Viral structural proteins as targets for antivirals

Abstract

Viral structural proteins are emerging as effective targets for new antivirals. In a viral lifecycle, the capsid must assemble, disassemble, and respond to host proteins, all at the right time and place. These reactions work within a narrow range of conditions, making them susceptible to small molecule interference. In at least three specific viruses, this approach has had met with preliminary success. In rhinovirus and poliovirus, compounds like pleconaril bind capsid and block RNA release. Bevirimat binds to Gag protein in HIV, inhibiting maturation. In Hepatitis B virus, core protein allosteric modulators (CpAMs) promote spontaneous assembly of capsid protein leading to empty and aberrant particles. Despite the biological diversity between viruses and the chemical diversity between antiviral molecules, we observe common features in these antivirals' mechanisms of action. These approaches work by stabilizing protein-protein interactions.

Figure 1.. Antiviral binding sites on a picornavirus capsid.

(a) A coxsackie virus capsid (1cov), with the viral structural proteins Vp1–3 colored blue, red, and green, respectively; Vp4 is inside the capsid and not visible in this image. Two distinct sites for capsid-directed antiviral compounds are denoted with arrows. (b) The binding site for compound 17 (CP17, pink), from Abdelnabi et al. 2019 (6gzv), is at an interface of a Vp3 and the two adjacent Vp1 molecules (one shown in white). The view is from the outside of the capsid. (c) In contrast, the binding location for pleconaril (magenta), binding the same site as a WIN compound, only contacts a single copy of Vp1 and is not located at an intermolecular interface (1ncr). (d) A surface representation of the same complex as (c) emphasizes that pleconaril is completely enclosed by Vp1.

Figure 2.. The effects of CpAM binding to HBV core protein.

(a) Electron micrographs of assembled HBV Cp dimer. (left panel) Virus-like particles are ~35nm icosahedra. Inset: an model of a T=4 capsid structure with chains colored according to quasi-equivalent environment. When assembled in the presence of a HAP (right panel), reaction products include large misassembled structures (40). (b) A hexamer fragment of a T=4 capsid structure in complex with a HAP (44). Coloring is the same as in the inset in the left panel of a, except with the HAP molecules colored gray. The HAP molecules bind at the interface between two adjacent subunits. (c) One site, multiple chemistries: a selection of CpAMs which target the interface between HBV capsid protein subunits. Despite occupying the same site, each molecule has a distinct binding mode, and associated phenotype (, , ••52) (d) A simplified reaction schematic for how CpAMs disrupt pre-formed capsids. Increasing the per-contact association energy causes favorable, but modified pairwise orientation that induces global strain. Strained capsids will rupture to relieve strain (53). Components in this figure are extracted from or adapted from figures in references (40), (44), and (••52).

Figure 3.. The HIV CA domain with hexamer-stabilizing small molecules.

(a) A ribbon diagram of a top view of a mature CA hexamer with bound P74 (PDB: 4u0e). The P74 molecule (space filling with carbons in gray) fits into a pocket largely formed by the NTD (top) and completed by the CTD from the neighboring subunit (CTDs are partially obscured by the NTD). (b) A CTD-SP1 construct representing a fragment of immature Gag with bound bevirimat (PDB: 6n3u). The CTD is on top. The last residue in the CTD, L363 (black), and SP1 (gray) are partially obscured. Electron density for bevirimat (magenta), contoured at 1.1 σ, fills a cylindrical gap in the ring of helices, presumably stabilizing them and blocking accessibility to protease.