Molecular modeling and epitopes mapping of human adenovirus type 3 hexon protein

Download Molecular modeling and epitopes mapping of human adenovirus type 3 hexon protein

Post on 30-Oct-2016




0 download

Embed Size (px)


<ul><li><p>Vaccine 27 (2009) 51035110</p><p>Contents lists available at ScienceDirect</p><p>Vaccine</p><p>journa l homepage: www.e lsev ier .com</p><p>Molecular modeling and epitopes mapping of hutype 3 hexon protein</p><p>Xiaohui Y nga</p><p>Hong Ga huYanhonga Department o d 157,b Key Laborator 15004c Microbiology</p><p>a r t i c l</p><p>Article history:Received 21 DReceived in reAccepted 10 JuAvailable onlin</p><p>Keywords:HAdVHexon proteinEpitope Mapping</p><p>irusle apHAdVdynam</p><p>called reverse ET (RET) was used to predict the type-specic B-cell neutralizing epitopes. An epitope-screening algorithmbasedon analyzing the solvent accessibility surface (SAS) area from the3Dmodel andcalculationof siteshomologyusingRETwasdesignedand implemented in theBioPerl script language. Fiveepitope polypeptide segments were predicted and mapped onto the 3D model. Finally ve polypeptideswere synthesized and the predicted epitopes were identied by enzyme-linked immunosorbent assay</p><p>1. Introduc</p><p>Adenoviviruses withvertex capsosied intogenomic prby neutralizof human intract infectiare particulmajor epideyet no effecto treat the</p><p> Supporte30771909), Th20070226007)(No. ZJY0701),Department (N</p><p> CorresponE-mail add</p><p>0264-410X/$ doi:10.1016/j.v(ELISA) and Neutralization Test (NT). It was found that the type-specic neutralizing epitopes of HAdV3are located at the top surface of hexon tower regions (residue numbers: 135146, 169178, 237251,262272, 420434). This work is of great signicance to the molecular design of a multivalent HAdVsvaccine.</p><p> 2009 Elsevier Ltd. All rights reserved.</p><p>tion</p><p>ridae viruses are nonenveloped, double-stranded DNAan icosahedral capsid comprising 240 hexons and 12meres [1,2]. Thehumanadenovirus (HAdV) canbe clas-</p><p>6 species (AF) on the basis of hemagglutination andoperties [3], consisting of 51 serotypes dened mainlyation criteria [4,5]. HAdVs can cause a broad spectrumfective diseases [69], among which upper respiratoryon and pedo-pneumonia caused by serotypes 3 and 7arly serious [5,10,11]. Especially in northern China, themic strains are HAdV3 and HAdV7 [12,13]. There is astive curative antiviral medicine or vaccine with whichse diseases.</p><p>d by The National Natural Science Foundation of China (No.e Doctoral Co-nancing Project of Chinese Ministry of Education (No., The Natural Science Foundation Key Project of Heilongjiang ChinaScience and Technology Project of Heilongjiang Provincial Educationo. 11521172).ding author. Tel.: +86 451 87502965; fax: +86 451 86667248.resses:, (Z. Qu).</p><p>Themajor coat protein of HAdV is hexon (i.e., a homotrimer pro-tein comprising threemonomers A, B and C) [14]. It has been shownthat antibodies stimulated by a hexon can neutralize an HAdV-mediated viral infection, and that this neutralization reaction istype-specic [15]. The tower region of this hexon homotrimer con-tains a large number of type-specic neutralizing epitopes (B-cellepitopes). Identifying these type-specic neutralizing epitopes isof great signicance in several areas of HAdV research, includingthe molecular design of a HAdV vaccine [1618], the developmentof a rapid HAdV diagnostic agent and preparing an antiadenovirusmedicine [19,20]. But very little is currently known about themap-ping of type-specic B-cell neutralizing epitopes of hexons inmanyserotypes of HAdV. Although it is possible to locate hypervari-able regions (HVRs) [14,21] using the multiple sequence alignment(MSA) method [22], it is difcult to identify the type-specic neu-tralizing epitopes and to obtain the specic three-dimensional (3D)conformation of the epitope peptides for a specic serotype. The 3Dconformation of hexon protein can provide relevant informationabout epitopes. The hexon structures of HAdV type 2 (HAdV2) andHAdV type 5 (HAdV5) [23,24] are available in the Protein Data Bank(PDB) [25], but these structures have inherent amino acid deletionsand disruption of the peptide chains. In addition, related docu-ments showthat these structuresareonly theconservedcore region</p><p>see front matter 2009 Elsevier Ltd. All rights reserved.accine.2009.06.041uana, Zhangyi Qua,, Xiaomin Wub, Yingchen Waoa, Lei Shanga, Hongyan Zhanga, Hongbo Cuic, YueTanga, Le Qina</p><p>f Hygienic Microbiology, Public Health College, Harbin Medical University, Baojian Roay of Forest Plant Ecology, Ministry of Education, Northeast Forestry University, HarbinLaboratory, Harbin Medical University, Harbin 150081, PR China</p><p>e i n f o</p><p>ecember 2008vised form 26 April 2009ne 2009e 30 June 2009</p><p>a b s t r a c t</p><p>The hexon protein of human adenovWe developed a new effective, reliabthree-dimensional (3D) model of themolecular mechanics and molecular/ locate /vacc ine</p><p>man adenovirus</p><p>, Lei Liua, Fengxiang Weia,i Zhaoc, Na Wua,</p><p>Harbin Heilongjiang 150081, PR China0, PR China</p><p>(HAdV) processes type-specic B-cell neutralizing epitopes.proach to map these epitopes on hexon protein of HAdVs. A3 hexon was obtained by homology modeling and rened byics simulations. A modied evolutionary trace (ET) analysis</p></li><li><p>5104 X. Yuan et al. / Vaccine 27 (2009) 51035110</p><p>of the HAdV hexon protein [15], and do not provide informationof the complete tower structure, which contains the type-specicneutralizing epitopes.</p><p>This study investigated two characteristics of epitopes onHAdVshexons nepitopes nology [26analysis. Thwas determ[26,29,30],the modelthat we namuse of MSAepitope-scrSAS analysicandidate e3D modelwere synthto prove theimmunosortope peptidwas used tosera.</p><p>2. Materia</p><p>2.1. HAdV3</p><p>The HAdfrom clinicchildren intotal of 384ment of PeUniversity.fully isolatemorphologidentied thbeen deposEU078562)</p><p>Anti-HAZealand rabHAdV3 partparticles emwas collectpared for acontrol. Theby ELISA as</p><p>2.2. Homol</p><p>Homololar dynamicsimulationDiego, USA)employed facid sequensequence d</p><p>2.2.1. MolecThe Insi</p><p>structure oPDB (http:/ate templatThe chimpa2obe) at 2.1ogy at 85.6%</p><p>HAdV5 hexon (PDB code: 1p30) [24], and was thus considered tobe the most appropriate template. MSA based on the Needlemanand Wunsch Algorithm [37] was performed with 2obe, 1p30, 1p2z,and HEX3 to conform to the structurally conserved region (SCR).</p><p>II/Mmodenein thd intde cst-de(CG).01kcng aerfoce wthe e</p><p>edtow</p><p>izatice walityprocrime</p><p>strs 3Dasce wnvessch</p><p>Solveana</p><p>e iswasradiuAS o</p><p>e groreateresidere u</p><p>verse</p><p>nalyologgatetainHex</p><p>hodd frowhicing Mitop</p><p>ultiA is ces oseroies Dnder007008615612610amely B-cell neutralizing epitopes and type-specicusing a combination of molecular simulation tech-] and bioinformatics evolutionary trace (ET) [27,28]e 3D structure of the HAdV type 3 (HAdV3) hexonined using molecular simulation/homology modelingand the solvent-accessible surface area (SAS) [31] ofwas calculated. In addition, a modied ET methoded reverse ET (RET) was employed. This involved the</p><p>, sites homology calculation, and a purpose-designedeening algorithm that combines the results from thes and sites homology calculation. The presence of vepitope segments was predicted and mapped onto theof the hexon. Finally, the predicted epitope peptidesesized. Two serological experiments were performedcorrectness of epitopes prediction: (1) enzyme-linkedbent assay (ELISA)was used to detect the afnity of epi-es and anti-HAdV3 serum; (2) Neutralization Test (NT)test the neutralizing effect to HAdV3 of antipeptides</p><p>ls and methods</p><p>and anti-HAdV3 serum</p><p>V used in this study was an isolated strain obtainedal throat-swab specimens. In 2003 and 2004, manythe Harbin area of China contracted fever [13], and athroat swabs were taken from them in the Depart-</p><p>diatrics, No. 1 Subsidiary Hospital of Harbin MedicalA strain of HAdV (namely Harbin04B) was success-d in our laboratory, and cell culture, immunology, andical, PCR, and sequencing analyses of the hexon genee virus asHAdV3, the nucleotide sequence ofwhich hasited in the NCBI GenBank database (accession number:.dV3 serumwasobtained froma6-week-old femaleNewbit that had been injected intramuscularly with 1013</p><p>icles and then boosted subcutaneously with 1013 viralulsied in complete Freunds adjuvant (Sigma). Blood</p><p>ed from the ear fringe vein plexus, and serum was pre-n ELISA, with preimmune serum used as a negativeantibody titer and theHAdV3 specicitywere detecteddescribed previously [32].</p><p>ogy modeling</p><p>gy modeling, energy minimization (EM) and molecu-s (MD) simulations were performed using a molecularsoftware package InsightII 2005 (Accelrys Inc., San. The consistent-valence force eld (CVFF) [3335] wasor EM and MD simulations. The HAdV3 hexon aminoce was deduced from the corresponding nucleotide</p><p>erived from Harbin04B and was named by HEX3.</p><p>ular modeling and structure renementghtII/Homology module was applied to build the 3Df the HAdV3 hexon. The web-FASTA tool [36] of the/ was used to search for an appropri-e for the homology modeling using HEX3 as a probe.nzee adenovirus 68 (AdC68) [15] hexon (PDB code: resolution exhibited the highest degree of homol-higher than HAdV2 hexon (PDB code: 1p2z) [23], and</p><p>Insightinitial</p><p>To rcuteddivideand sisteepedientthan 0by xithen pvergensolveperformfor theminimvergenthe quabovehomot</p><p>TheProlegram wsequengram ion Kab</p><p>2.2.2.SAS</p><p>residumodelprobein the Sresiduwere gwhosedata w</p><p>2.3. Re</p><p>ET aof hominvestifor cerHAdV3ETmetderiveHEX3,includdate ep</p><p>2.3.1. MMS</p><p>serotyping allof specBank uAC 000AC 000DQ149DQ149DQ149odeler programwas used to automatically construct theel of HEX3.the structure, the EM and MD simulations were exe-</p><p>e InsightII/Discover 3 module. The entire process waso the following steps: rst, all the hydrogen atomshains were optimized in a vacuum by a 500-stepscent (SD) minimization followed by conjugate gra-minimization until the nal convergence was loweralmol1 1. Second, the loop regions were optimizedll atoms except for those in the tower region, andrming 500 steps of SD and CG until the nal con-as lower than 0.01kcalmol1 1. Because EM cannotnergy-barriers problem [38], a MD simulation was(1000ps at 310K) to achieve the stable conformationer region (residues: 115310 and 400510). Third, CGon of the full protein was performed until the nal con-as lower than 0.01kcalmol1 1. This step improvedof the initial model of the HEX3 homotrimer. The</p><p>edure produced the 3D model of the HAdV3 hexonr.ucture was further checked using the InsightII/and InsightII/ProStat programs. The Proles 3D pro-used to examine the compatibility of an amino acidith a known 3D protein structure [39]. The ProStat pro-tigated the secondary structural of the 3D model basedSander method [40].</p><p>nt accessibility surface analysis (SAS)lysis is commonly used to evaluate how deep a givenburied [31]. The SAS of the entire HEX3 homotrimercalculated with the InsightII/Access Surf program. As of 1.4 was used for all calculations. The differencef each residue in the HEX3 model was determined. Twoupswere created: (1) an exposed group, whose residuesr than 25% of themaximumSAS; and (2) a buried group,ues were less than 10% of the maximum SAS. The SASsed in the subsequent epitopes screening.</p><p>evolutionary trace analysis (RET)</p><p>sis can extract the information obtained from the MSAous proteins onto a certain 3D molecule and therebywhich amino acid residues are likely to be crucialfunctions [27,28]. Because the neutralizing epitopes ofon are all type-specic, we utilized a special modiedtomap the candidate type-specicneutralizing epitopesm a epitope-screening algorithm onto the 3D model ofh we called reverse evolutionary trace (RET) analysisSA, sites homology calculation and a designed candi-es screening.</p><p>ple sequence alignment (MSA)ommonly the rst step of ET analysis. We obtained 23f complete hexon amino acid sequences of HAdV includ-types of species A, B, C, F and a few typical serotypes. The sequence data were all derived from NCBI Gen-the following accession numbers: AC 000017 (HAdV1),(HAdV2), X76549 (HAdV3), NC 003266 (HAdV4),(HAdV5), DQ149613 (HAdV6), AC 000018 (HAdV7),(HAdV10), AC 000015 (HAdV11), X73487 (HAdV12),(HAdV14), DQ149617 (HAdV15), AY601636 (HAdV16),(HAdV18), AY008279 (HAdV21), DQ149611 (HAdV31),</p></li><li><p>X. Yuan et al. / Vaccine 27 (2009) 51035110 5105</p><p>Fig. 1. Phylogeproteins. The ereected the dAccession num</p><p>AB052911 (X51782 (HDQ149643using the Cand adjustededuced amrelationshipstructed bythe MEGA4Fig. 1.</p><p>2.3.2. SitesSites ho</p><p>ferent sequis conservethe sites oferent hexothe alignedhomology=of total seqemployed ithe sites hothree rangereect the t</p><p>2.3.3. CandBecause</p><p>cell neutralalgorithm toless than45the followinbetween 6 asites, (3) inthree, and (residues gredate epitopHEX3 moderonment. Alepitopes scBioperl scri</p><p>Table 1Synthesized polypeptides and the location of S1S5 and two control polypeptidesP1 and P2.</p><p>Segment Locationa Sequence</p><p>locatioerlinin</p><p>itope</p><p>ynthve</p><p>menith thsizermea948peciis, whis bu</p><p>ancrrectinalolypeprothemethcatio</p><p>PeptiELIS</p><p>nd biAdV, 1/2of PBS1Seconetic relationships of deduced amino acid sequences of HAdV hexonvolutionary historywas inferred using theMPmethod. Unrooted treeistance relationship of all HAdV hexon amino acid sequences in it.ber of reference sequences is shown in the tree.</p><p>HAdV34), AB052912 (HAdV35), DQ149632 (HAdV37),AdV40), DQ315364 (HAdV41), EF153473 (HAdV48),(HAdV50). Besides the HEX3, MSA was performedlustalW1.83 program [41] with the Clustal algorithmd manually to conform the optimized alignment ofino acid sequences. In order to show the evolutionaryof these sequences, a phylogenetic tree was con-the maximum-parsimony (MP) method [42] using</p><p>.0 package [43]. The phylogenetic tree is shown in</p><p>homology calculationmology refers to the similarity of the same site in dif-ences aligned by MSA. It reects how a certain sited in aligned homologous protein sequences. Takingf HEX3 as standard sites, the sites homology in dif-n amino acid sequences were calculated according toresult. The calculation method was as follows: sitenumberof conservedaminoacids on samesite/number</p><p>S1S2S3S4S5P1P2</p><p>a Theb Und</p><p>tides.</p><p>2.4. Ep</p><p>2.4.1. SThe</p><p>above-[45]wSyntheIn theP2: 46type-sanalysanalysperformand coN-termEach pcarrierlimpetment mtheir lo</p><p>2.4.2.The</p><p>and biparedH1/1000100ltides (negativuences100%. Then a color mapping scheme wasn InsightII environment to map the information frommology calculation onto the 3D model of HEX3, we sets of site homology values (90%, 60%, and 30%) torend of sites homology.</p><p>idate epitopes screeningthe epitopes on the hexon protein are type-specic B-izing epitopes [4,5], we designed an epitope-screeningscreen the candidate epitopes: sitewith a homology of</p><p>%weredenedashypervariable site. Segments fulllingg standards were selected as candidates: (1) length ofnd 15, (2) more than half of them being hypervariableterval between candidate sequences not shorter than4) 90% of residues belonging to the exposed group withater than 25% of the maximum SAS. Finally, ve candi-e segments were screened out and mapped onto the 3Dl using a color mapping scheme in the InsightII envi-l above of the sites homology calculation and candidatereening were implemented in programs written by thept language [44].</p><p>in our lab [a positive cblockedwitfor 2h at 37was added tantibody wgoat serum&amp; Services Cment with...</p></li></ul>