Potent Antagonist of Alpha-2, 3-Sialyltransferase of

___________________________________________________________________________________

6 | P a g eCopyright-IJCSME

Special Issue on Computational Science, Mathematics and BiologyIJCSME- SCSMB-16-March-2016

ISSN-2349-8439

Potent Antagonist of Alpha-2, 3-Sialyltransferase of Neisseria MeningitidisShaik Parveen1, Katari Sudheer Kumar2 and Amineni Umamaheswari3

_______________________________________________________________________________________________ABSTRACT:Neisseria meningitidis is a Gram-negative bacterium causes life-threatening sepsis, meningitis and otherforms of meningococcal diseases. It initially produces symptoms like fatigue, fever, headache and can rapidly invades intothe blood stream causing a systemic infection, sepsis, disseminated intravascular coagulation, breakdown of circulation andseptic shock progress to neck stiffness, coma and even death. Multi drug resistance nature of N. meningitidis and vitalfunctioning in the pathogenesis of meningitis imposed as target for intervention in meningitis in the present study. SmallRNAs (sRNAs) are non-coding RNAs that regulate various metabolic activities by forming metabolic complexes in thebacteria. Complete genome of N. meningitidis has 249 sRNAs. Among which 68 were enzymes. CMP-N-acetylneuraminate-beta-galactosamine-alpha-2,3-sialyltransferase (LST) was selected as target to design antagonist as itplays a crucial role in cell wall synthesis, non-homologous to Homo sapiens and no alternative way to synthesize theproduct. A grid of 10 x 10 x 10 Å was generated around CDP of the crystal structure of LST. Seven published inhibitorswere subjected to shape based similarity screening using PHASE v3.9 of Schrodinger-2015 against more than 1 millioncompounds resulted LST inhibitors library with 3540 compounds. Multi-level rigid receptor docking was performed withthe library of inhibitors, which results 34 leads and were subjected to MM-GBSA calculations. 11 the best leads wereobtained by comparing the derived 34 leads with 7 published inhibitors. The best leads were further submitted to QPLD,MM-GBSA calculations and ADME predictions. Lead1 showed the least docking score with lowest binding free energy (-93.26 kcal/mol) and was subjected to IFD with MM-GBSA. MD simulations were performed up to 50 ns for lead1 dockcomplex using Desmond v3.8 of Schrodinger-2015. The potential energy, root mean square deviation, root mean squarefluctuations and protein-lead1 interactions revealed the stability of the complex in physiological environment. The proposedinhibitors would provide a frame work for development of therapeutics towards meningitis caused by N. meningitidis.Keywords: Neisseria meningitidis, meningitis, sRNA, CMP-N-acetylneuraminate-beta-galactosamine-alpha-2,3-sialyltransferase, docking, MM-GBSA, MD simulations.

___________________________________________________________________________________

IntroductionNeustria meningitides is Gram-negative aerobic,diplococci, responsible for a variety of meningococcaldiseases like cerebrospinal fever, Septicaemia, lifethreatening sepsis and bacterial meningitis. Annually, about500,000 cases of meningococcal meningitis are reportedworldwide and while survivors may suffer from seizures,hearing impairments and brain damage. It can cause severebrain damage and too fatal in 50% of cases if untreated.Meningococcal septicaemia typically causes a purpuric rashthis means that the condition may not be ignored [1]. N.meningitidis MC58 serogroup B has attained greaterinterest to researchers for developing effective vaccines anddrugs molecules. Herein, in silico analysis was carried outto find 249 sRNAs from N. meningitidis. CMP-N-acetylneuraminate-beta-galactosamine-alpha-2,3sialyltransferase (LST) is one of the sRNA candidateswhich is non-homologous to humans and plays a crucialrole in oligosaccharide synthesis, cell wall synthesis andalso not having the alternative pathway leading to the riseof meningitis symptoms.

The KEGG analysis revealed that LST of N. meningitidis isan important enzyme involved in oligosaccharide synthesisand cell wall synthesis mechanism [2].Glycosyltransferase catalyze the transfer of activatedcarbohydrate moieties from donor molecules to an acceptorsubstrate may be as simple as a second monosaccharide.The other organisms may have metal independent enzymesand may utilize many other nucleotide sugar donors. Someglycosyltransferases use lipid linked glycosyl donors,frequently a terpenoid such as dolichol or polyprenoltherefore LST of Glycosyltransferase family-52 plays mainrole in the N. meningitidis which involve in the bacterialproliferation [3]. There are some known inhibitors of LSTsuch as penicillin, ceftriaxone, vancomycin, gentamicin,chloramphenicol, rifampicin, ciprofloxacin, meropenem,which are available in literature database, PubMed [4].They are causing side effects like allergies and also havepoor pharmacological properties. Hence, an attempt wasmade in the present study to design novel inhibitors usingcomputer aided drug design to LST of N. meningitidiswhich is involved in synthesis of lipopolysaccharidebiosynthesis in outer membrane of cell wall. The methodbegins with known ligands or inhibitors of a drug target tofind their structural analogs from large to small moleculerepositories followed by structural analysis, computationaldocking and binding free energy calculations along with theMD simulations. Existing treatment showing the multi-drugresistance nature of the meningococcal strains N.meningitides, we need the deeper understanding of the

Bioinformatics Centre, Department of Bioinformatics, SriVenkateswara Institute of Medical Sciences University,

Tirupati, Andhra Pradesh – 517507; Ph: +91-877-2287727 Email: [email protected]

SVCE_RMD

Typewritten text

DOI: 10.18645/IJCSME.SPB.0002

___________________________________________________________________________________



ISSN-2349-8439

virulence factors hence, this project was selected andfurther work was carried out through computationalapproaches.

Material and MethodsA. Identification of unique sRNA in N. meningitidis:The whole genome sequence of N. meningitidis wasretrieved from the National center for Biotechnologyinformation (NCBI). sRNAs are crucial role in regulatingvarious cellular processes of pathogenic organisms. Thegenome of N. meningitidis was submitted to sRNAPredicttool to predict the sRNA candidates [5].The sRNAcandidates were checked for non homology against Homosapiens using BLASTp analysis. The enzyme targets thatare non-homologous to Homo sapiens with <30% identity[6], plays crucial role in essential metabolic pathways andunique to the pathogen having no other alternative pathwayfor biosynthesis of the product were considered as theputative drug targets of N. meningitidis using KyotoEncyclopedia of Genes and Genome (KEGG) [7], [8].Protein preparation and inhibitor binding siteprediction:CMP-N-acetylneuraminate-beta-galactosamine-alpha-2,3-sialyltransferase (LST) co-crystallized with inhibitorcytosine 51 triphosphate (CDP), 2YK6 was selected as thedrug target and retrieved from protein data bank (PDB).The protein was prepared by adjusting bond orders,replacing missing sidechains and loops by Prime, optimizedby Epik and minimized with OPLS 2005. The inhibitorbinding site residues of LST were considered usingliterature and lig-plot analysis [9]. A grid of 10 x 10 x 10 Åwas defined around the inhibitor binding site residues ofLST.C.Virtual screening and dockingEight existing published inhibitors of 2YK6 were searchedas query against 4.5 Lakhs small molecules of ASINEXdatabase through shape based similarity screening usingPHASE v3.9. Ligand preparation was done for the shapescreened hits using LigPrep v3.0 module of Schrödinger.The Epik v2.8 was employed to enumerate tautomers(10,000 for each ligand) and protonation states of ligands[10]. Reactive filters and Lipinski’s filters were applied torefine the generated tautomers using QikProp v3.6. All theeight published inhibitors and the hits were docked withinthe grid of LST using grid based ligand docking withenergetics (GLIDE) v6.3 [11]. Glide offers virtualscreening workflow or rigid receptor docking, thatimplements three tier docking protocol such as highthroughput virtual screening (HTVS), standard precision(SP) and extra precision (XP) docking methods respectively[12]. As the binding free energy was more accurate todefine the binding affinity than the docking score, bindingfree energy (ΔG) by MM-GBSA were calculated for theligand receptor dock complexes through Prime v3.6 [13].

QikProp v3.6 was used to calculate ADME/T properties(absorption, distribution, metabolism, excretion andtoxicity) for all the ligands employed in the virtualscreening workflow protocol. Qikprop is a quick andaccurately predicts physico-chemical significant descriptorsand pharmaceutically relevant properties of organicmolecules [14], [15]. The ADME/T properties of leads andeight published inhibitors were considered and allow forcomparing with those of 95% known FDA approved drugs.D. Molecular dynamicsimulations:Molecular dynamics (MD) simulations of 50 ns for the bestdock complex was performed using Desmond v3.8implemented in Schrödinger package 2014 [16]. The initialsteps of MD simulations were performed by applyingoptimized potential liquid simulations (OPLS) 2005molecular mechanics force field and the water moleculeswere placed to the ligand-receptor complex with simplepoint charge water model simulated through the multistepMD protocols of Maestro v9.8 [17]. Briefly, full systemminimization with restraints on solute was performed formaximum iterations of a hybrid of the sharpest lineage andthe limited memory Broyden-Fletcher-Gold farb-Shanno(LBFGS) algorithms, with a merging threshold of 50.0kcal/mol/Å2. Two similar minimizations without anyrestraints were performed with a convergence threshold of5.0 kcal/mol/Å2. The non-hydrogen solute atoms wererestrained in the NVT ensemble (constant number of atomsN, volume V and temperature T) using 10 ps simulationtime and temperature of 10 K. Simulations restraining nonhydrogen’s solute atoms were performed in the NPTensemble (constant number of atoms N, pressure P andtemperature T) for 12 ps simulation time and temperatureof 10 K. Further, NPT ensemble for a simulation time of 24ps was carried out with restraining all non-hydrogen soluteatoms at a temperature of 300 K and then the system iscarried for 50 ns MD simulations in NPT ensemble [18],[19].

Results and DiscussionPrediction of unique sRNA in N. meningitidis:The whole genome of Neisseria meningitidis has a size of2.18 Mb with the GC content of 51.18% and contains 2,065coding genes that encode 1,909 proteins. Small RNAs(sRNAs) are non-coding genes that constitute large andheterogeneous bacterial gene expression regulators.sRNAPredict tool resulted 249 sRNA candidates, amongwhich 101 proteins, 62 hypothetical proteins, 10 trRNAs, 8pseudo genes and 68 enzymes were defined. As enzymesare known to be involved in crucial metabolic pathways ofpathogens and have definite site for inhibition, henceenzymes were taken for further study. Among the 68enzymes, LST was found to be non-homologous to Homosapiens, unique to the pathogen without any alternative

___________________________________________________________________________________



ISSN-2349-8439

pathway and also plays a crucial role in the cell wallsynthesis. Thus CMP-N-acetylneuraminate-beta-galactosamine-alpha-2,3-sialyltransferase (LST) wasselected for designing novel antagonists to inhibit thefunction of LST which further halts the propagation of N.meningitidis.Protein preparation and inhibitor binding sitepredictionThe co-crystal structure of LST was retrieved from thePDB, visualized by PDBsum and PyMOL, prepared andminimized in the protein preparation wizard. The residuessuch as Asp165, Gly-166,Thr-167, Ile-214, Phe-215, Lue-234, Phe-235, Lue-254, Gly-255, Ala-278, Pro-279, His-280, Pro-281, Val-298, Ser-322, Gly-323, Ala-324 and Thr-327 were found to be present with in 4 Å region around theCDP of LST. The carboxylate side chain of Glu-300 formsa bidentate anchor to the C2 and C3 hydroxyl groups on theribose ring. The cytidine ring is sandwiched between thealkyl side chains of Pro-281 and Ile-299 and held by twohydrogen bonds from the backbone Fig. 1(A) [20]. Theimportant residues of LST like Ala-278, Arg-280, Ile-299,Glu-300 and Gly-323 were targetted by generating grid forligand docking using glide.Virtual screening and dockingShape screening of eight published inhibitors of LST withASINEX database contributed 408 structural analogs.LigPrep was employed to generate multiple conformationsfrom the published inhibitors and structural analogmolecules. The 408 conformations were passed through theLipinski’s filter, reactive filters and subsequently 387ligands were obtained. The grid of LST and 387 ligandswere applied for three levels of docking (HTVS, SP andXP) approaches. In high throughput virtual screening mode(HTVS) 96 ligands were obtained, 46 ligands weregenerated in standard precision (SP) docking and in extraprecision mode (XP) which is more accurate dockingmethod generated 34 ligands. Further, binding free energiesfor LST dock complexes of 34 ligands and eight publishedinhibitors were calculated for through Prime/MM-GBSA.

Fig.1: Molecular interactions (A). LST-CDP (B). LST-lead1 docking

The results revealed that eleven leads possess betterdocking scores (XPG) and binding free energies (∆G)(TABLE Ι) with hydrogen bonds, van der Waalsinterctionsand pharmocological properties in maintainingthe stability of the docking complex compared with theeight published inhibitors of LST (TABLE .ΙΙ and ΙΙΙ).Among the nineteen dock complexes, the complex with theleast glide and ∆G values signifies the stability andspontaneous formation of LST-lead complex, thereforelead1 possess the least XPG score of -13.18 kcal/mol andthe lowest binding free energy ∆G of -93.26 kcal/moltowards LST of N. meningitidis. Lead1 exerts threehydrogen bonds with LST in docking, one hydrogen bondby the nitrogen group of indole moiety with the backboneof Ile-299 and two hydrogen bonds by the amine groups ofpropanamide moiety with the sidechain atoms of Glu-300.The inhibitor binding site residues Leu-254, Pro-279,His-280, Pro-281, Val-298, Ser-322 and Ala-324 of LSTwere also well contributed van der Waals interactionswithin 4 Å region of lead1 Fig. 1(B).

TABLE 1: XPG AND ∆G SCORES OF LEADS ANDPUBLISHED INHIBITORS

___________________________________________________________________________________



ISSN-2349-8439





___________________________________________________________________________________



ISSN-2349-8439





___________________________________________________________________________________



ISSN-2349-8439

TABLE 2: ADME/T PROPERTIES OF 11 LEADS

TABLE 3: ADME/T PROPERTIES OF PUBLISHEDINHIBITORS

Molecular dynamics simulations

The inhibitor binding site residues of LST showedinteractions with lead1 during 10,416 trajectories of 50 nsMD simulations. The atoms of key catalytic residue, Glu-300 sidechain were involved in hydrogen bond formation in88% trajectories and the backbone atom was involved inwater mediated hydrogen bond in 78% trajectories withlead1 Fig. 2(B). The potential energy plot showed that theenergy of the LST-lead1 complex was consistent during 50ns MD simulations run with an overall average energy of -206360.16kcal/mol Fig. 2(C). The catalytic residues of LSTsuch as Ile-299, Glu-300 are interacting with lead1.Thereby blocking of these residues will further hinder thesynthesis of LOS involved in lipopolysaccharides (LPS).The inhibitor binding site residues in LST-CDP ligplot,Leu-254, Ala-278, Pro-279, His-280, Pro-281, Arg-282,Val-298, Ser-322, Gly-323 and Ala-324 were observed inthe LST-lead1 complex during 50 ns MD simulations Fig.2(A).

Fig.2: 50 ns molecular dynamics simulations of LST-lead1 (A). Lead1interactions (B). Lead1 contacts (C). Energy plot

The interactions fractions of the LST-lead1 complex in 50ns MD simulations revealed the lowest potential energy,root mean square deviations and root mean squarefluctuations to maintain the stability of the complex duringphysiological conditions. The bacterial outer membrane isfound in Gram-negative bacteria and its composition isdistinct from that of the inner cytoplasmic cell membrane.The outer leaflet of the outer membraneofN.

___________________________________________________________________________________



ISSN-2349-8439







___________________________________________________________________________________



ISSN-2349-8439







___________________________________________________________________________________



ISSN-2349-8439

meningitidesacts asendotoxinsand is responsible forsepticshockand hemorrhage due to the destruction of red bloodcells.The polysaccharidecapsulewhich preventshostphagocytosis aid in evasion of the host immuneresponse and fimbriaemediates attachment of the pathogento theepithelial cellsof the nasopharynx thus inducing theinfection. The LPS transport machinery is composed ofLptA, LptB, LptC, LptD, LptE and were organized by LPSassembly pathway. The depletion of any one of theseenzyme blocks the ordered LPS assemblypathway andresults with the defects in outer membrane biogenesis. Thusresults from virtual screening, docking and dynamicsstudies infers that, the proposed inhibitors hinders theorganized LPS assembly pathway, in turn blocks thebacterial cell wall synthesis that is necessary for thesurvival of the pathogen. Thus by hindering the cell wallsynthesis, proliferation and multiplication of the N.meningitidescan be reduced by treating with the proposedleads against bacterial meningitis.

ConclusionThe study on LST of Neisseria meningitides

involves proteomic analysis and followed by in silicoapproach to find out the more potential inhibitors to blockits pathogenic activity. The proteomic analysis of LSTrevealed the crucial residues for targeting the functionalityof the enzyme and the rational drug designing approachwas employed by virtual screening, molecular docking andmolecular dynamics simulations. Upon virtual screeningand docking, eleven leads with LST possess better bindingfree energies compared to the existing eight inhibitors ofLST were proposed as potential inhibitors. Among them,lead1 showed the best binding free energy ∆G score of -93.26 kcal/mol and binding orientation by forminghydrogen bonds, van der Waal interactions and goodpharmacological properties as par with 95% of the FDAapproved drug molecules to block the functional activity ofLST by blocking the cell wall synthesis mechanism in N.meningitides and further decreases the proliferation ofpathogen causing meningitis. The potential energy andLST-lead1 interactions also revealed the stability of thecomplex in 50 ns MD simulations. Therefore, lead1 mightbe a potent antagonist against LST, an essential enzyme forcell wall biosynthesis and it was targeted to arrest variouslife threatening diseases like meningitis and sepsis inducedby N. meningitides.

AcknowledgmentThe author SP is highly thankful for DBT, Ministry ofScience and Technology, Govt. of India for providingDBT-studentship. The authors are acknowledged to DBTfor the support of providing necessary infrastructurefacilities to carry out the research work at BIF(No.BT/BI/25/001/2006).

References1. C. Genco, L. Wetzler, “Neisseria Molecular

Mechanisms of Pathogenesis,” Caister AcademicPress. 2010, 5: 51-56.

2. P. Stefanelli, “EmergingresistanceinNeisseriameningitidisand Neisseria gonorrhoeae,” ExpertRev Anti Infect Ther.2011, 2:237-244.

3. J.E. Sadler, J.I. Rearick, J.C. Paulson, R..L. Hill"Enzymatic characterization of beta D-galactosidealpha2 leads to 3 sialyltransferase from porcinesubmaxillary gland". J. Biol. Chem. 1979, 254(11): 4444–51. PMID438198.

4. J.I. Alcala, J.C. Rearick, R.L. Paulson, "Enzymaticcharacterization of beta D-galactoside alpha-2,3-sialyltransferase from porcine submaxillarygland," J. Biol. Chem, 2004, vol. 25, pp. 4476-68.

5. N. Sivakumari, P. Chiranjeevi, D. Pradhan and A.Umamaheswari, “Discovery of Potent Inhibitorsagainst GTP Pyrophosphokinase of NeisseriameningitidisSerogroup B,” International Journalof Scientific & Engineering Research, Volume 6,Issue 2, February-2015, 273 ISSN 2229-5518.

6. N. Pradeep, I.V. Priyadarshini, D. Pradhan, M.Munikumar, S. Sandeep, K. Hema, B. Vengammaand A. Umamaheswari, ‘E-pharmacophore-basedvirtual screening to identify GSK-3β inhibitors,”Journal of Receptors and Signal Transduction.2015, 35 : 1-14.

7. I.V. Priyadarshini, D. Pradhan, M. Munikumar, S.Swargam, A. Umamaheswari * and D. Rajasekhar,Genome-based approaches to develop epitope-driven subunit vaccines against pathogens ofinfective endocarditis. Journal of BiomolecularStructure and Dynamics, 2014, 32(6): 876-889.

8. M. Munikumar, I.V. Priyadarshini, D. Pradhan, A.Umamaheswari, B. Vengamma, “Computationalapproaches to identify common subunit vaccinecandidates against bacterial meningitis,”Interdisciplinary Sciences: Computational LifeSciences, 2013, vol. 5, pp. 155-164.

9. S. Parveen, N. Pradeep, K. Hema, A.Umamaheswari, “Prediction of Novel Inhibitorsagainst Exodeoxyribonuclease І of H. influenzaethrough In Silico Approach”. InternationalJournal of Scientific and Engineering Research;2015, 6(2): 217-221.

10. S. Sandeep, D. Pradhan, N. Pradeep, K. Hema, V.Siva Krishna and A. Umamaheswari, “Structureguided novel lead molecules against ERKproteins: application of multiple docking andmolecular dynamics studies,” Journal of

___________________________________________________________________________________



ISSN-2349-8439

Biomolecular Structure and Dynamics; 2015, 33(supplement 1): 134-135.

11. K. Hema, S. Sandeep, N. Pradeep and A.Umamaheswari “In silico agonist for humanextracellular superoxide dismutase SOD3,” OnlineJournal of Bioinformatics, 2016, 17(1): 29-40.

12. A. Umamaheswari, D. Pradhan, I.V. Priyadarshini,M. Munikumar and P.V.L.N. SrinivasaRao,“Computational analysis of K-Hefutoxininteraction with Kv channels and L-carnitinemolecule for scorpion envenomation,” OnlineJournal of Bioinformatics. 2011,12(2): 304-322

13. A. Umamaheswari, D. Pradhan. and MHemanthkumar, “In silico identification ofcommon putative drug targets inLeptospirainterrogans,”Journal of ChemicalBiology. (2010) 3(4):165-173.

14. W.L. Jorgensen, D.S. Maxwell, J. Tirado Rives“Development and Testing of the OPLS All-AtomForce Field on Conformational Energetics andProperties of Organic Liquids,” J. Am. ChemSoc;1996: 118 (45): 11225–11236.

15. C.A. Lipinski, F. Lombardo, B.W. Dominy, P.J.Feeney, “Experimental and computationalapproaches to estimate solubility and permeabilityin drug discovery and development settings,” Adv.Drug Delivery Rev; 1997: 23(3): 4-25.

16. K. Sudheer Kumar, N. Pradeep, S. Sandeep, K.Hema, P. Chiranjeevi and A. Umamaheswari,“Inhibitor design against JNK1 through e-pharmacophore modeling docking and moleculardynamics simulations,” Journal of Receptors andSignal Transduction (2016).10.3109/10799893.2016.1141955

17. D. Pradhan, I.V. Priyadarshini, M. Munikumar, S.Sandeep and A. Umamaheswari, “161 Discoveryof potent KdsA inhibitors ofLeptospirainterrogans through homologymodeling, docking, and molecular dynamicssimulations,” Journal of Biomolecular Structureand Dynamics, 2013, ISSN: 0739-1102; 1538-0254.

18. G.A. Kaminski, R.A. Friesner, J. Tirado Rives,W.L. Jorgensen, “Evaluation andreparametrization of the OPLS-AA force field forprotein via comparison with accurate quantumchemical calculations on peptides,” J PhysChemB, 2001; 105:6474.

19. Maestro, version 9.8, Schrödinger, LLC, NewYork, NY, 2014.

20. L.Y. Lin,B. Rakic,C.P. Chiu,E. Lameignere,W.W.Wakarchuk,S.G. Withers, N.C.Strynadka.“Structure and mechanism of the

lipooligosaccharidesialyltransferase fromNeisseria meningitidis,”. Oct-2011, J Biol Chem.28;286(43):37237-48.

Documents

Potent Antagonist of Alpha-2, 3-Sialyltransferase of