Crystal structure of a four-layer aggregate of engineered TMV CP implies the importance of terminal residues for oligomer assembly

Abstract

Background: Crystal structures of the tobacco mosaic virus (TMV) coat protein (CP) in its helical and disk conformations have previously been determined at the atomic level. For the helical structure, interactions of proteins and nucleic acids in the main chains were clearly observed; however, the conformation of residues at the C-terminus was flexible and disordered. For the four-layer aggregate disk structure, interactions of the main chain residues could only be observed through water-mediated hydrogen bonding with protein residues. In this study, the effects of the C-terminal peptides on the interactions of TMV CP were investigated by crystal structure determination.

Methodology/principal findings: The crystal structure of a genetically engineered TMV CP was resolved at 3.06 Å. For the genetically engineered TMV CP, a six-histidine (His) tag was introduced at the N-terminus, and the C-terminal residues 155 to 158 were truncated (N-His-TMV CP(19)). Overall, N-His-TMV CP(19) protein self-assembled into the four-layer aggregate form. The conformations of residues Gln36, Thr59, Asp115 and Arg134 were carefully analyzed in the high radius and low radius regions of N-His-TMV CP(19), which were found to be significantly different from those observed previously for the helical and four-layer aggregate forms. In addition, the aggregation of the N-His-TMV CP(19) layers was found to primarily be mediated through direct hydrogen-bonding. Notably, this engineered protein also can package RNA effectively and assemble into an infectious virus particle.

Conclusion: The terminal sequence of amino acids influences the conformation and interactions of the four-layer aggregate. Direct protein-protein interactions are observed in the major overlap region when residues Gly155 to Thr158 at the C-terminus are truncated. This engineered TMV CP is reassembled by direct protein-protein interaction and maintains the normal function of the four-layer aggregate of TMV CP in the presence of RNA.

Figure 1. Crystal structure and function of the four-layer aggregate disk of N-His-TMV CP¹⁹.

(A) Electrostatic surface presentation of the N-His-TMV CP¹⁹ four-layer aggregate of ∼95 Å in height. The electrostatic surface was calculated in pyMOL. The complete four-layer aggregate is shown in four rings in the following order: b-ring, a-ring, a-ring, b-ring. (B) Electrostatic surface presentation of the N-His-TMV CP¹⁹ four-layer aggregate disk of ∼176 Å in diameter. (C) Electrostatic surface presentation of the ba-ring and aa-ring pairs. The aa-ring pair is sandwiched between the two b-rings. The a-ring and a-ring chains are anti-parallel and pack more tightly to ∼45 Å in height; the b-ring and a-ring chains are parallel and pack to ∼50 Å in height. (D) The slewed (radial) disk of the b-ring defines a 64 Å diameter axial pore based on residue Val 114 (bottom panel); (E) the slewed (radial) disk of the a-ring defines a 46 Å diameter axial pore based on residue Thr107 (bottom panel). (F) Assembly of the N-His-TMV CP¹⁹ four-layer aggregate, as measured by SEC and native-PAGE electrophoresis. (G) TMV RNA integrity was examined by 1% agarose gel electrophoresis. (H) The N. glutinosa was inoculated with WT TMV and reconstituted TMV. The local lesions that were inoculated with reconstituted TMV are labeled with red arrows.

Figure 2. Characterization of the disk and helical stack height using TEM in 10 mM PB, 100 mM NaCl, pH 7.2 with increasing concentrations of protein, temperature and time.

As concentration, temperature and time increased, the dominant structure changed from disks to rods. (A) 1.8 mg/mL N-His-TMV CP¹⁹, 4°C, 10 h: monomers, dimers and disks. (B) 1.8 mg/mL N-His-TMV CP¹⁹, 20°C, 15 h: disks. (C) 1.8 mg/mL N-His-TMV CP¹⁹, 20°C, 20 h: disks. (D) 14 mg/mL N-His-TMV CP¹⁹, 20°C, 10 h: disks and rods. (E) 14 mg/mL N-His-TMV CP¹⁹, 20°C, 15 h: disks and rods. (F) 14 mg/mL N-His-TMV CP¹⁹, 20°C, 20 h: rods.

Figure 3. Characterization of reconstituted N-His-TMV CP¹⁹ virus using TEM in 10 mM PB, 100 mM NaCl, pH 7.2 after 20 h at 22°C.

1 mL purified self-assembled N-His-TMV CP¹⁹ disks (1.8 mg/mL) in 10 mM PB, 100 mM NaCl, pH 7.2 was mixed with 0.2 mL purified TMV RNA for 20 h at 22°C. Four TEM images of the reconstituted virus are shown in A–D.

Figure 4. N-His-TMV CP¹⁹ interactions involved in b-ring and a-ring inter-subunits.

(A) Inter-chain interactions between Asn25-Ser15, Tyr72-Thr28, Phe35-Asp88, Gln36-Asp88 and Arg122-Asp88 in slewed disks of the b-ring were observed. These interactions are different from the previously reported disk inter-chain interactions (mediated by water). (B) Because the aa-ring is sandwiched between the b-rings, the a-ring inter-subunits in the LR region loop residues have stronger interactions than the b-ring inter-subunits, except for the observed inter-chain interactions between Asn25-Ser15, Tyr72-Thr28, Phe35-Asp88, Gln36-Asp88 and Arg122-Asp88.

Figure 5. Side view of the major overlap region between layers of atomic chains.

(A) The Asp19-Arg134 and Asp66-Arg134 protein–protein interactions between the b-pair and a-pair of atomic chains; the Lys53-Glu22 and Thr59-Thr59 protein–protein interactions between the a-pair and a-pair of atomic chains. (B) The N-His-TMV CP¹⁹ four-layer aggregate disk interaction model involves inter-subunits between layers. The horizontal dotted line is the central axis of the aa-pair. A particular slide range is observed because of the Thr59-Thr59 interaction.

Figure 6. Superposition of two monomers of N-His-TMV CP¹⁹.

(A) Superposition of monomers between the a₁-chain and a₂-chain in N-His-TMV CP¹⁹. (B) Superposition of monomers between the b₂-chain and a₂-chain in N-His-TMV CP¹⁹. (C) Superposition of monomers between the b₁-chain and b₂-chain in N-His-TMV CP¹⁹.

Figure 7. The inter-chain and intra-chain interaction model of N-His-TMV CP¹⁹ and the previously reported TMV CP disk (PDB code: 1EI7).

(A) The inter-chain and intra-chain interaction model of N-His-TMV CP¹⁹. (B) The inter-chain and intra-chain interaction model of the previously reported TMV CP disk structure (PDB code: 1EI7).

Figure 8. Structure comparisons between the previously reported TMV CP structure (PDB code 1EI7) and N-His-TMV CP¹⁹; and virus structure (PDB code 3JO6) and N-His-TMV CP¹⁹.

(A) The B factors of the atomic model are based on the atomic structure. Compared with TMV CP (PDB code 1EI7) and TMV CP (PDB code 3JO6), the N-His-TMV CP¹⁹ residues at the C-terminus loop (residues 148–154) have a low atomic B factor; the N-His-TMV CP¹⁹ in the low-radius inner loop [residues 104–111 of the a-ring (residues 107–111 of the b-ring)] have a high atomic B factor, indicating that the residues in these loops are disordered. In this structure, the disordered residues are stabilized by protein–protein hydrogen bonding interactions in the adjacent subunit. (B) Structure comparison between the TMV CP monomer (PDB code 1EI7) and the N-His-TMV CP¹⁹ monomer. The shifting N-His-TMV CP¹⁹ LR cylindrical helix is highlighted by an arrow. (C) Structure comparison between the TMV CP monomer (PDB code 3JO6) and the N-His-TMV CP¹⁹ monomer. The shifting N-His-TMV CP¹⁹ LR cylindrical helix is highlighted by an arrow. (D) Alignment of the side view of the TMV CP monomer (PDB code 1EI7) and the N-His-TMV CP¹⁹ monomer to show the different conformations in the main- and side-chain residues. The key residues in oligomer assembly are highlighted and labeled. (E) Alignment of the side view of the TMV CP monomer (PDB code 3JO6) and the N-His-TMV CP¹⁹ monomer to show the different conformations in the main- and side-chain residues. The key residues in oligomer assembly are highlighted and labeled.