samples were automatically injected with 2 μL volume at 8°C.

A linear gradient of solvent B (negative ion mode) or solvent D (positive ion mode) was programmed as follows: 0∼1 min, 2% B/D; 1–9 min, 2–50% B/D; 9–12 min, 50–98% B/D; 12–13.5 min, 98% B/D; 13.5–14 min, 98–2% B/D; 14–20 min, 2% D or 14–17 min, 2% B. The electrospray ionization mass detection (ESI-MSⁿ) in positive and negative ion modes was performed by using a Mass spectrometer (Q Exactive, Thermo). The positive ion spray voltage was 3.50 kV, the negative ion spray voltage was 2.50 kV, Sheath gas was at 30 arbitrary units, and auxiliary gas was 10 arbitrary units. The capillary temperature was at 325°C. A full scan was performed at a resolution of 70,000 with a mass to charge ratio range of 81–1,000. The higher-energy collisional dissociation (HCD) secondary lysis was performed with a collision voltage of 30 eV, while dynamic exclusion was used to remove unnecessary MS/MS information. Data-dependent experiments were performed on mass spectrometry/mass spectrometry (MS/MS) with scanning mode of higher-energy collisional dissociation (HCD). The standard collision voltage was 30 eV, and the dynamic exclusion was used to remove unnecessary information from the MS/MS spectrum.

The raw data from ultra-performance liquid chromatography-mass spectrometry/mass spectrometry (UPLC-MS/MS) were converted to mzXML format using Proteo Wizard software. The XCMS package (v3.3.2) in R language was used for peak identification, filtering, and alignment. And the parameters were set as follows: bw = 2, ppm = 15, mzwid = 0.015, mzdiff = 0.001, peakwidth = c (5, 30), method = centWave. The data were subjected to batch normalization of the peak area after obtaining the mass to charge ratio (m/z), retention time and intensity information, and the prerequisite molecules under positive and negative ion conditions.

Then the metabolite identifications were performed based on accurate mass spectra and mass spectrometry fragment patterns, and metabolites were identified by using the Human Metabolome Database (HMDB) (http://www.hmdb.ca/) and Metlin Metabolite Database (METLIN) (http://metlin.scripps.edu/). The biochemical reactions and Metabolite diagram involved in differential metabolites were queried using the Kyoto Encyclopedia of Gene and Genome Database (KEGG) (http://www.genome.jp/kegg/). The pathway enrichment analysis was performed using MetaboAnalyst (https://www.metaboanalyst.ca/) to identify the metabolic pathways and metabolites in the present study.

Studies have shown that the number of red blood cells in the heart (red blood cell staining intensity) is negatively correlated with the length or area of the tail vein thrombosis, and both of them can be used to evaluate the degree of thrombus in vivo (Zhu et al., 2016). To study the effect of LEO on thrombosis in vivo, O-Dianisidine was used to stain zebrafish. Then the antithrombotic activity of LEO was analyzed by calculating the erythrocytes staining intensity (pixel) in the heart and tail vein (thrombus area) in zebrafish. As shown in Figure 3A and Figure 4A, compared with the control group, the RBCs staining intensity in the heart was decreased significantly (p < 0.001), and the caudal vein thrombus was increased significantly (p < 0.001) in the AH group, indicating that AH could induce zebrafish thrombus successfully. Interestingly, the erythrocyte staining intensity in heart were dramatically improved in LEO (2.5, 5, 10 μM) and ASA (20 ng/ml) groups, and tail vein thrombus were notably decreased when treat with LEO (2.5, 5, 10 μM) and ASA (20 ng/ml). In addition, the thrombosis inhibition rates of the different LEO (2.5, 5, 10 μM) groups were 72.10%, 81.28%, and 88.00%, respectively, the inhibition rate of thrombus in ASA group was 93.82%. These results indicated that LEO had a certain therapeutic effect on zebrafish thrombosis induced by AH. The activities and concentrations exhibited a dose-effect relationship, and the antithrombotic effect of the LEO (10 μM) group was similar to that of ASA, as shown in Figure 4C.

While AH in vivo may induce oxidative damage via promoting ROS production, which will lead to disorders of lipid metabolism, activation of the coagulation cascade as well as the activation and aggregation of platelets, and finally induce thrombus formation. Caccese et al. (2000); Pignatelli et al. (1998); Wachowicz et al. (2002) therefore, to assess the effect of LEO on oxidative injury in zebrafish, the LDH, MDA, SOD, GSH levels were measured in the tissue homogenate. As shown in Figures 5A,B, after treatment with AH, the signal intensity and range of DCFH-DA staining increased (p < 0.001), and reduced significantly after treated with LEO (2.5, 5, 10 μM), indicating that the balance of oxidative stress in zebrafish was disturbed by AH and then improved by the treatment of LEO. Besides, after AH stimulation, the endogenous concentrations of LDH and MDA increased (p < 0.01), and the activities of SOD and GSH decreased (p < 0.05), indicating that the antioxidant capacity of zebrafish was impaired. While the intervention of LEO effectively improved the changes of these indicators, as shown in Figures 5C–F. These results suggested that LEO can interfere with the process of hypercoagulability and thrombosis through antioxidant damage.

Nitric oxide (NO), a heteroatomic free radical, is a key determinant of vascular homeostasis. The decrease of NO availability in vivo can lead to impaired endothelium-dependent vasodilation (Taddei et al., 2001), abnormal erythrocyte adhesion (Space et al., 2000), and increased platelet activation (Loscalzo, 2001). Therefore, to further study the antithrombotic mechanism of LEO, the content of NO and ET-1 were measured by biochemical kits. As shown in Figure 6, after treatment with AH, the content of NO decreased significantly (p < 0.01) and increased considerably after treating with LEO. In contrast, the content of ET-1 increased notably (p < 0.01) in AH group and reduced after treating with LEO. The result indicated that AH might injured endothelial cells and induced the imbalance of NO and ET-1. However, treatment with LEO can slightly improve the levels of NO and ET-1 caused by AH. These results indicated that LEO can increase the content of NO and reduce the secretion of ET-1, and then dilate blood vessels and improve the hypercoagulable state.

As shown in Figure 7, H&E staining showed that the caudal vein of zebrafish larval was abnormal in the AH group compared with the control group, which showed a significant accumulation of erythrocyte (a typical characteristic of thrombosis). Interestingly, the LEO (10 μM) treatment group significantly reduced erythrocyte aggregation, suggesting that LEO could inhibit erythrocyte aggregation from preventing thrombosis on AH-induced zebrafish larval.

To investigate whether LEO had a multi-targeted antithrombotic effect, we further examined the genes involved in platelet activation and coagulation cascade by RT-qPCR, respectively. Our results indicated that, after incubation with AH, the mRNA expression of platelet activating factors (pkcα, pkcβ, vwf), fibrinogen factors (fga, fgb, fgg) and thrombin factor (f2), were remarkably up-regulated, and significantly decreased by LEO (2.5, 5, 10 μM) protection, as presented in Figure 8. These results indicated that the mechanism of the antithrombotic effect of LEO might be associated with the inhibition of the related mRNA expression of platelet activity and coagulation cascade.

As mentioned previously, the PI3K/Akt signaling pathway represented an essential pathway involved in thrombosis (Hadas et al., 2013). Akt may phosphorylate the vascular eNOS and enhance NO production (Yang and Loscalzo, 2000). ERK is a member of the MAPK family and plays a critical role in platelet activation (Garcia et al., 2007; Flevaris et al., 2009). FIB is a crucial material for platelet aggregation, inducing platelet aggregation by binding to αIIbβ3 and promoting thrombus formation (Shattil et al., 1985). As shown in Figure 9, AH can significantly activate the protein expression of phospho-PI3K, phospho-Akt, phospho-ERK, and FIB, while LEO intervention can dramatically inhibit the expression of those proteins. These results indicated that the anti-thrombosis mechanism of LEO might be related to PI3K/Akt-ERK signaling pathway.