Surveillance of Avian Reoviruses in a Single Broiler Chicken Company from the California Central Valley Open Access

Avian reovirus (ARV) is a ubiquitous poultry pathogen that belongs to the genus Orthoreovirus, species Orthoreovirus avis in the Spinareoviridae family (1). It is a nonenveloped virus with a segmented, double-stranded RNA genome. The viral genome consists of 10 segments classified based on their electrophoretic mobility in large (L1, L2, L3), medium (M1, M2, M3), and small (S1, S2, S3, S4) (2,3,4), encoding for both structural and nonstructural proteins. Of these segments, the S1 gene encodes for three proteins. One of them is sigma C, which is involved in virus–host interactions and elicits specific neutralizing antibodies. Moreover, it is responsible for part of the genetic variability of the virus and is extensively used as a genetic marker for genotyping (5,6). Other segments, particularly M2 and L3, have also been found to play a role in the viral genetic variability (6). Because of its RNA-segmented genome, ARV is prone to mutations, reassortments, and genetic recombinations. These events can lead to the emergence of new genetic and antigenic variants that bypass the immunity induced by conventional vaccine strains (6,7). Seven genotypic clusters (GC) have been identified based on partial S1 gene characterization (8) to date. Although, molecular analysis of sigma C has offered valuable insights into ARV epidemiology, the information provided can still be used in evaluating ARV prevention programs. Additionally, supplementation with pathological, serological, and antigenic information is crucial for virus characterization and selection for autogenous vaccine preparation.

ARV primarily infects meat-type chickens and turkeys, causing malabsorption, runting-stunting syndrome, myocarditis, and immunosuppression (9,10). The most common clinical outcome of current ARV antigenic and genetic variant infections in broilers (since 2011 in the United States) is viral arthritis/tenosynovitis, which causes hock joint swelling due to lymphocyte infiltration, edema of the gastrocnemius and digital flexor tendons and, in chronic cases, fibrosis that could end in tendon rupture, hemorrhages, and bruising (11). These issues lead to a lack of uniformity in broiler flocks and increased condemnations at processing facilities, causing significant economic losses for poultry producers in addition to welfare issues in the affected birds (12). Viral arthritis/tenosynovitis was initially reported in poultry in the 1950s, at that time poultry flocks were protected with the development and use of commercial live modified and inactivated vaccines derived from ARV S1133 (13). For nearly 60 yr, this vaccination strategy helped in preventing vertical transmission and providing maternal antibodies to their progeny protecting them from early infection (14,15).

Because of the economic significance of the disease, impact on animal welfare, and to design preventative measures, it is essential to conduct thorough and consistent surveillance of the circulating ARV field strains. In some countries, inactivated autogenous vaccines are used in combination with live vaccine priming, to protect against pathogenic ARV. Autogenous vaccines should be continuously reformulated to match circulating field stains, making surveillance studies necessary to detect and monitor them (2,11,14). Previous studies have indicated that the use of autogenous vaccines helped in controlling the frequency of ARV cases over the past few years, lowered ARV loads in infected birds, and reduced the incidence of clinical ARV manifestations (8,13,16).

In this study we genotyped, using partial S1 sequences, ARV isolates obtained from field outbreaks in broiler flocks from a single company in the California Central Valley between 2020 and 2022. We also analyzed the data from reported cases including frequency of detected GCs per year, autogenous vaccine composition, number of cases per year, age of birds at submission, and viral loads per case per year from 2017 to 2022. The goal of this study is to contribute to the understanding of the genetic variability of ARV in California, look at patterns regarding the effectiveness of surveillance and prevention programs, and provide insights to develop better prevention and control plans for genetic, pathogenic, and antigenic ARV variants in commercial broiler flocks.

Cases were submitted to the California Animal Health and Food Safety laboratory (CAHFS) based on clinical signs and symptoms associated with clinical ARV in the field for necropsy and sample collection. Each case submission consisted of 5–8 chickens. After a thorough necropsy and gross pathology assessment, tissue samples (tendons, joints, hearts, and intestines) were collected for real-time reverse-transcriptase PCR (RT-qPCR), from birds with tenosynovitis and pericarditis.

Viral RNA extraction was done using the QIAamp Viral RNA Mini Kit (Qiagen®, Valencia, CA) on a QIAcube. RT-qPCR was performed using the OneStep RT-PCR kit (Qiagen) to confirm isolates obtained from cell cultures showing CPEs (cytopathic effects) consistent with ARV. For this RT-qPCR, primers targeting the M1 gene of ARV were used (3).

Samples with confirmed presence of ARV by RT-qPCR were processed for virus isolation. Tissues were minced and homogenized individually in viral transport media (VTM) using a gentleMACS™ Octo dissociator (Milteny Biotech, Bergisch Gladbach, Germany). The homogenized samples were diluted with VTM to a 1:10 w/v ratio and filtered with a 0.2-micron sterile syringe filter. Confluent chicken embryo liver (CEL) cells were produced according to a published protocol (17) and inoculated with 1 ml of filtered tissue homogenate in 12.5-cm² tissue culture flasks. Cultures were incubated for 1 hr at 37 C. Hank’s balanced salt solution (2 ml) was used to rinse the cells. Each flask with cells was supplemented with 1% fetal bovine serum (2.5 ml) and placed in a 5% CO₂ incubator at 37 C for 5 days. The flasks were observed daily for distinct cytopathic effect, specifically syncytia formation (18). Confirmatory RT-PCR was performed amplifying the conserved region of the S4 gene of the ARV genome (19). Samples showing no detectable CPEs after 5 days were frozen and thawed three times and passaged a second time onto fresh CELs (6). Submissions with samples from which ARV was isolated were categorized as cases (19).

A 1088-bp segment of the ARV S1 gene was amplified using the primers designed by Kant et al. (20) through an RT-PCR assay formulated by Sellers (8) with the following modifications: Reverse transcription was performed using the Superscript™ IV Reverse Transcriptase kit (Invitrogen, Carlsbad, CA) according to the manufacturer’s protocol with the addition of RNaseOUT (Invitrogen). The PCR step was performed using the Qiagen TAQ Master Mix kit adjusting for MgCl₂ to a final concentration of 3.2 mM, P1/P4 primers (10 µM each) and thermal cycling conditions of initial denaturation at 94 C for 2 min followed by 35 cycles (denaturation of 1 min at 94 C, annealing for 30 sec at 55 C, and extension for 2 min at 72 C) with a final extension of 10 min at 72 C. A 1% agarose gel containing GelGreen® Nucleic Acid Stain (Biotium, Fremont, CA) was prepared to visualize RT-PCR products. Bands were visualized in a Gel Doc™ EZ imager (Bio-Rad Laboratories, Richmond, CA). The PCR products obtained from positive samples were purified using the QIAquick® PCR purification kit according to the manufacturer’s protocol and nucleic acid concentration was determined using the Qubit™ dsDNA BR Assay Kit (Invitrogen).

The purified PCR products were submitted for Sanger sequencing with forward and reverse primers (P1 and P4). The obtained sequences were corrected for ambiguities and aligned with 130 previously published sequences along with four commercial vaccine sequences, S1133 (L39002, AF330703), 1733 (AF004857), and 2408 (AF204945) using MAFFT v7.490 in Geneious Prime® 2023.2.1 (21). The RAxML method was used for phylogenetic analysis using the GTR GAMMA I nucleotide model. A “rapid Bootstrapping and search for best-scoring ML tree” algorithm was applied with 1000 bootstrap replicates using 456 parsimony random seed value (22).

In this study, clinical data, that is, case numbers, age of the submitted chickens, and viral load of the collected samples, between 2017 and 2022, were analyzed and compared with autogenous vaccine composition and GC frequencies looking for trends that might denote the effectiveness of the prevention/control strategy.

One-way ANOVA followed by Tukey’s multiple-comparison test with a single pooled variance were performed for statistical analysis of the data using GraphPad Prism Version 9.5.1 (23).

Between 2020 and 2022, 65 ARV suspect submissions from a single broiler company in the California Central Valley were received at CAHFS. From the submitted birds’ 54 tendons (72.9%), 10 intestines (13.5%), 8 hearts (10.8%), 1 joint fluid (1.3%), and 1 fecal sample (1.3%) were collected for ARV testing for a total of 74 samples. After RT-qPCR confirmation, ARV was detected by RT-qPCR in 50/74 samples (67.5%) and processed for virus isolation in CEL cells. ARV was successfully isolated from 40/50 (80%) samples. Thirty-nine out of 40 (97.5%) isolates were obtained from tendons and only 1 (2.5%) was obtained from an intestine.

Partial S1 gene sequences of 1088 bp, covering most of the sigma C coding region (20), were obtained from 35 out of 40 samples (87.5%) using Sanger sequencing. These sequences were submitted to GenBank and published with accession numbers: OR556002 to OR556028 and OR815308 to OR815315. Fig. 1 shows a phylogenetic tree that was constructed using these sequences along with 130 ARV reference and 4 vaccine sequences from previous publications. Reference sequences belonged to all seven genotypic clusters. The sequences obtained from 2020 to 2022 were grouped in six out of the seven reported clusters. Their distribution in percentages were as follows: GC1 (5.71%), GC2 (31.43%), GC3 (5.71%), GC4 (17.14%), GC5 (25.71%), and GC6 (14.30%).

In this study, 2 out of 35 sequences (5.7%), were grouped in this cluster. These two sequences shared 75% and 76% homology with the reference strain, S1133. Amino acid identity between these two sequences was 86% and overall homology within the cluster ranged from 73% to 100%.

Most of the viruses isolated between 2020 and 2022 (n = 11, 31.4%) were represented in this cluster. Three isolates were identified in 2020, five in 2021, and three in 2022. Homology within this cluster was 66%–100% and, among the studied isolates, it ranged from 67%–99%. Amino acid identity of these 11 isolates with S1133 vaccine strain was very low (50%–54%).

Two isolates (5.7%), both from 2021, were grouped in this cluster and showed 86% amino acid homology between each other. Amino acid identity of these two sequences with S1133 was 50%. One of the isolates grouped with and shared 91% amino acid identity with an ARV variant (OP038530_118583/USA/AR/2017), which was previously classified as GC7. The other isolate only shared 86% homology with the same sequence.

Six viruses (17.1%), all isolated in 2021, grouped in this genotypic cluster. Amino acid homology within GC4 ranged from 64% to 99%. Similarity among the six isolates was 65%–98% and they shared 46%–47% identity with S1133.

GC5 was the second most prevalent cluster after GC2 (n = 9, 25.7%). All these viruses were isolated in 2021 and possessed 92%–99% amino acid similarity with one another. They are distant from S1133, showing only 45% similarity. All viruses in this genotypic cluster show high homology (92%–99%) except one isolate (AF354219_GEI10 97M/DEU/1997), a 1997 German isolate from a case of malabsorption, that shares 79% similarity with the other isolates grouped in the same cluster.

Five isolates (14.3%) were grouped in this cluster with amino acid homologies between 94% and 99%. Four of these viruses were isolated in 2021 and one was identified in 2022. They are 48% identical to S1133 and the overall similarity between most isolates within this cluster range between 90% and 100%. The exception are two reference sequences, AF297215_918/TWN/1992, isolated from a malabsorption case in Taiwan, and MG822697_15-1221/CAN/AB/2015, a Canadian isolate obtained from tendons, that showed 64% to 69% homology with all other viruses in cluster 6 and 88% similarity with each other.

None of the isolates from this study were grouped in this cluster.

Homologies to S1133 throughout the investigated years are shown in Table 1. A trend of reduction of the homologies to S1133 in all clusters can be seen throughout the years.

The detection frequency of genetic variant viruses belonging to each of the genotypic clusters between 2017 and 2022 as well as the respective autogenous vaccine composition are shown in Fig. 2. In 2017, most cases were associated with ARVs belonging to GC1 (n = 18) and GC6 (n = 15). The autogenous vaccine used in 2017 contained ARVs from genotypes GC1 and GC5. Because GC6 detections increased and this virus was not in the vaccine, a GC6 genetic variant isolated from the field was included in the autogenous vaccine formulated for 2018. Subsequently GC6 detections reduced in frequency in subsequent years. Similarly, surveillance during 2018 and 2019 showed an increased prevalence of ARVs belonging to GC2. Therefore, the vaccine was upgraded by adding a field-isolated GC2 genetic variant for 2020. This caused a reduction in the detection of GC2 in 2022. As shown in Fig. 2 genetic variants have been incorporated to the vaccine without eliminating strains in the vaccine.

A decreasing trend can be observed in ARV cases from 2017 to 2022 except for 2020 and 2021 (Fig. 3a). Age at submission was not clearly affected by the interventions made in the ARV prevention program (Fig. 3b). It ranged from 5 to 50 days with an average and median of 26 days. The average ages fluctuate between 21 ± 13 days (2020) and 30 ± 11 days (2019).

Viral loads in the processed samples showed a reduction trend from 2017 to 2020. The significance of this reduction (P ≤ 0.05) can be seen in Fig. 4. From 2020 to 2022, viral loads remained stable. Overall, there was a continued numerical reduction in viral loads year after year throughout the study period.

Avian reovirus is responsible for many clinical problems in infected chickens, including malabsorption, runting–stunting syndrome, and myocarditis. ARV genetic variants, isolated from broiler and breeder flocks in the United States since 2011, have been associated with clinical arthritis and tenosynovitis (10). There are many published reports of detection and characterization of these genetic variants around the world (6,8,10,11,20,24,25). In California, ARV genetic variants were first described in broilers in 2015 and, to date, belong to six genotypic clusters based on sigma C sequences (6).

In this study, ARV variants were detected, isolated, and genotyped from samples received between 2020 and 2022 from a single broiler company. A representative number of all ARV samples obtained from field cases of arthritis/tenosynovitis (50 out of 74) were selected for virus isolation and genotyping. The selection was made based on clinical significance with the assumption that broiler chickens suffering from this condition come from the same breeder flocks vaccinated with the same ARV program. In addition, broilers were raised in farms with standard biosecurity and management systems. We believe that this makes our sampling a good representation of the company situation. From the selected 50 samples, we were able to isolate 40 variant viruses. It is important to notice that isolation was performed in CELs. Some avian reoviruses grow at higher titers in chicken embryos, which might be a better substrate for isolation to perform characterization. The use of embryonated eggs should be considered when isolation in CELs is not successful. Thirty-five isolates (88%) were successfully amplified and partially sequenced. Lack of amplification in the remaining isolates might be attributed to poor quality and/or quantity of the extracted RNA leading to failed sequencing. Our findings show that the current circulating genotypic clusters belong to the same six GCs described in 2019 (6). However, the distribution shows a shift with GC2 and GC5 being the predominant strains, compared with GC1 and GC6 in the previous report (6). Moreover, a considerable reduction in the amino acid identity compared with S1133 was perceived. On average, this homology reduction was 4.5%, with GC1 being the lowest, 1.6%, whereas GC5 sequences showed the highest homology (6.6%) difference (Table 1). This observation illustrates ARV’s constant genetic variability and highlights the poor virus neutralization in the current control strategy. It is important to add that genotyping methods are not standard; for example: GC3 and GC7 are very close genetically; the use of smaller segments because of trimming during the analysis might create these discrepancies. Future research should focus on strategies that effectively reduce shedding to lower chicken infection and, subsequently, the virus genetic variability potential in the population.

As commercial live attenuated and inactivated vaccines have failed to protect flocks from the disease caused by variant ARV strains, autogenous vaccines have been extensively used in the United States and Canada as a measure to control the disease (14,16). Interesting trends were detected when we compared genotypic cluster frequencies throughout the years. Despite its inclusion in the autogenous vaccine, almost one-third of the ARV isolates belonged to GC2 in this study. This is also reported by other studies, where the addition of GC2 genetic variants in autogenous vaccines were unable to control ARV cases caused by strains of the same genotype (8,16). This might be because of the use of partial sequences of only one gene in the ARV’s genetic characterizations. On the other hand, studies have shown reduction of ARV cases when the same genotype strains were included in autogenous vaccines (8,16). Though overall reduction in ARV cases was observed, in the current report we saw an increased number of cases from the genetic variants that belonged to GC5. The persistence of GC5 highlights flaws in the ARV control strategy. It might be that the GC5 isolate selected to be part of the vaccine was not the best match for the GC5’s causing disease in the field. We also observed high differential homology among the distinct genogroups. The highest difference was 6.6% among GC5 isolates between 2017 and 2022. This suggests that these viruses accumulate changes that might help them become more fit and able to bypass the host immunity. These results emphasize the need for a comprehensive classification of the strains selected for vaccine production, that is, molecular, pathogenic, and antigenic characterizations to select the best representatives in the field. Poor isolate selection might alter vaccine effectiveness, reduce protection, and allow the viruses to escape from the immunity elicited by these products. In addition, it is important to emphasize that vaccine quality plays a crucial role in vaccination success and disease control. Virus titers before inactivation, correct inactivation to maintain intact virus epitopes, adjuvant quality, shipping, storage conditions, and so on also play an important role in vaccine effectiveness. The vaccines used for the control exert selective pressure on the ARV variant populations inducing limited cross protection and exponentially multiplying the genetic and antigenic diversity of the virus (26). Similar events have been reported in the continuous use of vaccines against RNA viruses (14,27).

The analysis of ARV trends by company is crucial in the evaluation of the effectiveness of an ARV control strategy. These strategies not only involve vaccines but also management to reduce ARV load in the environment. The use of this preventative strategy has contributed to the reduction of the ARV cases in the California Central Valley in recent years. It is important to note that ARV incidence shows a reduction trend (Fig. 3a). It is common to see fluctuating outbreaks of ARV in broiler chickens where, in some years, the disease is very prevalent and in subsequent years cases are scarce (8). Our results show a reduction trend starting in 2017 and ending in 2022. Cases in 2020 and 2021 were moderately elevated. Age at submission should also be studied, because one of the goals of the control strategy is to push the challenge after 3 wk of age when flock susceptibility to ARV lies below 25% (13). Our results show a fluctuation of ages at submission throughout the years and might be related with inconsistencies in the prevention program. Palomino-Tapia found that the age of birds at submission was distributed similarly to what was reported in this investigation (16). It is important to notice that this variable can be affected by parameters that are not related with the viral infection and disease. Despite this, it provides information that can add to other, less variable, parameters. The last trend worth following is viral load of the submitted samples. Although indirect and affected by different variables such as time of infection and sample collection, it can be also associated with environmental viral load and can be used to fine-tune the program in case viral loads are high. Our results showed a consistent reduction of the viral load in these samples over the years, which is consistent with the case reduction seen between 2017 and 2022 and might be related with better virus selection to formulate autogenous vaccines. A good autogenous (inactivated) vaccine represents the challenge strains in the field and can neutralize infective virus in chickens and reduce shedding (28,29,30). This reduction in shedding can be also correlated with the viral load reduction we observed (Fig. 4).

A good ARV prevention program should rely on surveillance, vaccination, and management for the reduction of the viral load in the field. In addition, trends in case numbers, age at submission, viral load, and prevalent genotypes should be observed to make sure the program is working. Virus neutralization can be used to test the effectiveness of vaccination programs using breeder serum to measure neutralization capabilities against vaccine strains. In addition, commercial ELISA kits can be used to check vaccine application through seroconversion, keeping in mind that these ELISAs cannot differentiate between vaccine, field challenge, variant, or conventional-strain-generated antibodies (26).

The findings shared in this manuscript illustrate an ARV control strategy used in a single company in the U.S. West Coast. Factors like biosecurity, geographical location, and breeder source change for every company and should be also taken into consideration when designing control and prevention strategies. The information gathered might help veterinarians understand the complexities of this endemic poultry disease and use the information in establishing company-specific ARV prevention programs.

The authors gratefully acknowledge California Animal Health & Food Safety (CAHFS) laboratory and the UC Davis Poultry Medicine Lab members, Sofia Egaña-Labrin, Ana da Silva, and Laura Flores for their technical assistance in this project.

ARV =

avian reovirus;

CAHFS =

California Animal Health and Food Safety laboratory;

CEL =

chicken embryo liver;

CPE =

cytopathic effects;

GC =

genotypic cluster;

RT-qPCR =

real-time reverse-transcriptase polymerase chain reaction;

VTM =

viral transport media