Design and Evolution

Structures A half century of structural studies have shown the design principles controlling self-assembly of viral proteins into icosahedral virus capsids. A large repertoire of virus structures, mainly determined using X-ray crystallography and cryo-electron microscopy, are now conveniently available and carefully annotated in the VIPER database. These structures are not only important to the study of virus structure/function relationships, but they also help to decipher virus evolution.

Unlike cellular organisms, viruses defy conventional classification schemes based on bioinformatics analysis of macromolecular sequences. The huge selective pressure to keep viruses functional results in immense genomic variation that prevents a meaningful and informative taxonomy. Furthermore, icosahedral viruses are intrinsically limited in “structure space” by the structural constraints associated with assembling a coat protein into a 60-fold symmetric protein cage, the capsid, which can be achieved only in a limited and well-defined number of combinations.

The realization that capsid structure persists much longer than viral genomes or protein sequences led to the use of experimentally determined virus structures to define virus taxonomy and identify virus lineages. Amazingly, the tremendous variety and exhilarating complexity of icosahedral viruses existing in nature can be rationalized to four combinations by which one or a few major capsid proteins (MCPs) self-assemble to form an icosahedral lattice. These “modes” of assembly were used to define four viral lineages populating the virosphere:

  1. Picornoviridae, like the single-stranded foot-and-mouth disease virus in which the building block is an eight-stranded β-barrel MCP usually arranged parallel to the surface of the virus.
  2. Reoviruses, such as the double-stranded RNA Bluetongue virus that contains a mostly α-helical MCP arranged to form a core of 60 homodimers.
  3. Tailed bacteriophages like the Salmonella-phage P22, herpesviridae, and even certain archaeal viruses which build their capsid around the flexible HK97 MCP fold.
  4. PRD1-adenovirus type viruses in which a capsid lattice is made of vertical double β-barrels stacked against each other and orthogonal to the capsid that trimerize to form pseudo-hexameric capsomers. This fourth lineage, possibly the least well characterized, includes viruses that infect eukaryotes, bacteria, and archaeal and that often contain a lipid membrane underneath the icosahedral protein capsid.

A Greasy Aid to Capsid Assembly: Lessons from a Salty Virus. (2015) Structure, 23(10): 1777-1779.
Archaeal viruses constitute the least explored niche within the virosphere. Structure-based approaches have revealed close relationships between viruses infecting organisms from different domains of life. Here, using biochemical and cryo-electron microscopy techniques, we solved the structure of euryarchaeal, halophilic, internal membrane-containing Haloarcula hispanica icosahedral virus 2 (HHIV-2). We show that the density of the two major capsid proteins (MCPs) recapitulates vertical single β-barrel proteins and that disulfide bridges stabilize the capsid. Below, ordered density is visible close to the membrane and at the five-fold vertices underneath the host-interacting vertex complex underpinning membrane-protein interactions. The HHIV-2 structure exemplifies the division of conserved architectural elements of a virion, such as the capsid, from those that evolve rapidly due to selective environmental pressure such as host-recognizing structures. We propose that in viruses with two vertical single β-barrel MCPs the vesicle is indispensable, and membrane-protein interactions serve as protein-railings for guiding the assembly.

This entry was posted in Microbiology and tagged , , , , , , . Bookmark the permalink.

3 Responses to Design and Evolution

  1. Ed Rybicki says:

    “immense genomic variation that prevents a meaningful and informative taxonomy”?? Huh? The ICTV has done pretty well with that diversity, and is slowly building bridges between families to encompass obviously-related orders – so I for one think that the taxonomy that has been erected so far is both meaningful and informative. Relating ALL viruses is a pipe dream; I think there are still too many origins for viruses for there to be a truly unified taxonomy that reflects evolutionary relationships – but that doesn’t mean we haven’t gone a long way!

    • AJ Cann says:

      Ah, but they don’t have an rRNA equivalent.

      • Ed Rybicki says:

        Sure! Because they don’t have a common origin. They are also possibly more modular than cellular organisms, in mixing structural modules and replication machinery modules, but the several unrelated taxons representing virus diversity are pretty organised?

Comments are closed.