Si

RMa

b

c

d

e

ARR1AA

KAHDSD

1

11fcwIfa

iT

0h

Journal of Virological Methods 189 (2013) 362– 369

Contents lists available at SciVerse ScienceDirect

Journal of Virological Methods

j our nal homep age: www.elsev ier .com/ locate / jv i romet

election and characterization of DNA aptamers for use in detection of aviannfluenza virus H5N1

onghui Wanga, Jingjing Zhaob, Tieshan Jiangc, Young M. Kwonc, Huaguang Lud, Peirong Jiaoe,ing Liaoe, Yanbin Lia,b,c,∗

Department of Biological and Agricultural Engineering, University of Arkansas, Fayetteville, AR 72701, USACell and Molecular Biology Program, University of Arkansas, Fayetteville, AR 72701, USADepartment of Poultry Science, Center of Excellence for Poultry Science, University of Arkansas, Fayetteville, AR 72701, USAAnimal Diagnostic Laboratory, Pennsylvania State University, State College, PA 16802, USACollege of Veterinary Medicine, South China Agricultural University, Guangzhou 510642, China

rticle history:eceived 6 August 2012eceived in revised form0 December 2012ccepted 11 March 2013vailable online 21 March 2013

eywords:vian influenza5N1NA aptamerELEXetection

a b s t r a c t

Aptamers are artificial oligonucleotides (DNA or RNA) that can bind to a broad range of targets. In diag-nostic and detection assays, aptamers represent an alternative to antibodies as recognition agents. Theobjective of this study was to select and characterize DNA aptamers that can specifically bind to avianinfluenza virus (AIV) H5N1 based on Systematic Evolution of Ligands by EXponential enrichment (SELEX)and surface plasmon resonance (SPR). The selection was started with an ssDNA (single-stranded DNA)library of 1014 molecules randomized at central 74 nt. For the first four selection cycles, purified hemag-glutinin (HA) from AIV H5N1 was used as the target protein, and starting from the fifth cycle, entire H5N1virus was applied in order to improve the specificity. After 13 rounds of selection, DNA aptamers thatbind to the H5N1 were isolated and three aptamer sequences were characterized further by sequencingand affinity binding. Dot blot analysis was employed for monitoring the SELEX process and conductingthe preliminary tests on the affinity and specificity of aptamers. With the increasing number of selectioncycles, a steady increase in the color density was observed, indicating that the aptamers with good bind-ing affinity to the target were enriched. The best aptamer candidate had a dissociation constant (KD) of

4.65 nM as determined by SPR, showing a strong binding between the HA and the selected aptamer. Thespecificity was determined by testing non-target AIV H5N2, H5N3, H5N9, H9N2 and H7N2. Negligiblecross-reactivity confirmed the high specificity of selected aptamers. The developed aptamer was thenapplied for detection of AIV H5N1 in spiked poultry swab samples. The obtained aptamers could openup possibilities for the development of aptamer-based medical diagnostics and detection assays for AIV

this

H5N1. (The H5N1 used in

. Introduction

The avian influenza virus (AIV) H5N1, discovered in the late990s, is a long filamentous or spherical virus (Shortridge et al.,998) with a segmented genome that belongs to type A in theamily Orthomyxoviridae. It has been reported in more than 60ountries for animal cases and in 15 countries for human casesith 607 people infected and 358 deaths since 2003 (WHO, 2012).

t has drawn global attentions due to the potential pandemic threator public health and enormous economic losses. Rapid detectionnd identification of AIV is crucial for public health protection and

∗ Corresponding author at: Department of Biological and Agricultural Engineer-ng, University of Arkansas, Fayetteville, AR 72701, USA.el.: +1 4795752881/4795752424; fax: +1 4795752846.

E-mail address: yanbinli@uark.edu (Y. Li).

166-0934/$ – see front matter © 2013 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.jviromet.2013.03.006

study was inactivated virus.)© 2013 Elsevier B.V. All rights reserved.

controlling the outbreaks. In addition to viral isolation, RT-PCR andELISA are used commonly for AIV detection with bio-recognitionligands of nucleic acid probes and antibodies, respectively. Cur-rently, diagnostic techniques for in-field detection of AI virus arebased on rapid diagnostic test kits and/or strip tests (Cho et al.,2013; Dusek et al., 2011). At present, some direct antigen detec-tion assays are available commercially for detection of AIV, suchas Directizen EZ Flu A and B (Becton Dickinson, Sparks, MD), BinaxNow Influenza A/B antigen kit (Binax, Portland, ME), and HumasisInfluenza A/B antigen test (Humasis, Anyang, ROK). These assaysuse antibodies as bio-recognition ligands for virus detection. Inrecent years, aptamers have been also investigated as an alter-native of bio-recognition ligands. Aptamers are artificial nucleic

acid ligands that can bind to a wide range of target molecules(Tuerk and Gold, 1990; Jayasena, 1999; Ellington and Szostak, 1990;Huizenga and Szostak, 1995; Zhou et al., 2010; Nadal et al., 2012;Duan et al., 2012), ranging from large molecules such as protein to

ogical

ssgEr1tlbatataAsaoi

sa(tE(a22scgrioh(Htad

asoHwstTdda

2

2

s5(uta

R. Wang et al. / Journal of Virol

imple organic small molecules like ATP, dyes, amino acids orimple small cations, with high affinity and specificity. They areenerated by an in vitro selection process called SELEX (Systematicvolution of Ligands by EXponential enrichment) which was firsteported in 1990 (Ellington and Szostak, 1990; Robertson and Joyce,990; Tuerk and Gold, 1990). Aptamers show a very high affinity forheir targets, with dissociation constants typically in the micromo-ar to low picomolar range, comparable to those monoclonal anti-odies and sometimes even better (Jenison et al., 1994). In addition,s bio-recognition ligands, aptamers possess numerous advan-ages, including small size, rapid and reproducible synthesis, simplend controllable modification to fulfill different diagnostic andherapeutic purposes, slow degradation kinetics, nontoxicity, and

lack of immunogenicity (Duan et al., 2012; Famulok et al., 2007).ptamers have been successfully employed in strip tests as sub-titutes (Wang et al., 2011a,b; Xu et al., 2009) and aptamer-basedssays show great potentiality for in-field applications. A numberf recent reviews have positioned aptamers to make a significantmpact in many areas (Binning et al., 2012; Syed and Pervaiz, 2010).

So far, several high-affinity DNA and RNA aptamers have beenuccessfully selected against viral proteins or whole virus, suchs human immunodeficiency virus HIV glycoprotein 120 (gp120)Zhou et al., 2009), human hepatitis B virus polymerase (P pro-ein) (Feng et al., 2011), hepatitis C virus envelope glycoprotein2 (Chen et al., 2009), and whole human cytomegalo virus and RSVKonopka et al., 2000). A number of aptamers have been isolatedgainst influenza A HA that inhibit viral infectivity (Dhar and Datta,009; Jeon et al., 2004; Gopinath et al., 2006a; Misono and Kumar,005; Cheng et al., 2008; Park et al., 2011). Jeon et al. (2004) haveelected a DNA aptamer that prevents influenza infection by effi-iently blocking the receptor binding region of the viral HA. Kumar’sroup also reported an RNA aptamer that discriminates the closelyelated HA derived from the H3N2 strain of influenza A virus andnhibits membrane fusion (Gopinath et al., 2006a). They also devel-ped an efficient RNA aptamer against human influenza B virusemagglutinin (Gopinath et al., 2006b). In the studies by Cheng et al.2008) and Park et al. (2011), DNA and RNA aptamers targeted toA1 proteins of influenza virus H5 subtypes were selected, respec-

ively. Their studies focused on potent inhibition of influenza virus,nd therefore, their selected aptamers may be promising candi-ates for treatment and prophylaxis of influenza virus infection.

The objective of this study was to select and characterize DNAptamers that can specifically bind to AIV H5N1, and use theelected aptamer as a bio-recognition ligand in the developmentf aptamer-based detection methods for specific detection of AIV5N1 for in-field or on-site application. Thirteen selection cyclesere performed to obtain the DNA aptamers with high affinity and

pecificity. The best aptamer candidate was evaluated and charac-erized using dot blot assay and SPR (surface plasmon resonance).he results presented in this report provide the first-step in theevelopment of a simple, rapid, robust, sensitive and cost-effectiveetection method based on aptamers, offering a possible viablelternative to current methods for AIV H5N1 detection.

. Materials and methods

.1. Target protein, virus and plasmid

The full-length glycosylated recombinant HA protein of AIVubtype H5N1 (A/Vietnam/1203/04) with a concentration of24 �g/ml was purchased from Protein Science Corporation

Meriden, CT, USA). The protein was produced in insect cellssing the baculovirus expression vector system and purifiedo >90% purity under conditions that preserve its biologicalctivity and tertiary structure. The inactivated AIV H5N1 strain

Methods 189 (2013) 362– 369 363

(A/Chicken/Scotland/59) (GenBank accession number X07826.1)was obtained from USDA-APHIS National Veterinary ServicesLaboratories (NVSL, Ames, IA), and an inactivated recent AsianH5N1 field strain (A/Duck/Guangdong/383/2008) was obtainedfrom College of Veterinary Medicine, South China Agricul-tural University, China. Related tests on the Asian H5N1stain (A/Duck/Guangdong/383/2008) were conducted in theAvian Research Laboratory at College of Veterinary Medicine,South China Agricultural University, China. The non-targetinfluenza viruses used in evaluation tests were the inactivatedAIV subtypes of A/H5N2/PA/chicken/85, A/H5N3/WileyLab/87,A/H5N9/WileyLab/85, A/H1N1/WileyLab/87, A/H2N2/PA/chicken/1117-6/04, A/H7N2/PA/chicken/3779-2/97, A/H9N2/WileyLab/87,which were prepared at Wiley Laboratory at Pennsylvania StateUniversity (University Park, PA). Because of biosecurity require-ment in handling live AIV strains and the biosafety limitation of ourlaboratory condition (biosafety level 2), only inactivated AIV strainswere used in this study. The pGEM-T easy vector was obtained fromPromega (Falls Church, VA).

2.2. DNA library and primers

The oligonucleotide template was synthesized as a single-stranded 115 bases with the following sequence: 5′-CCG GAA TTCCTA ATA CGA CTC – (N)74 – TAT TGA AAA CGC GGC CGC GG –3′ where the central 74 nucleotides represent random oligonu-cleotides based on equal incorporation of A, T, G and C at eachposition. The dsDNAs were obtained by PCR amplification usingForward 5′-CCG GAA TTC CTA ATA CGA CTC-3′ and Reverse 5′-CCG CGG CCG CGT TTT CAA TA-3′ primers. Biotinylated forwardprimer with 5′ biotinylation was for Dot blot hybridization. Reverseprimer with 5′ phosphorylation was used to obtain PCR products forlambda exonuclease digestion to produce ssDNA. Both the libraryand primers were synthesized by Integrated DNA Technologies(Coralville, IA).

2.3. In vitro selection procedure

Aptamer candidates against target AIV H5N1 were selectedusing SELEX technique (Tuerk and Gold, 1990). The detailed pro-tocol for SELEX procedure is described as follows. To excludefilter-binding ssDNA sequences from the pool, the ssDNAs werepassed three times prior to the selection cycle through a pre-wettednitrocellulose acetate membrane (0.45 �m HAWP filter, Millipore,MA, USA) in a filter holder (“pop-top”, a diameter of 13 mm, Milli-pore). For the first cycle of selection, 35.5 �l (1 �g/�l) DNAs wereadded in 114.5 �l binding buffer solution.

The molecules of first selection cycle are:

35.5 �l × 1 �g/�l35441.4 g/mol

= 1 nmol = 6.02 × 1014

(DNA Oligo Molecular weight : 35, 441.4 g/mol). (1)

This randomized pool contained approximately 1014 (14 ordersof magnitude) units DNA species under experimental conditions.

To initiate in vitro selection, ssDNA library was denatured at95 ◦C for 10 min and was allowed to cool down at room tempera-ture for 30 min. Denatured DNAs were incubated with the targetprotein HA/virus for 1 h and 45 min at room temperature in a bind-ing buffer (50 mM Tris–HCl, 25 mM NaCl, 5 mM MgCl2, 10 mM DTT(dithiothreitol), pH 7.5). This reaction mixture was filtered over

a HAWP filter (Millipore, MA) and washed three times with thebinding buffer. ssDNAs that were retained with HA/virus on the fil-ter were eluted twice with elution buffer (0.4 M sodium acetate,5 mM EDTA, 7 M urea; pH 5.5) at 70–80 ◦C for over the course of

364 R. Wang et al. / Journal of Virological

Table 1Concentrations of DNA and HA protein/AIV H5N1 used in selection cycles.

Cycle Target ssDNA pool (�M)

HA protein (�M)1 0.5 6.72 0.25 5.03 0.125 5.04 0.063 5.0

H5N1virus (�l)5 30 1.76 25 1.77 20 1.18 15 1.19 10 1.1

10 5 0.9

5oge9liGsemp

2s

sdo1oTt1ip(fC

a11DlS1

2

wGtWwG

11 2 0.912 1 0.813 0.5 0.8

min. Afterwards the eluted DNA was diluted with an equal volumef deionized water (dH2O) and was precipitated (0.12 mg glyco-en, equal volume of 7.5 M ammonium acetate and 1 ml of 100%thanol) and incubated for 1 h at −80 ◦C. After centrifugation at447 × g at 4 ◦C for 1 h, supernatant was discarded and the pel-

et was washed twice with 75% ethanol solution and resuspendedn dH2O. Selected ssDNAs were amplified by PCR (Mastercyclerradient, Eppendorf) and used for the next round of selection. Sub-equent selection cycles were the same as the earlier ones with thexception that the stringency of selection was increased to pro-ote competition between binding species. Therefore, as the cycles

rogress, the molar ration of HA/virus to DNA increased (Table 1).

.4. PCR amplification and conversion of oligonucleotides intosDNA

The recovered ssDNA pools were used as a template in theubsequent PCR reaction using a thermocycler (Mastercycler Gra-ient, Eppendorf). The reaction was carried out in a total volumef 50 �l of ThermoPol reaction buffer (20 mM Tris–HCl, pH 8.8,0 mM (NH4)2SO4, 10 mM KCl, 2 mM MgSO4) containing 0.4 �Mf each primer (forward and reverse), 200 �M dNTPs, and 2.5 U ofaq polymerase. The amplification conditions were as follows: ini-ial denaturation at 95 ◦C for 5 min and final extension at 72 ◦C for0 min; and 30 cycles of denaturation at 94 ◦C for 45 s; anneal-

ng at 64 ◦C for 45 s and extension at 72 ◦C for 45 s. Amplifiedroducts were analyzed on non-denaturing 6% Tris–borate EDTATBE) – polyacrylamide gels (Invitrogen, Carlsbad, CA) at 200 Vor 20 min after binding with SYBR Green 1 (Invitrogen, Carlsbad,A).

The PCR products were converted into ssDNA by incubation in total volume of 50 �l of lambda exonuclease reaction buffer with0 U of lambda exonuclease (New England Biolabs, Ipswich, MA) for

h at 37 ◦C, followed by 10 min at 75 ◦C to inactivate the enzyme.igested products were precipitated as above (resuspension of pel-

et in binding buffer not in water) and used for the next round ofELEX. The concentration of ssDNA was measured using NanoDrop000 Spectrophotometer (ThermoScientific, Wilmington, DE).

.5. Cloning and sequencing

After 13 rounds of selection cycle, the PCR product poolas purified using Qiaquick PCR Purification Kit (Qiagen, Hilden,ermany) and then was cloned into the pGEM-T easy vec-

ors according to the manufacturer’s manual (Promega, Madison,I). Twenty colonies were randomly picked. Plasmid DNAas extracted using the QIAprep Miniprep kit (Qiagen, Hilden,ermany). Ampicillin and IPTG were purchased from Calbiochem

Methods 189 (2013) 362– 369

(San Diego, CA), and X-Gal (5-bromo-4-chloro-3-indolyl-�-D-galactoside) and S.O.C medium were purchased from Invitrogen(Invitrogen, Carlsbad, CA). The plasmids were analyzed by Nano-Drop 1000 Spectrophotometer (ThermoScientific, Wilmington, DE)and then were sent for sequencing.

Sequencing of plasmid DNA of the selected transformants wasdone by automated DNA sequencing using ABI 3130xl analyzerBigDye 3.1 diemistry (ABI 7300 Sequence Detector, Foster City, CA).

Secondary structures of sequenced aptamers were predicted byweb-based UNAFold software in OligoAnalyzer 3.1 program fromIDT (Integrated DNA Technologies, Coralville, IA), which was basedon free energy minimization algorithm. Sequences were alignedusing ClustalX v.1.81 (Thompson et al., 1997) and pattern analysiswas performed using ABI software (Applied Biosystems, Foster City,CA).

2.6. Dot blot analysis with aptamers and antibodies

Dot blot assay (Vivekananda and Kiel, 2006) was used for a rapidanalysis of the affinity of aptamers from selected SELEX rounds.Biotinylated forward primer and phosphorylated reverse primerwere used to obtain dsDNA which were further digested by lambdaexonuclease to obtain biotinylated ssDNA for Dot blot analysis.

Target AIV H5N1 (5 �l) with different titers were spotted ontonitrocellulose membrane (BA85 Protran, 0.45 �m, Whatman, USA)and allowed to air dry. Then these samples were followed byblocking with blocking buffer (12.5 g casein; 4.5 g NaCl; 605 mgTris; 100 mg Thimerosal) for 30 min and incubated with biotiny-lated aptamers (2.2 �g/ml) from selected cycles for 45 min atroom temperature. After washing three times with 1× KPL (KPL,Gaithersburg, MD, USA) washing solution (0.002 M imidazole,0.02% Tween 20, 0.5 mM EDTA, 160 mM NaCl), the membrane wasreacted with streptavidin-conjugated alkaline phosphatase (1:500dilution) for 30 min. Excess enzyme was removed by three subse-quent washes with 1× KPL washing solution. Finally, the membranewas coated in BCIP/NBT (5-bromo-4-chloro-3-indoxyl-phosphateand nitrobluetetrazolium) (KPL, Gaithersburg, MD) substrate in thedark for color development. PBS buffer was served as a negativecontrol and monoclonal antibody (mAb) specifically against AIVH5 subtypes (provided by Animal Diagnostic Laboratory at Penn-sylvania State University) with a concentration of 4.4 �g/ml wasemployed in the dot blot analysis to replace the aptamer as a com-parison.

2.7. Surface plasmon resonance

BIAcore 3000 SPR instrument with CM5 chips (BIAcore, Pis-cataway, NJ) were used in all experiments that were conductedat 25 ◦C with a flow rate of 10 �l/min. The running buffer usedfor the experiments was HEPES and 10 mM NaOH (10 �l) wasused for regeneration to remove the bound target protein. Selectedaptamers were immobilized onto the surface of SPR chips usingbiotin-streptavidin method. The SPR chip was first coated withstreptavidin, and then biotin-labeled aptamers with a concen-tration of ∼1 �M was injected for 10 min (10 �l/min, 25 ◦C) andfollowed by injecting biotin (10 �M) for 5 min and ethanolamine(1 M, pH 8.5) 7 min as blocking reagents. The channel with strep-tavidin coating and ethanolamine blocking, but without aptamerimmobilization, was used as a reference channel.

For binding assay, HA proteins with different concentrations (2,5, 10, 20 and 40 �g/ml) were injected onto the sensor chip andthe affinity binding was monitored for 2 min followed by washing

with running buffer. After measuring SPR signals, the associationand dissociation rate constants (Ka and Kd), and equilibrium disso-ciation constants (KD) were calculated by a 1:1 [Langmuir] fittingmodel using the data analysis software (BIA evaluation software).

ogical Methods 189 (2013) 362– 369 365

afwH

2

aMM(oipfmwtttvbDli

2

f

3

3

wlipwg(iutfiaTsw

Stwp5gPpeeql

R. Wang et al. / Journal of Virol

Further experiments were conducted to study the bindingffinity between the selected aptamers and the H5N1 virus with dif-erent titers of 0.064, 0.128, 0.32 and 0.64 HAU. Then the specificityas investigated by testing non-target AIV H5N2, H5N9, H9N2,7N2 and H2N2 with a titer of 0.64 HAU.

.8. Swab sample preparation

Two tracheal swabs were taken from two healthy chickensnd then placed in per tube containing 3 ml of Viral Transportedia (VTM). The VTM included 500 ml of Minimum Essentialedia (MEM), 7.5 ml of HEPES buffer (1 M), 10 ml of Gentamicin

10 mg/ml), 2.5 ml of Kanamycin (10 mg/ml), 5 ml of Antibi-tic/Antimycotic (Pen-Strep-Amp), and 5 ml of Horse Serum (heatnactivated at 56 ◦C for 30 min). Four tubes of tracheal swab sam-les were mixed together to prepare the uniform swab sample forurther tests. First, each tube of VTM solution with two swabs was

ixed sufficiently using a vortex mixer. After mixing, each swabas pressed against the tube wall several times to squeeze the solu-

ion in swabs and then discarded. Four tubes of the solution werehen put into one tube and mixed sufficiently. Finally, the solu-ion was filtered using syringe filter (0.45 �m) and spiked with AIirus for further use. Virus titers of inactivated AIVs were providedy USDA-APHIS National Veterinary Services Laboratories, Animaliagnostic Laboratory at Pennsylvania State University and the Col-

ege of Veterinary Medicine at South China Agricultural Universityn China. Swab solution without spiking was used as a control.

.9. Chemicals

All chemicals used in this study unless specified were purchasedrom Sigma–Aldrich (St. Louis, MI).

. Results

.1. Selection of H5N1-biniding aptamers by SELEX

In the first selection cycle, about 1014 DNA random sequencesere incubated with target protein. A commercially available full-

ength glycosylated recombinant hemagglutinin (HA) protein ofnfluenza A virus H5N1 (A/Vietnam/1203/04) was used as targetrotein. Started from the fifth cycle, entire inactivated H5N1 virusas applied in order to improve the specificity. As the cycles pro-

ressed, the molar ratio of target HA/virus to DNA was increasedTable 1) to promote competition between binding species. Follow-ng incubation, the bound oligonucleotides were separated fromnbound oligonucleotides via nitrocellulose filtration. Then, thearget-bound aptamers were eluted from the filters, and ampli-ed by PCR. All DNAs were quantified using spectrophotometernd analyzed with PAGE (polyacrylamide gel electrophoresis) gel.he observation of a 115-bp band on the gel after each round ofelection and PCR amplification suggested that the target HA/virusere able to bind to a pool of aptamer sequences.

Single-stranded DNA (ssDNA) templates are required during theELEX and the preparation of ssDNA is an essential and impor-ant step. In this study, the lambda exonuclease digestion methodas used for generation of ssDNA. Lambda exonuclease is a highlyrocessive 5′–3′ exodeoxyribonuclease that selectively digest the′-phosphorylated strand of dsDNA. Therefore, a 5′-phosphorylatedroup was introduced into one strand of dsDNA by performingCR where only the reverse primer was 5′-phosphorylated. Thishosphorylated strand was removed by digestion with lambda

xonuclease while the other strand was retained. After lambdaxonuclease digestion, phenol/chloroform exaction and subse-uent ethanol precipitation were performed to eliminate the extra

ambda exonuclease from the aptamer pool. Finally, the generated

Fig. 1. Dot blot analysis of affinity binding between aptamer pools from the 4th,8th and 10th selection cycles and target AIV H5N1, and the cross-reactivity withnon-target AIV H5N2 and H7N2.

ssDNA was used as the input for the next selection cycle. A total ofthirteen repeated separation-amplification cycles were completedin order to yield high affinity and specificity DNA aptamers.

3.2. Binding affinity and specificity of aptamer pools followingselection cycle

After two or three rounds of selection, the aptamer pools werechecked for binding affinity and specificity to the target AIV H5N1using dot blot assay for a rapid assessment. In this format, targetAIV H5N1 was spotted onto nitrocellulose membrane and air driedat room temperature. PBS buffer without target virus was used as acontrol. To test the binding specificity of the aptamer pools, paralleldot blot assay was performed using non-target AIV subtypes H5N2and H7N2. The results of dot blot assay for the aptamer pool at 4th,8th and 10th cycles are presented in Fig. 2. For the 4th cycle aptamerpool, the target AIV H5N1 with a titer of 25.6 HAU developed a cleardot; however, the titer of 12.8 HAU only displayed a faint dot. Forthe aptamer pools of the 8th and 10th cycles, the stronger dotswere observed for the H5N1 even with the titer of 12.8 HAU, andthe bigger size of dots was observed at the 10th cycle than that ofthe 8th cycle. The dot blot results (Fig. 1) clearly showed that withthe increase of SELEX cycles, the selected aptamer pools displayedstronger binding to the target AIV H5N1 with the increase of bothdot size and intensity. No cross-reaction was observed with non-target AIV subtypes H5N2 and H7N2, indicating good specificitytoward the target virus.

3.3. Cloning and sequence analysis

Twenty clones were sequenced, since there were some repeatsequences and wrong matches, from which three unique sequenceswere obtained as the follows: (1) 5′-GAC GGG TAA CGT ATG TTTTAC ATT ACG AAA TTT AGA GCA CCC TTA CAG CGA GAC TCG TTGACC TGT AGC AGT G-3′, (2) 5′-GTG TGC ATG GAT AGC ACG TAACGG TGT AGT AGA TAC GTG CGG GTA GGA AGA AAG GGA AATAGT TGT CCT GTT G-3′, and (3) 5′-GGC CGA ATT GGT TCG TCGAGC GAG TCA CAC CAA CAA TGC TGC GAT AGA AAC TTC GTACGA GCT TTC TTA CGC TG-3′. Among these three sequences, onlysequence (3) has 74-nucleotide sequence while the other two have73-nucleotide sequences. The one base was probably missing dur-ing the PCR amplification process. According to the mutagenic PCR

theory (Cadwell and Joyce, 1994), the error rate of Taq polymeraseis high in the range 0.1 × 10−4 to 2 × 10−4 per nucleotide per pass ofthe polymerase. Secondary structures of the sequenced aptamers(as shown in Fig. 2) were predicted by web-based UNAFold software

366 R. Wang et al. / Journal of Virological Methods 189 (2013) 362– 369

Fig. 2. The predicted secondary structures of the three aptamers: (a) aptamer sequence (1); (b) aptamer sequence (2); and (c) aptamer sequence (3).

R. Wang et al. / Journal of Virological Methods 189 (2013) 362– 369 367

Table 2Dot blot analysis for the comparison study between aptamer sequence (2) and H5N1monoclonal antibody (anti-H5 mAb): (a) Detection of target H5N1 in the titers ran-ging from 0.0128 to 128 HAU; and (b) specificity test using non-target AIV subtypes.

AIV H5N1 titer (HAU)

128 12.8 1.28 0.128 0.0128

(a)Anti-H5 mAb ++ ++ + − −Aptamer sequence (2) ++ ++ + − −

Specificity test

H5N1 H5N2 H5N3 H5N9 H7N2

(b)

igT−sohlnTattd2t

3

cb

wrtcaHsmw

ctadswiTsada(a

n

Anti-H5 mAb ++ ++ + ++ −Aptamer sequence (2) ++ − − − −

n OligoAnalyzer 3.1 program from IDT (Integrated DNA Technolo-ies), which was based on free energy minimization algorithm.he �G value for the three structures was −5.03, −7.03 and6.24 kcal/mole, respectively; and the GC contents for the three

equences were 45.2%, 47.9% and 51.4%, respectively. The sec-ndary structure of sequence (1) had one external loop and threeairpin loops. The structure of sequence (2) contained one external

oop and two hairpin loops. Structure of sequence (3) had one exter-al loop, one interior loop, one bulge loop and two hairpin loops.hese secondary structures of the nucleotides such as hairpin loopsnd bulge loops played an important role of contacting and bindinghe target (Zhou et al., 2010). The binding between aptamers andheir targets relies on aptamers’ secondary structures and three-imensional (3D) structures (Zhou et al., 2010, 2011; Song et al.,008), hydrogen bonding, electrostatic and hydrophobic interac-ions rather than Watson–Crick base pairing (Citartan et al., 2012).

.4. Evaluation and characterization of DNA aptamers

The preliminary evaluation of the obtained three aptamers wasarried out using dot blot assay to check the binding-interactionetween the selected aptamers and the target AIV H5N1.

Non-target AIV subtypes H5N3, H5N9, H7N2, H2N2 and H9N2ith a titer of 25.6 HAU were also tested to check any cross-

eactivity. The results showed that there is no binding betweenhe selected three aptamers and the non-target AIV subtypes, indi-ating good specificity of the aptamers. Among the three aptamers,ptamer sequence (2) showed the best binding affinity to the target5N1 virus, displaying the largest dot size and the highest inten-

ity, which was employed for the comparison study with H5N1onoclonal antibody and then further used for characterizationith surface plasma resonance.

Table 2 shows the results of dot blot using both H5N1 mono-lonal antibody (anti-H5 mAb) and aptamer sequence (2) for thearget AIV H5N1. Table 2(a) indicates that both anti-H5 mAb andptamer sequence (2) could detect H5N1 with dot blot, and theetection limit was observed to be 1.28 HAU for both ligands. Thepecificity of the aptamer for detection of H5N1 was also comparedith anti-H5 mAb by testing other non-target AIV subtypes includ-

ng H5N2, H5N3, H5N9 and H7N2, and the results are presented inable 2(b). It was found that the developed aptamer (2) had greaterpecificity to H5N1, and other AIV subtypes (H5N2, H5N3, H5N9,nd H7N2) did not bind with it. The anti-H5 mAb displayed visibleots for all the tested H5 subtypes (H5N1, H5N2, H5N3, and H5N9)nd no cross-reactivity was observed for other non-H5. subtype

H7N2). The result indicated that the selected aptamer was specificgainst AIV H5N1.

Surface plasma resonance (SPR) is one of the most sensitive tech-iques particularly suitable for study the surface binding reaction

Fig. 3. An overlay of the SPR signals from the interaction between HA protein andthe selected aptamer.

of molecules. In this study, SPR was used to measure the affin-ity of the aptamer sequence (2) against recombinant HA proteinof the subtype H5N1 and AIV H5N1 virus. Immobilization of theselected aptamer on SPR chip surface was carried out using thebiotin-streptavidin method, which resulted in increasing of 997RU (resonance units) SPR signal, indicating success of the aptamerimmobilization. Then, HA proteins with different concentrations(2, 5, 10, 20 and 40 �g/ml) were injected onto the sensor chip andthe affinity binding was monitored. The analytical signal, recordedin resonance units (RUs), was computed as the difference betweenthe aptamer and corresponding reference channel. We evaluatedthe affinity binding of HA protein to the selected aptamer by inject-ing HA with the concentrations from low to high over the aptamerand reference surfaces. In this way, the portion of the SPR signalthat was attributed to specific affinity binding could be calculatedby subtracting the non-specific binding and bulk refractive indexeffects detected on the reference surface from the total SPR signalmeasured on the aptamer surface. An overlay of the SPR sensor sig-nals from the interactions between HA and the selected aptamer isshown in Fig. 3. As expected, increasing HA concentration resultedin an increase of the SPR signal. The calculated Ka (associationrate constants) and Kd (dissociation rate constants) was 4.69 × 104

(M−1 s−1) and 2.18 × 10−4 (s−1), respectively, and the KD (dissoci-ation constants) was 4.65 nM, indicating strong binding betweenthe HA protein and the selected aptamer.

Further experiments were conducted to study the bindingbetween the selected aptamer and the H5N1 virus. A linear relation-ship was found in the virus titer range from 0.064 to 0.64 HAU andthe corresponding equation was described as: y = 208.39x + 2.23(R2 = 0.99), where y is SPR signal in RU and x is virus titer in HAU.The result showed that the selected aptamer could be used as arecognition ligand to detect target AIV H5N1.

3.5. Detection of AIV H5N1in poultry swab samples using thedeveloped aptamer

Dot blot assay was applied further to the detection of trachealswab samples experimentally spiked with inactivated AIV H5N1viruses (strain A/Chicken/Scotland/59 and A/Duck/Guangdong/383/2008), and other subtypes (A/H5N2/PA/chicken/85, A/H5N3/

WileyLab/87, A/H5N9/WileyLab/85, A/H1N1/WileyLab/87, A/H2N2/PA/chicken/1117-6/04, A/H7N2/PA/chicken/3779-2/97, andA/H9N2/WileyLab/87) using the developed aptamer as recognitionagent. AIV H5N1 strain A/Duck/Guangdong/383/2008 belongs to

368 R. Wang et al. / Journal of Virological Methods 189 (2013) 362– 369

Fig. 4. Dot blot assay for detection of tracheal swab samples experimentally spiked with AIV: (a) AIV H5N1 strain A/Chicken/Scotland/59 with the titer range from 0.0128H ange

H

c2frnAiH

4

aatSeaimtrgafb

Acknowledgments

AU to 64 HAU; (b) AIV H5N1 strain A/Duck/Guangdong/383/2008 with the titer r2N2, H9N2 and H1N1).

lade 2.3.2, which have been circulating widely in China since008 and may cause a new wave of cross-continental spreadingrom Asia to Europe (Sun et al., 2011). As shown in Fig. 4, theesults indicated that the developed aptamer was able to recog-ize target AIV H5N1 (both strains A/Chicken/Scotland/59 and/Duck/Guangdong/383/2008) in chicken swab samples and no

nterference was observed from other AIV subtypes, such as H5N9,7N2, H2N2, H9N2 and H1N1.

. Discussion

The development of aptamers that display good binding affinitynd high specificity against AIV H5N1 suggests that the selectedptamer has great potential to be a viable alternative for detec-ion and screening of AIV H5N1 for in-field or on-site application.ince aptamers can be chemically synthesized in an easy andconomical manner, manufacturing and assay costs of aptamersre expected to be significantly lower than the antibody-basedmmunoassays currently used for rapid AIV detection. It is esti-

ated that the production cost of aptamers could be about 10–50imes less than antibodies (Low et al., 2009). In addition, theesults from this study indicated that the selected aptamer had

reat specificity for AIV H5N1, whereas the anti-H5 monoclonalntibody showed binding-interaction not only for H5N1, but alsoor other H5 AIV subtypes. The aptamer-based assay would likelye less prone to false positive results in AIV H5N1 detection.

from 1 HAU to 64 HAU; (c) specificity test with other AIV subtypes (H5N9, H7N2,

Moreover, the thermal stability of DNA aptamers and their tol-erance of harsh chemical, physical and biological conditions, ascompared to antibodies, add attractive advantages for the devel-opment of aptamer-based assay for AIV detection, specifically forin-field use. Aptamers can also readily be labeled with reporterand quencher moieties to form molecular beacons and functionalgroups for immobilization on microarrays, facilitating the con-struction of a biosensor which represents great potential for thedevelopment of aptamer-based detection technologies. The selec-tion of these sensitive and specific aptamers is the first step indeveloping a simply, rapid, robust, cost-effective and reliable detec-tion method.

In summary, DNA aptamers that target specifically AIV H5N1were generated using SELEX technology in the study. Resultsshowed that the selected aptamer displayed strong binding affinityand high specificity against target AIV H5N1. Future research willfocus on the development of a new detection technology using theselected aptamer as a bio-recognition ligand for in-field or on-sitedetection of AIV H5N1.

This research was supported in part by National Science Foun-dation (NSF/STTR #0932661) and Arkansas Biosciences Institute(ABI).

ogical

R

B

CC

C

C

C

D

D

D

E

F

F

G

G

H

J

J

J

K

L

R. Wang et al. / Journal of Virol

eferences

inning, J.M., Leung, D.W., Amarasinghe, G.K., 2012. Aptamers in virology: recentadvances and challenges. Front. Microbiol. 3, 29.

adwell, R.C., Joyce, G.F., 1994. Mutagenic PCR. Genome Res. 3, S136–S140.hen, F., Hu, Y., Li, D., Chen, H., Zhang, X.L., 2009. CS-SELEX generates high-affinity

ssDNA aptamers as molecular probes for hepatitis C virus envelope glycoproteinE2. PLoS One 4, e8142.

heng, C., Dong, J., Yao, L., Chen, A., Jia, R., Huan, L., Guo, J., Shu, Y., Zhang, Z., 2008.Potent inhibition of human influenza H5N1 virus by oligonucleotides derivedby SELEX. Biochem. Biophys. Res. Commun. 366, 670–674.

ho, C.H., Woo, M.K., Kim, J.Y., Cheong, S., Lee, C.K., An, S.A., Lim, C.S., Kim, W.J.,2013. Evaluation of five rapid diagnostic kits for influenza A/B virus. J. Virol.Meths. 187, 51–56.

itartan, M., Gopinath, S.C., Tominaga, J., Tan, S.C., Tang, T.H., 2012. Assays foraptamer-based platforms. Biosens. Bioelectron. 34, 1–11.

har, G.S., Datta, A., 2009. Targeting pseudoknots in H5N1 hemagglutinin usingdesigned aptamers. Bioinformation 4, 193–196.

uan, N., Wu, S., Chen, X., Huang, Y., Wang, Z., 2012. Selection and identificationof a DNA aptamer targeted to Vibrio parahemolyticus. J. Agric. Food Chem. 60,4034–4038.

usek, R.J., Hall, J.S., Nashold, S.W., TeSlaa, J.L., Ip, H.S., 2011. Evaluation of Nobuto fil-ter paper strips for the detection of avian influenza virus antibody in waterfowl.Avian Dis. 55, 674–676.

llington, A.D., Szostak, J.W., 1990. In vitro selection of RNA molecules that bindspecific ligands. Nature 346, 818–822.

amulok, M., Hartig, J.S., Mayer, G., 2007. Functional aptamers and aptazymes inbiotechnology, diagnostics, and therapy. Chem. Rev. 107, 3715–3743.

eng, H., Beck, J., Nassal, M., Hu, K.H., 2011. A SELEX-screened aptamer of humanhepatitis B virus RNA encapsidation signal suppresses viral replication. PLoS One6, e27862.

opinath, S.C.B., Misono, T., Kawasaki, K., Mizuno, T., Imai, M., Odagiri, T., Kumar,P.K.R., 2006a. An RNA aptamer that distinguishes between closely relatedhuman influenza viruses and inhibits hemagglutinin-mediated membranefusion. J. Gen. Virol. 87, 479–487.

opinath, S.C.B., Sakamaki, Y., Kawasaki, K., Kumar, P.K.R., 2006b. An efficient RNAaptamer against human influenza B virus hemagglutinin. J. Biochem. (Tokyo)139, 837–846.

uizenga, D.E., Szostak, J.W., 1995. A DNA aptamer that binds Adenosine and ATP.Biochemistry 34, 656–665.

ayasena, S.D., 1999. Aptamers: an emerging class of molecules that rival antibodiesin diagnostics. Clin. Chem. 45, 1628–1650.

enison, R.D., Gill, S.C., Pardi, A., Polisky, B., 1994. High-resolution molecular dis-crimination by RNA. Science 263, 1425–1429.

eon, S.H., Kayhan, B., Ben-Yadidia, T., Arnon, R., 2004. A DNA aptamer preventsinfluenza infection by blocking the receptor binding region of the viral Hemag-

glutinin. J. Biol. Chem. 279, 48410–48419.

onopka, K., Lee, N.S., Rossi, J., Düzgünes , N., 2000. Rev-binding aptamer and CMVpromoter act as decoys to inhibit HIV replication. Gene 255, 235–244.

ow, S.Y., Hill, J.E., Peccia, J., 2009. DNA aptamers bind specifically and selectivelyto (1 → 3)-�-d-glucans. Biochem. Biophys. Res. Commun. 378, 701–705.

Methods 189 (2013) 362– 369 369

Misono, T.S., Kumar, P.K., 2005. Selection of RNA aptamers against human influenzavirus hemagglutinin using surface plasmon resonance. Anal. Biochem. 342 (2),312–317.

Nadal, P., Pinto, A., Svobodova, M., Canela, N., O’Sullivan, C.K., 2012. DNA aptamersagainst the lup an 1 food allergen. PLoS One 7, e35253.

Park, S.Y., Kim, S., Yoon, H., Kim, K.B., Kalme, S.S., Oh, S., Song, C.S., Kim, D.E., 2011.Selection of an antiviral RNA aptamer against hemagglutinin of the subtype H5avian influenza virus. Nucleic Acid Ther. 21, 395–402.

Robertson, D.L., Joyce, G.F., 1990. Selection in vitro of an RNA enzyme that specifi-cally cleaves single-stranded DNA. Nature 344, 467.

Shortridge, K.F., Zhou, N.N., Guan, Y., Gao, P., Ito, T., Kawaoka, Y., Kodihalli, S., Krauss,S., Markwell, D., Murti, K.G., Norwood, M., Senne, D., Sims, L., Takada, A., Webster,R.G., 1998. Characterization of avian H5N1 influenza viruses from poultry inHong Kong. Virology 252, 331–342.

Song, S., Wang, L., Li, J., Zhao, J., Fan, C., 2008. Apatmer-based biosensors. TrendsAnal. Chem. 27, 108–117.

Sun, H., Jiao, P., Jia, B., Xu, C., Wei, L., Shan, F., Luo, K., Xin, C., Zhang, K., Liao, M., 2011.Pathogenicity in quails and mice of H5N1 highly pathogenic avian influenzaviruses isolated from ducks. Vet. Microbiol. 152, 258–265.

Syed, M.A., Pervaiz, S., 2010. Advances in aptamers. Oligonucleotides 20 (5),215–224.

Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F., Higgins, D.G., 1997. TheClustalX windows interface: flexible strategies for multiple sequence alignmentaided by quality analysis tools. Nucleic Acids Res. 24, 4876–4882.

Tuerk, C., Gold, L., 1990. Systematic evolution of ligands by exponential enrich-ment: RNA ligands to bacteriophage T4 DNA polymerase. Science 249 (4968),505–510.

Vivekananda, J., Kiel, J.L., 2006. Anti-Francisellatularensis DNA aptamers detecttularemia antigen from different subspecies by Aptamer-Linked ImmobilizedSorbent Assay. Lab. Invest. 86, 610–618.

Wang, L., Chen, W., Ma, W., Liu, L., Ma, W., Zhao, Y., Zhu, Y., Xu, L., Kuang, H., Xu,C., 2011a. Fluorescent strip sensor for rapid determination of toxins. Chem.Commun. (Camb.) 47, 1574–1576.

Wang, L., Ma, W., Chen, W., Liu, L., Ma, W., Zhu, Y., Xu, L., Kuang, H., Xu, C.,2011b. An aptamer-based chromatographic strip assay for sensitive toxin semi-quantitative detection. Biosens. Bioelectron. 26, 3059–3062.

WHO (World Health Organization). Cumulative Number of Confirmed HumanCases of Avian Influenza A/(H5N1) Reported to WHO, 2003–2012.http://www.who.int/influenza/human animal interface/EN GIP 20120706.CumulativeNumberH5N1cases.pdf (accessed 06.06.12).

Xu, H., Mao, X., Zeng, Q., Wang, S., Kawde, A.N., Liu, G., 2009. Aptamer-functionalizedgold nanoparticles as probes in a dry-reagent strip biosensor for protein analysis.Anal. Chem. 81, 669–675.

Zhou, J., Swiderski, P., Li, H., Zhang, J., Neff, C.P., Akkina, R., Rossi, J.J., 2009. Selection,characterization and application of new RNA HIV gp 120 aptamers for faciledelivery of Dicer substrate siRNAs into HIV infected cells. Nucleic Acids Res. 37,

3094–3109.

Zhou, J., Battig, M.R., Wang, Y., 2010. Apatmer-based molecular recognition forbiosensor development. Anal. Bioanal. Chem. 398, 2471–2480.

Zhou, J., Soontornworajit, B., Snipes, M.P., Wang, Y., 2011. Structural prediction andbinding analysis of hybridized aptamers. J. Mol. Recognit. 24, 119–126.