Molecular docking studies for the identification of novel melatoninergic inhibitors for acetylserotonin-O-methyltransferase using different docking routines
© Azam and Abbasi; licensee BioMed Central Ltd. 2013
Received: 30 August 2013
Accepted: 14 October 2013
Published: 24 October 2013
N-Acetylserotonin O-methyltransferase (ASMT) is an enzyme which by converting nor-melatonin to melatonin catalyzes the final reaction in melatonin biosynthesis in tryptophan metabolism pathway. High Expression of ASMT gene is evident in PPTs. The presence of abnormally high levels of ASMT in pineal gland could serve as an indication of the existence of pineal parenchymal tumors (PPTs) in the brain (J Neuropathol Exp Neurol 65: 675–684, 2006). Different levels of melatonin are used as a trait marker for prescribing the mood disorders e.g. Seasonal affective disorder, bipolar disorder, or major depressive disorder. In addition, melatonin levels can also be used to calculate the severity of a patient’s illness at a given point in time.
Seventy three melatoninergic inhibitors were docked with acetylserotonin-O-methyltransferase in order to identify the potent inhibitor against the enzyme. The chemical nature of the protein and ligands greatly influence the performance of docking routines. Keeping this fact in view, critical evaluation of the performance of four different commonly used docking routines: AutoDock/Vina, GOLD, FlexX and FRED were performed. An evaluation criterion was based on the binding affinities/docking scores and experimental bioactivities.
Results and conclusion
Results indicated that both hydrogen bonding and hydrophobic interactions contributed significantly for its ligand binding and the compound selected as potent inhibitor is having minimum binding affinity, maximum GoldScore and minimum FlexX energy. The correlation value of r2 = 0. 66 may be useful in the selection of correct docked complexes based on the energy without having prior knowledge of the active site. This may lead to further understanding of structures, their reliability and Biomolecular activity especially in connection with bipolar disorders.
Molecular Docking is an effective and competent tool for in silico screening. It is playing an important and ever increasing role in rational drug design [7, 8]. Docking is a computational procedure of searching for an appropriate ligand that fits both energetically and geometrically the protein’s binding site. In other words, it is a study of how two or more molecules e.g. ligand and protein, fit together. The problem is like solving a 3D puzzle . During the past decade, for understanding the formation of intermolecular complexes, the application of computational methods in this arena has been subjected to intensive research. It is commonly known that molecular binding of one molecule (the ligand) to the pocket of another molecule (the receptor), which is commonly a protein, is responsible for accurate drug activity. Molecular docking has been proved very efficient tool for novel drug discovery for targeting protein. Among different types of docking, protein-ligand docking is of special interest, because of its application in medicine industry . Protein-ligand docking refers to search for the accurate ligand conformations within a targeted protein when the structure of proteins is known .
Docking procedures are basically the combination of search algorithms and scoring function. The largest number of search algorithms and scoring functions are available. Search algorithms predict the ligand binding orientation and conformations commonly referred to as posing . Some common search algorithms are : Monte Carlo methods, Genetic algorithms, Fragment-based methods, Point complementary methods, Distance geometry methods, Tabu searcher and Systematic searches. In order to differentiate between the active and random compounds, the scoring functions are employed. The scoring functions predict binding free energies in ligand-protein docking generally in 7–10 kJ/mol . Numbers of molecular docking software are employed in drug research industry . The most popular and commonly used softwares for molecular docking are AutoDock [13–15], AutoDock/Vina , GOLD [17, 18], FlexX , FRED , DOCK  and ICM . For docking purpose, AutoDock/Vina employed Broyden-Fletcher-Goldfarb-Shanno algorithm and it significantly improves the average accuracy of the binding mode predictions compared to AutoDock 4 . FlexX employed an IC algorithm. IC algorithm attempts to reconstruct the bound ligand by first placing a rigid anchor in the binding site and later using a greedy algorithm to add fragments and complete the ligand structure. GOLD considers the degree of freedom in the binding site that corresponds to reorientation of hydrogen bond donor and acceptor groups. This degree of freedom represents only a very small fraction of the total conformational space that is available but should account for a significant difference in binding energy values .
In connection with efforts rendered in searching for novel inhibitors of ASMT, we perform a comparative docking study with four extensively used programs: AutoDock/Vina, GOLD, FlexX and FRED. The docking accuracy and scoring reliability of the selected docking approaches were evaluated by docking seventy three melatoninergic ligands with ASMT and correlating the predicted binding affinities with the experimental values.
ASMT and melatoninergic inhibitors
Active site residues of ASMT
One letter code
TRP 11, 117, 285
LYS 107, 223
TYR 108 131, 336
GLY 110, 263
THR 112, 144, 207, 336
LEU 142, 186, 308, 326
PHE 26,143, 156, 237
ILE 277, 310
ASP 238, 268, 284
ARG 210, 280
GLN 253, 334
Molecular docking protocols are widely used for predicting the binding affinities for a number of ligands. In current work, our aim was to examine the possibility of an existing relationship between the experimental bioactivities of the inhibitors under study and the docking scores. In order to get accurate results, all the docking experiments were performed with the default parameters. The time to dock one ligand was approximately 1–2 min. Docking with AutoDock/Vina, GOLD and FRED was performed on a Linux workstation (openSUSE11.4) with an Intel Pentium D processor (3.0 GHz) and 1 GB of RAM where as FlexX was run on windows 7 equipped with an Intel® Atom™ processor (1.67 GHz) and 1GB of RAM.
Docking using AutoDock/Vina
Intermediary steps, such as pdbqt files for protein and ligands preparation and grid box creation were completed using Graphical User Interface program AutoDock Tools (ADT). ADT assigned polar hydrogens, united atom Kollman charges, solvation parameters and fragmental volumes to the protein. AutoDock saved the prepared file in PDBQT format. AutoGrid was used for the preparation of the grid map using a grid box. The grid size was set to 60 × 60 × 60 xyz points with grid spacing of 0.375 Å and grid center was designated at dimensions (x, y, and z): -1.095, -1.554 and 3.894. A scoring grid is calculated from the ligand structure to minimize the computation time. AutoDock/Vina was employed for docking using protein and ligand information along with grid box properties in the configuration file. AutoDock/Vina employs iterated local search global optimizer [34, 35]. During the docking procedure, both the protein and ligands are considered as rigid. The results less than 1.0 Å in positional root-mean-square deviation (RMSD) was clustered together and represented by the result with the most favorable free energy of binding. The pose with lowest energy of binding or binding affinity was extracted and aligned with receptor structure for further analysis.
Docking using GOLD (Genetic Optimization for Ligand Docking)
Where Shb_ext is the protein-ligand hydrogen bonding and Svdw_ext are the van der waals interactions between protein and ligand. Shb_int are the intramolecular hydrophobic interactions whereas Svdw_int is the contribution due to intramolecular strain in the ligand.
Docking using FlexX
FlexX (which is now a part of LeadIT) is a flexible docking method that uses an Incremental Construction (IC) algorithm and a pure empirical scoring function similar to the one developed by Böhm and coworkers  to place ligands into the active site. IC algorithms first dissect each molecule into a set of rigid fragments according to rotatable bonds, and then incrementally assemble the fragments around the binding pocket . For docking studies, the pdb files of ligands were transformed into a SYBYL mol2 file format and a ligands library was generated. A receptor description file was prepared through the FlexX graphic interface. An active site was defined by selecting the residue of the protein. The active site includes protein residues around 10 Å radius sphere centered on the center of mass of the ligand. Based on energy values, top ten ranked poses for each ligand in data set were selected for further analysis.
Here, f (ΔR, Δα) is a scaling function penalizing deviations from the ideal geometry and Nrot is the number of free rotatable bonds that are immobilized in the complex. The terms ΔGhb, ΔGio, ΔGar and ΔG0 are adjustable parameters. ΔGlipo is lipophilic contact energy (Rarey et al., ).
Docking with FRED (Fast Rigid Exhaustive Docking)
FRED uses multi-conformer docking algorithm which separately generates a set of low-energy conformers, and then do rigid docking for each conformer . In order to carry out correct docking, FRED required accurately prepared receptor file as well as a ligand conformer library. The receptor file was prepared by using make-receptor file provided in FRED whereas ligand conformer library was created in Omega 2.3.2 (OpenEye Scientific Software) with default settings. The volume of the docking box centered on the receptor was expanded in all directions until it was approximately 31671 Å3. The dimensions of the box were: 28.10 Å × 32.91 Å × 34.25 Å. FRED with a Gaussian type fitting scoring function Chemgauss4 was used to dock ASMT with ligands conformer library in order to obtain a potent inhibitor against ASMT. Chemgauss4 uses the potentials between the chemically matched positions around the ligand docked pose. Those chemical positions are complementary to the nearby specific groups in the receptor. Generally, the interactions are either hydrogen bond donors or acceptors and a favorable hydrogen bond score is obtained when a polar hydrogen position on one molecule overlaps a lone pair position on another molecule. The interactions which can be scored by Chemgauss functions are: steric, acceptor, donors, coordinating groups, metals, lone pairs, polar hydrogens and chelator coordinating groups .
Results and discussion
Number of good, fair and poor ASMT docked complexes obtained by the different docking routines
Time required for the docking of a single ligand
Important interactions between the active residues of ASMT and Ligands within 5 Å
Figure 5b shows the binding mode of top pose B22ASMT complex generated by GOLD with GoldScore of 64.88. Compound B22 was mediating hydrogen bond interactions with the side chain residues of Arg280, Ala159, Tyr327 and Tyr108. The phenol and indole ring of B22 formed favorable hydrophobic contacts with Tyr108, Tyr327, Tyr366, Trp117 and Leu326 and ionic interaction with Arg280.
The FlexX generated 10 solutions for B22. The highest-ranking solution has a binding energy of -25.45 kJ/mol. Hydrogen bonds with a backbone and side chain residues of Arg210, Lys223 and Thr207, back bone residues of Val211, back bone residues of Phe212 and Glu224 are observed. The strong hydrophobic interactions were with Thr195, Lys223, Leu228, Arg169, pyrrole ring of His209, Val211 and aromatic ring of Phe212. Ionic interactions were with Arg210 and Glu224 (Figure 5c).
The top ranked pose of ASMT-A3 generated by FRED has Chemgauss4 score of -9.87. Phenolic side chain of Tyr327, Tyr108, amide side chain of Asn330 and side chain residues of Lys107 formed strong hydrogen bonding interactions with electronegative atoms of ligand. Phenol group of Tyr108, and the aromatic ring of Trp117 and Tyr 327 were highlighted as major contributor of hydrophobic interactions. A3 also resulted in favorable ionic interactions through the active site with the back bone residues of Asn 330 and side chain residues of Lys107 (Figure 5d) (Table 4). The lowest binding affinity, high GoldScore, a low binding energy and number of observed hydrogen bonding interactions between the backbone residues and the various important residues at the entrance of the pocket enables B22 to be a strongly binding inhibitor of ASMT.
Comparison: showing the PIC50 values, chemgauss scores, binding affinities, gold scores and binding energies
Binding affinities (kcal/mol)
Binding energies (kJ/mol)
A3 Fred Best
The performance of FRED with respect to ranking the inhibitors in the list was poor. FRED ranked inhibitor B22 on the 11th place whereas all the remaining routines ranked B22 in 1st place. FRED docked fifty three inhibitors out of seventy three provided inhibitors which is a low number as compared to other routines. Kellenberger et al., concluded in their work that FRED had problems with small, polar and buried ligands . FlexX docked sixty eight inhibitors out of seventy three. The most severe restriction of FlexX is that it treats receptor as a rigid entity. Leach in 1994 reported an approach handling receptor flexibility .
In the light of the above analysis, the B22ASMT docked pose generated by AutoDock/Vina produced the best results. It forms hydrogen bonds, hydrophobic and ionic interactions with the important residues of the binding pocket of ASMT thus stabilizing the structure of target receptor. The dock pose with least binding energy has the highest affinity and hence is the best docked conformation.
Docking and scoring have evolved significantly over the past years. It has become a valuable tool in drug discovery process. Our goal of this study was to explore the feasibility of four different docking approaches: (AutoDock/Vina, GOLD, FRED and FlexX) for our target ASMT and to find out the lead compound. We compared the predictive power of each docking and scoring function. Our results suggest that all docking programs studied here do a reasonable job in docking and should aid significantly the drug discovery process. However, AutoDock/Vina consistently outperformed as compared to other programs and was found to be relatively more useful in blind docking pose prediction. Moreover, analysis of the docked ligands with the protein brought into focus some important interactions operating at the molecular level. The results of the ligand docking showed that the binding pocket involves the amino acid residues Ser213, Ser98, Val97, Thr100, Val211, Ser227, Arg210, Arg280, Phe212, Leu198, Ile198, Ser104, Thr195, Leu160, Tyr327, Tyr108, Trp117, Leu326, Phe29, Phe19, Gln334, Asn330, Ala159, Lys107, Met105. The important hydrogen bond forming amino acid residues was Arg280, Thr100, Tyr108, Asn330, Trp117 and Tyr327. In conclusion we have discovered a highly potent lead compound which will be useful for the design of novel less toxic and highly efficient drug for the treatment of bipolar disorders and PPTs.
This work is supported by Higher Education Commission (HEC) of Pakistan. The authors would also like to thank OpenEye Scientific Software and LeadIT for providing us academic license. We also thank Saad Raza for fruitful discussion.
- Kanehisa M, Goto S, Hattori M, Kinoshita KF, Itoh M, Kawashima S, Katayama T, Araki M, HiraKawa M: From genomics to chemical genomics: new developments in KEGG. Nucleic Acids Res. 2006, 34: D354-D357.PubMed CentralView ArticlePubMedGoogle Scholar
- Donohue SJ, Roseboom PH, Illnerova H, Weller JL, Klein DC: Human hydroxyindole-O-methyltransferase: presence of LINE-1 fragment in a cDNA clone and pineal mRNA. DNA Cell Biol. 1993, 12 (8): 715-727.View ArticlePubMedGoogle Scholar
- Rodriguez IR, Mazuruk K, Schoen TJ, Chader GJ: Structural analysis of the human hydroxyindole-O-methyltransferase gene. Presence of two distinct promoters. J Biol Chem. 1994, 269 (50): 31969-31977.PubMedGoogle Scholar
- Montange MF, Champier J, Szathmari A, Wierinckx A, Mottolese C, Guyotat J, Branger DF, Jouvet J, Lachuer A: Microarray analysis reveals differential gene expression patterns in tumors of the pineal region. J Neuropathol Exp Neurol. 2006, 65 (7): 675-684.View ArticleGoogle Scholar
- Srinivasan V, Smits M, Spence W, Lowe AD, Kayumov L, Perumal SRP, Parry B, Cardinali DP: Melatonin in mood disorders. World J Biol Psychia. 2006, 7 (3): 138-151.View ArticleGoogle Scholar
- Pyne S, Gayathri P: Geometric methods in molecular docking. Bioinformatics India Journal. 2005, III: 11-12.Google Scholar
- Kuntz DI: Structure-based strategies for drug design and discovery. Science. 1992, 257: 1078-1082.View ArticlePubMedGoogle Scholar
- Drews J: Drug discovery: a historical perspective. Science. 2000, 287: 1960-1964.View ArticlePubMedGoogle Scholar
- Kaapro A, Ojanen J: Protein docking. 2002, Available at http://www.lce.hut.fi/teaching/S-114.500/k2002/Protdock.pdf (accessed 28/12/2012)Google Scholar
- Muegge I, Rarey M: Small molecule docking and scoring. Rev Comp Chem. 2001, 17: 1-60.View ArticleGoogle Scholar
- Sousa SF, Fernandes PA, Ramos MJ: Protein-ligand docking: current status and future challenges. Proteins: Struct Funct Bioinf. 2006, 65: 15-26.View ArticleGoogle Scholar
- Bissantz C, Flokers G, Rognan D: Protein-based virtual screening of chemical database.1. Evaluation of different docking/ scoring combinations. J Med Chem. 2000, 43: 4759-4767.View ArticlePubMedGoogle Scholar
- Morris GM, Goodsell DS, Halliday RS, Huey R, Hart WE, Below RK, Olson AJ: Automated docking using a Lamarckian genetic algorithm and an empirical binding free energy function. J Comput Chem. 1998, 19: 1639-1662.View ArticleGoogle Scholar
- Goodsell DS, Olson AJ: Automated docking of substrates to proteins by simulated annealing. Proteins: Struct Funct Bioinf. 1990, 8: 195-202.View ArticleGoogle Scholar
- Morris GM, Goodsell DS, Huey R, Olson AJ: Distributed automated docking of flexible ligands to proteins: parallel applications of AutoDock 2.4. Comp Aid Mol Des. 1996, 10: 293-304.View ArticleGoogle Scholar
- Trott O, Olson AJ: AutoDock/Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization and multi threading. J Comput Chem. 2010, 31: 455-461.PubMed CentralPubMedGoogle Scholar
- Jones G, Willett P, Glen RC: Molecular recognition of receptor sites using a genetic algorithm with a description of desolvation. J Mol Biol. 1995, 245: 43-53.View ArticlePubMedGoogle Scholar
- Jones G, Willett P, Glen RC, Leach AR, Taylor R: Development and validation of a genetic algorithm for flexible docking. J Mol Biol. 1997, 267: 727-748.View ArticlePubMedGoogle Scholar
- Rarey M, Kramer B, Lengauer T, Klebe G: A fast flexible docking method using an incremental construction algorithm. J Mol Biol. 1996, 261: 470-489.View ArticlePubMedGoogle Scholar
- McGann M: FRED and HYBRID docking performance on standardized datasets. J Comp Aid Mol Design. 2012, 26: 897-906.View ArticleGoogle Scholar
- Ewing TJA, Kuntz ID: Critical evaluation of search algorithms for automated molecular docking and database screening. J Comput Chem. 1997, 18: 1175-1189.View ArticleGoogle Scholar
- Abagyan R, Totrov M, Kuznetzov D: ICM-a new method for protein modelling and design: application to docking and structure prediction from the distorted native conformation. J Comput Chem. 1994, 15: 488-506.View ArticleGoogle Scholar
- Ou SC, Chung CY, Chung HY, Sung WT, Tsai CC, Chien CC, Su DY, Lin SY: Molecular Docking for Protein Folding Structure and Drug-likeness Prediction. WSEAS conference. Int J Biol Biomed. 2005, 1 (2): 57-63.Google Scholar
- Azam SS, Saroosh A, Zaman N, Raza S: Role of N-acetylserotonin O-methyltransferase in bipolar disorders and its dynamics. J Mol Liq. 2013, 182: 25-31.View ArticleGoogle Scholar
- Volkamer A, Kuhn D, Rippmann F, Rarey M: DoGSiteScorer: a web server for automatic binding site prediction, analysis and druggability assessment. Bioinformatics. 2012, 28 (15): 2074-2075.View ArticlePubMedGoogle Scholar
- Ettaoussi M, Pe´res B, Klupsch F, Delagrange P, Boutin JA, Renard P, Caignard DH, Chavatte P, Berthelot P, Lesieur D, Yous S: Design and synthesis of benzofuranic derivatives as new ligands at the melatonin-binding site MT3. Bioorg Med Chem. 2008, 16: 4954-4962.View ArticlePubMedGoogle Scholar
- Du H, Wang J, Zhang X, Hu Z: A novel quantitative structure activity relationship method to predict the affinities of MT3 melatonin binding site. Eur J Med Chem. 2008, 43: 2861-2869.View ArticlePubMedGoogle Scholar
- Markl C, Attia MI, Julius J, Sethi S, Enderby PAW, Zlotos DP: Synthesis and pharmacological evaluation of 1, 2, 3, 4-tetrahydropyrazino [1, 2-a]indole and 2-[(phenylmethylamino)methyl]-1H-indole analogues as novel melatoninergic ligands. Bioorg Med Chem. 2009, 17: 4583-4594.View ArticlePubMedGoogle Scholar
- Elsner J, Boeckler F, Davidson K, Sugden D, Gmeiner P: Bicyclic melatonin receptor agonists containing ring-junction nitrogen: Synthesis, biological evaluation, and molecular modeling of the putative bioactive conformation. Bioorg Med Chem. 2006, 14: 1949-1958.View ArticlePubMedGoogle Scholar
- Leclerc V, Ettaoussi M, Rami M, Farce A, Boutin JA, Delagrange P, Caignard DH, Renard P, Berthelot P, Yous S: Design and synthesis of naphthalenic derivatives as new ligands at the melatonin binding site MT3. Eur J Med Chem. 2011, 46: 1622-1629.View ArticlePubMedGoogle Scholar
- Li G, Zhou H, Jiang Y, Keima H, Topiol SW, Poda SB, Ren Y, Chandrasena G, Doller D: Design and synthesis of 4-arylpiperidinyl amide and N-arylpiperdin-3-yl-cyclopropane carboxamide derivatives as novel melatonin receptor ligands. Bioorg Med Chem Lett. 2011, 21: 1236-1242.View ArticlePubMedGoogle Scholar
- Li Z, Wan H, Shi Y, Ouyang P: Personal Experience with Four Kinds of Chemical Structure Drawing Software: Review on ChemDraw, ChemWindow, ISIS/Draw, and ChemSketch. J Chem Inf Comput Sci. 2004, 44 (5): 1886-1890.View ArticlePubMedGoogle Scholar
- Pettersen EF, Goddard TD, Huang CC, Couch GS, Greenblatt DM, Meng EC, Ferrin TE: UCSF Chimera—a visualization system for exploratory research and analysis. J Comput Chem. 2004, 25: 1605-1612.View ArticlePubMedGoogle Scholar
- Baxter J: Local optima avoidance in depot location. J Oper Res Soc. 1981, 32 (9): 815-819.View ArticleGoogle Scholar
- Blum C, Blesa MJ, Roli A, Sampels M: Hybrid Metaheuristics: An Emerging Approach to Optimization. Studies in Computational Intelligence. 2008, Berlin Heidelberg: Springer-Verlag, 114-View ArticleGoogle Scholar
- Böhm HJ: Prediction of binding constants of protein ligands: a fast method for the prioritization of hits obtained from de novo design or 3D database search programs. J Comput Aided Mol Des. 1998, 12: 309-323.View ArticlePubMedGoogle Scholar
- Kellenberger E, Rodrigo J, Muller P, Rogna D: Comparative evaluation of eight docking tools for docking and virtual screening. Proteins: Struct Funct Bioinf. 2004, 57: 225-242.View ArticleGoogle Scholar
- Kitchen DB, Decornez H, Furr JR, Bajorath J: Docking and scoring in virtual screening for drug discovery: methods and applications. Nat rev Drug discov. 2004, 3 (11): 935-949.View ArticlePubMedGoogle Scholar
- Perola E, Walters WP, Charifson PS: A detailed comparison of current docking and scoring methods on systems of pharmaceutical relevance. Proteins: Struct Funct Bioinf. 2004, 56: 235-249.View ArticleGoogle Scholar
- Leach AR: Ligand docking to proteins with discrete side-chain flexibility. J Mol Biol. 1994, 235: 345-356.View ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.