research article peptide-ligand binding modeling of sirna...

Research ArticlePeptide-Ligand Binding Modeling of siRNA withCell-Penetrating Peptides

Alfonso T Garciacutea-Sosa1 Indrek Tulp1 Kent Langel2 and Uumllo Langel23

1 Institute of Chemistry University of Tartu Ravila 14a 50411 Tartu Estonia2 Institute of Technology University of Tartu Nooruse 1 50411 Tartu Estonia3 Department of Neurochemistry Stockholm University 106 91 Stockholm Sweden

Correspondence should be addressed to Alfonso T Garcıa-Sosa alfonsogutee

Received 27 February 2014 Accepted 15 May 2014 Published 24 July 2014

Academic Editor Maria Jerzykiewicz

Copyright copy 2014 Alfonso T Garcıa-Sosa et al This is an open access article distributed under the Creative Commons AttributionLicense which permits unrestricted use distribution and reproduction in any medium provided the original work is properlycited

The binding affinity of a series of cell-penetrating peptides (CPP) was modeled through docking and making use of the numberof intermolecular hydrogen bonds lipophilic contacts and the number of sp3 molecular orbital hybridization carbons The newranking of the peptides is consistent with the experimentally determined efficiency in the downregulation of luciferase activitywhich includes the peptidesrsquo ability to bind and deliver the siRNA into the cellThe predicted structures of the complexes of peptidesto siRNA were stable throughout 10 ns long explicit water molecular dynamics simulations The stability and binding affinity ofpeptide-siRNA complexes was related to the sidechains and modifications of the CPPs with the stearyl and quinoline groupsimproving affinity and stability The reranking of the peptides docked to siRNA together with explicit water molecular dynamicssimulations appears to be well suited to describe and predict the interaction of CPPs with siRNA

1 Introduction

Cell-penetrating peptides (CPPs) are short (generally five to40 amino acid) peptide sequences that are able to deliverbiologically active cargos into the cell cytoplasm and nucleusby means of their ability to cross cell membranes [1 2]Molecules of particular interest for delivery across mem-branes are drugs and nucleic acids such as small interferingribonucleic acids (siRNA) given that this allows the normallyinactive siRNA access to bind to a cellrsquos specific nucleotidesequence that performs a given task such as regulatingendogenous genes [3] To better develop silencing genetechnology and its associated benefits a better understandingof themechanism inwhich CPPs bind to geneticmaterial andhelp introduce it into cells is neededThis would also providesuggestions on how to design peptides with better efficiency

Positively charged (basic) groups on amino acids likelysine and arginine provide features that are helpful for bind-ing to siRNA CPPs can bind covalently or noncovalently tosiRNA Arginine-rich motifs zinc fingers RNA recognitionmotifs small molecules and tethered approaches among

others have been used to bind RNA [4] Recent tools canhelp in profiling peptide and chemical compounds in theirbinding and delivery such as ligand efficiency indices [5ndash11] fraction of sp3 orbital hybridization carbons [12] aswell as the atomic binding interactions between peptide-ligand including explicit water [13] and dynamic effects [13]The guanidino group on an arginine residue is especiallyvaluable in binding nucleic acids given that it can performelectrostatic hydrogen bond cation-120587 and 120587-120587 interactionsArtificial neural networks and principal components analysishave been employed to study cell-penetrating peptides inan attempt to classify them according to their permeability[14] Boltzmannian stochastics have also been used to cal-culate populations of 3D structures of CPPs using PepLookcalculating both intra- and intermolecular interactions [15]Molecular dynamics simulations have also been carried outon penetratin and the TAT peptide with lipid bilayers [16ndash18] as well as of dimer peptides [19] or zinc-fingers [20]with DNA and the CPP CADY in complex with siRNA [21]Molecularmodeling can also discover ligands to nucleic acids[22] Some CPPs have been developed to improve their load

Hindawi Publishing CorporationBioMed Research InternationalVolume 2014 Article ID 257040 7 pageshttpdxdoiorg1011552014257040

2 BioMed Research International

delivery such as in the case of NF51 PF3 PF6 and TP10 [23ndash26] Recently determinedX-ray crystal structures of siRNA incomplex with peptides provide structural information abouttheir binding Docking coupled with molecular dynamicssimulations can provide clues on the structural energeticand dynamic effects of CPP to nucleic acid binding Bindingpartner atoms and functional groups their conformationalrearrangements and persistence over time are part of theseclues which in turn allow proposing suggestions for furthermodification of peptides for increased affinity andor speci-ficity to particular nucleotide sequences

2 Methods

21 Modeling The structure of the double-stranded 21 nucle-otide-luciferase siRNA (luc-siRNA) was generated usingMaestro version 92 [27] using as a template the structureof Tav2bsiRNA complex from the Protein Data Bank [28]structure file 2ZI0 The template has the closest crystal struc-ture to luc-siRNAwith 8matching base pairs aswell as helicalpeptides (Tomato aspermy virus 2b (TAV2b) proteins) boundto the siRNA This structure allows using a bound (holo)conformation of the siRNA in order to improve docking cal-culations to the receptor with respect to those using an apo orunbound structure Mismatching nucleotides were mutatedand the resulting structure (in complex with two alpha-helical peptides in the major groove) was energy minimizedusing MacroModel version 99 [29] prior to conducting a12 ns molecular dynamics simulation to relax and equilibratethe complex structure The relaxed and minimized siRNAstructure (sequence 51015840-ACGCCAAAAACAUAAAGAAAGand antisense 51015840-UUCUUUAUGUUUUUGGCGUCU) wasthen extracted from this complex for further use in dockingthe CPP peptides NF51 PF3 PF6 and TP10 [23ndash26]

22 Docking Peptide structureswere generatedwithMaestrov 92 [27] and energy-minimized The peptides were thendocked flexibly (flexible ligands rigid target) with GOLDv 501 [30] using ChemScore [31] and employing thefollowing conditions suited for flexible ligands autoscale =2 Population popsiz = auto select pressure = auto n islands= auto maxops = auto niche siz = auto Genetic operatorspt crosswt = auto allele mutatewt = auto migratewt = autoFlood fill radius = 40 A The auto option allows adjustingthe conformational sampling according to the number ofrotatable bonds in the ligands and this provides for theflexibility in the peptide ligands

A modified score of S(hbond ext) + (135lowastS(lipo)) [32]was further developed by incorporating the number of sp3carbons and used to rerank the peptides S(hbond ext)measures the intermolecular hydrogen bond contributionsto binding and S(lipo) is a lipophilic term that is calculatedbetween nonpolar carbon nonionic chlorine bromine andiodine and nonaccepting sulphur atoms [29]These terms aredependent on the distance between two atoms pairs and onhow much they differ from ideal values for interaction Sp3carbons were calculated with Marvin Beans [33]

23 Downregulation Experiments siRNA downregulationexperiments were performed as previously reported [25 26]Briefly in the case of TP10 PF3 and PF6 100 120583M CPP stocksolution was mixed with siRNA (10120583M stock solutionmdashthesame siRNA sequence as used in the modeling) in MQ waterin one-tenth of final treatment volume (ie 50120583L) usingmolar ratio (MR) MR30 in serum free media or MR40 inserum experiments Complexes were formed for 30min at RTand added toHEK cells grown to 60 confluence in a 24-wellplate in 450mL growth media After 4 h 1mL of fresh mediawas added to wells and cells were incubated for the indicatedtimes Further the cells were lysed using 100mL of 01TritonX-100 inHepesKrebbsRinger bufferAfter 30min lysison ice luciferase expressionwasmeasured using the PromegaLuciferase Kit on a 96-well Glomax luminometer (Promega)

In the case of NF51 EGFP-CHO cells were seeded in24-well plates 24 h prior to experiments siRNA was mixedwith CPP at different molar ratios (MR 5ndash30) in MQ waterComplexes were formed and cells were treated as describedabove After the indicated time media were removed andthe cells were rinsed with PBS detached from the plate andsuspended with PBS containing 5 FBS and FACS analysiswas performed

24 Molecular Dynamics The program Desmond version 31[34 35] was used to performmolecular dynamics (MD) sim-ulations Structures were first energy-minimized in Macro-Model version 99 [27] in an implicit water environment usingthe OPLS2005 force field and GBSA continuum solvationmodel with 2000 steps of Conjugate-Gradient and SteepestDescent or until a gradient threshold of 05 kcalmolA wasreached Further salt (Na+ and Clminus) ions were inserted toneutralize the system until reaching a 015M concentrationComplex structures were then solvated using TIP3P [36]water molecules in a periodic orthorhombic box of 8 Aadded to each direction extending from the solute to givetotal dimensions of 734 times 488 times 803 A Simulations wereconducted with a constant temperature of 300K using aNose-Hoover chain thermostat [37] and Martyna-Tobias-Klein barostat methods [38] A timestep of 1 fs was employedand simulations were run for 10 ns in total The first 1 nswere used as equilibration time The particle mesh Ewaldalgorithm was used for calculating long-range electrostaticinteractions [39] Trajectories were further analyzed withDesmond

In addition the peptide-siRNA complexes were modeledin explicit water and with an excess of 40 peptides to onesiRNA molecule

3 Results and Discussion

The CPPs studied have sequences that are shown in Table 1GOLD docked the peptides using a genetic algorithm

that explores different conformations and positions of thepeptide structures in the biomolecular target After dockingthe structures of the docked ligands appear to bewell attachedto the siRNA double-helix making extensive use of interac-tions between the peptide and the major groove of siRNA

BioMed Research International 3

Table 1 Amino acid sequences for cell-penetrating peptides

Name Sequence

TP10 AGYLLGKINLKALAALAKKIL-NH2

PF3 R1-AGYLLGKINLKALAALAKKIL-NH2

PF6 R1-AGYLLGK(120576NH-K(K(R2) 2) 2)-INLKALAALAKKIL-NH2

NF51 R1-AGYLLG-R3-INLKALAALAKKIL-NH2

where

R1

CH3

O8

R2

N

CF3

NH

N

CH3

NH

O

O

R3

O

NH2

2

Table 2 Docking score components modified scores and ranks of peptides bound to siRNA

Rank CPP 119878(hbond ext) 119878(lipo) Modified score from GOLD sp3 carbons SusiScore1 PF6 333 13338 18339 144 295922 NF51 292 15006 20550 93 170553 PF3 359 12884 17752 93 165094 TP10 488 7297 10339 77 7961

Hydrogen bonding and electrostatic as well as lipophilicinteractions are dominant as shown by their binding posesand scores The binding conformations are such that thepeptides prefer them to retain their secondary structure theyhad prior to binding that is a nine residue alpha-helix Thebinding modes and ligand poses for the best complexes areshown in Figure 1 The best complexes were those with thedeepest binding energy and with the most realistic bindingmode and plausible intermolecular interactions and werechosen from at least four different independent runs

The ranking of the peptides docked to the siRNA isshown in Table 2 The external hydrogen bonding term andthe lipophilic term components of the ChemScore scoringfunction as well as the total number of sp3 carbons in themolecule were used and recomposed to create a modifiedscore called SusiScore This modified score was composedas SusiScore = Number of sp3 carbonslowast(S(hbond ext) +135lowast(S(lipo))) SusiScore makes use of the most importantcomponents of the ChemScore scoring function that havebeen correlated to binding affinity [31] as well as the sp3carbons in the molecule which have been correlated to

Table 3 siRNA downregulation in HEK cells using luc-siRNA (51015840-ACGCCAAAAACAUAAAGAAAG and antisense 51015840-UUCUUUAUGUUUUUGGCGUCU)

CPP Serum containing mediaPF6 (MR40) 80NF51 (MR10) 70PF3 0TP10 0

the complexity and solubility of a compound [12] in orderto create a new scoring and ranking procedure

The calculation results are comparable to the experi-mentally determined efficiency of the different peptides ofdelivery of siRNA into cells which are shown in Table 3[25 26]There is the same trend between the downregulationexperiments and the binding scores

The reason for the better binding with siRNA for NF51PF6 and PF3 as compared to TP10 would appear to bethe stearyl tail present in the former three but lacking in


(a)

(b)

(c)

(d)

Figure 1 Docked binding modes for the complexes of siRNA andcell-penetrating peptides (a) NF51 (b) PF3 (c) PF6 and (d) TP10

TP10 This lipophilic group acts as an anchor which fixesthe peptide to the siRNA providing a means for the peptidestructure to stretch and embrace the majority of the siRNAhelix In addition the quinoline groups of PF6 interact withthe nucleic acid basepairs giving the complex PF6sdotsiRNAadded stability

An apparent explanation for the ability of the differentpeptides to deliver the siRNA into the interior of the cells maybe the strength of their binding complexes If as speculatedthe peptides can form pores in the cell membrane a strongerbinding to the siRNA would allow the peptide to drag thesiRNA with it as it transverses the membrane Peptides witha lower binding affinity would probably be able only to formpores and the lower delivery of siRNAwould be due to it onlyentering the cell through diffusion through the pores in themembrane created by the peptide Stronger binding peptidesmay prolong the interaction with the siRNA and thereforeaccompany it through the pores in the cell membrane it hascreated

The lipophilic stearyl tail that provided stronger bindingfor NF51 PF3 and PF6 with respect to TP10 may alsoplay a role in the insertion into the lipid bilayer of the cellmembrane providing a better interaction with the lipophilic

groups inside the lipid bilayer resulting in likely easierformation of pores in the cell membrane as well as bindinginteractions with siRNA

A mechanism for CPP delivery includes the associationin clusters with glycosaminoglycans on the cell membranesurface [40] It seems likely that the CPP complexes withnucleic acids are then internalized into the cell throughendocytosis [25] and even suggested being pinocytosis [18]Given the stronger binding and stronger downregulationprovided by NF51 and PF6 as compared to PF3 and TP10as shown in the present study and in biological experimentsit also adds support to the thesis that the binding betweenCPPs and nucleic acids (at least for those peptides consideredhere) is not released until inside the cell Perhaps their releaseoccurs by localization in the lysozyme [41] given that thelower pH there allows for acid or ionic competition for thebinding of the nucleic acid and disrupts the binding complexCPPsdotnucleic acid

MD simulations of single CPPs bound to siRNA showedthat the bound structures of peptides and siRNA obtainedfrom docking were stable The simulation quality analysisfrom Desmond showed stable and regular values for totalenergy potential energy pressure and temperature (standarddeviation lower than 2K for all cases)

The complexes did not fall apart and remained stronglybound throughout the 10 ns simulations The stearyl tail ofPF6 bound to siRNA folded more closely to the siRNA as thesimulation progressedThe secondary structure of the siRNAwas stable and maintained its double-helix throughout all ofthe simulations The RMSDs of the peptide backbone atomswere 1 A for NF51 15 A for PF6 15 A for PF3 and 2 A forTP10The RMSDs of the siRNA backbone atoms were 2 A forNF51 15 A for PF6 25 A for PF3 and 45 A for TP10 TheRMSDs of the siRNA and peptide backbone atoms togetherwere 2 A for NF51 4 A for PF6 4 A for PF3 and 45 A forTP10 Figure 2 shows a close-up view of the structure of thecomplex between the peptide PF6 and siRNA

The MD simulation of TP10 bound to siRNA showeda higher RMSD for peptide and siRNA compared to theother CPPs probably related to the weaker binding affinityfor the complex The RMSDs of the peptide sidechains andmodification groups were small and ranged from 15 A forthe K

3QN4group in PF6 05 A for the stearyl group in

PF6 075 A for the stearyl group in PF3 and 12 A for thestearyl group inNF51These values indicate the stability of thecomplex for siRNA bound to NF51 and PF6 and to a smallerextent PF3 and even less stable was the complex with TP10

During the simulation of the PF6 complex the stearylatedtail of the peptide approached siRNA from an extended intoa folded conformation forming a lid over the other residuesand bases by generating two stable intermolecular hydrogenbonds the peptide glycine (the ldquofirstrdquo glycine ie that closestto the stearyl tail) backbone NH donating a hydrogen bondto the siRNA guanine (G) N7 (distance shortened to 29 A)and between the peptide leucine backbone NH donating ahydrogen bond to the phosphoryl (nonester) oxygen in thesiRNA backbone (distance varied from over 5 to 26 A) Theformer hydrogen bond is specific to certain siRNA basesspecifically guanine and adenine and by modifying this


Phosphoryl O

Leu NHGly NH

GN7

Figure 2 Molecular dynamics snapshot structure of the CPPsdotsiRNA complex for PF6 Peptide is shown in ball and stick siRNA in wireframeand green and yellow ribbons Hydrogen bonds between the tail of PF6 and siRNA are shown in green dashed lines from left Leu NH rarrPhosphoryl O and Gly NH rarr Guanine N7 For clarity explicit water molecules are not shown

specific amino acid in the peptide (such as modifying thepeptide backbone to a group less susceptible to hydrolysis orwithout a hydrogen bond donor) or modifying the guanineor adenine nucleotides in this position in the siRNA bindingcan be enhanced or suppressed leading to specific bindingaffinity for designed peptides and siRNA sequences Thelatter hydrogen bond on the other hand given that it occursbetween the peptide and the backbone of the siRNA is lessspecific and would be expected to be conserved in differ-ent siRNA sequences These hydrogen bonds and contactsbetween peptide and nucleic acid give suggestions for furthermodification of the peptides and nucleotide sequences fortuning of specificity and binding affinity

The simulation of 40CPPs and siRNA in a ball showedthat the structures were stable over 1 ns The mechanism ofbinding allows one CPP peptide to interact strongly in themain groove while the subsequent peptides have less specificand less strong interactions with the charged backbonearranged in subsequent binding shells similar to solvationshells The structure of this ball for PF6 in complex withsiRNA can be seen in Figure 3 The ball simulations showedthe structures to be stable where the constructs of PF6showed a longer stability thanTP10 whose structure fell apartbefore 1 ns

Given the docking rescores and the molecular dynamicstrajectories which show stability for the peptides and therankings corresponding to experiment the importance ofappropriate sidechains and modifications for the CPPs isdemonstrated in the improved binding affinity for siRNA for

Figure 3 40 peptide units of PF6 surrounding siRNA

PF6 and NF51 as well as greater stability for the complex inMD simulations and perhaps an improved ability to cross thecell membrane In addition the sidechains andmodificationscan be further optimized to improve delivery downregu-lation of activity and safety and for the quinoline groupssuch as those in PF6 to possibly improve selectivity forparticular nucleic acid sequences The procedure developedin the present work may be used to design modifications tothe CPP core and sidechains or extending groups and followor predict their interactions with siRNA


4 Conclusions

The newly proposed reranking developed in the presentstudy SusiScore together with flexible ligand docking andexplicit water molecular dynamics simulations appear to beable to describe the interactions of cell-penetrating peptideswith siRNA The same guidelines that rule the binding ofsmall molecules to receptors are present in peptide bindingIn addition the present study shows that the number of sp3carbons improve the ranking of calculated results Thereforethe use of the number of sp3 carbon atoms that has beenshown to be important in small molecule binding wouldalso be appropriate for peptide binding The stearyl grouppresent in PF3 PF6 and NF51 and the quinoline grouppresent in PF6 increase the binding affinity and stability oftheir complexes with siRNA compared to TP10 which lacksthese groups Further modifications to the core sidechainsand extending groups of CPPs and other cell-penetratingcompounds may be designed and their interactions modeledthrough the procedure described here The impact of thedifferent cores sidechains and extending groups will be seenin binding affinity to siRNA as well as solubility stability andefficiency in delivering and downregulating nucleic acids

Abbreviations

CPP Cell-penetrating peptidesiRNA Small interfering ribonucleic acidsMQ water Milli-Q waterHEK cells Human embryonic kidney cellsRT Room temperatureEGFP-CHO cells Enhanced green fluorescent

protein-Chinese hamster ovary cellsPBS Phosphate-buffered salineFBS Fetal bovine serumFACS Fluorescence-activated cell sortingMD Molecular dynamicsMR Molar ratio

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

Thanks are due to the Estonian Ministry for Education andResearch (Grant SF0140031Bs09) for funding The fundingagency had no role in the preparation of the paper nor inthe study design none in the collection analysis and inter-pretation of data none in the writing of the report and nonein the decision to submit the paper for publication

References

[1] F Heitz M C Morris and G Divita ldquoTwenty years of cell-penetrating peptides frommolecular mechanisms to therapeu-ticsrdquoBritish Journal of Pharmacology vol 157 no 2 pp 195ndash2062009

[2] H M Aliabadi B Landry C Sun T Tang and H UludagldquoSupramolecular assemblies in functional siRNA deliverywhere do we standrdquo Biomaterials vol 33 no 8 pp 2546ndash25692012

[3] R W Carthew and E J Sontheimer ldquoOrigins and mechanismsof miRNAs and siRNAsrdquo Cell vol 136 no 4 pp 642ndash655 2009

[4] A C Cheng V Calabro and A D Frankel ldquoDesign of RNA-binding proteins and ligandsrdquo Current Opinion in StructuralBiology vol 11 no 4 pp 478ndash484 2001

[5] I D Kuntz K Chen K A Sharp and P A Kollman ldquoThemax-imal affinity of ligandsrdquo Proceedings of the National Academy ofSciences of the United States of America vol 96 pp 9997ndash100021999

[6] A L Hopkins C R Groom and A Alex ldquoLigand efficiency auseful metric for lead selectionrdquo Drug Discovery Today vol 9no 10 pp 430ndash431 2004

[7] A T Garcıa-Sosa S Sild and U Maran ldquoDocking and virtualscreening using distributed grid technologyrdquo QSAR and Com-binatorial Science vol 28 no 8 pp 815ndash821 2009

[8] A T Garcıa-Sosa C Hetenyi and U Maran ldquoDrug efficiencyindices for improvement of molecular docking scoring func-tionsrdquo Journal of Computational Chemistry vol 31 pp 174ndash1842010

[9] A T Garcıa-Sosa S Sild K Takkis and U Maran ldquoCombinedapproach using ligand efficiency cross-docking and antitargethits for wild-type and drug-resistant Y181C HIV-1 reversetranscriptaserdquo Journal of Chemical Information and Modelingvol 51 no 10 pp 2595ndash2611 2011

[10] A T Garcıa-Sosa M Oja C Hetenyi and U MaranldquoDrugLogit logistic discrimination between drugs and non-drugs including disease-specificity by assigning probabilitiesbased on molecular propertiesrdquo Journal of Chemical Informa-tion and Modeling vol 52 no 8 pp 2165ndash2180 2012

[11] A T Garcıa-Sosa ldquoHydration properties of ligands and drugs inprotein binding sites tightly-bound bridging water moleculesand their effects and consequences on molecular design strate-giesrdquo Journal of Chemical Information andModeling vol 53 no6 pp 1388ndash1405 2013

[12] F Lovering J Bikker and C Humblet ldquoEscape from flatlandincreasing saturation as an approach to improving clinicalsuccessrdquo Journal of Medicinal Chemistry vol 52 no 21 pp6752ndash6756 2009

[13] A T Garcıa-Sosa and R L Mancera ldquoFree energy calculationsof mutations involving a tightly bound water molecule andligand substitutions in a ligand-protein complexrdquo MolecularInformatics vol 29 no 8-9 pp 589ndash600 2010

[14] D A Dobchev I Mager I Tulp et al ldquoPrediction of cell-penetrating peptides using artificial neural networksrdquo CurrentComputer-Aided Drug Design vol 6 no 2 pp 79ndash89 2010

[15] A Thomas L Lins G Divita and R Brasseur ldquoRealisticmodeling approaches of structure-function properties of CPPsin non-covalent complexesrdquo Biochimica et Biophysica ActamdashBiomembranes vol 1798 no 12 pp 2217ndash2222 2010

[16] M F Lensink B Christiaens J Vandekerckhove A Prochi-antz and M Rosseneu ldquoPenetratin-membrane associationW48R52W56 shield the peptide from the aqueous phaserdquoBiophysical Journal vol 88 no 2 pp 939ndash952 2005

[17] H D Herce and A E Garcia ldquoMolecular dynamics simulationssuggest a mechanism for translocation of the HIV-1 TATpeptide across lipid membranesrdquo Proceedings of the NationalAcademy of Sciences of the United States of America vol 104 no52 pp 20805ndash20810 2007


[18] S Yesylevskyy S Marrink and A E Mark ldquoAlternative mech-anisms for the interaction of the cell-penetrating peptidespenetratin and the TAT peptide with lipid bilayersrdquo BiophysicalJournal vol 97 no 1 pp 40ndash49 2009

[19] G Papadopoulos A I Grigoroudis and D A KyriakidisldquoDimerization of the AtoC response regulator and modellingof its binding to DNArdquo Journal of Molecular Graphics andModelling vol 29 no 4 pp 565ndash572 2010

[20] M Bitar M G DrummondM G S Costa et al ldquoModeling thezing finger protein SmZF1 from Schistosoma mansoni insightsinto DNA binding and gene regulationrdquo Journal of MolecularGraphics and Modelling vol 39 pp 29ndash38 2013

[21] J-M Crowet L Lins S Deshayes et al ldquoModeling of non-covalent complexes of the cell-penetrating peptide CADY andits siRNA cargordquo Biochimica et Biophysica Acta Biomembranesvol 1828 no 2 pp 499ndash509 2013

[22] D-LMaD S Chan P LeeMH Kwan andC Leung ldquoMolec-ular modeling of drug-DNA interactions virtual screening tostructure-based designrdquo Biochimie vol 93 no 8 pp 1252ndash12662011

[23] K Kilk S El-Andaloussi P Jarver et al ldquoEvaluation of trans-portan 10 in PEI mediated plasmid delivery assayrdquo Journal ofControlled Release vol 103 no 2 pp 511ndash523 2005

[24] M Mae S EL Andaloussi P Lundin et al ldquoA stearylated CPPfor delivery of splice correcting oligonucleotides using a non-covalent co-incubation strategyrdquo Journal of Controlled Releasevol 134 no 3 pp 221ndash227 2009

[25] S El Andaloussi T Lehto I Mager et al ldquoDesign of a peptide-based vector PepFect6 for efficient delivery of siRNA in cellculture and systemically in vivordquoNucleic Acids Research vol 39no 9 pp 3972ndash3987 2011

[26] P Arukuusk L Parnaste N Oskolkov et al ldquoNew generationof efficient peptide-based vectors NickFects for the deliveryof nucleic acidsrdquo Biochimica et Biophysica Acta Biomembranesvol 1828 no 5 pp 1365ndash1373 2013

[27] Maestro version 92 Schrodinger LLC New York NY USA2011

[28] Protein Data Bank Research Collaboratory for Structural Bioin-formatics 2013 httpwwwpdborgpdbhomehomedo

[29] MacroModel Version 99 Schrodinger LLC New York NYUSA 2011

[30] G Jones P Willett and R C Glen ldquoMolecular recognition ofreceptor sites using a genetic algorithm with a description ofdesolvationrdquo Journal ofMolecular Biology vol 245 no 1 pp 43ndash53 1995

[31] M D Eldridge C W Murray T R Auton G V Paolini and RP Mee ldquoEmpirical scoring functions I The development of afast empirical scoring function to estimate the binding affinityof ligands in receptor complexesrdquo Journal of Computer-AidedMolecular Design vol 11 no 5 pp 425ndash445 1997

[32] Cambridge Crystallographic Data Centre Support amp Resour-ces httpwwwccdccamacukproductslife sciencesgoldfaqsscientific faqphp

[33] Marvin version 5601 ChemAxon Budapest Hungary 2011httpwwwchemaxoncom

[34] Desmond Molecular Dynamics System version 31 D E ShawResearch New York NY USA 2012

[35] Maestro-Desmond Interoperability Tools Version 31 Schro-dinger New York NY USA 2012

[36] W L Jorgensen and J D Madura ldquoSolvation and conformationofmethanol in waterrdquo Journal of the American Chemical Societyvol 105 no 6 pp 1407ndash1413 1983

[37] G J Martyna M L Klein and M Tuckerman ldquoNose-Hooverchains the canonical ensemble via continuous dynamicsrdquo TheJournal of Chemical Physics vol 97 no 4 pp 2635ndash2643 1992

[38] G J Martyna D J Tobias and M L Klein ldquoConstant pres-sure molecular dynamics algorithmsrdquo The Journal of ChemicalPhysics vol 101 no 5 pp 4177ndash4189 1994

[39] T Darden L Perera L Li and L Pedersen ldquoNew tricksfor modelers from the crystallography toolkit the particlemesh Ewald algorithm and its use in nucleic acid simulationsrdquoStructure vol 7 no 3 pp R55ndashR60 1999

[40] A Ziegler and J Seelig ldquoBinding and clustering of glycosamino-glycans a common property of mono- and multivalent cell-penetrating compoundsrdquo Biophysical Journal vol 94 no 6 pp2142ndash2149 2008

[41] A Erazo-Oliveras N Muthukrishnan R Baker T Wang andJ Pellois ldquoImproving the endosomal escape of cell-penetratingpeptides and their cargos strategies and challengesrdquo Pharma-ceuticals vol 5 no 11 pp 1177ndash1209 2012

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Anatomy Research International

PeptidesInternational Journal of


Hindawi Publishing Corporation httpwwwhindawicom

International Journal of

Volume 2014

Zoology


Molecular Biology International

GenomicsInternational Journal of


The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014


BioinformaticsAdvances in

Marine BiologyJournal of



Signal TransductionJournal of


BioMed Research International

Evolutionary BiologyInternational Journal of



Biochemistry Research International

ArchaeaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014


Genetics Research International


Advances in

Virolog y

Hindawi Publishing Corporationhttpwwwhindawicom

Nucleic AcidsJournal of

Volume 2014

Stem CellsInternational



Enzyme Research



Microbiology


delivery such as in the case of NF51 PF3 PF6 and TP10 [23ndash26] Recently determinedX-ray crystal structures of siRNA incomplex with peptides provide structural information abouttheir binding Docking coupled with molecular dynamicssimulations can provide clues on the structural energeticand dynamic effects of CPP to nucleic acid binding Bindingpartner atoms and functional groups their conformationalrearrangements and persistence over time are part of theseclues which in turn allow proposing suggestions for furthermodification of peptides for increased affinity andor speci-ficity to particular nucleotide sequences

2 Methods

21 Modeling The structure of the double-stranded 21 nucle-otide-luciferase siRNA (luc-siRNA) was generated usingMaestro version 92 [27] using as a template the structureof Tav2bsiRNA complex from the Protein Data Bank [28]structure file 2ZI0 The template has the closest crystal struc-ture to luc-siRNAwith 8matching base pairs aswell as helicalpeptides (Tomato aspermy virus 2b (TAV2b) proteins) boundto the siRNA This structure allows using a bound (holo)conformation of the siRNA in order to improve docking cal-culations to the receptor with respect to those using an apo orunbound structure Mismatching nucleotides were mutatedand the resulting structure (in complex with two alpha-helical peptides in the major groove) was energy minimizedusing MacroModel version 99 [29] prior to conducting a12 ns molecular dynamics simulation to relax and equilibratethe complex structure The relaxed and minimized siRNAstructure (sequence 51015840-ACGCCAAAAACAUAAAGAAAGand antisense 51015840-UUCUUUAUGUUUUUGGCGUCU) wasthen extracted from this complex for further use in dockingthe CPP peptides NF51 PF3 PF6 and TP10 [23ndash26]

22 Docking Peptide structureswere generatedwithMaestrov 92 [27] and energy-minimized The peptides were thendocked flexibly (flexible ligands rigid target) with GOLDv 501 [30] using ChemScore [31] and employing thefollowing conditions suited for flexible ligands autoscale =2 Population popsiz = auto select pressure = auto n islands= auto maxops = auto niche siz = auto Genetic operatorspt crosswt = auto allele mutatewt = auto migratewt = autoFlood fill radius = 40 A The auto option allows adjustingthe conformational sampling according to the number ofrotatable bonds in the ligands and this provides for theflexibility in the peptide ligands

A modified score of S(hbond ext) + (135lowastS(lipo)) [32]was further developed by incorporating the number of sp3carbons and used to rerank the peptides S(hbond ext)measures the intermolecular hydrogen bond contributionsto binding and S(lipo) is a lipophilic term that is calculatedbetween nonpolar carbon nonionic chlorine bromine andiodine and nonaccepting sulphur atoms [29]These terms aredependent on the distance between two atoms pairs and onhow much they differ from ideal values for interaction Sp3carbons were calculated with Marvin Beans [33]

23 Downregulation Experiments siRNA downregulationexperiments were performed as previously reported [25 26]Briefly in the case of TP10 PF3 and PF6 100 120583M CPP stocksolution was mixed with siRNA (10120583M stock solutionmdashthesame siRNA sequence as used in the modeling) in MQ waterin one-tenth of final treatment volume (ie 50120583L) usingmolar ratio (MR) MR30 in serum free media or MR40 inserum experiments Complexes were formed for 30min at RTand added toHEK cells grown to 60 confluence in a 24-wellplate in 450mL growth media After 4 h 1mL of fresh mediawas added to wells and cells were incubated for the indicatedtimes Further the cells were lysed using 100mL of 01TritonX-100 inHepesKrebbsRinger bufferAfter 30min lysison ice luciferase expressionwasmeasured using the PromegaLuciferase Kit on a 96-well Glomax luminometer (Promega)

In the case of NF51 EGFP-CHO cells were seeded in24-well plates 24 h prior to experiments siRNA was mixedwith CPP at different molar ratios (MR 5ndash30) in MQ waterComplexes were formed and cells were treated as describedabove After the indicated time media were removed andthe cells were rinsed with PBS detached from the plate andsuspended with PBS containing 5 FBS and FACS analysiswas performed

24 Molecular Dynamics The program Desmond version 31[34 35] was used to performmolecular dynamics (MD) sim-ulations Structures were first energy-minimized in Macro-Model version 99 [27] in an implicit water environment usingthe OPLS2005 force field and GBSA continuum solvationmodel with 2000 steps of Conjugate-Gradient and SteepestDescent or until a gradient threshold of 05 kcalmolA wasreached Further salt (Na+ and Clminus) ions were inserted toneutralize the system until reaching a 015M concentrationComplex structures were then solvated using TIP3P [36]water molecules in a periodic orthorhombic box of 8 Aadded to each direction extending from the solute to givetotal dimensions of 734 times 488 times 803 A Simulations wereconducted with a constant temperature of 300K using aNose-Hoover chain thermostat [37] and Martyna-Tobias-Klein barostat methods [38] A timestep of 1 fs was employedand simulations were run for 10 ns in total The first 1 nswere used as equilibration time The particle mesh Ewaldalgorithm was used for calculating long-range electrostaticinteractions [39] Trajectories were further analyzed withDesmond

In addition the peptide-siRNA complexes were modeledin explicit water and with an excess of 40 peptides to onesiRNA molecule

3 Results and Discussion

The CPPs studied have sequences that are shown in Table 1GOLD docked the peptides using a genetic algorithm

that explores different conformations and positions of thepeptide structures in the biomolecular target After dockingthe structures of the docked ligands appear to bewell attachedto the siRNA double-helix making extensive use of interac-tions between the peptide and the major groove of siRNA



Name Sequence





where

R1

CH3

O8

R2

N

CF3

NH

N

CH3

NH

O

O

R3

O

NH2

2











(a)

(b)

(c)

(d)














Phosphoryl O

Leu NHGly NH

GN7








4 Conclusions


Abbreviations





Acknowledgments


References


















































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology



Name Sequence





where

R1

CH3

O8

R2

N

CF3

NH

N

CH3

NH

O

O

R3

O

NH2

2











(a)

(b)

(c)

(d)














Phosphoryl O

Leu NHGly NH

GN7








4 Conclusions


Abbreviations





Acknowledgments


References


















































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology


(a)

(b)

(c)

(d)














Phosphoryl O

Leu NHGly NH

GN7








4 Conclusions


Abbreviations





Acknowledgments


References


















































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology


Phosphoryl O

Leu NHGly NH

GN7








4 Conclusions


Abbreviations





Acknowledgments


References


















































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology


4 Conclusions


Abbreviations





Acknowledgments


References


















































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology

































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology

research article peptide-ligand binding modeling of sirna...

Documents