Screening of Potential Anti-Thrombotic Ingredients from Salvia miltiorrhiza in Zebrafish and by Molecular Docking

Background: Danshen (DS), the dry root of Salvia miltiorrhiza Bge., has been used in traditional Chinese medicine (TCM) for many years to promote blood circulation and to inhibit thrombosis. However, the active ingredients responsible for the anti-thrombotic effect and the underlying mechanisms are yet to be fully elucidated. Methods: Molecular docking was used to predict the active ingredients in DS and their potential targets by calculating the scores of docking between DS ingredients and thrombosis-related proteins. Then, a chemical-induced zebrafish thrombosis model was applied to confirm their anti-thrombotic effects. Result: The molecular docking results indicated that compared to the control ligand, higher docking scores were observed for several compounds in DS, among which salvianolic acid B (SAB), lithospermic acid (LA), rosmarinic acid (MA), and luteolin-7-O-β-d-glucoside (LG) could attenuate zebrafish caudal vein thrombosis and recover the decrease in heart red blood cells (RBCs) in a dose-dependent manner. Conclusions: Our study showed that it is possible to screen the potential active components in natural products by combining the molecular docking method and zebrafish in vivo model.


Introduction
Thrombosis is characterized by intravascular thrombus formation and vessel occlusion, which may result in myocardial infarction, ischemia, and stroke, and it can be especially dangerous if it blocks blood flow to organs or tissues [1]. Thrombus formation in blood vessels is usually associated with systemic disorders such as hypercoagulability [2]. As the leading cause of death worldwide [3,4], thrombosis, which plays an important role in cardiovascular disease, seriously threatens human health and life [5,6]. Therefore, new drug development in thrombosis treatment is in great demand.
Traditional Chinese medicine (TCM) has been widely applied in clinical therapy in China and other countries [7][8][9] for a long time. According to the TCM theory, thrombosis belongs to the syndrome of blood stasis. Many herbal formulae and extracts are used to treat blood stasis syndrome (BSS) via promoting blood circulation, inhibiting platelet aggregation, and promoting the hemostasis release reaction [10,11]. Danshen (DS), the dry root of Salvia miltiorrhiza Bge. which can also be used in food and healthy supplement [12,13], has attracted increasing attention as a treatment option for cardiovascular and cerebrovascular diseases [13,14]. DS, which contains more than 200 chemical constituents including salvianolic acid B (SAB), danshensu (DSS), lithospermic acid (LA), rosmarinic acid (MA) [15], and luteolin-7-O-β-D-glucoside (LG), is a commonly used medicine for "invigorating" the blood and reducing blood clotting [16]. For example, many studies have shown that DS can inhibit platelet aggregation and adhesion, improve blood microcirculation, and reduce thrombus formation [17]. Therefore, DS has been extensively used to cure cardiovascular and cerebrovascular diseases, inflammatory diseases, and neurodegeneration diseases [15]. Furthermore, it has been demonstrated that the extract of the herb has effects against hemostasis as well as eliminating blood stasis. However, the active ingredients responsible for the anti-thrombotic effect in DS and their target proteins are yet to be fully elucidated.
Virtual screening based on molecular docking is a computational technique used in drug discovery research. It can deal with the quick search of large molecular libraries to identify the structures that are most likely to bind the drug target, which is typically a protein receptor or enzyme [18]. In recent years, molecular docking, which plays an important role in drug screening and design [19], is one of the most frequently and effectively used methods to predict the optimal binding mode (binding sites and binding poses) and affinity between a small molecule drug candidate (ligand) and a target protein (receptor) [20][21][22]. Several compounds including lithospermate B, salvianolic acid C, lithospermate acid B, magnesium lithospermate B, danshensuan B, salvianolic Acid B, monomethyl lithospermate, and oleoyl neocryptotanshinone in DS have been identified as ingredients with potential anti-thrombotic activity with this approach [18].
Zebrafish (Danio rerio) has been used in the past few years to identify novel factors of hemostasis and thrombosis, and to analyze their functions in greater detail [23][24][25][26]. Compared to the conventional mammal in vivo models, which are usually laborious, costly, and time consuming [27], the zebrafish model organism has a large number of advantages including high fecundity, small size, rapid development, and rapid generation time [28]. In fact, zebrafish is considered to be especially appropriate for the research of angiocardiopathy than other models because the endogenous blood circulation can be dynamically observed without invasive methods [16], and there is remarkable similarity in the cardiovascular system between zebrafish and humans [29,30]. Furthermore, with the capability of percutaneous oxygen absorption, zebrafish embryos could survive for a long time without functional blood circulation, during which the investigation of antithrombotic agents could be done [31].
In the present study, virtual screenings based on molecular docking between components in DS and thrombosis-related protein targets were performed. Potential active ingredients and possible protein targets were predicted according to the docking scores. Then, candidates of active compounds in DS with potential anti-thrombotic effect were treated to chemical-induced thrombosis zebrafish to validate their ability of promoting blood circulation and eliminating blood stasis.

Analysis of Molecular Docking
The structure information of 202 compounds in DS were obtained from the Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform (TCMSP) (https: //old.tcmsp-e.com/tcmsp.php, accessed on 4 November 2021). The relevant target proteins of thrombosis were collected in the Therapeutic Target Database (TTD) (http://db.idrblab. net/ttd/, accessed on 4 November 2021), and 31 target proteins were obtained, including a successful target, clinical trial target, discontinued target, and literature-reported target. Then, the corresponding X-ray crystallographic structures with higher resolution were searched through the UniProt (https://www.uniprot.org, accessed on 4 November 2021) database, and 25 crystal structures were selected from the Protein Data Bank (PDB) (http: //www.rcsb.org/, accessed on 4 November 2021) database for subsequent experiments. Finally, 202 active ingredients in DS were docked with 25 target proteins respectively, and the results are shown in Tables 1 and 2. On one hand, a thrombosis-related target protein could combine well with multiple ligands, e.g., antithrombin-III (ATIII, PDB ID:1R1L) could bind to 160 ingredients in DS; on the other hand, one compound in DS would bind to 10-15 thrombosis-related targets. According to the docking result, many small molecules in DS could bind to the same proteins including 1R1L, 5AFY, 4C2A, 6TS4, 3QX3, 3T3M, 2JKH, and 5AFN. Among all the compounds that could combine with multiple thrombosisrelated targets in Table 2, SAB, MA, LA, and LG (structures shown in Figure 1), which were characteristic compounds of DS and can be commercially obtained with high purity, were selected for further in vivo pharmacological tests. The docking scores of SAB, MA, LA, and LG are shown in Table 3.     To further explore the binding mode between the receptor and ligands, SAB, MA, LA, and LG were selected as active compounds for the molecular docking study. The interaction diagrams of the best-docked conformations are shown in Figure 2. The hydrogen bonds and amino acid residues of the interaction between coagulation factor XI (F11) (PDB ID: 6TS4) and the compounds are presented in Table 4. Hydrogen bonds play an important role in the binding of ligands to F11 that exist between SAB and TYR59A, ARG37D, GLY193, ALA195, GLY216, ALA97, GLY218, CYS219, CYS40, LEU39; MA and LEU39, CYS40, ASP189, CYS219, LYS192; LA and CYS58, ARG37D, ALA195, GLY193, ASP194, ASP189, GLY218, CYS219; as well as LG and ASP189, HIS57, GLU98, GLY193, ALA195, ASP194, TRP215, SER214, VAL227. Particularly, ALA195, GLY193, GLY216, CYS219, ASP189, and LYS192 are the most frequent amino acid residues, which may be the important residues for the interactions between ligands and F11. Moreover, the compounds containing hydroxyl and carbonyl, especially phenolic hydroxyl groups, are vital for hydrogen bond formation with F11. These results indicated that these compounds may be the potential active ingredients in DS with an anti-thrombotic effect. To further explore the binding mode between the receptor and ligands, SAB, MA, LA, and LG were selected as active compounds for the molecular docking study. The interaction diagrams of the best-docked conformations are shown in Figure 2. The hydrogen bonds and amino acid residues of the interaction between coagulation factor XI (F11) (PDB ID: 6TS4) and the compounds are presented in Table 4. Hydrogen bonds play an important role in the binding of ligands to F11 that exist between SAB and TYR59A, ARG37D, GLY193, ALA195, GLY216, ALA97, GLY218, CYS219, CYS40, LEU39; MA and LEU39, CYS40, ASP189, CYS219, LYS192; LA and CYS58, ARG37D, ALA195, GLY193, ASP194, ASP189, GLY218, CYS219; as well as LG and ASP189, HIS57, GLU98, GLY193, ALA195, ASP194, TRP215, SER214, VAL227. Particularly, ALA195, GLY193, GLY216, CYS219, ASP189, and LYS192 are the most frequent amino acid residues, which may be the important residues for the interactions between ligands and F11. Moreover, the compounds containing hydroxyl and carbonyl, especially phenolic hydroxyl groups, are vital for hydrogen bond formation with F11. These results indicated that these compounds may be the potential active ingredients in DS with an anti-thrombotic effect.   TYR59A, ARG37D, GLY193, ALA195, GLY216, ALA97, GLY218,  CYS219, CYS40, LEU39  TRP215, LYS192, HIS57, LEU146, CYS191, ALA190   MA  LEU39, CYS40, ASP189, CYS219, LYS192  ALA195, GLY226, GLY216, ALA190, SER214, TRP215,  CYS191   LA  CYS58, ARG37D, ALA195, GLY193, ASP194, ASP189, GLY218,  CYS219  ALA97, HIS57, LYS192, TRP215, ALA190, GLY216 LG ASP189, HIS57, GLU98, GLY193, ALA195, ASP194, TRP215, SER214, VAL227 GLY226, CYS219, ALA190, LYS192
In recent years, polyphenols contained in fruits and vegetables had attracted considerable attention because of their benefits in human health, including the reduction of the risk of cardiovascular disease. SAB, MA, LA, and LG were all polyphenolic constitutions of DS and were proved to exert an anti-thrombotic effect in our study. Therefore, the result may suggest to a certain degree that the application of DS in TCM is coherent with the use of polyphenols in Western medicine for the treatment of cardiovascular disease.

Maintenance and Husbandry of Zebrafish
Wild-type AB strain adult zebrafish were purchased from the Shanghai FishBio Co., Ltd. (Shanghai, China). Zebrafish were raised and kept in a flow-through farming system provided by Nanjing Eze Rinka Biotechnology Co., Ltd. (Nanjing, China) according to the standard procedure of China Zebrafish Resource Center. Zebrafish were maintained in light (under a 14:10 h light/dark cycle) and temperature (25-28 • C)-controlled conditions in water (pH 7.2-7.8 and conductivity 450-550 µs/cm) and fed twice daily with live brine shrimp and dry flake food. Embryos were obtained from spawning adults in a breeding chamber overnight with a sex ratio of 2:1 (male/female) according to the standard zebrafish breeding protocol.

Molecular Docking Studies
The structures of molecules of active ingredients in DS were obtained from TCMSP and saved to files in mol2 format. With Gasteiger-Huckel charge, Tripos force field, and the Powell energy gradient method, all molecules were optimized in SYBYL 2.0 software (TRIPOS, Inc., St. Louis, MO, USA). The Powell energy gradient method was used to optimize the molecules by setting the maximum number of times to 10,000 and the energy convergence level difference to 0.005 Kcal/mol [22,32] and the rest of the parameters were set to default values. According to this method, the lowest energy small molecule structure is regarded as the active conformation.
Target proteins related to thrombus were searched from the TTD and UniProt database. The crystal structure of the protein came from the PDB database. Prior to docking studies, proteins needed to be pre-treated, including residual repairs, energy minimization, the addition of hydrogen atoms, as well as the removal of co-crystallized ligands, water molecules, and metal ions from the complexes.
Molecular docking was carried out using a standard Surflex-Dock protocol in SYBYL 2.0 software. The docking pocket was formed with amino acid residues at 0.5 Å in the range of the original ligand molecule, and the remaining parameters were set as default. The original ligand was redocked into the binding pocket to verify the reliability of the docking method [22,32]. Then, compounds were docked into the binding pocket using the same parameters. Afterwards, by comparing the docking score with the original ligand, compounds with a higher value were screened out.

Preparation of Solutions and Samples
First, 2.64 mg of AH was dissolved in 50 µL DMSO to prepare 240 mM stock solution, and 2.36 µL PHZ was dissolved in 997.64 µL absolute ethanol to prepare 24 mM stock solution. These stock solutions were further diluted to working solutions with 0.2 mM PTU-containing water, and the final concentrations of DMSO and absolute ethanol were kept below 0.1%. Thirty milligrams O-dianisidine and 41 mg CH 3 COONa were dissolved in 20 mL of absolute ethanol as the stock solution; and then, 128 µL 30% H 2 O 2 and 20 mL of water were added to form O-dianisidine working solution. ASP, SAB, MA, LA, and LG were dissolved in DMSO to form the stock solutions; then, they were further diluted to working solutions (12.5, 25, 50, 100, and 200 µg/mL) with 0.2 mM PTU-containing water.
After DS compounds treatment, zebrafish were stained in the dark with O-dianisidine staining solution for 30 min. Then, the O-dianisidine working solution was abandoned, and the zebrafish were washed by DMSO three times. All of the images were taken at the same magnification, with an upright fluorescence microscopy imaging system. The software used to capture the images was cellSens Standard.
Zebrafish from each group were picked out for image capture of the heart and caudal vein RBCs, respectively. The images were quantitatively analyzed using Image-Pro Plus 6.0 according to IOD value. The anti-thrombotic effects of the test compounds were evaluated based on the formula below: IC 50 was calculated with GraphPad Prism 8.

Ethics Statements
All zebrafish experiments were conducted according to the guidelines of the Animal Ethics Committee of the School of Pharmacy and Bioengineering, Chongqing University of Technology.

Statistical Analysis
All data were expressed as the mean ± the standard error of mean (SEM) of three different experiments. Multiple group comparison was conducted by one-way analysis of variance (ANOVA) of IBM SPSS Statistics 19. A p-value of less than 0.05 was considered as statistically significant.

Conclusions
In this study, a rapid and effective method for anti-thrombotic effects screening in DS was established based on the zebrafish in vivo assay and computational studies of molecular docking. Molecular docking studies showed that SAB, MA, LA, and LG from DS exhibited the good binding affinity to the thrombosis-related proteins. Furthermore, SAB, MA, LA, and LG exhibited anti-thrombotic effects in the chemical-induced zebrafish thrombosis model. This study demonstrated a feasible approach for the screening of pharmacological effect of small molecules by combining a zebrafish in vivo model with molecular docking techniques. Meanwhile, our research also provided references for the further study of therapeutic effects of DS to thrombosis-related diseases.  Informed Consent Statement: Not applicable.

Data Availability Statement:
The data presented in this study are contained within the article.

Conflicts of Interest:
All the authors declare that they have no conflict of interest.
Sample Availability: Samples of all of the compounds are available from the authors.