BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, researchlibraries, and research funders in the common goal of maximizing access to critical research.

H5 Antibody Detection by Blocking Enzyme-Linked Immunosorbent Assay Usinga Monoclonal AntibodyAuthor(s): Yi-Cheng Chen, Chien-Hao Chen, Ching-Ho WangSource: Avian Diseases, 52(1):124-129. 2008.Published By: American Association of Avian PathologistsDOI: http://dx.doi.org/10.1637/8076-071807-RegURL: http://www.bioone.org/doi/full/10.1637/8076-071807-Reg

BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the biological, ecological, andenvironmental sciences. BioOne provides a sustainable online platform for over 170 journals and books publishedby nonprofit societies, associations, museums, institutions, and presses.

Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance ofBioOne’s Terms of Use, available at www.bioone.org/page/terms_of_use.

Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercialinquiries or rights and permissions requests should be directed to the individual publisher as copyright holder.

H5 Antibody Detection by Blocking Enzyme-Linked Immunosorbent AssayUsing a Monoclonal Antibody

Yi-Cheng Chen, Chien-Hao Chen, and Ching-Ho WangA

Department of Veterinary Medicine, National Taiwan University, No. 1 Sec. 4, Roosevelt Road, Taipei 10617, Taiwan

Received 21 July 2007; Accepted and published ahead of print 26 November 2007

SUMMARY. Many commercial enzyme-linked immunosorbent assays (ELISAs) are unable to differentiate antibody responsesto different avian influenza virus (AIV) subtypes. Developing an ELISA for specifically detecting the H5 antibody is the purpose ofthis study. Four monoclonal antibodies (Mabs) were raised using A/duck/Yunlin/04 (H5N2). They were confirmed as beingspecific to H5. Two of these antibodies showed hemagglutination inhibition (HI) activity using the HI test. Using immunodot blotassays, three Mabs recognized both Eurasian and American H5, whereas the other Mab recognized only the tested Eurasian H5virus. When testing denatured H5 antigen, one of the Mabs lost its antigen binding activity using Western blotting. For detectingthe H5 humoral response in serum, one monoclonal antibody was purified and labeled with horseradish peroxidase to set up ablocking ELISA. Chicken sera that blocked H5 Mab binding by .29% were considered H5 antibody positive. Inhibitionpercentages for sera from chickens infected with other AIV subtypes, H1 to H15, were ,29%. This blocking ELISA was used for478 field chicken serum samples. The results showed that the sensitivity and specificity of this ELISA were 98.3% (232/236) and95.9% (232/242), respectively. This blocking ELISA could be used specifically for detecting the H5 humoral responses in chickens.

RESUMEN. Deteccion de anticuerpos contra H5 mediante inmunoensayo asociado a enzimas de bloqueo utilizando unanticuerpo monoclonal.

Muchas pruebas de inmunoensayo asociado a enzimas (ELISA) no pueden diferenciar las respuestas de anticuerpos contra losdiferentes subtipos del virus de influenza aviar. El proposito de este estudio fue desarrollar una prueba ELISA para la deteccionespecıfica de anticuerpos contra H5. Utilizando el virus A/pato/Yunlin/04, se desarrollaron cuatro anticuerpos monoclonales. Losanticuerpos se confirmaron especıficos para H5. Utilizando la prueba de inhibicion de la hemoaglutinacion, dos de estosanticuerpos demostraron su capacidad para inhibir la hemoaglutinacion. Utilizando pruebas de inmunotransferencia puntual, tresde los anticuerpos monoclonales reconocieron tanto a los virus H5 Euroasiaticos como a los Americanos, mientras que el otroanticuerpo monoclonal solo reconocio los virus H5 Euroasiaticos. Cuando se analizo el antıgeno H5 desnaturalizado, uno de losanticuerpos monoclonales perdio la capacidad de union antigenica demostrada en la prueba de inmunotransferencia puntual. Parala deteccion de la respuesta inmune humoral en suero, se purifico un anticuerpo monoclonal y se marco con peroxidasa de rabanopicante para establecer una prueba de ELISA de bloqueo. El suero de las aves que bloqueo mas del 29% de la union del anticuerpomonoclonal con el H5 se considero positivo para H5. Los porcentajes de inhibicion para los sueros de aves con otros subtipos delvirus de influenza aviar (H1 a H15) fueron menores del 29%. Esta prueba ELISA de bloqueo se utilizo en 478 muestras de campode suero de pollo. Los resultados demostraron que la sensibilidad y especificidad de esta prueba fue del 98.3% (232/236) y 95.9%(232/242), respectivamente. Esta prueba ELISA de bloqueo puede ser utilizada para detectar la respuesta humoral especıfica contravirus H5 de influenza aviar en pollos.

Key words: avian influenza virus, blocking ELISA, H5 antibody

Abbreviations: AHRI 5 Animal Health Research Institute; AI 5 avian influenza; AIV 5 avian influenza virus; ELISA 5 enzyme-linked immunosorbent assay; H 5 hemagglutinin subtype; HA 5 hemagglutinin; HI 5 hemagglutination inhibition; HPAIV 5highly pathogenic avian influenza virus; LP 5 low pathogenicity; LPAIV 5 low pathogenic avian influenza virus; Mab 5 mono-clonal antibody; N 5 neuraminidase subtype; NA 5 neuraminidase; NC 5 nitrocellulose; O.D. 5 optical density; PBS 5 phos-phate-buffered saline; PBST 5 phosphate-buffered saline + 0.1% Tween 80; SDS 5 sodium dodecyl sulfate; SDS-PAGE 5 sodiumdodecyl sulfate-polyacrylamide gel electrophoresis; SPF 5 specific-pathogen-free; TMB 5 3,39,5,59-tetramethylbenzidine dihydro-chloride

Avian influenza (AI) is caused by an A type influenza virus thatbelongs to the family Orthomyxoviridae, genus Influenza virus A (12).Many bird species have been shown susceptible to the AI virus (AIV)(1). AIV is classified into subtypes based on antigenic differences intheir surface glycoproteins hemagglutinin (HA) and neuraminidase(NA) (2). To date, 16 HA subtypes (H1–H16) and nine NAsubtypes (N1–N9) have been recognized (4,10,15). AIV can becategorized as having high pathogenicity and low pathogenicity (LP)based on the severity of the clinical signs and mortality rate inexperimentally inoculated chickens (13). The highly pathogenicavian influenza virus (HPAIV) usually results in a systemic disease,with high morbidity and mortality rates. Until now, only certain H5and H7 isolates have been characterized as highly pathogenic inchickens (2,13). Low pathogenic avian influenza virus (LPAIV) is

more commonly isolated from domestic poultry, and it may beassociated with mild respiratory diseases and reduction in eggproduction (15).

The traditional methods for AIV serologic surveillance are theagar gel precipitation test, enzyme-linked immunosorbent assay(ELISA), and hemagglutination inhibition (HI) test (12). The HItest is considered the ‘‘gold standard’’ for AI antibody subtyping,and it remains a World Organization for Animal Health-approvedmethod (9). Although some reports described ELISAs for detectingsubtype-specific antibodies (3,11), commercial ELISA kits detectAIV common type antibodies but not antibodies from differentsubtypes (2). In this study, we produced anti-H5 monoclonalantibodies (Mabs) for specific H5 antibody detection using ablocking ELISA. This assay is sensitive and specific for antibodydetection in chicken sera infected with an American H5 strain,A/chicken/Taiwan/1209/03.ACorresponding author. E-mail: chingho@ntu.edu.tw

AVIAN DISEASES 52:124–129, 2008

124

MATERIALS AND METHODS

Viruses. Strain A/duck/Yunlin/04 (a Eurasian strain, H5N2) wasobtained in January 2004 at a duck farm in Taiwan, and it ischaracterized as LPAIV. This strain was used for HA gene cloning in theproduction of Mabs, HI test antigen, and blocking ELISA coatingantigen. This virus and A/chicken/Taiwan/1209/03 (an Americanlineage strain, H5N2, GenBank accession no. AY573917), A/chicken/Taiwan/2838V/00 (H6N1, GenBank accession no. EF681878), and A/chicken/Taiwan/2838N/00 (H6N1, GenBank accession no. EF681870)(6,7,14) were used for immunodot blot assay and western blot assay.The sequence percentage identity between A/duck/Yunlin/04 andH5N1 occurring recently in Southeast Asian countries, such as A/chicken/Indonesia/Soppeng/1631/71/07, reached 82% but between A/chicken/Taiwan/1209/03 and recent H5N1 reached only 75%. Thereason this strain was selected is that it is closer to recent Asian H5N1strains than A/chicken/Taiwan/1209/03, because the latter belongs tothe American lineage (8).

AIV concentration and purification. The viruses used in this studywere propagated in specific-pathogen-free (SPF) embryonated eggs. Fivehundred milliliters of allantoic fluid harvested 7 days subsequent to AIVinoculation was frozen and thawed three times. The virus solution wasclarified by centrifugation at 2000 3 g for 15 min (5804R centrifuge;Eppendorf, Hamburg, Germany) at 4 C and centrifuged at 70,000 3 gfor 2 hr (Avanti J-25 centrifuge; Beckman Coulter, Fullerton, CA). Thepellet was resuspended in 1 ml of Tris-ethylenediaminetetraacetic acidsaline (TEN: 10 mM Tris-base, 1 mM EDTA, and 100 mM NaCl,pH 8.0) buffer. The virus solution was then centrifuged at 150,000 3 gfor 3 hr in a 30% to 50% sucrose gradient. Subsequently, the virus bandwas pelleted by centrifugation at 50,000 rpm for 2 hr (Optima MAX-Eultracentrifuge; Beckman Coulter). The purified virus was resuspendedin 0.3 ml of TEN buffer. The used virus was pure, as indicated by oneclear band in the sucrose gradient. In sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gel, no chicken embryoallantoic fluid proteins were noted in these preparations (data notshown).

HA recombinant protein. A/duck/Yulin/04 viral RNA was extractedfrom allantoic fluid using the TRIzol reagent (Invitrogen, Carlsbad,CA). A partial HA fragment encoding amino acids 72–300 wasamplified with a primer set. This fragment encoded the major HA geneantigenic epitopes according to the H3 sequence (16). The sequences ofthis primer set were 5W1, 59-GAGGATCCTTTGATAATGTGGG-GAATCC-39 (containing BamHI restriction site) and 5W2: 59-TTGAATTCCTTCCCAACAGCCT-39 (containing EcoRI restrictionsite). The amplified DNA fragments were purified using the Gel-M gelextraction system (Viogene, Sunnyvale, CA), according to themanufacturer’s instructions. Subsequently, the purified products werecloned in pRSET B expression vector (Invitrogen) and introduced intoBL21 (DE3) Escherichia coli recipient cell (Invitrogen). The fusionprotein was induced by adding 1mM isopropyl-b-D-galactopyranosidefor 4 hr at 37 C. The bacterial cells were collected by centrifugation at12,000 3 g for 10 min. Then, they were treated with lysozyme andsonicated three times at 15 kHz for 10 sec in ice using a sonicator(Vibra-Cell; Sonics & Materials Inc., Newtown, CT). The pellet wastreated with 4.5 M, 6.0 M, and 8.0 M buffered urea (10 mM Tris-HCl, 0.1 M 2-mercaptoethnol, and 5% glycerol, pH 7.0) aftercentrifugation at 12,000 3 g for 10 min. The greatest protein yieldeluted in 6 M urea solution. Urea was dialyzed in phosphate-bufferedsaline (PBS), pH 8.0, for 3 days at 4 C. The purified protein was thencollected, aliquoted, and stored at 220 C.

Monoclonal antibodies. For Mab production, 8-wk-old BALB/cmice were injected intraperitoneally with 0.1 mg of recombinant HAprotein with complete Freund’s adjuvant, and they were boosted threetimes at day 21, day 35, and day 49 with the concentrated H5N2 strainA/duck/Yulin/04 virion in incomplete Freund’s adjuvant. The use ofmice was approved by the Institutional Animal Care and UseCommittee, National Taiwan University. The mice were then injectedintravenously with 50 mg of strain A/duck/Yulin/04 at day 63. Four daysafter the last injection, the splenocytes in the mice were fused with

myeloma cells (SP2/O). Cell fusion and cloning of positive hybridomaswere performed following the standard procedure (5). Briefly, a weekbefore cell fusion, myeloma cells were grown in RPMI 1640 medium(HyClone Laboratories, Logan, UT) with 10% fetal bovine serum at37 C. The spleen cells from immunized mice were mixed with thepreviously prepared myeloma cells and fused by gradually adding 50%polyethylene glycol-1400. Only the fused cells were grown inhypoxanthine-thymidine medium (Invitrogen). Positive hybridomaselection was based on ELISA and western blot assay results against A/duck/Yunlin/04 virus and HA recombinant protein. Selected Mabs wereanalyzed using the immunodot blot assay, western blot assay, and HItest.

After purification, Mabs were isotyped using the ImmunoPuremonoclonal antibody isotyping kit (Pierce Chemical, Rockford, IL)according to the manufacturer’s procedures. The antigen for capturingMabs was the concentrated A/duck/Yulin/04.

Immunodot blot assay. One hundred microliters of concentratedvirus was dotted onto a 0.45-mm nitrocellulose (NC) membrane(Protran nitrocellulose membrane; Whatman Schleicher and Schuell,Shelton, CT). After drying, the membrane was blocked with 0.25% (wt/vol) gelatin-NET (0.25% gelatin, 0.15 M NaCl, 5 mM EDTA-Na,0.5% Tween 20, and 50 mM Tris) at room temperature for 1 hr. It wasthen reacted with diluted antibody (1:1000 in gelatin-NET) at roomtemperature for 1 hr. After the membrane was washed three times withwashing buffer (0.05% Tween 20 in 1 liter of 13 PBS, pH 7.4), goatanti-mouse-horseradish peroxidase (HRP) conjugate (Kirkegaard andPerry Laboratories, Gaithersburg, MD) (1:2500 dilution) was added andincubated for 1 hr at room temperature. The preparations were thendeveloped using a 3,39,5,59-tetramethylbenzidine dihydrochloridesolution (TMB) peroxidase substrate (TMB component membranesubstrate; Kirkegaard and Perry Laboratories).

Western blot assay. AIV concentrated allantoic fluid was denaturedin sample buffer (125 mM Tris-base, 2 mM EDTA, 2% SDS, and 5%b-mercaptoethanol, pH 6.8) by boiling for 5 min. Denatured proteinswere separated in a 20% SDS-PAGE gel at 100 V for 60 min. Theproteins were then transferred onto an NC membrane (Bio-Rad,Hercules, CA), and the blots were blocked with 10% (wt/vol) skim milkfor 2 hr at room temperature. Anti-H5 Mabs (10003 dilution) wereadded and incubated for 1 hr at room temperature, and then they werewashed three times with washing buffer (13 PBS and 1% Tween 20,pH 7.4). The peroxidase-conjugated goat anti-mouse IgG (1:25003;Jackson ImmunoResearch Laboratories Inc., West Grove, PA) wasadded and incubated for 1 hr at room temperature. After washing, themembrane was developed with TMB (TMB component membranesubstrate; Kirkegaard and Perry Laboratories) for 15 min at roomtemperature. The reaction was stopped by rinsing with distilled water.

HI test. The HI test was performed essentially as described previously(9), using 1% chicken erythrocytes for hemagglutination and 4hemagglutinating units of A/duck/Yunlin/04. An HI titer .24 wasconsidered positive (9). Seven hundred and twelve chicken serumsamples (234 of them were for cut-off value determination) from aslaughterhouse were tested using the HI test. The serum sample originsand chicken ages are described below.

Blocking ELISA. Purified A/duck/Yunlin/04 (H5N2) viral antigendiluted in a coating buffer (15 mM Na2CO3 and 35 mM NaHCO3,pH 9.6) as 200 ng/0.1 ml per well was coated onto plates (Nalge Nunc,Rochester, NY) at room temperature overnight. The plate was washedonce with PBS + 0.1% Tween 80 (PBST; pH 7.4) and subsequentlyblocked with 100 ml of blocking buffer (PBST + 5% skim milk) for30 min at 37 C. The serum was diluted fivefold with blocking buffer.After washing three times with washing buffer, 100 ml of diluted serumwas added and incubated for 30 min at 37 C. The plate was washed fivetimes with PBST and added with 100 ml of horseradish peroxidase-conjugated MabaH5-6 (in house labeling with SureFIRE HRPconjugate kit; Kirkegaard and Perry Laboratories) dilution. The platewas incubated for 30 min at 37 C and then washed five times withwashing buffer. One hundred microliters of TMB (SureBlue TMBmicrowell peroxidase substrate; Kirkegaard and Perry Laboratories) wasadded and incubated in the dark for 10 min. The reaction was stopped

H5 antibody by ELISA 125

by adding 100 ml of TMB stop solution (Kirkegaard and PerryLaboratories) and read in a spectrophotometer (EL312e Bio-Kineticsreader; Bio-Tek Instruments, Winooski, VT) using the filter at 450 nm.The optical density (O.D.) value was used for calculating the inhibitionpercentage, including the positive and negative controls from each plateusing the following formula. Inhibition percentage 5 100% – (O.D. ofsample – O.D. of positive control)/(O.D. of negative control – O.D. ofpositive control).

Two chicken hyperimmune sera to each subtype from H1 to H15AIVs (total 30 sera) provided by Animal Health Research Institute(AHRI, 376, Chung-Cheng Road, Tamsui 251, Taiwan) were used fortesting a cross-reaction of different subtypes using the blocking ELISA.In addition, two sera from two 6-wk-old SPF chickens (AHRI) wereused for the negative control.

Field sera for cut-off value determination, sensitivity, andspecificity of the blocking ELISA. Two hundred and thirty-four seracollected from 12 noninfected Taiwan Country chickens flocks in 2004were used to determine the cut-off value in the blocking ELISA. Thoseflocks were confirmed to be noninfected by virus isolation in chickenembryonating eggs for two passages. In addition, the sera were confirmedto be AI-antibody negative by a commercial AI antibody ELISA (IDEXXLaboratories, Inc., Westbrook, ME) and then by HI test with H5 antigensincluding A/duck/Yunlin/04 and A/chicken/Taiwan/1209/03.

The H5N2 infection occurring in chicken flocks in Taiwan in 2004was caused by an American strain, A/chicken/Taiwan/1209/03 (8).Several other H5N2 viruses were isolated from chicken farms, and theirsequences were similar to A/chicken/Taiwan/1209/03 (data not shown).Thus, these flocks were infected with a single strain, A/chicken/Taiwan/1209/03. The 236 H5-positive sera were collected from these infectedflocks during that outbreak for determining blocking ELISA sensitivity.The sera were from flocks confirmed H5-positive by HI test. Virusisolation from tracheal swabs was also done by chicken embryonatingegg inoculation with negative results when the sera became 100%positive to AI antibody. Nearly 100% of the chickens in a flock showedthe anti-H5 antibody if they were from an infected flock. Thus, the timebetween infection and blood collection was .2 wk because nearly 100%of the chickens showed seroconversion 2–3 wk after infection (2).

Two hundred and forty-two sera collected from noninfected flockswere tested to determine specificity. These flocks were confirmed to benoninfected by virus isolation in chicken embryonating eggs for twopassages. The sera were confirmed AI-antibody negative using ELISAand HI testing with H5 antigens. All chickens used in blood collectionwere 9 to 14 wk old.

RESULTS

Production and characteristics of Mabs. Four Mabs, aH5-1,aH5-3, aH5-4, and aH5-6, were produced and used in this study.

Both A/duck/Yunlin/04 and A/chicken/Taiwan/1209/03 wererecognized by aH5-1, aH5-3, and aH5-4 using immunodot blotassay, but only A/duck/Yunlin/04 was recognized by aH5-6 (Fig. 1).In western blot assay, the HA1 protein of A/duck/Yunlin/04 and A/chicken/Taiwan/1209/03 were recognized using aH5-1 and aH5-4,but only the HA1 protein of A/duck/Yunlin/04 was recognized byaH5-6 (Fig. 2). The band above HA1 may be the precursor HA0.Although the size of the HA1 protein from A/duck/Yunlin/04 wassimilar to that of A/chicken/Taiwan/1209/03, the HA1 of the formercontains 341 amino acids (theoretical 38.7 kDa) and that of thelatter contains 342 (theoretical 38.7 kDa). In spite of this similarity,the HA1 band of A/duck/Yunlin/04 HA1 was lower than that inA/chicken/Taiwan/1209/03 (Fig. 2). They migrated differently inSDS-PAGE. This difference is most likely due to their differingglycosylation patterns. From Vector NTI version 9.0 N-glycosyla-tion analysis (Program protein, analysis, N-glycosylation motif;InforMax Inc., Bethesda, MD), three sites in the HA protein fromA/duck/Yunlin/04 show N-glycosylation potential and four sites inA/chicken/Taiwan/1209/03 HA (data not shown).

In the HI test, two Mabs, aH5-3 (HI titer 5 29) and aH5-6 (HItiter . 212) showed HI activity. Using an antigen-dependentisotyping kit, the Mabs aH5-1, aH5-3, aH5-4, and aH5-6 weretyped to IgG2a, IgG1, IgG2a, and IgG2b, respectively.

Cut-off value determination in blocking ELISA. After reagentand test protocol optimization, samples with known status, regardedas negative to AIV antibodies, were used for deciding the cut-offvalue. For this, 234 H5-negative serum samples from a slaughter-house were tested. These samples had a mean inhibition of 9.4%,with a standard deviation of 9.8%. Therefore, the cut-off value wasset as 29% (mean + 2 SDs) for deciding the serum sample statuswith respect to the AIV antibody.

Experimental sera. According to the detection sera (SPF) andhyperimmune sera (15 subtypes) results using blocking ELISA, theH5 hyperimmune sera inhibited at .90% (Fig. 3), whereas non-H5hyperimmune sera inhibited at ,20%, with 1–6% SD (n 5 2).These results showed that the blocking ELISA could differentiate H5subtype hyperimmune sera from 14 other subtypes.

Chicken sera from the field. The present blocking ELISA,coating with an Eurasian H5 virus (A/duck/Yunlin/04), detectedantibody caused by an American H5 (A/chicken/Taiwan/1209/03),because the field chickens were infected with A/chicken/Taiwan/1209/03 only (as mentioned above).

Testing field chicken sera with both the blocking ELISA and HItest showed a high concordance between the two methods (Table 1).Among the discordant, the four sera showing positive HI and

Fig. 1 Immunodot blot assay using H5 monoclonal antibodies (10003 dilution) aH5-1 (A), aH5-3 (B), aH5-4 (C), and aH5-6 (D). Spot 1:A/duck/Yunlin/04 (H5N2), spot 2: A/chicken/Taiwan/1209/03 (H5N2), spot 3: A/chicken/Taiwan/2838V/00 (H6N1), spot 4: A/chicken/Taiwan/2838N/00 (H6N1), and spot 5: SPF allantoic fluid (negative control). Secondary antibody is HRP-conjugated goat anti-mouse IgG(25003 dilution).

126 Y.-C. Chen et al.

negative ELISA gave HI titer/ELISA inhibition of 26/25%, 27/21%,24/18%, and 24/25% in each serum, respectively. The 10 serashowing negative HI and positive ELISA gave HI titer/ELISAinhibition of 20/39%, 20/74%, 20/31%, 20/45%, 22/54%, 20/47%,

23/49%, 20/30%, 22/33%, and 23/36% in each serum, respectively.The results showed that the sensitivity and specificity of thisELISA were 98.3% (232/236) and 95.9% (232/242), respectively(Table 1).

Fig. 2. Western blot assay by using H5 monoclonal antibodies (10003 dilution) aH5-1 (A), aH5-3 (B), aH5-4 (C), and aH5-6 (D). Lane 1:protein marker, lane 2: AIV A/duck/Yunlin/04 (H5N2), lane 3: AIV A/chicken/Taiwan/1209/03 (H5N2), lane 4: A/chicken/Taiwan/2838V/00(H6N1), lane 5: A/chicken/Taiwan/2838N/00 (H6N1), and lane 6: SPF allantoic fluid (negative control). Primary antibodies used in A, B, C, andD are aH5-1, aH5-3, aH5-4, and aH5-6, respectively. Secondary antibodies used goat anti-mouse IgG5labeled HRP conjugate (25003 dilution).Arrowheads indicate HA1 protein.

Fig. 3. Inhibition percentages of sera to different HA subtypes by blocking ELISA. Two hyperimmune sera to each subtype from H1 to H15and two SPF chicken sera were used for this test. According to the results of the detection of SPF and hyperimmune sera (15 subtypes) usingblocking ELISA, the inhibition percentages of the hyperimmune sera other than H5 subtype are ,20%. The standard deviation of each value rangesfrom 1% to 6% (n 52).

H5 antibody by ELISA 127

DISCUSSION

Active surveillance for the presence of antibodies against AIV inchicken sera is routine in Taiwan. Commercial ELISA kits are usedfor this antibody screening. Positive reactor sera are subtyped usingHI tests. The commercial ELISA kits do not differentiate betweenantibodies from different subtypes, because they detect the type Agroup-specific nucleoprotein and matrix antibodies elicited by allsubtypes (12). Antibodies to some LPAIVs, such as H6 AIV, whichhas occurred in Taiwan (6,7), also react positive to commercialELISA kits. The HI test needs a national laboratory to perform theH5-specific serology tests. In this study, we used an H5 Mab as thekey reagent to develop a blocking ELISA for the detection of specificH5 antibodies. In screening antibodies to AIV, ELISA has severaladvantages over the HI test. ELISA can be automated and used on alarge scale (11). It produces an objective result because it is readusing a chemical reaction rather than naked eye observation as in theHI test. In addition, ELISA could be used for large-scale bloodtesting.

The occurrence of LPAI H5N2 in Taiwan in 2004 adverselyaffected the poultry industry, resulting in the culling of 383,852chickens (8). In addition, media reports caused the public to severelycurtail chicken consumption. The 2004 outbreak in Taiwan wasactually caused by an American lineage H5N2 strain (A/chicken/Taiwan/1209/03) that may have originated from the illegal use of avaccine strain imported from South America (data not shown). Todetect the humoral response from Eurasian H5 AIV infections, thecoating antigen and Mab used in the present ELISA were from anEurasian H5N2 (A/duck/Yulin/04). Even so, this ELISA coulddetect the American H5N2-induced antibodies in chickens frominfected flocks. This infers that this blocking ELISA may detect H5-specific antibodies in a wider range of H5 AIVs.

In accordance with World Organization for Animal Healthrequirements, all isolated H5 AIVs must be reported (9). Thus, theoccurrence of notifiable AI, even LP, is a potential threat to thepoultry industry. The countries surrounding Taiwan have reportedH5N1 HPAI. Communication between Taiwan and other Asiancountries, especially China, occurs continuously. Therefore, sero-logic detection of H5 AIV at the earliest time is important forTaiwan maintaining a disease-free status. In LPAI monitoring,antibody testing is more effective than virus detection becauseantibodies persist after AIV infection in chicken flocks. AIV spreadsquickly in a flock. Within 2 wk, nearly all chickens in the same flockwill show positive for antibodies (2). Thus, for LPAI, identifyingantibodies is easier.

In an H5N2-infected flock in 2004, 10 serum samples were taken1 wk after infection. The number of positive sera inside a flock withthe HI test and blocking ELISA was compared, and the resultsshowed that the latter is more sensitive than the former where HItiter .24 was considered positive (data not shown). One week afterinfection, only 30% (3/10) chickens had positive H5 antibody by HItest, but 70% of them had H5 positive antibody by the blocking

ELISA. Thus, the ELISA test might be earlier to detect H5seroconversion than the HI test.

One Mab, aH5-3 reacted to A/chicken/Taiwan/1209/03 inimmunodot blot assay (Fig. 1) but not in the western blot assay(Fig. 2). This antibody might be conformation-dependant becauseof the HA protein, detected by this antibody, was not denatured inthe former but denatured in the latter. In this study, aH5-6 reactedto A/duck/Yunlin/04 but not to A/chicken/Taiwan/1209/03 inimmunodot blot assay (Fig. 1) or in the western blot assay (Fig. 2).This antibody is specific to A/duck/Yunlin/04, which belongs to theEurasian lineage, because A/chicken/Taiwan/1209/03 belongs to theAmerican lineage. These two viruses are about 20% heterogeneous inHA sequence. This antibody might recognize the epitope, which isdifferent between these two strains.

In practice, the cut-off value was determined by adding 2 SDs fromthe negative mean value, which permits a specificity value of 97.5% ifthe values are normally distributed. The specificity of the proposedmethod is good because there is no cross-reaction between H5 subtypeand other 14 subtypes (Fig. 3). The sensitivity and the specificity ofthis blocking ELISA are good for anti-H5 antibody detection in thefield in Taiwan. Further studies are necessary to determine whetherthis blocking ELISA is effective in the detection of H5-specificantibodies elicited by more geographically diverse H5 AIVs.

REFERENCES

1. Alexander, D. J. A review of avian influenza in different bird species.Vet. Microbiol. 74:3–13. 2000.

2. Alexander, D. J. Orthomyxoviridae—avian influenza. In: Poultrydiseases, 5th ed. F. Jordan, M. Pattison, D. Alexander, and T. Faragher, eds.W. B. Saunders, London, United Kingdom. pp. 281–290. 2001.

3. De Marco, M. A., E. Foni, L. Campitelli, M. Delogu, E. Raffini, C.Chiapponi, G. Barigazzi, P. Cordioli, L. Di Trani, and I. Donatelli.Influenza virus circulation in wild aquatic birds in Italy during H5N2 andH7N1 poultry epidemic periods (1998 to 2000). Avian Pathol. 34:480–485.2005.

4. Fouchier, R. A., V. Munster, A. Wallensten, T. M. Bestebroer, S.Herfst, D. Smith, G. F. Rimmelzwaan, B. Olsen, and A. D. Osterhaus.Characterization of a novel influenza A virus hemagglutintin subtype (H16)obtained from black-headed gulls. J. Virol. 79:2814–2822. 2005.

5. Galfre, G., and C. Milstein. Preparation of monoclonal antibodies,strategies and procedure. Methods Enzymol. 73:3–53. 1981.

6. Hsu, C. N., and C. H. Wang. Sequence comparisons between twoquasi strains of H6N1 with different pathogenicity from a single parentalisolate. J. Microbiol. Immunol. Inf. 39:292–296. 2006.

7. Lee, H. C., C. N. Hsu, and C. H. Wang. Molecular epidemiology ofavian influenza virus H6N1 in Taiwan. Taiwan Vet. J. 31:230–239. 2005.

8. OIE. World Organization for Animal Health. Disease information,avian influenza in Taipei China. [modified 2004 April 15; cited 2007 Nov12]. Available from: http://www.oie.int/downld/AVIAN%20INFLUENZA/A_AI-Asia.htm.

9. OIE. World Organization for Animal Health. Avian influenza. In:Manual of diagnostic tests and vaccines for terrestrial animals. [modified2007 Oct 16; cited 2007 Nov 12]. Available from: http://www.oie.int/eng/normes/mmanual/A_00037.htm.

10. Rohm, C., N. Zhou, J. Suss, J. Mackenzie J, and R. G. Webster.Characterization of a novel influenza hemagglutinin, H15: criteria fordetermination of influenza A subtype. Virology 217:508–516. 1996.

11. Sala, G., P. Cordioli, A. Moreno-Martin, M. Tollis, E. Borcchi, A.Piccirillo, and A. Lavazza. ELISA test for the detection of influenza H7antibodies in avian sera. Avian Dis. 47:1057–1059. 2003.

12. Swayne, D. E., D. A. Senne, and C. W. Beard. Avian influenza. In: Alaboratory manual for the isolation and identification of avian pathogens,4th ed. D. E. Swayne, J. R. Glisson, M. W. Jackwood, J. E. Pearson, and W.M. Reed, eds. American Association of Avian Pathologist, Kennett Square,PA. pp. 150–155. 1998.

Table 1. Comparison of the presences of antibodies to H5 obtainedfrom HI test and blocking ELISA.A

HI test

+ –

Blocking + 232 10ELISA – 4 232

ASera are from Taiwan Country chickens at a slaughterhouse in 2004.The market ages range from 9 wk old to 14 wk old.

128 Y.-C. Chen et al.

13. Swayne, D. E., and D. L. Suarez. Highly pathogenic avian influenza.Rev. Sci. Tech. Off. Int. Epizoot. 19:463–482. 2000.

14. Wang, C. W., and C. H. Wang. Experimental selection ofvirus derivatives with variations in virulence from a single low-pathogenicity H6N1 avian influenza virus field isolate. Avian Dis. 47:1416–1422. 2003.

15. Webster, R. G., W. J. Bean, O. T. Gorman, T. M. Chambers, and Y.Kawaoka. Evolution and ecology of influenza A viruses. Microbiol. Rev.56:152–179. 1992.

16. Wiley, D. C., I. A. Wilson, and J. J. Skehel. Structural identificationof the antibody-binding sites of Hong Kong influenza hemagglutinin andtheir involvement in antigenic variation. Nature 289:373–378. 1981.

ACKNOWLEDGMENT

Financial support from the Council of Agriculture, Taiwan, is highlyappreciated.

H5 antibody by ELISA 129