Development of an indirect enzyme-linked immunosorbent assay test for detecting antibodies to chicken astrovirus in chicken sera

Abstract

The development of an indirect enzyme-linked immunosorbent assay (ELISA) for the serological diagnosis of Group B chicken astrovirus (CAstV) infections is described. The test was based on the use of an affinity-purified capsid antigen, specific to CAstV isolate 11672, produced as a glutathione-S-transferase N-terminal fusion protein by a recombinant baculovirus. Strongly positive ELISA signals were elicited against experimentally produced antisera raised to CAstVs from Group B (subgroups i and ii) but were negative for antisera raised to a Group A CAstV. Using a panel of 240 selected serum samples, 99% agreement was observed when the results obtained by ELISA were compared to those from an indirect immunofluorescence test for CAstV 11672. The ELISA test was applied to 68 serum sets comprising 1864 samples, which were obtained from parent and grandparent flocks originating mainly in the UK. Of the 52 sets containing ELISA-positive samples, 24 sets had >75% samples positive and nine sets had <25% samples positive and were regarded as having high and low seropositivities, respectively. Of the 1864 serum samples tested 1090 (58.5%) were ELISA positive and of these, 234 sera (21.5%) produced strongly positive signals, whereas moderately positive and weakly positive signals were produced by 562 (51.5%) and 294 (27%) sera. When used for flock screening purposes, this ELISA test can be used to (i) investigate the occurrence of first-time CAstV infections of parent flocks during lay and the possible adverse effects caused by vertically transmitted CAstV infections on broiler hatchability and performance and (ii) diagnose Group B CAstV infections within specific pathogen free flocks.

Introduction

Two different avian astrovirus species are known to infect chickens, avian nephritis virus (ANV) and chicken astrovirus (CAstV). CAstV was first characterized as a novel astrovirus by Baxendale and Mebatsion (2004), who reported the isolation of three antigenically similar viruses, exemplified by the P22-18.8.00 isolate, from outbreaks of runting stunting syndrome that occurred in the Netherlands in the 1980s. Since then CAstVs have been detected by reverse transcription-polymerase chain reaction (RT-PCR) in intestinal samples from growth-retarded and normal broilers reared in the USA, India, UK and other European countries (Day et al., 2007; Smyth et al., 2009, 2010; Bulbule et al., 2013). The existence of two antigenically different CAstVs, typified by the 612 and 11672 isolates, which share low levels of cross-reactivity as determined by indirect immunofluorescence (IF), has been reported by Todd et al. (2009a). Thus, the 612 isolate of CAstV, described by McNeilly et al. (1994), shares a close antigenic and genetic relationship with the first characterized P22-18.8.00 isolate, but differs antigenically and genetically from the 11672 and FP3 CAstV isolates (Todd et al., 2009a). The FP3 isolate, originally recognized as an enterovirus-like virus (ELV), which was isolated from the meconium of dead-in-shell chicks during an investigation of early broiler mortality in the UK (Spackman et al., 1984), is closely related antigenically and genetically to the 11672 isolate, which was isolated in 2005 in our laboratories from one-day-old chicks as part of an investigation into hatchability problems in the UK. A recent study into the capsid protein sequence diversity of 25 CAstVs sourced mainly from the UK, but also including CAstVs from the USA, Europe (isolate P22-18.8.00) and South Africa (isolate 612), showed that CAstVs could be assigned to one of two major capsid groups, designated A and B, which shared 38–40% amino acid identities (Smyth et al., 2012). Not unexpectedly, the antigenically different 612 (Group A) and 11672 (Group B) were identified as belonging to different capsid groups. The sequence comparison also showed that the Groups A and B comprised three and two subgroups, respectively, with 612 being placed in Subgroup A i, 11672 in B i and a third CAstV strain, VF08-29, being representative of Subgroup B ii. The intra-subgroup amino acid identities of the B i and B ii subgroups were 84–85% and those of the A i, A ii and A iii subgroups ranged from 77% to 82% (Smyth et al., 2012).

Our interest in CAstVs relates to their possible involvement in causing enteritis and growth depression in chickens (Guy et al., 2009). They have been detected in a US investigation of gut and faeces samples from healthy broiler chicken flocks and from flocks affected by enteritis and growth problems (Pantin-Jackwood et al., 2007), and in gut content and kidney samples from UK flocks with growth-retardation problems (Smyth et al., 2009, 2010). In addition, CAstVs, originally recognized as ELVs, have also been shown to cause varying degrees of growth retardation following experimental infections of one-day-old specific pathogen free (SPF) or broiler chicks (McNulty et al., 1990; McNeilly et al., 1994). The detection of CAstVs in dead-in-shell embryos (Spackman et al., 1984) and hatched chicks (Smyth et al., 2013) indicates that these viruses can be vertically transmitted. However, the nature and extent of the disease problems caused by CAstVs have not been determined, largely due to the absence of convenient, sensitive and specific diagnostic tests.

The detection of virus-specific antibodies in serum samples provides a convenient way of determining the extents of virus infection within animal populations. We previously reported an investigation into the seroprevalence of CAstV infections in commercial chickens (Todd et al., 2009b). This involved the use of separate indirect IF tests based on the 11672 and 612 isolates, which show low levels of indirect IF cross-reactivity (Todd et al., 2009b). This study showed that infections with the two antigenically different CAstVs were (i) very common in broiler and parent flocks within the UK, (ii) detectable at lower levels in grandparent (GP) and great grandparent flocks, and (iii) widespread in parent flocks belonging to poultry organizations located throughout Europe. Indirect IF tests performed with virus-infected cells grown on glass coverslips are labour-intensive, time-consuming and require input from experienced personnel. Antibody-detecting enzyme-linked immunosorbent assay (ELISA) tests do not have these drawbacks and are much more suitable for testing large sample numbers that are required in flock screening programmes. CAstVs grow relatively poorly in cell culture and attempts to use cell culture-derived virus to produce ELISA antigen of the required purity and in sufficient amounts have met with inconsistent success. We have recently described the production of the CAstV capsid protein (specified by the 11672 isolate) as a glutathione S-transferase (GST) fusion protein in Sf9 insect cells infected with a recombinant baculovirus (Lee et al., 2013). That paper also reported the potential of the affinity-purified GST-11672 capsid to serve as an ELISA antigen for the detection of CAstV-specific antibodies that were induced by experimentally infecting SPF chickens. In this paper we describe the development and evaluation of an ELISA based on this recombinant CAstV capsid antigen.

Materials and Methods

Chicken sera

The samples of chicken serum used in this investigation were derived from five different sources.

1. SPF chickens. SPF eggs, supplied by VALO BioMedia GmbH (Osterholz-Scharmbeck, Germany), were hatched in our laboratory and chickens were reared to different ages in either negative pressure isolators (a) or in poultry houses (b).

2. SPF chickens experimentally infected with either the ANV-1 isolate G4260 (described in McNulty et al., 1990) or with three different CAstV isolates (c). Separate groups of one-day-old SPF chicks were orally inoculated with different CAstV isolates including 11672, 612 and VF08-29 (origins of isolates detailed in Smyth et al., 2012). With each group, blood samples (for serum production) were collected from five birds at 7, 14 and 21 days post-infection before the birds were killed. The remaining 22-day-old chickens were parenterally inoculated with their respective CAstV before the final blood samples were collected two weeks later.

3. Sets of serum samples were collected from GP chickens at different ages from organizations A and B. These included 138 samples from two farms comprising three and four houses that were shown to be completely negative by indirect IF and 150 samples from a single farm obtained at 1 day, 24 weeks and 52 weeks. There were also 30 positive samples in three sets from organization A.

4. Sera from broiler parent chickens at different ages from organization C.

5. Clinical serum samples from problem flocks with kidney disease and gout in the Middle East.

The serum samples are summarized in Table 1.

Table 1. Summary of chicken serum samples used.

Serum group	Bird type	Age	Number of serum samples	Number of sets/houses	Table no.
(1a) SPFa	Non-infected controls	14–42 days	40	3	2
(1b) SPFb	Non-infected controls	12 weeks	27	1	2
(2) SPFc	CAstV experimentally infected birds	36 days	15	1	3
(3a) GP org A	Grandparents	12–14 weeks	168	10	2, 4
(3b) GP org B	Grandparents	1 day–52 weeks	150	4	2
(4) P org C	Broiler parents	Various	1864	68	5, 6
(5) P org D (Middle East)	Problem flocks	18.5–36.4 weeks	72	4	4

Note: All sera are derived from UK sources unless otherwise specified.

Indirect IF test

Indirect IF tests to detect CAstV-specific antibodies in chicken sera were performed as described previously (Todd et al., 2009b). Briefly, this involved using acetone-fixed cultures of primary chicken embryo liver cells that had been grown in glass coverslips and infected with selected isolates of CAstV including 11672, 612 and VF08-29.

Recombinant baculovirus/antigen

The recombinant baculovirus, designated AcNPV-gst11672, was produced previously in our laboratory (Lee et al., 2013). This recombinant virus expressed the capsid gene specified by the 11672 CAstV isolate as a fusion protein in which GST is fused at the N-terminal of the capsid gene. Following infection of Sf 9 cells or Hi 5 cells, the GST-11672 capsid protein was purified by affinity chromatography using the BD BaculoGold GST purification kit (BD Biosciences, Oxford, UK) according to manufacturer's instructions.

ELISA protocol

ELISAs were performed using Immulon 1 B plates (Thermo Scientific, Hemel Hempstead, UK) and comprised three successive addition and incubation steps before reaction with the enzyme substrate. Each step involved 100 μl volumes per well and was followed by thorough washing cycles using washing buffer composed of phosphate buffered saline (pH 7.2) containing 0.05% Tween 20. The three additions and incubation steps were as follows.

Plates were coated overnight at 4°C with purified antigen (5 μg/ml) diluted in 0.05 M carbonate–bicarbonate buffer with addition of 0.7 M NaCl (pH 9.4).

Test and control sera diluted (1:100) in dilution buffer (phosphate buffered saline containing 0.05% of Tween 20 and 0.35 M NaCl, pH 7.2) were added and incubated for 1 h and 10 min at 25°C on a horizontal shaker (Labnet, Edison, NJ, USA) at 60 rpm or overnight at 4°C.

Rabbit anti-chick Ig peroxidase (Jackson ImmunoResearch, Newmarket, UK) diluted 1:5000 in dilution buffer was added to each well and incubated for 1 h at 25°C on a horizontal shaker (Labnet; 60 rpm).

To each well 100 μl of the enzyme substrate 3,3′,5,5′-tetramethybenzidine (TMB/E) solution (EMD Millipore, Nottingham, UK) was added and the plate was incubated at 37°C for 10 min in the dark. The reaction was stopped by the addition of 50 μl 2 M H₂SO₄ to each well and absorbance was read at 450 nm with correction filter 620 nm using a Sunrise microplate reader (Tecan UK Ltd., Reading, UK). The test results were expressed as ELISA ratio values (i.e. S/P ratio), which were calculated using the following equation:(OD, optical density)

Results

Optimization of ELISA test

The GST-11672 capsid protein antigen used for coating the ELISA plates was prepared by affinity chromatography. Analysis by sodium dodecyl sulphate polyacrylamide gel electrophoresis indicated that the fusion protein was ∼100 kDa in size, but additional minor protein bands were also observed (Lee et al., 2013). The concentration of the eluted antigen was originally determined empirically to be ∼5 μg/ml but varied from batch to batch. ELISA plates were coated with this antigen, diluted to 1:200 or 1:400, according to the dilution which best maintained the positive control at an OD of ∼1, on a per batch basis using carbonate–bicarbonate buffer (pH 9.4) containing 0.7 M NaCl; the added NaCl was found to improve the reproducibility of the results.

Positive and negative control sera were used on each plate. The negative serum comprised a pool of serum samples from SPF chickens, while the positive control serum, used at a dilution of 1:200, was obtained by experimentally infecting SPF chickens with the 11672 CAstV isolate. This positive control serum usually generated an ELISA absorbance value in the range 0.800–1.00. When the positive control serum failed to generate this strength of signal the results were discarded. The result for each serum tested was presented as an S/P ratio as described above.

Selection of positive “cut-off” value

Serum sets from SPF chickens at different ages (Table 2, groups 1 and 2) and from GP flocks from two different poultry organizations (Table 2, groups 3 and 4) were tested by the ELISA.

Table 2. Summary of ELISA results obtained with negative serum samples.

Serum group	Age	Number of samples	Number of sets/houses	Mean ratio	Standard deviation
1. SPF (1a)	14–42 days	40	3	0.00	0.01
2. SPF (1b)	12 weeks	27	1	0.01	0.01
3. GP organization A (3)	12–14 weeks^a	138	7	0.04	0.03
4. GP organization B (3)	1 day–52 weeks	150	4	0.12	0.06

Note: Numbers in parentheses denote serum sample source as summarized in Table 1.

^aConfirmed as negative by indirect IF (see Table 4).

The mean ELISA ratio values and standard deviations were calculated for each group. These ranged from 0.00 to 0.20 and from 0.01 to 0.06, respectively, and were lower for the sets from SPF chickens than those for those from commercial GP chickens. Since the test was likely to find greatest application with samples from commercial chickens, a decision was taken to base the selection of the “cut-off” value on the higher mean (0.12) and standard deviation (0.06) values. Thus, a “cut-off” value of 0.30 (mean + 3× standard deviations) was selected, with serum samples producing ELISA ratios >0.3 being considered positive.

Specificity

The specificity of the test was evaluated by investigating the ELISA results for sets of experimental sera, which were separately raised to ANV-1 and three different CAstV isolates, namely 11672, 612 and VF08-29. None of the eight serum samples, which were positive for antibody to ANV-1 by an indirect IF test, were positive by ELISA. Of the five serum samples raised to each of the three CAstV isolates, positive ELISA results were obtained with three (CAstV11672), four (VF08-29) and one (CAstV 612), respectively (Table 3). It was noted that three of the serum samples obtained by infecting with the 11672 and VF08-29 isolates had ELISA ratios that we have designated as “high” (>1.00) or “medium” (0.50–0.99), whereas the single positive sample obtained by infecting with the 612 isolate had a “low” (0.31–0.49) ratio. The two 11672 sera that were negative by the ELISA test were double-checked by indirect IF test and were confirmed as negative.

Table 3. ELISA ratio values obtained with serum samples obtained from SPF chickens experimentally infected with 3 CAstV isolates.

Sera group SPF (2c)	11672	VF08-29	612
1	0.06	0.37	0.07
2	1.27	1.21	0.45
3	0.01	0.84	0.03
4	1.59	0.06	0.00
5	1.71	1.66	0.00

Note: Number in parenthesis denotes serum sample source as summarized in Table 1.

Comparison of test results obtained by ELISA and indirect IF test

The ELISA and indirect IF results obtained with a total of 240 serum samples were compared (Table 4). The 240 samples included 138 samples from two GP farms (seven houses) that were completely negative by the indirect IF test and 30 samples from one GP farm (three houses) that comprised positive and negative samples as determined by the indirect IF test. The comparison also included two serum sets that were collected at different ages from each of two parent flocks from the Middle East (Table 4; groups 3 and 4 and groups 5 and 6). With each of these parent flocks, all or the vast majority of the serum samples collected at the earlier time-points were negative by both tests whereas all of the samples collected at the later time-points were positive by both tests. It was concluded that in each case the parent flocks had seroconverted.

Table 4. Serum sets compared by ELISA and indirect IF.

Serum group	Age	Number of samples (submissions)	No positive by indirect IF	No positive by ELISA	S/P ratio range of positive samples
1 (3a)	12–14 week	138 (7)	0	0	<0.30
2 (3a)	12–14 week	30 (3)	28	26	0.31–1.96
3^a (5)	18.5 week	21 (1)	1	0	<0.30
4^a (5)	34 week	21 (1)	21	21	0.56–2.06
5^b (5)	31.5 week	15 (1)	0	1	0.32
6^b (5)	36.4 week	15 (1)	14	14	0.33–1.55

Note: Numbers in parentheses under Serum group denote serum sample source as summarized in Table 1.

^aGroups 3 and 4 are from same parent flock.

^bGroups 5 and 6 are from the same parent flock.

Although the ranges of ELISA S/P ratios of the ELISA-positive samples differed for the three serum groups containing ELISA positives (Table 4, groups 2, 4 and 6), the range of the ELISA positives overall was 0.31–2.06. The single ELISA-positive sample in group 5, which was negative by the indirect IF test, had an S/P ratio of 0.32.

A close examination of the results obtained for the 240 samples showed that 175 of 176 samples that were negative by indirect IF were also negative by ELISA, while 62 of 64 samples that were positive by indirect IF were positive by ELISA. Taken together the results obtained with 237 of 240 (99%) were found to be in agreement.

Flock seropositivities

Sixty-eight serum sets from parent birds from organization C (Table 1, source 4), which were received as part of 11 different submissions, have been tested by ELISA. These serum sets differ with regard to either the flock sampled or the age at which the flock was sampled. Of the 68 sets tested, 16 (23.5%) were seronegative. The remaining 52 sets displayed variable seropositivities (proportions of positive samples in set) ranging from 9.8% to 100%. For analysis purposes, serum sets in which >75% samples were ELISA positive were considered to have “high seropositivity”, flocks in which <25% samples were ELISA positive were considered to have “low seropositivity” and flocks in which the proportion of ELISA-positive samples was in the 25–75% range were considered to have “medium seropositivity”. Of the 68 sets tested, the numbers of sets with low, medium and high seropositivities were nine (13.2%), 19 (27.9%) and 24 (35.3%), respectively. Of the 52 seropositive sets, the proportions that could be categorized as having low, medium and high seropositivities were 17.3%, 36.5% and 46.2%, respectively (Table 5).

Table 5. Seropositivity results obtained for serum sets.

Serum group (4)	No. sets	No. sera	No. sets seronegative	No. sets with low seropositivity	No. sets with medium seropositivity	No. sets with high seropositivity
1	6	320	2	0	2	2
2	2	92	0	0	0	2
3	4	440	0	0	1	3
4	11	289	2	5	1	3
5	12	120	5	1	1	5
6	1	22	1	0	0	0
7	14	280	0	0	9	5
8	6	120	1	0	1	4
9	6	61	3	2	1	0
10	2	40	1	0	1	0
11	4	80	1	1	2	0
Totals	68	1864	16 (23.5%)	9 (13.2%)	19 (27.9%)	24 (35.3%)

Notes: Variability in ELISA S/P ratios.

Number in parenthesis after Serum group indicates the serum source in Table 1.

Variability in ELISA S/P ratios

The ELISA S/P ratios determined in our testing of over 1800 serum samples ranged from 0.35 to 2.06. For result interpretation purposes, we have adopted a semi-quantitative approach by assigning the ELISA-positive samples to one of three groups depending on their S/P values. Serum samples with S/P ratios ranging from 0.31 to 0.50 were considered to be in the “weak” category, samples with S/P ratios ranging from 0.51 to 1.00 were considered to be in the “moderate” category and samples with S/P ratios above 1.00 were considered to be in the “strong” category. Of 1864 samples that comprised the 68 serum sets, 774 (41.5%) samples had S/P ratios of 0.30 or less and were designated negative (Table 6). The numbers of samples in the weak, moderate and strong categories were 294 (15.8%), 562 (30.2%) and 234 (12.5%), respectively. These three categories accounted for 27% (weak), 51.5% (moderate) and 21.5% (strong) of the total number (1090) of ELISA-positive samples.

Table 6. Comparison of S/P ratios for serum sets with high and low seropositivities.

Serum group (4)	No. sets	No. sera	No. negatives	No. weak S/P ratios	No. moderate S/P ratios	No. strong S/P ratios
Overall	68	1864	774 (41.5%)	294 (15.8%)	562 (30.2%)	234 (12.5%)
Sets/flocks with high seropositivity	24	840	90 (10.7%)	160 (19.1%)	407 (48.5%)	183 (21.8%)
Sets/flocks with low seropositivity	9	190	168 (88.4%)	16 (8.4%)	5 (2.6%)	1 (0.6%)

Note: Number in parenthesis after Serum group indicates the serum source in Table 1.

The S/P ratio values obtained with sets that were regarded as having high and low seropositivity were compared (Table 6). With the 24 flocks (sets) in the high seropositivity grouping, the numbers of samples with strong and moderate S/P ratios totalled 590 (70.2%), whereas the numbers of samples that were negative (90; 10.7%) or that had weak S/P ratios (160; 19.1%) were comparatively small. In contrast, with the nine flocks (sets) in the low seropositivity grouping, the numbers of samples with strong and moderate S/P ratios were small (6; 3.2%), while the total number of samples that were either negative (168; 88.4%) or that had weak S/P ratios (16; 8.4%) were very high (184; 96.8%). These results indicated that samples with high levels of ELISA reactive antibodies (strong S/P ratios) were more prevalent in serum sets with high proportions of ELISA-positive samples.

Discussion

This paper describes the development and evaluation of the first ELISA for the serological diagnosis of CAstV infections in commercial chickens. The described test detects antibodies to CAstVs from Group B and the results produced correspond closely with those obtained by indirect IF.

There have been very few reports describing the serological diagnosis of CAstV. These have involved the application of indirect IF tests (McNeilly et al., 1994; Todd et al., 2009b) and a virus neutralization (VN) test, which was based on the first characterized CAstV isolate (Baxendale & Mebatsion, 2004). The use of an ELISA for detecting CAstV antibodies in experimentally vaccinated chickens was also described by Sellers et al. (2010). On the basis of recently published capsid sequence diversity data and the implications that these data have for antigenic diversity (Smyth et al., 2012), the indirect IF and VN tests are likely to have limitations in relation to the range of CAstV strains serologically diagnosed. For example, the VN test based on the first characterized CAstV, a Group A isolate, is unlikely to detect antibodies to the Group B CAstVs and may not be effective in detecting antibodies to other sequence-distinct Group A CAstVs, which are thought to belong to different serotypes (Smyth et al., 2012).

A different problem exists with regard to indirect IF tests. Although separate indirect IF tests, based on the 612 CAstV isolate (Group A) and the 11672 CAstV isolate (Group B), were developed and applied to detect antibodies to these two antigenically different CAstVs, the low level of cross-reactivity that has been shown to exist (Todd et al., 2009b) means that the indirect IF test based on the Group B 11672 isolate produces weak positive reactions in serum samples that contain high levels of antibodies induced by Group A CAstV infections. We suspect that the cross-reactivity detected by indirect IF may be due to antibodies specific to non-structural CAstV proteins, such as the RNA polymerase, and not due to the antibodies specific to the capsid proteins. The amino acid identities, ∼85%, shared by the RNA polymerase proteins of the CAstVs of Groups A and B is likely to be sufficient to account for the cross-reactivity, whereas this is not likely to be the case for the capsid proteins, which share a comparatively low (38–40%) level of identity. Since the ELISA test described in this paper is based on the use of the baculovirus-derived capsid protein of the 11672 isolate, a Group B CAstV, it will not detect antibodies to the non-structural CAstV proteins, and, on the basis of capsid protein diversities, we would predict it would detect antibodies to Group B CAstVs and would not detect antibodies to Group A CAstVs. Our findings with experimentally infected chickens support these predictions. Thus, the ELISA based on the 11672 capsid produced strong ELISA signals with serum samples collected from chickens experimentally inoculated with 11672 and VF08-29, Group B CAstVs belonging to different subgroups (B i and B ii, respectively), but strong ELISA signals were not detected with serum samples collected following infection with the 612, a Group A CAstV isolate (Table 3).

At this stage we have no explanation as to why positive ELISA responses were observed in only three of the five CAstV 11672 serum samples. It would appear that the two negative birds did not seroconvert despite the double inoculation of CAstV 11672 and this was also confirmed by negative indirect IF results. The low ELISA ratio (0.45) obtained with one of the five sera collected following infection with the 612 isolate may be due to a non-specific reaction, which may have resulted from an individual reaction to the media used during the booster inoculation. Greater background reactivity is observed with commercial birds compared to SPF birds and this tends to increase as the flock ages. This ELISA test has been designed as a flock screening tool and results should be interpreted with regard to all of the results for that flock. For instance, one low positive value within a flock for which all other values are below the cut-off value would suggest a negative flock containing a single non-specific reaction of slightly greater strength in one bird compared to the others.

Comparison of the results obtained with 240 selected serum samples (Table 4) showed that there was close agreement between the results obtained by ELISA and indirect IF test, which was based on Group B 11672 CAstV isolate. If applied to sera from commercial poultry in general, complete agreement between the tests would not be expected, since, due to the low level cross-reactivity between CAstVs from Groups A and B, which is thought to involve the conserved non-structural proteins, high levels of antibody elicited by infection with Group A CAstVs will be detected by the indirect IF test but will not be detected by ELISA. It is probable that the positive flocks (Groups 2, 4 and 6, Table 4) selected for comparison of the two tests may have been infected with Group B CAstV only.

In this study the ELISA test was evaluated with sets of serum samples that were largely obtained from broiler parent and GP flocks (Tables 4 and 5). An earlier study in our laboratory, involving the use of separate indirect IF tests based on the 11672 (Group B) and 612 (Group A) isolates, reported that the large majority of European parent flocks, which were tested, were seropositive for both CAstV groups, but that the proportions of seropositive GP flocks were substantially lower (Todd et al., 2009b). In the present study using the Group B CAstV ELISA, we found that 76.4% of the serum sets contained positive samples, indicating that about one quarter of the flocks were uninfected at the time of testing. Closer analysis indicated that almost half (46.2%) of the seropositive flocks had more than 75% of the serum samples positive and, as such, were considered to be highly seropositive. These findings can be factored into the design of ELISA-based flock screening programmes, when selecting the numbers of serum samples that need to be tested per flock.

Another factor to consider relates to the levels of antibody detectable in serum samples. An examination of the ELISA values obtained with the 68 serum sets investigated showed that there were 1090 ELISA positives in the 1864 samples tested (Table 5). Of these 1090, 796 (72.9%), comprising 562 (51.5%) with ratios in moderate category (range 0.5–0.99) and 234 (21.4%) with ratios in the strong category (1.00 or greater) could be regarded as having substantial ELISA reactivity. The detection of substantial antibody levels in a relatively high (72.9%) proportion of samples may permit the pooling of serum samples, thereby reducing testing costs.

More detailed examination showed that the proportions of samples in the moderate (48.5%) and strong (21.8%) categories in sets/flocks displaying high seropositivity (>75% samples positive in the set) were substantially greater than the values, 2.6% (moderate) and 0.6% (strong), observed in sets/flocks displaying low seropositivity (<25% samples positive in the set) (Table 6). Thus, the levels of ELISA-detectable antibody in serum samples increase as flock seropositivity increases. These findings are consistent with the general views that both seroconversion and antibody levels within individual animals increase with time post-infection.

In the testing undertaken for our study we have evidence of strong seroconversions observed in two broiler parent flocks (Table 4). In one of these the parent flock was seronegative by ELISA at 18.5 weeks and 100% seropositive at 34 weeks, and in the other, one (6.7%) of 15 samples was weakly positive (S/P value of 0.32) when the parents were tested at 31.5 weeks and 14 (93.3%) of 15 were positive at 36.4 week-old. In the latter case, the very pronounced (6.7–93.3%) seroconversion within a relatively short time (<5 weeks) indicated that the spread of CAstV infection through this flock had been rapid. When the ages of these parent flocks are considered it is evident that birds within both flocks were infected when they were in lay, thereby creating the opportunity for infection of the egg and vertical transmission of CAstV to the progeny chicks.

Evidence that CAstV is vertically transmitted was first reported by Spackman et al. (1984), who described the isolation of an ELV, later characterized as a Group B CAstV (Smyth et al., 2012) from dead-in-shell embryos during an investigation into early broiler mortality. More recently, we have detected Group B CAstVs by RT-PCR in dead embryos and day-old chicks in cases of reduced hatchability (Smyth et al., 2013). Although the prevalence of CAstV vertical transmissions is unknown, we suspect that it will only occur when chickens become infected with CAstV for the first time during lay. Since CAstV infections appear to be common (Todd et al., 2009b), it is likely that the majority of parent flocks will have been infected during rear and that, due to protective immune responses, these birds will be resistant to subsequent challenge with an antigenically related CAstV. However, with higher levels of biosecurity now being applied within the poultry industry, the possibility of uninfected parent flocks coming into lay is increasing and this may lead to increasing incidence of problems that are associated with vertically acquired CAstV infections. Further investigation is required to determine whether vertically acquired CAstV infections can cause reduced hatchability and early broiler mortality problems and, should this be the case, whether the causal CAstV belongs to a particular CAstV group or subgroup. As a consequence, vaccines have yet to be developed. However, it is envisaged that a vaccine administered to broiler parents during rear will protect against challenge with field virus during lay thereby preventing vertically acquired CAstV infections and related problems. Such an approach will also result in the hatched chicks having uniformly high levels of maternal antibody which will better protect them against early CAstV infections that have been linked with broiler growth problems.

The ELISA developed in this study provides a convenient serological tool with which to diagnose Group B CAstV infections. It is likely to find limited application in commercial broilers due to the widespread occurrence of Group B CAstVs in these flocks (unpublished results). However, it is likely to find uses when applied to broiler parent flocks. For example, retrospective serological testing can be used to investigate the roles that Group B CAstV infections during lay have in causing conditions such as visceral gout and “White chick”, with which these viruses have been associated (Bulbule et al., 2013; Smyth et al., 2013). The routine testing of broiler parents and GP flocks at point of lay (∼20 weeks) by this ELISA will identify seronegative flocks that will be susceptible to infection during lay. Continued serological monitoring will detect infections occurring during lay and allow possible correlations with hatchability or early broiler mortality problems to be investigated. The ELISA test will also be useful if breeder vaccines become available in relation to identifying seronegative flocks that need to be vaccinated and to assessing antibody response induced by vaccination. The ELISA test is also likely to be useful to organizations producing SPF chickens. On the basis of its vertical transmission and pathogenicity for young chickens, CAstV, like its fellow avian astrovirus, ANV, has the potential to contaminate avian vaccines that are produced in SPF eggs. Therefore, it is probable that CAstV will be recognized as an additional “specified pathogen”. Flock screening by ELISA provides a convenient means of demonstrating freedom from Group B CAstV infections.