International Journal of Scientific & Engineering Research, Volume 6, Issue 4, April-2015 1382

ISSN 2229-5518

Identification of potent VEGFR-2 Inhibitors of

Angiogenesis through homology modeling, structure based virtual screening, docking and molecular dynamics simulations

Madhu Sudhana Saddala, J. Obaiah, A. Usha Rani

Abstract— VEGFR-2 is a known protein target for antiangiogenic agents, was used in this study. In order to understand of the mechanisms of ligand binding and the interaction between the ligand and the VEGFR-2. A three-dimensional (3D) model of the VEGFR-2 is generated based on the crystal structure of the complex containing KIT and sunitinib (PDB code 3G0E) by using the Moduller9v11 software. W ith the aid of the MD simulation methods, the final refined model is obtained and is further assessed by SAVES server, which shows that the refined model is reliable. W ith this model, a flexible docking with screened compounds from Zinc database was performed by AutoDock in PyRx Virtual Screening tool. Ten lead molecules having better binding energy than query (Pazopanib). The binding interactions of compounds with active site of VP-3 model suggested that the amino acid residues (Glu13, Glu15, Leu57, Glu134, Cys136, Lys137, Gly215, Arg218, Asp 223, Arg247 and Tyr249) play a key role for drug design. The MD simulations were performed for VEGFR-2- Zinc14945139 (lead1) docking complex in NAMD v2.9. The Trajectory analysis showed docking complex and inter molecular interactions was stable throughout the entire production part of MD simulations. The predicted pharmalogical properties of best compounds, among them Zinc14945139 is well within the range of a drug molecule with good ADME profile. Hence, would be intriguing towards development of potent inhibitor molecule against VEGFR-2.

Index Terms— VEGFR-2, MD simulations, Molecular docking, Homology modeling, Zinc database, NAMD, VMD.

—————————— ——————————


admium (Cd) is classified as a Type I human carcinogen and its causes angiogenesis. It is the process of growing new blood vessels from the preexisting ones [1]. It is not
only important for normal physiological functioning but also critical for many pathological processes such as tumor initia- tion, proliferation, and metastasis [2, 3, 4]. It has been well established that angiogenesis is a fundamental requirement for tumor development. Studies have demonstrated that vas- cular endothelium could be a primary target of Cd toxicity [1,
4]. Some studies reported disruptive effects of Cd on human endothelial cells [5, 6]. The effect of Cd on reactive oxygen species (ROS), mitogen-activated protein kinases/extracellular signal-regulated kinases (MAPK/ERK), AKT, and p70S6K1 signaling pathways and the downstream proangiogenic fac- tors hypoxia-inducible factor-1 (HIF-1) and VEGF in human lung epithelial cells. VEGF exerts its biological actions by binding to its two receptor tyrosine kinases expressed on en- dothelial cells, namely, VEGFR-1 (Flt-1) and VEGFR-2 (KDR/Flk-1). VEGFR-1 is poorly autophosphorylated in re- sponse to VEGF in endothelial cells and is weakly involved in transducing the VEGF angiogenic signals. The evidences sup- port the concept that VEGFR-1 might act as a decoy receptor rather than as a signal transducing molecule [7, 8], whereas ligand-induced homodimerization of VEGFR-2 leads to a strong autophosphorylation of several [9] tyrosine residues of VEGFR-2. It is essential for the morphogenesis of vascular en- dothelium and is the primary receptor mediating the angio- genic activity of VEGF through distinct signal transduction pathways that regulate endothelial cell proliferation, migra- tion, differentiation and tube formation [10, 11]. The VEGFR-2
signaling pathway is a promising target of angiogenesis, be- cause it is a common pathway for tumor-induced angiogenesis [12]. VEGF is viewed as an attractive therapeutic target for the development of novel anticancer agents [41]. There are several angiogenesis inhibitors in phase I or phase II clinical trials, including antibodies aimed at VEGF or VEGFRs [13, 14], solu- ble decoy receptors that sequester ligands [40] and small mol- ecule inhibitors that inhibit [15] kinase activity. To exploit more efficient and safer agents for the treatment of angiogene- sis-related diseases such as cancer, a large number of scholars have been actively pursuing small molecule therapeutic strat- egies targeted at VEGFR-2-mediated [16, 17] signal transduc- tion pathway. In this study we use Pazopanib natural drug as a query screening compounds from drug databases. Pazo- panib, which is being developed by GlaxoSmithKline plc, is an oral, second-generation multi-targeted tyrosine kinase inhibi- tor that targets VEGFR, platelet-derived growth factor recep- tor and c-kit, key proteins responsible for tumor growth and survival. Pazopanib exhibited good potency against all of the human VEGFRs and closely related tyrosine receptor kinases in vitro, and demonstrated antitumor activity in several hu- man tumor xenografts, including renal cell carcinoma (RCC), and breast and lung cancer. In phase I and II clinical trials, Pazopanib was generally well tolerated with the main side effects being hypertension, fatigue or gastrointestinal disor- ders. Pazopanib alone caused a decrease in tumor size and stable disease in a significant number of patients, including those with RCC, NSCLC and gynecological tumors [18]. Final- ly, a virtual screening experiment was carried out to validate the effectiveness of the model and its ability to retrieve actives

IJSER © 2015

International Journal of Scientific & Engineering Research, Volume 6, Issue 4, April-2015 1383

ISSN 2229-5518

seeded in a compound collection of 35000 decoys.
Therefore, an attempt was made to develop the novel and
potent inhibitors against VEGFR-2 in angiogenesis through
computational approaches such as homology modeling, molec-
ular docking, molecular dynamics (MD) simulations, QSAR and
structure based virtual screening studies.


2.1 Physicochemical properties

Protein sequence of VEGFR-2 (Accession: P35968.2) of human was retrieved in the FASTA format from NCBI database. Physicochemical properties such as aliphatic index, Grand average of hydropathy (GRAVY), Theoretical pI value of VEGFR-2 were calculated using PROTPARAM.

2.2 VEGFR-2 model

The X-ray structures of VEGFR-2 in the PDB (3EWH) do not contain the loop after the DFG motif, which includes resi- dues from 1048 to 1068 [19]. The X-ray structure denoted by
3EWH in the PDB contains some residues of this loop; their positions are similar to the positions of the analogue residues in the tyrosine-protein kinase KIT (PDB: 3G0E) (Fig.1). Keep- ing in view, a new VEGFR-2 model consists of this loop was built using homology modeling by using MODELLER 9v11 [20] on the basis of co-ordinates of KIT (PDBID: 3G0E) as a template. The resulting loop structure is shown in Figure1B. Computations were run on a quad core Intel 3.0 GHz Xeon X5472 processor.
Fig. 1: Comparison between VEGFR-2 and KIT. (a) Superim- position of the analogous loops of VEGFR-2 and KIT (PDB:
3G0E). (b) Superimposition of the analogous loops of VEGFR-
2 and KIT after homology modeling (VEGFR-2 in red and KIT
in blue).

2.3 Homology modeling

The crystal structure of KIT (PDB: 3GOE) was selected [19] as a template to model the VEGFR-2. Sequence alignment be- tween 3EWH and 3GOE sequences was performed using [21]. ClustalW and was further optimized manually (Fig.2). This alignment was employed for model construction with the comparative [20, 22] modeling software MODELER9v11. All models were built with a medium-level loop refinement, and the discrete optimized protein energy (DOPE) potentials [23] were calculated for subsequent evaluation. The last modeller objective (low dope score) obtained was analyzed through Ramachandran plot to check the stereo chemical quality of protein structure using PROCHECK [24], environment profile using verify_3D [25] and ERRAT [26].The residue packing and


Prof. A. Usha Rani, Dept. of Zoology, Sri Venkateswra University, Tirupa- ti-517502. E-mail:

Madhu Sudhana Saddala is currently pursuing Doctor of Phylosophy program in Bioinformatics, Sri Venkateswara University, India, PH-

09542412717. E-mail:
atomic contacts were analyzed using WHATIF [27] and Z

score of Ramachandran plot was analyzed using WHATCHECK [28]. The resulting VEGFR-2 model was fur- ther refined by the energy minimization and MD simulations.

2.4 Refinement of homology model

The initial model was refined with MD simulation which was carried out with the Visual Molecular Dynamics (VMD) tool [29]. The CHARMM 27 field [30] was used for parameter- ization and the program NAMD [31] was used for all energy minimization and molecular dynamics (MD) simulations. All of the MD simulations were carried out in explicit water, em- ploying periodic boundary conditions. The system was first energy minimized for 1000 steps with atoms of VEGFR-2 fixed, and then energy minimization was performed for 2500 steps.

2.5 Simulation parameters

The MD simulations system was equilibrated at 250 k for 10 ps with VEGFR2 atoms fixed, followed by 20 ps MD without re- straints. The system was subsequently simulated for 100 ps at
300 k with the following parameters. The classical equations of motion were integrated by a leap frog integrator using a time step at 1fs. The impulse based ver let-I/r-RESPA method was used perform multiple time stepping: 4 fs long-range electro- static: 2fs for short range non-bonded forces, and 1 fs for bonded force [31]. The swift function was used to cutoff the Lennard-Jones potential, with the first cut off at 10 Å and the second cutoff at 12 Å. Short range interactions were calculated at intervals of 4 fs. All bonds involving hydrogen atoms were constrained to their equilibrium bond parameters using the SHAKE along them. Langevin dynamics were employed to maintain the pressure at 1 atm, with a Langevin pisten period of 100 fs and oscillation decay time of 50 fs. Trajectories were recorded every 200 fs. Subsequently the dynamics behavior and structural changes of the protein was analyzed by the cal- culation of energy and the root mean square deviation (RMSD).

2.6 Assessment of the homology model

Models were generated using the ClustalW alignment of the template and the sequence depicted in Fig.2, followed by re- finement with energy minimization and limited constrained MD simulations with an implicit solvent model. To assess the models before and after the simulations, we employed DOPE score, ERRAT (Fig.4), Verify_3D (Fig.5) scores and

IJSER © 2015

International Journal of Scientific & Engineering Research, Volume 6, Issue 4, April-2015 1384

ISSN 2229-5518

PROCHECK (Table1). DOPE is an atom-based statistical po- tential, representing the conformational energy that measures the relative stability of a conformation with respect to other conformations of the same protein. Verify_3D scores were used to assess whether a residue is in the desired 3D environ- ment. The “environment” is based on buried area, secondary structure, and fraction of polar contacts. Furthermore, having included the “Add Membrane” routine, the low-dielectric en- vironment of the bilayer was taken into consideration in the Verify_3D calculations; amino acids embedded in the implicit membrane boundary were considered fully buried. A low score is given to a hydrophobic residue on a protein’s surface and a polar residue in the protein’s core. A model region with a score higher than the expected values are likely to be correct, but a score lower than the expected values should be carefully examined. Comparison of all constructed models led us to select the one with the lowest DOPE and highest Verify_3D scores (data not shown). It should also be mentioned that a substantial improvement of both DOPE and Verify scores was observed for this model following refinement (Verify_3D scores rose from 65.43 to 90.66, as shown in Table1). It was pleasing to see that the resultant model was within the ex- pected range of the Profiles-3D scoring function, thus suggest- ing that the residues are placed in a physically acceptable en- vironment after refinement. PROCHECK was also used in order to check the stereochemical quality of the model-built structures. Ramachandran plots were generated within PROCHECK to highlight regions with unusual geometry and provide an overall assessment of the quality of the model. It can be seen that almost all residues in the model structures are in favored regions, and only 0.4% falls in the disallowed re- gions (Fig.3). Also RMSD values between superimposed mod- el and crystallographic structure were calculated by using su- perpose program. VMD was used to analyze the molecular dynamics trajectory. The protein atoms in all of the frames were superimposed on the first frame of the trajectory to re- move global rotational and translational movements. The root mean square deviation (RMSD) was calculated with reference to the starting structure. The PyMol molecular viewer ( was employed to analyze the docked structures.
Fig. 2: Sequence alignment of 3G0E with VEGFR-2 (PDB code
3EWH). Identical residues, similar residues and conservative
residues highlight different colours.
Fig.3: A Ramachandran plot calculation of VEGFR-2 was per- formed using PROCHECK server.
Fig. 4: 3D profiles of VEGFR-2 were verified using ERRAT server. Overall quality score indicates residues are reasonably folded.
Fig.5: 3D profiles of VEGFR-2 verified with Verify_3D server. Fig.6: Residues comprising the active sites of VEGFR2 and KIT in complex
Fig.7: Both Zinc and PubChem lead compounds superimposed in active site of VEGFR-2
Fig.8: VEGFR-2 domains represented in different colours, red indicates Phosphorylase Kinase domain and green indicates Phosphotransferase domain.
Fig.9: RMSD of protein backbone atom and ligand with re- spect to time over the course of the 10 ns MD simulations run (VEGFR-2- Zinc14945139 (lead1) complex).

2.7 Active site Identification

Active site of VEGFR-2 was identified using CASTp server (Computer Atlas of Surface Topology of protein) [32]. A new program, CASTp, for automatically locating and measuring protein pockets and cavities, is based on precise computation- al geometry methods, including alpha shape and discrete flow theory. CASTp identification and measurements of surface accessible pockets as well as interior inaccessible cavities by locating, delineating and measuring concave surface regions on three-dimensional structure of proteins (Fig.6). The meas- urement includes the area and volume of pocket or void by solvent accessible surface model (Richards’ surface) and by molecular surface model (Connolly’s surface), calculated ana- lytically. It can also be used to study surface features and func- tional regions of protein.

2.8 Screening Ligands

Commercially available ligands are listed in public databases, such as ZINC database [39] that contains more than 4.6 million compounds in ready to dock and provide 3D formats at the URL Virtual screening has been emerged as a complementary approach to high throughput screening and has become an important in silico technique in the pharmaceutical industry [33]. The structure based virtual screening begins with the identification of potential ligand binding sites on the target proteins. Usually, molecules that meet the criteria for biological activity fulfill characteristics contained in the Lipinski’s rule of five [34], or the more re- laxed rules revised by Veber et al., 2002 [35].
In our work, we have selected 35000 from Zinc database based on structure similarity of query Pazopanib natural com- pound. The AutoDock in PyRx Virtual Screening Tool URL [36, 37] was used for the energy minimization of screened ligands.

2.9 Ligands-Protein docking studies

Docking is a computational method which predicts the pre- ferred orientation of one molecule to a second when bound to
ER © 2015

International Journal of Scientific & Engineering Research, Volume 6, Issue 4, April-2015 1385

ISSN 2229-5518

each other to form a stable complex. Docking has been widely used to suggest the binding modes of protein inhibiters. Most docking algorithms are able to generate a large number of possible structures, thus they also require a means to score each structure to identify those that of greatest interest. Dock- ing was performed using AutoDock in PyRx Virtual Screening tool [36, 37].
Zinc screened compounds were docked into active site of refined model. Lamarkian genetic algorithm was used as number of individual population (150), max number of energy evaluation (25000000), max number of generation (27000) [24], Gene mutation rate (0.02), crossover rate (0.8), Cauchy beta (1.0) and GA window size (10.0).The grid was set whole pro- tein due to the multi binding pocket at X=3.42, Y=-56.23, Z=98.32 and dimension Å) at X=89.92, Y=98.56, Z=98.32 and exhaustiveness 8. The pose for a given ligands identified on the basis of highest binding energy. Only ligand flexibility was taken into account and the proteins were considered to be rig- id bodies. The resulting complexes were clustered according to their root mean square deviation (RMSD) values and bind- ing energies, which were calculated using the AutoDock scor- ing function. Further characterization via molecular dynamics (MD) simulations was conducted using complexes that were selected according to their binding energy values and the in- teractions made with the surrounding residues. The PyMol molecular viewer ( was employed to analyze the docked structures.

2.10 MD simulation of of the VEGFR-2 - ligand complex

The VEGFR-2-ligand complex was placed in TIP3P water box, with distance of at least 15Å imposed between the solute at- oms and the edge of the box. The position of the solute atoms were fixed during the first energy minimization for 1000 steps and no atoms were fixed during the second energy minimiza- tion for 1000 steps. MD simulation of the complex was carried out with the aforementioned protocol. The equilibrated system was subjected to 10 ns MD at 300K.


3.1 Physicochemical Properties of VEGFR-2

Protein sequence of VEGFR-2 (Accession: P35968.2) of human was retrieved in FASTA from NCBI database. Physicochemi- cal properties of VEGFR-2 contain 334 residues with the mo- lecular weight of 2542.01, Aliphatic index, Grand average of hydropathy (GRAVY) and Theoretical pI value of the protein were found to be 43.63, 0.311 and 9.21 respectively.

3.2 Secondary structure prediction of VEGFR-2

The secondary structure elements of model VEGFR-2 consist of total 334 amino acids and the results of predicted structure with three different servers were shown in Table2. The sec- ondary structure analysis has shown 25.38% alpha helix with
84 residues by GOR4, 41.99% with 139 residues by SOPMA,
42.90% with 142 residues by SOPM. Extended beta strands
exhibit 12.99% with 43 residues by using SOPMA, 12.99% with
43 residues using SOPM and 21.75% with 72 residues using
GO4. The predicted values for beta turn were found to be
8.46% with 28 residues using SOPMA, 10.88% with 36 residues
using SOPM and GOR4 fails to provide any prediction for beta
turn. Random coils strands exhibit 36.56% with 121 residues
by using SOPMA, 33.23% with 110 residues using SOPM and
52.87% with 175 residues using GO4 (Table1).

3.3 3D model contruction and validation

In this sense, a new model containing this loop was built using homology modeling by program MODELLER 9v11 (Sali and Blundell, 1993) and the structure of KIT as a template. Coordi- nates of structurally conserved regions (SCRs), structurally variable regions (SVRs), N-termini and C-termini from the template were assigned to the target sequences based on the satisfaction of spatial restraints. All side chains of the model protein were set by rotamers. Hundred models were generat- ed using MODELLER9v11. Least DOPE score structure was chosen for structural evaluation.
The secondary structure prediction of VEGFR-2 by above mentioned servers. We identified secondary structural ele- ments such as 2-sheets, 4-beta hairpins, 3-beta bulges, 7- strands, 15-helices, 22-beta turns, 3-gamma turns. Stereo chemical quality of the model was elucidated by PROCHEK sever. Ramachandran plot showed that 92.4% residues were plotted within the most favored region, 6.6% of residues were located within the additionally allowed region, 1.0% of resi- dues within generally allowed region and 0.0% of residues within disallowed region (Fig.3). By adapting Whatcheck pro- gram the calculated values of Z-score for bond lengths-0.941, bond distances-0.019, bond angle-1.203, Omega angle re- straints-0.735, Improper dihedral distribution-0.856, In- side/Outside distribution-1.059, dihedral RMS Z-score-0.856 were found respectively. Thus, these z-score values are posi- tive than average. The overall quality factor of 75.466% was observed by using of ERRAT (Fig.4) environmental profile. Varify_3D (Fig.5) showed that 90.66% of all residues had an average 3D-1D score greater than 2 indicating that model was highly reliable. The constructed 3D model was superimposed with template using SPDBV software (Fig.1). RMSD of the model structure was 0.25 Å which indicates the reliability of the good model. Fig.6 showed active site of VEGFR-2, residues such as Glu13, Glu15, Leu57, Glu134, Cys136, Lys137, Gly215, Arg218, Asp 223, Arg247 and Tyr249 respectively.
Table1: ERRAT, Verify_3D, and PROCHECK Scores, before and after Minimization of VEGFR-2 (DOPE score = -38712. 25)

3.4 Sructure of VEGFR-2

VEGFR-2 is a type III receptor tyrosine kinase of the

IJSER © 2015

International Journal of Scientific & Engineering Research, Volume 6, Issue 4, April-2015 1386

ISSN 2229-5518

PDGFR family [PDGFRα/β, c-Kit, FLT3, and CSF-1 (cFMS)]. The majority of VEGFR-2 intracellular domains contain tyro- sine residues (Tyr or Y) that are involved in redundant actions on vasculogenesis or angiogenesis [10]. The VEGFR-2 model was also divided into two parts viz. Phosphorylase Kinase domain and Transferase (Phosphotransferase) domain. The Phosphorylase Kinase domain (815-918residues; in our model indicates 1-103 residues) consists of 4 alpha helices and 5 beta sheets. This domain may involve in regulation of endothelial cell proliferation. The Transferase (Phosphotransferase) do- main (919-1166; in our model indicates 104-331). It contains 10 alpha helices and 2 beta sheets. This domain may also involve in regulation of endothelial cell migration. These two domains are playing a key role in angiogenesis mechanism. In our study both domains were structurally, functionally illustrated and developed novel inhibitors. Fig.8 represented the N, C terminals and two domains.

Table2: Secondary structure Elements of VEGRF-2 was pre- dicted by using different servers.
Fig.10: Representative examples of Zinc lead compounds con- sistent with binding mode.
MD simulation of this VEGFR-2 was carried out to assess its stability and to optimize the interactions between the pro- tein and ligand. The RMSD plot showed a general trend for a stable MD trajectory (Fig.9). An initial rise in RMSD was ob- served for the first 1ns, indicating equilibration of the system. After this equilibration period, the RMSD of the protein-ligand complex over around 6.0Å for the rest of the simulation peri- od, indicating fluctuations in the structure of the protein. MD simulation suggesting that the VEGFR-2 structure is fairly sta- ble.

3.5 Sreening ligands

Structure based virtual screening is a proficient approach in discovering inhibitors with novel chemical scaffolds. Two- dimensional structure of Pazopanib was used as query to search for similar compounds in the Zinc database. Then, ap- proximately 35000 compounds were screened and the all compounds were saved for further molecular docking studies.

3.6 Docking studies

Processing of the VEGFR-2 included energy minimized and molecular dynamics simulations. The refinement of structure of protein was used for the dock. AutoDock was used for the docking studies. The docked conformation corresponding to the lowest binding energy was selected as the most probable binding conformation. The total screened 35000 compounds were docked into the active site of VEGFR-2. The docking en- ergies of all the compounds were represented in kcal/mol. The best ten zinc compounds showed best binding energies and significant affinities with target protein of VEGFR-2 the values are represented in Table3. Which all the ligands were embedded within the active site of target protein were ob- served forming hydrogen bonds with same position as Pazo- panib established active site of target protein (Fig.7). The best docked compounds such as Zinc17042891, Zinc14945139, Zinc45339421, Zinc59048015, Zinc45319275, Zinc45228864, Zinc17042880, Zinc17010211, Zinc17026002, Zinc49917676 were found to be shown highest binding energies viz., -11.6, -
10.2, -9.9, -8.7, -8.2, -7.6, -7.2, -7.1, -6.4, -6.2kcal/mol respective- ly (Table3). The zinc lead compounds and their interactions with active site of residues are represented in graphically (Fig.10).
Hydrogen bonds play a role in stabilizing the protein- ligand complex [38]. The screened compounds exhibit several hydrogen bonding moieties. The best binding affinity com- pounds were obtained from the Zinc database through the structure based virtual screening and molecular docking stud- ies. The screened compounds were docked with active site of VEGFR-2 protein. The compound Zinc14945139 was bound with the affinity -11.0 kcal/mol by the formation of five hy- drogen bonds with Asp225, Arg146, Arg146, Asp225 and Asp223 respectively within the active site of target protein. Table3 showed binding energies, protein ligand interactions,

IJSER © 2015

International Journal of Scientific & Engineering Research, Volume 6, Issue 4, April-2015 1387

ISSN 2229-5518

bond angle and bond lengths.

Table3: Binding affinity, bond lengths, and bond angles of the best ten compounds from ZINC database with VEGFR-2

Zinc ID’s Compound structure (2D)



Zinc45339421 -9.9

Zinc59048015 -8.7

1.Cys 136 CO-----N2 H28

2.Lys 137 CO-----N1H33

3.Leu 57 CO-----N3 H29

1.Asp 223 CO-----O36C 25

2.Asp 223 CO-----O38H43

3.Glu 134 CO-----N8H40













Zinc45319275 -8.2

Zinc45228864 -7.6

Zinc17042880 -7.2

Zinc17010211 -7.1

1.Arg 247 CN-----O33S 34

2.Tyr 249 CO-----N4 H37

3.Arg 218 CO-----H41 N7

4.Arg218 CO------- N6 H39

5.Gly 215 CO-------H39 N6

1.Arg 218 CO-----N7 H44

2.Arg 218 CO-----N6 H6

3.Gly 215 CO-----H41 N6

4.Arg247 CN------- O34 C19

5.Glu 13 CO------- N2 H43

1.Glu-13 CO-----N2H39

2.Glu-13 CO-----N3H35

3.Gly 15 CO-----H37N5

4.Arg-247 SO--N1992 C1993

1.Arg247 CO ------O32 S34

2.Arg247 CN ------O32 S34

3.Arg247 CN ------O32 S34

4.Glu 13 CO------ N2H38

5.Glu13 CO------N1H35







































IJSER © 2015

International Journal of Scientific & Engineering Research, Volume 6, Issue 4, April-2015 1388

ISSN 2229-5518

Zinc17026002 -6.4

Zinc49917676 -6.2



-4.7 1.Lys-137 CO------H34N3 161.9 2.03

Table4: Molecular physical-chemical properties and Lipinski properties of lead molecules of lead compounds with OSIRIS and

M: Mutagenisity, T: Tumorigenicity, I: Irritant, R: Reproductive effect, S: Solubility.

The MD simulations results revealed that the VEGFR-2- Zinc14945139 (lead1) docking complex maintained high con- formational stability (Fig.9) as well as consistent structural flexibility. MD simulations showed novel insight into the nat- ural dynamics of VEGFR-2- Zinc14945139 docking complex by the presence of Water Bridge and additional hydrogen bonds on different time scales. The predicted pharmalogical proper- ties of Zinc14945139 (lead1) was well within the range of a drug molecule with good ADME profile (Tables4) (Lipinski and Hopkins, 2004), hence could be considered for designing inhibitor against VEGFR-2 which in turn would be useful for developing new therapeutic insights on angiogenesis.


In our work, we have constructed a 3D model of VEGFR-2 using the Modeller9v11. After MD simulations, this refined model structure is obtained. The final refined model was as- sessed further by SAVES server and the results show that this model is reliable. The stable structure is further used for dock- ing of Screened compounds. Through the docking studies, the model structures of the ligand-protein complex were obtained. The docking results indicate that active site amino-acid resi- dues in VEGFR-2 play an important role in maintaining a functional conformation. The interactions of protein and lig-

IJSER © 2015

International Journal of Scientific & Engineering Research, Volume 6, Issue 4, April-2015 1389

ISSN 2229-5518

ands proposed in this study are useful for understanding the potential mechanisms of VEGFR-2 and the ligands. In particu- lar, with the aid of H2O some hydrogen bonds are formed in the docked complex. As is well known, hydrogen bonds play an important role for the structure and function of biological molecules. On the other hand, the results reported here lead to the proposal of Glu13, Arg247 and Asp225 as key residues because they have strong H-bond contacts with the ligands. These residues are suggested that as candidates for further experimental studies of structure-function relationship. Our investigation revealed that the top ten compounds have exhib- ited significant binding energies when compare to query (Paz- opanib). Zinc14945139 (lead1) showed the best binding energy and pharmalogical profiles among selected leads. The VEGFR-


[1] J.M.Woods, M.Leone, K.Klosowska, P.C.Lamar, T.J.Shaknovsky, W.C. Prozi- aleck, “Direct antiangiogenic actions of cadmium on human vascular endo- thelial cells”, Toxicol In Vitro, vol. 22, pp. 643–651, 2008.

[2] J.Folkman, “Role of angiogenesis in tumor growth and metastasis”, Semin

Oncol, vol. 29, pp. 15–18, 2002.

[3] M.V.Blagosklonny, “Antiangiogenic therapy and tumor progression”, Cancer

Cell, vol. 5, pp. 13–17, 2004.

[4] W.C.Prozialeck, J.R.Edwards, and J.M.Woods, “The vascular endothelium as a target of cadmium toxicity”, Life Sci, vol. 79, pp. 1493–1506, 2006.

[5] S.Majumder, A.Muley, G.K.Kolluru, S.Saurabh, K.P.Tamilarasan, S.Chandrasekhar, H.B.Reddy, S.Purohit, and S.Chatterjee, “Cadmium reduc- es nitric oxide production by impairing phosphorylation of endothelial nitric oxide synthase”, Biochem Cell Biol, vol. 86, pp. 1–10, 2008.

[6] M.Helmestam, A.Stavreus-Evers, and M.Olovsson, “Cadmium chloride alters mRNA levels of angiogenesis related genes in primary human endometrial endothelial cells grown in vitro”, Reprod Toxicol, vol. 30, pp. 370–376, 2010.

[7] L.Seetharam, N.Gotoh, Y.Maru, G.Neufeld, S.Yamaguchi et al., “A unique signal transduction from FLT tyrosine kinase, a receptor for vascular endothe- lial growth factor VEGF”, Oncogene, vol. 10, pp. 135–147, 1995.

[8] T.V.Petrova, T.Makinen, K.Alitalo, “Signaling via vascular endothelial growth factor receptors”, Exp Cell Res, vol. 253, pp. 117–130, 1999.

[9] S.Rousseau, F.Houle, H.Kotanides, L.Witte, J.Waltenberger et al., “Vascular endothelial growth factor (VEGF)-driven actin-based motility is mediated by VEGFR-2 and requires concerted activation of stress-activated protein kinase-

2 (SAPK2/p38) and geldanamycin-sensitive phosphorylation of focal adhe-

sion kinase”, J Biol Chem, vol. 275, pp. 10661–10672, 2000.

[10] A.K.Olsson, A.Dimberg, J.Kreuger, L.Claesson-Welsh, “VEGF receptor sig- naling - in control of vascular function”, Nat Rev Mol Cell Biol, vol. 7, pp. 359–

371, 2006.

[11] M.Shibuya, L.Claesson-Welsh, “Signal transduction by VEGF receptors in regulation of angiogenesis and lymph angiogenesis”, Exp Cell Res, vol. 312, pp. 549–560, 2006.

[12] J.H.Baek, J.E.Jang, C.M.Kang, H.Y.Chung, N.D.Kim et al., “Hypoxia induced VEGF enhances tumor survivability via suppression of serum deprivation- induced apoptosis”, Oncogene, vol. 19, pp. 4621–4631, 2000.

[13] N.Ferrara, “Vascular endothelial growth factor as a target for anticancer ther- apy”, Oncologist, vol. 9, pp. 2–10, 2004.

[14] R.S.Kerbel, B.A.Kamen, “The anti-angiogenic basis of metronomic chemo- therapy”, Nat Rev Cancer, vol. 4, pp. 423–436, 2004.

[15] M.E.M.Noble, J. A.Endicott, and L.N.Johnson, “Protein kinase inhibitors:

insights into drug design from structure”, Science, vol. 303, pp. 1800–1805,

2 – lead1 docking complex MD simulations revealed that the molecular system was highly stable and its interactions were also highly stable. Therefore, Zinc14945139 is a proposed as best potential inhibitor to designing anti angiogenesis drug.


Author, Madhu Sudhana Saddala is especially grateful to University Grants Commission, New Delhi for their financial assistance with the award of BSR-Meritorious fellowship. This work was carried out in DBT-Bioinformatics Infrastructure Facility (BIF), Department of Zoology, Sri Venkateswara Uni- versity, Tirupati (BT/BI/25/001/2006).


[16] K.Spiekermann, F.Faber, R.Voswinckel, W.Hiddemann, “The protein tyrosine kinase inhibitor SU5614 inhibits VEGF-induced endothelial cell sprouting and induces growth arrest and apoptosis by inhibition of c-kit in AML cells”, Exp Hematol, vol. 30, pp.767–773, 2002.

[17] S.R.Wedge, D.J.Ogilvie, M.Dukes, J.Kendrew, R.Chester et al., “ZD6474 inhib- its vascular endothelial growth factor signaling, angiogenesis and tumor growth following oral administration”, Cancer Res, vol. 62, pp. 4645–4655,


[18] B.Sloan, N.Scheinfeld, S.A.Pazopanib, “VEGF receptor tyrosine kinase inhibi- tor for cancer therapy”, Curr Opin Investig Drugs, vol. 9, pp. 1324-35, 2008.

[19] A.Sali, T.L.Blundell, “Comparative protein modelling by satisfaction of spatial

restraints”, J Mol Biol, vol. 234, pp. 779−815, 1993.

[20] M.Camila, A.Francisco, H.Jans, M.Alzate, A.Vergara-Jaque, T.Kniess, J.Caballero, “Study of differences in the VEGFR2 inhibitory activities between semaxanib and SU5205 using 3D-QSAR, docking, and molecular dynamics simulations”, J Mol Graph Model, vol. 32, pp. 39–48, 2012.

[21] J.D.Thompson, D.G.Higgins, T.J.Gibson, “CLUSTAL W: Improving the sensi- tivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice”, Nucl Aci Res, vol. 22, pp. 4673−4680, 1994.

[22] M.A.Marti-Renom, A.C.Stuart, A.Fiser, R.Sanchez, F.Melo, A.Sali, “Compara- tive protein structure modeling of genes and genomes”, Annu Rev Biophys Biomol Struct, vol. 29, pp. 291−325, 2000.

[23] M.Y.Shen, A.Sali, “Statistical potential for assessment and prediction of pro-

tein structures”, Protein Sci, vol. 15, pp. 2507−2524, 2006.

[24] R.A.Laskoswki, M.W.MacArthur, D.S.Moss, J.M.Thornton, “PROCHECK: A program to check the stereo chemical quality of protein structures”, J. Appl. Cryst, vol. 26, pp. 283-291, 1993.

[25] M.D.Eisenberg, R.Luthy, J.U.Bowie, “VERIFY3D: assessment of protein mod-

els with three-dimensional profiles”, Methods Enzymol, vol. 277, pp. 396-404,


[26] C.Colovos, T.O.Yeates, “Verification of protein structures: Patterns of non- bonded atomic interactions”, Protein Sci, vol. 2, pp. 1511-9, 1993.

[27] G.Vriend, “WHAT IF: a molecular modeling and drug design program”, J Mol Graph, vol. 8, pp. 52–56, 1990.

[28] R.W.W.Hooft, C.Sander, and G.Vriend, “Verification of protein structures: Side-chain planarity”, J Appl Crystallogr, vol. 29, pp. 714–716, 1996.

[29] W.Humphrey, A.Dalke, K.Schulten, “VMD: visual molecular dynamics”, J Mol Graph, vol. 14, pp. 33-8, 27-8, 1996.

[30] A.D.MacKerell et al., “All-atom empirical potential for molecular modeling and dynamics studies of proteins”, J Phys Chem, vol. 102, pp. 3586-3616, 1998.

[31] J.Dundas, Z.Ouyang, J.Tseng, A.Binkowski, Y.Turpaz, J.Liang, “CASTp:

computed atlas of surface topography of proteins with structural and topo-

IJSER © 2015

International Journal of Scientific & Engineering Research, Volume 6, Issue 4, April-2015 1390

ISSN 2229-5518

graphical mapping of functionally annotated residues”, Nucleic Acids Res, vol. 34, pp. W116–W118, 2006.

[32] J.Dundas, Z.Ouyang, J.Tseng, A.Binkowski, Y.Turpaz, J.Liang, “CASTp: computed atlas of surface topography of proteins with structural and topo- graphical mapping of functionally annotated residues”, Nucleic Acids Res, vol. 34, pp. W116–W118, 2006.

[33] C.A.Lipinski, F.Lombardo, B.W.Dominy, P.J.Feeny, “Experimental and com- putational approaches to estimate solubility and permeability in drug discov- ery and development settings”, Adv Drug Delivery Rev, vol. 23, pp. 3–25,


[34] D.F.Veber, S.R.Johnson, H.Y.Cheng, B.R.Smith, K.W.Ward, K.D.Kopple, “Molecular properties that influence the oral bioavailability of drug candi- dates”, J Med Chem, vol. 6, pp. 2615-23, 2002.

[35] L.K.Wolf, “PyRx”, C&EN, vol. 87, pp. 31, 2009.

[36] O.Trott, A.J.Olson, “AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithread- ing”, J Comput Chem, vol. 31, pp. 455–461, 2010.

[37] F.Gao, N.Bren, T.P.Burghardt, S.Hansen, R.H.Henchman, P.Taylor, J.A.McCammon, S.M.Sine, “Agonist-mediated conformational changes in ac- etylcholine-binding protein revealed by simulation and intrinsic tryptophan fluorescence”, J Biol Chem, vol. 280, pp. 8443–8451, 2005.

[38] J.J.Irwin, B.K.Shoichet, “ZINC-a free database of commercially available com- pounds for virtual screening”, J Chem Inf Model, vol. 45, pp. 177–182, 2005.

[39] J.J.Irwin, B.K.Shoichet, “ZINC-a free database of commercially available com- pounds for virtual screening”, J Chem Inf Model, vol. 45, pp. 177–182, 2005.

[40] N.Ferrara, R.S.Kerbel, “Angiogenesis as a therapeutic target”, Nature, vol. 438,

pp. 967–974, 2005.

IJSER © 2015