Cardioprotective Effect of Troxerutin: An In silico Study

Received 19 Sep, 2019

Accepted 16 Dec, 2019

Published 15 Jan, 2020

ABSTRACT

Background and Objective: Cardiovascular diseases are the leading causes of death worldwide. Therefore, the search for therapeutic alternatives is necessary, and in silico studies constitute an important tool for the development of cardioprotective drugs. The aim of this study was predicting the possible target which could explain the cardioprotective effects of the natural chemical marker troxerutin. Materials and Methods: Compounds with cardioprotective effects were selected using PubMed, ScienceDirect, and SciELO. The Molispiration server was used for oral bioavailability analysis, and Pred-HERG and Pro Tox-II were used to predict cardiac and systemic toxicities. Analyses of biological activities were performed in PASS online and those of molecular targets in Swiss Target Prediction and Superpred. Finally, pharmacophoric analysis and molecular docking were performed. To validate the model employed in molecular docking, redocking was also performed. Troxerutin, a natural flavonoid derivative, is found in the Brazilian Cerrado, extracted from Sophora japonica and Dimorphandra gardneriana. This molecule had the best scores and was thus selected for further analyses. To define the role of troxerutin as a thrombin antagonist, a pharmacophoric mapping was performed using five thrombin inhibitors with the lowest IC₅₀ values. Results: The results of PASS online corroborated with the cardioprotective effects of troxerutin, which were further confirmed by using thrombin, one of the targets obtained in the prediction studies. This interaction was verified by troxerutin docking with the human thrombin crystallographic structure (PDB ID: 2A2X). Redocking analysis established the parameters of the interaction model. As a result of thrombin antagonist, troxerutin interacted strongly with glycine and serine residues. Conclusion: Troxerutin is a promising thrombin inhibitor as demonstrated by pharmacophoric model and molecular docking analysis, which account for its use as a medical tool against cardiovascular disease.

INTRODUCTION

Worldwide, cardiovascular diseases (CVD), are the leading causes of death, and the total number of deaths from CVD was approximately 31% or 17.7 million in 2015. According to epidemiological data, CVDs kill twice more than cancers, 2.5 times more than accidents and deaths from violence, and 6 times more than infectious causes¹. The main cause of death from CVDs can be due to either ischemic heart disease or cerebrovascular disease². The other categories of CVDs include rheumatic heart disease, hypertensive heart disease, heart disease and myocarditis, atrial and flutter fibrillation, aortic aneurysm, peripheral vascular disease, and endocarditis³.

The socioeconomic impact of these pathologies is significant. In Brazil, the cost of hospitalizations for CVDs is the highest among the causes of hospitalizations. Total direct spending on hospitalizations and consultations for CVD in 2015 was more than R $ 5 billion^4,5. With the increase in life expectancy and consequent aging of the population, there is a tendency towards an increase in the incidence of CVD and, in turn, and its exponential cost.

Thrombin is a serine protease that catalyzes the conversion of fibrinogen to fibrin and regulates the coagulation cascade with the procoagulant effect, acting on hemostasis. It is closely related to thrombogenesis caused by some heart diseases, such as arrhythmias, heart failure, and valvulopathies. Thrombin inhibitors are increasingly recognized as therapeutic agents for CVD, preventing acute coronary syndrome, stroke, and venous thromboembolism⁶.

The search for alternative therapies through in silico studies of natural plant products aids in the development of drugs with cardioprotective properties and thus fosters a better therapy for CVD. For instance, acetylsalicylic acid was discovered from natural sources and is currently used to treat this set of diseases. In silico approach is based on studies of structure-activity relationship, and also explores the pharmacokinetic aspects of the candidate compounds. In addition, knowledge about the etiology and pathophysiology of diseases helps to optimize the process of discovery and development of new drugs, as it allows the identification of possible therapeutic targets that the substance may act on to stop the pathological process⁷.

Troxerutin, known as vitamin P4, is derived from a natural glycosidic and flavonoid extracted from the fruits of the plant species Sophora japonica and Dimorphandra gardneriana⁸.

Drugs from natural sources have nearly 50 % FDA approval for commercialization. Supplementation using natural products with cardioprotective action needs to be considered in populations with a high prevalence of CVD and in silico screening systems contribute highly to the identification of these compounds⁸. Thus, the main objective of this study was predicting the possible target to natural chemical marker troxerutin which could explain its cardioprotective effects.

MATERIALS AND METHODS

Study area and literature search: This study was carried out from January to October 2019 in Catholic Pontifical University of Goiás, Brazil.

The flavonoids from medicinal plant species with cardioprotective potential investigated in this article were selected from PubMed, Science Direct and SciELO databases. Seven compounds from different species were selected and their molecular structures were confirmed in the Pubchem database (Table 1).

Table 1: Compounds with cardiovascular potential selected for in silico analyses
Compound	Species
Troxerutin	Sophora japonica and Dimorphandra gardneriana
Catechin Gallet	Grape pomace
Epicatechin Gallet	Grape pomace
GalocatechinGallet	Grape pomace
Quercetin	Grape pomace
Kaempferol	Grape pomace
Aristolane	Nardostachyschinensis Batal

Drug likeness classification: To predict the pharmacokinetics properties from the molecules selected, the Molinspiration web server was used⁹. From properties identified in this server, the investigated compounds could be classified into drug likeness or non-drug likeness. The properties included hydrophobicity, electronic distribution, hydrogen bonding characteristics, molecule size and flexibility, and the presence of pharmacophoric characteristics that influence the behavior of the molecule in a living organism, including bioavailability, transport properties, protein affinity, reactivity, toxicity, stability, and others. On this server, Lipinski's rule of five was used. It states that for a compound to have good oral bioavailability (comparable to drugs already used), it must have values for four multiple parameters, such as log P greater than or equal to 5, molecular mass less than or equal to 500 Da, acceptors of hydrogen bond less than or equal to 10, and number of hydrogen bond donors less than or equal to 5¹⁰.

Prediction of toxicity analysis was performed by Pred-HERG^11–13, which predicts cardiac toxicity via HERG inhibition, and ProTox-II^14,15, which analyzes molecular similarity to assess toxicity including acute toxicity, hepatotoxicity, cytotoxicity, carcinogenicity, mutagenicity, immunotoxicity.

In silico prediction of biological properties: To predict biological activities, the PASS webserver was employed^16,17. From the structure of the molecules, their pharmacological effects, mechanisms of action, toxicity and adverse effects, interaction with enzymes and metabolic transporters, influence on gene expression and other characteristics were explored. Further, the effects that were linked to some cardioprotective activity of these molecules were also analyzed.

Molecular targets selection and docking: After the prediction of biological activities, the molecular targets were studied on two platforms, the SwissTargetPredicition¹⁸ to predict molecular targets using a combination of 2D and 3D similarity measurements and comparison of the query molecule to a library of 280,000 active compounds, and then the SuperPred webserver was used¹⁹ for target prediction with 2D and 3D similarity analysis of the compounds.

Pharmacophore perception: Pharmacophoric models were generated using the PharmaGist web server which generates 3D images of pharmacophores from a set of molecules that are known to bind to a common target receptor. The method works for possible spatial features (pharmacophores) and reports the one with the highest score. The selected pharmacophores were identified by multiple flexible alignments of the input molecule ligands, where the flexibility of these moleculeswasprocessed and in a deterministic model in the alignment process²⁰.

In relation to the parameters, a minimum of three to a maximum of six features including hydrogen-bond acceptor, hydrogen bond donor, hydrophobic, and ring aromatic features were selected for generating the pharmacophore models. The scoring weight was assigned as 3.0 for aromatic rings, 1.0 for a charge (anion/cation), 1.5 for hydrogen bond (donor/acceptor) and 0.3 for hydrophobic, which represented the default parameters of the webserver.

First, the active molecules against the same target, troxerutin, were selected for pharmacophore perception. From a total of 30 compounds, five molecules with the lowest IC₅₀ values were selected in Binding DB²¹. Next, to obtain the fitting pharmacophore models, troxerutin was imputed into the dataset, and the top-scored pharmacophore models were selected.

Molecular docking: To further prepare receptors for docking, the Hermes visualizer in the GOLD Suite 5.7.0 was used²². The region of interest used for GOLD docking was defined as all the protein residues within the 10 Å radius of the reference ligands that accompanied the downloaded protein complexes. The default values of all other parameters were used, and the complexes were submitted to 10 genetic algorithm runs using the CHEMPLP fitness function.

Statistical analysis: The Piecewise Linear Potential (fPLP) was used to model the steric complementarity between protein and ligand, while for ChemPLP additionally the distance and angle-dependent hydrogen and metal bonding terms from ChemScore are considered²².

RESULTS

Among the molecules analyzed for cardioprotective activities, classified as drug likeness and with favorable pharmacokinetic profiles, troxerutin (Fig. 1A) was the most promising and presented the highest scores in the employed servers.

As presented in Fig.1, the flavonoidspossess some known characteristics, such as antioxidant, anti-inflammatory, antineoplastic, antithrombotic and antifibrinolytic. By predicting the biological activity of this molecule in the online PASS server, the effectsthat corroborated the hypothesis that troxerutin has a cardioprotective activity, such as hemostatic effect, vasoprotection, free radical scavenging, antioxidant, antithrombotic, vasodilator, platelet adhesion inhibitor, therapy for peripheral vascular disease and other descriptions that support the applicability of this species was confirmed.

The aglyconated part of troxerutin is absorbable and thus does not present any cardiotoxicity. Evaluation of troxerutin on Pro - Tox II, revealed its toxicity at level 5, with 6 being the lowest. The average similarity was 82.08% and the prediction accuracy was 70.97%. The analysis further revealed inactive hepatotoxicity, cytotoxicity, mutagenicity, carcinogenotoxicity and other toxicities with a propensity of 61% to 97%.

Troxerutin fitted with required features of the pharmacophore model of thrombin inhibitors, and a greater part of the molecules was aligned (Fig.1B). The three features shared between troxerutin and the thrombin inhibitors with the lowest IC₅₀ values indicate that troxerutin may present this inhibitor activity, which was corroborated by molecular docking analysis.

Redocking showed that the values of the Root mean standard deviation (RMSD), used to measure the distance between the 5 best poses and the crystallized structure, were less than 1 (Fig.2), confirming the docking analysis. Thus, the model obtained here was robust to be used in the dockingas can be seen in Fig. 3 (A-B), probably the most stable pose of the interaction between troxerutin and the target selected (thrombin).

Fig. 1: (A)Structure of the troxerutin. (B)Pharmacophore features are color-coded in red for hydrogen-bond acceptors, blue for aromatic rings

Fig. 2: Redocking image for ligand ID NA09 on thrombin protein (2A2X). The image is a result of the 5 best poses

NA09: 2-((R)-1-(((S)-1-((6-CARBAMIMIDOYLPYRIDIN-3-YL)METHYLAMINO)-1-OXOPROPAN-2-YL)(METHYL)AMINO)-3-CYCLOHEXYL-1-OXOPROPAN-2-YLAMINO)ACETIC ACID (C22 H34 N6 O4).

DISCUSSION

This study showed that thrombin can be considered a potential target to the natural chemical marker troxerutin, which could explain its cardioprotective effects and this effect of the plant species that this substance is present. Besides, there is a few works in literature regarding in silico studies with troxerutin, however its binding with DNA, corroborated by in vitro assays was performed²³.

Furthermore, there is relevant evidence about the cardioprotective effects of the troxerutin, such as anti-thrombotic, anti-fibrinolytic and rheological activities. Anti-arrhythmogenic and anti-inflammatory effects was also showed by Zhang et al.²⁴ and Najafi et al.²⁵. These facts suggest the importance of elucidation of the probably mechanism of the troxerutin in cardioprotective effect.

To explore the key ligand-enzyme intermolecular interactions between troxerutin and human thrombin, molecular docking simulations were performed using a previously reported crystal structure of thrombin (PDB ID: 2A2X)²⁶. Then, molecular docking was carried out to illustrate the binding mode of the selected flavonoid and the target. As presented in Fig. 3a, troxerutin could be docked in the catalytic site of human thrombin, indicating that this compound could serve as competitive inhibitor of thrombin, explaining its cardioprotective effects²⁷.

The key interactions between troxerutin and thrombin were analyzed. Fig. 3b shows that in the catalytic site of human thrombin, troxerutin could strongly interact with two glycine (GLY255 and GLY257), and two serine (SER232 and SER253) residues by hydrogen bonding. In addition, two cysteines interact via sulfur-π (CYS228 and CYS258), and the anion-π (GLU229) interactions contribute to establishing the docking and stabilize the ligand in the catalytic site. In the protein structure obtained from PDB (2A2X), the crystalized inhibitor also presents the hydrogens bond acceptors, and displays an important role of the enzyme active site.

Perhaps the cardioprotective effects of the species Sophora japonica e Dimorphandra gardneriana could be justified by the presence of flavonoid compounds, including troxerutin, which probably contributes to beneficial properties of the extracts of this plants^8-10. Thus, this study proposed a potential mechanism that would explain the effects of these species and others that present troxerutin in its composition, besides future assays employing this chemical marker could be corroborated by the data raised in this work. The standardization of the extracts could be employing the chemical marker troxerutin to control the several processes to obtainment of liquid and dried extracts.

Fig. 3: Docked pose 1 of troxerutin within thrombin active site. This figure was generated using Pymol 1.1r1 software (A). 2D interaction diagram of troxerutin pose 1 within thrombin. This figure was generated using Discovery Studio 3.5 Visualizer (B)

CONCLUSION

CVDs have increased at an alarming rate in recent years. This has led to an increasing interest in the cardiopreventive effects of medicinal plants. The pharmacophoric approach in this study demonstrated that troxerutin is a promising thrombin inhibitor because it shares chemical features with different thrombin inhibitors. Thus, some medicinal plants such as S. japonica and D. gardneriana, which are rich in troxerutin, protect against CVDs; further studies of other plants that possess troxerutin must be encouraged to explore its cardioprotective activity. In vitro and in vivo assays using thrombin showed that it can be a starting point for the development of novel therapeutic options with thrombin-inhibitory and CVD-preventing activities. Finally, with in silico studies, the use of animals in experimental studies could decrease significantly since they would only be performed to confirm the researcher’s hypothesis.

ACKNOWLEDGEMENT

All individuals listed as authors must have contributed substantially to the design, performance, analysis, or reporting of the work and are required to indicate their specific contribution. Anyone (individual/company/institution) who has substantially contributed to the study for important intellectual content, or who was involved in the article’s drafting the manuscript or revising must also be acknowledged.

SIGNIFICANCE STATEMENT

This study discover the model to interaction between troxerutin and the possible target thrombin, that can be beneficial for rise future studies in vitro and in vivo, highlighting this natural compound as a promisor substance with therapeutically apply.

REFERENCES

1. OPAS. and OMS., 2017. Doenças cardiovasculares. Organização Pan-Americana da Saúde and Organização Mundial da Saúde, Brasilia, DF, Brasil.
2. Rocha, R.M. and W. Martins, 2017. Manual de Prevenção Cardiovascular. 1st Edn., SOCERJ., Rio de Janeiro.
3. Brant, L.C.C., B.R. Nascimento, V.M.A. Passos, B.B. Duncan and I.J.M. Bensenõr et al., 2017. Variações e diferenciais da mortalidade por doença cardiovascular no Brasil e em seus estados, em 1990 e 2015: Estimativas do Estudo Carga Global de Doença. Rev. Bras. Epidemiol., 20: 116-128.
   
4. Siqueira, A.D.S.E., A.G.D. Siqueira-Filho and M.G.P. Land, 2017. Analysis of the economic impact of cardiovascular diseases in the last five years in Brazil. Arq. Bras. Cardiol., 109: 39-46.
5. Stevens, B., L. Pezzullo, L. Verdian, J. Tomlinson, A. George and F. Bacal, 2018. The economic burden of heart conditions in Brazil. Arq. Bras. Cardiol., 111: 29-36.
6. Periayah, M.H., A.S. Halim and A.Z.M. Saad, 2017. Mechanism action of platelets and crucial blood coagulation pathways in hemostasis. Int. J. Hematol.-Oncol. Stem Cell Res., 11: 319-327.
7. Martins, A.C.V., P.D.L. Neto, M.G.D.V. Silva and V.N. Freire, 2013. Estudo in silico de potenciais compostos anticâncer derivados de biflorina. Sociedade Bras. Química.
8. Manivannan, J., T. Silambarasan, R. Kadarkarairaj and B. Raja, 2015. Systems pharmacology and molecular docking strategies prioritize natural molecules as cardioprotective agents. RSC Adv., 5: 77042-77055.
9. Jarrahpour, A., J. Fathi, M. Mimouni, T.B. Hadda, J. Sheikh, Z. Chohan and A. Parvez, 2012. Petra, Osiris and Molinspiration (POM) together as a successful support in drug design: Antibacterial activity and biopharmaceutical characterization of some azo Schiff bases. Med. Chem. Res., 21: 1984-1990.
10. Lipinski, C.A., 2000. Drug-like properties and the causes of poor solubility and poor permeability. J. Pharmacol. Toxicol. Methods, 44: 235-249.
11. Banerjee, P., A.O. Eckert, A.K. Schrey and R. Preissner, 2018. ProTox-II: A webserver for the prediction of toxicity of chemicals. Nucleic Acids Res., 46: W257-W263.
12. Banerjee, P., F.O. Dehnbostel and R. Preissner, 2018. Prediction is a balancing act: Importance of sampling methods to balance sensitivity and specificity of predictive models based on imbalanced chemical data sets. Front. Chem.
13. Drwal, M.N., P. Banerjee, M. Dunkel, M.R. Wettig and R. Preissner, 2014. ProTox: A web server for the in silico prediction of rodent oral toxicity. Nucleic Acids Res., 42: W53-W58.
14. Braga, R.C., V.M. Alves, M.F. Silva, E. Muratov and D. Fourches et al., 2015. Pred-hERG: A novel web-accessible computational tool for predicting cardiac toxicity. Mol. Inform., 34: 698-701.
15. Braga, R.C., V.M. Alves, M.F.B. Silva, E. Muratov, D. Fourches, A. Tropsha and C.H. Andrade, 2014. Tuning HERG out: Antitarget QSAR models for drug development. Curr. Top. Med. Chem., 14: 1399-1415.
16. Poroikov, V.V., D.A. Filimonov, Y.V. Borodina, A.A. Lagunin and A. Kos, 2000. Robustness of biological activity spectra predicting by computer program PASS for noncongeneric sets of chemical compounds. J. Chem. Inf. Comput. Sci., 40: 1349-1355.
17. Poroikov, V.V. and D.A. Filimonov, 2002. How to acquire new biological activities in old compounds by computer prediction. J. Comput.-Aid. Mol. Des., 16: 819-824.
18. Daina, A., O. Michielin and V. Zoete, 2019. SwissTargetPrediction: Updated data and new features for efficient prediction of protein targets of small molecules. Nucleic Acids Res., 47: W357-W364.
19. Nickel, J., B.O. Gohlke, J. Erehman, P. Banerjee and W.W. Rong et al., 2014. SuperPred: Update on drug classification and target prediction. Nucleic Acids Res., 42: W26-W31.
20. Schneidman-Duhovny, D., O. Dror, Y. Inbar, R. Nussinov and H.J. Wolfson, 2008. PharmaGist: A webserver for ligand-based pharmacophore detection. Nucleic Acids Res., 36: W223-W228.
21. Gilson, M.K., T. Liu, M. Baitaluk, G. Nicola and L. Hwang et al., 2016. BindingDB in 2015 A public database for medicinal Chemistry, computational Chemistry and systems Pharmacology. Nucleic Acids Res., 44: D1045-D1053.
22. Cole, J.C., J.W.M. Nissink and R. Taylor, 2005. Protein-Ligand Docking and Virtual Screening with GOLD. In: Virtual Screening in Drug Discovery, Shoichet, B. and J. Alvarez (Eds.)., Taylor & Francis CRC Press, Boca Raton, FL., USA.
23. Panat, N.A., B.G. Singh, D.K. Maurya, S.K. Sandur and S.S. Ghaskadbi, 2016. Troxerutin, a natural flavonoid binds to DNA minor groove and enhances cancer cell killing in response to radiation. Chem.-Biol. Interact., 251: 34-44.
24. Zhang, Z.F., S.H. Fan, Y.L. Zheng, J. Lu, D.M. Wu, Q. Shan and B. Hu, 2014. Troxerutin improves hepatic lipid homeostasis by restoring NAD+-depletion-mediated dysfunction of lipin 1 signaling in high-fat diet-treated mice. Biochem. Pharmacol., 91: 74-86.
25. Najafi, M., E. Noroozi, A. Javadi and R. Badalzadeh, 2018. Anti-arrhythmogenic and anti-inflammatory effects of troxerutin in ischemia/reperfusion injury of diabetic myocardium. Biomed. Pharmacother., 102: 385-391.
26. Ibarra, A., J. Cases, A. Bily, K. He and N. Bai et al., 2010. Importance of extract standardization and in vitro/ex vivo assay selection for the evaluation of antioxidant activity of botanicals: A case study on three Rosmarinus officinalis L. extracts. J. Med. Food, 13: 1167-1175.
27. Landim, L.P. and J.G.M. da Costa, 2018. Dimorphandra gardneriana Tulasne (Fava d’anta)-Uma abordagem etnobotânica e riscos de extinção. Rev. Biol., 9: 6-11.

This work is licensed under a Creative Commons Attribution 4.0 International License.