CC BY
57# Research Results in Pharmacology
Research Results in Pharmacology 8(1): 95-102 UDC: 615.252.349.7 DOI 10.3897/rrpharmacology. 8.76237
3 Research Article
Validation of structural-based virtual screening protocols with the PDB Code 3G0B and prediction of the activity of Tinospora crispa compounds as inhibitors of dipeptidyl-peptidase-IV
Andri Prasetiyo1, Shirly Kumala1, Esti Mumpuni1, Raymond R. Tjandrawinata2
1 Doctoral Program of Pharmaceutical Sciences, Faculty of Pharmacy, Pancasila University, Jalan Raya Lenteng Agung No.56-80, Srengseng Sawah, Jagakarsa, Jakarta 12640, Indonesia
2 Atma Jaya Catholic University, Jalan Raya Cisauk Lapan, Cisauk, Tangerang, Banten 15345, Indonesia
Corresponding author: Shirly Kumala ([email protected])
Academic editor: Oleg Gudyrev ♦ Received 7 October 2021 ♦ Accepted 9 March 2022 ♦ Published 31 March 2022
Citation: Prasetiyo A, Kumala S, Mumpuni E, Tjandrawinata RR (2022) Validation of structural-based virtual screening protocols with the PDB Code 3G0B and prediction of the activity of Tinospora crispa compounds as inhibitors of dipeptidyl-peptidase-IV. Research Results in Pharmacology 8(1): 95-102. https://doi.org/10.3897/rrpharmacology.8.76237
Abstract
Introduction: Brotowali (Tinospora crispa) has been traditionally used as an antidiabetic drug. DPP-IV inhibitor as an antidiabetic will increase insulin secretion. It indirectly escalates incretin hormones, such as Glucagon-Like peptide-1 (GLP-1) which depends on glucose. This study predicts potential compounds from the Brotowali plants, such as DPP-IV inhibitors, using the Molegro Virtual Docker (MVD).
Materials and methods: Before the molecular docking simulation, internal validation and external validation are necessary. Internal validation was carried out by re-docking the native ligands in the DPP-IV enzyme crystal structure (PDB codes 3G0B, 3W2T, and 3BJM). The external validation was carried out by simultaneous docking of 59 active compounds and 1918 inactive compounds (decoys) from the A Directory of Useful Decoys (DUD) database with PDB code 3G0B on 16 combinations, four search algorithms, and four functions scoring.
Results and discussion: The molecular docking simulation was carried out on 50 compounds from the Brotowali plant and alogliptin as standard compounds with PDB code 3G0B. The best results of the docking method validation yielded the RMSD values of 0.43 and EF1% of 20.34 and EF20% of 3.1 (the combination of search algorithm Moldock optimizer and scoring function Moldock score). The re-rank score of 5 compounds from the Brotowali plant (Rumphioside C, Borapetoside E, Borapetoside F, Rumphioside I, and 6'-O-Lactoyl Borapetoside B) were -107.7 kcal/mol; -105.4 kcal/mol; -104.2 kcal/mol, and -102.8 kcal/mol. Alogliptin (standard ligands) had a re-rank score of -101.6 kcal/mol. The combination of search algorithms MolDock optimizer and scoring function MolDock score is a valid protocol with a good result. The similarity of the binding sites of Borapetoside E and 6'-O-Lactoyl Borapetoside B is 75% when compared to the alogliptin binding sites (Glu 205, Glu 206, Tyr 547).
Conclusion: Based on the re-rank score and binding sites similarity, Borapetoside E and 6'-O-Lactoyl Borapetoside B have potential as an antidiabetic drug with a mechanism of action of DPP-IV inhibitors.
Keywords
antidiabetic, DPP-IV inhibitors, Tinospora crispa, validation, Molegro Virtual Docker.
Copyright Prasetiyo A etal. This is an open access article distributed under the terms of the Creative Commons Attribution License (CC-BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Introduction
Diabetes is a chronic condition where the glucose level in the blood increases because it cannot produce the hormone insulin or this hormone does not work effectively (IDF 2017). The incretin hormone, glucagon-like peptide-1 (GLP-1), stimulates insulin secretion in response to glucose from food, inhibits glucagon secretion, and promotes proliferation of pancreatic beta cells. Active GLP-1 can be degraded rapidly by DPP-IV. DPP-IV inhibitors increase the plasma concentration of active GLP-1 and induce insulin secretion in response to increased blood glucose levels (Yoshida et al. 2012). DPP-IV inhibitors have a favorable weight-neutral profile with the minimal hypoglycemia risk (Taylor and Lam 2020).
Brotowali has been used in traditional medicine to treat diabetes (Klangjareonchai et al. 2015; Ahmad et al. 2016). The in vitro test results on the Brotowali extract activity in DPP-IV inhibitors indicate that the extract has activities as a DPP-IV inhibitor (Riyanti et al. 2016). Studies have been carried out on the constituents from Brotowali stems plant. More than 50 compounds have been isolated and identified as terpenoids (Cycloeucalenol, Cycloeucalenone, Tinocrispol A, Bo-rapetol A, Borapetol B, 2-O-Lactoylborapetoside B, 6'-O-Lactoylborapetoside B, Borapetoside A, Borape-toside B, Borapetoside C, Borapetoside D, Borapetosi-de E, Borapetoside F, Borapetoside G, Borapetoside H, Rumphioside A, Rumphioside B, Syringin, Columbin), flavonoids (Apigenin, Diosmetin, Genkwanin), alkaloids (Magnoflorin, N-Formylanonaine, N-Acetylanonaine, N-Formylnornuciferine, N-Acetylnornuciferine, Lysica-mine, Tyramine, Higenamine, N-cis-Feruloyltyramine, N-trans-Feruloyltyramine, Paprazine, Columbamine, Di-hydrodiscretamin, Palmatine, Jatrorrhizine, Berberine, Salsolinol, (-)-Litcubinine), lignan (Secoisolariciresinol, Syringaresinol, Adenosine, Uridine, Adenine), and sterol (P-sitosterol, Stigmasterol, Makisterone C) (Praman et al. 2012; Ahmad et al. 2016).
The DPP-4 inhibitors that have been FDA approved are vildagliptin, saxagliptin, teneligliptin, sitagliptin, alogliptin, and linagliptin. The co-crystal structures of five inhibitors DPP-IV with native ligand have been reported PDB: 1X70, 3G0B, 3BJM, 3W2T, and 3VJK (Nabeno et al. 2013). Virtual Screening based on molecular docking has been used in drug discovery design to understand drug-receptor interaction. However, many issues often occur in these studies, including errors in selected location of the binding site and docking poses (Bielska et al. 2011; Chen 2015). In some cases, docking accuracy may change from 0% to 92.66% (Chen 2015). The virtual screening predictions should be evaluated to ensure reliability and determine the proper docking protocols to use (Lagarde et al. 2015). The retrospective virtual screening method assesses two main criteria: the accuracy of the binding site predictions and the enrichment of the benchmarking data in the active compound (Lagarde et al. 2015).
Molegro Virtual Docker (MVD) is a protein-ligand docking simulation program that allows us to carry out docking simulations in a fully integrated computational package. MVD is proven to apply to hundreds of protein docking performances similar to other docking programs, such as AutoDock4 and AutoDock Vina (Bitencourt-Ferreira and de Azevedo 2019). The MVD version 6 has four search algorithms: MolDock Optimizer (MDO) (based on differential evolution), MolDock Simplex Evolution (MDSE) (a modified algorithm based on Nelder-Mead local search algorithm), Iterated Simplex (IS) (based on Nelder-Mead algorithm), and iterated simplex with ant colony optimization (ISACO). It is also possible to choose four scoring functions in each search algorithm (Bitencourt-Ferreira and de Azevedo 2019). In summary, we have 16 combinations of four search algorithms and four scoring functions.
This research aimed to validate docking protocol using MVD (Sixteen docking protocol combinations of four search algorithms and four scoring functions), SBVS, and to find dpp4 inhibitor compounds, such as antidiabetic agents from the Brotowali plants.
Materials and methods
Hardware, software, and webserver
Personal computer, Intel Core i5-9400F [email protected] 2.90GHz, RAM 16.0 Gb, NVIDIA GeForce GTX 1660. Microsoft Windows 10 Pro 64-bit, OS Ubuntu 18.04.4 LTS, Open Babel GUI 3.1.1, Molegro Virtual Docker (MVD) 6.0, ChemBioDraw Professional 16.0, ChemBio3D Professional (PerkinElmer Inc. Cambridge, MA, USA), Protein Data Bank (https://www.rcsb.org/), PubChem (https://pubchem.ncbi.nlm.nih.gov/), A Directory of Useful Decoys (DUD) (http://dude.docking.org).
Selection of proteins
We use the X-ray crystal structure of the DPP-IV enzyme (PDB code 3G0B, 3BJM, and 3W2T) was retrieved from the protein data bank https://www.rcsb.org/. Proteins were selected for the docking study by the following criteria: the obtained structure with X-ray diffraction from a human cell has a resolution below 2.5A (Chakraborti et al. 2021) and has a native ligand that has been approved by FDA.
Internal validation
Validation of the docking (Castro-Alvarez et al. 2017) method was carried out by re-docking the native ligands in the DPP-IV enzyme crystal structure (PDB Code 3G0B, 3BJM, and 3W2T). We used Molegro Virtual Docking software with a default setting, whereby appropriate missing hydrogen atoms were added, missing bonds were assigned, partial charges were added if necessary, and
flexible torsions in ligands were detected. We identified five cavities as potential binding sites. However, only one cavity was used for the ligand-docking study. The maximum number of poses in redocking was five with ten replications and it was run by sixteen combinations of four search algorithms and four scoring functions. The validation method was successful when the Root Mean Square Deviation (RMSD) value was less than 1 A (Dhananjayan 2015).
External validation
The validation of the docking method was carried out by concurrent docking of 59 active compounds and 1918 inactive compounds in the DPP-IV enzyme crystal structure (The best DPP-IV enzyme crystal structure was obtained from the internal validation). The active and inactive compounds (decoys) were downloaded in the SMILES format from the DUD-E website and then stored on a local server with actives_final.ism and decoys_final.ism. These compounds (active and decoys) are then converted into forms *.smi and added hydrogen. Then, the preparation was conducted by adjusting the pH to 7.4 and all * .smi files are converted into 3D (in the form * mol2). These stages were carried out using the Open Babel 3.11 program. The active binding site region was defined as a spherical region that encompasses all proteins within 10.0 A of bound crystallographic ligand atom with the selected coordinates of X (42.21), Y (34.47), and Z (14.97) axes, respectively. Default settings were used for all the calculations. The maximum number of poses in redocking was five with ten replications and it was run by sixteen combinations of four search algorithms and four scoring functions.
For the enrichment results, we were using three enrichment indicators: EFmax (maximum enrichment factor), EF1 (enrichment factor at 1% of the ranked database), and EF20 (enrichment factor at 20% of the database). EF max and EF1 present the early enrichment, while EF20 presents the late-stage database screening (Granchi et al. 2015).
Ligand preparation
The structures of the compounds in the Brotowali plant and standard compounds were drawn using the Chem 2D Professional 16.0 program or downloaded on Pubchem, then continued in 3D with the ChemDraw 3D 16.0 program. The energy was optimized using MMFF94 tools on Chem3D 16.0, aiming to obtain minimum energy. The file was stored in MDLMolFile (* mol) and used for the docking process (Bitencourt-Ferreira and de Azevedo 2019).
Molecular docking
A total of fifty compounds in the Brotowali plants and alogliptin (standard ligands) were docked simultaneously in the DPP-IV enzyme crystal structure (PDB Code 3G0B). The active binding site region was defined as a spherical
region encompasing within 10A of bound crystallographic ligand atom with a size of X: 42.21, Y: 34.47, and Z: 14.97 axes, respectively. Default settings were used for all the calculations. Docking was performed using MolDock Optimizer algorithms, and for each of the 50 independent runs, a maximum number of2000 iterations were executed on a single population of 50 individuals. Re-docking five poses with fifty replications run best combinations of search algorithms MolDock Optimizer (based on differential evolution) and scoring function MolDock score. The resulting conformations were clustered, and only the negative lowest-energy representation from each cluster was returned when the docking run was completed. For analysis, one pose with the lowest value of Re-rank Score was selected as the best solution for each complex.
Results and discussion
Internal validation
This study employs three DPP-IV inhibitor crystal structures bound to different native ligands; PDB code 3G0B (Zhang et al. 2011), PDB code 3BJM (Metzler et al. 2008), and PDB code 3W2T (Nabeno et al. 2013). The internal validation of the docking method was done by re-docking the native ligands with the DPP-IV inhibitors (3G0B, 3BJM, and 3W2T), using 16 combinations of the four search algorithms and four scoring functions. Root Mean Square Deviation (RMSD) is the most commonly used quantitative measure of the similarity between two superimposed atomic coordinates (Kufareva and Abagyan 2012). We used the following the criterion to evaluate re-docking success: the root-mean-square deviation (RMSD) between the crystallographic position for the ligand and the pose (generated by MVD). The best results could be obtained when RMSD values were less than 1.0 A° compared to crystallographic structures in docking simulations.
The RMSD value of re-docking the native ligand with proteins PDB code 3G0B showed the RMSD values < 1A in all the docking protocol combinations (Table 1). It is shown that the sixteen docking protocols can accurately position a DPP-IV inhibitor on the DPP-IV binding site.
The RMSD value of re-docking the native ligan with PDB Code 3BJM re-docking showed an RMSD value < 1A on several docking protocols (Table 2). It is shown that the six docking protocols can accurately position a DPP-IV inhibitor on the DPP-IV binding site. We chose PDB Code 3G0B for external validation using MVD, considering that it has an RMSD value < 1A in all the docking protocol combinations and has a higher resolution value than PDB code 3BJM. High-resolution structures are highly ordered, and it is easy to see the accuracy of every atom in the electron density map level (Castro-Alvarez et al. 2017). We did not choose PDB Code 3W2T for external validation using MVD, considering that it has an RMSD value> 1A in all the docking protocol combinations.
Table 1. Root-Mean-Square Deviation value for Protein Data Bank Code 3G0B (native ligands alogliptin, resolution 2.25 A)
Table 3. Enrichment performance for each matrix unit
Algorithm RMSD (Angstrom)
Score MolDock MolDock Itereted GPU Screening
Optimizer SE Simplex (CUDA)
MolDock 0.43 0.43 0.42 0.48
MolDock 0.68 0.84 0.85 0.53
(GRID)
PLANTS 0.45 0.48 0.44 0.59
PLANTS 0.39 0.39 0.39 0.55
(GRID)
Algorithm RMSD (Angstrom)
Score MolDock MolDock Itereted GPU Screening
Optimizer SE Simplex (CUDA)
MolDock 0.45 1.32 1.89 2.03
MolDock 0.88 0.88 0.89 2.13
(GRID)
PLANTS 1.36 1.23 1.89 1.98
PLANTS 0.60 0.60 1.37 2.12
(GRID)
Enrichment performance EF EF20%
Very good > 30 >3
Good 20-30 2.5-3
Medium 10-20 2-2.5
Poor < 10 < 2
Note: RMSD - Root-Mean-Square Deviation; PDB - Protein Data Bank; GPU - Graphics Processor Unit; CUDA - Compute Unified Device Architecture; SE - Simplex Evolution.
Table 2. Root-Mean-Square Deviation value for Protein Data Bank Code 3BJM (Native ligands saxagliptin, resolution 2.35 Â
Note: EF — Enrichment Performance.
1% of the ranked database), and EF20% (enrichment factor at 20% of the database). EFmax and EF1% were used for the early enrichment, while EF20% was used for the late-stage database screening.
Based on EFmax and EF20% (Table 3 a good validation result was obtained with six combinations of docking protocols because the EF Max value > 30 and EF20% value > 3 (Table 4).
Table 4. External validation for PDB Code 3 G0B, EF Maximal 33.5
Note: RMSD - Root-Mean-Square Deviation; PDB - Protein Data Bank; GPU - Graphics Processor Unit; CUDA - Compute Unified Device Architecture; SE - Simplex Evolution.
Algorithm EF i% and EF 20%
Score MolDock MolDock Itereted GPU Screening
OptimIzer SE Simplex (CUDA)
MolDock 20.4/3.1 13.5/3.1 15.3/3.1 3.4/2.46
MolDock 10.2/3.1 15.3/2.9 13.4/3.1 10.2/2.29
(GRID)
PLANTS 11.6/2.7 8.5/2.8 10.17/2.7 5.1/2.54
PLANTS 8.5/2.7 10.2/2.80 6.8/3.1 10.2/3.0
(GRID)
External validation
Note: PDB - Protein Data Bank; EF - Enrichment Performance; GPU - Graphics Processor Unit; CUDA - Compute Unified Device Architecture; SE - Simplex Evolution.
The DPP-IV PDB code 30B enzyme crystal structure was chosen in the external validation. Docking external validation was also performed on docking between 59 active compounds and 1918 decoys on 16 combinations of four search algorithms and four scoring functions at the DPP-IV PDB code 30B enzyme crystal structure.
In this study, we calculated three enrichment indicators: EFmax (maximum enrichment factor), EF1% (enrichment factor at 1% of the ranked database), and EF20% (enrichment factor at 20% of the database). EFmax and EF1% were used for the early enrichment, while EF20% was used at the late-stage database screening. Enrichment is defined as the number of active compounds detected at a given percentage of total decoys set by score-ranked poses (Mishra and Basu 2013). The formula used to calculate the enrichment factor is as follow (Wang et al. 2020):
EFX%
Actives
sampled
N„.
N.
sampled
Actives
Actives sampled is the number of actives found at X% of the screened database, N sampled is the number of compounds at X% of the database, N total is the number of compounds in the database, and actives total is the number of actives in the database. In this study, we calculated three enrichment indicators: EFmax (maximum enrichment factor), EF1% (enrichment factor at
We used an EF value of 1% to find the best combination of docking. The combination of docking protocols with MolDock Optimizer algorithms and scoring Function MolDock has an EF max value > 30, an EF20% value > 3, and the highest value EF1% compared to other docking protocol combinations. So, we chose the combination of docking protocol with the MolDock optimizer algorithm and scoring function MolDock for molecular docking simulations using MVD with PDB code 3G0B. Other research experiments show that MolDock has powerful properties of high docking accuracy to the identification of ligand-binding mode. MolDock was able to identify the correct binding sites of 87% of the complexes. In comparison, the accuracy of Glide and Surflex is 82% and 75%, respectively (Kufareva and Abagyan 2012).
Molecular docking simulation
The molecular docking simulation of fifty compounds from the Brotowali plant and alogliptin (standard compounds) was carried out using algorithms the MolDock Optimizer and the score function MolDock (derived from the Piecewise Linear Potential). Alogliptin is an antidia-betic drug with a DPP-IV inhibitor mechanism already on the market and approved by the Food Drug Association (FDA). Re-rank scores are calculated in Molegro Virtual
1 ^ Glu 205
, r
sèr209
'Glu 206
\ |
X \ Ser^30
rw;
Glu 20.5
\
Ah1
'N"
rGlu 206 Glu 20.5
I, Tyr 662
Lys 554 y, ,
, \ / -..Tyr 547 , 7 ^ \J*r209
/Vat 546
|i \
Gly £32
i ^ \ Serj630 Trp 629
.Arg 125
44Ai
_Tyr547 A. »)
I ?fï> 629
V Tyr b0
Tyr 662
Jyr 547
--Arg 669
s À f O
J ^ r
Glu 205
\f - r«, ^
VKTyr662
i *
^Glu206
élu 205 " tbjyresa
Figure 1. The hydrogen bond interaction of ligands and amino acid of receptor. Note: a. alogliptin b. rumphioside C c. borapetoside E d. borapetoside F e. rumphioside I f. 6'-O-Lactoyl borapetoside B g. rumphioside F and h. borapetoside D.
Docker to estimate ligand binding, where lower values are associated with higher affinity.
The seven compounds of the Brotowali plant had a lower rank than alogliptin as standard compounds (Table 5). The lower the re-rank score, the more stable and active bond between ligands and receptors. Based on the re-rank score, we predicted that the seven compounds were more active than alogliptin as the standard ligand.
The protein-ligand complex hydrogen interactions of the seven compounds after docking molecular
simulations are shown in Fig. 1. Alogliptin (standard compound) has the binding sites on four amino acid residues (Glu 205, Glu 206, Tyr 547, and Tyr 631). Alogliptin has bridges with Glu 205 and Glu 206 in the S2 subsite which have vital roles in the inhibitory activity of enzyme DPP-IV (Nabeno et al. 2013). The similarity of Rumphioside C and Borapetoside B binding sites is 50% with the alogliptin binding sites (Glu 205 and Glu 206). Whereas, the similarity of Borapetoside E and 6'-O-Lactoyl Borapetoside B
Table 5. Docking result; re-rank score, and hydrogen interaction of amino acid residues
No
Compounds
Rerank Score (Kcal/mol)
Binding sites
Binding sites similar to standard ligands
Rumphioside C
HO^ ^OH
OH
-107.7 Ser 630, His 740, Arg 125,
Glu 205, Glu 206, Ser 209
50%
Borapetoside E
-105.4 Gly 632, Glu 205, Glu 206,
Tyr 547, Tyr 662
75%
Borapetoside F
-104.2 Lys 554, Tnp 629, Val 546,
Gly 632, Tyr 547, Ser 209, Arg 125, Ser 630
25%
Rumphioside I
-102.8
Tyr 547, Tyr 662
25%
1
2
3
OH
4
6'-O-Lactoyl borapetoside B
-102.8 Tnp 629, Tyr 547, Ser 630,
Tyr 666, Glu 206, Glu 205
75%
Rumphioside F
-102.8 Tyr 662, Glu 205, Arg 669,
Tyr 547
50%
5
6
OH
No
1
Compounds
Rerank Score (Kcal/mol)
Binding sites
Binding sites similar to standard ligands
Borapetoside D
-101.8 Gly 632, Glu 206, Glu 205,
Arg 125, Tyr 662
50%
Alogliptin
-101.6 Glu 205, Glu 206, Tyr 547,
Tyr631
100%
OH
8
N
NH
binding sites is 75% with the alogliptin binding sites (Glu 205, Glu 206, and Tyr 547). Based on the re-rank score and binding sites similarity, we predicted that Borapetoside E and 6'-O-Lactoyl Borapetoside B are the best compounds as DPP-IV inhibitors.
Conclusions
Structure-Based Virtual Screening with the crystal structure of DPP-IV (PDB 3G0B) using a combination of the MolDock Optimizer algorithm and the MolDock score function on the MVD proved valid with excellent criteria. Based on re-rank score and binding sites similarity, Borapetoside E and 6'-O-Lactoyl Borapetoside B have
the potential as antidiabetic drugs with a mechanism of action DPP-IV inhibitors.
Conflict of interest
The authors have declared that no competing interests exist.
Acknowledgement
The authors would like to thank Prof. Dr. Siswandono, Apt, Department of Pharmaceuticals Chemistry, Airlangga University, for permission and support in software facilities Molegro Virtual Docker.
References
■ Ahmad W, Jantan I, Bukhari SNA (2016) Tinospora crispa (L.) Hook. f. & Thomson: A review of its ethnobotanical, phytochemical, and pharmacological aspects. Frontiers in Pharmacology 7: 1-19. https://doi.org/10.3389/fphar.2016.00059 [PubMed] [PMC]
■ Bielska E, Lucas X, Czerwoniec A, Kasprzak JM, Kaminska KH, Bujnicki JM (2011) Virtual screening strategies in drug design -methods and applications. Biotechnologia 92: 249-264. https://doi. org/10.5114/bta.2011.46542
■ Bitencourt-Ferreira G, de Azevedo Jr WF (2019) Molegro virtual docker for docking. Methods in Molecular Biology 2053: 149-167. https://doi.org/10.1016/j.bmc.2012.08.012 [PubMed]
■ Castro-Alvarez A, Costa AM, Vilarrasa J (2017) The performance of several docking programs at reproducing protein-macrolide-like
crystal structures. Molecules 22(1): e136. https://doi.org/10.3390/ molecules22010136 [PubMed] [PMC]
■ Chakraborti S, Hatti K, Srinivasan N (2021) 'All that glitters is not gold': High-resolution crystal structures of ligand-protein complexes need not always represent confident binding poses. International Journal of Molecular Sciences 22(13): e6830. https://doi.org/10.3390/ ijms22136830 [PubMed] [PMC]
■ Chen YC (2015) Beware of docking! Trends in Pharmacological Sciences 36(2): 78-95. https://doi.org/10.1016/j.tips.2014.12.001 [PubMed]
■ Dhananjayan K (2015) Molecular docking study characterization of rare flavonoids at the Nac-Binding site of the first Bromodomain of BRD4 (BRD4 BD1). Journal of Cancer Research 1: 1-15. https://doi.org/10.1016Zj.bmc.2012.08.012
■ Granchi C, Capecchi A, Del Frate G, Martinelli A, Macchia M, Minutolo F, Tuccinardi T (2015) Development and validation of a docking-based virtual screening platform for the identification of new lactate dehydrogenase inhibitors. Molecules 20(5): 8772-8790. https://doi.org/10.3390/molecules20058772 [PubMed] [PMC]
■ IDF (International Diabetes Federation) (2017) IDF Diabetes Atlas. Eighth edition. International Diabetes Federation, Belgium, 147 pp.
■ Klangjareonchai T, Putadechakum S, Roongpisuthipong C (2015) Review of anti-hyperglycemic effect of Tinospora crispa. Walailak Journal of Science and Technology 12: 403-406.
■ Kufareva I, Abagyan R (2012) Methods of protein structure comparison. Methods in Molecular Biology 857: 231-257. https://doi. org/10.1007/978-1-61779-588-6_10
■ Lagarde N, Zagury JF, Montes M (2015) Benchmarking data sets for the evaluation of virtual ligand screening methods: review and perspectives. Journal of Chemical Information and Modeling 55(7): 1297-1307. https://doi.org/10.1021/acs.jcim.5b00090 [PubMed]
■ Metzler WJ, Yanchunas J, Weigelt C, Kish K, Klei HE, Xie D, Zhang Y, Corbett M, Tamura JK, He B, Hamann LG, Kirby MS, Marcinkeviciene J (2008) Involvement of DPP-IV catalytic residues in enzyme-saxagliptin complex formation. Protein Science 17(2): 240-250. https://doi.org/10.1110/ps.073253208 [PubMed] [PMC]
■ Mishra N, Basu A (2013) Exploring different virtual screening strategies for acetylcholinesterase inhibitors. BioMed Research International Similarity, 123(7): 471-474. https://doi. org/10.1007/978-3-540-68706-1_546 [PubMed] [PMC]
■ Nabeno M, Akahoshi F, Kishida H, Miyaguchi I, Tanaka Y, Ishii S, Kadowaki T (2013) A comparative study of the binding modes of recently launched dipeptidyl peptidase IV inhibitors in the active site. Biochemical and Biophysical Research Communications 434(2): 191-196. https://doi.org/10.1016Zj.bbrc.2013.03.010 [PubMed]
■ Praman S, Mulvany MJ, Williams DE, Andersen RJ, Jansakul C (2012) Hypotensive and cardio-chronotropic constituents of
Tinospora crispa and mechanisms of action on the cardiovascular system in anesthetized rats. Journal of Ethnopharmacology 140(1): 166-178. https://doi.org/10.1016/j.jep.2012.01.006 [PubMed]
■ Riyanti S, Suganda AG, Sukandar EY (2016) Dipeptidyl pepti-dase-IV inhibitory activity of some Indonesian medicinal plants. Asian Journal of Pharmaceutical and Clinical Research 9: 375-377.
■ Taylor OM, Lam C (2020) The effect of dipeptidyl peptidase-4 inhibitors on macrovascular and microvascular complications of diabetes mellitus: A systematic review. Current Therapeutic Research - Clinical and Experimental 93: e100596. https://doi.org/10.1016/j. curtheres.2020.100596 [PubMed] [PMC]
■ Wang S, Jiang JH, Li RY, Deng P (2020) Docking-based virtual screening of TpR1 inhibitors: Evaluation of pose prediction and scoring functions. BMC Chemistry 14(1): 1-8. https://doi.org/10.1186/ s13065-020-00704-3 [PubMed] [PMC]
■ Yoshida T, Akahoshi F, Sakashita H, Kitajima H, Nakamura M, Sonda S, Takeuchi M, Tanaka Y, Ueda N, Sekiguchi S, Ishige T, Shima K, Nabeno M, Abe Y, Anabuki J, Soejima A, Yoshida K, Takashina Y, Ishii S, Kiuchi S, Fukuda S, Tsutsumiuchi R, Kosaka K, Murozono T, Nakamaru Y, Utsumi H, Masutomi N, Kishida H, Miyaguchi I, Hayashi Y (2012) Discovery and preclinical profile of teneligliptin (3-[(2S,4S)-4-[4-(3-methyl-1-phenyl-1H-pyrazol-5-yl)piperazin- 1-yl]pyrrolidin-2-ylcarbonyl]thiazolidine): a highly potent, selective, long-lasting and orally active dipeptidyl peptidase IV inhibitor for the treatment of type 2 diabetes. Bioorganic & Medicinal Chemistry 20: 5705-5719. https://doi.org/10.1016/j. bmc.2012.08.012 [PubMed]
■ Zhang Z, Wallace MB, Feng J, Stafford JA, Skene RJ, Shi L, Lee B, Aertgeerts K, Jennings A, Xu R, Kassel DB, Kaldor SW, Navre M, Webb DR, Gwaltney SL (2011) Design and synthesis of pyrimidinone and pyrimidinedione inhibitors of dipeptidyl peptidase IV. 21 Journal of Medicinal Chemistry 54(2): 510-524. https://doi. org/10.1021/jm101016w [PubMed]
Author contributions
■ Andri Prasetiyo, Doctoral Program in Pharmaceutical Sciences, Assistant Professor, Faculty of Pharmacy, Pancasila University, Jakarta, Indonesia; [email protected], ORCID ID https://orcid.org/0000-0002-9936-6058. The main author played a major role in conducting the experiments, processing the data, and writing the article.
■ Shirly Kumala, Professor, Faculty of Pharmacy, Pancasila University, Jakarta, Indonesia; fskumala@univpancasila. ac.id, ORCID ID https://orcid.org/0000-0002-5444-4438. The author played a major role in providing ideas for writing, compiling and designing the article.
■ Esti Mumpuni, Associate Professor, Faculty of Pharmacy, Pancasila University, Jakarta, Indonesia; esti.mumpuni@ univpancasila.ac.id, ORCID ID https://orcid.org/0000-0002-9208-8040. The author played a leading role in the design of the experiment, analysis of the data, and participated in conducting the experiments.
■ Raymond R Tjandrawinata, PhD in Biochemical Sciences, Associate Professor, Faculty of Biotechnology, Atma Jaya Catholic University of Indonesia, Tangerang, Banten, Indonesia; [email protected], ORCID ID https:// orcid.org/0000-0003-0461-9874. The author played a role in interpreting and analyzing the data, writing and editing the article.
