Structure of the Capsid Size-Determining Scaffold of "Satellite" Bacteriophage P4

Abstract

P4 is a mobile genetic element (MGE) that can exist as a plasmid or integrated into its Escherichia coli host genome, but becomes packaged into phage particles by a helper bacteriophage, such as P2. P4 is the original example of what we have termed "molecular piracy", the process by which one MGE usurps the life cycle of another for its own propagation. The P2 helper provides most of the structural gene products for assembly of the P4 virion. However, when P4 is mobilized by P2, the resulting capsids are smaller than those normally formed by P2 alone. The P4-encoded protein responsible for this size change is called Sid, which forms an external scaffolding cage around the P4 procapsids. We have determined the high-resolution structure of P4 procapsids, allowing us to build an atomic model for Sid as well as the gpN capsid protein. Sixty copies of Sid form an intertwined dodecahedral cage around the T = 4 procapsid, making contact with only one out of the four symmetrically non-equivalent copies of gpN. Our structure provides a basis for understanding the sir mutants in gpN that prevent small capsid formation, as well as the nms "super-sid" mutations that counteract the effect of the sir mutations, and suggests a model for capsid size redirection by Sid.

Keywords: Caudovirales; Psu; Sid; bacteriophage P2; capsid assembly; mobile genetic elements; molecular piracy; sir mutants; size determination.

Figure 1

(A) SDS-PAGE from sucrose gradient purification of P4 procapsids. The lanes are labeled by fraction number, and the arrow shows the direction of sedimentation. Positions of bands corresponding to gpN, gpO and Sid are indicated. M = marker, with pertinent molecular weights (kDa) listed. (B) Cryo-electron micrograph of P4 procapsids. The arrows point to additional proteinaceous material assumed to be Sid. Scale bar, 100 nm. (C) Fourier Shell Correlation (FSC) plots for the icosahedral (blue) and asymmetric (red) reconstructions calculated in RELION, for the asymmetric map calculated in Phenix (green) and the correspondence between the atomic model and the reconstruction (yellow), calculated in Phenix. The gray lines show the 0.143 and 0.5 cutoff levels.

Figure 2

Reconstruction of the P4 procapsid. (A) Isosurface representation of the icosahedrally symmetric reconstruction, colored radially from the inside (red) to the outside (blue). The triangle represents one icosahedral face (three asymmetric units); the icosahedral symmetry axes are indicated. The schematic diagram shows the arrangement of gpN subunits in the T = 4 lattice. (B) Isosurface representation of the asymmetric reconstruction, viewed as in (A). (C) The atomic model, showing two complete icosahedral asymmetric units, including 8 copies of gpN (A–D and A2–D2) and two copies of Sid. One of the asymmetric units is shown with lighter color. The inset shows the orientation of the model within the capsid. (D) Ribbon diagram showing the entire dodecahedral cage formed by Sid.

Figure 3

The gpN capsid protein. (A) Ribbon diagram of gpN_B (the most complete subunit), colored by structural feature: N-arm, red; E-loop, pink; P-domain β-sheet, blue; P-domain α-helices, yellow (α3 is the spine helix); A-domain, green. The right-hand panel is rotated by 180°. Key structural features are labeled, and the gpO cleavage site between residues 31 and 32 is indicated (circle). (B) Intra-capsomer interaction between the gpN_B (yellow) and gpN_D (blue) subunits, showing how the gpN_B E-loop overlays the spine helix of the neighboring (gpN_D) subunit. The right-hand panel is rotated by 180° and shows how the gpN_B N-arm and α0 grasp the P-domain of gpN_D from underneath. (C) Electrostatic surfaces for the gpN_B and gpN_D subunits from panel B, separated and rotated by 120° in opposite directions, as indicated by the insets. Interaction surfaces of opposite charge are indicated by the dashed lines. (D) Inter-capsomeric interaction between three gpN_D and their neighboring gpN_B subunits related by the icosahedral threefold axis. The insets show details of the interactions between the gpN_D and gpN_B subunits.

Figure 4

The Sid scaffold. (A) Ribbon diagram for one monomer of Sid. Relevant structural features are labeled. (B) Ribbon diagram of the Sid dimer, viewed from the side, perpendicular to the twofold axis (left) and from the top, looking down the twofold axis (right). Inset: detail of the twofold interaction that includes the disulfide bond (C140). The map is shown as a transparent gray isosurface. (C) The Psu dimer (PDB ID: 3RX6) [45], viewed perpendicular to (left) and down (right) the twofold axis. (D) Ribbon diagram of the Sid trimer, viewed down the icosahedral threefold axis. The insets on the right show details of the threefold interaction. Two Sid subunits are shown as electrostatic surfaces, with the third subunit as a ribbon diagram. The left panel shows the view from the outside of the capsid, while the right panel is viewed from the inside. The surfaces that interact are indicated by the dashed squares.

Figure 5

Sid-gpN interaction. (A) Ribbon diagram of the Sid dimer (purple) and the underlying gpN hexamer; subunits colored as in Figure 2. In the right panel, only gpN_B is depicted. (B) Detail of the interaction between the Sid α3–α4 loop and gpN_B. Residues that make contact are shown in stick representation, and contacts closer than 4 Å are shown as gray lines. (C) Interactions between the C-terminal helix α6 of Sid and the gpN_B sir loop. (D) Same view as (C). The sir loop is shown in orange, and the residues mutated in sir and nms mutations are labeled in. (E) Same view as (D), showing how nms mutations (E215G, Q227R, G234R) might facilitate novel charged interactions with gpN (residues D189, E190, E191).

Figure 6

Model for the Sid-induced capsid redirection. Initially, Sid dimers interact with gpN hexamers, forming a Sid-gpN complex (top). Once Sid trimerizes and forms a Sid-gpN trimer complex (middle), continued growth of the shell is only compatible with formation of a T = 4 lattice (left). If the Sid-gpN complexes were to assemble into a T = 7 lattice (right), trimerization of Sid would be impossible, leading to a highly unfavorable configuration of Sid.