Variable and conserved features of copy-back viral genome populations generated de novo during Sendai virus infection
Yanling Yang, Yuchen Wang, Carolina B. López
TL;DR
This study examines how copy-back viral genomes form during Sendai virus infection, revealing both consistent and variable patterns in their generation and accumulation.
Contribution
The study provides the first detailed analysis of de novo copy-back viral genome populations using clean virus stocks.
Findings
Polymerase drop-off positions are distributed throughout the genome, but reattachment occurs preferentially near the trailer end.
Dominant cbVG species remain stable across passages, while low-abundance cbVGs are dynamic and variable.
cbVGs originating near nucleotide 1 are conserved across all stocks, suggesting a replication-specific mechanism.
Abstract
Copy-back viral genomes (cbVGs) are generated during the replication of negative-sense RNA viruses when the polymerase drops off from the genome and reattaches to the nascent strand. cbVGs have strong immunostimulatory properties and impact infection outcomes. Despite their importance, the composition and mechanisms of de novo cbVG generation and accumulation remain unclear due to challenges in obtaining cbVG-free virus stocks (clean stocks). Here, we obtained several clean stocks by independently rescuing recombinant Sendai virus (SeV) six times and verified their cleanliness through PCR, RNA sequencing, and absence of immunostimulatory activity. High multiplicity-of-infection (MOI) passaging of clean stocks produced six high-MOI passaged stocks, each with distinct cbVG populations. Among them, polymerase drop-off (break) positions occurred throughout the genome, while polymerase…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Fig 1
Fig 2
Fig 3
Fig 4
Fig 5- —National Institute of Allergy and Infectious Diseaseshttp://dx.doi.org/10.13039/100000060
- —National Institute of Allergy and Infectious Diseaseshttp://dx.doi.org/10.13039/100000060
- —BJC Investigator Program
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsVirology and Viral Diseases · Animal Virus Infections Studies · Bacteriophages and microbial interactions
INTRODUCTION
Copy-back viral genomes (cbVGs) are one of the most extensively studied classes of non-standard viral genomes (nsVGs) (1). Although initially considered artifacts that propagated along with standard viral genomes (stVGs), cbVGs are now recognized as key modulators of virus–host interaction during the course of infections (2). Moreover, the kinetics and dynamics of cbVG accumulation correlate with disease progression and clinical outcomes (3). The biological activity of cbVGs is accomplished through four major effects: interfering with standard virus replication, activating innate immune pathways via RIG-I and MDA5, promoting persistence through prosurvival signaling, and inducing cellular stress responses that suppress viral protein expression (4–6).
cbVGs are generated during the replication of negative-strand RNA viruses when the polymerase drops off (breaks) from the genome and reattaches (rejoins) to the nascent strand, using it as a template to complete replication (2, 7). Unlike standard genomes, cbVGs contain a unique junction region formed by the combined break and rejoin sites, as well as complementary trailer ends that are present only in cbVGs (2, 8). cbVGs have been observed in several mononegavirales, including Sendai virus (SeV), measles virus, parainfluenza virus 5, respiratory syncytial virus (RSV), and Ebola virus (9–13). SeV, especially the Cantell strain, has long served as a model for studying cbVGs, as a single well-characterized cbVG, cbVG-546, has been consistently detected in SeV Cantell virus stocks and in samples from SeV Cantell infections (8). cbVG-546 is considered the prototypical immunostimulatory cbVG, activating RIG-I/MDA5 or TLR3 signaling, and its derivative, a defective viral genome-derived oligonucleotide, has been used as a vaccine adjuvant in pre-clinical studies (8, 14–20).
Beyond their direct applications, cbVGs play critical roles in viral replication and host responses, making the regulation of their generation an attractive target for therapeutic development (7). However, our understanding of viral and host factors that regulate cbVG generation and accumulation is limited to data showing that the SeV and measles virus C proteins suppress the production and accumulation of cbVGs (21–23), as well as a single amino acid substitution in the SeV N protein drives cbVG production (24). A key challenge in identifying factors that impact the generation of cbVGs and the composition of the cbVGs population in a particular stock is the difficulty of obtaining cbVG-free viral stocks as cbVGs are often carried over from early isolates leading to biased populations. This fact complicates efforts to distinguish the de novo generation of cbVGs from the selection and accumulation of already existing species. For example, it remains unclear whether cbVG-546 in the SeV Cantell strain has been persistently maintained as the virus is passaged and expanded or whether it is newly generated during infection with cbVG-depleted SeV Cantell stocks (25). Resolving such questions is essential for a comprehensive understanding of cbVG biology and for effectively harnessing their therapeutic potential.
Toward this goal, we successfully generated multiple pure stocks of newly rescued recombinant SeV Cantell reporter virus (rSeVC^eGFP^) that were confirmed to be cbVG-free. By passaging these clean stocks at high multiplicity of infection (MOI), we obtained six distinct cbVG populations. Sequencing analysis revealed that break points occurred throughout the genome, with rejoin events preferentially occurring near the trailer end, and that every stock contained cbVGs with a break at position 1. Although diverse cbVG populations arose independently from the same virus, a small subset of cbVG species was conserved across passages, and unexpectedly, shorter cbVGs did not preferentially accumulate as dominant species. Together, these findings reveal both diversity and reproducibility in cbVG generation and provide a foundation for defining the mechanisms that govern this process.
RESULTS
Reverse genetics enables the generation of cbVG-clean Sendai virus Cantell stocks
To obtain SeV stocks free of cbVGs, we independently rescued six recombinant rSeVC^eGFP^ virus stocks, which we named rS1-rS6. rSeVC^eGFP^ was designed based on the SeV Cantell strain, with eGFP inserted between the NP and P genes, as previously described (26) (Fig. 1A). Viruses were rescued by co-transfecting the full-length plasmid and three helper plasmids into BSR-T7 cells to generate passage 0 (P0) stocks. Passage 1 (P1) stocks were expanded from P0 by infecting LLC-MK2 cells at an MOI of 0.1 (Fig. 1B). To evaluate whether the P1 stocks were free of cbVGs, we used three different methods. First, we utilized a well-described PCR assay to amplify cbVGs present in the stocks (25). In this assay, both the stVG and cbVG are reverse transcribed using a single RT primer. The stVG is detected upon amplification of a segment of the virus with specific primers, while cbVGs are detected using the original RT primer and a primer complementary to the trailer of the virus. Because these two primers are oriented in the same direction on the stVG, no stVGs can be amplified (Fig. S1). Using this strategy, the PCR product of cbVG-546 is 278 bp. No obvious cbVG bands were detected in any of the six newly rescued P1 virus stocks infected cells, whereas the SeV genome was detected in all SeV stock-infected samples (Fig. 1C). Next, we used RNA sequencing, followed by analysis with the Viral Opensource DVG Key Algorithm 2 (VODKA2) to detect cbVGs in each stock by precisely detecting their junctions (Break position_Rejoin position) with high sensitivity and specificity (27). VODKA2 analysis revealed only one cbVG junction in each P1 stock, with fewer than five reads, compared to 678,462 reads in our SeV Cantell control stock, which has been passaged in conditions to maintain a high content of cbVGs (Fig. 1D). Dots near the diagonal indicate cbVGs with break and rejoin sites in close proximity, whereas points farther from the diagonal reflect more distant sites. Lastly, we compared the immunostimulatory activity of our stocks to that of the SeV Cantell wild-type control. Our newly rescued clean stock (rS1 P1) triggered much lower antiviral gene expression than SeV Cantell and our cbVG-low SeV Cantell stock, which was prepared by growing SeV Cantell under conditions that deplete cbVGs (25), while maintaining similar replication levels (Fig. 1E and F). Together, these data demonstrate that SeV Cantell can be prepared to contain nearly undetectable levels of cbVGs when launched from plasmids.
*Generation of Sendai virus cbVG-clean stocks using reverse genetics. (A) Schematic diagrams of the SeV genome and the recombinant reporter virus rSeVCeGFP. The lengths of the genomes are indicated. (B) Schematic illustration of the SeV rescue process and the development of passage 1 (P1) clean virus stocks. (C) cbVG and genomic PCR of LLC-MK2 cells infected at an MOI of 1.5 with each of the P1 stocks for 24 h. SeV Cantell stock containing high levels of cbVG-546 (indicated by the red star) was used as a positive control. Mock (M) indicates mock-infected cells. (D) Break-rejoin plots displaying cbVG junctions detected by RNA Seq and VODKA2 analysis for SeV Cantell stock and four independently rescued P1 rSeVCeGFP clean virus stocks. Each dot represents a unique cbVG junction, with dot size indicating the read count. The red arrows point to the only cbVG junction present in each clean virus stock. (E, F) SeV NP, IL29, and ISG56 mRNA levels of A549 cells mock-treated or infected with rSeVCeGFP clean stocks rS1 P1, SeV Cantell, or cbVG-low SeV Cantell stocks at an MOI of 1.5 for 6 h (E) or 24 h (F). The data represent three independent experiments. Statistical analysis was performed using ordinary one-way ANOVA, followed by Dunnett’s multiple comparisons post hoc test. ns: not significant, *: P < 0.05, **: P < 0.01, ***: P < 0.001, ***: P < 0.0001. (G) Schematic illustration depicting the process of passaging SeV stocks at low MOI from passage 2 (P2) to passage 4 (P4). (H) Virus titers from P1 to P4 for the indicated stocks. (I) cbVG and genomic PCR of LLC-MK2 cells infected at an MOI of 1.5 with P1 to P4 of the indicated stocks for 24 h. SeV Cantell stock was used as a positive control, with cbVG-546 indicated by the red star. Mock (M) indicates uninfected cells.
To evaluate whether the stocks could maintain a clean status over successive passages, we used two stocks, rS1 and rS2, to infect the highly susceptible cell line LLC-MK2 cells at an MOI of 0.01 for 2 days. The viruses were then harvested, and infection was repeated, passaging up to the fourth passage (P4) (Fig. 1G). Throughout these passages, we observed no significant changes in the virus titers for both stocks (Fig. 1H), and no clear cbVG species were detected by PCR, indicating that the clean status of the stocks was maintained through these passages (Fig. 1I).
Unique populations of cbVG are generated de novo from clean virus stocks
To test if high-MOI passaged stocks generated from clean rSeVC^eGFP^ P1 stocks contain the dominant cbVG-546 species present in the original SeV Cantell (21), we generated high-MOI passaged stocks from all our independent rescues. To do this, we passaged rS1, rS2, rS4, and rS6 P1 stocks at an MOI of 10 (24) until passage 8 (P8), while rS3 and rS5 were only passaged to P3 (Fig. 2A). RNA sequencing and VODKA2 analyses showed that all P3 (rS1–rS6) and all P8 (rS1, rS2, rS4, and rS6) high-MOI passaged stocks contained significantly higher levels of cbVGs compared to clean P1 stocks, although the amount of cbVG varied among different high-MOI passaged stocks. rS1 and rS2 had higher levels of cbVGs than the other four stocks at P3 (Fig. 2B). As expected, the ratio of total cbVG reads to total viral reads was considerably higher in high-MOI passaged stocks than in the clean stocks (Fig. 2C). As cbVGs are known to interfere with standard viral genome replication by competing for the viral polymerase and other replication resources, inducing antiviral immune signaling pathways, and inhibiting viral protein translation (28), we next measured infectious particles production across passages. As expected, we found fewer infectious viral particles in rS1 and rS2 compared to rS4 and rS6 at earlier passage (P4), corresponding with the higher amount of cbVGs detected in rS1 and rS2 at P3. In contrast, rS4 and rS6 did not show a significant decrease in infectious particles (Fig. 2D). Although no change in viral titer was observed at P3, qPCR analysis showed that SeV NP mRNA levels in cells infected with the rS1 and rS2 P3 stocks were significantly lower than those infected with the P1 stock (Fig. 2E). cbVG PCR of P3 showed that all stocks except rS4 had detectable and different cbVG amplicons, while the cbVGs in rS4 were undetectable (Fig. 2F). None of the P3 stocks showed a dominant amplicon corresponding to cbVG-546, demonstrating that diverse populations of cbVGs are generated de novo by SeV Cantell by P3.
*Unique cbVG populations generated by high MOI passaging of clean stocks. (A) Schematic illustration of the process of creating SeV high-MOI passaged stocks from P2 to passage 8 (P8), starting from P1 clean stocks. (B) Raw number of cbVG reads from total RNA-Seq of clean P1 stocks, and high-MOI passaged P3 and P8 stocks. (C) Ratio of total cbVG reads to total viral reads for clean P1 stocks, and high-MOI passaged P3 and P8 stocks. (D) Infectious virus titers for rS1, rS2, rS4, and rS6 from P1 to P8. # indicates passages where the titers were too low to conduct an infection at 10 MOI, and an MOI of 0.1 was used instead. (E) A549 cells were infected with the rS1 P1 stock, parental Cantell stock, rS1 P3, or rS2 P3 stocks at an MOI of 2. SeV NP mRNA levels were quantified by RT-qPCR at 6, 12, and 24 hpi. Bars represent mean ± SD of biological replicates (n = 3). Statistical significance was assessed by two-way ANOVA with multiple comparisons. ns, not significant; **P < 0.01; ***P < 0.001; ***P < 0.0001. (F) cbVG and genomic PCR of LLC-MK2 cells infected at an MOI of 1.5 with each high-MOI passaged P3 virus stock or SeV Cantell for 24 h. cbVG-546 was indicated by the red star. Mock (M) indicates uninfected cells. (G) Break-rejoin plots displaying cbVG junctions identified in the six independently high-MOI passaged P3 stocks. Each dot represents a unique cbVG junction. The X-axis and Y-axis represent the break positions and rejoin positions in the parental SeV genome, respectively. (H) Comparison of break positions from (G) across stocks in the genome. Each red line indicates a unique break position. Genes are shaded in gray. (I) Comparison of rejoin positions in the genome from (G) across stocks. Each blue line indicates a unique rejoin position. Genes are shaded in gray.
The PCR results are subject to technical limitations of the assay, including primer-template binding preferences and competition during amplification, which may reduce the sensitivity for detecting the full diversity of cbVG species. To further characterize the de novo-generated cbVG populations, we identified cbVGs present in the high-MOI passaged P3 stocks by RNA-Seq. VODKA2 identified populations of cbVGs that significantly differ in each stock, with only a few cbVGs with similar break_rejoin positions detected in more than one stock (Fig. 2G). These data corresponded well with the diversity of species detected by PCR (Fig. 2F). The dominant *de novo-*generated cbVG species in each stock (>1% of the cbVG population) were formed by interrupting replication (break) throughout the viral genome, including at nucleotide 1 (Fig. 2H), and rejoining near the trailer end of the genome (Fig. 2I). Of note, the dominant cbVGs in rS1 and rS2 P3 stocks, including those with break at nucleotide 1 of the genome, were recently validated by our group using PCR-independent direct RNA nanopore sequencing (29). Our results agree with observations in infections with RSV or measles virus, which showed a large diversity in sequences for the polymerase break, while a much more conserved region was observed for the polymerase rejoin during cbVG generation (23, 30, 31).
cbVG formation initiates with genome-wide breaks, with position 1 as a conserved feature
To assess the degree of conservation of the positions used for break and rejoin during cbVG generation across the different stocks, we first quantified the number of breaks, rejoins, and junctions (break_rejoin pairs) and compared their occurrences in the high-MOI passaged P3 stocks. From a total of 484 break positions, 536 rejoin positions, and 846 junctions identified across all stocks, we identified 25 break positions, 72 rejoin positions, and 25 junctions that occurred in two or more stocks (Fig. 3A through C). The specific locations and percentages of the 25 recurring break positions in each stock are shown in Fig. 3D with consecutive nucleotides indicated with a horizontal line below the X-axis. Of note, all stocks have cbVG breaks at position 1 at a relatively high percentage. The rejoins occurring in two or more stocks showed a distinctive hotspot from positions 16,173 to 16,203 on the rSeVC^eGFP^ genome (Fig. 3E), which also includes the rejoin position 16,185 corresponding to the rejoin of cbVG-546 (position 15,290 in the SeV Cantell genome). No similarity in the genomic sequences at the break or rejoin positions was identified (Fig. S2). Strikingly, all 25 junctions that were found in two or more stocks had a break at the beginning of the genome in positions 1–3 (Fig. 3F), indicating that early positions of the genome are frequently rejoined with SeV internal genomic sequences. cbVG-546 was detected at a low percentage (0.0025% among all cbVG reads) in rS2, but not in other P3 stocks. Overall, these analyses demonstrate the limited conservation of cbVG species generated at early virus passage and highlight the diversity of positions for polymerase break and initiation of cbVG formation within the SeV genome, with species breaking at or close to nucleotide 1 as a conserved feature during cbVG formation.
cbVG break, rejoin, and junction positions in high-MOI passaged P3 stocks. (A–C) Number of unique breaks (A), rejoins (B), or junctions (break_rejoin) (C) for each high-MOI passaged P3 stock. “≥2” indicates the number of events that appear in two or more stocks. The numbers above the bars indicate the actual number of events. (D–F) Depiction of breaks (D), rejoins (E), and junctions (F) that appear in two or more stocks. The colors indicate the percentages of those breaks, rejoins, or junctions over the total cbVG reads in their respective stocks following the corresponding scale on the right. Underlines indicate adjacent break or rejoin positions. The rejoin position in red font in (E) corresponds to the rejoin of cbVG-546 in SeV Cantell.
A subset of cbVGs dominates in each stock, independent of the cbVG predicted length
To identify properties that could explain cbVG enrichment during viral passaging, we next analyzed the dominant species in each high-MOI passaged P3 stock. As shown in Fig. 4A, most of the cbVG junctions in P3 were detected at a low number of reads, with a raw median read count around 10. In each stock, the number of cbVG reads that reached 1% of the total cbVG reads ranged from 4 to 13 (Fig. 4B). Interestingly, some junctions were highly represented at this early passage, suggesting that these cbVG species rapidly reached high abundance within the population. The dominant cbVGs (>1% of the total cbVG reads) in each stock at this passage varied with no obvious selection of cbVGs of predicted short length (Fig. 4C and D and Fig. S3B). In fact, rS3 and rS5 had dominant cbVGs with a break at positions 1, predicted to be long cbVGs. Dominant cbVG junctions are shown in Table S1. Some of them have similar break and rejoin positions (break_rejoin is ±1 or 2), suggesting that they belong to the same cbVG species.
Dominant break positions, rejoin positions, and junctions in each high-MOI passaged viral stock. (A) Number of reads per unique cbVGs junction for each high-MOI passaged P3 stock. Each dot represents a unique cbVG junction. The black horizontal lines indicate the median number of cbVG reads for each stock. (B) Percentage of reads for each unique cbVG in each high-MOI passaged P3 stock. Each dot represents a unique cbVG junction, and the black horizontal line indicates the median percentage for each stock. The number on top indicates the number of dominant unique cbVG junctions defined as species with >1% of the total cbVG reads. (C) Mapping of dominant cbVGs from (B) on the rSeVCeGFP genome. Red dots indicate break positions, blue dots indicate rejoin positions. Thick blue lines represent cbVGs with frequencies greater than or equal to 3% of the total cbVG reads. Thin gray lines represent cbVGs with frequencies of 1% to 3% of the total cbVG reads. (D) Dominant cbVGs and their predicted length for all high-MOI passaged P3 stocks. Each dot represents a unique cbVG species. The black horizontal line indicates the median dominant cbVG length for each stock. The dashed line indicates the length of the standard genome. (E) Percentage of unique cbVG junctions (red) and dominant (reads% > 1%) cbVG junctions (blue) in each high-MOI passaged P3 stock that meet the “rule of six.”
We then assessed whether the dominant cbVGs meet the “rule of six,” which establishes that a genome length of a multiple of six is required for optimal paramyxovirus replication (32, 33). As shown in Fig. 4E, 40.1%–99.0% of the unique cbVGs met the “rule of six," and all of the dominant cbVGs met this rule, indicating that meeting the rule of six is a critical factor in their selection.
Dominant cbVGs remain stable through serial passaging, while low-abundance cbVGs continue to diversify
To assess how stable the population of cbVGs is after initial selection, we analyzed the cbVGs of rS1, rS2, rS4, and rS6 up to P8. We selected these stocks as representatives of two stocks with a high level of cbVGs (rS1, rS2) and two with a low level of cbVGs (rS4, rS6) based on P3 sequencing and cbVG PCR results. As shown in Fig. 5A and B, cbVGs were first detected by PCR in P2 in rS1 and rS2, and the dominant species remained stable. However, no clear dominant cbVG bands were seen in rS4 at any passage (Fig. 5C). rS6 showed an intermediate phenotype where amplicons were detected beginning in P3 (Fig. 5D).
cbVG profiles across passages. (A–D) cbVG and genomic PCR results of high-MOI passaged stocks rS1 (A), rS2 (B), rS4 (C), and rS6 (D) from passages P1 to P8 in LLC-MK2 cells infected at 1.5 MOI for 24 h. SeV Cantell (C) containing the predominant cbVG-546 was used as a positive control. Mock (M) indicates uninfected cells. (E) Venn diagrams show the number of unique cbVG junctions in P3 and P8, as well as shared cbVGs between P3 and P8, for each stock. Numbers indicate the cbVG junction counts in each group. (F) Percentage of total cbVG reads in each stock contributed by reads either shared (teal) between P3 and P8 or unique to each passage (yellow). Numbers inside bars indicate the percentage of shared cbVG reads over total cbVG reads. (G–J) Composition of dominant cbVGs (≥1% of total cbVG reads) in P3 and P8 for rS1 (G), rS2 (H), rS4 (I), and rS6 (J). Dominant cbVGs present in P3, P8, or both are shown. Each color represents a different cbVG junction. Numbers within bars indicate the percentage of each cbVG junction relative to the total cbVG reads (only those ≥8% of total cbVG reads are shown). The dashed box indicates all the other cbVGs. (K) Comparison of break positions between P3 and P8 in the genome. Dark gray lines indicate shared breaks between P3 and P8. Each orange or magenta line indicates a unique break position only in P3 or P8. Genes are shaded in gray. (L) Comparison of rejoin positions between P3 and P8 in the genome. Dark gray lines indicate common rejoins between P3 and P8. Each purple or teal line indicates a unique rejoin position in P3 or P8. Genes are shaded in gray.
We next analyzed RNA-Seq data on P8 stocks of rS1, rS2, rS4, and rS6, and looked for cbVG junctions present only in P3 or P8, or those shared in both passages. Shared unique cbVG junctions were present in both P3 and P8 in all stocks (Fig. 5E), and the fraction of total shared cbVG reads reached 86.8%–99.8% among total cbVG reads (Fig. 5F). Although not all dominant cbVGs were shared between P3 and P8, and two dominant cbVGs showed a marked decline in abundance from P3 to P8 in rS6, the most abundant cbVGs in rS1, rS2, and rS4 remained consistent between P3 and P8. (Fig. 5G through J). Comparison of breaks and rejoins of the unique cbVGs (those present in only one of the passages) revealed breaks and rejoins that were present in P3 and absent in P8, accompanied by newly generated breaks present only in P8 across the genome (Fig. 5K and L). These results suggest that in most cases, dominant cbVGs remain stable during passaging, while low-abundance cbVGs continue to diversify. However, notable changes in the cbVG population were observed in stock rS6, suggesting that cbVG population stability can vary depending on the cbVG population composition.
cbVG dominant species do not associate with specific SeV genomic variants
Lastly, we assessed whether the generation or accumulation of dominant cbVG species was associated with SeV genomic variants that arose during passages. To identify SeV variants, we used the bioinformatics tool iVar (34) that calls variants from sequencing data by aligning reads to a reference genome. As shown in Table S2, a few variants were detected in each stock, with P8 containing more variants than earlier passages, as expected. At P3, only rS1 carries a high-frequency non-synonymous variant (A9996C) in the L protein, resulting in a methionine to leucine substitution. In addition, T20A and T24A variants were detected in the leader. Non-coding region variants were detected in all four stocks, with a higher frequency in P8, indicating that these sites are highly variable. In conclusion, no significant variants were identified that could explain differences in cbVG species or abundance between stocks.
DISCUSSION
Although nsVGs have been observed in many RNA viruses, including positive-sense RNA viruses, negative-sense RNA viruses, double-stranded RNA viruses, and retroviruses as previously reviewed (2), for over six decades, they were initially considered to be replication artifacts with no real physiological value. More evidence has shown that nsVGs occur during natural infections in humans and play crucial roles in driving viral processes and virus–host interactions. For example, in the case of influenza viruses, it has been shown that patients with high severity or fatal outcomes have reduced nsVG levels than those with mild disease (35). Our investigations on RSV have revealed that immunostimulatory cbVGs, naturally generated during RSV replication, can induce an innate antiviral response in both mice and humans (12). Moreover, the kinetics of nsVG accumulation and duration could predict the clinical outcome of RSV infection in humans (3). In measles virus infection, cbVGs have been detected in subacute sclerosing panencephalitis, with a small subset capable of spreading across multiple regions in the brain (31, 36). Together, these studies highlight the importance of understanding how cbVGs are generated, how they accumulate, what species exist, and what functions they have.
In this study, we addressed a major challenge in cbVG research: the inability to distinguish between de novo cbVG generation and carryover from earlier virus stocks. By generating cbVG-free recombinant Sendai viruses and passaging them under defined conditions, we demonstrate that cbVGs arise reproducibly but with stock-specific compositions, indicating both conserved and stochastic features of their formation. These results advance our understanding of cbVG biology and establish an experimental system to study the mechanisms underlying their generation and accumulation.
Our six high-MOI passaged stocks, generated from independently rescued rSeVC^eGFP^, exhibited distinct cbVG profiles and levels with few conserved features. In the past, the lack of truly cbVG-clean viral stocks limited studies of cbVG generation to identifying the cbVGs present in virus populations and tracking the accumulation of specific cbVGs during serial passaging (37, 38). By contrast, the study of newly rescued recombinant viruses overcomes this limitation, enabling a more systematic characterization of cbVG formation. Notably, studies examining the role of the SeV and measles virus C protein in cbVG generation also revealed the presence of de novo generated cbVG species in newly rescued viruses (21, 23); although cbVGs were not the main focus of those studies, their findings support our observation that cbVG generation is highly variable yet retains certain conserved features.
In this study, we observed cbVGs with junctions of position 1, or close to it, and a sequence closer to the trailer of the genome across all six rescued virus stocks, indicating that this species is a frequent occurrence. Similar potentially long cbVGs were detected in RSV-infected in vitro samples, as well as in clinical samples, but remain understudied (39). Moreover, our recent work using direct RNA sequencing of rSeVC^eGFP^ stocks has further validated their existence (29). Although long cbVGs have often been debated as sequencing or assembly artifacts, our findings provide direct evidence that they are generated during infection and may represent a conserved feature of paramyxovirus replication. Importantly, the presence of these predicted long cbVGs raises the possibility that they play a role in modulating the generation of shorter cbVGs, viral replication dynamics, and/or host antiviral responses, warranting further investigation.
While short cbVGs were previously thought to preferentially accumulate as dominant species due to their shorter replication cycles, our results demonstrate that cbVG length is not necessarily linked to dominance. Instead, cbVGs that conform to the “rule of six” showed a clear replication advantage, and all dominant cbVGs in our study adhered to this rule, consistent with previous reports (32). As our analysis was limited to eight passages, whether extended passaging would alter cbVG dynamics remains an open question for future studies.
Comparison of cbVG populations in P3 and P8 stocks showed that new cbVGs continued to arise even in the presence of dominant cbVGs, but they did not accumulate to high levels by passage 8. Together with the persistence of cbVG-546 in the SeV Cantell strain, these findings suggest that once established, dominant cbVGs are difficult to replace, likely due to a combination of replication dynamics, host selection pressures, and potential functional advantages. Although many novel cbVGs were generated in our study, we did not investigate whether they share functional properties, such as activating innate immune responses, interfering with genome replication and protein expression, or contributing to persistence. Functional differences among cbVGs could explain discordant results between viral titers and cbVG reads. We predict that some cbVGs are more interfering than others, and therefore, the specific cbVG population in each stock will differentially impact the standard virus. If conserved functions are observed, it would be important to assess if common sequence or structural features are involved. For example, our recent work demonstrated that specific nucleotides in the terminal loop of the cbVGs are necessary for strong activation of RIG-I signaling in cells (40), raising the possibility that similar motifs underlie the activity of newly generated cbVGs.
Since a virus is a community of particles containing different standard and non-standard viral genomes (1), our stocks contained not only cbVGs but also variants of the standard virus, with higher frequencies of variants observed at later passages, as expected. None of the variants could explain the frequency or characteristics of the different cbVG populations. Among the six P3 stocks, two (rS1 and rS2) had high cbVG levels; however, only rS1 carried a high-frequency mutation in the polymerase L gene, indicating that this mutation is not directly linked to cbVG abundance. All stocks had variants at positions 20 and 24 of the leader region, which lie within the first element of the SeV bipartite replication promoter, consisting of an initial element spanning nucleotides 1–31 and a second element located at nucleotides 79–96 (41). These variants were more frequent at P8 than at P3, suggesting that this site represents a high-frequency adaptive mutation in LLC-MK2 cells. Previous studies on SeV reported that a single G576T mutation (N protein D153Y) is associated with cbVG production, likely by weakening N–N interactions and reducing nucleocapsid density (24); however, we did not detect this mutation in our stocks, highlighting that the variant landscape may depend on both passage history and host cell context. A recent study on RSV found that a certain cbVG rejoin hotspot is linked to the trailer sequence. When a poly-U mutation was introduced into the trailer, cbVG generation was reduced, but during passaging, a viral variant arose that restored cbVG production (42). In our study, however, we did not observe any mutations in the trailer sequence related to cbVG abundance. This difference may be virus-specific. Such arising variants could influence cbVG generation or viral adaptation in ways that extend beyond replication efficiency alone, an area that warrants further investigation.
In summary, by establishing cbVG-free SeV stocks, we were able to directly examine the de novo generation and accumulation of cbVGs. Our findings demonstrate that cbVG generation is both frequent and stochastic; yet certain features, such as adherence to the “rule of six” and breakpoints near position 1 of the genome, appear conserved. These insights highlight the dual nature of cbVG biology, shaped by both random polymerase errors and selective forces that favor cbVG species. Future studies should investigate how cbVG diversity translates into functional outcomes, including their roles in modulating viral replication, shaping host immune responses, and influencing infection persistence. A deeper understanding of these processes will not only clarify the role of cbVGs in viral pathogenesis but may also open new avenues for harnessing cbVGs as antiviral or immunomodulatory tools.
MATERIALS AND METHODS
Cell lines
BSR-T7/5 cells (kindly provided by Dr. Conzelmann) (43), LLC-MK2 cells (ATCC, #CCL-7), and A549 cells (ATCC, #CCL-185) were cultured in tissue culture medium (Dulbecco’s modified Eagle’s medium) (Invitrogen,#11965092) supplemented with 10% fetal bovine serum (Sigma, #F0926), gentamicin 50 ng/mL (ThermoFisher, #15750060), L-glutamine 2 mM (Invitrogen, #G7513), and sodium pyruvate 1 mM (Invitrogen, #25-000-C1) at 5% CO_2_ 37°C. Cells were treated with mycoplasma removal agent (MP Biomedical, #3050044) and tested monthly for mycoplasma contamination using the MycoAlert Plus mycoplasma testing kit (Lonza, #LT07-318).
Virus rescue and passage
Six independent rSeVC^eGFP^ viruses, named rS1 to rS6, were rescued as previously described (26) and titered as described elsewhere (44). High-MOI passaged virus stocks were developed from the rescued viruses as described elsewhere (29, 45). rS1 and rS2 have been reported previously as rSeVB and rSeVA (29, 46).
Copy-back viral genome-PCR
To detect cbVGs by PCR, LLC-MK2 cells were infected with SeV at 1.5 MOI for 24 h. Total RNA of infected cells and mock cells was extracted using Kingfisher and a Mag-MAX mirVana Total RNA Isolation Kit (Thermo Fisher, #A27828) following the manufacturer’s guidelines. Two hundred fifty nanograms of total RNA was used for cDNA synthesis with a high-capacity RNA to cDNA kit (Thermo Fisher, #18080051) with our RT primer (25) to capture viral RNA, including cbVGs. SeV cbVG RT-PCR was done as described in references 25, 46.
Quantitative reverse transcription PCR
Total RNA from infected cells and control samples was extracted and transcribed as described above. qPCR was performed on an Applied Biosystems QuantStudio 5 system using SYBR Green (Thermo Fisher, #S7564) and 5 μM forward and reverse primers specific for SeV NP (forward: 5′-TGCCCTGGAAGATGAGTTAG-3′; reverse: 5′-GCCTGTTGGTTTGTGGTAAG-3′) (19), IL29 (forward: 5′-CGCCTTGGAAGAGTCACTCA; reverse: 5′-GAAGCCTCAGGTCCCAATTC-3′) (47), and ISG56 (forward: 5′-GGATTCTGTACAATACACTAGAAACCA-3′; reverse: 5′-CTTTTGGTTACTTTTCCCCTATCC-3′) (47). Relative transcript levels were normalized to human GAPDH and β-actin expression as previously described (48).
RNA sequencing, VODKA2, and iVar
RNA sequencing libraries were prepared using the Illumina TruSeq Stranded Total RNA kit and sequenced on a NextSeq platform (High Output, 2 × 150 bp), generating ~30 million reads per sample. Data preprocessing and cbVG detection were performed with VODKA2 as previously described (27). In some cases, consecutive nucleotides were identified by VODKA2 as break or rejoin sites. These may represent different cbVGs or a single cbVG that was split into multiple ones during library preparation, sequencing, or bioinformatic processing. For this analysis, we considered them as individual break or rejoin sites. Viral variants were detected using iVar (34).
Statistics
Statistics were calculated using GraphPad Prism Version 10 (GraphPad Software, San Diego, CA).
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1González Aparicio LJ, López CB, Felt SA. 2022. A virus is a community: diversity within negative-sense RNA virus populations. Microbiol Mol Biol Rev 86:e 0008621. doi:10.1128/mmbr.00086-2135658541 PMC 9491172 · doi ↗ · pubmed ↗
- 2Vignuzzi M, López CB. 2019. Defective viral genomes are key drivers of the virus-host interaction. Nat Microbiol 4:1075–1087. doi:10.1038/s 41564-019-0465-y 31160826 PMC 7097797 · doi ↗ · pubmed ↗
- 3Felt SA, Sun Y, Jozwik A, Paras A, Habibi MS, Nickle D, Anderson L, Achouri E, Feemster KA, Cárdenas AM, Turi KN, Chang M, Hartert TV, Sengupta S, Chiu C, López CB. 2021. Detection of respiratory syncytial virus defective genomes in nasal secretions is associated with distinct clinical outcomes. Nat Microbiol 6:672–681. doi:10.1038/s 41564-021-00882-333795879 PMC 9098209 · doi ↗ · pubmed ↗
- 4Xu J, Sun Y, Li Y, Ruthel G, Weiss SR, Raj A, Beiting D, López CB. 2017. Replication defective viral genomes exploit a cellular pro-survival mechanism to establish paramyxovirus persistence. Nat Commun 8:799. doi:10.1038/s 41467-017-00909-628986577 PMC 5630589 · doi ↗ · pubmed ↗
- 5Manzoni TB, López CB. 2018. Defective (interfering) viral genomes re-explored: impact on antiviral immunity and virus persistence. Future Virol 13:493–503. doi:10.2217/fvl-2018-002130245734 PMC 6136085 · doi ↗ · pubmed ↗
- 6Gonzalez Aparicio LJ, Yang Y, Hackbart MS, López CB. 2023. Copy-back viral genomes induce a cellular stress response that interferes with viral protein expression without affecting antiviral immunity. P Lo S Biol:2023.05.17.541157. doi:10.1371/journal.pbio.3002381 PMC 1069536237983241 · doi ↗ · pubmed ↗
- 7Genoyer E, López CB. 2019. The impact of defective viruses on infection and immunity. Annu Rev Virol 6:547–566. doi:10.1146/annurev-virology-092818-01565231082310 · doi ↗ · pubmed ↗
- 8Mercado-López X, Cotter CR, Kim W-K, Sun Y, Muñoz L, Tapia K, López CB. 2013. Highly immunostimulatory RNA derived from a Sendai virus defective viral genome. Vaccine (Auckl) 31:5713–5721. doi:10.1016/j.vaccine.2013.09.040PMC 440609924099876 · doi ↗ · pubmed ↗