INTRODUCTION:

Kaposi’s sarcoma-associated herpesvirus (KSHV) Human herpesvirus 8 (HHV-8) belongs to the gamma herpesvirus subfamily. HHV-8 is a double-stranded DNA tumor virus that is responsible for causing three malignancies, namely, Kaposi’s sarcoma, primary effusion lymphoma and multicentric Castleman’s disease (MCD)¹^,²^,³HHV-8 is associated with all forms of Kaposi sarcoma (KS) including classic, endemic, transplant-related, and AIDS-related, as well as rare neoplastic disorders and the lymph proliferative disorder known as multicentric Castleman’s disease (MCD)⁴.The mechanism of how HHV-8 associated with the tumors is still not clear, but the infection of HHV-8 precedes with tumor development⁵.

The mode of transmission of HHV-8 still not well established. However, epidemiologic and virologic data suggest that saliva is one of the sources of the infectious virus and may be an important route of transmission⁶^,⁷. The currently approved treatments for herpes virus suffer from poor efficacy because of viral resistance mutations, or severe dose-limiting side effects such as myelosuppression or nephrotoxicity⁸.

In all human herpesviruses (HHV) protease enzyme is present, which play a critical role in the formation of the mature capsid. Previous study by Gao et al⁹. showed that genetic deletion of protease in HSV-1 lead to prevention of capsid maturation, that is essential for viral replication, thus make proteases as a potential target for HHV inhibition¹⁰.

Stroud and co-workers crystallized the structures of KSHV protease¹⁰^,¹¹ to explore the active site residues and to understand the mechanism of action of the inhibitors at the binding site. They also reported the series of broad-spectrum HHVs inhibitors using nuclear magnetic resonance (NMR) assays as well as X-ray crystallographic studies. The 3D-structure showed that KHSV protease structure consisting of seven β strands and six major helices and a relatively shallow substrate binding pocket with a strict preference for alanine at P1 and serine at P1′. It also contains a conserved aromatic hot spot, which is present in all nine human herpes virus proteases, suggesting the potential for the development of broadly therapeutic small molecules that could inhibit HHV protease enzyme activity by disrupting the protease dimer formation or disrupting the interactions between formed dimer.

The present study aimed at identification of novel potent HHV protease inhibitors by utilizing the already known potent HHV protease inhibitors.This will help to identify the chemical features that are necessary required for broad-spectrum allosteric inhibition of human HHV proteases. To achive or aim we have empoyled pharmacophore modeling and molecular docking techniques.

MATERIALS AND METHODS:

Ligand preparation using LigPrep:

All the compounds used in the study were imported in the Maestro and prepared using Ligprep module of Maestro [12]. LigPrep uses the following parameters to minimize the ligands 1) Ligprep converts the 2d to 3d structures 2) adds hydrogen atoms 3) generate stereoisomer’s and ionization states at pH 7.0 4) and generates low energy conformations followed by energy minimization using the OPLS-2005 force field.

Protein preparation and Receptor grid generation:

The 3D structure of KSHV protease (PDB ID: 4P3H) was downloaded from the protein data bank. The protein was prepared using the protein preparation wizard using the Maestro software (Schrödinger Suite 2019-1 Protein Preparation Wizard; Epik). During protein preparation the protein was pre processed to remove the water molecules in the protein except in the active site region, hydrogen atoms were added, assigns the correct bond orders and chargers to the protein, missing side chains, and loops were added using Prime. Out of 314 water molecules present in the crystal structure, only 12 water molecules were retained because these water molecules show significant water-mediated contacts with the protein. Later the protein was minimized using the OPLS 2005 force field and minimization was terminated when the protein RMSD range exceeds the 0.3 Å. Soon after protein minimization, the protein was taken for receptor grid generation. The receptor grid was generated by selecting the co-crystal ligand, 25G present in the active site of the protein using the grid dimensions i.e 10 × 10 × 10 Å.

Glide XP docking:

Docking was performed using Glide extra precision (XP) module of Maestro software [13]. The ligands were docked onto the structure of KSHV protease (PDB ID: 4P3H). Glide uses a series of hierarchical filters to search for possible locations of the ligand in the active-site region of the receptor. The maximum of 10 poses for each ligand were created and docked. The best pose was selected based on the Glide score e is based on ChemScore, which includes a steric-clash term and also adds buried polar terms devised by Schrödinger to penalize electrostatic mismatches

GScore = 0.065*vdW + 0.130*Coul + Lipo + Hbond + Metal + BuryP + RotB + Site

Where vdW is vdW energy, Coul is coulombic energy, Lipo is the lipophilic term, Hbond is the hydrogen bonding term, Metal is the metal binding term, BuryP is a penalty for buried polar groups, RotB is a penalty for freezing rotatable bonds, and Site is polar interactions in the active site. The docking calculations were performed using the OPLS2005 force field. The results of the docking were then quantified in terms of the Glide score and Glide energy. In this study, the best pose was also selected by evaluating the Root mean square deviation of predicted pose with crystal ligand pose, and an active site residue interactions.

Validation of Glide XP docking:

The Glide XP module was validated by re-docking of the co-crystal ligand with the 3D structure of KSHV protease (PDB ID: 4P3H). A program that is able to return the poses below root mean square deviation of (RMSD) 1.5Å to 2Å are considered to have performed successfully. Morover, the best pose was chosen for further analysis.

Energy based pharmacophore modeling (E-pharmacophore):

E-pharmacophore modeling was performed using the crystal structure of the Kaposi’s sarcoma-associated herpesvirus (KSHV) protease in complex with N-[2-benzyl-4-(1H-tetrazol-5-yl) phenyl]-6- (cyclohexylmethyl) pyridine-2-carboxamide. The crystal structure for KSHV protease (PDB ID: 4P3H) was downloaded from the PDB website. E-pharmacophore can only be performed when the crystal structure of protein in complex with substrate or inhibitor was available [14]. Before generating the structure-based pharmacophore model using E-pharmacophore script, the 3D structure of KSHV protease was prepared using the protein preparation wizard of Schrodinger software. Further grid was generated using a receptor grid generation module of Glide software. After grid generation, the 25G ligand molecule was extracted from the optimized crystal structure to perform glide docking. Docking was performed using the grid file and ligand, 25G using Glide XP docking. The docked structure (*pv.mae) was taken for E-pharmacophore modeling. The pose viewer file (*pv.mae) possess ligand poses and interactions of the ligand with active site residues of the protein. The best pose was chosen manually and then taken for E-pharmacophore model generation.E-pharmacophore generates the pharmacophore based on the protein-ligand energetic terms computed by the glide scoring function. Each pharmacophore site was ranked based on the energy (Kcal/mol). The H-bond acceptor (HBA), H-bond donor (HBD), Hydrophobic (H), negative (N), positive (P), and aromatic ring (R) features were used to create pharmacophore sites. The best pharmacophore features were chosen based on the energy and also manual analysis of protein-ligand complex and further exported as a hypothesis file (*.xyz) using write hypothesis with selected features option.

Pharmacophore screening:

Structure-based pharmacophore screening was performed using Phase Find matches to hypothesis module of maestro software [15]. The KEGG phytochemical compounds and ZINC chemical databases were given as an input for pharmacophore screening. Phase find the hits by taking the pharmacophore hypothesis as a template for 3D chemical databases to search the novel potential ligands. Phase uses finding and fetching approaches to search the ligand or hits. The hits obtained were ranked based on the fitness score, a measure of how well the pharmacophore sites matches with the reference ligand hypothesis. Fitness score range from 0 to 3. Higher the fitness score indicates that the hits retrieved from the database were perfectly mapped with the hypothesis.

RESULTS AND DISCUSSION:

Validation of Glide docking:

Molecular docking was performed to predict the binding pose of the ligand 25G, onto the active site of the KSHV protease. The ligand 25G, was extracted from 3D structure of the protein and minimized using LigPrep software. The minimized ligand was redocked using Glide XP to determine the lowest energy pose versus co-crystal ligand pose using RMSD. The RMSD was calculated by superimposing the two poses, and the RMSD obtained was 0.201Å. [16] This indicates that Glide XP docking program was successful in reproducing the native pose. Figure 1 shows the alignment of redocked (cyan) and co crystal ligands (green). The glide score of the redocked pose was -8.388 Kcal/mol. The ligand binds with the active site residues SER 191, LEU 79, LEU 193, PRO 192, PHE 76, TRP 109, LEU 110, ILE 105 and LEU 83 as similar to co-crystal ligand.

Figure 1: Docking validation. Superimposition of co-crystal ligand (green) with redocked pose (cyan). Dotted line indicated H-bond interactions

Development of E-pharmacophore model:

The docked complex which has similar pose as the co-crystal ligand was taken as an input for E-pharmacophore script. The script predicts six pharmacophore features or sites which includes one H-bond acceptor (HBA), one negative (N), three-ring aromatics (RA) and one hydrophobic (HY). The generated pharmacophore model was mapped with the bonded and non-bonded interactions to analyze the essential amino acid residue interactions with the active site residues of the protein. The generated pharmacophore was manually mapped with the active site residues to check the accuracy of the model. Among six pharmacophore sites, Negative group (N) was omitted since it does not form bonded or non bonded interactions with the active site. The refined pharmacophore model contains one hydrophobic, one HBA and three RA groups. Table 1 shows the scores for pharmacophoric sites based on the energetic terms from glide XP docking. The HBA2 form accepts one H-bond from the Ser 191. The pharmacophore has three RA (RA11, RA12, and RA13) where RA11 forms hydrophobic interactions with Leu 193, Leu 83 and Leu 79, RA12 share hydrophobic interactions with Pro 192, Phe 76, Leu 79 and Ile 44 whereas RA13 display hydrophobic interactions with Trp 109, Val 89 and Ala 90. The HY8 share hydrophobic interactions with the Ile 105 and Leu 110. Figure 2 shows the refined pharmacophore model with inter site distances and Figure 3 displays the mapping of the pharmacophore model with the active site residues of KSHV protease.

Figure 2

Figure 3

Figure 2 and 3: Refined structure-based pharmacophore hypotheses developed using e-pharmacophore module of Maestro software. Pharmacophore features are represented as brick red for H-bond acceptor A2, and R11, R12, and R13 brown color indicates the ring aromatic features green color for hydrophobic feature H8. Distances between the pharmacophore sites are denoted in angstroms (Å). Figure 3: displays the mapping of pharmacophore features with the active site of KSHV protease.

Table.1: The scores for pharmacophoric sites based on the energetic terms from glide XP docking

Pharmacophore sites	Energy Score (Kcal/mol)
H8	-1.22
A2	-0.70
R11	-0.51
R12	-0.73
R13	-0.99

Pharmacophore-based screening:

The refined pharmacophore model was screened against KEGG phytochemical compounds database using Phase module. Out of 2000 compounds, 8 compounds were retrieved with the fitness score ranging from 0.7 to 1.6. Compounds with the above 1.5 fitness score were taken for docking studies.

Molecular docking studies:

Out of 8 compounds 1 compound, namely Ginkgetin show above 1.5 fitness score. Molecular docking studies were performed for the Ginkgetin along with known ligand, 25G to predict the binding mode. Ginkgetin found to show similar glide score as a known ligand, 25G. Ginkgetin shares hydrophobic interactions with the conserved active site residues amino acid residues Ile 44, Pro 192, Trp 109, Phe 76 and Leu 83 thereby inhibiting the function of KSHV protease and also other HHV proteases. The glide score obtained for Ginkgetin is -8.014 Kcal/mol. Figure 4 shows the binding mode of the Ginkgetin in the active site of KSHV protease.

Figure 4: Binding of Ginkgetin in the active site of KSHV protease. Green dotted lines show hydrophobic interactions.

Our docking results were in agreement with the previous experimental studies of Ginkgetin with HSV-1[17]. Ginkgetin, which is a biflavone originally isolated from Ginkgo biloba.

CONCLUSION:

In this study, we developed an energy-based pharmacophore model by utilizing the crystal structure of KSHV protease (PDB ID: 4P3H). The pharmacophore model developed was refined by keeping in mind the conserved active site residues of HHV proteases. The refined model has one H-bond acceptor, three-ring aromatic and one hydrophobic group (ARRRH). This refined model was used as a 3D search for identification of novel inhibitors. The novel inhibitors were shortlisted based on the high fitness score and glide score. Docking studies revealed the novel compound, Ginkgetin bound at the conserved active site of HHV proteases, which was in aggrement with experimental studies. This study explored the essential chemical features of known inhibitor that can be useful to identify the novel broad-spectrum HHV proteases.

ACKNOWLEDGEMENTS:

SK would like to thank Vignan’s University for their constant support. NA would like to thank Center for High Performance and Computing (CHPC), South Africa for providing the computational resources.

REFERENCES:

1. Cohen A, Wolf DG, Guttman Yassky E, Sarid R. KAPOSI'S Sarcoma–associated Herpesvirus: Clinical, Diagnostic, and Epidemiological Aspects. Critical reviews in clinical laboratory sciences. 2005; 42(2):101-153.

2. Wen KW, Damania B. Kaposi sarcoma-associated herpesvirus (KSHV): molecular biology and oncogenesis. Cancer letters. 2010; 289(2): 140-150.

3. Jacobs SR, Gregory SM, West JA, Wollish AC, Bennett CL, Blackbourn DJ, Heise MT, Damania B. The viral interferon regulatory factors of Kaposi's sarcoma-associated herpesvirus differ in their inhibition of interferon activation mediated by Toll-like receptor 3. Journal of virology. 2013; 87(2):798-806.

4. Sunil M, Reid E, Lechowicz MJ. Update on HHV-8-associated malignancies. Current infectious disease reports. 2010;12(2): 147-154.

5. Beral V, Newton R, Sitas F. Human herpesvirus 8 and cancer. Journal of the National cancer Institute. 1999; 91(17): 1440-1441.

6. Pica F, Volpi A. Transmission of human herpesvirus 8: an update. Current opinion in infectious diseases. 2007; 20(2):152-156.

7. HHV-8 Herpes Adult and Adolescent Opportunistic Infection. https://aidsinfo.nih.gov/guidelines/html/4/adult-and-adolescent-opportunistic-in.

8. Kimberlin DW, Whitley RJ. Antiviral therapy of HSV-1 and-2. In Human herpesviruses: biology, therapy, and immunoprophylaxis, Cambridge University Press: 2007.

9. Gao M, Matusick-Kumar L, Hurlburt W, Di Tusa SF, Newcomb WW, Brown JC, McCann P, Deckman I, Colonno RJ. The protease of herpes simplex virus type 1 is essential for functional capsid formation and viral growth. Journal of virology. 1994; 68(6): 3702-3712.

10. Reiling KK, Pray TR, Craik CS, Stroud RM. Functional consequences of the Kaposi's sarcoma-associated herpesvirus protease structure: regulation of activity and dimerization by conserved structural elements. Biochemistry. 2000; 39(42): 12796-12803.

11. Gable JE, Lee GM, Jaishankar P, Hearn BR, Waddling CA, Renslo AR, Craik CS. Broad-spectrum allosteric inhibition of herpesvirus proteases. Biochemistry. 2014; 53(28): 4648-4660.

12. Good AC, Oprea TI. Optimization of CAMD techniques 3. Virtual screening enrichment studies: a help or hindrance in tool selection? Journal of Computer Aided Molecular Design. 2008; 22(34):169–178.

13. Friesner RA, Murphy RB, Repasky, MP, Frye LL, Greenwood JR, Halgren TA, Sanschagrin PC, Mainz DT. Extra Precision Glide: Docking and Scoring Incorporating a Model of Hydrophobic Enclosure for Protein-Ligand Complexes. Journal of Medicinal Chemistry. 2006; 49: 6177–6196

14. Salam NK, Nuti R, Sherman W. Novel Method for Generating Structure-Based Pharmacophores Using Energetic Analysis. Jorunal of Chemical Information and. Modeling. 2009; 49: 2356–2368

15. Dixon SL, Smondyrev AM, Rao SN. PHASE: A Novel Approach to Pharmacophore Modeling and 3D Database Searching. Chemical Biology & Drug Design. 2006; 67: 370-372

16. Hevener KE, Zhao W, Ball DM, Babaoglu K, Qi J, White SW, and Lee RE. Validation of molecular docking programs for virtual screening against dihydropteroate synthase. Journal of chemical information and modeling, 2009; 49(2): 444-460.

17. Hayashi KYOKO, Hayashi T, and Morita N. Mechanism of action of the antiherpes virus biflavone ginkgetin. Antimicrobial agents and chemotherapy. 1992; 36(9): 1890-1893.

Received on 01.04.2019 Modified on 20.05.2019

Research J. Pharm. and Tech. 2019; 12(11):5177-5181.

DOI: 10.5958/0974-360X.2019.00896.5