A novel stabilization mechanism accommodating genome length variation in evolutionarily related viral capsids

Tailed bacteriophages are one of the most numerous and diverse group of viruses. They store their genome at quasi-crystalline densities in capsids built from multiple copies of proteins adopting the HK97-fold. The high density of the genome exerts an internal pressure, requiring a maturation process that reinforces their capsids. However, it is unclear how capsid stabilization strategies have adapted to accommodate the evolution of larger genomes in this virus group. Here we characterized a novel capsid reinforcement mechanism in two evolutionary-related actinobacteriophages that modifies the length of a stabilization protein to accommodate a larger genome while maintaining the same capsid size. We used cryo-EM to reveal that capsids contained split hexamers of HK97-fold proteins with a stabilization protein in the chasm. The observation of split hexamers in mature capsids was unprecedented, so we rationalized this result mathematically, discovering that icosahedral capsids can be formed by all split or skewed hexamers as long as their T-number is not a multiple of three. Our results suggest that analogous stabilization mechanisms can be present in other icosahedral capsids, and they provide a strategy for engineering capsids accommodating larger DNA cargoes as gene delivery systems.


Figure S1 .
Figure S1.Genomes of Patience and Adjutor.A) shows the full genome of Adjutor and Patience.Orange/yellow lines between the two genomes shows genes that have nucleotide sequence identity.The length of the genomes is shown to the right.B) shows the zoomed in area of the structural genes.Certain genes are annotated.The number above each gene is the pham number.The number in brackets is the number of other phages with the same pham protein.For example, the major capsid protein is Pham 8871 with 98 other bacteriophages containing the same Pham protein.Genome cartoon was created with Phamerator 1 .

Figure S2 .
Figure S2.Unmodelled density in Patience.The areas in grey are unmodelled density observed on the underside of the capsid near the local three-fold (black triangle) axis.Blue shows major capsid protein while pink shows gp4.The helix can be observed on the right hand side.

Figure S3 .
Figure S3.Free Patience gp4 exists as a random coil.A CD spectra of free Patience gp4 at 1 mg/mL.

Figure S4 .
Figure S4.Alphafold predicted structures of putative gp4 homologs.Patience and Adjutor cryo-EM models and their predictions are shown for comparison.

Figure S5 .
Figure S5.Peg-in-hole interaction of Patience gp4.Shows the ribbon diagram of the Patience hexamer capsomer from the outside looking into the capsid.This is in flipped 180o so that the underside of the ribbon diagram of the hexamer is seen.The surface is then shown and zoomed in to highlight the peg-in-hole interaction.Finally, the ribbon diagram of the peg-in-hole is shown with Ala330 labelled as the center of the major capsid protein peg.The hydrogen bonds and salt bridges are revealed.

Figure S6 .
Figure S6.Straightening of the E-loop of Chain A to bridge the 2-fold chasm.Shows the overlay of chains A and B of the hexamer capsomer (A) as well as the two chains separately with Arg370 and Glu78 highlighted (B and C).

Figure S8 .
Figure S8.Metal co-ordination in Patience and Adjutor.The cryo-EM derived models are shown on the left with the putative metal binding domain highlighted with a black box.On the right the putative metal binding domain is shown closer up with the cryo-EM map overlaid (grey color).Predicted contacts are represented by dashed lines and residues potentially involved in the metal coordination are numbered.

Figure S9 .
Figure S9.Predicted structures of all the major capsid proteins found in the Patience-like major capsid protein family.All structures were predicted with Alphafold.The N-terminus was truncated to where it crosses behind the spine helix due to the poor prediction.

Figure S10 .
Figure S10.Metal co-ordination in the Patience-like major capsid protein.All structures shown were predicted with Alphafold.No metal ion has been modeleld in.Every major capsid protein of the Patience-like major capsid protein family shows similar metal co-ordination.In Onyinye, the five amino acids involved in the co-ordination are labelled.The first row all show similar co-ordination and have the same amino acids.The second row highlights the major capsid proteins that have different amino acids in positions 4 and 5.The final row shows Patience and Beckerton that lack the position 5 amino acid.

Table S1 .
Cryo-EM collection parameters, analysis, and final resolutions.

Table S2 .
Members of the Patience-like major capsid protein family of Patience and Adjutor.Gp4 homologs are all putative and not confirmed by cryo-EM (apart from Patience and Adjutor that have been identified in this paper).

Table S3 .
Interactions between Adjutor gp4 and hexamer major capsid proteins.Interactions predicted with the PDBsum server.

Table S4 .
Interactions between Patience gp4 and hexamer major capsid proteins.Interactions predicted with the PDBsum server.

Table S5 .
Metal Ion-Binding site prediction scores of the putative metal binding sites in the Patience-like family of major capsid proteins.The Alphafold prediction pdb files were used with the online server.