Journal of Veterinary Diagnostic Investigation Vol. 18 Issue 5, 437-442
Copyright © 2006 by the American Association of Veterinary Laboratory Diagnosticians
Detection of Antibodies in Serum and Egg Yolk Following Infection of Chickens with an H6N2 Avian Influenza Virus
Darrell W. Trampel1,
En-Min Zhou,
Kyoung-Jin Yoon and
Kenneth J. Koehler
Correspondence: 1 Corresponding Author: Dr Darrell W Trampel, 1802 Elwood Drive, VMRI-Bldg 1, Iowa State University, Ames, Iowa 50011-1240
 |
Abstract
|
|---|
Active serologic surveillance programs to detect avian influenza viruses (AIVs) in table egg-laying chickens have been initiated by several states as a response to the economic threat posed by these viruses. Most outbreaks of avian influenza in domestic poultry are caused by mildly pathogenic AIVs. In the study reported here, infection by an H6N2 AIV was used as a model of mildly pathogenic AIV infections in egg-type chickens. The total number of eggs laid by 5 control hens was 619 or 0.904 eggs/day/hen, whereas the total number laid by 10 infected hens was 1,018 or 0.743 eggs/day/hen. The difference in egg production between the 2 groups was not statistically significant (P = 0.38). Anti-influenza antibodies were monitored by use of an agar gel immunodiffusion test and an ELISA for a period of 20 weeks after inoculation. Antibodies in serum developed sooner, peaked at higher levels, and remained at higher levels than did antibodies found in egg yolk, as indicated by ELISA results. For infected chickens, the correlation between serum and egg yolk ratios was 0.66. Serum samples would appear to be preferable to egg yolk samples for surveillance programs intended to identify chicken flocks that may have been infected by an AIV weeks or months before samples are collected.
Key Words: Agar gel immunodiffusion • antibodies • avian influenza virus • chickens • eggs • ELISA • serum • yolk
|
Introduction
|
|---|
Commercial table egg-laying chicken flocks have become infected with mildly pathogenic avian influenza (MPAI) viruses, also called low pathogenic avian influenza viruses (AIVs),14 with increasing frequency in recent years. Furthermore, there is some evidence that these viruses may be adapting to layer populations.11,18 Isolation of an H4N8 subtype (A/chicken/Alabama/75) from laying hens in 1975 represented the first diagnosis of influenza in chickens since a fowl plague outbreak in 1929.7,13 In 1978, commercial laying hens in central Minnesota were found to be infected with H6N1virus.5 Pennsylvania has experienced 3 episodes of MPAI involving egg-type chickens since 1980: 1) H5N2 in 1983–84 that later evolved to become highly pathogenic; 2) H7N2 between 1996 and 1998; and 3) H2N2 in 2004 that did not induce clinical signs of disease.6,12 Subtype H6N2 was isolated from chickens at egg farms in Southern California between February 2000 and September 2003.9,17 Three distinct genotypes of H6N2 viruses were identified in California chickens, and all contained an 18-amino acid deletion in the neuraminidase, a characteristic associated with other influenza viruses isolated from chickens.16 In Connecticut, an H7N2 influenza virus was isolated from chickens at 4 egg farms owned by the same company between February and June of 2003.1 Even though these viruses were classified as having low pathogenicity, significant mortality, morbidity, and economic losses were associated with most of these outbreaks.
Active serologic surveillance programs to detect infection by AIVs in table egg-laying chickens have been initiated by several states as a response to the economic threat posed by these viruses. A voluntary US Avian Influenza Clean program offered by the National Poultry Improvement Plan (NPIP) is available to assist in prevention and control of influenza in egg-type chicken breeding flocks.2 A national program administered by the NPIP to include commercial table egg-laying chickens has been proposed. These programs are intended for early detection of inapparent infections so that control procedures to prevent spread of the virus can be quickly implemented. Surveillance programs typically use serologic screening tests to detect antibodies against type-A influenza viruses, and positive serum samples are sent to the National Veterinary Services Laboratories in Ames, Iowa for subtyping. The ELISA and agar gel immunodiffusion (AGID) tests have been found to be more sensitive than hemagglutination-inhibition tests in detecting anti-influenza antibodies in serum and egg yolk from laying hens.3
In the study reported here, infection by an H6N2 AIV was used as a model of MPAI virus infections in egg-type chickens. Objectives were to: 1) determine the effects of infection by the H6N2 virus on egg production; 2) compare the level and duration of the antibody response in serum and egg yolk; and 3) evaluate the ability of an ELISA to detect antibodies against MPAI viruses in egg yolk.
|
Materials and Methods
|
|---|
Experimental Design
Fifteen 27-week-old Hy-Line W36 white Leghorn chickens in peak egg production were purchased from a commercial egg production facility with no history of avian influenza and good biosecurity. Hens were placed in individual cages and were provided feed and water ad libitum. On the day after their arrival, blood samples were taken from all 15 chickens and were tested for antibodies against AIV by use of the AGID test. All chickens tested negative. The chickens were allowed to acclimate to their new environment for 4 weeks prior to initiation of the experiment. After the acclimation period, 10 chickens were inoculated with H6N2 AIV by placing 0.1 ml of inoculum on each eye and on both external nares. Five uninoculated hens served as controls and were housed in a separate room. Serum samples from all chickens were tested for anti-influenza antibodies by the AGID test and the ELISA at postinoculation (PI) wk 1–8, 12, 16, and 20. Eggs from inoculated and control chickens were collected daily and labeled with the date and hen number until PI wk 20, when the experiment was terminated. Yolk from all eggs was tested for antibodies against type-A influenza viruses by the AGID test and the ELISA.
Virus
An MPAI virus of subtype H6N2, designated A/chicken/California/342/03,a was used for this study. The virus was isolated from a trachea/lung tissue pool from 95-wk-old egg-type chickens experiencing a field case of avian influenza. The hemagglutination (HA) titer of the original material received was 1 : 128. On receipt, the virus was propagated in 10-day-old, specific-pathogen-free, embryonating eggs to 107.5 Egg infectious dose-50 (EID50)/0.2 ml. The titer of inoculum was adjusted to 104 EID50/0.2 ml.
Preparation of Egg Yolk
The shell on top of each egg was opened, and 1 ml of yolk was collected with a syringe. The yolk was mixed with equal amount of 0.01 M phosphate-buffered saline (PBS; pH 7.4) and was centrifuged at 1,000 x g for 30 min. The supernatant was collected and used in the AGID test and the ELISA.
Agar Gel Immunodiffusion Test
The AGID test for AIV antibody detection was performed as follows: 50 µl of undiluted serum and yolk samples (prepared as described previously) was added to wells of an agar gel plate, which was incubated at 25°C for 24 hr before the gel was read for antigen-antibody reaction.15
Enzyme-linked Immunosorbent Assay
Antibodies directed against AIVs were detected by use of a commercial ELISA kitb according to the manufacturer's instructions. The format of the kit is an indirect ELISA. According to the manufacturer, the kit has been validated using serum reactive to AIVs of various subtypes (H1–H14) and is able to detect all influenza-A virus exposures in chickens.
Serum and egg yolk samples were diluted 1 : 500 and were incubated at ambient temperature for 30 min. After a washing step, goat antichicken conjugate was added to each well, and the plates were incubated at ambient temperature for 30 minutes. Absorbance at 650 nm values were measured using an ELISA plate reader.c Results were expressed as the ratio between the optical density (OD) generated by the serum or yolk sample being tested (S) and the OD in a well containing a positive-control serum sample (P) provided by the manufacturer (i.e., the sample-to-positive ratio (S/P). The manufacturer recommends that samples with S/P 
0.5 be considered negative for AIV antibody and that samples with S/P > 0.5 be considered positive for AIV antibody when sera are tested.
Statistical Methods
All computations were performed with statistical package.d The ELISA S/P values were plotted against time, and the LOESS nonparametric curve fitting procedure4 was used to estimate mean response curves for infected and control chickens. Various sets of mean response curves were estimated for the serum and egg yolk samples. Least squares regression was used to fit a line for describing the relationship between serum and egg yolk ELISA S/P values. To account for correlation among ELISA S/P values measured for time and the same chicken, the MIXED procedure in SAS was used to compare serum and egg yolk ELISA S/P means for infected and control chickens, with chickens incorporated into the model as random effects.10 Logistic regression8 was used to classify infected and noninfected chickens on the basis of either serum ELISA S/P or egg yolk ELISA S/P. Estimates of sensitivity, specificity, and false-positive and false-negative classification rates were obtained from an approximate cross-validation procedure.
|
Results
|
|---|
Effects of H6N2 Mpai on Egg Production
Average daily egg production of chickens infected with mildly pathogenic H6N2 influenza virus and uninoculated control hens for 20 weeks is shown in Fig. 1. The total number of eggs laid by the 5 control hens was 619 or 0.904 eggs/day/hen. The total number of eggs laid by the 10 infected hens was 1,018 or 0.743 eggs/day/hen. Average daily egg production of infected chickens was 18% less than that of noninfected control hens. However, the difference in egg production between the 2 groups was not statistically significant (P = 0.38) because of variability between infected chickens and the small numbers of chickens in the experiment. Clinical signs of disease were not observed in inoculated or control chickens.
View larger version (11K):
[in this window]
[in a new window]
|
Figure 1 Daily average egg production from 10 chickens infected with mildly pathogenic H6N2 avian influenza virus and from 5 noninfected control chickens for 20 weeks after inoculation.
|
|
Antibody Response in Sera and Yolk
Average serum and egg yolk ELISA S/P values for the 15 hens in the study are shown in Fig. 2. Antibodies against influenza virus were detected by the ELISA in 100% of serum samples at 7 days after virus exposure. At PI day 10, 50% of tested egg yolk samples yielded positive ELISA results. Mean serum S/P values were significantly higher than mean S/P values for egg yolk from infected chickens (P < 0.0001). Antibody levels in serum peaked at PI week 4, whereas antibodies in yolk reached their highest levels at PI week 5. Mean S/P values for the serum and yolk of control chickens remained at < 0.3 throughout the experiment. For infected chickens, the correlation between serum and egg yolk ratios was 0.66. Mathematically, this relationship can be expressed by the following regression formula:
| (18) |
View larger version (15K):
[in this window]
[in a new window]
|
Figure 2 Average serum and egg yolk ELISA S/P values for 10 chickens infected with mildly pathogenic H6N2 avian influenza virus and for 5 noninfected control chickens for 20 weeks after inoculation. S/P = the ratio between the optical density (OD) generated by the sample being tested (S) and the OD in a well containing a positive-control sample (P).
|
|
Results of AGID tests of serum samples from 8 of 10 inoculated chickens were positive for influenza antibodies at PI week 1, and chickens of this group continued to test positive for the duration of the experiment (Fig. 4). At PI week 20, serum from 9 of 10 infected chickens still tested positive by the AGID test. Serum from all 5 uninoculated control chickens tested negative by the AGID test at PI weeks 1–8, 12, 16, and 20. The first AGID-positive egg yolks were detected on PI day 11 in eggs from 4 of 10 infected chickens. By PI day 13, yolks of 7 of the 8 eggs available for AGID testing were positive for influenza antibodies. Yolks of only 3 of 7 eggs from infected chickens available for testing at PI week 20 tested positive by the AGID test.
View larger version (11K):
[in this window]
[in a new window]
|
Figure 4 Average percentage of positive agar gel immunodiffusion test results for serum and egg yolk samples from 10 chickens for 20 weeks after inoculation with H6N2 AIV.
|
|
Diagnostic Performance of the Elisa
The threshold S/P value selected to differentiate between a negative and a positive result markedly affected the outcome and validity of serum and yolk ELISA results. Data for serum samples taken from chickens known to be infected with H6N2 AIV (inoculated group) or known to be uninfected (control group) on the basis of negative serum AGID test results were used to measure the diagnostic performance of the ELISA (Table 1). For example, if the recommended S/P value of 0.5 was used to separate positive and negative test results for serum samples, it was determined that 94.3% of serum samples tested (181/192) were correctly identified as being negative or positive for antibodies against AIV (Table 2).
Approximately 91.7% of all serum samples from inoculated and AGID-positive chickens tested positive (111/121; sensitivity) and 98.6% of all negative samples from uninoculated, AGID-negative chickens tested negative (70/71; specificity). Only 0.9% of all positive ELISA results were incorrect (1/112; false positives) and 12.5% of all negative ELISA results were incorrect (10/80; false negatives). In addition, an egg yolk ELISA classification table was created from data generated by testing eggs from chickens, the infection status of which was known from serum AGID tests (Table 3). If the S/P value of 0.5 recommended for serum samples was used to separate positive and negative tests done on yolk samples, it could be expected that only 81.3% of all test results will be correct, sensitivity will be 70.9%, specificity will be 96.8%, 3.0% of results will be false positives, and 30.8% of yolk samples tested will yield false-negative results.
|
Discussion
|
|---|
Reduced egg production has been associated with field outbreaks of mildly pathogenic H6N2 virus in egg-type laying hens. On 12 commercial table-egg farms in California, egg production decreased 7–40% after infection of the chickens with AIV subtype H6N2.9 Commercial egg production chicken flocks infected with MPAI viruses in Alabama (H4N8), Minnesota (H6N1), and Pennsylvania (H7N2) experienced similar decreases in egg production.5–7 In the study reported here, the degree to which H6N2 virus affected egg production was highly variable among chickens. Each control chicken laid an average of 124 eggs during the 20 weeks of the experiment, and 80% of infected hens produced an average of 126 eggs during the same period. However, egg production in 2 infected hens was severely reduced: only 5 and 6 eggs over the 20 weeks of the experiment. These 2 hens also had serum ELISA S/P consistently >3.0, much higher than that in other infected birds. Hens in an infected flock with the highest serum ELISA S/P values could remain undetected by an egg-testing program because those hens tend to stop producing eggs.
Antibody production of experimental chickens in response to infection by the H6N2 influenza virus was similar to that induced by other mildly pathogenic AIVs. It has been reported that inoculation of white Leghorn chickens with H7N2 MPAI virus resulted in positive serum ELISA results but negative yolk ELISA results at PI day 7.3 By PI day14, serum and yolk had positive ELISA results. Epidemiology of a field outbreak of MPAI H7N2 in chickens indicated that antibody in serum tends to persist for longer than does antibody in egg yolk.6 In one flock, 97% of eggs collected early in the outbreak were positive for avian influenza antibodies by the AGID test, but only 10% of eggs were positive 6 weeks later, and none of 30 eggs was positive 11 weeks after initial testing. In the study reported here, at 20 weeks after infection, AGID test results for serum samples from 90% of inoculated chickens were positive, but only 43% of eggs (3/7) from infected hens had positive AGID results.
Anti-influenza antibodies in serum developed sooner, peaked at higher levels, and remained at higher levels than antibodies found in egg yolk, as indicated by ELISA results (Fig. 1). Testing of yolk samples by the ELISA procedure tends to yield a high percentage of false-negative results, which could allow some infections to remain undetected. Furthermore, less than half of egg yolk samples from infected hens tested positive by the AGID test at 20 weeks after infection. Obtaining serum samples for influenza testing requires more time and effort than does submitting eggs for testing. However, serum samples would appear to be preferable to egg yolk samples for surveillance programs intended to identify chicken flocks that may have been infected by an AIV weeks or months before samples were collected.
The sensitivity and specificity of ELISAs for serum and yolk were dependent on the S/P value selected as the threshold level for defining a positive test result. When a sample was tested, 4 outcomes were possible: 1) correct identification of positive samples, 2) false-positive results for negative samples, 3) false-negative results for positive samples, and 4) correct identification of negative samples (Table 2). As can be seen from Tables 1 and 3, an inverse relationship exists between sensitivity and specificity and between false positives and false negatives. A lower S/P threshold will increase false positives and reduce false negatives, whereas a higher S/P cut-off value has the opposite effect. For the ELISA of serum samples generated in this experiment, selecting 0.45 as the S/P value for defining a positive test instead of 0.50 provided a higher percentage of correct test results (96.9% vs. 94.3%) and markedly lowered the false negatives (5.5% vs. 12.5%). When testing egg yolks, selecting the S/P value (0.25) that yields the highest percentage of correct results (88.9%) still results in 10.9% false-positive and 11.3% false-negative ELISA results (Table 3).
View larger version (9K):
[in this window]
[in a new window]
|
Figure 3 Relationship between mean ELISA S/P in egg yolk and mean ELISA S/P in serum. S/P = the ratio between the optical density (OD) generated by the sample being tested (S) and the OD in a well containing a positive-control sample (P).
|
|
|
Acknowledgments
|
|---|
The authors acknowledge and thank Dr. P. R. Woolcock (University of California, Fresno Laboratory) for providing the H6N2 isolate of AIV used in this research.
|
Sources and manufacturers
|
|---|
From the Department of Veterinary Diagnostic and Production Animal Medicine, College of Veterinary Medicine (Trampel, Zhou, Yoon), and the Department of Statistics, College of Liberal Arts and Sciences (Koehler), Iowa State University, Ames, IA 50011-1240. 
a. Provided by Dr. Peter Woolcock, Fresno Laboratory, Fresno, CA.
b. FlockCheck Avian Influenza Antibody Test Kit, IDEXX Laboratories, Inc., Westbrook, ME.
c. Universal Microplate Reader (EL 800), Bio-Tek Instrument, Inc., Winooski, VT.
d. SAS, version 9.1, SAS Institute, Inc., Cary, NC.
|
References
|
|---|
- Adriatico N.A.: 2005, Experiences using H7 avian influenza vaccination in Connecticut. In: Proceedings of the 54th Western Poultry Disease Conference, April 25–27, 2005. pp. 1–3. Vancouver, BC, Canada: American Association of Avian Pathologists, Athens. GA.
- Anonymous. National poultry improvement plan and auxiliary provisions: February 2004, United States Department of Agriculture, Animal and Plant Health Inspection Service, Washington, DC, 91-55-063.
- Beck J.R., Swayne D.E., Davison S., et al.: 2003, Validation of egg yolk antibody testing as a method to determine influenza status in white Leghorn hens. Avian Dis 47:1196–1199.[Medline]
- Cleveland W.S., Devlin S.J., Grosse E.: 1988, Regression by local fitting. J Econom 37:87–114.
- Halvorson D.A., Karunakaran D., Newman J.A.: 1980, Avian influenza in caged laying chickens. Avian Dis 24:288–294.
- Henzler D.J., Kradel D.C., Davison S., et al.: 2003, Epidemiology, production losses, and control measures associated with an outbreak of avian influenza subtype H7N2 in Pennsylvania (1996–98). Avian Dis. 47:1022–1036.[Medline]
- Johnson D.C., Maxfield B.G.: 1976, An occurrence of avian influenza virus infection in laying chickens. Avian Dis 24:422–424.
- Johnson R.A., Wichern D.W.: 2002, Applied multivariate statistical analysis, 5th ed. Prentice Hall, Upper Saddle River, NJ.
- Kinde H., Read D.H., Daft B.M., et al.: 2003, The occurrence of avian influenza A subtype H6N2 in commercial layer flocks in southern California (2000–2002): clinicopathologic findings. Avian Dis 47:1214–1218.[Medline]
- Little R.C., Milliken G.A., Stroup W.W., Wolfinger R.D.: 1996, SAS system for mixed models. SAS Institute Inc, Cary, NC.
- Liu M., Guan Y., Peiris M., et al.: 2003, The quest of influenza A viruses for new hosts. Avian Dis 47:849–856.[Medline]
- Senne D.A.: 2005, Avian influenza diagnostic summary. In: Proceedings of the 108th Annual Meeting of the US Animal Health Association, October 21–28, 2004. pp. 544–547. US Animal Health Association, Richmond, VA, Greensboro, NC.
- Shalaby A.A., Slemons R.D., Swayne D.E.: 1994, Pathological studies of A/Chicken/Alabama/7395/75 (H4N8) influenza virus in specific-pathogen-free laying hens. Avian Dis 38:22–32.[Medline]
- Swayne D.E., Halvorson D.A.: 2003, Influenza. In: Diseases of poultry, 11th ed. pp. 135–160. Iowa State Press, Ames, IA.
- Swayne D.E., Senne D.A., Beard C.W.: 1998, Avian influenza. In: A laboratory manual for the isolation and identification of avian pathogens, 4th ed. pp. 150–155. Iowa State Press, Ames, IA.
- Webby R.J., Woolcock P.R., Krauss S.L., et al.: 2003, Multiple genotypes of nonpathogenic H6N2 influenza viruses isolated from chickens in California. Avian Dis 47:905–910.[Medline]
- Whiteford A.: 2005, Symposium on avian influenza. In: Proceedings of the 108th Annual Meeting of the US Animal Health Association, October 21–28, 2004. p. 558. Greensboro, NC: US Animal Health Association, Richmond, VA.
- Woolcock P.R., Suarez D.L., Kuney D.: 2003, Low-pathogenicity avian influenza virus (H6N2) in chickens in California, 2000–02. Avian Dis 47:872–881.[Medline]
This article has been cited by other articles:
|
|
|
|
 
S. Gharaibeh, K. Mahmoud, and M. Al-Natour
Field Evaluation of Maternal Antibody Transfer to a Group of Pathogens in Meat-Type Chickens
Poult. Sci.,
August 1, 2008;
87(8):
1550 - 1555.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
Abstract
Introduction
