Mol Biol Rep DOI 10.1007/s11033-014-3816-z

Molecular docking and molecular dynamics study on SmHDAC1 to identify potential lead compounds against Schistosomiasis Raghvendra Singh • Paras Nath Pandey

Received: 12 May 2014 / Accepted: 4 November 2014 Ó Springer Science+Business Media Dordrecht 2015

Abstract Schistosomiasis, a disease caused by helminth parasites of genus Schistosoma. Its treatment intensively depends on single drug, praziquantel which increases the risk of development of drug-resistant parasite. Inhibitors of human HDAC are profoundly reported as novel anti-cancer drugs and used as new anit-parasitic agents. Schistosoma monsoni class I HDACs are expressed in all stages of life cycle and indicating that this enzyme is most likely a major target for the designing specific inhibitors. In order to find novel target for the treatment of Schistosomiasis, three dimensional structure of SmHDAC1 was generated, using homology modelling. Features of the generated structure, was then deduced with respect to conformation of peptide backbone, local compatibility of the generated structure in terms of energy and molecular dynamics study. Considering these features of the generated structure, we selected all the class 1 inhibitors reported so far, which showed interactions with HDACs. Virtual screening was done using reported inhibitors (70) and using SmHDAC1 and HsHDAC1 as the targets. On the basis of binding affinity and IC50 value, 24th ligand was selected for the molecular docking purpose. In this study, out of all the reported inhibitors, 24th inhibitor (N,8-dihydroxy-8-(naphthalen-2-yl) octanamide zinc id- ZINC13474421) showed Electronic supplementary material The online version of this article (doi:10.1007/s11033-014-3816-z) contains supplementary material, which is available to authorized users. R. Singh (&) Center of Bioinformatics, Institute of Interdisciplinary Studies, Nehru Science Center, University of Allahabad, Allahabad, India e-mail: [email protected] P. N. Pandey Deaprtment of Mathematics, University of Allahabad, Allahabad, India e-mail: [email protected]

better binding with SmHDAC1 (-8.1 kcal/mol) as compared to HsHDAC1 (-6.4 kcal/mol) in terms of binding energy and supported by IC50 value. This paper throws light on the reliable model for further structure based drug designing, concerning SmHDAC1 of S. mansoni. Molecular docking studies highlighted advantages of comparative in silico interaction studies of SmHDAC1 and HsHDAC1. N,8-dihydroxy-8(naphthalen-2-yl) octanamide can further use for the clinical trial. Keywords Schistosomiasis
Homology modeling
SmHDAC1
Virtual screening
Molecular docking

Introduction Schistosomiasis (bilharzia) caused by platyhelminth parasites from the genus Schistosoma (S. mansoni, S. japonicum, S.haematobium, S. intercalatum, S. mekongi) is one of the major human neglected parasitic diseases [1, 2]. More than 250,000 deaths and around 200 million infections by Schistosomiasis get reported worldwide yearly, with about 800 million people further at risk of infection [3]. Praziquantel is reported as the only drug used for mass treatment of schistosomiasis [4]. So, availability of the single drug for this neglected serious disease draws a sincere attention towards the development of novel compound for the treatment of schistosomiasis [5]. The Human epigenetic players are usually taken to be as intense target to develop anti-cancer compounds/scaffolds because of the increasing involvement of epigenetic players in cancer genesis and progression [6, 7]. Interestingly, several human eukaryotic parasites share common features such as development of tumers due to, increased metabolic activity, using lactate fermentation as an energy source, uncontrolled cell division, and development of host resistance [8]. Therefore, it

123

Mol Biol Rep

gives the positive expectation with approach targeting parasitic epigenetic players that can be successful in treating human parasitic diseases caused by eukaryotic pathogens. A ‘‘piggyback’’ strategy can be used to speed up the search for novel anti-parasitic drugs, that builds on chemical scaffolds previously validated for other diseases [9]. Histone deacetylases are among the most studied epigenetic targets as evident from the studies made so far [10]. Histone deacetylase inhibitors have also been tested to fight major human parasitic diseases such as leishmaniasis, malaria, schistosomiasis, toxoplasmosis, and trypanosomiasis [11]. Yet, the risk of off-target effects of the developed drug with host (human) has also been highlighted by reported studies. So to cross this hurdle, the use of structural data, obtained either from modeling or crystallographic/NMR studies can be of great importance and does seem promising [11]. Reportedly, with valuable progression this strategy has been applied on Plasmodium falciparumHDAC1 (pfHDAC1), where chemical library screening and drug design studies, based on a homology model of this enzyme, have yielded inhibitors with anti-parasitic activity [12]. Earlier it was reported that member of Class I HDAC play major role in parasitic infectivity projecting it as potential epigenetic drug target [11], reinforcing it as a potential epigenetic drug target. As structure is important key to functionality of any biomolecule, as crystal structure of SmHDAC1 is not yet resolved by any experimental tools. Here we have used bioinformatics approach to model the SmHDAC1 structure and screened the derivatives of hydroxamtes, biphenyl and benzamide to fined best lead compound.

Materials and methods Sequence alignment and homology modeling Protein sequences of class I HDACs of human (Q13547, Q92769, O15379 and Q9BY41) and SmHDAC1 (G4V604) of S. mansoni were retrieved from UniProtKB database (http://www.uniprot.org). All sequences, were then imported into the ClustalX program [13] for multiple sequence alignments (MSA). Alignment result was examined manually to find out conservation pattern of Zn?? binding residues, catalytic residues and cavity forming residues, using the structural information available in PDBs. Conservation of metal binding residues were further crossvalidated by MSA result of HDAC1 of S. mansoni, human (UniProt acc. no. G4V604), H. sapience (UniProt acc. no. Q13547), M. musculus (UniProt acc. no. O09106), D. rerio (UniProt acc. no. Q8JIY7) and A. aeolicus (UniProt acc. no. O67135). To find the new target for the treatment of Schistosomiasis, the three-dimensional structure of human HDAC1 (HsHDAC1) (PDB id: 4BKX) was used as a template for

123

homology modeling of S. mansoni HDAC1 (SmHDAC1) (G4V604) in the MODELER9.10 program [14]. Electrostatic Surfaces of the modelled structure and class I HDACs were generated using APBS plugin [15] which is an interface to the popular adaptive Poisson–Boltzmann solver [16] in PyMol with the amber force field, a grid ˚ , and dielectric constants of 2 and 78 for spacing of 0.5 A the protein and the solvent, respectively. Evaluation and validation of protein model The quality of the generated protein model was evaluated considering stereo chemical geometry (RAM-PAGE server [17] ), energy (ProSA [18] web server), packing environment WHAT IF [19] and RMSD difference between modeled and template structure. Molecular dynamics The MD simulations of modeled SmHDAC1 protein was performed with the Gromacs 4.6.5 Gromacs 4.6.5 [20] using amber99 force field [21]. Protein was kept in the centre of a cubic box, that was filled with TIP3P [22] water model. The minimum distance between any atom of the ˚. protein and the boundary of the cubic box was kept 10 A The physiological condition of the system was maintained by adding 0.01 M concentration of NaCl solution. System was stabilized by energy minimization, for this purpose 50,000 steps of the time step of steepest descent were run with step size 0.01 ps. After energy minimization, system was subjected to equilibration to achieve optimum stable temperature by NVT ensemble. Equilibration of pressure is conducted under an NPT ensemble to stabilize the pressure. The particle mesh Ewald (PME) [23] method used for ˚ calculating long-range electro static interactions, a 0.9 A cut-off fixed for non bonded Van der Waals interactions, and short-range electrostatic cut-off. LINCS [24] constraints were performed for all bonds, keeping the whole protein molecule fixed and allowing only the water molecule to move to equilibrate with respect to the protein structure. For maintaining the constant temperature and pressure 300 K and 1 bar respectively, the modified Berendsen thermostat (V-rescale) and Parrinello–Rahman barostat were used to couple the system. Virtual screening and molecular docking Considering the features of the modelled structure, we selected all inhibitors of the class I HDAC reported [25] (Table S1). These inhibitors were treated as ligand and were docked with the active site of SmHDAC1 (modelled protein structure) and HsHDAC1 (PDB id: 4BKX). AutoDock Vina [26] was used for virtual screening and the docking computation. To

Mol Biol Rep

Fig. 1 Alignment of HDAC1 of Schistosoma mansoni and class I human HDACs (HDAC1, HDAC2, HDAC3 and HDAC8). Catalytic ˚ residues, Zn2? binding residues, and residues forming the 14 A

cavities are marked as green circle, red quadrilaterals and blue triangles, respectively. (Color figure online)

generate the docking input files, AutoDockTools [27] package was used. Zn ion present in vicinity of active site was selected as centre for grid preparation. Grid box has drawn 24 x 22 x ˚ spacing. Further docking result were analysed 22 with 1 A by PyMol [28]. LigPlot [29] was used for analysis of protein ligand complex interaction. SAHA (slandered inhibitor of class I HDACs of human) was taken as reference to validate the docking results.

(Figure S1; Table S3) and catalytic residues as the members of human class I HDACs contained. Conservation of metal binding residues were further cross validated by MSA of HDAC1 protein sequence of S. mansoni, H. sapience, M. musculus, D. rerio and Aquifecous (Fig. 2a). To model the 3D structure of SmHDAC1, template was identified by submitting SmHDAC1 (target sequences) to PSIBLAST search (http://blast.ncbi.nlm.nih.gov/) against the Protein Data Bank PDB (www.rcsb.org/). Human HDAC1 (PDB id: 4BKX) has 65.67 % identity (as expected) with target sequences, was selected as a template for homology modeling in this study. Total ten models of SmHDAC1 were generated by using MODELER 9.10. Template contains cofactor (Zn ion) in the catalytic core which is important for the functionality of protein. Homology modeling does not provide facility to model the cofactor like metal ions. Py Mol software used to align the modeled structure along with template (which contains coordinates of Zn co-factor). Residue environment near the Zn ion in modeled protein are similar as in template. So, PyMol used to rewrite pdb of

Results and discussion Sequence analysis and homology model validation To find similarity between SmHDAC1 and class I HDACs of human, we performed MSA (Fig. 1). SmHDAC1 shows highest identity (65.67 %) similarity (69.46 %) with HsHDAC1, as compared to the rest of the members of class I HDACs of human (Table S2). It was also noticed that SmHDAC1 shared similar pattern of zinc binding residues

123

Mol Biol Rep Fig. 2 Conserved Zn?? and K? ions binding residues of SmHDAC1. a The highly conserved Zn ion and K ions binding residues of HDAC1 among H. sapience (Hs_HDAC1), M. musculus (mouse_HDAC1), D. rerio (ZebraFish_HDAC1), S. mansoni (SmHDAC1) and Aquifecous (Aquifecous_HDAC1) are shown here in the multiple sequence alignment. The solid orange circles represent Zn ions binding residues while red and purple coloured solid circles represent the two different K ions binding residues. b 3D representation of SmHDAC1. The close up c shows active site residues as well as Zn ion and one K ion while Close-up d represents the second K ion and its binding residues. (Color figure online)

modeled protein along with coordinates of Zn ion that further used in downstream study. The quality of modeled structure (SmHDAC1) were then evaluated by generation of Ramachandran plot by RAMPAGE server, It showed, the torsion angles of the modeled structure of SmHDAC1 showed 96.0 % amino acid residues in the most favoured regions, 3.4 % in the allowed regions and 1.5 % in the generously allowed regions whereas 0.6 % amino acid residues were in the disallowed region. Ramachandran plot was satisfactory while comparing with HsHDAC1 (4BKX) (Figure S2). The constructed model of SmHDAC1 was validated by ProSA program in terms of Z-score representing the overall quality and measuring the deviation of the total energy of the protein structure. In this plot, the Z-score value of the obtained model of SmHDAC1 was -8.42, which was located within the space of protein structure solved by X-ray. This value is very close with the value of the template (4BKX) viz -10.69, suggesting that the obtained model is reliable and closes to experimentally determined structure (Figure S3 upper panel). Knowledge based energy profile of the modeled SmHDAC1 in comparison to the crystal structure of HsHDAC1 (4BKX) show

123

good harmony (Figure S3 lower panel). The packing environments for residues of the modeled SmHDAC1were compared with experimental structures HsHDAC1 by WHAT IF. A score of -5.0 or inferior typically specify poor packing of structure. In general, our SmHDAC1model has similar packing scores compared to the template X-ray structure. A few residues (Figure S4) are poorly packed in these models based upon scores less than -5.0. Radius of gyration (Figure S5) support the stability and compactness of modeled structure. ˚ , between The back bone RMSD difference was 0.202 A the SmHDAC1 (Figure S6C) and the template HsHDAC1 (4BKX) (Figure S6B) crystal structure, which indicates that spatial arrangement of the generated model is quite similar to the template (Figure S6A). It was ascertain that modeled structure has reasonably better quality. RMSD plot (Figure S7) supports the better quality of modeled structure respectively. In summary, the quality of modeled SmHDAC1 was evaluated by above specified methods and analysis suggested that good quality of model was generated by homology modeling.

Mol Biol Rep Fig. 3 Comparision of conserved residues of active site and cavity forming residues of SmHDAC1 and HsHDAC1. a1 and b1 show cavity forming residues (sticks) of HsHDAC1 and SmHDAC1 respectively Zinc ?? ion is shown in solid sphere (transparent) of HsHDAC1 (SmHDAC1). a2 and b2 represent active site binding residues (sticks) of HsHDAC1 and SmHDAC1 respectively

Fig. 4 Electrostatic surfaces of modeled SmHDAC1 and X-ray structures of HDACs 1, 2, 3, and 8

123

Mol Biol Rep Table 1 Filtered inhibitors: showing better binding affinity with modeled structure of Schistosoma as compare to Human HDAC1with corresponding IC50 values

0.2

IC50 value(uM)[24] 0.006

23rd

-6

-6.7

0.7

0.025

24rth

-6.4

-8.1

1.7

0.035

27th

-6.8

-7.1

0.3

0.1

2nd

-6.6

-6.8

0.2

0.0714

10th

-5.9

-6.4

0.5

0.095

13th

-6.1

-6.6

0.5

0.045

17th

-6

-6.6

0.6

0.008

44rth

-6

-6.1

0.1

52.56

45th

-5.4

-5.5

0.1

12.25

48th

-6.6

-6.8

0.2

2.4

55th

-7

-7.7

0.7

100

5th

-5.8

-6.5

0.7

0.065

60th

-6.8

-7.7

0.9

44

61st

-6.8

-7.7

0.9

40

62nd

-7.3

-8.2

0.9

2.8

63rd

-7.4

-7.8

0.4

4.6

6th

-5.9

-6.4

0.5

0.135

70th

-7.4

-7.7

0.3

5

First

-6.7

-6.9

0.2

0.103

21st

-6.5

-6.7

0.2

0.6

Ligand 22nd

Binding Energy of HsHDAC1 (kcal/mol) -7.1

Binding Energy of SmHDAC1 (kcal/mol) -7.3

Difference

Fig. 5 2D structure and physical representation of ligand number 24th

Comparison of the homology model of SmHDAC1 and HsHDAC1 (PDB id: 4BKX) During homology modeling of SmHDAC1, X-ray structure of HsHDAC1 in complex with the dimeric ELM2-SANT ˚ resolution [30]. domain of MTA1were reported at 3 A Analysis of secondary structure and fold comparison of model and X-ray structure of SmHDAC1 and HsHDAC1 was done by STRIDE [31]. Overall folds of model are very much similar with x-ray crystal structure of HsHDAC1. Half of the residues are involved in forming secondary

123

structure, while rest of the residues are engaged in forming loops and turns which are critical for analysis and appear in loops adjacent to the active site. Loop and turn forming residues play essential role in selection of ligands. Single canonical a/b fold [32] contains six parallel b sheet embedded between 10 a helices is displayed (Figure S8A) by STRIDE while HsHDAC1, a/b domain contains 8 parallel b sheet embedded between 11 a helices (Figure S8B) DSSP analysis of trajectory generated by molecular dynamics, support the stability of conserved canonical a/b fold present in SmHDAC1 with some interchangeable

Mol Biol Rep

Fig. 6 Docking of HsHDAC1 with SAHA inhibitor. a Ribbon representation of docked complex of SAHA (pink) with HsHDAC1 (rainbow), b SAHA enters in hydrophobic core of the protein (surface

view), c 2D ligplot view represent hydrophobic interactions and hydrogen bonds. (color figure online)

Fig. 7 Comparison of docking of ligand 24th with SmHDAC1 and HsHDAC1. a1 and b1 are cartoon representation of docked ligand 24th with SmHDAC1 and HsHDAC1 respectively. Zn ion (solid sphere) is present in centre of the complex. Ligand and interacting

residues are shown in sticks form. a2 and b2 are surface representation of complex red colour is representing hydrophobic cavity. a3 and b3 are 2D ligplot of which show D2 representation of ligand complex interaction. (Color figure online)

123

Mol Biol Rep Fig. 8 Display hydrogen bonds formed in between ligand 24th (stick representation) and residues of SmHDAC1(lines representation). Hydrogen bonding between atoms is shown in dotted lines (yellow in colour). Distances are labelled in angstrom. (Color figure online)

secondary structure states during the dynamics (Figure S9B,C). Major fluctuations observed in loops regions (Figure S9A). Architecture of the SmHDAC1 catalytic pocket Comparison of active site forming residues and catalytic pocket forming residues of SmHDAC1 and HsHDAC1 are shown in Fig. 3. Cavity forming residues which are commonly present in HsHDAC1 and SmHDAC1 are Gly87, Gly88, Leu89, Phe100, Cys101 and Tyr253. Gly86 is present in SmHDAC1 instead of Ala135 as compare to HsHDAC1 (Fig. 3 a1, b1). Residues lining the hydrophobic pocket of SmHDAC1 (Phe59, Tyr102, Gly250, Gly251 and Leu89) facilitate the interaction of ligand. In HsHDAC1, an extra loop which is present nearby catalytic pocket which make a major difference in terms of selectivity of ligands. These loop residues are Asn23, Tyr24, Gly27, Gln28, His30, Met32, Arg36, Ile31 and Arg38 (Fig. 3 a1). The differential properties of SmHDAC1 and HsHDAC1 is due to the major differences in cavity forming residues, hydrophobic lining residues and additional loop nearby catalytic pocket in HmHDAC1 play essential role in selectivity of inhibitors of two different homologous proteins. Electrostatic surface of SmHDAC1, HsHDAC1, 2, 3 and 8 were generated by APBS tool (Fig. 4). The surface charges between modeled SmHDAC1 and X-ray structure HsHDAC1 are fairly well (Fig. 4a, b). The surface of

123

HsHDAC1 and SmHDAC1 are more polar than HDAC8, while HDAC2 and HDAC3 are show intermediate polarity. Most of the potential differences have been shown away from the active site in members of class1 HDACs (Fig. 4 a, b, c, d and e). In Fig. 4, it can be seen that SmHDAC1, HsHDAC1 and HDAC2 demonstrate significantly more positive charge on the face containing the active site than HDAC3 or HDAC8. These differences could be accountable for the discriminating recruitment of other proteins that allow the HDACs to target specific genes or specific lysines on other proteins such as p53 [33]. Identification of potential inhibitor of SmHDAC1 by virtual screening AutoDock Vina has been used for virtual screening. Comparative results of HsHDAC1 and SmHDAC1 in terms of binding energy have been shown in (Table S3) with corresponding their IC50 value. Result of virtual screening shows that out of 70 ligands, twenty-one ligands (number 22nd, 23rd, 24th, 27th, 2nd, 10th, 13th, 17th, 44th, 45th, 48th, 55th, 5th, 60th, 61st, 62nd, 63rd, 6th, 70th, First, 21st) have better binding affinity with SmHDAC1 as compare to HsHDAC1 in terms of binding affinity (kcal/mol) shown in Table 1. These twenty-one ligands were further screened on the basis of IC50 value and best binding affinity. Among twenty-one of them, ligand number 62nd and 24th comparatively show better binding energy *8.2 kcal/mol but in terms of IC50 value ligand number 24th (IC50 = 0.035) is

Mol Biol Rep

more selective than ligand number 62nd (IC50 = 4.6uM). Ligand number 24th was further selected for the molecular docking purpose with HsHDAC1 and SmHDAC1 by using AutoDock Vina. Ligand number 24th (ZINC13474421) consist of eight carbon long aliphatic chain each side attached with hydroxamic acid group and two aromatic rings (Fig. 5). Popular name of ligand 24th is DNC009841and another name is N,8-dihydroxy-8-(naphthalen-2-yl) octanamide zinc database [34] id ZINC13474421and physical representation of the molecule has been given in Fig. 5 http:// zinc.docking.org/).

Molecular docking revels 24th ligand showed better binding with SmHDAC1 as compare to HsHDAC1 in terms of qualitatively as well as quantitatively. The selectivity (omitting) of SmHDAC1 (HsHDAC1) towards the ligand number 24th is because of difference in cavity forming residues between SmHDAC1 and HsHDAC1. After analyzing the docking results with respect to binding affinity, hydrophobic interactions and hydrogen bonding, we can conclude that, 24th ligand can be further used for clinical trial for the treatment of disease caused by S. mansoni.

Docking of Inhibitor 24th (ZINC13474421) with SmHDAC1 and HsHDAC1

Conclusion

For the validation, detail inside and to explore the molecular features of filtered ligand by virtual screening method, SAHA inhibitor has been taken (Fig. 6). It represents surface binding and hydrogen bond forming residues (His28, Pro29, Asp98, Glu99, His141, Phe150, His178, Phe205, Lue271, and Tyr303) which stabilize the complex by all non-bonded interactions and coordination of Zn?? ion. In case of SmHDAC1 (Fig. 7 (i) a, b and c), docking result shows that hydroxamic acid group of 24th ligand attached with eight carbon long aliphatic group, lying in the hydrophobic cavity formed by the residues Phe53, Phe59, Tyr102, Gly250, Gly251and Leu89 and bind with zinc ion at the bottom of cavity. Aromatic ring of the 24th ligand pointed outside in the complex. Cavity formed by hydrophobic residues are play crucial role to stabilize the complex. Figure 7 a1 and a3 indicating that aromatic rings of ligand were stabilized by forming hydrophobic interaction with Phe53 and Phe59 residues of SmHDAC1. Oxygen atom of carbonyl group, present in hydroxamic acid, formed hydrogen bonding with NE2 atom of His90 ˚ ). Another oxygen atom, directly attached with -NH (3.2 A of hydroxamic acid group formed hydrogen bonding with NE2 of His90, His91, Tyr253 and helped in anchor the ˚ ). Zn?? ion ligand by coordination with Zn?? ion(2.2 A plays key role in complex formation as well as complex stabilization by coordinating with Asp126, His128, Asp216 and ligand 24th (Fig. 8). Docking of 24th ligand with HsHDAC1 (Fig. 7 b1, b2 and b3) clearly indicating that it shows only surface binding. As docking score suggest it is very poor binding as compare to SmHDAC1. Ligand molecule was unable to access the catalytic pocket of HsHDAC1 Fig. 7 b2. Hydroxamic acid group of ligand was unable to bind and coordinate with catalytic residues and Zn?? ion respectively as in case of SmHDAC1.

This paper throws light on the reliable model for further structure based drug designing, concerning SmHDAC1 of S. mansoni. Molecular docking studies highlighted advantages of comparative In-silico interaction studies of SmHDAC1 and HsHDAC1. N,8-dihydroxy-8-(naphthalen-2-yl) octanamide may be used for the in vitro assays for enzyme inhibition and IC50 determination, as well as in vitro phenotypic assays. Acknowledgments This is my PhD work, supported by University Grant Commission, New Delhi, India. We thank, Dr. Navneet Mishra for is valuable suggestions, Mr. Surya Pratap singh for his critical reading and scientific discussions, and Miss Pallavi Gaur for editing the manuscript and Miss Swadha Singh for critical reading of early versions of the manuscript. Constructive comments from them helped us to make the manuscript more accurate. We apologize to our colleagues whose relevant work has not being cited because of the space limitations.

References 1. Brown M (2011) Schistosomiasis. Clin Med 11:479–482 2. Ross AG, Bartley PB, Sleigh AC, Olds GR, Li Y et al (2002) Schistosomiasis. N Engl J Med 346:1212–1220 3. Gray DJ, Ross AG, Li Y-S, McManus DP (2011) Diagnosis and management of schistosomiasis. BMJ 342:d2651 4. Domling A, Khoury K (2010) Praziquantel and Schistosomiasis. Chem Med Chem 5:1420–1434 5. Doenhoff MJ, Kusel JR, Coles GC, Cioli D (2002) Resistance of Schistosoma mansoni to praziquantel: is there a problem? Trans R Soc Trop Med Hyg 96:465–469 6. Kelly TK, De Carvalho DD, Jones PA (2010) Epigenetic modifications as therapeutic targets. Nat Biotech 28:1069–1078 7. Geutjes EJ, Bajpe PK, Bernards R (2011) Targeting the epigenome for treatment of cancer. Oncogene 31:3827–3844 8. Pierce RJ, Dubois-Abdesselem F, Lancelot J, Andrade L, Oliveira G (2012) Targeting schistosome histone modifying enzymes for drug development. Curr Pharm Des 18:3567–3578 9. Nwaka S, Hudson A (2006) Innovative lead discovery strategies for tropical diseases. Nat Rev Drug Discov 5:941–955

123

Mol Biol Rep 10. Lombardi PM, Cole KE, Dowling DP, Christianson DW (2011) Structure, mechanism, and inhibition of histone deacetylases and related metalloenzymes. Curr Opin Struct Biol 21:735–743 11. Andrews KT, Haque A, Jones MK (2012) HDAC inhibitors in parasitic diseases. Immunol Cell Biol 90:66–77 12. Mukherjee P, Pradhan A, Shah F, Tekwani BL, Avery MA (2008) Structural insights into the Plasmodium falciparumhistone deacetylase 1 (PfHDAC-1): a novel target for the development of antimalarial therapy. Bioorg Med Chem 16:5254–5265 13. Hompson JD, Gibson TJ, Plewniak F, Jeanmougin F, Higgins DG (1997) The ClustalX windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res 24:4876–4882 14. Sali A, Blundell TL (1993) Comparative protein modeling by satisfaction of spatial restraints. J Mol Biol 234:779–815 15. Lerner MG, Carlson HA (2008) Apbs plugin for pymol. University of Michigan, Ann Arbor 16. Baker N, Sept D, Joseph S, Holst M, McCammon J (2001) Electrostatics of nanosystems: application to microtubules and the ribosome. Proc Natl Acad Sci U S A 98(18):10037 17. Lovell SC, Davis IW, Arendall WB, Bakker PIW, Word JM, Prisant MG, Richardson JS, Richardson DC (2002) Structure validation by Calpha geometry: phi, psi and Cbeta deviation. Proteins: structure. Funct Genet 50:437–450 18. Wiederstein M, Sippl MJ (2007) ProSA-web: interactive web service for the recognition of errors in three dimensional structures of proteins. Nucleic Acid Res 3:407–410 19. Vriend G, Sander C (1993) Quality control of protein models: directional atomic contact analysis. J Appl Crystallogr 26:47–60. WHAT IF Web Interface. http://swift.cmbi.kun.nl/WIWWWI/ 20. Hu¨nenberger PH, Mark AE, van Gunsteren WF (1995) Fluctuation and cross-correlation analysis of protein motions observed in nanosecond molecular dynamics simulations. J Mol Biol 252:492–503 21. Wang J, Cieplak P, Kollman PA (2000) How well does a restrained electrostatic potential (RESP) model perform in calculating conformational energies of organic and biological molecules? J Comput Chem 21:1049–1074

123

22. Jorgensen WL, Chandrasekhar J, Madura JD, Impey RW, Klein ML (1983) Comparison of simple potential functions for simulating liquid water. J Chem Phys 79:926–935 23. Essmann U, Perera L, Berkowitz ML, Darden T, Lee H, Pedersen LG (1995) A smooth particle mesh Ewald method. J Chem Phys 103:8577–8593 24. Hess B, Bekker H, Berendsen HJC, Fraaije JGEM (1997) LINCS: a linear constraint solver for molecular simulations. J Comput Chem 18:1463–1472 25. Noureen N, Rashid H, Kalsoom S (2012) An efficient anticancer histone deacetylase inhibitor and its analogues for human HDAC8. Med Chem Res 21:568–577 26. Trott O, Olson AJ (2010) AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization and multithreading. J Comput Chem 31:455–461 27. Sanner M (1999) Python: a programming language for software integration and development. J Mol Graph Model 17:57–61 28. Delano WL (2002) The PyMol molecular graphics system. DeLano Scientific, Palo Alto 29. Wallace AC, Laskowski RA, Thornton JM (1995) LIGPLOT: a program to generate schematic diagrams of protein–ligand interactions. Protein Eng 8:127–134 30. Millard CJ, Watson PJ, Celardo I, Gordiyenko Y, Cowley SM, Robinson CV, Fairall L, Schwabe JW (2013) Class I HDACs share a common mechanism of regulation by inositol phosphates. Mol Cell 51:57–67 31. Frishman D, Argos P (1995) 75 % accuracy in protein secondary structure prediction. Proteins 27:329–335 32. Nardini M, Dijkstra BW (1999) a/b Hydrolase fold enzymes: the family keeps growing. Curr Opin Struct Bio 9:732–737 33. Roy S, Packman K, Jeffrey R, Tenniswood M (2005) Histone deacetylase inhibitors differentially stabilize acetylated p53 and induce cell cycle arrest or apoptosis in prostate cancer cells. Cell Death Differ 12:482–491 34. Irwin Sterling (2012) Mysinger, Bolstad and Coleman. J Chem Inf Model 52(7):1757–1768