Abstract
RNA viruses are the fastest evolving known biological entities. Consequently, the sequence similarity between homologous viral proteins disappears quickly, limiting the usability of traditional sequence-based phylogenetic methods in the reconstruction of relationships and evolutionary history among RNA viruses. Protein structures, however, typically evolve more slowly than sequences, and structural similarity can still be evident, when no sequence similarity can be detected. Here, we used an automated structural comparison method, homologous structure finder, for comprehensive comparisons of viral RNA-dependent RNA polymerases (RdRps). We identified a common structural core of 231 residues for all the structurally characterized viral RdRps, covering segmented and non-segmented negative-sense, positive-sense, and double-stranded RNA viruses infecting both prokaryotic and eukaryotic hosts. The grouping and branching of the viral RdRps in the structure-based phylogenetic tree follow their functional differentiation. The RdRps using protein primer, RNA primer, or self-priming mechanisms have evolved independently of each other, and the RdRps cluster into two large branches based on the used transcription mechanism. The structure-based distance tree presented here follows the recently established RdRp-based RNA virus classification at genus, subfamily, family, order, class and subphylum ranks. However, the topology of our phylogenetic tree suggests an alternative phylum level organization.
Highlights
RNA polymerases are key enzymes in RNA virus replication and transcription
We selected 42 high-resolution viral RNA-dependent RNA polymerases (RdRps) structures from the Protein Data Bank, one from each viral species (Supplementary Table S1). These structures represent polymerases of (+)RNA viruses from the viral families Caliciviridae, Coronaviridae, Flaviviridae, Leviviridae and Permutotetraviridae, (−)RNA viruses of the families Arenaviridae, Orthomyxoviridae, Peribunyaviridae, Pneumoviridae, Phenuiviridae and Rhabdoviridae, and double-stranded RNA (dsRNA) viruses of the families Birnaviridae, Cystoviridae, Picobirnaviridae and Reoviridae (Table 1)
The selected proteins were structurally aligned using homologous structure finder (HSF) [33], which progressively compares protein structures and merges the most similar structures by identifying the equivalent residues until all the structures are part of a hierarchical clustering and the common structural core is identified for the entire dataset
Summary
RNA polymerases are key enzymes in RNA virus replication and transcription. All currently known viral RNA-dependent RNA polymerases (RdRps) share a right-handshaped appearance with palm, fingers and thumb subdomains [1,2,3,4], suggesting a shared evolutionary origin. Stepwise sequence alignment of separate sequence clusters and subsequent analysis of inter-cluster similarities with profile-profile methods was used recently for the comparison of wide range of viral RdRp sequences [12] Based on these data, the International Committee on Taxonomy of Viruses (ICTV) has introduced megataxonomy for RNA viruses of the realm Riboviria [13,14,15]. High-resolution RdRp structures from (−)RNA viruses have become available since these studies, both from viruses having segmented (families Orhomyxoviridae, Peribunyaviridae, Arenaviridae and Phenuiviridae) and non-segmented (−)RNA genome (families Rhabdoviridae and Pnemoviridae) [2,26,27,28,29,30] All these new structures share similarities with the RdRps of (+)RNA and dsRNA viruses suggesting a common origin. Our results largely support the recently established RdRp-based RNA virus taxonomy and bring new insights into the functional specialization of RdRp during evolution
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
