AVIAN DISEASES 51:697–704, 2007
Improvements to the Hemagglutination Inhibition Test for Serological Assessment of
Recombinant Fowlpox–H5-Avian-Influenza Vaccination in Chickens and Its Use
Along with an Agar Gel Immunodiffusion Test for Differentiating Infected from
Noninfected Vaccinated Animals
David E. Swayne,AD Gloria Avellaneda,A Thomas R. Mickle,B Nikki Pritchard,B Julio Cruz,B and Michel BublotC
Southeast Poultry Research Laboratory, Agricultural Research Service, U.S. Department of Agriculture, 934 College Station Road,
Athens, Georgia 30605
Merial Select, Inc., 1168 Airport Parkway, Gainesville, Georgia 30501
Merial-SAS Discovery Research, Lyon?? 69007, France
Received 29 November 2006; Accepted and published ahead of print 9 March 2007
SUMMARY. In general, avian influenza (AI) vaccines protect chickens from morbidity and mortality and reduce, but do not
completely prevent, replication of wild AI viruses in the respiratory and intestinal tracts of vaccinated chickens. Therefore,
surveillance programs based on serological testing must be developed to differentiate vaccinated flocks infected with wild strains of
AI virus from noninfected vaccinated flocks in order to evaluate the success of vaccination in a control program and allow
continuation of national and international commerce of poultry and poultry products. In this study, chickens were immunized with
a commercial recombinant fowlpox virus vaccine containing an H5 hemagglutinin gene from A/turkey/Ireland/83 (H5N8) avian
influenza (AI) virus (rFP-H5) and evaluated for correlation of immunological response by hemagglutination inhibition (HI) or agar
gel immunodiffusion (AGID) tests and determination of protection following challenge with a high pathogenicity AI (HPAI) virus.
In two different trials, chickens immunized with the rFP-H5 vaccine did not develop AGID antibodies because the vaccine lacks AI
nucleoprotein and matrix genes, but 0%–100% had HI antibodies, depending on the AI virus strain used in the HI test, the HI
antigen inactivation procedure, and whether the birds had been preimmunized against fowlpox virus. The most consistent and
highest HI titers were observed when using A/turkey/Ireland/83 (H5N8) HPAI virus strain as the b-propiolactone (BPL)–
inactivated HI test antigen, which matched the hemagglutinin gene insert in the rFP-H5 vaccine. In addition, higher HI titers were
observed if ether or a combination of ether and BPL-inactivated virus was used in place of the BPL-inactivated virus. The rFP-H5
vaccinated chickens survived HPAI challenge and antibodies were detected by both AGID and HI tests. In conclusion, we
demonstrated that the rFP-H5 vaccine allowed easy serological differentiation of infected from noninfected birds in vaccinated
populations of chickens when using standard AGID and HI tests.
´ ´ ´ ´
RESUMEN. Mejoras a la prueba de inhibicion de la hemoaglutinacion para la evaluacion serologica de la vacunacion de aves ´
con virus recombinante de viruela expresando el gen H5 de influenza aviar y su utilizacion en conjunto con la prueba de
inmunodifusion en agar para diferenciar animales infectados de animales vacunados.
En general, las vacunas contra influenza aviar protegen a las aves contra morbilidad y mortalidad y reducen pero no previenen la
replicacion de cepas de campo de influenza aviar en los tractos respiratorios e intestinales de aves vacunadas. En consecuencia, los
programas de vigilancia basados en pruebas serologicas deben ser redimensionados para diferenciar parvadas vacunadas e infectadas
con virus de campo de influenza, de parvadas vacunadas pero no infectadas. Esto con la finalidad de evaluar el exito de la ´
vacunacion en un programa de control y permitir la continuidad del comercio avıcola nacional e internacional de aves domesticas y ´
productos avıcolas. En el presente estudio se inmunizaron aves con una vacuna recombinante comercial de viruela aviar expresando
el gen H5 de la hemaglutinina proveniente de un virus de influenza aviar H5N8 (A/pavo/Irlanda /83) y se evaluo la correlacion de ´
´ ´ ´
la respuesta inmune mediante la prueba de inhibicion de la hemoaglutinacion o la prueba de inmunodifusion en agar, ası como la ´
proteccion a un desafıo con un virus de influenza aviar de alta patogenicidad. En dos experimentos diferentes, las aves inmunizadas
con la vacuna recombinante no desarrollaron anticuerpos detectables mediante la prueba de inmunodifusion en agar, esto debido
´ ´ ´
a que la vacuna carece de los genes de la nucleoproteına y de la proteına de la matrız del virus de influenza aviar. Sin embargo,
dependiendo de la cepa de virus de influenza aviar utilizada en la prueba de inhibicion de la hemoaglutinacion, del procedimiento
utilizado para inactivar el antıgeno y de si las aves habıan sido preinmunizadas contra el virus de viruela aviar, entre el 0% y el 100%
de las aves presentaron anticuerpos inhibidores de la hemoaglutinacion. Los tıtulos mas altos y consistentes de inhibicion de la ´
´ ´ ´ ´
hemoaglutinacion se observaron cuando se utilizo como el antıgeno para la prueba de inhibicion de la hemoaglutinacion la cepa ´
H5N8 de influenza aviar de alta patogenicidad (A/pavo/Irlanda /83) inactivada con b-propiolactona, la cual es homologa con el´
inserto de genoma presente en la vacuna recombinante. Adicionalmente, cuando en lugar de solo virus inactivados con b-
´ ´ ´
propiolactona se utilizaron virus inactivados con eter o una combinacion de virus inactivados con eter y virus inactivados con b-
´ ´ ´ ´
propiolactona, se observaron tıtulos mas altos de inhibicion de la hemoaglutinacion. Las aves vacunadas con el virus de viruela aviar
recombinante sobrevivieron al desafıo con influenza aviar de alta patogenicidad y se les detectaron anticuerpos tanto por la prueba
´ ´ ´ ´ ´
de inhibicion de la hemoaglutinacion, como por la prueba de inmunodifusion en agar. En conclusion, se demostro que la vacuna
´ ´ ´
recombinante permitio la facil diferenciacion entre aves infectadas y aves no infectadas en poblaciones de aves vacunadas, utilizando
´ ´ ´ ´
las pruebas estandar de inhibicion de la hemoaglutinacion y prueba de inmunodifusion en agar.
Key words: avian influenza, DIVA, fowlpox virus, hemagglutinin, serology, vaccine
Abbreviations: AGID 5 agar gel immunodiffusion; AI 5 avian influenza; ANOVA 5 analysis of variance; BPL 5 b-
propiolactone; Ck/MX 5 A/chicken/Mexico/232/94 (H5N2) high pathogenicity avian influenza; Ck/SK 5 A/chicken/South
Korea/ES/03 high pathogenicity avian influenza; DIVA 5 differentiating infected from vaccinated animals; Dk/Pd 5 A/duck/
Corresponding author. E-mail: David.Swayne@ars.usda.gov
698 D. E. Swayne et al.
Pottsdam/1402-6/86 (H5N2) low pathogenicity avian influenza; FP 5 fowlpox; HA 5 hemagglutinin; HI 5 hemagglutination
inhibition; HP 5 high pathogenicity; HPAI 5 high pathogenicity avian influenza; IM 5 intramuscular; LP 5 low pathogenicity;
LPAI 5 low pathogenicity avian influenza; NP/M 5 nucleoprotein/matrix protein; rFP-H5 5 recombinant fowlpox virus with H5
influenza hemagglutinin gene; SPF 5 specific-pathogen-free; SQ 5 subcutaneous; Tk/Eng 5 A/turkey/England/N28/73 (H5N2)
low pathogenicity avian influenza; Tk/Ire 5 A/turkey/Ireland/83 (H5N8) high pathogenicity avian influenza; Tk/WI 5 A/turkey/
Wisconsin/68 (H5N9) low pathogenicity avian influenza; WL 5 white leghorn; WW 5 wing-web
Avian influenza (AI) is a disease of birds caused by type A through high-efficiency particulate air filters and located in rooms under
influenza viruses, which belong to the Orthomyxoviridae family high biocontainment (Biosafety Level 3 agriculture) (1). Light exposure
(20). Type A influenza viruses are subdivided into 16 hemagglutinin was continuous. Water and feed were provided ad libitum. At
(HA; H1–H16) and 9 neuraminidase (N1–N9) subtypes (6). In termination, all chickens were humanely euthanatized (100 mg/kg
sodium pentobarbital, intravenously).
addition, AI viruses are further classified into two pathotypes: high
Vaccines. A commercial recombinant fowlpox (FP) virus vaccine
pathogenicity (HP) or low pathogenicity (LP) (20). The HPAI containing a cDNA insert of the H5 hemagglutinin gene of A/turkey/
viruses cause very severe systemic disease in chickens and turkeys Ireland/1378/83 (T/Ire) AI virus (rFP-H5) (TROVACH-AIV H5, Merial
with up to 100% mortality, whereas LPAI viruses cause localized Select, Inc., Gainesville, GA) (23), a commercial FP virus vaccine (Merial
infection of respiratory or intestinal tracts with mild or no disease. Select, Inc.), and an experimental inactivated oil-emulsion vaccine
All HPAI viruses are of H5 and H7 subtypes, whereas LPAI can be containing A/turkey/Wisconsin/68 (H5N9) LPAI virus were adminis-
of any of the H1–H16 subtypes. However, because H5 and H7 tered subcutaneously in the nape of the neck. The latter vaccine (OE Tk/
LPAI viruses can mutate to HP, H5 and H7 LPAI and HPAI are WI) was prepared using previously described procedures, except
notifiable diseases to the World Organization for Animal Health and inactivation was by 0.1% b-propiolactone (BPL) instead of formalin in
subsequently have a negative impact on trade. order to better preserve antigenicity, and the vaccine was administered in
In the United States, AI outbreaks in poultry are controlled a 0.2-ml volume to give uniform dosage (14,16). The rFP-H5 was given at
13 (103.5 mean tissue culture infective doses [TCID50]) or 103 [104.5
through prevention, management, and eradication programs (15).
TCID50]) of manufacturer’s recommended dose in a 0.2-ml volume.
These programs have been accomplished through multidisciplinary Challenge viruses. The first egg passage of A/chicken/South Korea/
approaches including increased surveillance and diagnostics, en- ES/03 (H5N1) (Ck/SK) and a second egg passage of A/chicken/
hanced biosecurity including quarantine in the infected zone, Queretaro/14588-19/95 (H5N2) (Ck/MX) HPAI viruses were used for
education of poultry workers, and elimination of infected or challenge.
suspected infected poultry. Several vaccines have shown efficacy in Clinical evaluation. Birds exhibiting clinical signs, including resting
preventing clinical signs and death and reducing respiratory and on the sternum with reluctance to rise, drooping of the head, and
intestinal replication of LPAI and HPAI viruses. The major generalized ruffled feathers were scored as sick. Morbidity rates were
protective immune response is produced against the homologous calculated based on these minimal clinical signs. Many of the birds
hemagglutinin protein. In many instances, AI vaccination has had exhibited more severe clinical signs and died. The mortality rate was
minimal use because of concerns about a potential negative impact calculated based on the number of birds that died.
Serology. Serum samples were tested for influenza A–specific anti-
on trade of poultry and poultry products. Traditionally, inactivated
nucleoprotein/matrix protein (NP/M) antibodies by AGID and for anti-
AI virus vaccination has interfered with the surveillance programs hemagglutinin antibodies by HI test (22). The antigen used for the
because standard agar gel immunodiffusion (AGID) and hemagglu- AGID test was produced from A/turkey/WI/66 (H9N2) LPAI virus
tination inhibition (HI) serological tests cannot distinguish between strain as previously described (22). For the HI test antigen, the following
antibodies resulting from vaccination vs. those arising from an viruses were inactivated with 0.1% BPL and used as previously described
infection. (22): A/turkey/Ireland/83 (H5N8) HPAI (Tk/Ire), A/turkey/Wisconsin/
Three fowlpox virus recombinant vaccines containing H5 AI virus 68 (H5N9) LPAI (Tk/WI), A/duck/Pottsdam/1402-6/86 (H5N2) LPAI
gene inserts have been developed, evaluated, and licensed for use in (Dk/Pd), A/turkey/England/N28/73 (H5N2) LPAI (Tk/Eng), Ck/SK
chickens (2,12,18,23). In experimental studies, vaccinated birds were HPAI, and Ck/MX HPAI (Ck/MX) viruses. In addition, Tk/Ire was
negative for AGID antibodies, as expected, because these vaccines did used following ether treatment (11). Briefly, Tween 80 was added to
not contain inserts of AI nucleoprotein or matrix genes. However, in infectious allantoic fluid in a glass vial to make a final volume of 0.1%,
and was allowed to set for 15 min. This was followed by addition of
only one of the studies were HI antibodies consistently identified,
one-half volume of anhydrous ether with shaking for 30 sec; the aqueous
which, when identified, would be a good predictor of immunization layer was then allowed to separate. The aqueous phase was stored at 4 C
and protection. This study was undertaken to determine if the HI test in an explosion-proof refrigerator and used within a few days as the HI
to detect specific anti-hemagglutinin antibodies could be standardized antigen. All work with ether was done in total air–exhaust cabinetry. For
to measure and predict a consistent protective immunological combined ether–BPL antigen, the ether treatment was accomplished
response in vaccinated chickens and whether the AGID test could followed by 0.1% BPL treatment. All HI titers were reported as
be used to detect infected chickens within the vaccinated population geometric mean titers with 8 being the minimum positive titer.
after challenge with two different HPAI viruses. Phylogenetic analysis. An unrooted phylogenetic tree using a de-
duced amino acid sequence from the full-length nucleotide sequence of
the HA1 portion of the hemagglutinin gene was constructed with the
MATERIALS AND METHODS neighbor-joining (NJ) method (13) using 1000 bootstraps implemented
in MEGA 3 software (MEGA3, Center for Evolutionary Functional
Chickens and housing. One-day-old commercial white leghorn Genomics, Tempe, AZ) (9) and using a Jones-Taylor-Thornton ( JTT)
(WL) chickens (HyLine, Inc., Dallas Center, IA) and specific-pathogen- amino acid matrix (13). Alignment was performed with Clustal W (24)
free (SPF) WL chickens (SPAFAS, Inc., Storrs, CT, and Sunrise Farms, followed by manual editing and gap removal.
Catskill, NY) were used. All birds were housed in isolation during the Experimental design. Experiment 1. Effect of AI virus strain and
brooding and postvaccination period. For virus challenge, chickens were inactivation procedure for HI antigen production on the HI antibody titers
housed in negative-pressure, stainless steel isolation cabinets ventilated in chickens following vaccination and challenge. Groups of 12 1-day-old
DIVA for rFP-H5 AI vaccines 699
Table 1. Design for experiment 3. SPF WL chickens were vaccinated at 1, 21, and 42 days of age by subcutaneous (SQ), wing-web (WW), or
intramuscular (IM) routes with FP virus control or recombinant fowlpox (rFP-H5) vaccines. The rFP-H5 vaccine was given in 13 or 103 doses.
Groups 1–10 had 15 chickens, and Group 11 had 14 chickens.
Vaccines and age at vaccination
Experimental groups 1 day 21 days 42 days
1 None None None
2 13 rFP-H5 (SQ) None None
3 13 rFP-H5 (SQ) 13 rFP-H5 (SQ) None
4 None 13 rFP-H5 (SQ) None
5 None 13 rFP-H5 (SQ) 13 rFP-H5 (SQ)
6 FP (SQ) 13 rFP-H5 (SQ) None
7 FP (SQ) 13 rFP-H5 (SQ) 13 rFP-H5 (SQ)
8 FP (SQ) 103 rFP-H5 (SQ) None
9 FP (SQ) 103 rFP-H5 (SQ) 103 rFP-H5 (SQ)
10 FP (SQ) 103 rFP-H5 (IM) 103 rFP-H5 (IM)
11 FP (SQ) 103 rFP-H5 (WW) 103 rFP-H5 (WW)
commercial WL chickens were vaccinated with a 13 or 103 dose of with a 13 or 103 dose of rFP-H5 vaccine was positive for
rFP-H5 AI vaccine or with a 13 dose of OE Tk/WI vaccine. The 103 anti-NP/M antibodies at 3 wk postvaccination and before
dose was included to increase opportunity to detect HI serological challenge with HPAI virus. However, the chickens vaccinated with
response because prior studies using a 13 dose have given inconsistent
the Tk/WI oil-emulsion vaccine were positive for anti-NP/M
HI titers (18). A nonvaccinated group of 10 birds was inoculated with
diluent only. All birds were challenged at 3 wk of age with 103.3 mean antibodies before challenge (Table 2). After challenge all the
chicken embryo infectious doses (EID50) of Korea/03, which was immunized groups had one or more positives for anti-NP/M
equivalent to minimum dose for 100% lethality. Serum samples were antibodies.
collected in 3-wk-old chickens prior to challenge with HP Korea/03 and Six strains of H5 AI virus were used in HI tests between all 3
in 5-wk-old chickens 2 wk after challenge for analysis of anti-NP/M experiments. When examining the genetic data on the six HI antigen
protein antibodies and anti-H5 antibody titers using the AGID and HI strains, two viruses were the North American H5 lineage and four
tests, respectively. viruses were the Eurasian H5 lineage (Fig. 1). However, the viruses
Experiment 2. HI antibody titers and correlation with protection in birds were not closely related genetically except for the Dk/Pd and Tk/Eng
vaccinated with rFP-H5 vaccine with and without prior immunization
(Fig. 1). When examining the relatedness of the viruses to the
with FP virus. Groups of 10 or 11 1-day-old SPF WL chickens were
vaccinated subcutaneously with FP vaccine or remained unvaccinated hemagglutinin insert in the rFP-H5, the heterologous virus strains
and then were vaccinated at day 21 or day 42 of age with rFP-H5 had 93% or less similarity in the HA1 portion (Table 3). This
vaccine. At 70 days of age, all chickens were challenged intranasally with percentage of amino acid similarity did not correlate with the HI
a high-challenge dose of Ck/MX, approximately 1000 mean chicken antibody response (Tables 2, 3). In assessing vaccine response to
lethal doses (107.2EID50), to test maximal protection of the vaccine. rFP-H5 vaccine in Experiment 1, the best HI antibody response and
Serum samples were obtained at 70 and 84 days of age and tested by highest HI antibody titers were obtained when using the
AGID and HI tests. homologous Tk/Ire antigen (Table 2). By contrast, the other four
Experiment 3. HI antibody titers from birds immunized with FP antigen strains (Tk/Eng, Dk/Pd, Tk/WI, and Ck/SK) had 83.5% to
followed by different doses of rFP-H5 vaccine via different routes. Table 1
describes the experimental design. Serum samples from birds that had
93.2% similarity to the rFP-H5 vaccine hemagglutinin insert, but all
been experimentally vaccinated at 1 day of age with FP vaccine and with were equally poor at detecting HI antibody responses and the
13 or 103 doses of rFP-H5 given by different routes of inoculation resulting titers were very low (Tables 2, 3).
were collected and tested by AGID and HI tests. Also, serum samples When using BPL-inactivated HI virus, the number of positive
from birds immunized only with rFP-H5 AI virus vaccine were obtained birds for the HI test as well as the geometric mean titer (GMT)
and tested by AGID and HI tests. values in each treatment group varied with the HI test antigen used
Experiment 4. Field sample evaluation. Serum samples were obtained (Table 2). At prechallenge, the number of HI positives and the
from 10 chickens from each of the following four flocks in the field in GMT for the 13 and 103 rFP-H5 groups were significantly greater
Guatemala and tested by AGID and HI tests: 1) rFP-H5 vaccine, 2)
when the HI test antigen was Tk/Ire as compared to using the other
inactivated whole AI vaccine (n 5 2), and 3) nonvaccinated sentinel
chickens. HI test antigens. With Tk/Ire antigen, all rFP-H5 vaccinated birds
Statistical analysis. Normally distributed data sets were analyzed by had HI antibodies. When the HI test antigen was Dk/Pd, Tk/WI, or
analysis of variance (ANOVA). Data sets with significant differences (P Ck/SK, no HI positives were detected in the 13 group, whereas
, 0.05) were further analyzed by Student–Newman–Keuls multiple a few chickens were HI-positive but had low GMT titers in the 103
comparison test. Data sets not normally distributed were analyzed by rFP-H5 group. When using Tk/Eng HI test antigen, 50% of the
a nonparametric ANOVA test (Kruskal–Wallis) and, for significantly rFP-H5 vaccinated chickens had HI antibodies, but the titers were
different groups (P , 0.05), the Dunn multiple comparison test was low. After challenge, the number of chickens with HI titers was
performed. All statistical tests were performed using SigmaStat software significantly greater for Tk/Eng and Tk/Ire HI antigens as compared
(Jandel Scientific, San Rafael CA).
to the others, but only Tk/Ire had a significantly greater GMT. All
rFP-H5 groups had higher HI GMTs after, as compared to before,
challenge. Only the homologous Tk/Ire antigen was consistent in
assessing HI serological response in rFP-H5–vaccinated and AI
Experiment 1. Effect of AI virus strain and inactivation procedure virus–challenged chickens.
for HI antigen production on the HI antibody titers in chickens By contrast, 42%–100% of inactivated Tk/WI vaccinated
following vaccination and challenge. None of the chickens immunized chickens had HI antibodies prechallenge whereas 83%–100% had
700 D. E. Swayne et al.
Table 2. Experiment 1. HI serological response from chickens immunized with 13 or 103 doses of recombinant fowlpox (rFP-H5), 13 dose of inactivated Tk/WI, or diluent at 1 day of age and
Different uppercase superscript letters denote significant differences in the number of positives between groups. Different lowercase superscript letters indicate significant differences in HI GMT
tested at 3 wk of age (3 wk postvaccination) and 5 wk of age (2 wk post-challenge with Ck/SK). Serological response was measured using different BPL-inactivated AI virus strains for HI antigen.
12/12A (315)a 10/12A
Postchallenge HI test results at 5 wk of age
(no. positive/total tested [GMT])
Fig. 1. Phylogenetic tree based on deduced amino acid sequence of
the HA1 segments of H5 AI viruses. Abbreviations for viruses used for
this alignment are as follows: Ck/Mx/94 (A/chicken/Mexico/232/94
[H5N2]), Ck/Qu/95 (A/Chicken/Queretaro/14588-19/95 [H5N2]),
Tk/WI/68 (A/turkey/Wisconsin/68 [H5N9]), Ck/SK/03 (A/chicken/
Korea/ES/03 [H5N1]), Tk/Ire/83 (A/turkey/Ireland/1378/1983
[H5N8]), Tk/Eng/73 (A/turkey/England/N28/73 [H5N2]), Dk/Pd/86
(A/duck/Potsdam/1402-6/86 [H5N2]), Ck/Pe/83 (A/chick/Pennsylva-
nia/1/1983 [H5N2]), Ck/Tx/02 (A/chicken/TX/167280-4/02
[H5N3]), Ma/Wi/75 (A/Mallard/Wisconsin/34/75 [H5N6]), Ost/SA/
04 (A/Ostrich/South Africa/1/2004 [H5N2]), Tk/Mn/81 (A/turkey/
Prechallenge HI test results at 3 wk of age
12/12A (39)a 3/12B (5)b
Minnesota/3689-1551/81 [H5N2]), Dk/Ire/83 (A/duck/Ireland/113/
(no. positive/total tested [GMT])
1983 [H5N8]), Ck/Mx/95 (A/chicken/Mexico/28159-541/95
[H5N2]), Ck/It/97(A/Chicken/Italy/312/97 [H5N2]), Ck/It/98 (A/
chicken/Italy/8/98 [H5N2]), Ck/Th/04 (A/chicken/Chachoengsao/
Thailand/CU-10/04 [H5N1]), and Bird/Th/04 (A/bird/Thailand/3.1/
HI antibodies postchallenge when using different HI antigens
(Table 2). The Tk/WI HI test antigen, which matched the vaccine
strain, gave significantly higher titers both pre- and postchallenge as
compared to the other HI antigens.
To test the influence of antigen inactivation procedures, we tested
sera with the Tk/Ire virus inactivated by three different methods:
BPL, ether, and the combination of ether followed by BPL treatment
(Table 4). Prechallenge and postchallenge, chickens from the rFP-
2/11 6/12B (6)b
1/12 6/12 (6)
H5 vaccine groups were all positive for HI antibodies, but the titers
were significantly higher when using ether and ether/BPL compared
to the BPL-only inactivated virus. The HI antibody titers were lower
AGID test results
Table 3. Amino acid similarity of H5 AI viruses used as HI antigen
to the hemagglutinin gene insert in rFP-H5 vaccine based on deduced
amino acid sequence derived from full-length HA1.
antibody titers between groups.
HA1 amino acid sequence identity
Virus strain with Tk/Ire (%)
DIVA for rFP-H5 AI vaccines 701
Table 4. Experiment 1. Effect of different methods for inactivation (BPL, ether, or ether/BPL) of Tk/Ire virus on the measured HI antibody
serological response from chickens immunized with 13 or 103 doses of recombinant fowlpox (rFP-H5), 13 dose of inactivated Tk/WI, or diluent
at 1 day of age and tested at 3 wk of age (3 wk postvaccination) and 5 wk of age (2 wk postchallenge with Ck/SK). Serological response was
measured using Tk/Ire virus inactivated with BPL, ether, or ether/BPL. Different uppercase superscript letters denote significant differences in the
number of positives between groups. Different lowercase superscript letters indicate significant differences in HI GMT antibody titers
Prechallenge HI test results (no. positive/total tested [GMT]) Postchallenge HI test results (no. positive/total tested [GMT])
Experimental groups BPL Ether Ether/BPL BPL Ether Ether/BPL
A b A a A a A b A a
13 rFP-H5 12/12 (42) 12/12 (446) 12/12 (223) 12/12 (97) 12/12 (1261) 12/12A (512)a
103 rFP-H5 12/12A (39)b 12/12A (478)a 12/12A (208)a 11/11A (97)b 11/11A (1552)a 11/11A (955)a
Diluent 0/10 0/10 0/10 0/10 0/10 0/10
Inactivated Tk/WI 10/12A (11)c 12/12A (119)a 12/12A (56)b 12/12A (45)b 12/12A (478)a 12/12aA (274)a
for ether/BPL as compared to the ether-only inactivated virus, but 21 days of age (groups 2–5) were HI antibody positive at 42 days
the differences were not significant. For the inactivated Tk/WI when using BPL-inactivated Tk/Ire virus, whereas 100% were positive
vaccine, the number of HI antibody positives pre- and postchallenge using ether-inactivated Tk/Ire virus (Table 6). The corresponding HI
was not significantly different between the three inactivation titers were threefold to fourfold higher when using the ether-
procedures for HI antigen production, but titers were significantly inactivated virus. At 63 days, the number of HI positive birds was
higher for ether-inactivated or ether/BPL–inactivated virus as similar, but with slightly decreased HI titers, with the exception of
compared to BPL-inactivated virus. group 5 (23 vaccinated group on 21 and 42 days) where the mean HI
Experiment 2. Correlation of HI antibody titers with protection in antibody titer was similar to the 42-day time point. By contrast, of the
birds vaccinated with rFP-H5 vaccine with and without prior chickens given FP at 1 day of age and followed with rFP-H5 vaccines
immunization with FP vaccine. None of the vaccinated chickens at 21 and 42 days (groups 6–11), only group 9 vaccinates (103 rFP-
had anti-NP/M antibodies (AGID) postvaccination (Table 5). All H5 by subcutaneous route on 21 and 42 days of age) had HI
chickens vaccinated with only rFP-H5 vaccine at 21 days of age had antibodies in 100% of the chickens and only when using ether-
HI antibodies at 70 days of age, but HI titers were significantly inactivated Tk/Ire virus (Table 6).
higher using Tk/Ire ether-inactivated as compared to BPL- Experiment 4. Field serum samples evaluation. Nonvaccinated
inactivated virus as test antigen. All the rFP-H5 vaccinated chickens sentinels lacked HI antibody titers irrespective of which BPL-
survived the HPAI virus challenge (Table 5). By contrast, the three inactivated virus as the HI antigen was used (Table 7). HI antibodies
groups vaccinated with FP at 1 day of age followed either by rFP-H5 were in 50% of the serum samples from chickens vaccinated with
vaccine at 21 or 42 days of age or no rFP-H5 vaccine lacked HI inactivated oil-emulsified vaccines when using the homologous BPL-
antibodies regardless of HI test antigen used, and 80%–100% of the inactivated Ck/MX virus as the HI test antigen compared to 20%–30%
chickens died following lethal challenge with HPAI virus. with the BPL-inactivated Tk/Ire virus. HI antibodies were present in
Following challenge, all survivors had anti-NP/M and anti- 80% of the serum samples of rFP-H5–vaccinated chickens when using
hemagglutinin antibodies (Table 5). For HI tests, the postchallenge homologous BPL-inactivated Tk/Ire virus whereas 0% were HI positive
GMT titers were significantly greater than postvaccination titers. using the BPL-inactivated Ck/MX virus as the HI test antigen.
Experiment 3. HI antibody titers from birds immunized with FP
followed by different doses of FP-H5 vaccine via different routes. The
number of chickens with HI antibodies and the antibody GMT
varied greatly between the different groups. Eighty percent to 100% The successful use of vaccination in an AI control program
of chickens that received one or two doses of rFP-H5 at 1 and/or necessitates the design and use of testing strategies to monitor the
Table 5. Experiment 2. Effect of different methods for inactivation of Tk/Ire virus on the measured HI antibody serological response from
chickens immunized with recombinant fowlpox (rFP-H5) and control fowlpox (FP) vaccines at 1 day, 21 days, and 42 days of age and challenged
with 107.2 EID50 of Mexico/95 HPAI virus at 70 days. Sera were tested at 70 and 84 days of age in AGID and HI tests. HI tests used BPL- or ether-
inactivated Tk/Ire virus as test antigen. MDT 5 mean time to death in days. Different uppercase superscript letters denote significant differences in
the number of positives between groups. Different lowercase superscript letters indicate significant differences in HI GMT antibody titers
Serological response at 70 days of age Serological response at 84 days of age (no.
(no. positive/total tested [GMT]) positive/total tested [GMT])
Vaccines and age at vaccination HI using HI using HI using HI using
Mortality BPL- ether- BPL- ether-
1 day 21 days 42 days Morbidity (MDT) AGID Tk/Ire Tk/Ire AGID Tk/Ire Tk/Ire
— rFP-H5 — 0/11 0/11(0) 0/11 11/11A (24)b 11/11A (239)a 11/11 11/11A (3327)a 11/11A
FP rFP-H5 — 9/10 8/10 0/10 0/10A 0/10A 2/2 2/2A (362)a 2/2A (8192)b
FP — rFP-H5 10/11 9/11 0/11 0/11A 1/11A 2/2 2/2A (724)a 2/2A (23,171)b
(3.2) (0) (5)
FP — — 10/10 10/10 0/10 0/10 0/10 N/A N/A N/A
(3.0) (0) (0)
702 D. E. Swayne et al.
Table 6. Experiment 3. Effect of dose and route of vaccination on the HI antibody response from chickens vaccinated by SQ, WW, or IM routes
with FP virus control or recombinant fowlpox virus (rFP-H5) vaccines. The rFP-H5 vaccine was given in 13 or 103 doses. Sera were tested at 42
and 63 days of age in HI test using BPL- or ether-inactivated Tk/Ire virus as HI test antigen. Different uppercase superscript letters denote significant
differences in the number of positives between groups. Different lowercase superscript letters indicate significant differences in HI GMT antibody
titers between groups.
HI test results at 63 days of age
HI test results at 42 days of age (no. positive/total tested
Vaccines and age at vaccination (no. positive/total tested [GMT]) [GMT])
Group 1 day 21 days 42 days BPL Ether BPL Ether
A A A
1 None None None 0/14 0/14 0/15 0/15A
2 13 rFP-H5 (SQ) None None 13/15A 15/15A 11/15A 14/15A
(23)a (338)b (16)a (194)b
3 13 rFP-H5 (SQ) 13 rFP-H5 (SQ) None 12/15A 15/15A 12/15A 15/15A
(17)a (274)b (13)a (239)b
4 None 13 rFP-H5 (SQ) None 13/13A 13/13A 14/15A 15/15A
(32)a (478)b (16)a (239)b
5 None 13 rFP-H5 (SQ) 13 rFP-H5 (SQ) 13/14A 14/14A 15/15A 15/15A
(30)a (630)b (42)a (724)b
6 FP (SQ) 13 rFP-H5 (SQ) None 0/14A 1/14A (4)a 0/15A 4/15A (5)a
7 FP (SQ) 13 rFP-H5 (SQ) 13 rFP-H5 (SQ) 0/15A 2/15A (5)a 5/14A (6)a 10/14B (18)b
8 FP (SQ) 103 rFP-H5 (SQ) None 0/15A 5/15A (6)a 0/15A 5/15A (6)a
9 FP (SQ) 103 rFP-H5 (SQ) 103 rFP-H5 (SQ) 0/15A 8/15AB (7)a 7/13AB (11)a 13/13B (79)b
10 FP (SQ) 103 rFP-H5 (IM) 103 rFP-H5 (IM) 0/12A 0/12A 2/13A (5)a 8/13A (9)a
11 FP (SQ) 103 rFP-H5 (WW) 103 rFP-H5 (WW) 0/14A 1/14A (4)a 0/11A 1/11A (5)a
immunological response to the vaccine to serve as an indirect Surprisingly, these early trials reported inconsistent HI antibody
measure of protection and to detect naturally infected poultry within responses and low HI titers (2,18,26), which made monitoring the
the vaccinated population (i.e., ‘‘differentiating infected from vaccination program to assess immunological protection impractical.
vaccinated animals’’ [DIVA]) (20). For inactivated AI vaccines, HI By contrast, in the current rFP-H5 vaccine study, HI antibody
testing has been the standard for indirect assessment of protection in responses using BPL-inactivated Tk/Ire virus (i.e., the donor of the
chickens because hemagglutinin is the major protein that elicits hemagglutinin insert in the rFP-H5 vaccine) were consistently
a protective immune response (15). The determination of AI virus demonstrated in 100% and 80% of the chickens experimentally or
infection status within such a vaccinated population has utilized field-vaccinated with rFP-H5 vaccine, respectively, and the mean titers
either an existing serological test on unvaccinated sentinels for in these groups (13–42 GMT) were similar to those reported in
detection of antibodies against a specific hemagglutinin subtype a previous experimental study (26). However, when using North
using an HI test, or detection of antibodies against influenza type A American or other Eurasian lineages of H5 AI viruses as the source of
proteins (NP/M proteins) using AGID or ELISA tests or special BPL-inactivated virus as HI test antigen, results were inconsistent or
DIVA strategies which use vaccine strains with a different less consistent in HI antibody positive responses (0%–50%), and the
neuraminidase than the wild AI virus found in the field and mean titers were usually nonexistent or low (0–6 GMT). Previous
serological tests to detect antibodies against the field wild AI-virus studies reported inconsistent HI test results when using an HI test
neuraminidase (4,7,21). Recently, detection of antibodies against antigen that differed from the AI H5 hemagglutinin gene insert in the
NS-1 proteins has been proposed as a DIVA test, because such recombinant FP vaccine (2,18,26). Furthermore, in our study, the
antibodies were detected in AI virus-infected poultry but not in commonly used H5 HI test reference strain for North America (Tk/
inactivated AI virus–vaccinated poultry (25,28). WI) was not acceptable as an HI antigen because of the less than 25%
Recombinant FP vaccines with inserts of AI H5 hemagglutinin seropositive rates and low titers (5 GMT), whereas the HI antigen
genes have provided protection in chickens against both experimen- strain used in Europe and Asia, Tk/Eng, gave slightly better results
tal H5 LPAI and HPAI virus challenge (2,18,19,27). Such (50% seropositive and 6 GMT). However, the Tk/Eng antigen was
vaccinated chickens have uniformly lacked antibodies against NP/ still not acceptable as an HI antigen for assessing immunological
M proteins of the AI virus (i.e., AGID negative) because the response in rFP-H5 vaccinated chickens. Therefore, serological
recombinant FP vaccines contained only the hemagglutinin gene and documentation of effective vaccination with rFP-H5 was best achieved
thus provided a built-in DIVA strategy to detect infections in the by testing sera from vaccinated chickens using BPL-inactivated virus
vaccinated population through identification of AGID antibodies. derived from homologous Tk/Ire strain, i.e., the strain that donated
Table 7. HI antibody response in broiler chickens from Guatemala, which were nonvaccinated sentinels or vaccinated with one of two different
commercial inactivated oil-emulsion AI (OE Ck/MX A and B) vaccines or rFP-H5 vaccine at manufacturer’s recommended dose. HI test antigen
was derived from BPL-inactivated Ck/MX and Tk/Ire AI viruses.
HI serology (no. positive/total tested [GMT])
Vaccine group Age at vaccination (days) Age at serology (days) BPL-inactivated Ck/MX BPL-inactivated Tk/Ire
Nonvaccinated sentinels Not applicable Not applicable 0/10 0/10
OE Ck/MX – A Unknown 43 5/10 (9.2) 2/10 (4.9)
OE Ck/MX - B 7 33 5/10 (8.6) 3/10 (8)
rFP-H5 1 34 0/10 (0) 8/10 (13)
DIVA for rFP-H5 AI vaccines 703
the hemagglutinin gene to rFP-H5 vaccine. However, because Tk/Ire test using BPL-inactivated Tk/Ire virus as HI test antigen can be
is an HPAI virus, only special laboratories with high biocontainment used to assess the immunological response of chickens, which can be
can produce this antigen for HI testing. With the influenza virus used as an indirect measure of protection. Prior studies have
reverse-genetic technology, a suitable LPAI virus for antigen pro- demonstrated efficacy against a variety of H5 HPAI challenge viruses
duction can be produced by incorporating the hemagglutinin gene (19). Second, because the rFP-H5 vaccine lacks NP/M proteins, the
from the Tk/Ire virus without the coding region for the HPAI cleavage detection of antibodies against NP/M proteins in AGID or ELISA
site or a naturally occurring H5 LPAI of close genetic relationship, tests can serve as a serological test to identify infection among the
such as A/Duck/Ireland/113/83 (H5N8) (see Fig. 1). vaccinated population of chickens, i.e., an easy-to-use DIVA
Interestingly, use of ether-inactivated or ether/BPL-inactivated strategy.
viruses in HI tests of rFP-H5 vaccinated chickens produced five to 20
times higher HI titers than using BPL-inactivated Tk/Ire virus.
Previously, ether-inactivated influenza B virus in HI testing increased REFERENCES
sensitivity, but reduced specificity of HI tests to detect antibodies
1. Barbeito, M. S., G. Abraham, M. Best, P. Cairns, P. Langevin, W. G.
against influenza B virus in humans resulting from vaccination or Sterritt, D. Barr, W. Meulepas, J. M. Sanchez-Vizcaino, M. Saraza, E.
infection (8). Ether-inactivated viruses used as HI test antigen were Requena, M. Collado, P. Mani, R. Breeze, H. Brunner, C. A. Mebus, R. L.
required to identify primary and secondary responses to equine Morgan, S. Rusk, L. M. Siegfried, and L. H. Thompson. Recommended
influenza A virus vaccination (3). The mechanism of the increased biocontainment features for research and diagnostic facilities where animal
sensitivity using ether-inactivated virus may be the result of splitting pathogens are used. Rev. Sci. Tech. Off. Int. Epiz. 14:873–887. 1995.
the virion into multiple 30-mm spherical particles thereby making 2. Beard, C. W., W. M. Schnitzlein, and D. N. Tripathy. Protection of
more hemagglutinin available to react with antibodies (5,10). Such chickens against highly pathogenic avian influenza virus (H5N2) by
use of ether- or ether/BPL-inactivated viruses may enhance HI testing recombinant fowlpox viruses. Avian Dis. 35:356–359. 1991.
and improve evaluation of immunological responses in the field. 3. Burrows, R., P. R. Spooner, and D. Goodridge. A three-year
evaluation of four commercial equine influenza vaccines in ponies
However, the combustible nature of ether necessitates care in the
maintained in isolation. Dev. Biol. Stand. 39:341–346. 1977.
production of such antigen and quality control to assure removal of
4. Capua, I., C. Terregino, G. Cattoli, F. Mutinelli, and J. F. Rodriguez.
any residual ether in the final antigen product. Development of a DIVA (Differentiating Infected from Vaccinated
Because the rFP-H5 vaccine lacked the NP and M genes of an Animals) strategy using a vaccine containing a heterologous neuraminidase
AI virus, all vaccinated, prechallenged birds were negative for for the control of avian influenza. Avian Pathol. 32:47–55. 2003.
antibodies against NP and M proteins, i.e., AGID negative. On low- 5. Davenport, F. M., R. Rott, and W. Schafer. Physical and biological
dose H5N1 HPAI virus challenge, a few of the rFP-H5–vaccinated properties of influenza virus components obtained after ether treatment. J.
birds developed anti-NP/M antibodies, and on high-dose H5N2 Exp. Med. 112:765–783. 1960.
HPAI virus challenge, all rFP-H5–vaccinated chickens developed 6. Fouchier, R. A. M., V. Munster, A. Wallensten, T. M. Bestebroer, S.
anti-NP/M antibodies. These data suggest detection of anti-NP/M Herfst, D. Smith, G. F. Rimmelzwaan, B. Olsen, and A. D. M. E.
antibodies will identify infected chickens within an rFP-H5– Osterhaus. Characterization of a novel influenza a virus hemagglutinin
subtype (h16) obtained from black-headed gulls. J. Virol. 79:2814–2822.
vaccinated population, but the test should be interpreted as a flock
test and not as an individual bird test. Similarly, usage of 7. Halvorson, D. A. Avian influenza control in Minnesota. Poult. Dig.
a neuraminidase inhibition test on sera from inactivated AI– 54:12–19. 1995.
vaccinated chickens has demonstrated a similar DIVA concept as 8. Kendal, A. P., and T. R. Cate. Increased sensitivity and reduced
a flock test but not as an individual bird test, for determining specificity of hemagglutination inhibition tests with ether-treated influenza
infection status (21). B/Singapore/222/79. J. Clin. Microbiol. 18:930–934. 1983.
The presence of HI antibodies in individual rFP-H5–vaccinated 9. Kumar, S., S. Tamura, and M. Nei. MEGA3: integrated software for
birds was associated with protection from lethal challenge by two molecular evolutionary genetics analysis and sequence alignment. Briefings
different HPAI viruses (Tables 2, 5), but occasionally birds without in Bioinformatics 5:150–163. 2006.
measurable HI titers were protected (Table 5). The minimal GMT 10. Monto, A. S., and H. F. Maassab. Ether treatment of type B influenza
individual bird titer in our laboratory using BPL-inactivated virus as virus antigen for the hemagglutination inhibition test. J. Clin. Microbiol.
the HI antigen was 8. Vaccination with an FP vaccine on day 1
11. Norrby, E. Hemagglutination by measles virus. 4. A simple procedure
interfered with development of a protective immune response when for production of high potency antigen for hemagglutination-inhibition
followed by one or two doses of rFP-H5 vaccine, as evidenced by the (HI) tests. Proc. Soc. Exp. Biol. Med. 111:814–818. 1962.
lack of consistent protection from lethal HPAI challenge, which has 12. Qiao, C. L., K. Z. Yu, Y. P. Jiang, Y. Q. Jia, G. B. Tian, M. Liu, G.
been previously demonstrated (17). However, the current studies H. Deng, X. R. Wang, Q. W. Meng, and X. Y. Tang. Protection of chickens
had some interesting findings. In chickens free from exposure to FP, against highly lethal H5N1 and H7N1 avian influenza viruses with
antibody titers obtained 3 wk following vaccination at 1 or 21 days a recombinant fowlpox virus co-expressing H5 haemagglutinin and N1
of age were similar, and a second rFP-H5 vaccination 3 wk later did neuraminidase genes. Avian Pathol. 32:25–31. 2003.
not induce a boost in humoral immunity (Tables 4–6). In the case 13. Saitou, N., and M. Nei. The neighbor-joining method: a new method
of active FP immunity, there was severe inhibition of antibody for reconstructing phylogenetic trees. Mol. Biol. Evol. 4:406–425. 1987.
response after a single vaccination with 13 or 103 doses of rFP-H5. 14. Stone, H. D. Efficacy of avian influenza oil-emulsion vaccines in
chickens of various ages. Avian Dis. 31:483–490. 1987.
There was, however, partial induction of AI antibody response
15. Swayne, D. E. Application of new vaccine technologies for the control
following two subcutaneously administered 13 doses of rFP-H5 and
of transboundary diseases. Dev. Biologicals 119:219–228. 2004.
more consistently following two 103 doses of rFP-H5 (Table 6). In 16. Swayne, D. E., J. R. Beck, M. Garcia, and H. D. Stone. Influence of
all situations, the subcutaneous route of vaccination was superior to virus strain and antigen mass on efficacy of H5 avian influenza inactivated
intramuscular or wing-web vaccination (Table 6). vaccines. Avian Pathol. 28:245–255. 1999.
In conclusion, the rFP-H5 vaccine has potential use as a tool in 17. Swayne, D. E., J. R. Beck, and N. Kinney. Failure of a recombinant
a comprehensive avian influenza control program. First, the H5 HI fowl poxvirus vaccine containing an avian influenza hemagglutinin gene to
704 D. E. Swayne et al.
provide consistent protection against influenza in chickens preimmunized through sequence weighting, position specific gap penalties and weight
with a fowl pox vaccine. Avian Dis. 44:132–137. 2000. matrix choice. Nucleic Acids Res. 22:4673–4680. 2006.
18. Swayne, D. E., J. R. Beck, and T. R. Mickle. Efficacy of recombinant 25. Tumpey, T. M., R. Alvarez, D. E. Swayne, and D. L. Suarez. A
fowl pox vaccine in protecting chickens against highly pathogenic Mexican- diagnostic aid for differentiating infected from vaccinated poultry based on
origin H5N2 avian influenza virus. Avian Dis. 41:910–922. 1997. antibodies to the nonstructural (NS1) protein of influenza A virus. J. Clin.
19. Swayne, D. E., M. Garcia, J. R. Beck, N. Kinney, and D. L. Suarez. Microbiol. 43:676–683. 2005.
Protection against diverse highly pathogenic H5 avian influenza viruses 26. Webster, R. G., Y. Kawaoka, J. Taylor, R. Weinberg, and E. Paoletti.
in chickens immunized with a recombinant fowlpox vaccine containing an Efficacy of nucleoprotein and haemagglutinin antigens expressed in fowlpox
H5 avian influenza hemagglutinin gene insert. Vaccine 18:1088–1095. virus as vaccine for influenza in chickens. Vaccine 9:303–308. 1991.
2000. 27. Webster, R. G., J. Taylor, J. Pearson, E. Rivera, and E. Paoletti.
20. Swayne, D. E., and D. A. Halvorson. Influenza. In: Diseases of Immunity to Mexican H5N2 avian influenza viruses induced by a fowl pox-
poultry, 11th ed. Y. M. Saif, H. J. Barnes, A. M. Fadly, J. R. Glisson, L. R. H5 recombinant. Avian Dis. 40:461–465. 1996.
McDougald, and D. E. Swayne, eds. Iowa State University Press, Ames, IA. 28. Zhao, S., M. Jin, H. Li, Y. Tan, G. Wang, R. Zhang, and H. Chen.
pp. 135–160. 2003. Detection of antibodies to the nonstructural protein (NS1) of avian
21. Swayne, D. E., C. W. Lee, and E. Spackman. Inactivated North influenza viruses allows distinction between vaccinated and infected
American and European H5N2 avian influenza virus vaccines protect chickens. Avian Dis. 49:488–493. 2005.
chickens from Asian H5N1 high pathogenicity avian influenza virus. Avian
Pathol. 35:141–146. 2006.
22. Swayne, D. E., D. A. Senne, and C. W. Beard. Influenza. In: ACKNOWLEDGMENTS
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 We thank J. Beck, J. Doster, and K. Moresco for technical assistance.
Association of Avian Pathologists, Kennett Square, PA. pp. 150–155. 1998. Mention of trade names or commercial products in this publication is
23. Taylor, J., R. Weinberg, Y. Kawaoka, R. G. Webster, and E. Paoletti. solely for the purpose of providing specific information and does not
Protective immunity against avian influenza induced by a fowlpox virus imply recommendation or endorsement by the U.S. Department of
recombinant. Vaccine 6:504–508. 1988. Agriculture. This research was supported by Trust Agreement #58-
24. Thompson, J. D., D. G. Higgins, and T. J. Gibson. CLUSTAL W: 6612-1-0220 between the U.S. Department of Agriculture–Agricultural
improving the sensitivity of progressive multiple sequence alignment Research Service and Merial, Inc.