Comparison of intracellular and encapsidated viral RNA reveals selective constraints on reovirus recombination

Technical Summary: Comparison of Intracellular and Encapsidated Viral RNA Reveals Selective Constraints on Reovirus Recombination

Problem Statement
Recombination is a primary driver of genetic diversity, evolution, and host adaptation in RNA viruses. While recombination mechanisms are well-characterized for positive- and negative-strand RNA viruses, they remain poorly understood for segmented, double-stranded RNA (dsRNA) viruses like reovirus. Previous studies on reovirus, which possesses a genome of ten dsRNA segments, suggested low recombination frequencies and proposed that recombination occurs primarily during minus-strand synthesis. However, these prior analyses were limited to genomic dsRNA extracted from purified viral particles. Consequently, they characterized only the subset of recombined RNAs that successfully survived the processes of assembly, assortment, and packaging, potentially missing a broader population of recombination events generated during infection.

Methodology
To address the limitations of previous studies, the authors conducted a comparative analysis of recombination events in two distinct sample types derived from the same infections:

1. Intracellular Viral RNA: Total cellular material containing genomic dsRNA, replication intermediates, mRNAs, and recombination products that may not be eligible for packaging.
2. Encapsidated Viral RNA: Genomic dsRNA extracted from purified virions, treated with benzonase to remove non-encapsidated nucleic acids.

The study utilized three independent reovirus T1L clones in two separate experiments with a high multiplicity of infection (10 PFU/cell) for approximately one round of replication (24 hours). RNA was prepared using the ClickSeq approach, which employs 3'-azido-nucleotides for stochastic chain termination during reverse transcription, followed by click chemistry ligation of adapters. This method was chosen to minimize PCR bias and accurately capture RNA diversity. Libraries were sequenced on an Illumina NovaSeq platform.

Recombination junctions were identified using the Viral Recombination Mapper (ViReMa) algorithm, defined as insertions or deletions >5 bp flanked by high-quality alignment. The study focused on four segments (L1, L3, M1, M3) to provide a genomic snapshot. Analyses included:

• Quantification of junction frequency (JFreq).
• Principal Coordinate Analysis (PCoA) to compare recombination profiles.
• Detailed characterization of microhomology patterns (nucleotide identity surrounding junctions).
• Assessment of nucleotide biases and the utilization frequency of available microhomology sequences across the genome.

Key Results

• Recombination Frequency: Intracellular viral RNA exhibited a substantially higher recombination frequency (mean JFreq = 46.1) compared to particle-associated RNA (mean JFreq = 9.3). Intracellular RNA was found to be almost exclusively plus-strand (98.5%), whereas particle-associated RNA contained junctions from both strands, with a slight predominance of minus-strand junctions.
• Profile Similarity: PCoA revealed that recombination profiles clustered more tightly by sample type (intracellular vs. encapsidated) than by biological replicate or infection experiment. This indicates that the recombination landscape is fundamentally different between the intracellular environment and the final virion population.
• Microhomology Patterns:
  • Particle-associated RNA: Showed strong positional identity (microhomology) extending ~9 nucleotides upstream and ~3 nucleotides downstream of junctions. Approximately 17–33% of junctions contained six identical nucleotides, and 28–47% contained four identical nucleotides.
  • Intracellular RNA: Showed significantly lower microhomology, with positional identity largely confined to the −3 to −1 positions. Only ~1–5% of junctions contained six identical nucleotides, and ~6–11% contained four.
• Sequence Bias: Particle-associated RNA appeared sequence-independent regarding nucleotide enrichment at junctions. In contrast, intracellular plus-strand RNA exhibited a modest but consistent nucleotide bias (enrichment of uridine/thymine at −3 to −1 and adenine at +1), suggesting sequence context plays a role in intracellular recombination.
• Site Utilization: While four-nucleotide microhomology sequences are broadly distributed across the genome, particle-associated RNA utilized a very small subset of these sites (78–89% of sequences were unused). Intracellular RNA utilized available microhomology sequences much more broadly (53–75% unused), indicating a less selective process during generation.

Key Contributions
The study provides the first direct comparison of recombination landscapes in intracellular versus encapsidated reovirus RNA. It demonstrates that the low recombination rates previously reported for reovirus are likely an artifact of analyzing only the packaged viral population. The data reveals that recombination is more frequent and less selective during the intracellular phase (specifically during plus-strand synthesis) than previously recognized. Furthermore, the distinct microhomology signatures between the two sample types suggest that recombination mechanisms or constraints differ between transcription/replication and the subsequent packaging steps.

Significance
The authors conclude that recombination in reovirus occurs more frequently and promiscuously than previously thought, likely during plus-strand synthesis. However, the viral population is subject to a "selective filter" during assortment and packaging. Only a specific subset of recombination products—those that likely meet specific structural or sequence criteria for packaging—are retained in the final virion population. This finding challenges the view of reovirus recombination as a highly selective event occurring solely during minus-strand synthesis and highlights the importance of analyzing total cellular nucleic acids to understand the full scope of viral genetic diversity and evolution. The results suggest that defective viral genomes (DVGs) generated via recombination are more diverse intracellularly than the subset that successfully integrates into virions.