Turkey astrovirus (TAstV) is an important agent of poult enteritis. The diagnosis of astroviruses has been dependent mainly on electron microscopy (EM) or immune EM (IEM). To develop other simple, rapid, and reliable diagnostic assays, two antigen-capture enzyme-linked immunosorbent assays (AC-ELISAs), polyclonal AC-ELISA and monoclonal AC-ELISA, were developed in this study. Monoplex and multiplex reverse transcription-polymerase chain reactions (RT-PCRs) were also developed using nondegenerate primer sets specific to the capsid region and degenerate primer pairs specific to the polymerase area of two TAstV. EM was included for comparison. Fecal or intestinal contents samples from naturally and experimentally infected poults with enteritis were examined using the developed assays. The polyclonal AC-ELISA had higher sensitivity and wider detection spectrum than the monoclonal AC-ELISA with group-specific monoclonal antibody (MAb), whereas the monoclonal AC-ELISA had very high specificity but lower sensitivity, which was estimated at 0.06 μg of viral proteins. Small round viruses (SRV) that could be astroviruses or other small viruses were detected in 34.4% of the samples examined by EM. The monoplex RT-PCR results amplified with primers SRV-1-3 and SRV-1-5 revealed that the positive rate of astroviruses was 45.3%, which was 10.9% higher than that of EM even if other SRVs were not excluded. Multiplex RT-PCR with SRV-1-3 and SRV-1-5 and AFCP-F1 and AFCP-R1 and the monoplex RT-PCR with degenerate primers verified that the positive rate of astroviruses was 59.4%, which was 25% higher than that of EM. Both RT-PCRs showed good specificity and wider detection spectrum compared with earlier published data.
Astroviruses are etiologic agents of acute gastroenteritis in the young of a variety of species, including humans, cattle, sheep, swine, dogs, cats, deer, mice, minks, and turkeys (5,22). Extraintestinal illness caused by astroviruses has also been reported in some avian species (6,12). Human astroviruses were first described in 1975 by Appleton and Higgins (1) and Madeley and Cosgrove (21). Turkey astrovirus (TAstV) infections have been described by McNulty et al. (23), Saif et al. (30), and Reynolds and Saif (27). TAstVs have been incriminated as an etiologic agent of enteric disease and turkey poult enteritis and mortality syndrome (PEMS), a highly pathogenic multisystem disease (23,27,30). TAstV infections are geographically widespread (29). In an electron microscopy (EM) and electropherotyping survey of commercial turkey flocks with enteric disease, astroviruses were the most commonly detected viruses (28).
Astroviruses are small, round, nonenveloped RNA viruses of 28–30 nm diameter. The remarkable feature of astroviruses is that about 10% of the particles have a five- or six-pointed starlike surface structure when examined by direct EM. Previously, detection and diagnosis of astroviruses has been heavily dependent on EM, immune EM (IEM), and fluorescent antibody assays. However, the typical distinctive starlike morphology of astroviruses is pH dependent and can vary among different preparations (3,22). Yu et al. (38) reported that a turkey astrovirus they detected was featureless. Therefore, it is difficult sometimes to differentiate astroviruses from other small round viruses (SRVs), such as enteroviruses, or picornaviruses by EM or IEM on the basis only of their capsid profiles. Furthermore, the sensitivity of EM has been estimated to be 106–107 virus particles per gram of feces (22), which is relatively low. The fluorescent antibody assays are not practical for testing a large number of samples, and their sensitivity and specificity are also not very high (0.2–0.4 μg of antibody nitrogen). An enzyme immunoassay (EIA) for detection of human astrovirus was developed (11). Cubitt et al. (4) demonstrated that both EM and EIA as applied on stool specimens had a similar level of sensitivity, but EIA was more practical for monitoring an extensive outbreak. There is no EIA for detection of TAstVs yet. A reverse transcription polymerase chain reaction (RT-PCR) for human astrovirus has been used as a diagnostic tool and is associated with higher sensitivity than that of EM and EIA (13). RT-PCR methods for TAstV detection have also been developed by Yu et al. (38), and Koci et al. (16). However, the detection spectrum of the RT-PCR methods was not broad because all the primer pairs used in their RT-PCRs were designed on the basis of a single turkey astrovirus sequence in which no conserved region was found.
The objective of this study was to develop several diagnostic tests with improved parameters for detection of TAstVs. Two antigen-capture enzyme-linked immunosorbent assays (AC-ELISAs) were established with the use of polyclonal hyperimmune antisera against TAstV1987 and TAstV2001 isolates and monoclonal antibody (MAb) developed against TAstV2001. Monoplex and multiplex RT-PCRs with wide detection spectra were also developed. In addition, direct EM was included, and the results obtained from the above tests were compared.
MATERIALS AND METHODS
Fecal or intestinal contents specimens
Samples (n = 319) of fecal or intestinal content from turkey poults with enteritis from North Carolina commercial turkey farms were collected from May to August 2001. These specimens were treated as described previously (38). Briefly, samples were mixed well with 0.01 M pH 7.4 phosphate-buffered saline (PBS) at a ratio of 1:10 and frozen and thawed three times. The samples were then clarified by centrifugation at 3000 × g for 30 min at 4 C. Supernatants were collected and filtered through 0.8- and 0.45-μm disposable syringe filters. Filtrates were then stored at −70 C until examined.
Animals and embryonated eggs
Specific pathogen-free (SPF) chickens, turkeys, and turkey embryos used in this study were obtained from the SPF flocks maintained by the Food Animal Health Research Program (Wooster, OH). The flocks are free of detectable chicken and turkey pathogens, including enteric viruses as indicated by periodic testing. Guinea pigs used in the study originated from a certified commercial source. BALB/C mice used for developing MAbs were obtained from Charles River Laboratories (Wilmington, MA).
The astroviruses used in this study—turkey astrovirus 1987 (TAstV1987), turkey astrovirus 1997 (TAstV1997 or SRV p14), and turkey astrovirus 2001 (TAstV2001)—were isolated in our laboratory in the years indicated from feces of commercial poults with diarrhea and high mortality. TAstV1987 and TAstV2001 belong to two different serotypes, whereas TAstV1997 (SRV p14) was not serotyped (35). All viruses were stored at −70 C. TAstV1987 and TAstV2001 were used for production of hyperimmune antisera, which were used as capture antibody and detector antibody, respectively, and as the positive controls in the polyclonal AC-ELISA assay. TAstV2001 was employed as antigen to produce MAbs, which were used as detector antibodies in the monoclonal AC-ELISA. TAstV1987, TAstV1997 (SRV p14), and TAstV2001 were used as positive controls in the monoclonal AC-ELISA assay and all RT-PCRs. Turkey group A rotavirus (TGARV), turkey group D rotavirus (TGDRV), turkey atypical rotavirus (ATR), and turkey enterovirus (TEV) were maintained at our laboratory. Turkey coronavirus (TCV, Minnesota strain) was obtained from The American Type Culture Collection (Rockville, MD). Human group A rotavirus (Wa-Rota) was provided by Dr. Linda Saif of The Ohio State University. All six viruses were used to determine the specificity of the polyclonal and monoclonal AC-ELISA. TAstV2001 purified with CsCl was used to determine the detection limits of the monoclonal AC-ELISA.
Virus propagation and purification
The procedure used for the propagation and purification of TAstV1987 and TAstV2001 and the protocol for measuring the virus protein concentration have been described (35).
The capture antibodies used in the polyclonal AC-ELISA were prepared by inoculating turkey poults or chickens with purified TAstV1987 and TAstV2001. For turkey antiserum production, the procedure was a modification of that described by Hayhow and Saif (9). The procedure for production of chicken anti-TAstV1987 and -TAstV2001 was similar to that for production of turkey anti-TAstV1987 and -TAstV2001, except that the first inoculum for chickens was given by injection.
The detector antibodies, anti-TAstV1987 or -TAstV2001, were prepared in guinea pigs as described previously (35). To remove any possible anti–intestinal tissue antibodies, the detector antibodies were absorbed with freeze-dried SPF turkey intestinal tissue powder before use.
Production and characterization of MAbs against TAstV2001
The procedure for production of MAbs against TAstV2001 has been described previously (17). ELISA was used for isotyping MAbs in hybridoma supernatant fluids. A commercial mouse MAb isotyping reagent, including monospecific antiserum to mouse immunoglobulins (IgGs; IgG1, IgG2a, IgG2b, IgG3, IgM, and IgA), was obtained from Sigma Chemical Co. (St. Louis, MO). The tests were performed following the instructions of the manufacturer. To predetermine the reactivity of each MAb, three antibody-capture ELISAs (AB-ELISAs), with purified TAstV1987, TAstV1997 (SRV p14), and TAstV2001 as coating proteins, were developed as described (35).
The procedure used for the polyclonal AC-ELISA was similar to that described earlier (9) with modification. A checkerboard titration to determine the concentration of each reagent in AC-ELISA was first performed. Turkey anti-TAstV2001 or -TAstV1987 served as the capture antibody. Fecal specimens to be tested and controls were prepared as described above. Guinea pig anti-TAstV1987 or anti-TAstV2001 hyperimmune sera were used as detector antibodies. The conjugate was goat anti–guinea pig IgG (heavy and light chain) horseradish peroxidase–labeled antibody (Kirkegaard and Perry Laboratories, Gaithersburg, MD). The substrate used was the 3,3′,5,5′-tetramethylbenzidine (TMB) mixed with equal volumes of peroxidase solution B (TMB Microwell Peroxidase Substrate System, Kirkegaard and Perry Laboratories). Absorbance was read at 450 nm with a spectrophotometer (MAXlineTM microplate reader, Molecular Devices Corporation, Sunnyvale, CA). The cutoff value was determined by (3 × SD) + (2.1 × mean absorbance value of all negative controls) (9,20).
Monoclonal AC-ELISA was similar to that of polyclonal AC-ELISA described above. Turkey anti-TAstV2001 or anti-TAstV1987 hyperimmune antisera was used as the capture antibody. Two group-specific MAbs, 1-8E and 1-9D, were employed as detector antibodies, and goat anti-mouse IgG (heavy and light chain) horseradish peroxidase–labeled antibody was used as conjugate. Of 319 fecal samples, 116 were tested by this assay. The intestinal homogenates of 26-day-old turkey embryos inoculated with TAstV1987, TAstV1997 (SRV p14), and TAstV2001 severed as positive controls. The intestinal mock from SPF turkey embryos and the feces of SPF turkey poults and PBS were used as negative controls. The TAstV2001 purified by CsCl was used to determine the analytical sensitivity of the test, and six other enteric viruses mentioned above, including TGARV, TGDRV, ATR, TEV, TCV (Minnesota strain), and Wa-Rota, were used to verify the specificity of the monoclonal AC-ELISA.
RT-PCR for detection of astroviruses in fecal specimens
The RNA was extracted from filtrate specimens of fecal or intestinal contents with a commercial kit (TRIzol LS reagent, Life Technologies, Grand Island, NY) following the manufacturer's instructions.
Three sets of oligonucleotide primer pairs were used in the RT-PCR (Table 1). Information on the first set of primer pairs was kindly provided by Drs. Akbar Ali and Donald L. Reynolds, Iowa State University. They were referred to as SRV-1-3 and SRV-1-5, with an expected product of 473 nucleotides. The second set of primers was designed from the sequence of TAstV-2 (Accession no. AF206663), referred to as AFCP-F1 and AFCP-R1 with an expected product of 464 nucleotides. Both above primer sets are specific to the 3′-end capsid region of the turkey astrovirus RNA genome. The third primer pairs were degenerate primers covering a 601-nucleotide region of the polymerase genes in two available TAstV genomes (Accession nos. Y15936 and AF206663). They were designated TAPG-Ll and TAPG-R1. All primers were designed with primer design software Primer 3 ( http://www.yk.rim.or.jp/∼aisoai/tool.html) and were synthesized by Integrated DNA Technologies, Inc (Coralville, IA).
Monoplex RT-PCR and multiplex RT-PCR
The monoplex RT-PCRs with the above sets of primers were performed as described previously (38). The six amplicons selected randomly were confirmed by sequencing and BLAST analysis (38). Primer sets SRV and AFCP were used in multiplex RT-PCR. The reaction conditions of multiplex RT-PCR were similar to those of monoplex RT-PCR.
Direct EM was performed as described previously (38).
All the antisera prepared in guinea pigs, chickens, and turkeys were titrated with AB-ELISA (35) (Table 2.). The titers of these antisera ranged from 12,800 to 409,600. All antisera prepared against TAstV1987, TAstV1997 (SRV p14), and TAstV2001 aggregated the homologous and heterologous TAstV1987, TAstV1997, and TAstV2001 when examined by IEM, whereas sera from normal guinea pigs or SPF chickens and turkeys did not (data not shown).
Production and characterization of MAbs against TAstV2001
Twenty-five hybridomas secreting specific MAbs against TAstV2001 were obtained. Of these 25 positive hybridomas, 23 secreted group-specific MAbs, which reacted with TAstV1987, TAstV1997 (SRV p14), and TAstV2001, and two (2-6A and 2-8A) produced strain-specific MAbs, which only detected homologous virus TAstV2001. All MAbs secreted by the 25 hybridomas did not react with TGARV, TGDRV, ATR, TEV, TCV (Minnesota strain), and Wa-Rota. The isotypes (subisotype) of MAb 1-5G, 2-10H, and 5-7H were IgG2b, IgG2a, and IgM, respectively; the rest were IgG1. No MAbs belonged to IgG3 or IgA. Among these MAbs, 1-8E and 1-9D had the highest antibody titer and were used as detector antibody in the monoclonal AC-ELISA.
A total of 64 fecal specimens were tested by polyclonal AC-ELISA. Of these samples, 51 (79.6%) were astrovirus positive. A comparison between EM results and polyclonal AC-ELISA results is shown in Table 3. The specificity was determined by testing the known positive and negative intestinal samples, which were predetermined by EM and RT-PCR. All positive and negative controls tested were shown to be positive or negative, respectively. The detector antibody, guinea pig anti-TAstV2001, preabsorbed with SPF turkey embryo intestinal homogenates, resulted in a lower nonspecific background compared with the non-preabsorbed guinea pig anti-TAstV2001 antiserum. Polyclonal AC-ELISA was more sensitive compared with that of the monoclonal AC-ELISA.
Of 319 fecal samples, 116 samples, along with both positive and negative controls, were examined by this test. The positive controls had high optical density (OD) values, and the negative controls had low OD values (as low as the blank wells), indicating the test was specific. However, only three intestinal homogenate samples with a virus titer of 102.5 to 104.5 EID50 (dose infecting 50% of embryos) were strong positive; all the rest were negative. The lowest detection limit was determined to be 0.06 μg of viral protein.
Monoplex RT-PCR and multiplex RT-PCR
Fecal specimens or intestinal contents (n = 319) were examined by monoplex RT-PCR with the primer pairs SRV-1-3 and SRV-1-5. The expected 473-bp RT-PCR products were obtained from 43.2% of the samples or fecal or intestinal contents from poults with diarrhea and high mortality (Table 4; Fig. 1). The positive rate was 10% higher when compared with that of direct EM (Table 4), indicating that the sensitivity of the RT-PCR was higher than that of EM. However, some fecal samples (10 of 22) were detected only by EM but not by RT-PCR; that is, EM detection was positive, but RT-PCR with the SRV-1-3 and SRV-15 primers was negative, perhaps because of detection of all the SRVs, such as enterovirus, calicivirus, and other small enteric viruses by EM but not by RT-PCR. False negatives caused by inhibitors in the RT-PCR system might be another reason, which would lower the sensitivity of RT-PCR as well. Sequence analysis for six amplicon samples selected randomly from SRV primer set–amplified products had a high homology with published turkey astrovirus sequence data (data not shown), confirming that the RT-PCR was specific. It was shown that the SRV primer set could only amplify TAstV2001, but not TAstV1997 (SRV p14), and a weak amplicon band was visualized for TAstV1987 (Fig. 1). The results of monoplex RT-PCR with primer set AFCP-F1 and AFCP-R1 demonstrated that the AFCP could only amplify TAstV1987 and TAstV1997 (SRV p14), but not TAstV2001 (Fig. 2). The results of RT-PCR with degenerate primers TAPG-Ll and TAPG-R1 showed an expected 601-bp band (Fig. 3), with a positive rate of 59.4% of tested samples (n = 64), which is 25% and 14.1% higher than that of EM and RT-PCR with primers SRV-1-3 and SRV-1-5, respectively (Table 3). The samples tested with monoplex RT-PCR with the degenerate primer set were tested by multiplex RT-PCR with primer sets SRV and AFCP (Fig. 4). Similar results were obtained when the results from the multiplex RT-PCR were compared with the monoplex RT-PCR with a degenerate primer set (Table 3, Fig. 4).
The samples (n = 319) tested with the monoplex RT-PCR with SRV primer set were tested by EM as well. Representative viral particles detected in preparations from fecal or intestinal content specimens of poults with diarrhea and high mortality by EM with negative staining are illustrated in Fig. 5. The viral particles C-1 and C-2 are SRVs including turkey astrovirus. The average size of SRV particles is approximately 28 ± 2 nm. The positive rate of SRVs was 32.2% (Table 4). In addition to SRVs, other viral particles, such as rotavirus, coronavirus (TCV), minireovirus, and adenovirus, as well as bacterial phages were detected in some of the fecal samples. The frequency of these particles is summarized in Table 4. In addition, Alfalfa mosaic–like virus was observed in three of 319 samples. The majority of these fecal samples were shown to contain multiple viral types.
Until recently, the primary method of detecting SRVs, including astrovirus, in feces or intestinal contents was direct EM with negative staining. This conventional method is relatively sensitive enough to identify the agents because large numbers of viral particles are frequently shed with diarrhea at the acute stage of the disease caused by astroviruses (19). The sensitivity of direct EM has been estimated to be 106 to 107 virus particles per gram of specimen (22). However, only 10% of astrovirus particles might exhibit the five or six pointed starlike characteristic, so it is difficult sometimes to differentiate astroviruses from other featureless SRVs. The EM results from 319 tested fecal samples in this study showed that the positive rate of SRV was 32.2% (Table 4). Similarly, the EM results from 64 tested samples (out of 319 samples) showed a 34.4% positive rate of SRV (Table 3), which included turkey astrovirus and possibly other small round enteric viruses. Because of this limitation of EM, Reynolds (26) and others suggested the use of IEM techniques that are reliable and can enhance not only the specificity but also the sensitivity of the test if the corresponding antiserum is available (2,14,18,26). However, IEM can mask the characteristic morphologic star feature or fail to detect new serotypes if lacking the specific antiserum or because of existing cross-reactions between different viruses. Generally, both direct EM and IEM techniques are time consuming and labor intensive, especially when handling large numbers of field specimens, and the application of theses techniques requires sophisticated equipment and experienced personnel.
The AC-ELISA is a relatively quick and simple test to perform for detection of pathogen antigen(s). An antigen-capture EIA for detection of human astroviruses was developed by Herrmann and others (10,11) with the use of a group-specific MAb. A modification of this assay with a biotinylated detector antibody was developed by Moe et al. (25). Both EIAs were shown to be useful tools, especially when a large number of samples need to be tested during epidemiologic investigations (34). However, the sensitivity of EIA is not very high and is similar to IEM (105 to 106 viral particles per gram of stool) (11,25). For TAstV detection, Koci and Schultz-Cherry (15) analyzed the reactivity between a recombinant TAstV-2 capsid protein and a MAb against TAstV (produced by James Guy , North Carolina State University) by western blot, ELISA, and immunofluorescence in transfected cells (15). They found that all tests were positive. The results obtained with the monoclonal AC-ELISA developed in this study demonstrated that the assay possessed very high specificity; however, the assay only detected 0.06 μg of viral protein. A monoclonal AC-ELISA developed by Hassan et al. (8) did not detect an infectious bursal disease virus in samples that had a titer of less than 5 × 103 EID50. Sharma et al. (33) obtained similar results when the monoclonal AC-ELISA did not detect viral antigens in thymuses and bursas of experimentally infected chicks. The monoclonal AC-ELISA developed in this study could detect TAstV2001 and TAstV1987 easily in all intestinal homogenate samples with a titer of 5 × 102 EID50, which were from turkey embryos inoculated with these viruses. To some extent, it seems that our monoclonal AC-ELISA is more sensitive. The MAbs used in our system were cell culture supernatants, which usually have low antibody titers compared with those from ascites fluid, which contains high levels of antibodies. Our assay did not detect the viruses in any field fecal samples, which indicates that virus concentration in these fecal samples was under the detection limits of the monoclonal AC-ELISA. The low sensitivity of the monoclonal AC-ELISA might be attributed to the mono-specificity of binding. Improvements in the sensitivity of the monoclonal AC-ELISA might be achieved by pooling several MAbs directed against different epitopes (8), resulting in increased binding capacity of the MAbs. The use of the avidin-biotin system to immobilize the MAbs, resulting in greater capture capacity, might be another approach to improve the sensitivity of the monoclonal AC-ELISA. Monoclonal AC-ELISA provides an alternative, practical method for the detection of astrovirus antigens after the viruses are enriched via culture in embryos.
Polyclonal AC-ELISA showed a 45.2%, 34.3%, and 20.2% higher positive rate than EM, monoplex RT-PCR with SRV primer set, and multiplex RT-PCR, respectively. As shown in Table 3, the agreement between EM and polyclonal AC-ELISA was 85.3%, indicating that the detection spectrum of EM is wider than the polyclonal AC-ELISA but that the polyclonal AC-ELISA is more sensitive than EM. The higher positive detection rate of the polyclonal AC-ELISA over RT-PCR is a result of its wider detection spectrum than RT-PCR but not higher sensitivity because 5% of the fecal samples positive by RT-PCR were not positive by polyclonal AC-ELISA (data not shown). The polyclonal AC-ELISA can be used for diagnostic or epidemiologic purposes.
In addition to the above methods, molecular probes and RT-PCR were recently developed to detect both human astroviruses and TAstVs as the partial or complete gene or genomic sequence data became available (16,37). The sensitivity of probes is similar to EIA (25,37), but the sensitivity of RT-PCR was estimated to be 10–100 viral particles per gram of sample (31), which is much higher compared with EIA. For instance, the RT-PCR detected astroviruses in 32% of the samples tested, whereas EIA only had a detection rate of 10% in an investigation of an outbreak of astrovirus gastroenteritis at a daycare center (24). For human astrovirus detection, RT-PCR with serotype-specific primers (1–8) was developed, and it correlated well with EIA (32). For TAstV detection, Koci et al. (16) designed two sets of primers: one based on the polymerase gene and another from the capsid gene of the genome of TAstV-2 (Accession no. AF206663). Generally speaking, the nucleotide sequences of polymerase genes in RNA viruses from the same family are relatively conserved and can be used to design primers to detect virus members belonging to the same family, whereas the nucleotide sequences of capsid genes of theses viruses are relatively variable and enable us to design primers to detect different strains of astroviruses. However, a similarity analysis of the TAstVs (Accession nos. Y15936 and AF206663) by ClustalW software revealed that the polymerase regions of the two viruses had no conserved regions. Hence, the detection spectrum of RT-PCRs described earlier (16) could not be broad because the primer sets used were from only a single TAstV sequence (Accession no. AF206663). To develop RT-PCR with improved detection spectrum, two sets of novel primers were designed on the basis of these two available TAstV sequences. They were referred to as the AFCP set and TAPG set, respectively. The AFCP set was from the capsid gene of TAstV-2 genome (Accession no. AF206663) and was used in the monoplex RT-PCR and multiplex RT-PCR along with the SRV primer set. The TAPG set was a degenerate primer set covering the polymerase genes of two TAstVs (Accession no. Y15936 and AF206663) and was used in the monoplex RT-PCR. The previously available SRV primer set was also used in this study. The application of the monoplex RT-PCR with primer set SRV demonstrated that this assay was very sensitive and specific in spite of a narrow detection spectrum. For example, it can only amplify the TAstV2001 isolate, but not the TAstV1987 and TAstV1997 (SRV p14) isolates. On the contrary, the monoplex RT-PCR with primer set AFCP can amplify both TAstV1987 and TAstV1997 (SRV p14), but not TAstV2001. Monoplex RT-PCR with degenerate primers and multiplex RT-PCR with both SRV and AFCP primer sets overcame the pitfall of the monoplex RT-PCR with either the SRV or AFCP primer set alone and showed good specificity and a wider detection spectrum. A single-stranded RNA internal control (IC) reagent developed in our laboratory was used with multiplex RT-PCR as a coamplification template (36). The results from the multiplex RT-PCR in the presence of the IC showed that the IC is a good indicator to decrease the false negative rate of the assay.
Sensitivity and specificity of the tests developed in this report were not calculated because no gold standard test is currently available for TAstVs.
In general, for diagnostic purposes, the EIA with polyclonal antibodies is probably the better test to use at this time. The RT-PCR is good for identification of the two recognized serotypes of the viruses. In the future, when more information becomes available on possible unrecognized turkey astroviruses, the RT-PCR might be a more useful test.
We thank R. N. Dearth for his technical assistance and Ying Zhang (Molecular and Cellular Imaging Center, Ohio Agricultural Research and Development Center, The Ohio State University) for sequencing. We acknowledge Dr. H. John Barnes (College of Veterinary Medicine, North Carolina State University) for providing initial fecal samples. We thank Dr. Akbar Ali and Dr. Donald L. Reynolds (Department of Veterinary Microbiology and Preventive Medicine, Iowa State University) for providing primer information. We appreciate the valuable critical review provided by Dr. Srinand Sreevatsan and Dr. Jeff LeJeune. This work was supported by the U.S. Egg and Poultry Association (grant 545).
Primers used for RT-PCR
The AB-ELISA titers of polyclonal hyperimmune antisera against purified TAstV1987 or TAstV2001
Detection of turkey astroviruses by EM, monoplex RT-PCR, multiplex RT-PCR with internal control, and polyclonal AC-ELISA
Frequency of viruses detected by direct EM and RT-PCR in fecal samples of affected poults with enteritis