Homologous recombination is an event in which parts of a genome from different parent lineages are combined to form a new lineage. This can occur for coronaviruses when a single cell is co-infected, or infected by two virus strains at once. While the concept of recombination events in virology is not new, current phylogenetic and phylodynamic methods of representing shared ancestry of virus families often ignore recombination events. Specifically, phylogenetic methods assume that virus variants evolve through mutations and a single ancestor can explain the genetic basis of its “offspring” virus variants. To include recombination events in shared ancestry analyses, Drs. Nicola Müller, Kathryn Kistler and Trevor Bedford from the Bedford lab designed a BEAST2 compatible, open-source software package that uses a “Markov chain Monte Carlo approach to infer recombination networks from genetic sequence data under a template switching model of recombination,” reported the researchers. With this new tool in hand, the researchers analyzed the genetic data of different coronaviruses to determine the frequency of recombination events for seasonal, SARS and MERS coronaviruses. This work is forthcoming in Nature Communications and currently available as a preprint in bioRxiv.
“Recombination poses a great challenge to the field of phylogenetics in that it breaks one of the fundamental assumptions of the field (namely that how viruses are related can be described using a tree). Our work provides a way to account for recombination and to study how recombination impacts the evolution of viruses. This allows us, for example, to show just how important recombination is in shaping the evolution of SARS-like viruses, such as SARS-CoV-1 and 2 or to show how frequently recombination occurs in seasonal coronaviruses,” commented Dr. Müller. Genome recombination can be beneficial for viruses in different ways. For example, recombination can introduce genetic variability or purge toxic mutations to benefit viral fitness. Thus, study of recombined sequence and recombination rates during viral evolution can inform on favorable and toxic mutations depending on which segments of the genome have positive or negative selection pressure for inclusion or exclusion, respectively, during recombination events.
The researchers first applied the Markov chain Monte Carlo approach to infer how SARS-like coronaviruses are related. The resulting evolutionary history includes multiple recombination events and formed a network-like image of shared viral ancestry over time. Interestingly, these recombination events mostly occurred between closely related SARS-like coronaviruses. Evolutionary histories were also constructed for MERS-CoV and seasonal human coronaviruses. Similar to the SARS-like coronaviruses, recombination events were abundant over time. To compare recombination rates, SARS-like viruses undergo about 0.06 recombination events per lineage per year, which is less frequent than the pandemic H1N1 Influenza B virus at about 0.15 reassortment events per lineage per year. From the sampled genetic data, MERS-CoV had the highest rate of recombination at about 2.4 recombination events per lineage per year. These findings support the significance of homologous recombination events in the evolution of coronaviruses.
The researchers also looked at rates of recombination at specific sites within the viral genome, including a focused comparison between the two subunits of the spike protein. Their findings inferred a modest increase in recombination for spike S1 subunit that binds to host receptors for entry into cells as compared to the conformationally dynamic spike S2 subunit. This finding is intriguing as it may suggest that receptor-binding domain swapping is favored over the stem-helix region of spike during recombination in co-infected cells.
“Novel SARS-CoV-2 variants have so far been the result of mutations. The abundance of recombination events in seasonal human coronavirus now raises the question if and how recombination will contribute to creating novel SARS-CoV-2 variants,” stated Dr. Müller. The lab proposes that using Bayesian phylogenetic tracking of evolving SARS-CoV-2 viruses may inform on positively selected genome fragments and help guide our understanding of which virus variants may appear in the future.
The spotlighted research was funded by the Swiss National Science Foundation, NSF GRFP Fellowship, Pew Biomedical Scholarship, and the National Institutes of Health.
Müller NF, Kistler KE, and Bedford T. 2021. Recombination patterns in coronaviruses. Version 2. bioRxiv. Preprint. [revised 2022 Feb 8].