Pharmacologic hyperstabilisation of the HIV-1 capsid lattice induces capsid failure

  1. KM Rifat Faysal
  2. James C Walsh
  3. Nadine Renner
  4. Chantal L Márquez
  5. Vaibhav B Shah
  6. Andrew J Tuckwell
  7. Michelle P Christie
  8. Michael W Parker
  9. Stuart G Turville
  10. Greg J Towers
  11. Leo C James
  12. David A Jacques
  13. Till Böcking

  • Curated by eLife

    eLife logo

    eLife assessment

    In this important study, the authors propose that lenacapavir inhibits HIV-1 replication by inducing "lethal hyperstabilization" of the capsid, based on experiments that clearly demonstrate such an effect at high drug concentrations. Data supporting the model are incomplete at low drug concentrations, and a firm correlation between the in vitro effects and therapeutic mechanism of action has not yet been established.

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

The HIV-1 capsid has emerged as a tractable target for antiretroviral therapy. Lenacapavir, developed by Gilead Sciences, is the first capsid-targeting drug approved for medical use. Here, we investigate the effect of lenacapavir on HIV capsid stability and uncoating. We employ a single particle approach that simultaneously measures capsid content release and lattice persistence. We demonstrate that lenacapavir’s potent antiviral activity is predominantly due to lethal hyperstabilisation of the capsid lattice and resultant loss of compartmentalisation. This study highlights that disrupting capsid metastability is a powerful strategy for the development of novel antivirals.

Article activity feed

  1. Version published to 10.7554/elife.83605 on eLife
  2. eLife

    eLife assessment

    In this important study, the authors propose that lenacapavir inhibits HIV-1 replication by inducing "lethal hyperstabilization" of the capsid, based on experiments that clearly demonstrate such an effect at high drug concentrations. Data supporting the model are incomplete at low drug concentrations, and a firm correlation between the in vitro effects and therapeutic mechanism of action has not yet been established.

  3. eLife

    Reviewer #1 (Public Review):

    This is a review of the manuscript entitled "Pharmacologic hyperstabilisation of the HIV-1 capsid lattice induces capsid failure" by Faysal et al., in this manuscript the authors used an elegant single virion fluorescence assay based on TIRF to measure the stability of mature HIV cores. Virions were biotinylated and captured onto glass coverslips through specific Biotin-Avidin interactions. Immobilized virions were then introduced to the imaging buffer which contained the pore-forming protein DLY, and fluorescently labeled CypA. Mature virions were identified through the binding of CypA which had a red fluorescent tag allowing them to measure the dynamics of GFP trapped within the mature cores as well as the CypA bound outside the core. The authors show that the addition of LEN starting from about 50nM stabilized the mature cores even after cores have ruptured and released their internal GFP. Higher concentration of Len results in ultrastabilization of the cores and rapid rupture leading to the release of GFP at an earlier timepoint. A biochemical assembly assay was performed which showed uM quantities of Len synergized with IP6 to promote CA assembly. Purified mature virions were also treated with 700nM of Len and analyzed by CryoET, this analysis showed an increased representation of irregular cores within the Len-treated sample. Putting all of this together, the authors concluded that Len facilitates core rupture through hyperstabilization of HIV cores, as described in the title.

    While I have found this work technically well performed and well explained, I do not believe that the presented data supports the conclusions reached by the authors.

  4. eLife

    Reviewer #2 (Public Review):

    The authors set out to study the potent HIV capsid inhibitor lenacapavir (LEN) and how it alters capsid stability. They use a previously developed single-molecule fluorescence imaging assay to take two measurements of individual viral particles over time: 1) they track the release of GFP from GFP-loaded particles to determine whether the capsid is intact or open, and 2) they track the disassembly of the capsid lattice by measuring the signal intensity of a capsid binding fluorophore (AF568-CypA), which diminishes as the capsid lattice subunits disassociate.

    As in their previous work, the authors report that most of their capsids are "leaky" and rapidly lose GFP after the viral membrane is permeabilized, followed by disassembly of the capsid lattice. A subset of capsids maintain GFP signal for various periods of time until they spontaneously "open," and a smaller subset remains closed for the entire length of the imaging experiment (typically 30 min). Interestingly, the authors find that LEN has two effects in this assay: it not only promotes a more rapid release of GFP (interpreted to mean loss of capsid integrity), but it also prevents the capsid lattice from disassembling after opening. As expected, the cellular cofactor IP6 (which stabilizes capsids in cells and in vitro) was found to protect against capsid rupture and counteracted the effects of LEN (although high concentrations of LEN could override any protective effects of IP6).

    Their single-molecule experiments are nicely buttressed by in vitro assembly reactions of purified CA protein, with IP6 promoting cone formation and LEN promoting aberrant assembly into tubes. The authors go further to test the kinetics of LEN's effects on HIV infection and reverse transcription, and they perform experiments in comparison to other factors that target the FG binding pocket (BI-2, PF-74, and a peptide from the host factor CPSF6). They find that LEN works differently than these other capsid binders, and stabilizes the lattice structure much more effectively, which the authors suggest is due to how well LEN bridges between CA-CA monomers and rigidifies CA hexamers.

    It's particularly interesting that the results of their kinetic studies indicate that LEN's effects on capsid strain (which may ultimately promote rupture) may not happen immediately, but instead, take time to build as the drug occupies more and more binding sites. The authors estimate that roughly 30% of binding sites need to be occupied by LEN to reach half-maximal inhibition of infection, and based on their binding curves, it may take ~20h to reach this level of occupancy in the presence of sub nM concentrations of LEN. Although other mechanisms in addition to catastrophic rupture of capsids are likely at play during inhibition of infection (such as inhibition of host factor binding), these kinetics support previous reports that the most potent functions of capsid inhibition occur at or between the steps of nuclear entry and integration.

    It is important to note that although in vitro uncoating assays can help us understand the physical nature of HIV capsid and capsid inhibitor interactions, the assays in this paper might not accurately model the capsid dynamics that are experienced in a cell during infection. The authors report that more than half of their capsids are "leaky" at the start of their assay, but this could be an artifact of the experimental system. Several groups have now demonstrated that capsids remain intact or largely intact for several hours after infection. Thus, while their method is valuable to the research community and can provide insight into capsid stability (and how it can be influenced by capsid binding factors), the authors should be cautious about using pore-forming proteins to permeabilize the virion and interpreting the release of GFP in their single-molecule fluorescence system as an accurate reflection of HIV dynamics in vivo.

    In this regard, it would be helpful to establish whether the pore-forming proteins used in vitro to permeabilize the virus membrane have an impact on capsid integrity. It's possible that the concentration of pore-forming proteins used in this paper (200nM) actually promotes "leaky" capsids and rapid opening of capsids in vitro, whereas capsids in their native state in the cytoplasm could remain mostly intact until disrupted by host factors and/or small molecules. Determining whether lower concentrations of DLY/SLO (or PFO as used in Marquez et al., 2018) change the ratio of leaky to closed capsids, or delay the time to capsid opening (either in the presence of IP6 or in the presence of LEN) would be informative. It may be possible to optimize the concentration of pore-forming proteins (and other buffer constituents) to achieve permeabilization of the membrane with minimal disruption to capsid integrity, which could approximate conditions within the cell.

    Experiments with capsid mutations that stabilize or destabilize the lattice structure (and exhibit different sensitivities to IP6) could help support the authors' conclusions, as would testing mutations that confer resistance to LEN (e.g. Q67H+N74D, M66I, etc...). It would be of great interest to find if CA mutations affect either GFP release or the CypA paint signal, and whether resistance mutations mitigate the effects of LEN in single-molecule experiments.

    The discussion section of this paper is expertly written and places the work into the larger context of HIV research. The authors have thoughtfully analyzed their experiments with capsid inhibitors in relation to kinetics, occupancy, the potential for rigidification, and cofactor binding. They offer reasonable explanations for how LEN exhibits opposing effects on the HIV capsid at high occupancy through inducing capsid rupture while simultaneously preventing the dissociation of CA subunits. Many lines of evidence are now converging on the concept that the capsid evolved to be stable enough to protect its contents, yet flexible enough to navigate the steps of reverse transcription, nuclear entry, and uncoating. With this paper, the authors make a strong case that LEN functions as an antiviral, at least in part, through engaging "lethal hyperstabilization" of the capsid, promoting rigid lattice formations that are incompatible with closed cone structures.

  5. eLife

    Reviewer #3 (Public Review):

    In this article, Faisal et. al., use a combinatorial approach to look at the mechanisms of HIV-capsid inhibition by the highly potent drug Lenacepavir (LEN). The authors conclude that LEN induces capsid opening, but hyper-stabilizes the remaining capsid lattice during the early stages, and adversely affects the assembly of capsids during late stages of HIV-1 infection. Additionally, they suggest that hyper-stabilization effects of LEN on the capsid-lattice are induced by a low occupancy of this highly potent drug, while less potent inhibitors like PF74 need high occupancy on the lattice to induce similar effects. Taken together their findings shine a light on the importance of the capsid binding pocket targeted by multiple inhibitors including LEN, PF74, BI-2, and host-factor CPSF6 on overall capsid assembly, its stability in cells, and its role in HIV-1 infection.

    Strengths:
    1. Combinatorial approach using single-molecule imaging, cryoET and cellular assays show the adverse effects of LEN on HIV-1 capsid assembly, capsid disassembly, and block to virus infectivity.
    2. Several novel insights are obtained in this paper, including the cryoET-data showing 2-layers of capsid formation in the presence of LEN. CPSF6-peptide binding to capsids, and their effect on stability.

    Weakness:
    1. Interpretation of the capsid stability data is focused on single virus traces rather than population analysis, which might paint a different picture of the conclusions.
    2. The description and interpretation of the data in the results sections and the conclusions are inconsistent, and somewhat confusingly presented for the general non-expert audience.

  6. Version published to 10.1101/2022.09.21.508807v2 on bioRxiv
  7. Version published to 10.1101/2022.09.21.508807v1 on bioRxiv