key: cord-277804-ujabzic4
authors: Yuk, Seong-su; Kwon, Jung-Hoon; Noh, Jin-Yong; Hong, Woo-tack; Gwon, Gyeong-Bin; Jeong, Jei-Hyun; Jeong, Sol; Youn, Ha-Na; Heo, Yong-Hwan; Lee, Joong-Bok; Park, Seung-Yong; Choi, In-Soo; Song, Chang-Seon
title: Comparison between dot-immunoblotting assay and clinical sign determination method for quantifying avian infectious bronchitis virus vaccine by titration in embryonated eggs
date: 2016-01-21
journal: J Virol Methods
DOI: 10.1016/j.jviromet.2016.01.008
sha: 
doc_id: 277804
cord_uid: ujabzic4

A sensitive and specific method for measuring the vaccine titer of infectious bronchitis virus (IBV) is important to commercial manufacturers for improving vaccine quality. Typically, IBV is titrated in embryonated chicken eggs, and the infectivity of the virus dilutions is determined by assessing clinical signs in the embryos as evidence of viral propagation. In this study, we used a dot-immunoblotting assay (DIA) to measure the titers of IBV vaccines that originated from different pathogenic strains or attenuation methods in embryonated eggs, and we compared this assay to the currently used method, clinical sign evaluation. To compare the two methods, we used real-time reverse transcription-PCR, which had the lowest limit of detection for propagated IBV. As a clinical sign of infection, dwarfism of the embryo was quantified using the embryo: egg (EE) index. The DIA showed 9.41% higher sensitivity and 15.5% higher specificity than the clinical sign determination method. The DIA was particularly useful for measuring the titer of IBV vaccine that did not cause apparent stunting but propagated in embryonated chicken eggs such as a heat-adapted vaccine strain. The results of this study indicate that the DIA is a rapid, sensitive, reliable method for determining IBV vaccine titer in embryonated eggs at a relatively low cost.

Infectious bronchitis virus (IBV) is a gamma coronavirus that causes a highly contagious disease in chickens. The virus causes severe economic losses to the poultry industry worldwide because it can affect the upper respiratory and reproductive tracts, and some strains can cause nephritis in chickens (Jackwood, 2012) . Despite intensive vaccination using both live attenuated and killed vaccines to prevent the disease, the emergence of new variant strains that do not serologically cross-react has complicated disease control and demonstrates the importance of vaccinating chickens with the disease-causing IBV types (Cavanagh, 2005 (Cavanagh, , 2007 .

Vaccination is considered the most cost-effective approach for controlling IBV infection (Meeusen et al., 2007) . To prevent the economic losses caused by IBV, live attenuated vaccines and inactivated oil-emulsion vaccines containing both the KM91 and Massachusetts 41 (M41) strains are widely used in Korea. More recently, vaccine strains providing broad cross-protection have been developed, including the K2 and K40/09 strains (Kim et al., 2013; Lim et al., 2012 Lim et al., , 2015 . In Korea, the viral content of a vaccine preparation is quantified by IBV titration according to standard procedure approved by the Animal and Plant Quarantine Agency. For this procedure, the chicken embryos are inoculated with serial dilutions of the virus preparation, and then the embryos are examined for the presence of specific lesions caused by the virus, i.e., dwarfing, curling, and stunting (Doherty, 1967) . However, as a variety of live-attenuated vaccines have been developed for variant field strains, it is unclear whether the clinical sign determination method reflects viral propagation in IBV-infected embryos. Furthermore, as there are no specific standards for measuring clinical signs, the measurements of specific lesions can differ among observers.

Our previous study revealed that a novel dot-immunoblotting assay (DIA) using monoclonal antibodies against several IBVs could detect viruses propagated in embryonated chicken eggs (Song et al., 1998) . To accurately, reproducibly, and efficiently measure vaccine titers, we used the DIA to detect IBV propagated in inoculated http://dx.doi.org/10.1016/j.jviromet.2016.01.008 0166-0934/© 2016 Elsevier B.V. All rights reserved. embryonated eggs. The aim of this study was to evaluate and compare the sensitivity and specificity of the DIA to the clinical sign determination method for detecting IBV in inoculated embryonated eggs during titration of IBV vaccines.

A respiratory strain belonging to the Mass group (M41), a nephropathogenic strain belonging to the KM91-like subgroup (KM91), and a recombinant nephropathogenic strain belonging to Korean new cluster 1 (K40/09) were used to evaluate the titer of IBVs used in killed vaccines as described by Kim et al. (2013) . Two nephropathogenic strains that were attenuated by 170 serial passages (K2p170) or heat-adapted passages (K40/09HP40) in chicken embryos were used to evaluate the live-attenuated vaccine strains (Lee et al., 2010) . All viruses were propagated in 10-day-old specific-pathogen-free embryonated chicken eggs (ECE; Hy-Vac, Adel, IA) at 37 • C for 72 h. Allantoic fluid from each egg was harvested, aliquoted, and frozen at −70 • C until use.

All animal protocols used in this study were reviewed, approved, and supervised by the Institutional Animal Care and Use Committee of Konkuk University. To confirm the detection limit of the DIA, each vaccine strain was serially diluted 2-fold in phosphatebuffered saline (PBS; Invitrogen, Carlsbad, CA) and analyzed by DIA and real-time reverse transcription (RT)-PCR. Concurrently, triplicate 10-fold serially diluted samples were used to determine the detection limit of real-time RT-PCR. To measure vaccine titer based on the infectivity of chicken embryos, 10-fold serial dilutions of a 10 0 virus solution (10 −1 -10 −9 ) were generated by mixing 1 mL of the virus with 9 mL of PBS containing 50 g/mL gentamicin sulfate (Sigma-Aldrich, St. Louis, MO); all dilutions were kept on ice. Next, 0.1 mL of the 10 −4 through 10 −9 dilutions were inoculated into five 10-day-old SPF ECE (Hy-Vac), and the eggs were incubated at 37 • C. No eggs had to be discarded due to non-specific death of embryos within one day of inoculation. After three days of incubation, 500 L of allantoic fluid was extracted from the inoculated eggs using a 1 mL syringe and was used to detect propagated IBV by the DIA and real-time RT-PCR. Eggs were then resealed with paraffin for further incubation. Seven days after inoculation, the embryo:egg (EE) index was calculated for all eggs. Propagation of the inoculated virus was determined using real-time RT-PCR, DIA, and the EE index method, and the 50% egg infectious dose (EID 50 ) of the five vaccines was calculated based on the method of Reed and Muench (1938) .

RNA was extracted from each inoculated allantoic fluid sample using the ExiPrep viral RNA/DNA extraction kit (BIONEER Co., Daejeon, Korea) according to the manufacturer's instructions, and eluted in a 50 L volume. Real-time RT-PCR analysis of the extracted RNA samples was conducted as previously described (Callison et al., 2006) . The primers and TaqMan dual-labeled probe were synthesized by Macrogen (Seoul, Korea). The primers and probe were utilized in a 25 L reaction containing 12.5 L of Quantitect Probe RT-PCR 2× mix (Qiagen, Hilden, Germany), 0.25 L of RT enzyme (Qiagen), 0.5 mol of primers, 0.1 mol of probe, and 5 L of the RNA sample from 50 L of extracted RNA. Amplification was performed in an ABI Prism 7500 real-time PCR System using the following program: 50 • C for 30 min, 95 • C for 15 min, and 45 cycles of 94 • C for 1 s and 60 • C for 60 s, with emitted fluorescence measurement. Cycle threshold (Ct) values below the detection limit were considered positive for IBV.

The DIA was conducted as described previously with slight modifications (Song et al., 1998) . Allantoic fluids from ECEs inoculated with the serially diluted vaccines were centrifuged at 1000 × g for 1 min. Next, 150 L of the supernatant was dotted onto a nitrocellulose membrane (0.45 m pore size) using a Hybri-Dot 96-well filtration manifold (Bio-Rad Laboratories, Hercules, CA). Uninfected normal allantoic fluid was used as a negative control. The membrane was then blocked with 3% bovine serum albumin in Tris-buffered saline (TBS; 100 mM Tris, 0.9% sodium chloride, pH 7.5) at 37 • C overnight. The primary monoclonal antibodies (3F5) were diluted 1:1000 with Tris Tween-buffered saline (TTBS; 100 mM Tris, 0.2% Tween 20, 0.9% sodium chloride, pH 7.5) and incubated with the membrane for 30 min at 37 • C. After washing the membrane in TTBS three times for 15 min each with gentle agitation, the membrane was incubated for 30 min at 37 • C with biotinylated anti-mouse immunoglobulin G (Vectastain ABC kit; Vector Laboratories Inc., Burlingame, CA) diluted in TTBS. Following washing, the membrane was incubated for 30 min at 37 • C with biotin and avidin-conjugated peroxidase complex (Vectastain ABC kit) diluted in TTBS. After the final washing, the membrane was developed using 6 mg of diaminobenzidine (Pierce, Rockford, IL) in 10 mL of 50 mM Tris and 10 L H 2 O 2 for 1 min. The reaction was then stopped by rinsing with distilled water (three times) and the membrane was allowed to air-dry. Dark brown-colored dots were considered positive for IBV.

Dwarfing of the infected chicken embryos was detected by determining the EE ratio. The weights of the eggs were measured 7 days after inoculation of the serially diluted viruses. The respective eggs and embryos were weighed using an electronic balance (Ohaus, Florham Park, NJ). The EE ratio was defined as the weight of embryo divided by the weight of respective egg. A total of 120 SPF eggs (Hy-Vac) were used to determine the EE ratio of control ECE (mock-inoculated). Briefly, we inoculated 10-day-old eggs with 100 L of sterile PBS as a mock-inoculated control. The EE ratio of 20 eggs was determined at 7 days post-inoculation. The average EE ratio and the standard deviation of the mock control were calculated. The EE index was determined by dividing the EE ratio of inoculated eggs by the mean EE ratio of mock-inoculated eggs (Dhinakar Raj et al., 2004) .

To assess the receiver operating characteristics (ROC) using MedCalc ® version 15.8 statistical software (Mariakerke, Belgium), positive or negative results confirmed by real-time RT-PCR were assigned as "true-positive" or "true-negative," respectively. The area under the ROC-curve (AUC index) was then calculated for both the DIA and clinical sign detection methods. High values (close to 1) indicate a highly accurate test (Greiner et al., 2000) . The concordance of both assays to the properly classified samples as positive or negative was estimated by calculating the weighted kappa statistic (Ä test). Ä test values of 0.41-0.60 indicate moderate agreement, values of 0.61-0.80 indicate substantial agreement, and values of 0.81-0.99 indicate nearly perfect agreement (Viera and Garrett, 2005) .

The 2-fold diluted virus was detected concurrently by DIA and real-time RT-PCR (Supplementary material Fig. 1 ). The highest dilution that showed dark-brown dots in the DIA was 2 −7 for K2p170 and K40/09 and 2 −6 for M41, KM91, and K40/09HP40. The 10-fold diluted viruses were also examined by real-time RT-PCR (Supplementary material Fig. 2 ). For the tested pathogenic or attenuated IBVs, there were no detectable Ct values at the 10 −7 dilution. Therefore, the detection limit of the real-time RT-PCR assay was set to the Ct value of the 10 −6 dilution. The real-time RT-PCR method was found to be approximately 2 15 -fold more sensitive than the DIA method. Thus, real-time RT-PCR is suitable for detecting propagated viruses and can be used to compare the DIA and EE ratio detection methods for propagated IBV during titration. Because the DIA does not distinguish between infectious and noninfectious virus, we had to confirm that the assay does not detect inoculated virus in the eggs. As shown by the detection limit of the DIA, we observed that the input virus was not detected at dilutions higher than 2 −6 -2 −7 . Thus, none of the dilutions used in the titering (10 −4 -10 −9 ) would give a positive result without viral replication, especially after further dilution in the allantoic fluid. After 7 days of incubation, the mean EE ratio and standard deviation of mock-inoculated eggs was 0.375 ± 0.036. To detect dwarfism, the EE index was defined as the EE ratio of IBV-inoculated ECE divided by the mean EE ratio of mock-inoculated ECE. Embryos with EE indices below the lowest individual EE index for mock-inoculated ECE (0.880) were classified as positive for dwarfism. Supplementry material related to this article found, in the online version, at http://dx.doi.org/10.1016/j.jviromet.2016.01.008.

Based on the results of real-time RT-PCR, the titer of respiratory the type IBV M41 strain was 10 8.4 EID 50 /mL ( Table 1) . Four of the five eggs inoculated with the 10 −7 dilution showed positive results. The DIA results were identical to the real-time RT-PCR results, even the individual egg results. In contrast, only three of the five eggs inoculated with the 10 −6 dilution were positive by the EE index. The weight of the other two eggs exceeded the lowest value of the individual EE index of mock-inoculated ECE (0.885, 0.890). Because one embryo inoculated with the 10 −8 dilution died 3 days after inoculation, we excluded this egg from the EE index measurements. One egg inoculated with the 10 −9 dilution showed a positive result for the EE index (0.800). The titer of M41 obtained using the EE index was lower than that obtained using the other assays (10 8.3 EID 50 /mL).

According to real-time RT-PCR analysis, the titer of the Korean nephrogenic type IBV (KM91) and its attenuated form (K2p170) were both 10 7.0 EID 50 /mL. Based on the individual values, four of the five eggs inoculated with the 10 −5 dilution and three of the five eggs inoculated with the 10 −6 dilution showed positive results for both viruses. The DIA results were identical to the real-time RT-PCR results, including the individual egg results. Unexpectedly, in the KM91 titration, one egg at the 10 −4 dilution showed a negative result (0.933) for the EE index, despite showing positive results in the other assays. Because one embryo inoculated with the 10 −5 dilution died 3 days after inoculation, we excluded this egg from the EE index measurements. EE indices of two of the eggs inoculated with the 10 −7 virus dilution (0.878, 0.833) were slightly below the lowest individual EE index of the mock-inoculated EE index. The titer of KM91 determined using the EE index was higher than that determined using the other assays (10 7.3 EID 50 /mL). In the K2p170 titration, the results according to the EE indices were identical to those obtained by real-time RT-PCR, except for one egg inoculated with the 10 −7 dilution (0.742). Accordingly, the titer of K2p170 determined using the EE index was higher than that determined using other assays (10 7.2 EID 50 /mL).

Based on the real-time RT-PCR results, the titers of the recombinant nephrogenic type IBV K40/09 and its heat-attenuated form K40/09HP40 were 10 7.8 and 10 8.2 EID 50 /mL, respectively ( Table 1) . Two of the five eggs inoculated with the 10 −7 dilution of the K40/09 were positive. However, the Ct value of one egg inoculated with the 10 −7 dilution of K40/09HP40 (27.96) was higher than the detection limit of the DIA (21.07) but lower than that of real-time RT-PCR (35.05). Thus, the egg was negative according to the DIA. In the K40/09 titration, two embryos inoculated with the 10 −7 and 10 −8 dilutions died at four and five days after inoculation, respectively; therefore, we excluded these eggs from the EE index measurement. The EE index of one egg inoculated with the 10 −8 dilution (0.747) was positive. In the K40/09HP40 titration, one embryo inoculated with the 10 −6 dilution died at two days after inoculation and was excluded from the EE index measurement. The EE indices of two eggs inoculated with the 10 −5 dilution (0.881, 0.907) and one egg inoculated with the 10 −6 dilution (1.159) were negative. Two eggs inoculated with the 10 −7 dilution (0.984, 0.919) and one egg inoculated with the 10 −8 dilution (0.938), which were positive according to real-time RT-PCR, did not show clear dwarfism. Consequently, the three assays showed three different titers. When compared to the titer measured by real-time RT-PCR (10 8.2 EID 50 /mL), the titers measured by DIA (10 7.9 EID 50 /mL) and EE index (10 7.2 EID 50 /mL) were 2-fold and 10-fold lower, respectively.

The performance of the two methods was also compared ( Table 2) . DIA detected propagated IBV in 77 (97.25%) of 79 positive samples, with perfect specificity (100%) and an AUC index of 98.7%. The value for this test was 0.973, indicating that the values were highly consistent with the real-time RT-PCR results. Of the 79 true-positive samples, two samples in K40/09 and K40/09HP40 were false-negatives by DIA, which was also observed using the clinical signs method. The results of the EE index, which was defined by the cut-off value of the lowest EE index of the mockinoculated ECE (0.880), showed propagated IBV in 65 (87.84%) of 74 positive samples, with relatively low specificity (84.50%) and an AUC index of 86.2%. The Ä value for this test was 0.724, indicating moderate agreement with the real-time RT-PCR results. Of the nine false-negative results obtained using the clinical sign method, two samples were M41, four samples were KM91, two samples were K2p170, and one sample was K40/09. Of the 11 false-positive results obtained using the clinical sign method, three samples were M41, two samples were KM91, one sample was K2p170, and five samples were K40/09HP40. For the EE index results, which were obtained using the cut-off value of the lowest individual EE index of the mock-inoculated ECEs, we compared the cut-off value to the optimum value calculated based on the EE index distribution. The optimum cut-off (0.890) calculated based on the EE indices distribution suggested that the cut-off value of the lowest individual EE index of the mock-inoculated ECEs (0.880) was reliable and could be used to classify clinical signs by weight. Using this new cut-off, only one false-negative sample was corrected to a positive result, and the assay showed 89.19% sensitivity and 85.92% specificity (Fig. 1) . We compared the clinical signs and DIA results to the real-time RT-PCR results and found that the DIA results were highly consistent with the real-time RT-PCR results. Among the 150 eggs inoculated with serially diluted IBV samples, only two samples showed false-negative results. However, 20 results obtained using the clinical signs method did not match the real-time RT-PCR results. For the M41 respiratory strain, the titer calculated by the clinical signs method were lower than the titer calculated by real-time RT-PCR. The nephropathogenic strain KM91 and its attenuated form K2p170 showed higher titer when measured by the clinical signs method. Compared to KM91, K2p170 showed clearer signs of stunting. This demonstration that egg-adapted strains induce clearer embryo dwarfism than wild strains is in agreement with those of Dhinakar Raj et al. (2004) . However, the recombinant nephrogenic strain K40/09 and its attenuated counterpart K40/09HP40, which was egg-passaged with heat adaptation, showed opposite results. The overall EE indices were higher for K40/09HP40 than for K40/09. Moreover, the K40/09HP40 titer measured by the clinical signs method was approximately 10-fold lower than the titer measured by real-time RT-PCR. These results suggest that the clinical signs induced by this heat-adapted virus might not be sufficient to produce dwarfism in ECEs.

The extensive genetic diversity and high mutation rate of IBV generate many different virus types that the existing IBV vaccine strains do not protect against. Thus, there is great demand for a multiple serotype IBV vaccine that provides broad protection. In addition, several research groups have attempted to develop a multivalent vaccine against not only multiple IBV serotypes but also other avian infectious diseases such as Newcastle disease, avian influenza, and infectious laryngotracheitis, as well as nonrespiratory diseases such as Marek's disease and fowl pox (Ndegwa et al., 2015; Sharma et al., 2002; Vagnozzi et al., 2010; Winterfield, 1968) . For such multivalent vaccines, which must be mixed accurately to induce a protective immune response, it is thought that the higher or lower titers observed by the clinical sign method (e.g., for the nephropathogenic strain or heat-adapted strains) could prompt manufacturers to add less or more antigen than required, respectively. This is especially important for live vaccines, because the virus first infects the host and then induces an immune response via innate and acquired immunity, the virus quantitation must represent the true propagating ability within the host.

In Korea, the currently approved methods for quantifying freeze-dried live IBV vaccine or killed IBV vaccine before virus inactivation state that quantification should be determined by observing infectious bronchitis-induced lesions at seven days after inoculation of serially diluted vaccine into 10-day-old ECE (QIA, 2015) . Generally, the disease-induced lesions are clinical signs, including stunting, curling, and clubbed down. In the US, the method for titrating IBV vaccines also states that on the seventh day after virus inoculation, all eggs should be examined for IBV lesions (USDA, 2014) . In this method, specific lesions, such as bile stasis and kidney urates, are listed as general clinical signs. However, assessing specific lesions is laborious and time consuming. Moreover, well-trained observers who can recognize the clinical signs and can skillfully approach without damaging them are required. In this study, we choose stunting as a typical clinical sign because it is the most objective and quantifiable clinical sign, and it can be easily distinguished without observer error.

Real-time RT-PCR is a reliable method because its detection limit is lower than that of other methods. However, the fluorescence thermal cycler requires continuous maintenance and trained operators who can perform the experiments without introducing cross-contamination between samples, which is very costly. Accurate measurement of vaccine titer could help manufacturers increase vaccine production or prevent vaccine failure resulting from antigen content that is insufficient to induce an immune response. The DIA method could be used to measure the titer of IBV vaccines that do not elicit clear stunting signs but propagate in ECEs, such as the heat-adapted vaccine strain. In addition, this method is faster than clinical signs method, and could reduce inter-observer differences. The results indicate that the DIA presented in this study is a relatively low cost, timesaving, sensitive, and reliable method for detecting IBV during IBV vaccine titration in embryonated eggs.

Development and evaluation of a real-time Taqman RT-PCR assay for the detection of infectious bronchitis virus from infected chickens

Coronaviruses in poultry and other birds

Coronavirus avian infectious bronchitis virus

Egg:embryo weight ratio as an indicator of dwarfism induced by infectious bronchitis virus

Titration of avian infectious bronchitis virus in the tissues of experimentally infected chickens

Principles and practical application of the receiver-operating characteristic analysis for diagnostic tests

Review of infectious bronchitis virus around the world

Cross-protective immune responses elicited by a Korean variant of infectious bronchitis virus

Characterization of a novel live attenuated infectious bronchitis virus vaccine candidate derived from a Korean nephropathogenic strain

Live attenuated nephropathogenic infectious bronchitis virus vaccine provides broad cross protection against new variant strains

Successful cross-protective efficacy induced by heat-adapted live attenuated nephropathogenic infectious bronchitis virus derived from a natural recombinant strain

Current status of veterinary vaccines

Combined infectious bronchitis virus Arkansas and Massachusetts serotype vaccination suppresses replication of Arkansas vaccine virus

Korean Standard Assay of Veterinary Biological Products, Chicken Infectious Bronchitis Freeze-dried Live Vaccine Standard Examination

A simple method of estimating fifty per cent endpoints

Field trial in commercial broilers with a multivalent in ovo vaccine comprising a mixture of live viral vaccines against Marek's disease, infectious bursal disease, Newcastle disease, and fowl pox

Detection and classification of infectious bronchitis viruses isolated in Korea by dot-immunoblotting assay using monoclonal antibodies

United States Department of Agriculture Center for Veterinary Biologics Testing Protocol. In: Supplemental Assay Method for the Titration of Newcastle Disease Vaccine, Infectious Bronchitis Vaccine, and Combination Newcastle Disease-Infectious Bronchitis Vaccine in Chicken Embryos. United States Department of Agriculture Animal and Plant Health Inspection Service

Protection induced by infectious laryngotracheitis virus vaccines alone and combined with Newcastle disease virus and/or infectious bronchitis virus vaccines

Understanding interobserver agreement: the kappa statistic

Respiratory signs, immunity response, and interference from vaccination with monovalent and multivalent infectious bronchitis vaccines

This work was supported by a grant from the Agricultural Biotechnology Development Program, Ministry of Agriculture, Food and Rural Affairs, Republic of Korea (no. 314017-03).