Introduction

Avian astroviruses were first reported as a cause of fatal hepatitis [1] in ducklings and were subsequently reported in chickens [30], turkeys [14] and guinea fowl [5]. These viruses belong to the family Astroviridae, which is currently divided into two genera, Mamastrovirus and Avastrovirus, whose members infect mammals and birds, respectively.

Astroviruses have a single-stranded, positive-sense RNA genome containing three open reading frames (ORFs). ORF1a encodes a nonstructural protein, ORF1b encodes the RNA-dependent RNA polymerase, and ORF2 encodes the capsid protein [29]. The length of the ORFs varies among species and serotypes of the virus. Molecular characterization has shown that ORF1b is the least divergent, while ORF2 is the most divergent [27].

Multiple astroviruses have been reported in chickens, namely, avian nephritis virus types 1-3 (ANV1-3) and chicken astrovirus (CAstV). In turkeys, turkey astrovirus types 1 and 2 (TAstV-1 and TAstV-2) and ANV have been detected [3, 6]. ANV, which was formerly classified as a picornavirus, is primarily associated with tubulonephrosis, visceral gout, diarrhea, and retarded growth in young chicks [23] and can also infect turkeys, ducks, geese and pigeons [2, 20, 31]. Astroviruses of chickens and turkeys are associated with tenosynovitis and locomotion problems [6]. Chicken astrovirus (CAstV) is frequently detected in broiler flocks affected with runting and stunting syndrome [20, 24]. Serological surveys have shown that antibodies to CAstV are also widespread in broiler-breeder flocks and in some turkey flocks [3, 28]. Turkey- and chicken-origin astroviruses are immunologically and genetically different. In turkeys, TAstV-2 is more commonly detected than TAstV-1 [13, 22]. Of the chicken and turkey astroviruses, ANV appears to be the most phylogenetically distinct, but it is more closely related to TAstV-1 than to other avian astroviruses [9, 21].

Astrovirus, either alone or in combination with other enteric viruses, has been detected in enteric disease conditions of turkey poults, e.g., poult enteritis syndrome (PES), poult enteritis mortality syndrome (PEMS), and poult enteritis complex (PEC) [12, 15, 18, 19, 25, 26]. Light turkey syndrome (LTS) is a relatively new enteric disease syndrome of Minnesota turkeys in which the birds weigh less than their standard breed character at marketing age. Although a decrease in marketing weight is the primary criterion used for categorizing a flock as having LTS, other clinical signs such as depression and diarrhea, particularly during brooding, have also been reported [4, 16]. Inoculation of two-week-old turkey poults with a fecal suspension from birds with LTS resulted in clinical signs of depression, huddling, lack of uniformity, and reduction in body weight beginning at the age of seven weeks. It has been suggested that viral enteritis at an early age may set up conditions for the development of LTS later in adult turkeys [16].

Previous studies have indicated the presence of TAstV-2, orthoreovirus, and avian rotavirus in intestinal samples of LTS- and PES-affected turkeys [10, 11, 12, 16]. However, information on the presence of TAstV-1 and ANV in such flocks is lacking. This study was undertaken to compare the prevalence and genetic diversity of TAstV-1 and ANV in LTS and non-LTS flocks of Minnesota turkeys. The genetic diversity of both viruses was also compared with those detected in cases of PES.

Materials and methods

Samples and RNA extraction

In 2011, we collected 80 fecal pools from LTS flocks and 40 from non-LTS flocks in Minnesota. Five fecal pools were collected from four different LTS flocks at 2, 3, 5 and 8 weeks of age. Similarly, five fecal pools were collected from two non-LTS flocks at the same ages. For comparison with LTS and non-LTS strain sequences, we included 130 intestinal-content samples from cases of PES that were tested at the Minnesota Veterinary Diagnostic Laboratory between 2009 and 2014. The 120 pools from LTS and non-LTS flocks were tested for the presence of enteric viruses using a multiplex reverse transcription-polymerase chain reaction (mRT-PCR) assay [10], the results of which have been reported previously [16]. These same pools were tested in this study for the presence of TAstV-1 and ANV by RT-PCR. Briefly, the fecal pools were homogenized to make 10 % suspensions using sterile PBS (pH 7.4), followed by centrifugation at 1200 × g for 20 min at 4 °C. The supernatants were decanted, and total RNA was extracted using a QIAamp Viral RNA Mini Kit (QIAGEN, Valencia, CA).

Reverse transcription polymerase chain reaction (RT-PCR)

Primers were designed for the RdRp genes of TAstV-1 and ANV (Table 1). The RT-PCR was performed using a QIAGEN One Step RT-PCR Kit. The reaction was done in a volume of 25 µL. The RT step was performed at 50 °C for 30 min, which was followed by denaturation at 95 °C for 15 min and 35 polymerase chain reaction (PCR) cycles with denaturation at 94 °C for 1 min, annealing at the appropriate temperature (Table 1) for 1 min and elongation at 72 °C for 1 min. The final elongation was at 72 °C for 10 min. PCR products were confirmed by analysis in an ethidium-bromide-stained 1.2 % agarose gel followed by visualization under a UV transilluminator.

Table 1 Primers used for the detection of turkey astrovirus 1 and avian nephritis virus
Full size table

Sequencing and phylogenetic analysis

The PCR amplicons, after purification using a QIAquick PCR Purification Kit (QIAGEN), were submitted to the University of Minnesota Genomic Center (UMGC) for sequencing. Sanger sequencing was performed separately with forward and reverse primers for the respective genes. Forward and reverse sequences were aligned together to generate a consensus sequence using Sequencher software version 5.1 (http://genecodes.com/). The obtained sequences were compared with the reference strains available in GenBank and were aligned using CLUSTAL W in MEGA 6. Phylogenetic trees were constructed using the maximum-likelihood statistical method based on the general time-reversible model (GTR) for the nucleotide (nt)-based tree in MEGA 6.0 for both TAstV-1 and ANV. The reliability of different phylogenetic groupings was evaluated by the bootstrap method with 1,000 replications. The p-distance method was used to calculate the evolutionary distances (percent nt and amino acid [aa] identities).

Nucleotide sequence accession numbers

The obtained sequences were submitted to the GenBank database under the following accession numbers: TAstV-1, KT355292 to KT355320; ANV, KT355270 to 355291.

Results

Prevalence of TAstV-1 and ANV

Of the 80 LTS pools tested, 16 (20.0 %) and 11 (13.8 %) were positive for TAstV-1 and ANV, respectively. For non-LTS flocks, these numbers were 8 (20.0 %) and 5 (12.5 %), respectively. No dual infection with TAstV-1 and ANV was found in any of the samples from either LTS or non-LTS flocks. When stratified by age (2, 3, 5 and 8 weeks), the maximum number in LTS flocks was positive for viruses at 5 weeks of age (9 of 20, 45 %), followed by 8 weeks of age (7 of 20, 35 %). In the non-LTS flocks, 4/10 (40 %), 2/10 (20 %), 5/10 (50 %) and 2/10 (20 %) pools were positive at 2, 3, 5 and 8 weeks of age, respectively (Table 2). Of the 130 samples from PES flocks, 19 (14.1 %) and 11 (8.5 %) were positive for TAstV-1 and ANV, respectively. Dual infection with both TAstV-1 and ANV was found in three (2.3 %) samples from PES flocks. Thus, of the 130 PES samples, 16 and 8 were found singly positive for TAstV-1 and ANV, respectively.

Table 2 Prevalence of turkey astrovirus 1 and avian nephritis virus in LTS and non-LTS flocks according to agea
Full size table

Sequence comparison and phylogenetic analysis

Out of the 24 TAstV-1-positive samples from LTS and non-LTS flocks, the RdRp gene was sequenced for 10 samples (7 LTS and 3 non-LTS). Similarly, 11 (8 LTS and 3 non-LTS) of 16 samples that were positive for ANV were sequenced. Of the PES samples, 19 and 11 sequences of TAstV-1 and ANV, respectively, were determined. Thus, a total of 29 TAsV-1 and 22 ANV strains from all three types of flocks (LTS, non-LTS, and PES) were sequenced. A BLAST search of amplicon sequences obtained using primers specific for TAstV-1 ORF1b confirmed the identity of all 29 sequences as TAstV-1 (Fig. 1).

Fig. 1
figure 1

Phylogenetic tree of 29 Minnesota TastV-1 isolates based on partial nucleotide sequences (650 bp) of the RdRp region. The phylogenetic tree was constructed using the maximum-likelihood statistical method based on the general time-reversible (GTR) model with 1000 bootstrap replicates. Sequences from this study are in bold

Full size image

Phylogenetic analysis revealed that all TAstV-1 strains in this study grouped together with previously reported reference strains of TAstV-1 taken from GenBank, but they clustered differently from TAstV-2 (Fig. 1). Further, p-distance analysis of PES sequences showed nt sequence identity of 92-100 %. The nt sequence identity was 97-100 % and 99-100 % for LTS and non-LTS strains, respectively (Table 3). Collectively, the study sequences showed 90-100 % nt sequence identity with the reference strains of TAstV-1. The numbers of substitutions in TAstV-1 from LTS, PES and non-LTS cases, were 7, 12 and 6, respectively. Twelve amino acids substitutions were observed in the deduced aa sequences of our study when compared to those in the reference strain (Y15936) (Supplementary Table 1).

Table 3 Percent identity of nucleotide and amino acid sequences among TAstV-1 strains from different types of flocks
Full size table

Based on 10 % nt sequence divergence, sequences of ANVs from this study and 24 previously published sequences clustered in three groups (groups 1 to 3) in the polymerase gene phylogeny (Fig. 2). All of our ANV strains except one (KT355287/UMN18/PES/2014) clustered in group 1, along with turkey ANV-2 strains from the USA (HQ188697, HQ188695, DQ324838 and DQ 324837) and guinea fowl ANV-2 strains from Poland and Brazil (HQ317722 and HQ 378374, respectively). The ANV strain KT355287/UMN18/PES/2014 clustered in group 3 and was closely related to pigeon ANV (HQ889774) and chicken-origin ANV-1 strains from the USA, the UK, South Korea and China. Group 2 was formed by chicken ANV-1 strains from Japan.

Fig. 2
figure 2

Phylogenetic tree of 22 Minnesota turkey ANV isolates based on partial nucleotide sequences (650 bp) of the RdRp region. The phylogenetic tree was constructed using the maximum-likelihood statistical method based on the general time-reversible (GTR) model with1000 bootstrap replicates. Sequences from this study are in bold

Full size image

The nucleotide sequence identity determined by p-distance calculation of 22 ANVs from this study among themselves was 88-100 %. The ANVs from PES (n = 11), LTS (n = 8) and non-LTS (n = 3) flocks showed 88-100 %, 90-100 % and 100 % identity, respectively, among themselves (Table 4). Since a complete genome sequence of turkey ANV was not available in GenBank, we used a partial sequence of NC-TK-SEP ANV-670-2005 (HQ188697) from the US as a reference. The study sequences showed 84-96 % nt sequence identity to the reference strains, which resulted in a high level of identity of the deduced aa sequences (91-100 %) (Table 4). Using the HQ188697 sequence as a benchmark for alignment of aa sequences (positions 13 to 233), no insertions or deletions in the portion of RdRp studied was observed (Supplementary Table 2). In ANV from LTS samples, there were 16 amino acid substitutions compared to non-LTS samples, in which only five substitutions were observed. One substitution, D158N, was observed only in non-LTS samples. There were 29 aa substitutions, of which eight (K14R, R17K, D68V, L81F, D158N, V206S, K216I and T219I) were not observed in ANVs used for comparison. Five substitutions (N49D, R57K, Y118F, D164N and D183N/S) were observed in 20, 6, 8 and 16 ANV sequences of this study, respectively (Supplementary Table 2).

Table 4 Percent identity of nucleotide and amino acid sequences among ANV strains from different types of flocks
Full size table

Discussion

Enteritis is one of the most important diseases affecting turkeys in the U.S. Two syndromes, namely PES and LTS, have recently been recognized in Minnesota turkeys. This work is a continuation of our previous research in which we reported the presence of TAstV-2, RV and ARV from LTS-affected turkeys [10, 11, 16]. Such studies are needed to determine the virus(es) present in enteritis-affected birds so that preventive and control measures can be initiated to minimize the morbidity and mortality (if any), and subsequently, the economic losses to the turkey growers.

In this study, we determined the prevalence of TAstV-1 and ANV in LTS and non-LTS turkey flocks as well as PES-affected flocks in Minnesota. In addition, we characterized the detected viruses at molecular level. In a recent study, Mor et al. [16] tested Minnesota turkey flocks with historically high incidence of enteric disease and found the maximum load of enteric viruses such as TAstV-2, RV, and ARV, alone or in combination in LTS and non-LTS flocks at 2–3 weeks of age, which decreased at subsequent samplings. In the present study, however, the maximum number of samples was positive for TAstV-1 and ANV at 5 weeks of age in both LTS and non-LTS flocks. A decrease in the prevalence of these two viruses at eight weeks was similar to that described for other enteric viruses such as TAstV-2, orthoreovirus, and avian rotavirus [12, 16].

Domianska-Blicharz et al. [7, 8] conducted a molecular survey of astroviruses in 77 (1-19 weeks of age) and 156 (1 day to 20 weeks of age) turkey flocks in 2008 and 2009-2012, respectively, in Poland. The prevalence of TAstV-1 and ANV was reported to be 11.6 % (9/77 flocks) and 1.29 % (1/77 flock) in 2008 and 9.6-15.4 % and 0.0 % in 2009-2012, respectively. The prevalence was higher in Minnesota turkeys than in Polish turkeys, which could be due to collection of samples from young birds in the current study, whereas in the Polish study, samples were collected up to 20 weeks of age. Moura-Alvarez et al. [17] reported high prevalence of TAstV-1 (49.0 %) and ANV (35.5 %) in Brazilian turkeys. Pantin-Jackwood et al. [18] surveyed 33 commercial turkey flocks in the USA and reported a prevalence of 15.4 % and 12.5 % for TAstV-1 and ANV, respectively, which is similar to the prevalence in LTS and PES flocks in Minnesota (this study). Interestingly, none of the samples from LTS and non-LTS flocks showed co-infection with two viruses. However, co-infection was seen in PES-affected flocks. Under field conditions, it is possible that co-infection may lead to more-adverse effects than those induced by an individual virus. In fact, we have previously reported that infection with multiple enteric viruses greatly affected body weight gain in turkey poults in addition to causing severe enteritis [12]. This is also supported by a study by Spackman et al. [25], who reported that co-infection with three enteric viruses resulted in severe enteritis and other effects not observed in single infections with the same viruses. Future experimental studies with TAstV-1 and ANV are needed to determine their impact in cases of LTS and PES.

Phylogenetic analysis of TAstV-1 sequences based on 15 partial ORF1b sequences from the GenBank database revealed grouping of TAstV-1 sequences from the USA, Brazil, Poland and Croatia. Sequences from PES samples that were collected from 2009 to 2014 showed more divergence than samples taken at a single time point from LTS and non-LTS flocks. Also, the variation was higher in sequences from enteritis samples (LTS and PES) than in those from clinically healthy birds (non-LTS). Similarly, Domianska-Blicharz et al. [7, 8] reported 97.4-99.8 % and 90.1-98.0 % nt sequence identity among TAstV-1 isolates from a single year (2008) and from a four-year sampling period (2009-2012), respectively. However, in contrast to our findings, Pantin-Jackwood et al. [20] reported 89.9-98.2 % nt and 90.6-100 % aa sequence identity in TAstV-1 sequences detected in 2005 from clinically healthy poults (2-4 weeks of age) from North Carolina.

Phylogenetic analysis of ORF1b sequences of ANV with previously reported sequences in GenBank grouped these viruses with ANV-like sequences detected in turkeys, except for one PES sequence UMN18 (KT355287) from Dakota County in Minnesota. This UMN18 sequence grouped with ANV-like sequences detected in chickens from the US and other countries. This divergent sequence, detected for the first time in 2014, indicates the importance of continuous molecular screening of enteric viruses including the complete genome analysis of the detected viruses. Complete genome analysis will determine if this is a divergent strain or a recombinant of chicken and turkey ANV-like viruses. The ANVs from PES and LTS birds showed 88-100 % nt sequence identity, which is similar to the 86.2-99.6 % nt sequence identity reported in ANV sequences from different states of the USA [20].

In this study, we detected nt substitutions in the RdRp regions of both TAstV-1 and ANV. More substitutions were observed in sequences from LTS and PES samples than in those from non-LTS samples, which may indicate that these substitutions are related to the pathogenicity of these viruses and their ability to cause enteritis in turkey poults. Pantin-Jackwood et al. [20] also reported that despite being more conserved, the polymerase gene shows a large amount of genetic variation in astroviruses. These authors also measured the ratio of synonymous to nonsynonymous nucleotide substitutions to determine the extent of possible selective pressures influencing the evolution of the polymerase gene in avian astroviruses and found that this gene is subjected to positive selection (dN > dS). However, further experimental studies are needed to confirm or refute this hypothesis.

This study documented the prevalence, age-related distribution, and molecular characterization of TAstV-1 and ANV strains associated with LTS and non-LTS flocks. Also, a considerable proportion of TAstV-1 and ANV infection was observed in non-LTS cases, which emphasizes the need for further studies. We studied only a part of the gene. Even in this conserved gene, there were unique amino acid changes. Further studies of the capsid gene are needed to better understand the genetic diversity in avian astroviruses.