Diagnostic investigation of avian reovirus field variants circulating in broiler chickens in Pennsylvania of United States between 2017 and 2022

ABSTRACT

Avian reovirus (ARV) has been continuously affecting the poultry industry in Pennsylvania (PA) in recent years. This report provides our diagnostic investigation on monitoring ARV field variants from broiler chickens in Pennsylvania. Genomic characterization findings of 72 ARV field isolates obtained from broiler cases during the last 6 years indicated that six distinct cluster variant strains (genotype I-VI), which were genetically diverse and distant from the vaccine and vaccine-related field strains, continuously circulated in PA poultry. Most of the variants clustered within genotype V (24/72, 33.3%), followed by genotype II (16/72, 22.2%), genotype IV (13/72, 18.1%), genotype III (13/72, 18.1%), genotype VI (05/72, 6.94%), and genotype I (1/72, 1.38%). The amino acid identity between 72 field variants and the vaccine strains (1133, 1733, 2408, 2177) varied from 45.3% to 99.7%, while the difference in amino acid counts ranged from 1–164. Among the field variants, the amino acid identity and count difference ranged from 43.3% to 100% and 0 to 170, respectively. Variants within genotype V had maximum amino acid identity (94.7–100%), whereas none of the variants within genotypes II and VI were alike. These findings indicate the continuing occurrence of multiple ARV genotypes in the environment.

KEYWORDS:

• Avian reovirus
• autogenous vaccine
• amino acid identity
• multiple genotypes
• field variants
• genomic charcterization

Introduction

Avian reoviruses (ARVs) infect many avian species worldwide with varying degrees of pathogenicity. Infection in commercial poultry results in various clinical conditions, including tenosynovitis, arthritis, myocarditis, pericarditis, perihepatitis, malabsorption, and runting-stunting syndrome. Clinically, tenosynovitis/arthritis is characterized by swelling of the hock joint, footpad, and digital flexor tendon and an inability to walk or lameness. Although birds of all ages are generally prone to avian reovirus infection (Van der Heide, 2000), susceptibility tends to decrease as chickens age (Rosenberger & Olson, 1991). Consequently, the disease outcome is contingent upon factors such as the host’s immune status, route of exposure, pathogen type, and transmission route, including horizontal or vertical transmission. In endemic regions, therefore, a key aspect of prevention strategies involves managing the field challenge to delay infection until later in the chicken’s life, primarily relying on good management practices such as maintaining proper vaccination and biosecurity measures, and ensuring optimal environmental conditions for the birds (Elkadmiri et al., 2023). Around 20–40% morbidity and up to 10% mortality are commonly observed in ARV-affected broiler and layer chickens, but higher morbidity and mortality caused by the newly emerged ARV variants have been reported in recent years (Lu et al., 2015, Markis, 2022; Sellers, 2022). Live-modified and inactivated vaccine strains (1133, 1733, 2408, and 2177) are used to control the ARV infection in commercial poultry. However, owing to the emergence of novel or genetically diverged variants in the recent ARV outbreaks, several research studies concluded that these vaccine strains are heterologous to the circulating variants and do not generate sufficient neutralization titre in the vaccinates to protect from the challenge (Lublin et al., 2011; Troxler et al., 2013; Lu et al., 2015; Tang & Lu, 2016; Ayalew et al., 2017).

ARVs are members of the order Reovirales, family Spinareoviridae, and the genus Orthoreovirus (Matthijnssens et al., 2022). They are non-enveloped viruses and possess double-stranded segmented ribonucleic acid with an inherent potential for extreme variability (Liu et al., 2003; Banyai et al., 2011). The complete genome sequence is 23 kb long and is divided into 10 segments named large (L₁-L₃), medium (M₁-M₃), and small (S₁-S₄) based on the nucleotide length, and encodes four non-structural (µNS, σNS, P10, P17) and eight structural proteins (λA, λB, λC, µA, µBC, σA, σB, and σC) (Spandidos & Graham, 1976; Bodelon et al., 2001). Most genome segments are monocistronic, while the S1 segment is tricistronic and encodes for two non-structural (P10, P17) and one structural protein Sigma C (σC). Though several structural proteins contribute to the genetic heterogeneity and protection from ARV infection, the S1 segment, particularly the σC protein, holds significant importance due to its role in genetic diversity, mediating interactions between virion particles and host cells, and eliciting immunity (Wickramasinghe et al., 1993; Kant et al., 2003). This may be the reason that the genome sequence corresponding to the σC gene is generally used as a distinct genetic marker to assess amino acid variability and classify the field variants into different genotype groups (Wickramasinghe et al., 1993; Kant et al., 2003; Goldenberg et al., 2010; Lu et al., 2015; Ayalew et al., 2017; Egana-Labrin et al., 2019).

During the last decade, the number of clinical cases of ARV infections has increased in poultry flocks in Pennsylvania (PA) and other states of the USA, indicating that either the vaccination programme is not strictly followed or the field variants are heterogenic to vaccine strains. Our previous studies reported the newly emerged ARV variants or novel strains in PA poultry between 2011 and 2014 (Lu et al., 2015). Continuous monitoring and diagnostic investigations are prerequisites to strategize effective disease control interventions in an endemic setting when ARVs constantly evolve, and there is much genetic heterogeneity across different segments of the viral genome, particularly the σC protein. In the present report, we provide our update investigations for the isolation and σC gene-based characterization of ARV variants from clinical cases between 2017 and 2022, and comparisons of amino acid variability and divergence among field isolates and the reference vaccine strains. The research data indicate that multiple genotypic variants of ARVs are continuously circulating in PA poultry and the environment. Regardless of their relatedness to a given genetic cluster in the phylogeny tree, a significant amino acid variability exists among the isolates. Therefore, as an effective intervention in ARV control strategies, an autogenous vaccine should be developed from a variant isolated from a given flock at a given time. In this context, while the Sigma C gene is not the sole genetic marker for defining genetic variations among and within each genotype in the selection of autogenous vaccine candidates, it does offer valuable insights into antigenic shifts or the emergence of drift-based escape mutants. Hence, it is advisable to continuously monitor and screen the isolates for the Sigma C gene, and when required, employ a complete genome sequence-based analysis for the initial selection and update of the vaccine strain.

Materials and methods

ARV isolation and identification

Tendons were collected primarily from broiler necropsy cases showing joint-related clinical signs and lesions such as lameness, joint swelling, and tendon rupture. Other tissue specimens of proventriculus, liver, spleen, and intestine were also collected from cases exhibiting maldigestion, malabsorption, and stunted growth. The tissue specimens were each processed separately for virus isolation in LMH cell cultures (CRL2113, ATCC) as per routine diagnostic procedures at the Animal Diagnostic Laboratory, as described previously (Lu et al., 2015). Briefly, each tissue specimen was minced individually with a scalpel and scissor in a 20 ml sterile plastic container (VWR, Radnor, PA, USA) and diluted at 1:5 (w/v) with viral transport media (VTM). The tissue homogenate was placed in a Stomacher bag and homogenized in a stomacher blender (Model 80, Seward Laboratory Systems Inc., Islandia, NY, USA) for 2–3 min. The homogenized suspension was transferred to a sterile 15 ml falcon tube and centrifuged at 100 g for 10 min at 5°C. The supernatant was filtered through a 0.45 µm syringe filter and used for inoculation onto 70–90% confluent monolayer of LMH cells in cell culture flasks (T12.5 cm² or T25 cm²) or cell culture plates (6-well, 12-well or 24-well). The LMH monolayer cells were inoculated with the filtered supernatant (0.25 ml for T12.5 cm², 0.5 ml for T25 cm², 0.1 ml for 12-well plate), incubated at 37°C for 20–30 min for virus absorption, and then overlaid with 2% FBS of LMH cell maintenance media. The specimen-inoculated LMH cell cultures were incubated at 37°C supplemented with 5% CO₂ and examined daily for the appearance of typical giant or bloom-like cytopathic effect (CPE).

Genome extraction, RT–PCR, and sequencing

Viral RNA was extracted from the archived supernatant of each ARV isolate per the manufacturer’s instruction (RNeasy Mini Kit, QIAGEN, Valencia, CA, USA). A partial amplification of the genome corresponding to the σC gene (1088 bp) was carried out using the P1 (sense) and P4 (anti-sense) primer pair described previously (Kant et al., 2003). The RT–PCR master mix (One-Step RT–PCR Kit, QIAGEN) for an individual sample consisted of a final concentration of 5X buffer (1X), 10 mM dNTPs (400 µM), forward and reverse primers (0.4 µM each), RNAse inhibitor (0.08U) and Enzyme Mix at kit recommended volume (1.5 µl/sample). Nuclease-free water (NFW) was added to adjust the reaction volume to 40 µl. The extracted RNA was diluted (1:2) in NFW, pre-desaturated at 95°C for 5 min, and chilled immediately. A 10 µl of denatured RNA was added to 40 µl of the master mix. Thermocycling comprised one cycle of 50°C for 30 min, 95°C for 15 min, followed by 38 cycles of 94°C for 30 s, 50°C for 30 s, and 72°C for 90 s. The final extension was conducted at 72°C for 5 min. The amplified products were gel-electrophoresed (1.0%) and visualized under a UV-transilluminator. Once confirmed, the amplified PCR fragments were purified using a commercially available kit (Gel/PCR DNA Fragment Extraction Kit, IBI Scientific, Dubuque, IA, USA) per the manufacturer’s instructions. The Sanger sequencing (3730XL DNA Analyzer, Applied Biosystems, Grand Island, NY, USA) was performed in both directions with the P1 and P4 primers (1 µM) at the Genomic Core Facility, Pennsylvania State University, USA.

Bioinformatic analysis

The nucleotide sequence received for the forward and reverse primer was first assembled with Geneious prime version 2022.0.2 (accessible at https://www.geneious.com) using the reference gene (AAK18188.1). Ambiguous nucleotides were removed from each end of the sequence, and only a quality sequence (approximately 900 bp) was used for submission to the NCBI database (GenBank) and subsequent analysis. The 72 ARV isolates from broiler cases in this study are presented in supplementary Table S1. A segment of 300 amino acids (1-300 aa) corresponding to nucleotide positions 525–1424 of the σC gene of each ARV variant and vaccine strain was used for alignment, percent amino acid identity and variability, and phylogenetic analysis. Reference sequences and representative sequences of each genotype cluster (n = 39) were retrieved from GenBank (http://www.ncbi.nlm.nih.gov) to determine the phylogenetic relationship with the field isolates. Together, 111 amino acid sequences were aligned with BioEdit version 7.0.5.3 (Hall, 1999) using the default setting in the ClustalW programme. The phylogenetic relationship of field isolates with those of reference and representative variants was assessed through the Maximum Likelihood method and Tamura-Nei model with 2000 bootstrap replicates in MEGAX software (Kumar et al., 2018) (Figure 1). Multiple amino acid alignment, phylogenetic relationship (Figure 2), and percent amino acid identity and variability (Supplementary data S1 and S2) were made for field isolates (n = 72) and the vaccine strains (n = 4) by a ClustalW alignment with BLOSUM cost matrix technique (Geneious prime version 2022.0.2).

Figure 1. Consensus phylogeny of the 72 ARV field variants based on the inherent sequence variability in the Sigma C (σC) gene. The clustering pattern indicates the genetic relationship between the σC gene protein of field strains and 39 reference and variant strains from around the globe. The ARV isolates clustered into six distinct genotyping groups. The tree was built by the Maximum Likelihood method and Tamura-Nei model with 2000 bootstrap replicates in MEGAX software.

Figure 2. A multiple amino acid sequence alignment and clustering pattern of the 72 field isolates clustered in Sigma C (σC) genotypes I-VI and vaccine strains in genotype I. A ClustalW alignment drew the alignment with the BLOSUM cost matrix technique. Amino acid differences from the consensus sequence are highlighted.

Sequence submission in the GenBank database

The σC gene sequence of 72 field variants of ARV was deposited in the GenBank database through online Bankit submission and is available under accession numbers OK160042-OK160049, OM837662-OM837672, and ON782394-ON782456 (supplementary Table S1).

Results

Clinicopathological examination and virus confirmation

A diverse range of clinical symptoms and pathological findings were observed in this study encompassing 72 ARV cases from 2017 to 2022. Birds of different ages (7–75 days) exhibited varying susceptibility, with younger birds being more vulnerable. Specifically, most clinical cases were more prevalent in birds aged ≤ 21 days (n = 44), followed by those aged 22–40 days (n = 20) and ≥ 41 days (n = 5). Age-related details of three flocks were not available. The broiler case flocks predominantly exhibited lameness, reluctance to move, splayed legs, femoral head necrosis, tenosynovitis, arthritis, and inflammation of tendons and surrounding skeletal muscles, increased synovial fluid, enlarged proventriculus, uneven and stunted growth, and intestinal inflammation. Noteworthy clinical signs included retained yolk sac, dehydration, femoral head necrosis, and tendon abnormalities. Of the 72 ARV isolates, 42 were prepared from tendons and leg joint tissues, and 30 from proventriculus, liver, spleen, and intestine specimens. A brief history and other relevant details are provided in Supplementary Table S1. Each of the 72 case specimens produced giant or bloom-like typical CPE characterized for ARV infection in the specimen-inoculated LMH monolayer cell cultures. The appearance of CPE varied from one-cell passage to 3∼4 serial passages. The occurrence of CPE was observed from as early as 24 h post-inoculation to 3∼5 days post-inoculation.

ARV Sigma C (σC) gene sequencing, phylogenetic, and amino acid analysis of field variant and vaccine strains

The 72 ARV isolates were successfully processed for Sigma C (σC) gene amplified at 1088 bp with the P1 and P4 primer pair (Kant et al., 2003). The amplified products of the ARV isolates were sequenced in both directions, assembled, and translated into amino acids to compare the genotypic properties with those ARV strains reported previously and representing different genotypes. Phylogenetically, the 72 ARV field isolates were grouped into six genotypes: I, II, III, IV, V, and VI (Figure 1). One-third (24/72, 33.3%) of the isolates clustered within genotype V, followed by genotype II (16/72, 22.2%), genotype IV (13/72, 18.1%), genotype III (13/72, 18.1%), genotype VI (05/72, 6.94%), and genotype I (1/72, 1.38%). Year-wise distributions with different σC genotypes of the 72 ARV isolates are presented in Table 1.

Table 1. Occurrence frequency in Sigma C (σC) genotypes of the 72 ARV field variants isolated between 2017 and 2022.

σC genotype	Detection of occurrence in years						Total (%)
	2017	2018	2019	2020	2021	2022
I	1	0	0	0	0	0	1 (1.38)
II	2	1	2	7	4	0	16 (22.2)
III	1	5	3	0	3	1	13 (18.1)
IV	2	2	1	2	4	2	13 (18.1)
V	1	18	2	2	1	0	24 (33.3)
VI	0	0	1	1	1	2	5 (6.94)
Total (%)	7 (9.72)	26 (36.1)	9 (12.5)	12 (16.7)	13 (18.1)	5 (6.94)	72 (100)

Comparative analysis for amino acid identity and count difference included 300 amino acids of total length (326 aa) of σC protein starting from positions 1–300. The amino acid identities between the vaccine strains (1133, 1733, 2408, 2177) varied from 95.7% to 99.7%, with the difference in amino acid counts ranging from 1–13. The amino acid identity between the 72 ARV field isolates and vaccine strains varied from 45.3% to 99.7%, while amino acid count differences ranged from 1–164. The ARV field isolates were also variable for their amino acid identity, ranging from 43.3% to a maximum homology of 100%, while amino acid count differed from 0–170. In-field isolates, amino acid variability was high compared to vaccine strains for both N-terminus (aa 41-120) and the C-terminus (aa 233-266). However, the N-terminus variability was more significant than the C-terminus among these ARV field isolates. Indeed, the C-terminus was conserved at several sites of the σC protein (Figure 2).

Comparative amino acid analysis of isolates within each genotype and the vaccine strains

Multi-paired alignment and clustering of ARV field isolates and vaccine strains were conducted. Figure 2 shows a clustering pattern among ARV isolates and vaccine strains. Only one ARV isolate was grouped within genotype I. The percent identity among ARV field isolates and vaccine strains varied from 96.7% to 99.0%, while the amino acid count differed from 1–10. Within genotype II, the ARV isolates broadly grouped into three sub-clusters and percent identity between field isolates varied from 65.0% to 99.7%. The percent identity among ARV isolates and vaccine strains varied from 50.0% to 56.7%, while the difference in amino acid count ranged from 129 to 150. ARV field isolates within genotype III are also grouped into three sub-clusters. The percent identity and amino acid count among ARV field isolates and vaccine strains varied from 47.7% to 52.7% and 141–157, respectively. Within ARV field isolates, percent identity varied from 59.0% to 100%, while amino acid count differed from 0–123. ARV field isolates within genotype-IV are grouped as two sub-clusters. Percent identity and difference in amino acid count within ARV field isolates were 83.0% to 100.0% and 0–51, respectively. However, among ARV field isolates and vaccine strains, percent identity varied from 45.3% to 49.7%, while amino acid count differences ranged from 151 to 164. ARV field isolates within genotype V had the greatest sequence homology. This is evidenced by the clustering pattern of ARV field isolates (Figure 2), a high rate of percent identity (94.7% to 100.0%), and a smaller number of different amino acids (0–16). Moreover, the ARV field isolates within genotype V were much more heterogenic when compared with the vaccine strains, where percent identity varied from 45.7% to 48.0%, while the amino acids difference ranged from 155–163. Genotype VI comprised of two sub-clusters. Among ARV field isolates and vaccine strains, percent identity varied from 45.7% to 50.0%, while amino acid count differed from 150–163. Percent identity and difference in amino acid count within ARV field isolates varied from 81.0% to 94.0% and 18–57, respectively (Supplementary data S1 and S2).

Discussion

Over the past 10 years, a gradual increase in the number of clinical cases of ARV infections has been noticed in PA and other states of the United States, resulting in significant economic losses to the poultry industry, especially in broiler chickens. It is, therefore, essential to routinely monitor and characterize the prevalent variants of ARVs in a disease-endemic setting to facilitate disease control intervention. This is particularly important for ARVs because of their segmented RNA genomes and inherent potential for variations (Kant et al., 2003; Ayalew et al., 2017; Sellers, 2017). In this regard, studies on classification and genotypic relationships can provide insight into the origin, epidemiology, and ongoing evolution among emerging and re-emerging ARV variants circulating in poultry flocks.

The ARV field variants isolated from broiler chickens in the present report came from birds that had a history of lameness, tenosynovitis, and related complications similar to those reported in other studies (Goldenberg et al., 2010; Lu et al., 2015; Ayalew et al., 2017; Palomino-Tapia et al., 2018; Egana-Labrin et al., 2019; De Carli et al., 2020). Therefore, the isolation of ARVs was primarily focused on tendon tissues. While the virus was successfully retrieved from tendon and joint tissues in all cases of this study, virus isolation success from other tissue specimens was variable. Indeed, there was a varying tissue tropism of ARV observed for the liver, proventriculus, spleen, and intestine in this study. This concurs with the observation made by Egana-Labrin et al. (2019), who reported tendons as the preferred tissue specimen for virus isolation over the heart, intestine, liver, and pancreas. The specimen-inoculated LMH monolayer showed typical cell fusion characterized by syncytia formation. Such a fusogenic property of ARV is ascribed to the expression of the non-structural protein (p10) in the first open reading frame of the tricistronic S1 gene (Benavente et al., 2007; Liu et al., 2008).

The σC protein-based genotyping system was used to compare ARV field isolates with reference and representative strains of each genotype. The σC protein is the most variable cell attachment protein and is considered a selective genetic marker to characterize and classify field-prevalent variants into six genotypes (Liu et al., 2003; Lu et al., 2015; Ayalew et al., 2017). The phylogenetic analysis and multiple alignment analysis for ARV in this study were based upon deduced amino acid sequence (300 aa, 1-300) so that synonymous and non-synonymous nucleotide substitution could also be accounted for. Throughout the period 2017–2022, the wide distribution of the six detected genotypes signifies the continuing presence of diverse genotypes in the poultry population. However, there were variations in the frequency of isolation for each genotype. Phylogeny analysis with the addition of reference and representative strains yielded a well-defined and distinct clustering pattern. The 69 ARV variants of PA origin were grouped into six genotypes. Of the three ARV variants originating from NY state, two grouped into genotype III while one clustered within genotype IV. Genotype V was detected more frequently, followed by genotype II, genotype IV, genotype III, genotype VI, and genotype I. Our previous study reported that genotype II was the most common variant in PA poultry from 2011–2014, followed by genotype V, genotype I, genotype VI, genotype III, and genotype IV (Lu et al., 2015). Furthermore, although establishing a direct correlation between the identified genotype in avian cases and specific clinical symptoms remains challenging and may not be conclusive, certain pathological findings and clinical features consistently demonstrated associations with particular genotypes in the present study. Notably, genotype III displayed a frequent association with unilateral femoral head necrosis, mild tibial rotation, and hydropericardium. In contrast, genotype VI was consistently linked to lameness, runting, and stunting. Cases associated with Genotype II commonly manifested tenosynovitis, femoral head necrosis, and gastrointestinal abnormalities. Based on the findings of our previous study (Lu et al., 2015) and the present investigation, it is expected that variations in the distribution of ARV genotypes will continue. Therefore, continuing evaluation of control measures is crucial for better understanding of the genetic and pathotypic shifts of ARV throughout the genome, influenced by natural selection, mutations, recombination, and reassortment.

Generally, breeder flocks are vaccinated with specific ARV strains to eliminate or limit vertical transmission and protect progeny from the horizontal spread of field variants. Unfortunately, the vaccination history of the specific flock under investigation was not accessible to us, hindering the correlation between the identified genotype and the introduced vaccine strain within that particular flock. It is imperative to recognize this aspect as a limitation inherent in the present study; however, subsequent investigations should duly consider and address this factor when interpreting results. Nevertheless, necropsy findings and clinical history (Supplementary Table S1) suggest that the ARV field isolates represented field variant strains currently circulating in these poultry flocks, in which 71 of 72 field isolates clustered within genotype II-VI and were distinct from commercially-available vaccine strains. There was some temporal clustering of isolates over the study period. Maximum amino acid identity among vaccine strains and the 71 field variants was found to be 56.7%, suggesting that circulating variants are genetically heterogenic and have evolved and diverged significantly compared to vaccine strains.

In conclusion, multiple ARV genotypes are continuously circulating in commercial poultry flocks. The prevalent variants can be virulent and genetically heterogenic. A significant amino acid variability exists between the genotypes and within each genotype. Therefore, we need to continue our efforts to screen and monitor ARV infections in the field.

Author contributions

Conceived and designed research experiments: HL. Performed research experiments: MZS, HY, HL. Performed ARV case diagnosis: MEL, PAD, EAW. Analysed research data: MZS, HL. Wrote the manuscript: MZS, HL. Reviewed/edited the manuscript: HL, MEL, PAD, EAW.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

The US Egg Industry Center at Iowa State University, the Pennsylvania Soybean Board, and the Pennsylvania Center for Poultry and Livestock Excellence supported this research.