Abstract
The high mutation rate of RNA viruses enables a diverse genetic population of viral genotypes to exist within a single infected host. This in-host genetic diversity could better position the virus population to respond and adapt to a diverse array of selective pressures such as host-switching events. Deep sequencing was used to characterize genomic changes in coronavirus quasispecies during simulated host-switching. Three bovine nasal samples infected with bovine coronavirus (BCoV) were used to infect human and bovine macrophage and lung cell lines. The virus reproduced relatively well in macrophages, but lung cell lines were not infected efficiently enough to allow passage of non lab-adapted samples. Approximately 12 kb of the genome was amplified before and after passage and sequenced at average coverages of nearly 950× (454 sequencing) and 38,000× (Illumina). The consensus sequence of many of the passaged samples had a 12 nucleotide insert in the spike gene, and multiple point mutations were associated with its presence. Deep sequencing revealed that the insert was present but very rare in the unpassaged samples and could quickly shift to dominate the population when placed in a different environment. Analysis of the deep sequencing data indicated that two distinct genotypes circulated at different frequency levels in each sample, supporting the hypothesis that mutations in passaged strains were "selected" from a pre-existing pool rather than through de novo mutation.
Introduction: The Quasispecies Engine of Adaptation
RNA viruses exist within a single host not as a single sequence, but as a "quasispecies"—a cloud of genetically diverse genotypes. This diversity is thought to be the engine behind zoonotic "spillover" events, where a virus jumps from animals to humans, as seen with SARS-CoV and SARS-CoV-2. A fundamental question in viral evolution is whether a virus entering a new host waits for the "right" mutation to happen by chance (de novo) or if it is already "pre-adapted" with a library of variants ready for selection. This study investigates the genomic dynamics of BCoV during simulated host-switching to understand how viral populations navigate new selective pressures.
Methods and Experimental Design
The researchers utilized three natural BCoV samples from infected calves and one lab-adapted strain (Nebraska). These samples underwent serial passage in various human and bovine cell lines, including macrophages (THP-1, Bomac), lung cells (A549, EBL), and enteric cells (HRT-18). To track the evolution of the viral population at high resolution, they employed:
- Ultra-Deep Sequencing: Approximately 12 kb of the 30 kb genome was sequenced using both 454 and Illumina platforms to detect extremely rare variants.
- Haplotype Reconstruction: The software ShoRAH was used to estimate the genetic diversity of the heterogeneous population.
- Structural Modeling: A 3D model of the spike protein’s N-terminal domain was constructed to visualize mutation sites and their potential effects on cell entry.
Results: Selection from a Pre-existing Pool
The most significant finding is that adaptation was primarily driven by the enrichment of variants already present at low frequencies (<1%) in the original bovine samples. This suggests that the "winning" genotypes for new environments are often already present in the viral "cloud."
Phylogenetic Divergence
Phylogenetic trees (A-D) representing different genomic regions of BCoV, showing the clustering of passaged samples and their relationship to the original bovine strains. The Role of Viral Population Diversity in Adaptation of Bovine Coronavirus to New Host Environments, page 3
Analysis of the genomic sequences revealed distinct clustering patterns based on the passage environment. In the natural samples, 78% to 99% of the mutations that eventually became dominant in passaged strains were already present as sub-consensus variants in the original nasal swabs.
The Spike Insert "Switch"
A 12-nucleotide insert in the spike gene (encoding four amino acids: Ser-Arg-Arg-Arg) served as a key marker of adaptation. This insert creates a multibasic site, likely serving as a furin cleavage site or enabling binding to heparin sulfate, allowing the virus to infect a broader range of cell types. In some samples, this insert went from being undetectable by standard methods (0.01% frequency) to dominating the population (>90% frequency) after just a single passage.
Discussion and Structural Insights
Structural modeling showed that 7 out of 8 non-synonymous mutations in the receptor-binding domain were located on surface-accessible residues, indicating direct selective pressure on the virus's entry mechanism. While the virus replicated efficiently in macrophages, it struggled in lung cell lines, suggesting that the path to respiratory adaptation may be more complex than enteric or macrophage adaptation.
Limitations and Future Work
The study used immortalized cell lines, which lack the complex immune pressures of a living host. Furthermore, because the virus did not infect lung cell lines efficiently, the researchers could not fully characterize the mutations required for a transition to a purely respiratory phenotype. Future work using reverse genetics is required to definitively prove the functional impact of these mutations on host range.
Summary updated