Interactions of HIV-1 Capsid with Host Factors and Their Implications for Developing Novel Therapeutics

Abstract

The Human Immunodeficiency Virus type 1 (HIV-1) virion contains a conical shell, termed capsid, encasing the viral RNA genome. After cellular entry of the virion, the capsid is released and ensures the protection and delivery of the HIV-1 genome to the host nucleus for integration. The capsid relies on many virus-host factor interactions which are regulated spatiotemporally throughout the course of infection. In this paper, we will review the current understanding of the highly dynamic HIV-1 capsid-host interplay during the early stages of viral replication, namely intracellular capsid trafficking after viral fusion, nuclear import, uncoating, and integration of the viral genome into host chromatin. Conventional anti-retroviral therapies primarily target HIV-1 enzymes. Insights of capsid structure have resulted in a first-in-class, long-acting capsid-targeting inhibitor, GS-6207 (Lenacapavir). This inhibitor binds at the interface between capsid protein subunits, a site known to bind host factors, interferes with capsid nuclear import, HIV particle assembly, and ordered assembly. Our review will highlight capsid structure, the host factors that interact with capsid, and high-throughput screening techniques, specifically genomic and proteomic approaches, that have been and can be used to identify host factors that interact with capsid. Better structural and mechanistic insights into the capsid-host factor interactions will significantly inform the understanding of HIV-1 pathogenesis and the development of capsid-centric antiretroviral therapeutics.

Keywords: HIV-1 capsid; antiretroviral therapeutics; capsid-targeting inhibitors; high-throughput screening techniques; host factors.

Figure 1

Schematic overview of the early stages of Human Immunodeficiency Virus type 1 (HIV-1) replication, which is spatiotemporally regulated by diverse capsid-interacting host factors. The HIV-1 virion binds to the CD4 receptor and the CXCR4 and/or CCR5 coreceptors via the viral envelope glycoproteins, resulting in fusion with the target cell, which releases capsid into the cellular cytoplasm. The capsid traffics towards the nucleus along the microtubule network by employing opposing adaptor and motor proteins such as FEZ1, kinesin, BICD2, dynein, and MAP1, a process that may result in a bi-directional “tug-of-war”. The host cellular protein CypA and metabolite IP6 bind to capsid and maintain core stability during cytoplasmic trafficking. The host innate immune system can respond to capsid, resulting in interferon (IFN) expression and inducing the IFN-induced viral restriction factors TRIM5α and Mx2. TRIM5α binds to the capsid causing premature uncoating, while Mx2 binds to the capsid impeding the nuclear import. When capsid arrives at the nuclear envelope, it presumably binds Nup358, and other factors, to promote import into the nucleus via the nuclear pore complex as an intact or nearly intact capsid. The host factors CPSF6, Nup153, and TRN-1 also appear to facilitate capsid nuclear import via direct capsid interactions. Reverse transcription is completed in the nucleus, followed by complete uncoating of the capsid and integration of the viral DNA into the host genome. All processes are detailed in the text.

Figure 2

HIV-1 capsid architecture and summary of binding interfaces of capsid with diverse host factors. During the maturation of HIV-1 virion, ~1500 CA monomers assemble into ~250 hexamers and exactly 12 pentamers in alignment produce fullerene cone geometry, forming capsid (Figure 2A–D). The interfaces within the capsid which bind diverse host factors are highlighted in different colors, as detailed in the Figure 2E legend. (A) Structure of the CA monomer (PDB ID: 4XZF). The left structure is shown in surface representation and the right structure as ribbon representation. CA consists of two α-helical domains, an N-terminal domain (NTD) labeled in gold and a C-terminal domain (CTD) labeled in light blue, connected by a flexible linker (grey). All α-helices (H1-H11) and key structural elements are indicated by arrows with names. (B) Structure of a CA hexamer with top view (left) and side view (right) (PDB ID: 4XZF). (C). Structure of high-order (hexagonal) CA hexamers with top view (left) and side view (right) (PDB ID: 4XZF). (D). Architecture of the capsid (PDB ID: 3J3Y). The incorporation of pentamers (orange) at either end of capsid provides the curvature necessary to close the conical structure. (E) The interfaces within capsid that can bind diverse host factors are highlighted in different colors, which are shown in Figure 2A–D. Host proteins CypA, Nup358, and TRN-1 bind to the flexible CypA-binding loop (labeled in red) exposed on the top of CA-NTD that protrudes from the capsid outer surface. Host proteins CPSF6 and Nup153 share the same phenylalanine-glycine binding pocket (labeled in green) created by the NTD-CTD interface between two neighboring CA monomers in a hexamer. The restriction factor Mx2 specifically binds to the three-fold inter-hexamer interface (labeled in purple) of capsid; moreover, it is unable to bind to a CA monomer or a single hexamer. Host protein FEZ1 and metabolite IP6 can bind to the positively charged central pore of hexamers formed by R18 (labeled in dark blue) of CA. TRIM5α can form a hexagonal network (labeled in pink, shown in Figure 2D, also called TRIM5α cage) that avidly binds the capsid shell. The extent to which the TRIM5α cage can cover the capsid and how TRIM5α directly contacts the capsid surface have not been fully established.

Figure 3

HIV-1 capsid inhibitors and the comparison of their binding sites on capsid with host proteins. (A) Chemical structure of PF74. PF74 contains a polyphenyl core (red), a linker region (green), and an indole ring (dark blue). (B) Chemical structure of GS-CA1. (C) Chemical structure of GS-6207. Gilead GS-CA compounds possess a similar polyphenyl core (red) and linker region (green) with PF74, and a cyclopenta-pyrazole ring (light blue), as well as other chemical groups (black). GS-6207 differs from GS-CA1 by three modifications (magenta arrows), difluoroethyl groups on indazole ring are replaced by a trifluoroethyl group, a cyclopropane moiety on sulfonamide group is replaced by a methyl group, and a difluoromethyl group on cyclopenta-pyrazole ring is replaced by a trifluoromethyl moiety. (D) Comparison of the binding sites of capsid inhibitors with host proteins. The capsid inhibitors GS-6207 and PF74 bind to the same pocket occupied by host proteins CPSF6 and Nup153 within the NTD-CTD inter-subunit interface created by two adjacent CA monomers in a hexamer. The structural comparison shows that the difluorobenzyl ring (dark blue) of GS-6207 and the phenyl ring (purple) of PF74 superpose well on the F321 (red) of CPSF6 and F1417 (orange) of Nup153.

Figure 4

The workflow of yeast two-hybrid (Y2H) screen employed to identify virus–host interactions. The principle of a Y2H system is based on the reconstitution of a functional transcription factor driven by the interaction between a bait protein and a prey protein. The DNA-binding domain (DB) of the transcription factor is fused to a viral protein of interest (bait, in green), while the transcription activation domain (AD) of the transcription factor is fused to a host protein (prey, in yellow) coming from human cDNA libraries. Upon co-expression of the bait and prey fusions in yeast cells, if the bait and prey interact, DB and AD will be reconstituted (indicated in red), and thus activate the transcription of a reporter gene (top lane), whose expression can cause a visible color change on the selective plate. These positive clones can be isolated for sequencing to determine the prey proteins. If the bait and prey do not interact, DB and AD will remain separated, and transcription of the reporter gene does not occur (bottom lane).

Figure 5

The workflows for AP-MS combined with SILAC or chemical cross-linking employed to identify virus–host interactions. (A) The workflow of AP-MS. As shown in the top lane, a typical AP-MS experiment begins with the fusion of an affinity tag (red) into a viral protein of interest (bait, in green). The tagged bait protein is then introduced into the host cells by transfection. After a period of expression, the tag-targeting antibody (bronze) conjugated to a resin (light grey) is used to pull down the bait protein, along with any host proteins (preys) bound to it. The resin should be washed several times to remove non-specific binding proteins (black). The captured prey host proteins (yellow, brown, and grey) are subsequently eluted from the resin, digested, and analyzed by Liquid Chromatography-Tandem Mass Spectrometry (LC-MS/MS). In parallel, a control experiment is set up by only introducing the affinity tag into the host cells, followed by the exact same steps performed in the sample experiment, as shown in the bottom lane. By comparing the number of identified MS/MS spectra of the same protein from sample or control cells, the viral protein-specific interactors (yellow and brown) can be distinguished from the non-specific binding proteins (grey) attached to the resin. (B) The workflow of AP-MS combined with SILAC. Sample cells labeled with heavy isotopes (red) are transfected with tagged bait plasmid, whereas control cells labeled with light isotopes (blue) are transfected with tagged empty plasmid, followed by parallel affinity purification, washing, and elution. The eluted prey proteins from different host cells are mixed together, digested, then analyzed and quantified by LC-MS/MS. Each captured host protein owns a heavy/light ratio indicating its specificity of interaction with the bait protein. SILAC-based quantification is more accurate than spectral count-based quantification. (C) The workflow of AP-MS combined with chemical cross-linking. After the transfection and expression of tagged bait protein or tag in host cells, cross-linkers (red double ended arrows) are added into cells to preserve the weak and transient interactions (pink and blue) between bait and prey proteins, then followed by AP-MS, as detailed above.

Figure 6

The workflow of Proximity-Dependent Labeling (PDL) technology followed by MS to identify virus–host interactions. The viral protein of interest (bait, in green) is fused to an engineered biotin ligase, BirA*, or ascorbate peroxidase enzyme, APEX (indicated by red square), and expressed in host cells, whereas only bait is expressed in the cells as control. The bait-enzyme when supplied with its appropriate substrates generates reactive intermediates which then covalently label all proximal host proteins (prey) within a radius of ~10–20 nm with biotin (indicated by red balls). The biotinylated proteins (prey) are isolated from host cells and enriched by streptavidin-conjugated beads, proteins are then eluted from the beads, digested, and then identified by LC-MS/MS. The PDL approach can label any host proteins that are within the labeling radius of the viral protein-enzyme, whether directly (strong, weak, or transient interactions) or not directly associated with the viral protein.