Polymorphisms in the Most Oncolytic Reovirus Strain Confer Enhanced Cell Attachment, Transcription, and Single-Step Replication Kinetics

Reovirus serotype 3 Dearing (T3D) is in clinical trials for cancer therapy. Recently, it was discovered that highly related laboratory strains of T3D exhibit large differences in their abilities to replicate in cancer cells in vitro, which correlates with oncolytic activity in a murine model of melanoma. The current study reveals two mechanisms for the enhanced efficiency of T3D^PL in cancer cells. Due to polymorphisms in two viral genes, within the first round of reovirus infection, T3D^PL binds to cells more efficiency and more rapidly produces viral RNAs; this increased rate of infection relative to that of the less oncolytic strains gives T3D^PL a strong inherent advantage that culminates in higher virus production, more cell death, and higher virus spread.

ABSTRACT

Reovirus serotype 3 Dearing (T3D) replicates preferentially in transformed cells and is in clinical trials as a cancer therapy. Laboratory strains of T3D, however, exhibit differences in plaque size on cancer cells and differences in oncolytic activity in vivo. This study aimed to determine why the most oncolytic T3D reovirus lab strain, the Patrick Lee laboratory strain (T3D^PL), replicates more efficiently in cancer cells than other commonly used laboratory strains, the Kevin Coombs laboratory strain (T3D^KC) and Terence Dermody laboratory (T3D^TD) strain. In single-step growth curves, T3D^PL titers increased at higher rates and produced ∼9-fold higher burst size. Furthermore, the number of reovirus antigen-positive cells increased more rapidly for T3D^PL than for T3D^TD. In conclusion, the most oncolytic T3D^PL possesses replication advantages in a single round of infection. Two specific mechanisms for enhanced infection by T3D^PL were identified. First, T3D^PL exhibited higher cell attachment, which was attributed to a higher proportion of virus particles with insufficient (≤3) σ1 cell attachment proteins. Second, T3D^PL transcribed RNA at rates superior to those of the less oncolytic T3D strains, which is attributed to polymorphisms in M1-encoding μ2 protein, as confirmed in an in vitro transcription assay, and which thus demonstrates that T3D^PL has an inherent transcription advantage that is cell type independent. Accordingly, T3D^PL established rapid onset of viral RNA and protein synthesis, leading to more rapid kinetics of progeny virus production, larger virus burst size, and higher levels of cell death. Together, these results emphasize the importance of paying close attention to genomic divergence between virus laboratory strains and, mechanistically, reveal the importance of the rapid onset of infection for reovirus oncolysis.

IMPORTANCE Reovirus serotype 3 Dearing (T3D) is in clinical trials for cancer therapy. Recently, it was discovered that highly related laboratory strains of T3D exhibit large differences in their abilities to replicate in cancer cells in vitro, which correlates with oncolytic activity in a murine model of melanoma. The current study reveals two mechanisms for the enhanced efficiency of T3D^PL in cancer cells. Due to polymorphisms in two viral genes, within the first round of reovirus infection, T3D^PL binds to cells more efficiency and more rapidly produces viral RNAs; this increased rate of infection relative to that of the less oncolytic strains gives T3D^PL a strong inherent advantage that culminates in higher virus production, more cell death, and higher virus spread.

INTRODUCTION

Mammalian orthoreovirus (MRV; reovirus) has served as an important model system since being isolated by Albert Sabin in 1959. The three serotypes of reovirus are each represented by a prototypic strain: type 1 Lang (T1L), type 2 Jones (T2J), and type 3 Dearing (T3D). Although naturally nonpathogenic in humans, reovirus can cause myocarditis and central nervous system infection in immunocompromised mice in a serotype-specific manner (1–4), thus providing a safe model system to study the effects of virus and innate signaling in vivo. Numerous studies have also used reovirus serotype comparisons in vitro to understand fundamental aspects of viruses, such as how proteinaceous capsids assemble and disassemble. Most recently, interest in reovirus has grown owing to discovery that serotype 3 reovirus (T3D) possesses potent oncolytic properties (5). Specifically, T3D was found to replicate preferentially in tumors and cancer cells over nontransformed cells and healthy tissues. Accordingly, T3D is currently in phase III clinical trials for cancer therapy and is the subject of research worldwide in laboratories interested in virus oncolysis.

Even within the T3D prototypical serotype 3 reovirus, there is significant divergence between laboratory strains (6). Importantly, the laboratory strains of T3D exhibited distinct oncolytic potential in vivo, with the Patrick Lee laboratory strain (T3D^PL; identical at the amino acid level to the T3D^Reolysin strain being tested in clinical trials) being most effective at tumor regression. The study also found that T3D laboratory strains caused distinct plaque morphology, with large-plaque formation correlating with in vivo oncolytic potency. These findings not only raised the importance of using shared virus stocks for oncolytic studies worldwide but also afforded a new opportunity to understand features of reovirus that contribute to good oncolysis. Since differences in plaque size can arise from differential efficiency of many stages of virus infection, replication, and/or dissemination, we sought to understand which stages of virus replication contribute to the large-plaque phenotype of the most oncolytic T3D strain.

The reovirus genome is packaged within the viral core and is composed of 10 double-stranaded RNA (dsRNA) segments: 4 small (S1, S2, S3, and S4), 3 medium (M1, M2, and M3), and 3 large (L1, L2, and L3) (7). The 10 genome segments encode 12 proteins: 8 structural (σ1, σ2, σ3, μ1, μ2, λ1, λ2, and λ3) and 4 nonstructural (σ1s, σNS, μNS, and μNSC). Note that the naming of reovirus dsRNA segments and proteins was based on electrophoretic molecular weights and therefore can get confusing; for example, L1 (slowest-migrating L genome segment) encodes λ3 (fastest-migrating large reovirus protein). Using reassortment analysis and reverse genetics, the polymorphisms associated with the large-plaque phenotype of T3D^PL relative to that of the less oncolytic but commonly used Terence Dermody laboratory strain (T3D^TD) were mapped to the S4, M1, and L3 genome segments. The σ3 protein encoded by the S4 gene comprises the outermost virus capsid and plays a major role in maintaining virus stability in the environment (8–10). Aside from its structural role, σ3 binds dsRNA during virus infection functions to overcome PKR signaling and to maintain viral protein translation (11, 12). The μ2 encoded by the M1 gene also provides at least two functions during reovirus replication. Within reovirus core particles, μ2 is a λ3 polymerase cofactor, supplying nucleoside triphosphatase (NTPase) activity and supporting temperature-dependent core transcriptase activity. During reovirus replication, μ2 is also a determinant of virus factory morphology, bridging tubulin to μNS, which subsequently recruits other viral proteins and RNAs (13–18). The λ1 protein encoded by the L3 gene is a major component of the viral core inner capsid and has dsRNA binding, NTPase, and RNA helicase activity (18–21).

In this study, a comprehensive analysis of T3D laboratory strains revealed that the most oncolytic strain (T3D^PL) has much faster kinetics of infection than the less oncolytic strains (T3D^TD and the Kevin Coombs laboratory strain [T3D^KC]) in a single round of replication. The more rapid infection leads to more viral RNA, protein, and progeny production and, ultimately, to faster cell death and reovirus dissemination. These findings point to the importance of inherent differences in virus replication proficiency and intracellular permissively bestowed by single amino acid polymorphisms. Two specific mechanisms for rapid infection were identified: first, T3D^PL had superior binding to tumor cells, and, second, T3D^PL core particles had faster kinetics of transcription. Genetic assortment showed that the polymorphic M1-encoded μ2 was responsible for increased core transcription levels of T3D^PL, linking μ2 as an important determinant of transcription rate and timely infection.

(This article was submitted to an online preprint archive [53].)

RESULTS

T3D^PL replicates more rapidly and to higher burst size in a single round of infection.

The T3D^PL reovirus strain exhibited oncolytic activities superior to those of the T3D^TD strain in a murine melanoma model and also caused larger plaques than a variety of laboratory T3D strains including T3D^TD (see the accompanying manuscript [6]). The larger plaques produced by T3D^PL could stem from more rapid rates of virus replication, increased burst size, enhanced cell lysis and virus release, or increased cell-to-cell spread. To determine if T3D^PL replicates more rapidly and/or to higher burst size than T3D^TD in a single round of infection, single-step virus growth analyses were performed at a multiplicity of infection (MOI) of 3 (Fig. 1A, left). L929 cells were exposed to T3D^PL and T3D^TD at 4°C for 1 h to synchronize infection, washed to remove unbound viruses, and then incubated at 37°C for various time periods. Reovirus titers at onset of infection were similar between T3D^PL and T3D^TD, indicating equal input infectious virions as calculated. Following entry, reovirions uncoat to cores and become noninfectious, leading to reduced titers at 3 and 6 h postinfection (hpi) for both T3D^PL and T3D^TD. Between 8 and 12 hpi, ∼1,000-fold new infectious progeny was produced per input of T3D^PL. In contrast, titers increased by only ∼50-fold for T3D^TD. T3D^PL reached maximum saturation titers at 18 hpi, while T3D^TD titers were saturated at 24 hpi. Furthermore, saturation titers for T3D^PL were ∼9-fold higher than those for T3D^TD. The one-step growth curves therefore indicated that new progeny production for T3D^PL occurs faster and produces higher virus burst size than that of T3D^TD. Importantly, these findings suggest that oncolytic potential correlates with proficiency of virus replication in a single round of replication.

FIG 1.

T3D^PL replicates more rapidly and to higher burst size in a single round of infection. (A) L929 cells were infected with T3D^PL or T3D^TD at an MOI of 3 (left) or 0.01 (right) and incubated at 37°C. At each time point, total cell lysates were harvested, and virus titers were determined in duplicate on L929 cells. Data represents mean ± standard deviations (n = 3). (B) L929 cells were infected with T3D^PL or T3D^TD (MOI of 1) and incubated at 37°C. Polyclonal reovirus antibody was added directly to the well medium at 3 hpi (3 hpi anti-Reo Ab). Reovirus=infected cells were visualized by fluorescence microscopy using polyclonal anti-reovirus antibodies (green) and HOECHST staining of cell nuclei (blue). (C) H1299, ID8, and B16-F10 cancer cell lines were mock infected or infected with T3D^PL or T3D^TD at the indicated MOI (based on titers established on L929 cells) and incubated at 37°C for 12 or 24 h. Paraformaldehyde-fixed cells were subjected to immunofluorescence staining with reovirus-specific primary antibody and Alexa Fluor 488-conjugated secondary antibody and to flow cytometric analysis. Gates were set based on mock-infected cells.

When multistep virus growth curves were performed at an MOI of 0.01, the cumulative advantage of improved replication over several rounds of infection was striking: T3D^PL accumulated ∼4,000-fold higher virus titers than T3D^TD by 48 hpi (Fig. 1A, right).

As a complementary approach to monitor the rate of reovirus initial infection versus cell-cell spread, we used immunofluorescence staining of reovirus antigen-positive L929 cells (Fig. 1B). In order to decipher the extent of cell-cell spread versus initially infected cells, staining was compared in the absence or presence of reovirus neutralizing antibodies added at 3 hpi to prevent cell-cell spread. Cells became positive for T3D^PL staining more rapidly than for T3D^TD (e.g., compare viruses at 12 hpi) and equalized at ∼24 hpi. By 48 hpi, T3D^PL very effectively spread to neighboring cells (i.e., compare staining 48 hpi without antibody [Ab] to that with anti-reovirus Ab), while limited cell-cell spread was observed for T3D^TD. By 120 hpi, very few cells remained alive following T3D^PL infection, as indicated by a lack of Hoechst nuclear staining. In comparison, T3D^TD successfully underwent cell-cell spread by 120 hpi, but cells remained relatively intact. The 48-hpi time point represents only two rounds of reovirus replication yet demonstrates the strong compounding difference between T3D laboratory strains with respect to replication and spread on tumorigenic cells (i.e., in vitro oncolysis).

To determine if faster kinetics of establishing infection was an inherent property of T3D^PL, kinetics of infection were compared for T3D^PL and T3D^TD in different cancer cell lines. Human lung carcinoma cells (H1299), mouse ovarian carcinoma cells (ID8), and the murine melanoma cells previously used to compare oncolytic potency of T3D^PL and T3D^TD in vivo (B16-F10) were exposed to the two T3D strains at various MOIs. Note that these three cell lines show differential infectivities to reovirus relative to each other, and, hence, MOIs for a given cell line were chosen to give approximately 10%, 30%, and 60% infected cells. At 12 and 24 hpi, flow cytometric analysis was performed to quantify the proportion of reovirus protein-expressing cells (Fig. 1C). In all three cancer cell lines, T3D^TD infection was delayed, giving the same percentages of infected cells at 24 hpi as already achieved by T3D^PL at 12 hpi. Together, the data suggested that T3D^PL inherently replicates more rapidly and to higher burst size in a single round of infection.

T3D^PL shows minor enhancement in cell attachment.

Increased virus replication shown observed in Fig. 1 could result from higher efficiency of virus entry or postentry steps of virus replication and assembly. The first and commonly studied step of virus infection is attachment to cells. We therefore compared cell attachment efficiencies between CsCl gradient-purified T3D^PL, T3D^KC, and T3D^TD strains, with T3D^KC being a third laboratory strain of T3D that causes smaller plaques than T3D^PL (6). L929 cells were exposed to equivalent particle numbers of each laboratory strain at 4°C for 1 h to permit virus binding without entry and then washed extensively. Flow cytometric analysis was used to measure bound virions per cell; in this approach the mean fluorescence intensity has a linear relationship with the number of cell-bound virions, enabling standard curves to be generated with serial dilutions of T3D^PL and used to extrapolate relative numbers of bound particles for T3D^KC and T3D^TD strains (Fig. 2A). Reproducibly, T3D^KC and T3D^TD bound cells at 60% and 40% efficiencies relative to cell binding by T3D^PL (Fig. 2B).

FIG 2.

T3D^PL exhibits minor enhancement in cell attachment. (A) Using nonenzymatically detached L929 cells in suspension, T3D^PL dilutions were bound at 4°C, and following extensive washing to remove unbound reovirus, cell-bound reovirus was stained using polyclonal reovirus antibodies. Mean fluorescence intensity (MFI) was quantified using flow cytometry (BD FACSCanto). Black histogram, mock infection; dark orange, most concentrated; light orange, least concentrated. Linear regression analysis of MFI was determined from serial dilutions and corrected for mock-infected background. (B) Corrected for similar particle numbers using Coomassie staining, cell binding of T3D^KC and T3D^TD was performed as described for panel A (left) and standardized to T3D^PL using curve from panel A (right; n ≥ 5). One-way analysis of variance with Tukey’s multiple-comparison test was performed **, P < 0.0021; ****, P < 0.0001. (C) CsCl-purified reovirus preparations were separated using SDS-PAGE. Total protein was stained using Imperial Coomassie dye while specific reovirus proteins were identified using Western blot analysis with protein-specific antibodies. (D) Agarose gel separation and Imperial Coomassie dye staining of virions based on the number of σ1 trimers per virion. The table indicates percentage of virions with specific number of σ1 timers/virion calculated using densitometric band quantification obtained with an ImageQuantTL 1D gel analysis add-on (n ≥ 4). One-way analysis of variance with Tukey’s multiple-comparison test was performed ***, P < 0.0001; ****, P < 0.0001; ns, P > 0.05.

For reovirus, cell binding is mediated by the trimeric σ1 protein that extends from channels generated by pentameric λ2 proteins at vertices of reovirus particles. Previous studies showed that reovirus particles can have 0 to 12 σ1 trimers, and that 3 or more σ1 trimers per virion are necessary and sufficient for virion binding to cells (22, 23). We therefore queried the levels of σ1 on T3D laboratory strains. First, CsCl gradient-purified T3D^PL, T3D^KC, and T3D^TD virus particles were subjected to denaturing gel electrophoresis, and proteins were visualized by Imperial total protein staining (Fig. 2C, top) and Western blot analysis using reovirus protein-specific antibodies (Fig. 2C, bottom). The ratio of capsid proteins λ1/2, μ1C, σ2, and σ3 were similar between T3D^PL, T3D^KC, and T3D^TD. However, for equivalent capsid protein (σ3), T3D^KC and T3D^TD had lower average levels of σ1 than T3D^PL. Agarose gel separation of full virions was then applied to quantify the proportion of virions with fewer than three σ1 trimers, which would be incapable of cell binding (Fig. 2D). While 24% and 48% of T3D^KC and T3D^TD particles (respectively) had fewer than three σ1 trimers per virion, only 2% of T3D^PL particles had insufficient σ1 trimers for binding. The proportion of virions with ≥3 σ1 trimers per virion correlated closely to the percentage of cell bound particles shown in Fig. 2B. Together the analysis indicates that oncolytic potency of the most oncolytic laboratory strain T3D^PL correlates with enhanced attachment to tumor cells, which in turn can be explained by a higher proportion of virus particles with ≥3 σ1 trimers per virion.

Postentry steps of reovirus replication are strongly enhanced for T3D^PL.

Following attachment to cells, reovirus undergoes endocytosis and trafficking to lysosomes where cathepsins L and B mediate virus uncoating (24–28). Specifically, during reovirus uncoating, the outermost σ3 protein is degraded, and μ1C is cleaved to the membrane-destabilizing products δ, μ1N, and φ. The resulting intermediate subviral particle (ISVP) is capable of penetrating membranes and delivering the final reovirus core particle to the cytoplasm (29–34). Cleavage of σ3 and μ1C was monitored by Western blotting analysis at various time points postinfection. Complete cleavage of σ3 was observed by 3 hpi and coincided with initiation of μ1C to δ cleavage (Fig. 3A). The rate of μ1C cleavage to δ was quantified as a percentage of δ to the total of δ plus μ1C standardized to the level of β actin (housekeeping protein). All T3D strains had similar uncoating rates, and almost complete uncoating was attained by 5 hpi. De novo-synthesized μ1 (i.e., precursor to μ1C) and σ3 were observed at 5 hpi in T3D^PL but not in T3D^KC and T3D^TD.

FIG 3.

Postentry steps of reovirus replication are strongly enhanced for T3D^PL. (A) T3D^PL, T3D^KC, and T3D^TD (MOIs of 1 to 3) were bound to L929 cell monolayers at 4°C and following washing to remove unbound virions, cells were incubated at 37°C and cell lysates were collected at various time points postinfection. Total proteins were separated using SDS-PAGE, and Western blot analysis with specific antibodies was used to identify reovirus proteins and β actin (loading control) (left). Densitometric band quantification of μ1C and δ and percent virus uncoating (right) were calculated using the following formula; [δ/(μ1C + δ)] ×100 (n = 4 ± standard deviation; MOI of 1, n = 2; MOI of 3, n = 2). Linear regression analysis determined that the slopes are not significantly different. (B to D) L929 cells were exposed to T3D^PL, T3D^KC, and T3D^TD (MOI of 3) at 4°C for 1 h, washed, and incubated at 37°C. At indicated time points, following total RNA extraction and cDNA synthesis, reovirus S4 and M2 RNA expression levels relative to the level of the housekeeping gene GAPDH were quantified using RT-qPCR (n = 3 ± standard deviation; ***, P < 0.0001). Linear regression analysis determined that (i) the slope of T3D^PL differs from that of T3D^KC or T3D^TD and (ii) the slope of T3D^KC does not differ from that of T3D^TD. (C) Total proteins were separated using SDS-PAGE, and Western blot analysis with specific antibodies were used to identify reovirus proteins and β actin (loading control). (D) Adherent and detached cells were collected at 24 hpi and subjected to staining with 7-AAD and annexin V, followed by flow cytometric analysis. PE, phycoerythrin.

After removal of the outer capsid, reovirus cores that enter the cytoplasm become transcriptionally active. Positive-sense mRNAs are transcribed within reovirus cores and released into the cytoplasm for translation by host protein synthesis machinery (35–38). To assess overall accumulation of viral RNA, we used quantitative reverse transcription-PCR (RT-qPCR) for S4 and M2 as representative viral RNAs over the course of infection (Fig. 2B). Over 12 hpi, viral transcripts accumulated at a significantly higher rate for T3D^PL than T3D^KC/T3D^TD. Specifically, prior to viral RNA saturation at 8 hpi, T3D^PL exhibited 77- and 85-fold higher levels of S4 viral RNA than T3D^KC and T3D^TD, respectively (Fig. 2B, left). Differences in viral M2 RNA levels were even greater, with T3D^PL having 92- and 133-fold higher levels than T3D^KC and T3D^TD, respectively (Fig. 2B, right). As anticipated, viral protein levels closely reflected virus transcript levels; Western blot analysis with polyclonal anti-reovirus antibody showed substantially higher accumulation of T3D^PL viral proteins than of T3D^KC and T3D^TD proteins at every time point from 6 hpi onward (Fig. 2C). Not surprisingly given the large increase in kinetics and total level of viral RNA and proteins for T3D^PL, cell death measured by flow cytometry with annexin V (measures early apoptosis) or 7-amino-actinomycin D (7-AAD; measures cell death) was higher for T3D^PL than for T3D^TD, with, for example, values of ∼85% versus 19% at an MOI of 3 for these strains, respectively (Fig. 2D). The analysis indicates that kinetics of reovirus macromolecule synthesis are drastically faster for T3D^PL than for T3D^TD, which explains the accelerated replication and larger burst size for T3D^PL in the one-step growth curves shown in Fig. 1.

S4, M1, and L3 polymorphisms independently contribute to enhanced viral RNA and protein synthesis by T3D^PL.

Using reassortment and reverse genetics analysis, M1, S4, and L3 reovirus genes were previously associated with increased plaque size of T3D^PL (6). Specifically, monoreassortants that contained one of these genes independently from T3D^PL in an otherwise T3D^TD background (or vice-versa) caused plaque size that was intermediate between the two T3D strains. Only when all three gene segments were introduced together was plaque size restored to that of the parental virus phenotype. Our current analysis indicated that differences between T3D^PL and T3D^TD were already apparent in the first round of replication, so we determined if M1, S4, and L3 reovirus genes contributed to differential levels of viral RNA and protein between the T3D strains. RNA and protein analysis were chosen over later steps in virus replication in order to capture the earliest difference between T3D laboratory strains. Quantitative RT-PCR (Fig. 4A) and Western blot analysis (Fig. 4B) were repeated for single mono-reassortants at 12 hpi. M1, S4, and L3 genes from T3D^PL increased reovirus RNAs when placed in an otherwise TD^RG background relative to levels of T3D^TD. In the reciprocal mono-reassortment, M1, S4, and L3 genes from T3D^TD decreased reovirus RNAs in an otherwise PL background relative to the level for T3D^PL. In a T3D^TD background, it was evident that M1, S4, and L3 genes from T3D^PL could increase virus protein (μ1C) levels. Importantly, when introduced independently, each gene conferred only intermediate levels of RNA and protein relative to the levels of the parental strains, indicating that each gene confers only a portion of the replication advantage that T3D^PL exhibits over T3D^TD. Nevertheless, all three genes confer a replication advantage for the most oncolytic T3D^PL strain relative to replication of the less oncolytic T3D^TD strain in a single round of infection.

FIG 4.

S4, M1, and L3 polymorphisms independently contribute to enhanced viral RNA and protein synthesis by T3D^PL. (A) L929 cells infected with parental T3D^PL or T3D^TD and S4, M1 and L3 gene monoreassortant PL-RG or TD-RG viruses at an MOI of 3. At 12 hpi, S4 RNA relative to GAPDH was quantified using RT-qPCR (n ≥ 3). All values were normalized to the value for T3D^TD. Statistical significance was determined using one-way analysis of variance with Dunnett’s multiple-comparison test **, P < 0.01; ****, P < 0.0001. (B) At 12 hpi, Western blot analysis was conducted, and densitometric band quantification of μ1 and σ3, relative to that of nonspecific background band, was calculated and normalized to the value for T3D^TD (n ≥ 3).

M1-encoded μ2 confers accelerated core transcription.

Incoming reovirus cores generate RNAs and proteins that assemble into new progeny cores; these progeny cores then further amplify RNA and protein synthesis. The cyclical nature of reovirus RNA and protein amplifications makes all replication steps interdependent. Accordingly, the observation that virus transcripts accumulate at a higher rate for T3D^PL did not directly indicate whether T3D^PL cores transcribe more efficiently or whether subsequent steps are enhanced. To directly assess transcription activities of T3D laboratory strains, we therefore performed in vitro transcription assays (Fig. 5A). T3D^PL, T3D^KC, and T3D^TD particles were treated with chymotrypsin (CHT) to generate transcriptionally active cores, purified by high-speed centrifugation, and confirmed to have the characteristic loss of outer capsid proteins (σ1, σ3, and μ1) but maintenance of core inner capsid proteins (σ2, λ1, and λ2) (Fig. 5B). Cores were then subjected to transcription reactions for various time intervals (Fig. 5C), and levels of S4, M2, and L2 viral RNAs were measured by RT-qPCR (Fig. 5D). In vitro-synthesized mouse glyceraldehyde-3-phosphate dehydrogenase (GAPDH) RNA was used as a spike-in control for sample-to-sample processing variations. Negative-control reaction mixtures lacking recombinant ATP (rATP) produced no increases in RNA levels, confirming that RNA measured in the presence of all four recombinant nucleoside triphosphates (rNTPs) accurately reflected de novo core transcription. Furthermore, in vitro core transcription rates were linear, unlike the exponential rates observed during intracellular infection shown in Fig. 2B, reflecting the absence of secondary rounds of amplification that occur during reovirus infection but not in vitro. Importantly, the rates of S4, M2, and L2 viral RNA transcription were significantly higher for T3D^PL than for T3D^KC and T3D^TD. For example, S4 transcripts increased for T3D^PL at 16.8 ± 0.9 transcripts/min but at 3.6 ± 0.3 and 2.5 ± 0.3 transcripts/min for T3D^KC and T3D^TD, respectively. For M2, rates were 7.3 ± 0.5, 1.1 ± 0.2, and 0.9 ± 0.1 transcripts/min for T3D^PL, T3D^KC and T3D^TD, respectively. Therefore, T3D^PL has an inherent transcriptase advantage within the viral cores.

FIG 5.

M1-encoded μ2 confers accelerated core transcription. (A) Schematic of chymotrypsin reovirus digestion and viral core purification. (B) Aliquots at various stages during viral core generation and purification were separated by SDS-PAGE, and either total viral proteins were stained using Coomassie dye or σ1 proteins were identified using Western blot analysis. (C) Schematic of reovirus core transcription reactions shown in panels in D and E. (D) Equivalent purified T3D^PL, T3D^TD, or T3D^KC cores were subjected to in vitro transcription reactions as described in Materials and Methods, with either 4 NTPs or 3 NTPs (No ATP). Samples were spiked with GAPDH RNA as an internal control, and RNA extracted and subjected to RT-qPCR for reovirus S4, M2, or L2 relative to the level of GAPDH and normalized to value at the 0 min (input) (n = 3 ± standard deviation; **, P < 0.001). Linear regression analysis determined that (i) the slope of T3D^PL differs from that of T3D^KC or T3D^TD and (ii) the slope of T3D^KC does not differs from that of T3D^TD. (E) Equivalent purified cores from T3D^PL or T3D^TD and S4, M1 and L3 gene monoreassortants in an otherwise PL-RG (left) or TD-RG (right) background were subjected to transcription reactions with SYBR green II RNA dye, and fluorescence was monitored at 5-min intervals to monitor increases in RNA abundance over time. Note that in the legend virus labels correspond to the line they are next to and can be used to visualize the hierarchy of highest-to-lowest transcription rates. Fluorescence values were standardized to the value at 0 min (input) (n ≥ 4 ± standard errors of the means). (F) Linear regression analysis was performed on data from panel E (n ≥ 4). Statistical significance was determined using one-way analysis of variance with Tukey’s multiple-comparison test, with either parental T3D^PL or T3D^TD as a control data set. * P < 0.05; ***, P < 0.001; ****, P < 0.0001; ns, P > 0.05.

Having identified core transcription as a key mechanism for the superior replication of T3D^PL, we then asked if any of the three genes that segregated with the large-plaque phenotype of T3D^PL accounted for the inherent core transcription advantage. The M1-encoded μ2 protein and L3-encoded λ1 protein seemed the most likely candidates since both λ1 (21, 39, 40) and μ2 (13, 17, 41) have been previously implicated in modulating NTPase activities. The S4-encoded σ3 viral protein is completely removed from cores and therefore was not anticipated to affect core transcription in vitro. Using parental and mono-reassortant viruses described for the experiment shown in Fig. 4, we assessed the effects of M1, S4, and L3 genes on core transcription. Reassortants were treated with CHT, and purified cores were standardized to equivalent numbers of particles. To facilitate more rapid comparison among many samples, we developed an assay that measures RNA levels in real time throughout the reaction, using RNA-binding SYBR green II. Though less sensitive than RT-qPCR, the SYBR II method accurately recapitulated the enhanced transcriptional rates of T3D^PL relative to those of T3D^TD (Fig. 5E). As expected, no RNA was produced in the absence of rATP (negative control). The T3D^TD-derived M1 decreased RNA synthesis rates when placed into an otherwise T3D^PL background relative to the rate for T3D^PL (Fig. 5E, left). Reciprocally, when the T3D^PL-derived M1 was introduced into an otherwise T3D^TD background, it was sufficient to restore core transcription to rates similar to those of T3D^PL (Fig. 5E, right). When slopes of transcription rates were graphed for three independent experiments (Fig. 5F), it was evident that the M1 gene was necessary and sufficient to confer high transcription rates of T3D^PL. These data clearly implicate the M1-encoded μ2 protein as a rate-limiting determinant of reovirus transcription and a dominant basis for the superior plaque size and oncolytic potency of T3D^PL.

DISCUSSION

Laboratory strains of T3D exhibit distinct plaque sizes and in vivo oncolytic potentials (6). The distinction between T3D lab strains is important not only for bridging previously contradictory findings between laboratories but also to provide a powerful strategy to identify novel reovirus gene/protein functions and facilitate development of more potent oncolytic reoviruses. The current study reveals that the most oncolytic T3D^PL possesses replication advantages in a single round of infection. Two specific mechanisms were identified. First, T3D^PL has elevated attachment to cells, attributed to having fewer particles with ≤3 σ1 cell attachment proteins. Second, T3D^PL transcribes at rates superior to those of the less oncolytic T3D strains, which is ascribed to polymorphisms in the M1 genome segment encoding the μ2 protein. Accordingly, T3D^PL establishes rapid onset of viral RNA and protein synthesis, leading to more rapid kinetics of progeny virus production and a larger virus burst size. These results reveal the importance of rapid onset of infection for reovirus oncolysis.

Previous knowledge about σ1 can help explain why T3D^TD and T3D^KC strains produce more virus particles bearing too few (≤3) σ1 molecules for attachment to host cells. The S1-encoded σ1 proteins of T3D^TD and T3D^KC differ from the protein of T3D^PL at two positions: A22V and A408T. We along with others previously found that mutations in the σ1 anchor domain (amino acids 1 to 27) affect incorporation of σ1 trimers into the λ2 vertex (22, 42, 43), so we strongly anticipate that the A22V polymorphism plays an important role. The sequence difference at residue 408 is located in the σ1 head domain (amino acids 311 to 455), which interacts with cell surface JAM-A (44, 45). However, since the A408T polymorphism is located distal to the JAM-A binding interface and because binding deficiencies of T3D^KC and T3D^TD very closely correlated with having virions with fewer than 3 σ1 trimers, we predict that the A22V polymorphism is most important. Future experiments could segregate the two polymorphisms and test the relative importance of each. Interestingly, however, the S1 genome segment was not one of the three segments (S4, M1, and L3) that strongly segregated with a large-plaque phenotype in reassortant and reverse genetics studies (6). Mostly like, because differences in cell binding are only ∼2-fold, they have less impact on plaque size, viral replication, and oncolysis than the polymorphisms in M1, S4, and L3.

The increased transcription rate of T3D^PL relative to rates of other T3D strains was clearly associated with polymorphisms in the M1 genome segment. The M1-encoded μ2 protein serves at least two distinct functions: as a structural protein within virus cores, μ2 functions as an NTPase and transcription cofactor; and when expressed during virus replication, μ2 associates with tubulin and other viral proteins to accumulate viral macromolecules at sites of virus replication called factories (14, 46). Increased rates of transcription could reflect two alternative scenarios: either that transcription activity is faster or, alternatively, that a higher proportion of virus cores are capable of transcription at all. By affinity isolation of transcribing cores with bromouridine (BrU) antibodies, Demidenko and Nibert eloquently demonstrated that only a proportion of reovirus serotype 1 Lang (T1L) cores were engaged in transcription during in vitro transcription reactions (47). We favor the first scenario because even at a low virus dose, T3D^TD ultimately achieves the same number of infected cells as T3D^PL (Fig. 1). In other words, if T3D^TD had a higher proportion of cores that were transcriptionally inert, it would be expected that fewer cells would become productively infected. The current discovery that T3D^PL μ2 likely has higher rates of transcription provides a new understanding for the impact of μ2. Although NTPase and RNA triphosphatase (RTPase) activities were previously ascribed to μ2 (13, 15), the surprising increase in rate of transcription suggests that μ2 activities are rate limiting during reovirus RNA synthesis. It will be interesting to determine if the polymorphism affects NTP binding, NTPase/RTPase activities, or conformational change during NTP hydrolysis and to relate these activities to RNA unwinding and/or capping using the two T3D strains as a tool.

Four polymorphisms exist between μ2 proteins of T3D^PL and T3D^TD (R150Q, S208P, R342Q, and A528S), but only R150Q, S208P, and A528S are also polymorphic between T3D^PL and T3D^KC and therefore most likely to be implicated. Although no crystal structures are available for mammalian orthoreovirus μ2, a structure for the putative core protein NTPase of grass carp reovirus VP4, in complex with the polymerase, recently became available (PDB accession number 6M99) (48). Sequence comparison between VP4 and μ2 showed that R150Q, S208P, and A528S μ2 polymorphisms align to regions that are mildly conserved between VP4 and μ2 (Fig. 6A). Accordingly, we felt comfortable mapping these polymorphisms in the structure of VP4 to obtain clues about activity.

FIG 6.

Structure-function analysis of M1 polymorphisms and final model. (A) Sequence comparison between T3D^PL, T3D^TD and T3D^KC μ2 (6), versus grass carp reovirus NTPase/VP4 [GenBank accession number Q8JU68] assembled by multiple sequence alignment by Florence Corpet (52). Red indicates amino acid similarities. Stars indicate the locations of polymorphisms in μ2 between T3D^PL and T3D^TD and T3D^KC. (B) Crystal structure of grass carp reovirus VP3 polymerase (blue) and NTPase/VP4 (yellow) from PDB accession number 6M99. Black represents residues previously described to bear similarity to ATPase domains as described in the text. Green indicates the proposed location of the S208P polymorphism, blue indicates the proposed location of the R150Q polymorphism, and red indicates the proposed location of the A538S polymorphism, extrapolated from panel A. (C) Higher magnification of the proposed ATPase domains and location of the A538S polymorphism. (D) Model describes two mechanisms for enhanced replication of the most oncolytic T3D^PL revealed in this study.

Both the R150Q and S208P polymorphisms lie near the NTPase surfaces distal from key polymerase entry and exit portals or from domains previously described to show similarity to motifs of ATPases or NTPase activity site (13, 15, 49) (Fig. 6B). Other laboratories noticed that among different serotypes of reovirus, a proline at position 208 of μ2 produced a filamentous factory morphology while a serine at the same position produced a globular factory morphology (14, 16). The effects on factory morphology are unlikely to contribute to transcription differences in cores. However, variations at position 208 of μ2 were previously found to affect μ2 misfolding and ubiquitination (50). By affecting μ2 stability and folding, it is possible that the 208 variation contributes to transcription differences among T3D strains. The A528S polymorphism, however, lies in close proximity to the ATPase-like domains (Fig. 6C) and is therefore a likely determinant of rapid transcription by T3D^PL. Future exploration into residues surrounding A528S should help reveal further the regulation of NTPase and transcription activities. Very surprisingly, when 312 MRV NTPase sequences available in the NCBI database were assessed, position 528 was found in a highly conserved triad of F/Y/L-A/V-H. While the great majority of MRV NTPases contained an alanine or valine at position 528, T3D^PL μ2 was the only sequence with a serine, and two other MRV μ2 proteins had cysteines at this position. It would be interesting to determine the effect of introducing serines at position 528 into other MRV μ2 proteins and to understand why evolution of MRV favored alanine and valine instead.

Finally, in addition to M1, both S4 and L3 genes were implicated in enhanced oncolytic activity of T3D^PL (6). Our current analysis indicates that both S4 and L3 genes also contribute to the first round of reovirus replication and to postentry but before cell death processes that impact RNA and protein levels. Future studies are needed to determine the precise mechanistic advantages conferred by polymorphisms in these two genes, but studies can now be focused on specific steps early during infection. Altogether this study shows that single amino acid changes drastically affect reovirus replication and supports the idea that strong attention should be paid to the genome and phenotypic characteristics of specific oncolytic virus strains. This study specifically reveals that T3D strains possess inherent cell-independent differences in attachment and transcription. It is possible that a cell-dependent process such as interferon antiviral signaling also contributes to differences between T3D strains and could further affect oncolytic potency.

MATERIALS AND METHODS

Cell lines, reovirus stocks, reovirus passaging, large-scale reovirus amplification, reovirus plaque assays, T3D-PL and T3D-TD reverse genetics system, and reassortant reovirus generation.

Cell lines, reovirus stocks, reovirus passaging, large scale reovirus amplification, reovirus plaque assays, T3D-PL and T3D-TD reverse genetics system, and Reassortant reovirus generation are described in Mohamed et al. (6).

IHC and FIA.

In plaque assays, 4% paraformaldehyde was added for 60 min (min) prior to removing overlays and fixation again with methanol. For cell monolayers, paraformaldehyde fixation was used alone. Cells were washed twice with phosphate-buffered saline (PBS) and incubated with blocking buffer (3% bovine serum albumin (BSA–PBS–0.1% Triton X-100) for 1 h at room temperature. Primary antibody (rabbit anti-reovirus polyclonal antibody [pAb]) diluted in blocking buffer at a final concentration of 1:10,000 was added and incubated overnight at 4°C. Samples were washed three time for  5 min each time with PBS–0.1% Triton X-100. Secondary antibody diluted in blocking buffer (goat anti-rabbit alkaline phosphatase for immunohistochemistry [IHC] and goat anti-rabbit Alexa Fluor 488 for fluorescence infectivity assay [FIA]) was added and incubated for 1 to 3 h at room temperature. Samples were washed three times for 5 min each time with PBS–0.1% Triton X-100. For IHC, nitroblue tetrazolium (NBT; 0.3 mg/ml) (B8503; Millipore Sigma) and 5-bromo-4-chloro-3-indolylphosphate (BCIP; 15 mg/ml) (N6639; Millipore Sigma) substrates diluted in AP buffer (100 mM Tris-HCl, pH 9.5, 100 mM NaCl, 5 mM MgCl₂) were added, and infected cells were monitored for black/purple staining using microscopy. Reactions were stopped with PBS–5 mM EDTA. For FIA, nuclei were stained with 0.5 μg/ml Hoechst 33352 (H1399; ThermoFisher Scientific) for 15 min, and stained samples were visualized and imaged using EVOS FL Auto Cell Imaging System (ThermoFisher Scientific).

Immunofluorescence.

Cells were seeded on coverslips (no. 1.5 thickness). Cell monolayers were washed once with PBS, fixed with 4% paraformaldehyde, washed twice with PBS, and incubated with blocking buffer (3% BSA–PBS–0.1% Triton X-100) for 1 h at room temperature. Primary antibody (mouse anti-σNS 3E10 conjugated to Alexa Fluor 568, mouse-anti-σ3 10G10 conjugated to Alexa Fluor 647, and rabbit anti-μ2 pAb) diluted in blocking buffer was added and incubated overnight at 4°C. Primary monoclonal antibodies (MAbs) mouse anti-σNS and mouse-anti-σ3 10G10 were conjugated to Alexa Fluor 568 and Alexa Fluor 647, respectively, using Apex antibody labeling kits as per manufacturer’s protocol (ThermoFisher Scientific). Samples were washed three times for 5 min each time with PBS–0.1% Triton X-100. Secondary antibody diluted in blocking buffer (goat anti-rabbit Alexa Fluor 488) was added and incubated for 1 to 3 h at room temperature. Samples were three times for 5 min each time with PBS–0.1% Triton X-100. Nuclei were stained with 0.5 μg/ml Hoechst 33352 (H1399; ThermoFisher Scientific) for 15 min. Coverslips were mounted on microscope slides using 10 μl of SlowFade Diamond (S36967; ThermoFisher Scientific) and visualized using an Olympus IX-81 spinning-disk confocal microscope (Quorum Technologies).

RNA extraction and RT-PCR.

Cells were lysed in TRI Reagent (T9424; Millipore Sigma), and the aqueous phase was separated following chloroform extraction as per the TRI Reagent protocol. Ethanol was mixed with the aqueous phase, and an RNA isolation protocol was continued as per the protocol of the GenElute Mammalian Total RNA Miniprep kit (RTN350; Millipore Sigma). RNA was eluted using RNase-free water, and total RNA was quantified using Biodrop DUO (Biodrop). Using 1 μg RNA per 20-μl reaction volume, cDNA synthesis was performed with random primers (48190011; ThermoFisher Scientific) and Moloney murine leukemia virus (M-MLV) reverse transcriptase (28025013; ThermoFisher Scientific) as per the M-MLV reverse transcriptase protocol. Following a 1/8 cDNA dilution, RT-qPCRs were executed as per the protocol of SsoFast EvaGreen Supermix (1725204; Bio-Rad) using a CFX96 system (Bio-Rad).

Western blot analysis.

Cells were washed with PBS and lysed in radioimmunoprecipitation assay (RIPA) buffer (50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1% IGEPAL CA-630 [NP-40], 0.5% sodium deoxycholate) supplemented with protease inhibitor cocktail (11873580001; Roche) and phosphatase inhibitors (1 mM sodium orthovanadate, 10 mM β-glycerophosphate, 50 mM sodium fluoride). For each 12-well plate, 100 μl lysis buffer was used, and volume was scaled up or down according to well size. Following addition of 5× protein sample buffer (250 mM Tris-HCl, pH 6.8, 5% SDS, 45% glycerol, 9% β-mercaptoethanol, 0.01% bromophenol blue) for a final 1× protein sample buffer, samples were heated for 5 min at 100°C and loaded onto SDS-acrylamide gels. After SDS-PAGE, separated proteins were transferred onto nitrocellulose membranes using a Trans-Blot Turbo transfer system (Bio-Rad). Membranes were blocked with 3% BSA–Tris-buffered saline with Tween 20 (TBST) (blocking buffer) and incubated with primary and secondary antibodies (5 ml of blocking buffer and primary and secondary antibody per minimembrane [7 by 8.5 cm]). Membranes were washed three times for 5 min each time with TBST after primary and secondary antibodies (10 ml wash buffer per minimembrane [7 × 8.5 cm]). Membranes with horseradish peroxidase (HRP)-conjugated antibodies were exposed to ECL Plus Western blotting substrate (32132; ThermoFisher Scientific) (2 ml of substrate per minimemberane [7 × 8.5 cm]) for 2 min at room temperature. Prior to visualization, membranes were rinsed with TBT. Membranes were visualized using an ImageQuant LAS4010 imager (GE Healthcare Life Sciences), and densitometric analysis was performed by using ImageQuant TL software (GE Healthcare Life Sciences). Note that a “wash” in the above protocol constitutes addition of reagent, gentle rocking back and forth twice, and removal of reagent.

Flow cytometry binding assay.

L929 cells were detached with CellStripper (Corning), diluted, aliquoted (5 × 10¹⁰ cells/sample/ml), prechilled at 4°C for 30 min, and bound with normalized virions at 4°C for 1 h. Unbound virus was washed off, and cell-bound virus was quantified using flow cytometry following sequential binding with rabbit anti-reovirus pAb and goat anti-rabbit Alexa Fluor 488. After secondary antibody staining, samples were fixed with 4% paraformaldehyde for 30 min at 4°C. Following a wash with PBS to remove 4% paraformaldehyde, cell pellet was gently resuspended in 500 μl of PBS. Samples were processed using a FACSCanto (BD Biosciences) instrument, and data were analyzed using FSC Express, version 5 (De Novo Software). Total cells were gated using forward scatter area (FSC-A) and side scatter area (SSC-A), while single cells were gated using FSC-A and forward scatter height (FSC-H). A minimum of 10,000 total cells were collected for each sample. All steps were performed at 4°C on a microcentrifuge tube rotator, and fluorescence-activated cell sorter (FACS) buffer (PBS–5% fetal bovine serum [FBS]) was used as the antibody diluent and wash buffer. Prior to fixation, cells were centrifuged at 500 × g for 5 min at 4°C. After fixation, cells were centrifuged at 1,000  × g for 5 min at 4°C. Two wash steps were performed following virus binding and primary and secondary antibody incubation. Note that a “wash” in the above protocol constitutes centrifugation, aspiration of supernatant, resuspension of pellet in wash buffer, centrifugation, and aspiration of wash buffer.

Flow cytometry cell death assay.

Early apoptosis and cell death were measured by annexin V and 7-AAD staining precisely as previously described (51), prior to processing with FACSCanto (BD Biosciences) and analysis with FSC Express, version 5 (De Novo Software). Identical gates were applied to all samples of a given experiment and were applied based on discrimination of the live (annexin V and 7-AAD negative) population.

Agarose gel separation of reovirus.

Purified virions (5 × 10¹⁰ virus particles) diluted in 5% Ficoll and 0.05% bromophenol blue were run on a 0.7% agarose gel in TAE buffer (40 mM Tris-HCl, 5 mM sodium acetate, 1 mM EDTA [pH 7.5]) for 12 h at room temperature, stained with Imperial total protein stain for 2 h at room temperature, destained overnight in TAE buffer, and visualized on the ImageQuant LAS4010 imager (GE Healthcare Life Sciences)

In vitro reovirus core transcription assay.

Reovirus cores were generated by incubating purified virions with chymotrypsin (CHT) (C3142; Millipore Sigma) at 14 μg/ml for 2 h at 37 C. CHT digest reactions were halted by addition of protease inhibitor cocktail (11873580001; Roche) and incubation at 4°C. Reovirus cores were pelleted by centrifugation at 100,000 × g for 2 h at 4°C, and reconstituted in 100 mM Tris-HCl, pH 8. Transcription reactions were assembled on ice to include 100 mM Tris-HCl, pH 8, 10 mM MgCl₂, 100 μg/ml pyruvate kinase (P7768; Millipore Sigma), 3.3 mM phosphoenol pyruvate (P0564; Millipore Sigma), 0.32 units/μl RNaseOUT (10777019; ThermoFisher Scientific), 0.2 mM rATP, 0.2 mM rCTP, 0.2 mM rGTP, 0.2 mM rUTP, and 1 × 10¹¹ virus cores per 150-μl reaction volume. Negative-control samples were set up without rATP. Reactions were allowed to proceed at 40°C and at indicated time points, 40-μl transcription aliquots were added to 400 μl of TRI Reagent LS (T3934; Millipore Sigma) containing 3 ng of mouse GAPDH RNA (in vitro transcribed using T7 RiboMAX [Promega], as per the manufacturer’s protocol). Using 10 μg of glycogen (R0551; ThermoFisher Scientific) as a carrier according to manufacturer’s instructions, RNA was purified and converted to cDNA (28025013; ThermoFisher Scientific) using random primers (48190011; ThermoFisher Scientific), and RT-PCR (1725204; Bio-Rad) was performed to quantify reovirus S4, reovirus M2, and mouse GAPDH. Values were standardized to GAPDH and plotted relative to the value at 0 h posttranscription. For high-throughput transcription assays, reaction mixtures were set up in a similar manner, aliquoted into RT-PCR 96-well tubes, and spiked with a 10× final concentration of SYBR green II (S7564; ThermoFisher Scientific). Relative fluorescence was measured at 5-min intervals for 2 h in a CFX96 system (Bio-Rad).