|Year : 2020 | Volume
| Issue : 3 | Page : 317-323
Inhibition of pancreatic elastase in silico and in vitro by Rubus rosifolius leaves extract and its constituents
Yesi Desmiaty1, Esti Mulatsari2, Fadlina Chany Saputri3, Muhammad Hanafi4, Rini Prastiwi5, Berna Elya3
1 Department of Phytochemistry, Faculty of Pharmacy, Pancasila University, Jakarta, Indonesia; Department of Phytochemistry, Faculty of Pharmacy, Universitas Indonesia, Depok, Indonesia
2 Department of Phytochemistry, Faculty of Pharmacy, Pancasila University, Jakarta, Indonesia
3 Department of Phytochemistry, Faculty of Pharmacy, Universitas Indonesia, Depok, Indonesia
4 Department of Phytochemistry, Faculty of Pharmacy, Pancasila University, Jakarta, Indonesia; Research Centre for Chemistry Indonesian Institute of Sciences, Jakarta, Indonesia
5 Department of Pharmacognosy, Faculty of Pharmacy, Universitas Muhammadiyah Prof. Dr. Hamka, Jakarta, Indonesia
|Date of Submission||06-Nov-2019|
|Date of Decision||29-Apr-2020|
|Date of Acceptance||11-May-2020|
|Date of Web Publication||20-Jul-2020|
Prof. Berna Elya
Faculty of Pharmacy, Universitas Indonesia, Jakarta, 16424
Source of Support: None, Conflict of Interest: None
| Abstract|| |
Objective: Elastases are protease enzymes, which mainly hydrolyze proteins of the connective tissue, so they have a significant impact on human disease. Rubus rosifolius is one of the Rubus species found in Indonesian mountains, and it has potential as an elastase inhibitor. The objective of this research was to examine the in vitro elastase inhibitor activity of R. rosifolius leaves and to dock different ligands of its constituents against target protein of Porcine Pancreatic Elastase (PPE) receptor. Method: Dried leaves powder of R. rosifolius was extracted using Soxhlet apparatus with n-hexane, ethyl acetate, and methanol. The extract was evaporated, and in vitro elastase inhibitor activity was determined using PPE with the quercetin used as control positive. Selected nine constituents of R. rosifolius were evaluated on the docking behavior of elastase receptor using Protein–Ligand ANT System (PLANTS) computational software with PPE enzyme with Protein Data Bank (PDB) file 1BRU. Result: The methanol extract showed significantly inhibited elastase with IC50 186.13 μg/mL, but ethyl acetate extract showed weak activity, and n-hexane extract did not show any activity. Docking studies and binding free energy calculations and hydrogen bonding with some amino acids revealed that ellagic acid showed the least binding energy for the target enzyme. Conclusion: This research has opened new insights into understanding that constituents of R. rosifolius methanol extract are potential inhibitors against elastase, and suggested the active compound is ellagic acid.
Keywords: Docking studies, elastase inhibitor, ellagic acid, Rubus rosifolius
|How to cite this article:|
Desmiaty Y, Mulatsari E, Chany Saputri F, Hanafi M, Prastiwi R, Elya B. Inhibition of pancreatic elastase in silico and in vitro by Rubus rosifolius leaves extract and its constituents. J Pharm Bioall Sci 2020;12:317-23
|How to cite this URL:|
Desmiaty Y, Mulatsari E, Chany Saputri F, Hanafi M, Prastiwi R, Elya B. Inhibition of pancreatic elastase in silico and in vitro by Rubus rosifolius leaves extract and its constituents. J Pharm Bioall Sci [serial online] 2020 [cited 2021 Jul 25];12:317-23. Available from: https://www.jpbsonline.org/text.asp?2020/12/3/317/290122
| Introduction|| |
Elastases are proteolytic enzymes, a serine protease, that degrade a wide variety of connective tissue proteins in the lungs, arteries, skin, and ligaments such as elastin. Pancreatic elastase (PE) and neutrophil elastase (NE) represent two of these elastin-cleaving enzymes. Elastases predominate in the pathogenesis of emphysema, acute pancreatitis, rheumatoid arthritis, thrombosis, and stroke in the absence of suitable inhibitors. Elastase-1 is mainly expressed in skin, although it was first found as foreign material during phagocytosis.,, The development of in vitro methods to test the inhibition of elastase enzyme also can be used in antiaging activity screening tests,, and this research can also be used to develop a nutraceutical food. Some compounds are known to have acted as elastase inhibitor, that is, epigallocatechin gallate, catechin, procyanidin, and quercetin.,,,
Indonesia is a tropical country and has many mountains with a lot of plants unexplored, including chemical and biological activities. Rubus rosifolius is one of the Rubus species found in Indonesian mountains. This fruit is economically important as fruit crops and is commercially sold as fresh fruit in the Tangkuban Perahu mountain tourism area at West Java. R. rosifolius fruit was reported to contain flavonoid, sterol, triterpenoid, and had antibacterial activity, antinociceptive, and antiproliferative. Leaves of R. rosifolius were reported to have 5,7-dihydroxy-6, 8, 4’-trimethoxyflavonol, 5-hydroxy-3, 6, 7, 8, 4’-pentamethoxyflavone, and tormentic acid, which had antiproliferative activity against ovary cancer cell. The main isolated compound from the hexane extract of R. rosifolius herb was 28-methoxytormentic acid and was reported to have potential analgesic activity. Others reported that essential oil of R. rosifolius contained β-caryophyllene, dihydroagarofuran, isokessane and β-kessane, ellagic acid, euscaphic acid, pomolic acid, β-sitosterol d-glucoside-6′-acetate, trachelosperogenin A, and some had activity against the carcinogen-activating CYP1B1 enzyme.
Some Rubus were reported to have potential activity as inhibitor enzyme such as elastase, collagenase, hyaluronidase, and tyrosinase.,,,, So, the objective of this research was to examine the potential of R. rosifolius leaves as an elastase inhibitor and to dock different ligands of its constituents against target protein porcine pancreatic elastase (PPE) receptor. In this research, Suc-AAA-pNA (SANA) was used as a substrate; this method was established, and the hydrolysis mechanism also was known. After protease action, this substrate releases p-nitroaniline, which is detectable by a microplate reader at 410nm.
The docking process was carried out on the reported compound of R. rosifolius. Nine constituents hypothesized to inhibit the work of PPE enzymes, were chosen. The nine constituents were tormentic acid, pomolic acid, euscaphic acid, ellagic acid, rosifoliol, β-caryophyllene, trimethoxyflavonol, pentamethoxyflavonol, and dihidroagarofuran. Molecular docking can be used to know the potential of compounds to be a drug candidate based on the affinity of binding to target proteins. Binding affinity is a critical aspect to be considered in molecular and macromolecular interactions. This binding affinity was illustrated by the value of the docking score. Lower binding affinities indicate that a compound requires less energy to engage in binding or interaction. In other words, the lower affinity value of binding increases the potential for binding to the target protein., One docking method that has been widely used is the method of docking using software PLANTS (Protein–Ligand ANT System). This method has been widely used in several in silico studies such as in silico tests of antimalarial, anti-inflammatory, antioxidant, and antibacterial activity.,
| Materials and Methods|| |
Chemicals and reagents
The chemical and reagents used in this study were buffer Trizma base (T1503; Sigma–Aldrich, St. Louis, MO), PPE (E1250; Sigma–Aldrich), substrate N-succinyl-Ala-Ala-Ala-p-nitroanilide (SANA) (S4760; Sigma–Aldrich), quercetin (Sigma-Aldrich), and aquadest.
R. rosifolius leaves were collected from Mount Tangkuban Perahu, West Java. The specimen was authenticated by a botanist at Research Center for Biology, Indonesian Institute of Sciences (LIPI), Cibinong, Indonesia. Fresh samples were cleaned, air-dried, and grounded into a fine powder by laboratory mill.
The leaves were extracted using Soxhlet apparatus by using three different solvents, that is, n-hexane, ethyl acetate, and methanol. The organic solvents were evaporated using a rotary vacuum evaporator and then were dried using a vacuum oven.
Elastase inhibitory assay
The assay was performed as described by Kraunsoe et al. with slight modification. Elastase inhibitory assay using PPE with the substrate SANA. The product reaction, p-nitroaniline for 15 min at room temperature, was monitored by measuring the absorbance at 401nm with a microplate reader (VersaMax™ Microplate Reader (USA)). The reaction mixture contained 70mM Trizma–HCl buffer (pH 8.0), 16 mU PPE, and 0.29mM SANA, as shown in [Table 1]. It was preincubated for 15 min at 25°C, and the reaction was started by adding substrate. Blanks contained all the components except the enzyme. Quercetin was used as a positive control. The experiments were done in triplicate. The percentage of elastase inhibition was expressed as a percentage of inhibition of elastase activity.
Hardware and software required
The hardware that was used to docking process is a laptop with AMD A6 processor, 4 GB RAM, Windows 10, 64-bit operating system.
The software used in drawing ligand structures is MarvinSketch. The docking process was done using PLANTS 1.1_mingwm.exe. Protein and ligand preparation using YASARA free software. The docking results are visualized using PyMOL.
Chemical structures of two ligands as control positive were (1) gallic acid (CID370) and (2) quercetin (CID5280343). Chemical structures of the nine selected ligands, namely (1) tormentic acid (CID73193), (2) pomolic acid (CID382831), (3) euscaphic acid (CID471426), (4) ellagic acid (CID5281855), (5) rosifoliol (CID527256), (6) betacaryophyllene (CID5281515), (7) trimethoxyflavonol (CID14606539), (8) pentamethoxyflavonol (CID10453852), and (9) dihidroagarofuran (CID10775429). All structures were retrieved from PubMed compound databases (www.pubmed.com). All of the test ligands and 2-(2-hydroxy-cyclopentyl)pent-4-enal as native ligand were prepared to form their conformations using MarvinSketch and saved with file name ligand.mol2 and ref_ligand.mol2.
Target protein preparation
The three-dimensional structures of the PPE enzyme with PDB ID: 1BRU (DOI: 10.2210/pdb1BRU/pdb) were obtained from www.rcsb.org was pre-processed. Protein and ligand are separated by YASARA software. Protein saved as protein.mol2 and the ligand save as ref_ligand.mol2.
The docking process was done using the modification of the standard procedure of molecular docking using PLANTS operating system., The docking protocol validation is performed by calculating the root mean square deviation (RMSD) values between the reference ligand (2-(2-hydroxy-cyclopentyl)pent-4-enal) of PDB ID: 1BRU with the conformation of the redocked ligand. Docking protocol is considered good and can be used for the further docking process, if it has a value less than 3 Å; the alignment closer to 0 is considered better. Binding site center and binding site radius of protein obtained in the process of redocking the reference ligand was input in the plantsconfig.file configuration. It was used as a validated docking protocol. Validated protocol (plantsconfig.file), ref_ligand.mol2, protein.mol2, and ligand.mol2 were prepared as input data. The docking process was done by typing commands on cmd worksheet. PLANTS 1.1_mingwm.exe software will run to calculate the value of the docking score according to the docking protocol that has been validated. The docking process obtained docking scores as output data that showed the energy of the ligand in binding to the target protein. The more negative the docking scores, the affinity of the ligand binding to the protein was stronger. Inhibition constant was determined by the Gibbs equation that was ∆G = RTLnKi.
| Results and Discussion|| |
Extraction and elastase inhibitory assay
The dried extracts were weighed, and the yield of each extract was calculated. The extraction rendement of n-hexane, ethyl acetate, and methanolic extracts of R. rosifolius was 2.11%, 6.93%, and 6.23%, respectively. Elastase inhibitory assay showed that methanol extract gave the best activity in 100 µg/mL extract. The results are shown in [Table 2]. There are no previous reports available on the elastase inhibitor activity of this species. In vitro elastase activity was assayed using SANA as a substrate. This substrate was chosen because it is specific to the proteolytic activity of PPE, and it was used in similar works present in the literature.
Molecular docking evaluation
It has been reported in Rubus species that there are phenolic compounds such as ellagic acid (usually found as glycosylated glycosylation polymers), gallic acid, chlorogenic acid, and caffeic acid. The docking score of compounds in R. rosifolius leaves extract can be seen in [Table 3]. The RMSD value was obtained from the redocking 1BRU PDB.ID to form the validated docking protocol, which was 3 Å; this value is eligible for the protocol to be used for further docking process. This value ensures that the docked ligand was in the right position in the binding pockets of the receptor with a minimum shift to the reference ligand position. The alignment between the reference ligand and the conformation of the docked ligand of R. rosifolius inhibition active compound is shown in [Figure 1].
|Figure 1: Position of reference ligand (2-(2-hydroxy-cyclopentyl)pent-4-enal) (red), Rubus rosifolius inhibition active compound (ellagic acid in yellow color, rosifoliol in purple color) in the binding pocket of 1BRU.pdb receptor (drawn with PyMol)|
Click here to view
The docking process of 11 compounds (as depicted in [Figure 2]) showed that two compounds could inhibit PPE enzyme with PDB ID: 1BRU [Table 3]. The gallic acid and quercetin as control positive and rosifoliol had lower docking scores than 2-(2-hydroxy-cyclopentyl)pent-4-enal in inhibiting elastase enzyme (1BRU). Ellagic acid had the lowest docking score in inhibiting the elastase enzyme (1BRU).
|Figure 2: Chemical structure of 11 test compounds. (A) 2-(2-hydroxy-cyclopentyl)pent-4-enal. (B) Gallic acid. (C) Quercetin. (D) Euscaphic acid. (E) Tormentic acid. (F) Pomolic acid. (G) Rosifoliol. (H) β-Caryophyllene. (I) Ellagic acid. (J) Dihydroagarofuran. (K) Trimethoxy flavonol. (L) Pentamethoxyflavonol|
Click here to view
The binding interactions of the most active docked conformation of the test ligands and the target proteins have been identified using PyMol. By checking one by one all amino acids within 4 Å of the active site of the target protein, the critical binding interactions were identified. The interaction of ligands with the binding pocket receptor is given in [Figure 3].
|Figure 3: The ligand in the binding pocket of receptor 1BRU.pdb, the hydrogen bond in yellow line, protein chain in cyan, the ligand in green C–C bond. (A) Quercetin (control positive). (B) Gallic acid. (C) Ellagic acid. (D) Rosifoliol (drawn with PyMol)|
Click here to view
The negative value of binding energy change (ΔG) reveals that the binding process is spontaneous; it can fit well in the binding pocket receptor forming the most stable drug receptor energetically. The more negative of binding energy value of a compound is more feasible it is used as a drug. The molecular docking results showed that control positive and two of test compounds in R. rosifolius have a docking score that was more negative than reference ligand (2-(2-hydroxy-cyclopentyl)pent-4-enal), which shows that the constituent in R. rosifolius potentially inhibits PPE enzyme. Ellagic acid and rosifoliol have the lowest docking score and inhibition constant; it was indicated that they have the best activity compared to other compounds in inhibiting PPE [Table 3]. Lower score docking and inhibition constant make the formed complex protein–ligand to be more stable and stronger. The interaction of ligand (gallic acid, quercetin, ellagic acid, and rosifoliol) with amino acids on 1BRU binding pocket proteins showed some hydrogen bonds [Table 4]. Some amino acids play a role in the formation of hydrogen bonds between the ligand and active amino acids of protein, that is, SER_190, ASN_192, SER_214, GLY_216, SER_195, CYS_191, HIS_57, and SER_217. Ellagic acid has the most affinity because it has the most number of hydrogen bonds. The SER_195 and HIS_57 are active amino acids that in accordance with the active amino acids form hydrogen bonds between proteins and reference ligand (2-(2-hydroxy-cyclopentyl)pent-4-enal) in PDB data, as depicted in [Figure 4]. Visualization results of ligand–protein interaction showed that the presence of the –OH and O groups in the ligand allows more hydrogen bonds to form. Besides hydrogen bonding, electronic bonding, hydrophobic, and van der Waals interaction have influenced the activity of ligand inhibiting the receptor.
|Table 4: Hydrogen bond of ligand and active amino acid in binding pocket protein 1BRU|
Click here to view
|Figure 4: Hydrogen bond (blue color) of 2-(2-hydroxy-cyclopentyl)pent-4-enal with active amino acid residues of 1BRU.pdb ID protein from PDB database|
Click here to view
Some terpene constituents (such as β-caryophyllene, dihydroagarofuran, and tormentic acid) were isolated from n-hexane fraction. Trimethoxyflavonol and pentamethoxyflavonol were isolated from dichloromethane fraction., The docking test showed that all compounds have larger Ki, and the in vitro test showed that there is no inhibition activity from n-hexane extract. Pomolic acid and euscaphic acid were isolated from ethyl acetate extract. Both of these compounds showed a large Ki even though ethyl acetate extract showed weak activity. Ellagic acid and gallic acid were the most predominant of the phenolic compound isolated from methanol extract. Docking test and in vitro test showed that methanol extract has potential activity as an elastase inhibitor.
For the external preparation can be widely used in cosmetics to prevent skin aging and to maintain the skin in a youthful and healthy state; and also as a nutritional food. In these experiments, it was concluded that methanol extract showed potential activity as an inhibitor against elastase, and molecular docking to protein pancreatic porcine elastase (PPE) receptor (1BRU.pdb) indicated that the ellagic acid was suspected as one of the most active compounds. The result is also consistent with other studies that show that methanol is an efficient extraction medium for a broad spectrum of compounds.
We would like to acknowledge the financial assistance obtained from the “Publikasi Terindeks Internasional (PUTI) Doktor Universitas Indonesia” grant (NKB-525/UN2.RST/HKP.05.00/2020) in 2020.
Financial support and sponsorship
Financial assistance for this study was obtained from the “Publikasi Terindeks Internasional (PUTI) Doktor Universitas Indonesia” grant.
Conflicts of interest
There are no conflicts of interest.
| References|| |
Sivamani P, Singaravelu G, Thiagarajan V, Jayalakshmi T, Ramesh Kumar G Comparative molecular docking analysis of essential oil constituents as elastase inhibitors. Bioinformation 2012;8:457-60.
Stergiou PY, Foukis A, Gkini OA, Bieth JG, Papamichael EM Kinetic and computational analysis of the reversible inhibition of porcine pancreatic elastase: a structural and mechanistic approach. J Enzyme Inhib Med Chem 2016;31:131-9.
Brás NF, Gonçalves R, Mateus N, Fernandes PA, Ramos MJ, De Freitas V Inhibition of pancreatic elastase by polyphenolic compounds. J Agric Food Chem 2010;58:10668-76.
Mukherjee PK, Maity N, Nema NK, Sarkar BK Bioactive compounds from natural resources against skin aging. Phytomedicine 2011;19:64-73.
Desmiaty Y, Winarti W, Nursih AMM, Nisrina H, Finotory G Antioxidant and antielastase activity of Kaempferia rotunda
and Curcuma zedoaria
. Res J Chem Environ 2018;22:95-8.
Pangastuti A, Indriwati SE, Amin M Investigation of the anti-aging properties of allicin from Allium sativum
L bulb extracts by a reverse docking approach. Trop J Pharm Res 2018;17:635-9.
Thring TS, Hili P, Naughton DP Anti-collagenase, anti-elastase and anti-oxidant activities of extracts from 21 plants. BMC Complement Altern Med 2009;9:27.
Kanashiro A, Souza JG, Kabeya LM, Azzolini AE, Lucisano-Valim YM Elastase release by stimulated neutrophils inhibited by flavonoids: importance of the catechol group. Z Naturforsch C J Biosci 2007;62:357-61.
Petreanu M, Ferreira EK, Sagaz AP, Vendramini-Costa DB, Ruiz AL, De Carvalho JE, et al
. Uncommon trimethoxylated flavonol obtained from Rubus rosaefolius
leaves and its antiproliferative activity. Evid Based Complement Alternat Med 2015;2015:341216.
Kanegusuku M, Sbors D, Bastos ES, de Souza MM, Cechinel-Filho V, Yunes RA, et al
. Phytochemical and analgesic activity of extract, fractions and a 19-hydroxyursane-type triterpenoid obtained from Rubus rosaefolius
(Rosaceae). Biol Pharm Bull 2007;30:999-1002.
Southwell IA, Tucker DJ Isokessane and β-kessane from Rubus rosifolius
Sm. J Essent Oil Res 1996;8:143-8.
Campbell TF, McKenzie J, Murray J, Delgoda R, Bowen-Forbes CS Rubus rosifolius
varieties as antioxidant and potential chemopreventive agents. J Funct Foods 2017;37:49-57.
Akkol EK, Süntar I, Ilhan M, Aras E In vitro
enzyme inhibitory effects of Rubus sanctus
Schreber and its active metabolite as a function of wound healing activity. J Herb Med 2015;5:207-10.
Bravo K, Alzate F, Osorio E Fruits of selected wild and cultivated Andean plants as sources of potential compounds with antioxidant and anti-aging activity. Ind Crops Prod 2016;85:341-52.
Piwowarski JP, Kiss AK, Kozłowska-Wojciechowska M Anti-hyaluronidase and anti-elastase activity screening of tannin-rich plant materials used in traditional Polish medicine for external treatment of diseases with inflammatory background. J Ethnopharmacol 2011;137:937-41.
Iino M Elastase inhibitor contains raspberry. Patent No. US 8.491.946 B2. US8491946 B2. 2013.
Herrmann M, Joppe H, Franke H, Vielhaber G Blackberry (Rubus fruticosus
) extract as anti ageing. Patent Pub. No.: US 2008/0095719 A1. Pub. No.: US 2008/0095719 A1. 2008.
Mulatsari E, Mumpuni E, Herfian A Vitual screening and binding mode elucidation of curcumin analogues on cyclooxygenase-2 using AYO_COX2_V1. 1 protocol. J Phys Conf Ser 2017;835:1-8.
Kastritis PL, Bonvin AMJJ On the binding affinity of macromolecular interactions : daring to ask why proteins interact on the binding affinity of macromolecular interactions : daring to ask why proteins interact. J R Soc Interface 2013;10:1-28.
Seo MH, Park J, Kim E, Hohng S, Kim HS Protein conformational dynamics dictate the binding affinity for a ligand. Nat Commun 2014;5:1-7.
Baker JR, Woolfson DN, Muskett FW, Stoneman RG, Urbaniak MD, Caddick S Protein-small molecule interactions in neocarzinostatin, the prototypical enediyne chromoprotein antibiotic. Chembiochem 2007;8:704-17.
Tassa C, Duffner JL, Lewis TA, Weissleder R, Schreiber SL, Koehler AN, et al
. Binding affinity and kinetic analysis of targeted small molecule-modified nanoparticles. Bioconjug Chem 2010;21:14-9.
Tegar M, Purnomo H Tea leaves extracted as anti-malaria based on molecular docking PLANTS. Procedia Environ Sci 2013;17:188-94.
Kraunsoe JAE, Claridge TDW, Lowe G Inhibition of human leukocyte and porcine pancreatic elastase by homologues of bovine pancreatic trypsin inhibitor. Biochemistry 1996;2960:9090-6.
Korb O, Stützle T, Exner TE Empirical scoring functions for advanced protein−ligand docking with PLANTS. J Chem Inf Model 2009;49:84-96.
Morris GM, Goodsell DS, Halliday RS, Huey R, Hart WE, Belew RK, et al
. Automated docking using a Lamarckian genetic algorithm and an empirical binding free energy function. J Comput Chem 1998;19:1639-62.
Afif Chaouche T, Bendahou M, Arab K Antibacterial activity of two extracts from Rubus fruticosus
L. against resistant pathogens and their antioxidant potential. Afr J Microbiol Res 2015;9:1255-62.
Rangwala H, Karypis G fRMSDPred: predicting local RMSD between structural fragments using sequence information. Proteins 2008;72:1005-18.
Kumar KM, Anitha P, Sivasakthi V, Bag S, Lavanya P, Anbarasu A, et al
. In silico
study on penicillin derivatives and cephalosporins for upper respiratory tract bacterial pathogens. 3 Biotech 2014;4:241-51.
Balavignesh V, Srinivasan E, Babu NR, Saravanan N Molecular docking study ON NS5B polymerase of hepatitis c virus by screening of volatile compounds from Acacia concinna
and ADMET prediction. Int J Pharm Life Sci 2013;4:2548-58.
Kinoshita T, Yamaguchi A, Tada T Tris(hydroxymethyl)aminomethane induces conformational change and crystal-packing contraction of porcine pancreatic elastase. Acta Crystallogr Sect F Struct Biol Cryst Commun 2006;62:623-6.
Khalid A Design and molecular docking of antioxidant lead compound and its analogues acting as human tyrosine kinase inhibitors. IOSR J Pharm Biol Sci 2013;5:75-80.
Jin YJ, Pyo YH Effect of monascus-fermented soybean extracts on antioxidant and skin aging-related enzymes inhibitory activities. Prev Nutr Food Sci 2017;22:376-80.
[Figure 1], [Figure 2], [Figure 3], [Figure 4]
[Table 1], [Table 2], [Table 3], [Table 4]