Determination of a Suitable Avian Cell Line for the Quantification of Avian Reovirus via Plaque Assays Open Access

The avian reovirus (ARV, species Orthoreovirus avis, genus Orthoreovirus, and family Spinareoviridae) (1) is a ubiquitous double-stranded RNA virus that can infect domestic and wild birds, including chickens and turkeys used for poultry products (2). Despite biocontainment controls and vaccination with live-attenuated strains of ARV (e.g., strains S1133 and 1733), ARV infections resulting from novel pathogenic field isolates have increased in the past 20 yr, becoming a significant economic burden for poultry producers (3,4). Viral quantification is a fundamental component of virology research. In the current literature, ARV viral quantity is usually reported as a 50% tissue culture infectious dose (TCID₅₀) (2,5,6,7,8). This end point dilution assay measures the dilution at which 50% of inoculated wells show signs of viral infection (9,10). Although TCID₅₀ assays are useful when an infectious dose is required, TCID₅₀ assays and their variations do not provide specific information about the number of viable virions present in a sample, and TCID₅₀ values can differ from the values provided by plaque assays by up to 10-fold (11). This information can be critical when assessing more in-depth characteristics of viral pathogenesis such as tissue tropism, host viral burden, particle-to-plaque-forming unit (PFU) ratio, and replication efficiency and kinetics. Plaque assays are better suited for this and have been used since the 1970s to isolate and quantitate different strains of ARV (12,13). The first attempts used chicken embryonic liver and kidney cells (13). In 2009, Tran et al. (14) established a protocol for ARV quantification using plaque assays in quail myoblast clone 5 (QM5) cells, and an adaptation of this protocol has been used for ARV plaque purification in leghorn male hepatocellular carcinoma (LMH) cells (15,16). More recently, African green monkey kidney (Vero) cells have been used in plaques assays for the titration of ARV (17,18). We compare the suitability of LMH and QM5 cells for the titration of ARV via plaque assays because of the advantages of using an immortalized cell line. LMH and QM5 cells differ in growth and morphologic characteristics, which affect suitability for plaque assays.

Chicken hepatocellular carcinoma LMH cells (19) were the kind gift of Holly Sellers at the Poultry Diagnostic and Research Center, University of Georgia, (Athens, GA). Quail fibrosarcoma clone QM5 cells (20,21) were the generous gift of Kevin Coombs at the University of Manitoba (Canada). LMH and QM5 cells were maintained in T175 flasks (no. 353112; Corning, Glendale, AZ) with Dulbecco minimum essential media (DMEM; no. 11995-065; Thermo Fisher Scientific, Waltham, MA) supplemented with 0.1% gentamycin (no. 15710064; Gibco, Waltham, MA) and 8% fetal bovine serum (FBS; no. S1400; Biowest, Bradenton, FL) at 38 C, with 5% CO₂.

ARV-141045 (OR612103) and ARV-127720 (OR6122112) were also acquired from the Sellers’ laboratory collection at the University of Georgia. Prior to our acquisition, both field isolates were isolated according to a standardized protocol for the Sellers’ laboratory (3). ARV-S1133 Fort Dodge strain was acquired from the laboratory of Ruediger Hauck at Auburn University (Auburn, AL). Viruses were expanded on LMH cells by adding 20-μl virus cell lysate onto 95% confluent monolayer of LMH cells in a T175 flask (no. 353112; Corning). The ARVs were allowed to replicate for 5 days prior to collection. Cell culture medium (supernatant fraction) and the cell debris fraction were collected and evenly aliquoted into 38.5-ml Open-Top Thinwall Ultra-Clear centrifuge tubes (no. 344058; Beckman Coulter, Indianapolis, IN) and ultracentrifuged in an XL-80K Ultracentrifuge (Beckman Coulter) in an SW60 rotor (Beckman Coulter) at 175,000 × g for 2 hr at 4 C. Cell culture supernatant was removed, and the cell pellet was resuspended in 7 ml of DMEM with 0.1% gentamicin and no FBS. The pellet was subsequently pulse sonicated using a Digital Sonifier 450 ultrasonic processor (Branson, Brookfield, CT) on ice at 30% amplitude, 10 sec on, 30 sec off (3×) before aliquoting into 500-μl cryovials (no. RK-44351-02; Corning) and stored at −80 C.

Plaque assays were performed following a Tran et al. protocol (14) using the LMH and QM5 cell lines. Cells were plated at 85%–90% confluency in 24-well CellBIND culture plates (no. 3337; Corning) and incubated at 38 C in a humidified atmosphere of 5% CO₂ overnight. Viral stock solutions were 10-fold serially diluted in DMEM from 10⁻¹ to 10⁻⁶. A total of 200 μl of each serial dilution was added to a confluent monolayer of LMH or QM5 cells plated the previous day. Cells were placed in a 38 C incubator for 1 hr of viral absorption. After that, viral inoculum was gently removed and replaced with an overlay consisting of 500 μl of 1:1 mix of 2% hydroxypropyl methylcellulose (no. 09965; Sigma, Burlington, MA) in H₂O solution and the cell maintenance media previously mentioned (DMEM, 0.1% gentamycin, and 8% FBS). The final concentration of the overlay was 1% hydroxypropyl methylcellulose, 4% FBS, 0.05% gentamycin, and 0.5× DMEM. Cells were returned to the incubator and monitored for the development of cytopathic effects (CPE) and plaque formation for the next 5 days. Fresh DMEM-methylcellulose mixture was added to cells as needed. After the development of CPEs and maturation of plaques, the viral infection was halted with a solution of 1% crystal violet (no. C0775; Sigma), 10% formaldehyde (no. F8775; Sigma), and 5% ethanol (no. 64-17-5; Sigma) in H₂0 for 20 min. After a gentle rinse with H₂O, plates were allowed to air dry overnight prior to analysis of CPE and counting of plaques. Each plaque assay was repeated three times. Standard deviation was calculated in Microsoft® Excel® for Microsoft 365 Microsoft Office Online, version 2412 (22) (STDEV function) and used to determine the variation among technical replicates of each biological replicate.

We attempted to quantify each virus using standard plaque assays in LMH and QM5 cells to determine which cell line would produce well-formed countable plaques to use to calculate infectious titers by counting PFU. We used ARV-S1133 Fort Dodge, a highly virulent nonattenuated isolate that is ancestral to the attenuated S1133 strain vaccine and two field isolates, ARV-141045 and ARV-127720, from very antigenically distant σC³-genoclusters (GC1 and GC7, respectively) on the basis of the genocluster classification of Kant (23), updated by Sellers (3). The field strains were isolated from birds exhibiting tenosynovitis (H. Sellers, pers. comm.), but the virulence and ability to form plaques in cell culture had not been previously assessed.

Overall, when infected with ARV, LMH cells exhibited slower growth, rounding, clustering, and an unhealthy appearance, especially at higher ARV concentrations, but did not produce well-delineated plaques even when highly diluted (10⁻⁵; Fig. 1A). When inoculated onto QM5 cells, these ARV isolates showed general CPEs, such as rounding, clustering, and reduction of stellate projections at higher concentrations. However, when the virus was appropriately diluted, QM5 cells exhibited localized, discrete CPEs that resulted in the formation of well-delineated plaques (Fig. 1A).

Extreme CPEs and cell death were observed at the 10⁻¹ through 10⁻³ dilutions for all ARV isolates in both cell types (Fig. 1A). Starting at 10⁻⁴, plaques became visible on QM5 cells but were too numerous to count for all ARVs (Fig. 1A,B). ARV-141045 produced CPEs quickly, developing punctate, mature, and countable plaques by 48 hr postinfection (hpi) in QM5 cells (Fig. 1B). ARV-S1133 Fort Dodge produced mature plaques by 72 hpi in QM5 cells (Fig. 1A) that were oblong in shape and readily coalesced during infection (Fig. 1B). ARV-127720 induced CPEs slowly but eventually produced plaques at 120 hpi (Fig. 1B). The mature plaques produced by ARV-127720 were less defined but were still countable, with an overall circular shape. All of the ARV isolates used in this study produced countable plaques in QM5 cells. A dilution of 10⁻⁵ was necessary to accurately count plaques on QM5 and determine the PFU count with little variation among replicates (Fig. 2B). None of the ARV dilutions used produced clear plaques on LMH that could be counted.

In this work, we compared and evaluated the effectiveness of plaque assays for ARV using LMH and QM5 cells to establish a more quantitative method for viral titration. In the current literature for ARVs, traditional methods such as TCID₅₀ and PCR-based assays have been used for assessing viral tissue tropism and replication kinetics and leveling vaccine and challenge doses across an experiment. However, techniques such as TCID₅₀ can vary extensively between individual labs and experiments, contributing to ambiguity and conflicting data. A plaque assay is the preferred method to determine infectious virion counts on the basis of the number of PFU observed with serial dilutions of virus infecting susceptible cells (14,24). In addition to providing actual virion counts, plaque assays also provide useful information, such as plaque size and morphology. In-depth plaque morphology is not readily assessable with TCID₅₀ assays that yield an infectious dose at which 50% of wells display signs of viral infection. As new and emerging ARV isolates arise, it is important to understand the differences that contribute to ARV variation. To investigate ARVs more closely and with precision, we focused on plaque assays, not TCID₅₀, to help reinforce a useful metric to assess strain and isolate-specific differences. We are aware of the limitations of a cell culture system for virus titration, particularly in relation to certain ARV field isolates showing greater affinity for embryonated eggs than for tissue culture (25,26). For these particular isolates, a tissue culture titration system might not accurately represent the total infectious titer present in a sample, potentially producing misleading results. Hence, users must exercise discretion. Nevertheless, we also recognize the benefits of plaque assays beyond the utility of using immortalized cell lines. For instance, Kim et al. (27) showed that plaque morphology could be correlated with respiratory syncytial virus subtype, while Mandary et al. (28) have shown that in enterovirus 71 (EVA-71), the morphology of plaques could be correlated with mutations in the viral VP1 protein. These and other studies underlie the wealth of information that can be provided by close observation of the morphology and characteristics of a viral plaque.

The titration of ARV via plaque assays does not seem to be a common practice in the study of ARVs at present, as reflected by the lack of scientific publications that refer to its usage. In 2009, Tran et al. (14) established a protocol for ARV quantification using plaque assays in QM5 cells. Before this publication, QM5s were mainly used for the phenotypic and genomic characterization of ARV (29,30,31,32), and a recent study has used this cell line in an ARV infectivity assay (33). The lack of scientific reports using any of the avian myogenic cell lines derived from the quail fibrosarcoma cell line QT6 was surprising to us, especially because, in our hands, this cell line was superior to LMHs for ARV plaque assays. Possible explanations could be the lack of availability of the QM5 cell line in any cell repository, the traditional predominance of LMH cells, and the use of specific-pathogen-free chickens (SPF) primary cell culture methods. Another possible reason for the reduced use of QM5, in particular, and plaque assays, in general, is the widespread perception that many ARV strains do not produce CPEs suitable for a successful plaque assay. An adaptation of the Tran protocol has been used for ARV plaque purification in LMH cells (15,16).

However, it is important to develop a standard assay when assessing viral characteristics, especially when new and emerging viral isolates are causing concern; such is the case for ARVs. Remaining relevant to the host should always be considered, and SPF, such as chicken embryonic liver cells (CeLic), fulfill this basic requirement for ARV. However, CeLic require significant resources and can be costly and time consuming to produce. The alternative is immortalized cell lines that can be maintained with less resources and require instruments commonly found in virology-focused research labs. They can be maintained longer than SPF primary cells, stored for future use, and allow for more consistency among batches. A literature search revealed ARV has been propagated in green monkey kidney (Vero cells) (34,35,36), baby hamster kidney cells (BHK-21) (35), Crandall feline kidney cells (36), rabbit kidney cells (36), GBK cells (36), QT35 cells (37), as well has primary CeLic (36), and others (38). Collectively, the variation in cell types suggests that ARVs are versatile and can adapt to different immortalized cell lines. However, the fact that there are multiple reports of successful ARV propagation in different cell lines does not mean that all isolates would replicate in all cell lines equally. Hence, choosing the most appropriate cell line is important to ensure consistency and accuracy of results (39).

Cell culture–related variables such as type of host cell, type of culture media, ionic strength or pH might affect the apparent concentration of virus (40). In our study, we used a chicken-derived (LMH) and a quail-derived (QM5) cell line maintained with the same culture media under the same conditions, and clear plaques were only observed in the quail cells. We recently reported that ARV shows a high level of host-species specificity among the main production bird species (41). Thus, the observed differences in ARV growth could be attributed to the different host origin of the cells. The fact that LMH (hepatocellular carcinoma cells) and QM5 (myoblasts) are two distinct cell types might also directly affect viral growth and syncytia formation. As previously reported by others (19), our LMH cells grew in a mosaic cobblestone-like pattern with multiple pseudopodia and lacked strong contact inhibition. We observed that when infected with ARV, LMH cells develop syncytia, commonly associated with ARV infection (42,43,44), which can detach from the monolayer and float in suspension. However, we also observed clusters of cells that could easily be confused with syncytia in uninfected cultures grown to the point of post confluency, which complicated the ability to properly assess viral-induced CPEs (Supplemental Fig. S1A–H). In contrast, uninfected QM5 cells showed their described (14,20) morphology: flat and spindle shaped with occasional multinucleation exhibiting contact inhibition. We observed that they grew with occasionally round, cell-free patches, requiring a heavy seeding density to form a confluent monolayer and that this pattern of growth must be maintained by avoiding serum depletion to prevent differentiation, as previously described (20). In addition, our QM5 cells adhered more strongly to the cell culture substrate and did not exhibit a tendency to develop cell clusters, reminiscent of syncytia in uninfected culture, as we observed in LMH cells. Furthermore, throughout our experiments, QM5 cells were repeatedly able to show definite CPE and distinct mature ARV plaques independently of the strain used to infect them. This was not the case for LMH cells, which consistently failed to show clear, distinct, easily countable plaques, a necessity for a viral plaque assay. We also observed that LMH cells are only partially adherent, and normal, healthy cultures of LMH had many floating cells. Overall, in our hands, clear countable plaques were only achieved in the QM5 cell line. LMH cells proved to be more useful for viral propagation than for plaque assays due the rapid growth of all ARVs tested in this study.

Supplemental data associated with this article can be found at https://doi.org/10.1637/aviandiseases-D-24-00100.s1.

We thank Holly Sellers, Kevin Coombs, Ruediger Hauck, and their corresponding academic institutions for facilitating the cell lines and ARV strains used in this study. Telvin L. Harrell was responsible for the conceptualization, data curation, formal analysis, investigation, methodology, project administration, validation, visualization, and writing of the original draft. Sonsiray Alvarez-Narvaez handled the investigation, methodology, and writing review and editing. Steven J. Conrad contributed to the conceptualization, investigation, funding acquisition, resources, supervision, and review and editing of the manuscript.

ARV =

avian reovirus;

CeLic =

chicken embryonic liver cells;

CPE =

cytopathic effect;

DMEM =

Dulbecco minimum essential media;

FBS =

fetal bovine serum;

hpi =

hours postinfection;

LMH =

leghorn male hepatocellular carcinoma;

PFU =

plaque-forming units;

QM5 =

quail myoblast clone 5;

SPF =

specific-pathogen-free chickens;

TCID₅₀ =

50% tissue culture infectious dose