Near-atomic, non-icosahedrally averaged structure of giant virus Paramecium bursaria chlorella virus 1

Qianqian Shao, Irina V. Agarkova, Eric A. Noel, David D. Dunigan, Yunshu Liu, Aohan Wang, Mingcheng Guo, Linlin Xie, Xinyue Zhao, Michael G. Rossmann, James L. Van Etten, Thomas Klose, Qianglin FangNature Communications 2022 · DOI

TL;DR

By breaking away from traditional icosahedral averaging, researchers have achieved a near-atomic (~3.8 Å) resolution structure of the giant virus PBCV-1. This approach revealed that the viral shell is far more complex than previously thought, featuring five distinct variants of the major capsid protein that create specialized "microenvironments" for spikes, fibers, and internal lipid vesicles.

What Problem Does This Paper Address?

Giant viruses, or Nucleocytoplasmic Large DNA Viruses (NCLDVs), are massive—often larger than some bacteria. While we’ve known they generally follow an icosahedral (20-sided) symmetry, this symmetry is an approximation. Many giant viruses possess "asymmetric" features, such as a single unique vertex used for injecting DNA into a host or "stargate" structures for genome release.

The challenge is that standard Cryo-Electron Microscopy (cryo-EM) reconstruction techniques rely on "icosahedral averaging" to boost resolution. This process effectively "blurs out" any feature that doesn't appear identically at all 12 vertices of the virus. Consequently, the very components that drive the viral life cycle—the unique spikes and specialized portals—have remained structurally mysterious. Furthermore, the sheer size of these viruses (up to 500 nm) introduces technical hurdles like the "Ewald sphere effect," where the depth of the particle makes traditional 2D-to-3D projection inaccurate.

Methods

The team studied Paramecium bursaria chlorella virus 1 (PBCV-1) using several advanced cryo-EM strategies:

  • Symmetry Relaxation: Instead of assuming 60-fold (icosahedral) symmetry, they used five-fold (C5) averaging. This allowed them to distinguish the "unique vertex" from the other eleven "common" vertices.
  • Block-Based Reconstruction: To handle the virus's massive size and the resulting geometric distortions (Ewald sphere effect), they reconstructed the capsid in smaller "blocks" rather than as one giant sphere, significantly improving local resolution.
  • De Novo Modeling & AI: They identified 30 different types of proteins within the structure. For previously unresolved proteins, they used AlphaFold2 and RoseTTAFold to generate structural templates, which were then fitted into the cryo-EM density maps.

Datasets

The primary data consisted of:

  • Cryo-EM Movie Stacks: Approximately 16,200 movie stacks collected on a Titan Krios microscope.
  • Particle Data: A final set of ~56,500 viral particles used for the reconstruction.
  • Deposited Models: The resulting structures were deposited in the Electron Microscopy Data Bank (EMDB-34439, EMD-34438) and the Protein Data Bank (PDB 8H2I).

Key Insights and Ideas

  • The Myth of the Uniform Capsid: We used to think the outer shell was made of hundreds of identical Major Capsid Proteins (MCPs). This paper reveals five MCP variants (MCPv1–v5) and a penton protein variant (P1v1).
  • Molecular "Microenvironments": These protein variants aren't distributed randomly. They are strategically placed to create specific docking sites. For example, certain variants provide the "glue" for the virus's long fibers, while others stabilize the unique vertex.
  • The Unique Vertex Command Center: The unique vertex is a complex machinery. It features a spike (for host attachment) sitting atop a "crown" formed by the P1v1 variant. Below this sits a lipid vesicle supported by a network of minor capsid proteins (P9, P15–P18).
  • The "Finger Protein" (P19): The researchers identified 25 copies of a "finger protein" located below the capsomers. Structural analysis suggests these proteins might function as polysaccharide lyases—enzymes that help the virus digest the host's tough cell wall during infection.

Key Results

  • Resolution: Achieved a near-atomic resolution of ~3.8 Å for the five-fold averaged structure.
  • Complexity: The atomic model contains 7,135 polypeptide chains comprising 30 different types of proteins, 15 of which were previously unresolved.
  • Vesicle Dimensions: Identified an internal lipid vesicle with a lumen volume of approximately 7.3 × 10⁴ ų, anchored to the unique vertex by transmembrane helices.
  • Assembly Model: The data suggests a "bottom-up" assembly model where the unique vertex (and its associated vesicle) acts as the initiation point for the rest of the icosahedral shell.

Limitations and Future Work

While the resolution of the capsid is excellent, the spike structure itself remains "uninterpretable" in this map. This is likely due to a symmetry mismatch—the spike might have three-fold symmetry while the vertex has five-fold, causing it to blur even with C5 averaging.

Additionally, while the paper proposes how the capsid assembles, it doesn't fully capture the disassembly process—specifically how the internal vesicle and spike coordinate to open the "gate" for DNA release. Future studies using even lower symmetry (C1) or specialized sub-particle refinement will be needed to catch the spike in high definition.

Summary updated

Abstract

Giant viruses are a large group of viruses that infect many eukaryotes. Although components that do not obey the overall icosahedral symmetry of their capsids have been observed and found to play critical roles in the viral life cycles, identities and high-resolution structures of these components remain unknown. Here, by determining a near-atomic-resolution, five-fold averaged structure of Paramecium bursaria chlorella virus 1, we unexpectedly found the viral capsid possesses up to five major capsid protein variants and a penton protein variant. These variants create varied capsid microenvironments for the associations of fibers, a vesicle, and previously unresolved minor capsid proteins. Our structure reveals the identities and atomic models of the capsid components that do not obey the overall icosahedral symmetry and leads to a model for how these components are assembled and initiate capsid assembly, and this model might be applicable to many other giant viruses.
BibTeX
@inproceedings{shao2022near,
  author = {Qianqian Shao and Irina V. Agarkova and Eric A. Noel and David D. Dunigan and Yunshu Liu and Aohan Wang and Mingcheng Guo and Linlin Xie and Xinyue Zhao and Michael G. Rossmann and James L. Van Etten and Thomas Klose and Qianglin Fang},
  title = {{Near-atomic, non-icosahedrally averaged structure of giant virus Paramecium bursaria chlorella virus 1}},
  booktitle = {Nature Communications},
  year = {2022},
  doi = {10.1038/s41467-022-34218-4},
}
Figure