construction and characterization of human adenovirus serotype 3 packaged by serotype 7 hexon

7
Virus Research 160 (2011) 214–220 Contents lists available at ScienceDirect Virus Research journal homepage: www.elsevier.com/locate/virusres Construction and characterization of human adenovirus serotype 3 packaged by serotype 7 hexon Xingui Tian, Xiaobo Su, Haitao Li, Xiao Li, Zhichao Zhou, Wenkuan Liu, Rong Zhou State Key Lab of Respiratory Disease, The First Affiliated Hospital of Guangzhou Medical College, Guangzhou Medical University, Guangzhou 510120, China article info Article history: Received 10 February 2011 Received in revised form 16 June 2011 Accepted 22 June 2011 Available online 29 June 2011 Keywords: Chimera Hexon Fiber Human adenovirus Serotype Hypervariable regions abstract Human adenovirus serotype 3 (Ad3) and serotype 7 (Ad7) are important pathogens causing respiratory tract diseases such as acute respiratory disease in pediatric and adult patients, but the immunodominant targets of Ad3- and Ad7-specific neutralizing antibodies (NAbs) remain unclear. A chimeric Ad vector, Ad3/H7, was constructed by replacing the Ad3 hexon gene (H3) with the hexon gene (H7) of Ad7. The chimeric viruses were successfully rescued in HEp-2 cells, and the Ad7 hexon was able to encapsidate the Ad3 genome, and functioned as efficiently as the Ad3 hexon. Furthermore, we tested the host neutraliza- tion responses against the viruses using BALB/C mice. Up to 97% of the NAbs produced by mice that were infected with these viruses were specific for the hexon protein in vitro. Preimmunization of mice with one of Ad7 and Ad3/H7 significantly prevented subsequent intranasal infection of the other type in vivo. In contrast, preimmunization of mice with one of Ad3 and Ad3/H7 did not remarkably prevent subsequent infection of the other type. We next evaluated the functional significance of hexon and other structural proteins specific NAbs to suppress the immunogenicity of Ad3/H3 and Ad3/H7 vectors expressing EGFP in mice preimmunized with wild type Ad. Preimmunization of mice with Ad7 evidently suppressed EGFP- specific humoral immune responses elicited by Ad3/H7, and did not exert suppressive effects on Ad3/H3. But contrary to the in vitro neutralization results, EGFP-specific humoral immune responses elicited by Ad3/H7 was remarkably inhibited in Ad3-preimmunization mice. The whole genome of the Ad7 strain was sequenced and aligned with Ad3. The major differences between Ad3 and Ad7 were only observed in the fiber and hexon among all structural proteins, and the variation between the hexons only located in four hypervariable regions (HVRs), HVR-1, -2, -5, and -7. These results thus suggest that Ad3- and Ad7-specific NAbs are directed primarily against the hexon proteins both in vitro and in vivo. But high titer Ad3 fiber-specific NAbs may also play an important role in blunting Ad3 immunogenicity in vivo. These studies contribute to a more profound understanding of Ad immunogenicity and have relevance for the design of novel Ad vaccine. © 2011 Elsevier B.V. All rights reserved. 1. Introduction Human adenoviruses (HAdVs) in the alimentary canal and res- piratory tract were discovered in 1953, and have been found to cause a broad spectrum of diseases in pediatric and adult patients (Aoki and Tagawa, 2002; Arnold et al., 2010; Kunz and Ottolini, 2010; Louie et al., 2008; Rowe et al., 1953). HAdV can be classified into six species (A–F) consisting of 53 serotypes, based on serum neutralization antibodies and nucleotide sequence. Recently, a new serotype, HAdV52, was reported and defined as a seventh species, species G (Davison et al., 2003; Green et al., 1979; Ishiko et al., 2008; Corresponding author. Tel.: +86 20 34281614; fax: +86 20 34281614. E-mail addresses: [email protected] (X. Tian), [email protected] (X. Su), [email protected] (H. Li), [email protected] (X. Li), [email protected] (Z. Zhou), [email protected] (W. Liu), [email protected] (R. Zhou). Jones et al., 2007; Walsh et al., 2009). Recently, a broad recombina- tion phenomenon has been discovered between different serotypes of virus strains leading to lethal strains, some of which may be new serotypes (Kajon et al., 2010; Lukashev et al., 2008; Rebelo-de- Andrade et al., 2010). Members of species B, HAdV-3 and -7, have caused severe respiratory disease, such as acute respiratory dis- ease (ARD) pediatric pneumonia epidemics and outbreaks in Asia, Europe and America (Erdman et al., 2002; Frantzidou et al., 2005; Hierholzer, 1992; Li et al., 2005; Tang et al., 2011). As yet, there is no effective medicine or vaccine except for live, oral vaccines used by the US military from 1971 to 1996 (Gooch and Mogabgab, 1972; Lyons et al., 2008). For all these reasons, development of an effective vaccine directing HAdV-3 and -7 is required. An adequate knowledge of the immunodominant regions of adenovirus NAbs will be helpful for the development of novel ade- novirus vaccine. The Ad capsid consists of three major structural proteins: hexon, fiber, and penton base. Research carried out on adenovirus serotype 2 (Ad2) and Ad5 have identified the major 0168-1702/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.virusres.2011.06.017

Upload: xingui-tian

Post on 30-Oct-2016

218 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: Construction and characterization of human adenovirus serotype 3 packaged by serotype 7 hexon

Cb

XS

a

ARRAA

KCHFHSH

1

pc(2inss

(yz

0d

Virus Research 160 (2011) 214–220

Contents lists available at ScienceDirect

Virus Research

journa l homepage: www.e lsev ier .com/ locate /v i rusres

onstruction and characterization of human adenovirus serotype 3 packagedy serotype 7 hexon

ingui Tian, Xiaobo Su, Haitao Li, Xiao Li, Zhichao Zhou, Wenkuan Liu, Rong Zhou ∗

tate Key Lab of Respiratory Disease, The First Affiliated Hospital of Guangzhou Medical College, Guangzhou Medical University, Guangzhou 510120, China

r t i c l e i n f o

rticle history:eceived 10 February 2011eceived in revised form 16 June 2011ccepted 22 June 2011vailable online 29 June 2011

eywords:himeraexoniberuman adenoviruserotypeypervariable regions

a b s t r a c t

Human adenovirus serotype 3 (Ad3) and serotype 7 (Ad7) are important pathogens causing respiratorytract diseases such as acute respiratory disease in pediatric and adult patients, but the immunodominanttargets of Ad3- and Ad7-specific neutralizing antibodies (NAbs) remain unclear. A chimeric Ad vector,Ad3/H7, was constructed by replacing the Ad3 hexon gene (H3) with the hexon gene (H7) of Ad7. Thechimeric viruses were successfully rescued in HEp-2 cells, and the Ad7 hexon was able to encapsidate theAd3 genome, and functioned as efficiently as the Ad3 hexon. Furthermore, we tested the host neutraliza-tion responses against the viruses using BALB/C mice. Up to 97% of the NAbs produced by mice that wereinfected with these viruses were specific for the hexon protein in vitro. Preimmunization of mice with oneof Ad7 and Ad3/H7 significantly prevented subsequent intranasal infection of the other type in vivo. Incontrast, preimmunization of mice with one of Ad3 and Ad3/H7 did not remarkably prevent subsequentinfection of the other type. We next evaluated the functional significance of hexon and other structuralproteins specific NAbs to suppress the immunogenicity of Ad3/H3 and Ad3/H7 vectors expressing EGFP inmice preimmunized with wild type Ad. Preimmunization of mice with Ad7 evidently suppressed EGFP-specific humoral immune responses elicited by Ad3/H7, and did not exert suppressive effects on Ad3/H3.But contrary to the in vitro neutralization results, EGFP-specific humoral immune responses elicited byAd3/H7 was remarkably inhibited in Ad3-preimmunization mice. The whole genome of the Ad7 strainwas sequenced and aligned with Ad3. The major differences between Ad3 and Ad7 were only observed

in the fiber and hexon among all structural proteins, and the variation between the hexons only locatedin four hypervariable regions (HVRs), HVR-1, -2, -5, and -7. These results thus suggest that Ad3- andAd7-specific NAbs are directed primarily against the hexon proteins both in vitro and in vivo. But hightiter Ad3 fiber-specific NAbs may also play an important role in blunting Ad3 immunogenicity in vivo.These studies contribute to a more profound understanding of Ad immunogenicity and have relevance

vacc

for the design of novel Ad

. Introduction

Human adenoviruses (HAdVs) in the alimentary canal and res-iratory tract were discovered in 1953, and have been found toause a broad spectrum of diseases in pediatric and adult patientsAoki and Tagawa, 2002; Arnold et al., 2010; Kunz and Ottolini,010; Louie et al., 2008; Rowe et al., 1953). HAdV can be classified

nto six species (A–F) consisting of 53 serotypes, based on serum

eutralization antibodies and nucleotide sequence. Recently, a newerotype, HAdV52, was reported and defined as a seventh species,pecies G (Davison et al., 2003; Green et al., 1979; Ishiko et al., 2008;

∗ Corresponding author. Tel.: +86 20 34281614; fax: +86 20 34281614.E-mail addresses: [email protected] (X. Tian), [email protected]

X. Su), [email protected] (H. Li), [email protected] (X. Li),[email protected] (Z. Zhou), [email protected] (W. Liu),[email protected] (R. Zhou).

168-1702/$ – see front matter © 2011 Elsevier B.V. All rights reserved.oi:10.1016/j.virusres.2011.06.017

ine.© 2011 Elsevier B.V. All rights reserved.

Jones et al., 2007; Walsh et al., 2009). Recently, a broad recombina-tion phenomenon has been discovered between different serotypesof virus strains leading to lethal strains, some of which may benew serotypes (Kajon et al., 2010; Lukashev et al., 2008; Rebelo-de-Andrade et al., 2010). Members of species B, HAdV-3 and -7, havecaused severe respiratory disease, such as acute respiratory dis-ease (ARD) pediatric pneumonia epidemics and outbreaks in Asia,Europe and America (Erdman et al., 2002; Frantzidou et al., 2005;Hierholzer, 1992; Li et al., 2005; Tang et al., 2011). As yet, there isno effective medicine or vaccine except for live, oral vaccines usedby the US military from 1971 to 1996 (Gooch and Mogabgab, 1972;Lyons et al., 2008). For all these reasons, development of an effectivevaccine directing HAdV-3 and -7 is required.

An adequate knowledge of the immunodominant regions of

adenovirus NAbs will be helpful for the development of novel ade-novirus vaccine. The Ad capsid consists of three major structuralproteins: hexon, fiber, and penton base. Research carried out onadenovirus serotype 2 (Ad2) and Ad5 have identified the major
Page 2: Construction and characterization of human adenovirus serotype 3 packaged by serotype 7 hexon

search

ci2whbHa(pm

opt2taWo(awhsAt

2

2

EpewabpcK

2

w(tp(ra

fCCaBHt

taP

X. Tian et al. / Virus Re

oat protein of the HAdV hexon as the main target for neutral-zing antibodies (NAbs) (Pichla-Gollon et al., 2007; Sumida et al.,005; Wu et al., 2002; Youil et al., 2002); however, it is not clearhether this is the case for other HAdV serotypes. Some studiesave found that other structural proteins, such as fiber and pentonase, also could induce neutralizing antibodies (Fender et al., 1995;ong et al., 2003; Liebermann et al., 1998). Little is currently knownbout the neutralizing epitopes of the more than 50 HAdV serotypesPichla-Gollon et al., 2007; Rux and Burnett, 2000), although it isossible to locate HVRs using the multiple sequence alignmentethod (Crawford-Miksza and Schnurr, 1996; Ebner et al., 2005).Replacement of the immunogenic capsid proteins with those of

ther serotypes is one of the approaches used to circumvent theroblem of pre-existing immunity to an adenovirus delivery vec-or (Gall et al., 1998; Roberts et al., 2006; Rux et al., 2003; Wu et al.,002; Youil et al., 2002). Some studies based on the HAdV5 vec-or have suggested that replacement of the hexon may or may notffect the packaging and gene transfer efficiency (Gall et al., 1998;u et al., 2002; Youil et al., 2002). An Ad3 vector has been devel-

ped previously as a candidate for vaccine design and gene transferZhang et al., 2009). In the present study, we attempted to constructchimeric adenovirus, Ad3/H7, by replacing the Ad3 hexon geneith the hexon gene of Ad7 (H7) in the hope that this study couldelp understand the functional significance of NAbs against majortructure proteins, and the strategy used to modify the hexon fromd3 could be applied to adenovirus vaccine development and gene

herapy.

. Materials and methods

.1. Cells, virus strains and vectors

The wild type Ad3 strain HAdV3 GZ-01 (w-Ad3) and the3-defective replication-competent EGFP expression plasmid,BRAdV3dE3egfp (pAd3egf), corresponding virus Ad3/H3 (AdV3-gf) was got from Zhang et al. (2009). The Ad7 used in this studyas HAdV7-gz08 strain, originally isolated from a child with ARD

nd cultured in HEp-2 cells. The HEp-2 cells were grown in Dul-ecco’s modified Eagle’s medium (DMEM) supplemented withenicillin (100 IU/ml), streptomycin (100 �g/ml) and 10% (v/v) fetalalf serum. The viral DNA was extracted by the modified proteasedigestion method without virus purification (Hardy et al., 1997).

.2. Construction of recombinant plasmid

The shuttle vector, pBRH3S, for the Ad3 hexon replacementas constructed via two sub cloning steps. The left segment, L1

4.34 kb), containing nucleotides (nt) 16,764–21,140 of pA3egf andhe right segment, L2 (3.44 kb), containing nt 21125–24569 ofA3egf, were amplified separately by polymerase chain reactionPCR). L1 was cloned into pBR322 between the EcoRI and BamHIestriction endonuclease sites, with L2 ligated between the BamHInd SalI sites to obtain pBRH3S.

To create the shuttle vector, pBRH7S, H7 was obtained by PCRrom Ad7 with the primers Hexon-F (5′-CCG AGG CTG AGT TGC TTTAA GAT GGC-3′) and Hexon-R (5′-GCA CTC TGA CCA CGT CGA AGATT CG-3′). The H7 PCR product was digested with ClaI and BamHInd ligated into pBRH3S that had also been treated with ClaI andamHI. The resulting shuttle vector, pBRH7S, contained 2681 nt of7 from nt 12–2692. The resultant hexon was not different from

he wild-type full-length H7.

The pBRH7S shuttle vector was digested with EcoRI and SalI, and

he fragment containing the homologous recombination regionsnd the hexon genes was purified and recombined with AvRII- andacI-digested pAd3egf in Escherichia coli strain BJ5183. The resul-

160 (2011) 214–220 215

tant clones containing H7 and EGFP were designated pAd3egf/H7and selected by PCR using primers a7u (5′-ACA GCA GGA GAA GAAAGA G-3′), a7r (5′-CCA TCC ATT GTC TTT AGA T-3′), egfu (5′-GAAATC GAT GTG AGC AAG GGC GAG GAG CT-3′) and egfr (5′-CTA GATCCG GTG GAT CCC-3′). Constructs were confirmed by restrictiondigestion and sequencing analysis.

2.3. Virus rescue and preparation

To rescue modified Ad3 viruses, pAd3egf/H7 was digestedwith AsisI. Linearized genomic DNA was transfected into HEp-2cells grown in 30 mm dishes using Lipofectamine2000 (Invitrogen,Carlsbad, CA, USA). The transfected cells were cultured at 37 ◦C/5%CO2 for 6–10 days and were examined for evidence of cytopathiceffect daily. Cells were frozen and thawed for three cycles and freshcultures of HEp-2 cells were infected with viral suspension. At 96 hpost-infection, the virus was harvested and designated Ad3/H7.Finally a total of ten 100-mm dishes containing HEp-2 cells infectedwith Ad3/H7 were harvested. Viruses were purified by standardCsCl gradient centrifugation as described by Wu (Wu et al., 2002).The virus particle (VP) titers were determined by spectrophotome-try using a conversion factor of 1.1 × 1012 VPs per absorbance unitat 260 nm.

2.4. Growth characteristics and heat stability of Ad3/H7

To evaluate DNA replication efficiency of Ad3/H3 or Ad3/H7,quantitative PCR (qPCR) was performed with HAdV using a Q-PCR kit (South China IOV, Guangzhou, China). HEp-2 cells wereseeded into 24-well plates at a density of 100,000 cells/well the daybefore infection. The cells were infected with 1 × 106 DNA copiesof Ad3/H3, Ad3/H7 or Ad7, and the infected cells including themedium for each virus were harvested every 12 h. Viral genomicDNA copy numbers could then be determined by qPCR.

At the same time, HEp-2 cells were seeded into 96-well NUNCOptical bottom plates (ThermoFisher Scientific, Rochester, NY, USA)and infected with 1 × 105, 1 × 106 and 1 × 107 copies of Ad3/H3 orAd3/H7, and cultured in RPMI Medium 1640 (GIBCO, Carlsbad, CA,USA) without phenol red and without serum for 48 h. The fluo-rescence was measured with a Varioskan Flash Multimode Reader(Thermo Scientific, Waltham, MA, USA) by exciting at 488 nm andrecording the emitted light at 507 nm. Background fluorescencewas equalized to wells containing cells only.

To test the heat stability of Ad3/H7, Ad3/H3 and Ad7 viruses,1 × 107 copies/ml of viral DNA were incubated at 45 ◦C for differ-ent periods before infecting HEp-2 cells. Viral genomic DNA copynumbers in infected cells were analyzed at 48 h post-infection.

2.5. In vitro neutralization

Viruses were injected into mice by the intraperitoneal (i.p.)route at a concentration of 1010 VPs/mouse. Control mice wereinjected with phosphate-buffered saline (PBS). There were fivemice in each group. Mice were subjected to booster injections everytwo weeks for a total of three injections. At day 42 after the firstimmunization, blood samples were collected from the mice andsera were isolated and kept frozen for serology tests. Proceduresfor these animal experiments complied with all relevant federalguidelines and institutional policies.

Neutralization tests in vitro were performed to test the neu-tralizing effect of three antisera against Ad3/H3, Ad3/H7 and Ad7viruses. All the antisera and preimmune serum (negative control)

were serially diluted 2-fold in DMEM, and 50 �l aliquots of eachdilution were mixed with 50 �l of virus with 100 TCID50. Theantibody–virus mixtures were incubated for 1 h at 37 ◦C and thentransferred to 96-well plates containing 85–95% confluent mono-
Page 3: Construction and characterization of human adenovirus serotype 3 packaged by serotype 7 hexon

2 search

lpac

2

AiwAsilup

2

rEwpiiwww2T1jww(

F(PM

16 X. Tian et al. / Virus Re

ayers of HEp-2 cells. Monolayers were incubated for 72 h in theresence of 5% CO2. The serum neutralization titer was defineds the reciprocal of the serum dilution that inhibited 50% of theytopathic effect.

.6. In vivo neutralization

Each mouse was injected with PBS or 2 × 1010 VPs of w-Ad3 ord7, Ad3/H7. Each group included 9 mice. Mice received a booster

njection after two weeks. Twenty days later, each group of miceas subjected to an intranasal immunization of 2 × 109 VPs of w-d3 or Ad7, Ad3/H7 per mouse, resulting in three mice for eachubgroup. Three days later, the mice were sacrificed with CO2nhalation and the lungs were harvested, homogenized on dry ice,ysed in lysis buffer, and measured for Ad genome copies by qPCRsing a Q-PCR kit. Procedures for these animal experiments com-lied with all relevant federal guidelines and institutional policies.

.7. ELISA

For another in vivo neutralization experiment, each mouse alsoeceived twice i.p. injection of PBS or 2 × 1010 VPs of w-Ad3 or Ad7.ach group included six mice. Two weeks later, each group of miceas subjected to an intramuscular (i.m.) injection of 2 × 1010 VPser mouse of Ad3/H3 or Ad3/H7 vectors expressing EGFP, result-

ng in three mice for each subgroup. Two weeks after the thirdnjection, serum Ab titers from immunized mice specific for EGFP

ere measured by ELISA. Ninety-six well plates coated overnightith 100 �l/well of 1 �g/ml recombinant GFP (Millipore) in PBSere blocked for 2 h with PBS containing 2% BSA and 0.05% Tween

0. Sera were then added in serial dilutions and incubated for 1 h.he plates were washed three times with PBST and incubated forh with a 1/5000 dilution of Goat anti-Mouse IgG(H + L)-HRP con-

ugate affinity-purified secondary Ab (Bio-Rad). The plates wereashed four times, developed with tetramethylbenzidine, stoppedith 2 M H2SO4, and analyzed at 450 nm with ELISA plate reader

Thermo Scientific Multiskan MK3).

ig. 1. Confirmation of the Ad3/H7 chimera. (A) Restriction enzyme analysis of the reco3) pAd3egf/H7-4/KpnI, (4) pAd3egf/H7-7/BamHI, (5) pAd3egf/H7-7/EcoRV, (6) pAd3egf/HCR analysis with H7 specific primers, a7u and a7r. (9) Virus Ad3/H7-4, (10) virus Ad3/H71: DL-5000 DNA marker; M2: DL-15000 DNA marker; and M3: DL-2000 DNA marker.

160 (2011) 214–220

2.8. Comparison of Ad7 and Ad3 genome

The whole genome of HAdV7-gz08 strain was sequenced by PCRand annotated using a previously described method (Purkayasthaet al., 2005). The genome of HAdV3-GZ01 was sequenced before(GenBank Accession No. DQ099432.4) (Zhang et al., 2006, 2009).The whole genome comparisons and the main structural pro-tein sequence alignments were performed with blastn or blastpfrom NCBI (http://www.ncbi.nlm.nih.gov/BLAST/), and multiplesequence alignment was done with software MEGA4.1 and ClustalX(1.83).

3. Results

3.1. Generation of hexon chimeric vector pAd3egf/H7

The H7 gene obtained by PCR was used to replace the corre-sponding region of H3 in pBRH3S to obtain pBRH7S. This resultedin the replacement of H3 for amino acid residues 5–901. Becausethe N- and C-terminal regions of H3 are the same as that of H7, theresultant hexon gene was 100% identical to the native H7 gene. Thechimeric hexon plasmids, pAd3egf/H7, were obtained by homolo-gous recombination and confirmed by restriction enzyme analysiswith BamHI, EcoRV, AsisI and KpnI (Fig. 1A). Viruses Ad3/H7 wererescued in HEp-2 cells and confirmed by PCR with the Ad7-specificprimers, a7u and a7r (Fig. 1B). The hexon gene of virus Ad3/H7 was100% identical to the native H7 gene by sequencing. Two cloneswere determined and shown in Fig. 1, the clone Ad3/H7-4 was usedin the following work.

3.2. Growth characteristics of recombinant adenovirus

The replication of Ad3/H3, Ad7 and Ad3/H7 were comparedby quantifying genomic DNA using a quantitative PCR method

(Fig. 2A). The replication efficiency of the chimeric Ad3/H7 wasapproximately the same as that of Ad3/H3. We then examined theefficiency of gene transfer for Ad3/H7 in HEp-2 cells with the aidof the EGFP reporter. There was no apparent difference between

mbinant plasmid pAd3egf/H7. (1) pAd3egf/H7-4/BamHI, (2) pAd3egf/H7-4/EcoRV,7-7/KpnI, (7) pAd3egf/H7-4/AsisI and (8) pAd3egf/H7-7/AsisI. (B) Serotype-specific

-7, (11) negative control virus Ad3/H3, and (12) positive control virus HAdV7-gz08.

Page 4: Construction and characterization of human adenovirus serotype 3 packaged by serotype 7 hexon

search 160 (2011) 214–220 217

AA

tidcAti(s

Fan1Adcu

Table 1Neutralizing antibody titers from adenovirus-immunized mice.

Virus Titer of serum

Anti-Ad3/H3 Anti-Ad7 Anti-Ad3/H7

Ad3/H3 2048–4096 0a 64Ad7 0a 4096 4096Ad3/H7 64 4096 4096

a No detectable neutralization.

X. Tian et al. / Virus Re

d3/H3 and Ad3/H7, suggesting that the gene transfer ability ofd3/H3 was not reduced by hexon replacement with H7 (Fig. 2B).

By comparing the heat stabilities of Ad3/H3 and Ad3/H7 weested whether hexon replacement could affect the structuralntegrity of the virions. The viruses were incubated at 45 ◦C forifferent durations before infecting HEp-2 cells, and their repli-ation efficiencies were compared. The replication efficiency ofd3/H7 was not significantly reduced following a 10 min incuba-

ion at 45 ◦C, and its activity was retained at about 20% after a 30 min

ncubation at 45 ◦C, which was nearly the same as that of Ad3/H3Fig. 2C). This suggested that the chimeric Ad3/H7 virus was astable as Ad3/H3.

ig. 2. Characterization of Ad3/H7. (A) The DNA replication curve of Ad7, Ad3/H3nd Ad3/H7. Cells were harvested between 12 and 120 h. Viral genomic DNA copyumbers were determined by qPCR. (B) Fluorescence of HEp-2 cells infected with× 105, 1 × 106, 1 × 107 copies of Ad3/H3 or Ad3/H7 DNA. (C) Thermostability ofd3/H7. Approximately 1 × 107 DNA copies/ml of virus was incubated at 45 ◦C forifferent time periods before infecting HEp-2 cells. The relative infectivity was cal-ulated as the percentage of DNA copies of the heat-treated viruses compared withntreated viruses.

3.3. Ad3/H3 and Ad3/H7 were cross-neutralized in vitro

To determine whether Ad3/H3 and Ad3/H7, Ad7 could escapethe host neutralization response, we performed in vitro neutral-ization assays using serum obtained from mice experimentallyimmunized with 1010 VPs/mouse of Ad3/H3, Ad7 or Ad3/H7. Asshown in Table 1, the antisera were specific for the correspond-ing virus with titers ranging from 2048 to 4096. Serum fromAd3/H3-immunized mice had much lower but clearly detectablecross-reactive NAb titers against Ad3/H7. Serum from Ad3/H7-immunized mice had comparable NAb titers to Ad3/H7 and Ad7,but exhibited 30-fold lower NAb titers to Ad3/H3. Serum from Ad7-immunized mice had comparable NAb titers to Ad3/H7 and Ad7.As expected, there was no cross-neutralization between Ad3/H3and Ad7. These data clearly demonstrated that in vitro Ad3- andAd7-specific NAbs were principally directed against the hexons,although Ad3 fiber-specific NAbs were also present but at muchlower titers than Ad3 hexon-specific NAbs.

3.4. Ad3/H3 and Ad3/H7 were cross-neutralized in vivo

We next examined the in vivo cross-neutralization using micethat were pre-immunized twice with 2 × 1010 VPs of w-Ad3, Ad7,Ad3/H7 or PBS. As shown in Fig. 3A, the NAbs titer of these miceserum arrived about 2048–4096 in vitro.

Because Ad3 and Ad7 were ordinarily infected through respi-ration tract, the in vivo cross-neutralization of w-Ad3, Ad7 andAd3/H7 was firstly examined by measuring the Ad genome DNAcopies in lungs of mice immunized through intranasal infection. Asshown in Fig. 3C, preimmunization of mice with one of Ad7 andAd3/H7 significantly prevented subsequent infection of the othertype in vivo. In contrast, preimmunization of mice with one of w-Ad3 and Ad3/H7 did not remarkably inhibit subsequent infectionof the other type. These data suggest hexon-specific NAbs play amajor role in the protective effect against infection of Ad3 and Ad7in vivo.

Ad3/H3 and Ad3/H7 were also used as vectors expressing EGFP.We assessed EGFP-specific humoral immune responses followingAd3/H3 or Ad3/H7 infection using ELISA. Two weeks after the thirdinjection with 2 × 1010 VPs of Ad3/H3 or Ad3/H7, EGFP-specificserum Ab titers were measured by ELISA. As shown in Fig. 3B, anti-Ad7 immunity did not suppress Ad3/H3-elicited immune responsesin mice. Ad3/H3 immunogenicity was greatly inhibited by w-Ad3 pre-immunization. However, Ad3/H7 vector-elicited immuneresponses were greatly suppressed not only by anti-Ad7 immu-nity, but also by anti-w-Ad3 immunity, although the NAbs titersin vitro against Ad3/H7 of the mice preimmunized with w-Ad3were much lower than that preimmunized with Ad7 (Fig. 3A). Thesedata suggest that high titers of Ad3 fiber-specific NAbs may have animportant role in suppressing Ad vector immunogenicity in vivo,

which cannot be predicted by in vitro neutralization assays.
Page 5: Construction and characterization of human adenovirus serotype 3 packaged by serotype 7 hexon

218 X. Tian et al. / Virus Research

Fig. 3. Cross-neutralization in vivo of Ad3/H3 and Ad3/H7. BALB/c mice received twoinjections of 2 × 1010 VPs of w-Ad3, Ad7, Ad3/H7 or PBS. (A) Serum was obtainedfrom these mice following a third infection and assessed for NAbs to Ad3/H3 orAd7, Ad3/H7. (B) EGFP-specific ELISAs for serum obtained from mice immunizedwith Ad3/H3 or Ad3/H7 by i.m. injection. (C) Three days after booster injection with2 × 109 VPs of Ad3/H3 or Ad7, Ad3/H7 via intranasal infection, the mice were sac-rm

3H

G2o

ificed and their lungs were harvested for the qPCR assay. Results are shown as theeans (and standard deviations) for samples from three individual mice.

.5. HAdV7-GZ08 DNA sequence and comparison withAdV3-GZ01

The complete genome of HAdV7-gz08 (GenBank Accession No.Q478341) was 35,306 bp with a base composition of 25.2% G,5.8% C, 25.4% A and 23.6% T. The genome of Ad7 had a G + C contentf 51%, similar to the G + C content of other members of species B

160 (2011) 214–220

adenoviruses (50–52%). The genome of HAdV7-gz08 had only oneor two amino acid changes in the major structural proteins com-pared with the prototype HAdV-7 Gomen strain (Purkayastha et al.,2005). The genomes of strains HAdV7-gz08 and HAdV3-GZ01 share97% identity. The hexons, fibers, and all other structural proteinscoding sequences of strains HAdV7-gz08 and HAdV3-GZ01 were 96,57 and >99% identical at amino acid level, respectively. There werefour single amino acid substitutions in the penton base, three sub-stitutions in pVI, four in pVIII, five in pIVa2, a single change in pIIIaand no change in pIX. Hexon protein sequences of HAdV3-GZ01and HAdV7-gz08 were analyzed by multiple sequence alignment.Four regions of variability were identified; they were in amino acids139–143, 175–180, 268–275 and 416–441 of HAdV3-GZ01, whichwere respectively located in HVR-1, -2, -5 and -7 in the hexon struc-tures, and may contribute to immunogenicity (Crawford-Mikszaand Schnurr, 1996; Rux et al., 2003) (Fig. 4). In the rest of the hexonsequences, there were six single amino acid substitions.

The fiber sequences of HAdV3-GZ01 and HAdV7-gz08 were alsoanalyzed using the blastp program. It was found that the identityof them was low (58%), but the fibers of HAdV7-gz08 and somestrains of subspecies B2, such as HAdV11, HadV14 and HAdV34a,were greater than 92% identical.

4. Discussion

For adenovirus vaccine development, it is important to deter-mine the neutralizing antigens of pathogenic adenovirus types,such as Ad3, Ad4 and Ad7. Any of the major capsid proteins maybe recognized by NAbs (Fender et al., 1995; Hong et al., 2003;Liebermann et al., 1998). It has been previously reported that hexonis the major target of adenovirus NAbs (Pichla-Gollon et al., 2007;Sumida et al., 2005; Wu et al., 2002; Youil et al., 2002). The presentexperimental data in vitro showed that up to 97% of the NAbsproduced by mice that were infected with Ad3 and Ad7 werehexon-specific. The known genome sequences of the two strainsmade it simple to analyze the results of neutralization experiments.Comparative analysis of Ad3 and Ad7 structural proteins showedthat major differences were only observed in the fiber and hexon.The fiber of Ad3 led to weak neutralization and no cross-reactionwith Ad7. The penton base protein did not lead to neutralization andwas unable to react with NAbs. The fiber-specific NAbs accountedfor less than 3% of the total NAb pool in vitro, which partly explainsthe results of other investigators who have reported that fiber-specific NAbs could be detected at high titers in human sera, butnot in sera with lower neutralizing titers (Nanda et al., 2005). In thecurrent experiments it could not be concluded whether the fiber ofthe Ad7 virus can induce neutralizing antibodies. To construct therecombinant type 3 virus entirely escaping the host neutralizationresponse, the fiber domain needs to be reconstructed to delete suchactivity. The fiber of virus Ad3/H7 could be replaced with the fiber ofAd7 in the next stage, to produce a recombinant virus with almostidentical structural proteins of the Ad7 virus.

Inconsistent with the in vitro neutralization results and previousstudy using rAd5 vectors (Sumida et al., 2005), preimmunization ofw-Ad3 could markedly suppress Ad3/H7-elicited EGFP-specific Abresponse in vivo. The first possible reason is that Ad3 fiber-specificNAbs may play a more important role in vivo than in vitro, perhapsreflecting different mechanisms of virus neutralization from Ad5.Unlike members of other adenovirus species that bind to the cellsurface receptor, coxsackie and adenovirus receptor (CAR), mem-bers of species B recognize the membrane co-factor proteins CD46,

CD80 or CD86 as a cellular receptor (Short et al., 2006; Sirena et al.,2004). The second possible reason is the method. Different fromthe previous studies using adoptive transfer studies, we used micepreimmunized with Ad following injection. The suppression might
Page 6: Construction and characterization of human adenovirus serotype 3 packaged by serotype 7 hexon

X. Tian et al. / Virus Research 160 (2011) 214–220 219

F V-5. Ow

baeAvabi2iesvmwouaitd

tctBiaArSdwhoeosscg

ig. 4. Hexon protein sequence alignments of HAdV3-GZ01, HAdV7-gz08, and HAdas from GeneBank: AP 000211.1.

e led to not only by Ad-specific humoral immune responses butlso by Ad-specific cellular immune responses in this study. Differ-nt immunization routes may also contribute to different results.d3/H3 or Ad3/H7 vectors expressing EGFP were injected to miceia i.m. A previous study also suggested that in vitro neutralizationssays might not reliably predict the effects of virus-specific anti-odies on adenovirus vectors in vivo, and the mice were injected via

.m. with AdC68 orAdCDQ vectors in the study (Pichla-Gollon et al.,009). These data suggest that Ad3 and Ad7 neutralization in vivo

s a more complex process than in vitro neutralization. In anotherxperiment of Ad genome copies measurement in mice, hexon-pecific NAbs significantly prevented subsequent infection of Adia intranasal tract in vivo, but limited to the number of each groupice, it was not well-defined whether preimmunization of miceith one of w-Ad3 and Ad3/H7 could prevent subsequent infection

f the other type. More profound researches may be necessary fornderstanding the Ad3 and Ad7 fiber-specific neutralization mech-nism in vivo. However, the capability of the fiber of virus particlesnducing neutralization is still weak compared with hexon in vivo,herefore it may be more efficient to use hexon as a target in vaccineevelopment.

Although specific sites recognized by NAbs have been proposedo reside within small regions of the hexon that are exposed to theapsid surface and possess sequences that vary among serotypes,hey have only been identified in a few adenoviruses (Rux andurnett, 2000). The main difference was seven highlighted HVRs,

n which HVR7 could be further subdivided into three highly vari-ble regions (Crawford-Miksza and Schnurr, 1996; Rux et al., 2003).mong the HVRs of adenoviruses, only a few are type-specific NAbecognition sites (Pichla-Gollon et al., 2007; Rux and Burnett, 2000).equence alignment (Fig. 3) clearly demonstrated that there was noifference within HVR-3, -4 and -6 between the Ad3 and Ad7 hexon,ith variation only in HVR-1, -2, -5 and -7. Types 3 and 7 adenovirusexon-neutralizing epitopes should be determined by one or moref the four HVRs. Yuan et al. (2009) used homology modeling and anpitope mapping method to determine five neutralizing epitopesf the adenovirus type 3 Harbin04B strain hexon (GenBank Acces-

ion No. EU078562). A comparison with hexon sequences of thetrains used in this study indicate they may be the same hexon HVRsontained in strain Harbin04B serotype 7. Work is currently pro-ressing to exchange the HVR-1, -2, -5 and -7 between Ad3 and Ad7

nly N-terminal regions are shown. The HAdV-5 hexon protein sequence used here

to distinguish the neutralizing epitopes and construct a chimerichexon embodying neutralizing epitopes of both serotypes.

Previous studies have suggested a strategy for engineering novelAd5 vectors through replacement of the hexon to evade dominantAd5-specific NAbs (Gall et al., 1998; Wu et al., 2002). In this study,the chimeric recombinant Ad3 virus whose hexon was replacedwith the same subspecies Ad7 hexon could be generated, andthe replacement had no apparent effect on the replication effi-ciency, thermostability and gene transfer ability. In our anotherexperiment, a recombinant adenovirus plasmid pAd3egf/H4 wassuccessfully constructed by replacing Ad3 hexon gene of pAd3egfwith Ad4 hexon (Group E), but chimeric virus Ad3/H4 could notbe rescued, despite several attempts with several different plas-mid clones (data not shown). This may be the major differenceamong the virus structural proteins of various species, resultingin defective packaging. The incompatibility to hexon of other Adspecies agrees with a study based on the Ad5 vector (Youil et al.,2002). The potential incompatibility of various species of Ad mayinfluence Ad evolution mode. We also constructed a recombinantAd3 expressing H7 which gene was inserted into the E3 region ofA3/H3, but we did not know whether H7 was embeded in the Ad3capsid. Roberts et al. (2006) obtained a modified Ad5 that success-fully evaded the role of Ad5 neutralizing antibodies by using Ad48hexon HVRs to replace the original HVRs of the Ad5 hexon. Ourstudy and the results of others (Matthews et al., 2010; Roberts et al.,2006) indicate that the HVRs are the major targets of Ad-specificNAbs, and modification of specific HVRs may be a more effectivehexon gene-switch strategy. This strategy can be used to developadenovirus vectors able to evade pre-existing humoral immunity.A multivalent adenovirus vaccine inducing NAbs against severalserotypes could also be constructed in this way.

Acknowledgments

This work was supported by the National Nature Science Foun-dation of China (NSFC, 30770102), School of Guangzhou Research

Projects (No. 10A006D) and the Mérieux Foundation in France. Wethank Zifeng Yang, Wenda Guan, Tianhua Zhong, Rong Liu andSheng Qin of State Key Lab of Respiratory Disease for their kindhelp with this work.
Page 7: Construction and characterization of human adenovirus serotype 3 packaged by serotype 7 hexon

2 search

R

A

A

C

D

E

E

F

F

G

G

G

H

H

H

I

J

K

K

L

L

L

L

L

M

Gen. Virol. 87 (Pt 6), 1531–1541.Zhang, Q., Su, X., Seto, D., Zheng, B.J., Tian, X., Sheng, H., Li, H., Wang, Y., Zhou,

20 X. Tian et al. / Virus Re

eferences

oki, K., Tagawa, Y., 2002. A twenty-one year surveillance of adenoviral conjunctivi-tis in Sapporo, Japan. Int. Ophthalmol. Clin. 42 (1), 49–54.

rnold, J., Janoska, M., Kajon, A.E., Metzgar, D., Hudson, N.R., Torres, S., Harrach,B., Seto, D., Chodosh, J., Jones, M.S., 2010. Genomic characterization of humanadenovirus 36, a putative obesity agent. Virus Res. 149 (2), 152–161.

rawford-Miksza, L., Schnurr, D.P., 1996. Analysis of 15 adenovirus hexon proteinsreveals the location and structure of seven hypervariable regions containingserotype-specific residues. J. Virol. 70 (3), 1836–1844.

avison, A.J., Benko, M., Harrach, B., 2003. Genetic content and evolution of aden-oviruses. J. Gen. Virol. 84 (Pt 11), 2895–2908.

bner, K., Pinsker, W., Lion, T., 2005. Comparative sequence analysis of the hexongene in the entire spectrum of human adenovirus serotypes: phylogenetic, tax-onomic, and clinical implications. J. Virol. 79 (20), 12635–12642.

rdman, D.D., Xu, W., Gerber, S.I., Gray, G.C., Schnurr, D., Kajon, A.E., Anderson,L.J., 2002. Molecular epidemiology of adenovirus type 7 in the United States,1966–2000. Emerg. Infect. Dis. 8 (3), 269–277.

ender, P., Kidd, A.H., Brebant, R., Oberg, M., Drouet, E., Chroboczek, J., 1995. Anti-genic sites on the receptor-binding domain of human adenovirus type 2 fiber.Virology 214 (1), 110–117.

rantzidou, F., Pavlitou, A., Mataftsi, A., Dumaidi, K., Georgiadis, N., 2005. Molecularepidemiology of adenovirus strains isolated from patients with ocular diseasein the area of Thessaloniki, Greece (1998–2002). J. Med. Virol. 75 (3), 440–446.

all, J.G., Crystal, R.G., Falck-Pedersen, E., 1998. Construction and characterizationof hexon-chimeric adenoviruses: specification of adenovirus serotype. J. Virol.72 (12), 10260–10264.

ooch 3rd, W.M., Mogabgab, W.J., 1972. Simultaneous oral administration of liveadenovirus types 4 and 7 vaccines. Protection and lack of emergence of othertypes. Arch. Environ. Health 25 (6), 388–394.

reen, M., Mackey, J.K., Wold, W.S., Rigden, P., 1979. Thirty-one human adenovirusserotypes (Ad1-Ad31) form five groups (A-E) based upon DNA genome homolo-gies. Virology 93 (2), 481–492.

ardy, S., Kitamura, M., Harris-Stansil, T., Dai, Y., Phipps, M.L., 1997. Construction ofadenovirus vectors through Cre-lox recombination. J. Virol. 71 (3), 1842–1849.

ierholzer, J.C., 1992. Adenoviruses in the immunocompromised host. Clin. Micro-biol. Rev. 5 (3), 262–274.

ong, S.S., Habib, N.A., Franqueville, L., Jensen, S., Boulanger, P.A., 2003. Identificationof adenovirus (ad) penton base neutralizing epitopes by use of sera from patientswho had received conditionally replicative ad (addl1520) for treatment of livertumors. J. Virol. 77 (19), 10366–10375.

shiko, H., Shimada, Y., Konno, T., Hayashi, A., Ohguchi, T., Tagawa, Y., Aoki, K., Ohno,S., Yamazaki, S., 2008. Novel human adenovirus causing nosocomial epidemickeratoconjunctivitis. J. Clin. Microbiol. 46 (6), 2002–2008.

ones 2nd, M.S., Harrach, B., Ganac, R.D., Gozum, M.M., Dela Cruz, W.P., Riedel, B.,Pan, C., Delwart, E.L., Schnurr, D.P., 2007. New adenovirus species found in apatient presenting with gastroenteritis. J. Virol. 81 (11), 5978–5984.

ajon, A.E., Dickson, L.M., Murtagh, P., Viale, D., Carballal, G., Echavarria, M., 2010.Molecular characterization of an adenovirus 3–16 intertypic recombinant iso-lated in Argentina from an infant hospitalized with acute respiratory infection.J. Clin. Microbiol. 48 (4), 1494–1496.

unz, A.N., Ottolini, M., 2010. The role of adenovirus in respiratory tract infections.Curr. Infect. Dis. Rep. 12 (2), 81–87.

i, L., Phan, T.G., Nguyen, T.A., Kim, K.S., Seo, J.K., Shimizu, H., Suzuki, E., Okitsu,S., Ushijima, H., 2005. Molecular epidemiology of adenovirus infection amongpediatric population with diarrhea in Asia. Microbiol. Immunol. 49 (2), 121–128.

iebermann, H., Mentel, R., Bauer, U., Pring-Akerblom, P., Dolling, R., Modrow, S.,Seidel, W., 1998. Receptor binding sites and antigenic epitopes on the fiber knobof human adenovirus serotype 3. J. Virol. 72 (11), 9121–9130.

ouie, J.K., Kajon, A.E., Holodniy, M., Guardia-LaBar, L., Lee, B., Petru, A.M., Hacker,J.K., Schnurr, D.P., 2008. Severe pneumonia due to adenovirus serotype 14: anew respiratory threat? Clin. Infect. Dis. 46 (3), 421–425.

ukashev, A.N., Ivanova, O.E., Eremeeva, T.P., Iggo, R.D., 2008. Evidence of frequentrecombination among human adenoviruses. J. Gen. Virol. 89 (Pt 2), 380–388.

yons, A., Longfield, J., Kuschner, R., Straight, T., Binn, L., Seriwatana, J., Reitstetter,R., Froh, I.B., Craft, D., McNabb, K., Russell, K., Metzgar, D., Liss, A., Sun, X., Towle,A., Sun, W., 2008. A double-blind, placebo-controlled study of the safety andimmunogenicity of live, oral type 4 and type 7 adenovirus vaccines in adults.

Vaccine 26 (23), 2890–2898.

atthews, Q.L., Fatima, A., Tang, Y., Perry, B.A., Tsuruta, Y., Komarova, S., Timares,L., Zhao, C., Makarova, N., Borovjagin, A.V., Stewart, P.L., Wu, H., Blackwell, J.L.,Curiel, D.T., 2010. HIV antigen incorporation within adenovirus hexon hyper-variable 2 for a novel HIV vaccine approach. PLoS One 5 (7), e11815.

160 (2011) 214–220

Nanda, A., Lynch, D.M., Goudsmit, J., Lemckert, A.A., Ewald, B.A., Sumida, S.M., Truitt,D.M., Abbink, P., Kishko, M.G., Gorgone, D.A., Lifton, M.A., Shen, L., Carville, A.,Mansfield, K.G., Havenga, M.J., Barouch, D.H., 2005. Immunogenicity of recombi-nant fiber-chimeric adenovirus serotype 35 vector-based vaccines in mice andrhesus monkeys. J. Virol. 79 (22), 14161–14168.

Pichla-Gollon, S.L., Drinker, M., Zhou, X., Xue, F., Rux, J.J., Gao, G.P., Wilson, J.M.,Ertl, H.C., Burnett, R.M., Bergelson, J.M., 2007. Structure-based identification ofa major neutralizing site in an adenovirus hexon. J. Virol. 81 (4), 1680–1689.

Pichla-Gollon, S.L., Lin, S.W., Hensley, S.E., Lasaro, M.O., Herkenhoff-Haut, L., Drinker,M., Tatsis, N., Gao, G.P., Wilson, J.M., Ertl, H.C., Bergelson, J.M., 2009. Effect ofpreexisting immunity on an adenovirus vaccine vector: in vitro neutralizationassays fail to predict inhibition by antiviral antibody in vivo. J. Virol. 83 (11),5567–5573.

Purkayastha, A., Su, J., Carlisle, S., Tibbetts, C., Seto, D., 2005. Genomic and bioin-formatics analysis of HAdV-7, a human adenovirus of species B1 that causesacute respiratory disease: implications for vector development in human genetherapy. Virology 332 (1), 114–129.

Rebelo-de-Andrade, H., Pereira, C., Giria, M., Prudencio, E., Brito, M.J., Cale, E., Taveira,N., 2010. Outbreak of acute respiratory infection among infants in Lisbon,Portugal, caused by human adenovirus serotype 3 and a new 7/3 recombinantstrain. J. Clin. Microbiol. 48 (4), 1391–1396.

Roberts, D.M., Nanda, A., Havenga, M.J., Abbink, P., Lynch, D.M., Ewald, B.A., Liu, J.,Thorner, A.R., Swanson, P.E., Gorgone, D.A., Lifton, M.A., Lemckert, A.A., Holter-man, L., Chen, B., Dilraj, A., Carville, A., Mansfield, K.G., Goudsmit, J., Barouch, D.H.,2006. Hexon-chimaeric adenovirus serotype 5 vectors circumvent pre-existinganti-vector immunity. Nature 441 (7090), 239–243.

Rowe, W.P., Huebner, R.J., Gilmore, L.K., Parrott, R.H., Ward, T.G., 1953. Isolation ofa cytopathogenic agent from human adenoids undergoing spontaneous degen-eration in tissue culture. Proc. Soc. Exp. Biol. Med. 84 (3), 570–573.

Rux, J.J., Burnett, R.M., 2000. Type-specific epitope locations revealed by X-ray crys-tallographic study of adenovirus type 5 hexon. Mol. Ther. 1 (1), 18–30.

Rux, J.J., Kuser, P.R., Burnett, R.M., 2003. Structural and phylogenetic analysis ofadenovirus hexons by use of high-resolution x-ray crystallographic, molecularmodeling, and sequence-based methods. J. Virol. 77 (17), 9553–9566.

Short, J.J., Vasu, C., Holterman, M.J., Curiel, D.T., Pereboev, A., 2006. Members of ade-novirus species B utilize CD80 and CD86 as cellular attachment receptors. VirusRes. 122 (1–2), 144–153.

Sirena, D., Lilienfeld, B., Eisenhut, M., Kalin, S., Boucke, K., Beerli, R.R., Vogt, L., Ruedl,C., Bachmann, M.F., Greber, U.F., Hemmi, S., 2004. The human membrane cofactorCD46 is a receptor for species B adenovirus serotype 3. J. Virol. 78 (9), 4454–4462.

Sumida, S.M., Truitt, D.M., Lemckert, A.A., Vogels, R., Custers, J.H., Addo, M.M.,Lockman, S., Peter, T., Peyerl, F.W., Kishko, M.G., Jackson, S.S., Gorgone, D.A.,Lifton, M.A., Essex, M., Walker, B.D., Goudsmit, J., Havenga, M.J., Barouch, D.H.,2005. Neutralizing antibodies to adenovirus serotype 5 vaccine vectors aredirected primarily against the adenovirus hexon protein. J. Immunol. 174 (11),7179–7185.

Tang, L., Wang, L., Tan, X., Xu, W., 2011. Adenovirus serotype 7 associated with asevere lower respiratory tract disease outbreak in infants in Shaanxi Province,China. Virol. J. 8 (23).

Walsh, M.P., Chintakuntlawar, A., Robinson, C.M., Madisch, I., Harrach, B., Hudson,N.R., Schnurr, D., Heim, A., Chodosh, J., Seto, D., Jones, M.S., 2009. Evidence ofmolecular evolution driven by recombination events influencing tropism in anovel human adenovirus that causes epidemic keratoconjunctivitis. PLoS One 4(6), e5635.

Wu, H., Dmitriev, I., Kashentseva, E., Seki, T., Wang, M., Curiel, D.T., 2002. Con-struction and characterization of adenovirus serotype 5 packaged by serotype 3hexon. J. Virol. 76 (24), 12775–12782.

Youil, R., Toner, T.J., Su, Q., Chen, M., Tang, A., Bett, A.J., Casimiro, D., 2002. Hexongene switch strategy for the generation of chimeric recombinant adenovirus.Hum. Gene Ther. 13 (2), 311–320.

Yuan, X., Qu, Z., Wu, X., Wang, Y., Liu, L., Wei, F., Gao, H., Shang, L., Zhang, H., Cui, H.,Zhao, Y., Wu, N., Tang, Y., Qin, L., 2009. Molecular modeling and epitopes map-ping of human adenovirus type 3 hexon protein. Vaccine 27 (37), 5103–5110.

Zhang, Q., Su, X., Gong, S., Zeng, Q., Zhu, B., Wu, Z., Peng, T., Zhang, C., Zhou, R.,2006. Comparative genomic analysis of two strains of human adenovirus type3 isolated from children with acute respiratory infection in southern China. J.

R., 2009. Construction and characterization of a replication-competent humanadenovirus type 3-based vector as a live-vaccine candidate and a viral deliveryvector. Vaccine 27 (8), 1145–1153.