Introduction

Topoisomerases are the nuclear enzymes that induce transient breaks in the DNA. There are two types of Topoisomerases—Topoisomerase I and Topoisomerase II. Topoisomerase I seems not to be as essential as Topoisomerase II for the survival of eukaryotic cells (Caroline and Katherine, 1998; Smiley et al., 2007; Thakur, 2011). So, it has been reviewed that most attention has been paid on the drugs acting on Topoisomerase II (Molinski, 1993; Hangstler et al., 2002). This enzyme plays a critical role in transcription and replication of DNA (Cortés et al., 2003; Kumar and Rawat, 2011), and also maintains the DNA topology, distangles knotted DNA, maintains correct chromosome condensation, decondensation, and segregation (Sorensen et al., 1996; Wang, 1998). It has been well-reviewed that it is a validated target of various antineoplastic drugs like anthracyclines (doxorubicin, daunorubicin), epipodophyllotoxins but are limited by their tumor resistance mechanism, side effects profile and also by their sensitivity to P-gp receptor mediated efflux. Now it is well-established that several antineoplastic agents those act through intercalation also acts on Topoisomerase II (Lee, 1996, Hawtin et al., 2010).

Literature of traditionally occurring medicines shows that natural products have very wide role and they are valuable source for new drug discovery (Fabricant and Farnsworth, 2001; Butler, 2004; Harvey, 2008). Crystallographic data based molecular modeling has been used to aid the design of synthetic analogs of natural products (Corbett and Berger, 2004; Huang et al., 2011; Nematollahi et al., 2011). There are so many compounds possessing pyridoacridine skeleton having anti-HIV activity, Ca2+ releasing activity, metal chelating property, DNA intercalating activity, and Topoisomerase II inhibition property (Bhakuni and Rawat, 2005; Kumar and Rawat, 2011). Pyridoacridines are colored marine alkaloids having 7H- pyrido [2, 3, 4-kl] acridine skeleton (Molinski, 1993; Skyler and Heathcock, 2002) (Fig. 1).

Fig. 1
figure 1

General structure of pyridoacridines

Full size image

The tetracyclic members of this class are archetypical pyridoacridines. Nine cytotoxic tetracyclic alkaloids, Cystodytins A-I have been identified from yellow tunicate Cystodytes dellechiajei. Cystodytins A and Cystodytins F are shown in Fig. 2a, b Cystodytins A–C are the first member of this class (Bontemps et al., 2010; Kumar and Rawat 2011). Cystodytin A–C and Varamine A and C (Fig. 2c, d) have been found to be cytotoxic against L-1210 (Kumar and Rawat, 2011).

Fig. 2
figure 2

a–h Examples of various novel pyridoacridine alkaloids

Full size image

A novel pentacyclic alkaloid, ascididemin isolated from brown colored tunicate Didemnum sp. has been found to be cytotoxic against L-1210 marine leukemia cells (Kumar and Rawat, 2011). Ascididemin and its isomers also exhibited cytotoxicity against U-87MG, U-373MG, T-47D, MCF-7, HCT-15, A-549, A-427, T-24, and J-82 cell lines (Matsumoto et al., 2003; Bhakuni and Rawat, 2005) (Fig. 2e, f).

Cyclodercitin, a hexacyclic alkaloid shown in Fig. 2g, obtained from the extracts of a deep violet sponge Dercitus sp. inhibits the proliferation of P-388 murine leukemia cells. It has also been reviewed that Eilatin, a heptacyclic pyridoacridine exhibits cytotoxic activity against HCT cell line (Stanslas et al., 2000; Kumar and Rawat, 2011). Eilatin octacyclic analog does not show any activity against HT-29 (Fig. 2h).

Therefore, it can be said that mostly all pyridoacridines have an immense role as anticancer agents. It is supposed that these compounds show anticancer effect due to inhibition of Topoisomerase II (Dias et al., 2005; Sanchez-Carrasco et al., 2008; Cragg et al., 2009). So, we report herein the study describing binding of pyridoacridine alkaloids against Topoisomerase II, which has been carried out by molecular docking investigations.

Computational details

Regarding this issue the crystal structures of Topoisomerase II were obtained from the Brookhaven Protein Data Bank http://www.rcsb.org/pdb (PDB entry: 1ZXM, 1PVG, 1AJ6 and 1QZR). To carry out docking studies, the 2D structures of various pyridoacridine ligands (Kumar and Rawat, 2011; Menna et al., 2011) were drawn (Fig. 3a–c) and these were converted into 3D and their energy was minimized using MM2 method with RMS gradient of 0.1 centers. These compounds were saved in mdl mol and pdb files for further use. Docking studies were carried out by iGEMDOCK 2.1, AutoDock Vina 1.1.2, and AutoDock 4.2.1 programs. In order to perform the task, the various interactions formed by docked ligands were observed.

Fig. 3
figure 3figure 3figure 3

a–c Structures of various pyridoacridine ligands

Full size image

To insure that the ligand orientation and the position obtained from the docking studies were likely to represent valid and reasonable binding modes of the inhibitors, docking of co-crystallized ligands were carried out for all protein structures (1PVG, 1QZR, 1AJ6, and 1ZXM). The ligand conformation found in the crystal structure, was extracted and docked back to the corresponding binding pocket. Results of control docking showed the optimal orientation of the docked inhibitor, close to that of the original orientation found in the crystal as shown in Fig. 4. The RMS deviation of less than 0.2 Å between the docked and crystal ligand coordinates indicate very good alignment of the experimental and calculated positions.

Fig. 4
figure 4

ANP docked molecule (1PVG) (UCSF Chimera ver. 1.5.3)

Full size image

In iGEMDOCK (iGEMDOCK ver. 2.1), drug screening was used as default settings with population size 200, 70 generations, and 3 numbers of solutions. iGEMDOCK scoring function was chosen along with ligand intra energy with hydrophobic and electrostatic preference both as 1.

Finally, ranking of compounds were done by pharmacological energy i.e.,

$$ E_{\text{pharma}} = \, E_{\text{GEMDOCK}} + \, E(E)_{\text{pharma}} + \, 2E(H)_{\text{pharma}} + \, 0.5E(V)_{\text{pharma}} ; $$

whereas, E GEMDOCK is the docked energy of iGEMDOCK and E(E)pharma, E(H)pharma, and E(V)pharma are the pharmacological scores of electrostatics, hydrogen-bonding, and vdW interactions (Hsu et al., 2011), respectively.

For AutoDock Vina (Trott and Olson, 2010), ligands were removed from pdb files and protein molecules were prepared by deleting solvent molecules and non-complex ions. Incomplete side chains were replaced using Dun Brack Rotamer library (Dunbrack, 2002). Hydrogens were added and gasteiger charges were calculated using Antechamber. The prepared files were saved in pdb format and used for further studies. Similarly, ligand files were prepared in pdb format with explicit hydrogen addition. All pdb files were transformed into pdbqt format. Grid center was placed on the active site. The sizes and centers of grid box are given in Table 1.

Table 1 Size and center of grid box
Full size table

Exhaustiveness which influences the thoroughness of the global search algorithm was set to be 8. Then, finally docking results were viewed using PDB and PDBQT files.

To gain better insight, AutoDock (Morris et al., 1999) was also employed to dock the selected pyridoacridine ligands. The prepared ligand files were transformed to pdbqt format, non-polar hydrogens were merged and charges were defined. The grid calculations were setup and maps were calculated using the program AutoGrid. The Grid maps were centered on the ligand binding site and dimensions were noted. The grid spacing was 0.3750 Å and the default AutoDock parameter settings were used for docking. The grid centers and number of points are shown in Table 2. All docking runs were performed using Lamarckian genetic algorithm and the obtained Dock scores were reported in kcal/mol. The docking protocol utilized in the study consisted of 10 independent GA runs, using an initial population of 150 randomly placed individuals, a maximum number of 250,000 energy evaluation, mutation rate of 0.02, a crossover rate of 0.80, and an elitism value of 1.

Table 2 Grid center and number of points
Full size table

Results and discussion

Topoisomerase II inhibitors with varying structural features and inhibition constants were selected from the literature and were docked into the catalytic site of Topoisomerase II. Dock runs of pyridoacridine ligands on protein 1ZXM, 1PVG, 1QZR, and 1AJ6 using iGEMDOCK, AutoDock, and Auto Dock Vina resulted in few best compounds that were evaluated based on their binding compatibility [docked energy (kcal/mol)] with the receptor. The results of docking experiments with these inhibitors are summarized in Tables 3 and 4. These results are mainly evaluated by structure analysis of the docked complexes.

Table 3 Docking scores for PDB 1PVG and 1QZR
Full size table
Table 4 Docking scores for PDB 1AJ6 and 1ZXM
Full size table

The IC50 values (μM) were recorded for the lowest binding energy mode by AutoDock Tools (AutoDock Tools ver. 1.5.6 rc2). The calculated IC50 values could not be correlated with the experimental values as the later values are not from direct inhibition of Topoisomerase II. The predicted IC50 value and experimental IC50 values are shown in Table 5.

Table 5 Predicted IC50 values, experimental IC50 values of some pyridoacridines
Full size table

Hydrogen-bonding interactions of compounds were visualized using Discovery Studio Visualizer as shown in Fig. 5. Compound no. 6 and 7 are having hydroxyl group which showed hydrogen-bonding interactions with SER 149, ASN 150, ALA 167, and LYS 168. Compound No. 56 was sandwiched in between ARG 76 and ILE 78 through Sigma-Pi stacking interactions (Fig. 6a). Docked poses of some compounds showing hydrogen-bonding interaction and Van der Waal interactions are shown with the help of AutoDock Tools in Fig. 7. Most of the compounds showed hydrogen-bonding interactions with SER 128, SER-149, TYR 165, ALA 167, ILE 120, ARG 76, and Vander Waal interactions with ASN-91, ALA-92, ASN-95, ARG-98, ILE-125, ILE-141, PHE-142, and THR-215. Surface diagram of all the ligands docked on PDB structure 1QZR is shown in Fig. 6b (PyMol, 2008).

Fig. 5
figure 5

a, b H-bonding interactions of compound no. 16 and 12 with Topoisomerase II (1QZR and 1PVG)

Full size image
Fig. 6
figure 6

a The Sigma–Pi interactions of compound no. 56 with Topoisomerase II (1AJ6) and b Surface diagram of all docked molecules into Topoisomerase II (1QZR)

Full size image
Fig. 7
figure 7

a, b) Docking poses, interaction of compound no. 3 and 5 with Topoisomerase II (1AJ6 and 1ZXM)

Full size image

The compound no. 11 exhibits a good score with iGEMDOCK but not with other two programs and its predicted IC50 value is found to be very high as compared to experimental value which may be due to some physical properties of the molecule and/or some another mechanism responsible for anticancer activity. For molecules containing lesser bulkier groups attached to the ring, good scores were given by all the three programs.

The correlation of docking scores by the two programs AutoDock and AutoDock Vina is shown graphically in Fig. 8. It can be seen that most of the compounds shows a correlation in their docking scores e.g., Compound no. 14 and 15 but there are also some compounds like 46 and 47, which do not correlate in their docking scores. Almost all the compounds except 11, 22, and 35 gave better score then bounded ligand with AutoDock Vina and AutoDock in case of PDB 1AJ6. Scores of iGEMDOCK could not be correlated with docking scores of other two softwares used.

Fig. 8
figure 8

Graphical representation of docking scores by the two docking programs Autodock and Autodock Vina

Full size image

Conclusion

Docking programs allowed us to estimate the docking scores, binding modes, and inhibition constants for the molecules under study. Almost all the compounds chosen except a few are found to be active against Topoisomerase II. In case of PDB 1AJ6, some ligands showed better fitness even than the co-crystallized ligand. An idealized representation of each ligand that makes every possible potential interaction with the binding site and other data obtained from all three programs iGEMDOCK, AutoDock Vina, and AutoDock, conclude that pyridoacridines are successfully docked into the protein binding site. Furthermore, this study will help in designing of novel derivatives of pyridoacridines and in discovery of new chemical entities for anticancer therapy.