Lenacapavir-induced lattice hyperstabilization is central to HIV-1 capsid failure at the nuclear pore complex and in the cytoplasm

  1. Arpa Hudait
  2. Ryan C Burdick
  3. Ellie K Bare
  4. Vinay K Pathak
  5. Gregory A Voth  Is a corresponding author
  1. Department of Chemistry, Chicago Center for Theoretical Chemistry, Institute for Biophysical Dynamics, and James Franck Institute, The University of Chicago, United States
  2. Viral Mutation Section, HIV Dynamics and Replication Program, Center for Cancer Research National Cancer Institute at Frederick, United States
6 figures and 1 additional file

Figures

Figure 1 with 3 supplements
Overview of the coarse-grained (CG) molecular model of the nuclear pore complex (NPC), human immunodeficiency virus type 1 (HIV-1) capsid, and Lenacapavir (LEN).

(A) The left panel shows the composite membrane-embedded CG model of human NPC. The NPC is shown in gray spheres. The disordered NUP98 chains are shown in red chains. The nuclear membrane is modeled with a four-site CG lipid bilayer model. The CG bead representing the lipid headgroups are shown in pink spheres, interfacial and tail beads are shown in white spheres. In the right panel, the cross-section of the NPC is shown. The cytoplasmic ring (CR) and nuclear ring (NR) are shown in silver to distinguish from the inner ring (IR) (shown in gray spheres). The FG-NUP62 is modeled as a heterotrimeric subcomplex along with NUP54 and NUP58 is shown in cyan spheres (occluded by the FG-NUP98 mesh). The diameter of the central channel is 59 nm. (B) The left panel shows the CG representation of LEN with CG sites 1–8 labeled on the LEN chemical representation. The center panel shows the CG-mapped representation of LEN bound to a CA hexamer. The CG-mapped structure was generated from the X-ray crystal structure PDB: 6VKV. A CG LEN molecule is shown in orange beads. The CG sites of adjoining CA monomers CA1 and CA2 that are in contact with LEN are shown in red and green, respectively. In the right panel, the CG representation of the CA monomer is shown. The CANTD and CACTD domains are shown in cyan and silver spheres, respectively. In the snapshot of the cone-shaped capsid, the CANTD of the hexamer is shown in cyan, and the pentamer is shown in red. The CACTD domain of both hexamer and pentamer is shown in silver.

Figure 1—figure supplement 1
Dynamics of capsid translocation (without Lenacapavir [LEN]) to the nuclear pore complex (NPC) central channel.

(A) The time series of the distance (DCapNPC) between the geometric center of the capsid and equatorial midplane of the NPC inner ring along the channel axis. (B) To characterize the conformation of NUP98 chains, we characterize the probability distribution of radius of gyration (Rg) and end-to-end distance (Ree). Stages 1–4 are defined as equally dividing the simulation time (140×106 τCG) into four equal segments.

Figure 1—figure supplement 2
Top view of the cross-section of the nuclear pore complex (NPC) inner ring when the cone-shaped capsid is docked at the central channel and the NPC is ‘broken’.

The subunits of the inner ring are shown in gray spheres. The NTD domains of both CA hexamer and pentamer are shown in cyan spheres. The CTD domain is shown in white spheres. The rest of the NPC (Y complex and NUP98 chains) and the nuclear membrane is not shown for clarity.

Figure 1—video 1
Docking of the cone-shaped capsid (no Lenacapavir [LEN] bound) to the nuclear pore complex (NPC) central channel.

The NTD domains of CA hexamer and pentamer are shown in cyan and red spheres, respectively. The CTD domains of all CA monomers are shown in white spheres. The NPC is shown in gray, and the nuclear membrane is shown in pink (head group) and white (other groups). The rendered trajectory consists of coordinates saved every 0.5×106 τCG for a cumulative simulation time of 140×106 τCG (Figure 1—figure supplement 1). Initially, the NUP98 chains (shown in red) are extended. As the associative interactions between the FG-motifs of NUP98 chains and CA drive the capsid toward the nuclear end from the cytoplasmic end (via an electrostatic ‘ratchet’ mechanism), there is an increasing number of FG-CA interactions that provide the energetic driving force for capsid docking. During the docking trajectory, the NUP98 chains remain in extended conformation which continues to allow extensive FG-CA interactions, thereby providing the energetic driving force for capsid docking to the NPC central channel.

Figure 2 with 1 supplement
Competition between NUP98 and Lenacapavir (LEN) during docking of LEN-treated capsid.

(A) The time series of the distance (DCapNPC) between the geometric center of the LEN-bound capsid and equatorial midplane of the nuclear pore complex (NPC) inner ring along the channel axis (upper panel), the number of phenylalanine-glycine (FG) sites of NUP98 (fFGNUP98) directly in contact with the CA monomers (middle panel), and the number of LEN molecules bound to the capsid (NLEN) (lower panel) are plotted. Here, fFGNUP98 denotes the total number of FG sites that are in the vicinity of the CA hydrophobic FG-binding pocket (within 4 nm radius). (B) The upper panel shows the initial configuration and the final configuration of the capsid at the end of 200×106 τCG. In the snapshot at 200×106 τCG, one ribonucleoprotein (RNP) chain (blue spheres) partly extrudes out concomitant to the first appearance of defects at the pentamer-hexamer interface. The color scheme is the same as in Figure 1. In the lower panel, only the CTD domain is shown for each CA monomer. The CA monomers to which at least one LEN molecule is bound are shown in white spheres. The CA monomers to which no LEN molecule is bound are shown in orange spheres. Note the high density of orange spheres located at the narrow end of the capsid in the right panel (docked capsid). This indicates that some LEN molecules are displaced by the NUP98 mesh to allow capsid docking. A cutaway view of the nuclear membrane is also shown to represent the degree of capsid docking from the cytoplasmic side toward the nuclear side. The NPC, LEN, and NUP98 are not shown for figure clarity. For all the plots, the solid lines are the mean values calculated from the time series of two independent replicas, and the shaded region is the standard deviation at each time step.

Figure 2—figure supplement 1
Competition between FG-NUP98 and Lenacapavir (LEN) binding during capsid translocation for Replica 2.

The capsid is shown in a reduced representation (same as Figure 2). Only the CTD domain is shown for each CA monomer. The CA monomers to which at least one LEN molecule is bound are shown in white spheres. The CA monomers to which no LEN molecule is bound are shown in orange spheres. A cutaway view of the nuclear membrane is also shown to represent the degree of capsid translocation from the cytoplasmic side toward the nuclear side. The nuclear pore complexes (NPC), FG-NUP98, and LEN molecules are not shown for clarity.

Figure 3 with 2 supplements
Stepwise rupture of capsid treated with Lenacapavir (LEN) during nuclear pore complex (NPC) docking.

(A) The NTD domains of CA hexamer and pentamer are shown in cyan and red spheres, respectively. The CTD domains of all CA monomers are shown in white spheres. The ribonucleoprotein (RNP) chains are shown in blue. The NPC (cutaway view) is shown in gray, and the nuclear membrane is shown in pink (head group) and white (other groups). The initial events of CA-CA contact disruption at the hexamer-pentamer interface (CAHexCAPen) at the narrow end are labeled with arrow. In the leftmost snapshot, the RNA extrudes out of the defect as a result of partial dissociation of the pentamer. In the final stage, nucleation of cracks occurs at the hexamer-hexamer interface (CAHexCAHex) and extends from the narrow to the wide tip. (B) Snapshots of the defects at the hexamer-pentamer interface and hexamer-hexamer interface. CACTD domains of the hexamers that constitute the defect site are highlighted in dark orange spheres. Note that the CG beads constituting the CACTD domains at the defect site are disconnected from at least one nearest neighbor, leading to the locally cracked lattice (marked with red arrows). (C) Time series of the degree of defects at the pentamer-hexamer (NPenHex) and hexamer-hexamer (NHexH). Note that NPen and NHexHex are calculated by normalizing to the total number of CA pentamer (Burdick et al., 2020) and hexamer rings (209), respectively. The left panel shows the time series of the undercoordinated CA monomers at the pentamer-hexamer interface. The right panel shows the time series of the undercoordinated CA monomers at the hexamer-hexamer interface. (D) Time series of the radius of gyration (Rg) of the two RNP chains. The solid lines are the mean values calculated from the time series of two independent replicas, and the shaded region is the standard deviation at each time step.

Figure 3—video 1
Docking of the cone-shaped capsid (Lenacapavir [LEN]-bound) to the nuclear pore complex (NPC) central channel (Replica 1).

The NTD domains of CA hexamer and pentamer are shown in cyan and red spheres, respectively. The CTD domains of all CA monomers are shown in white spheres. The ribonucleoprotein (RNP) chains are shown in blue. LEN molecules bound to the capsid are shown in orange. LEN molecules that are not bound to the capsid are not shown for clarity. The NPC (cutaway view) is shown in gray, and the nuclear membrane is shown in pink (head group) and white (other groups). Initially, the CA-CA contacts are dissociated at the hexamer-pentamer interface, first at the narrow end and then at the wide end. The RNP complex extrudes from these defect sites. As the pentamers are dissociated, the integrity of the capsid is compromised. This leads to the formation of defects at the hexamer-hexamer interface. The movie frames correspond to snapshots every 250,000 τCG.

Figure 3—video 2
Docking of the cone-shaped capsid (Lenacapavir [LEN]-bound) to the nuclear pore complex (NPC) central channel (Replica 2).

The NTD domains of CA hexamer and pentamer are shown in cyan and red spheres, respectively. The CTD domains of all CA monomers are shown in white spheres. The ribonucleoprotein (RNP) chains are shown in blue. LEN molecules bound to the capsid are shown in orange. LEN molecules that are not bound to the capsid are not shown for clarity. The NPC (cutaway view) is shown in gray, and the nuclear membrane is shown in pink (head group) and white (other groups). Initially, the CA-CA contacts are dissociated at the hexamer-pentamer interface, first at the narrow end and then at the wide end. The RNP complex extrudes from these defect sites. As the pentamers are dissociated, the integrity of the capsid is compromised. This leads to the formation of defects at the hexamer-hexamer interface. The movie frames correspond to snapshots every 250,000 τCG.

Figure 4 with 1 supplement
Live-cell imaging of Lenacapavir (LEN)-induced rupture of human immunodeficiency virus type 1 (HIV-1) cores docked at the nuclear envelope.

(A) Dual-labeled viral cores: the capsid lattice is labeled with GFP-CA, and the capsid content is labeled with content marker HALO (cmHALO; JF646 dye). (B) Representative GFP-CA-labeled viral core in a cell expressing POM121-HALO (JF549 dye) that was stably associated with the nuclear envelope, retaining cmHALO in a DMSO-treated control cell (top) or losing cmHALO within 1 min of 100 nM LEN addition (bottom). (C) Number of GFP-CA-labeled viral cores per cell that remained at the nuclear envelope during the 15 min observation period. A total of 237 DMSO-treated cells and 192 LEN-treated cells were analyzed. (D) Percentage of GFP-CA-labeled viral cores that were cmHALO+ at the start of imaging. GFP-CA-labeled viral cores were analyzed for DMSO-treated (29 total) and nm LEN-treated (37 total) cells. (E) Percentage of GFP-CA-labeled viral cores that lost cmHALO during the 15 min observation period. (F) Time (min) of cmHALO disappearance following DMSO or LEN addition. p-Values were calculated using Fisher’s exact test; ns, not significant (p>0.05).

Figure 4—figure supplement 1
Characterization of dual-labeled human immunodeficiency virus type 1 (HIV-1) virions and live-cell imaging assay.

(A) HIV-1 vector design for virion labeling. HIV-1 plasmids expressing WT Gag (pHmNG) or Gag with internal GFP (pHGFP-GFPCA) or internal HALO (pHmNG-iHALO) were transfected at 40%, 10%, and 50%, respectively, of the total plasmid amount. Black triangles mark protease cleavage sites. Proteolytic cleavage of Gag from pHGFP-GFPCA generates GFP-CA upon virus maturation, while cleavage of Gag from pHmNG-iHALO produces fully processed HALO protein. Asterisk denotes a mutation in the env gene introducing a premature stop codon; virions were pseudotyped with VSV-G. The GFP and mNG reporters in the nef open reading frame are expressed in virus-producing cells but not incorporated into virions. (B) Virions containing cmHALO were labeled with Janelia Farm JF646 dye during transfection. Excess dye was removed during virus concentration. (C) Western blot analysis of viral lysates comparing unlabeled and dual-labeled virions. (D) Effect of labeling on virus infectivity. TZM-bl cells were infected with p24-normalized amounts of unlabeled or dual-labeled virus. Luciferase activity was measured 48 hours post-infection. (E) Representative images of dual-labeled virions. Virions containing GFP-CA and cmHALO (JF646) were centrifuged onto a chambered slide and imaged by confocal microscopy. The percentage of GFP-CA spots that colocalize with cmHALO is shown (AVG ± SD from five images; ~80 virions/image). (F) Live-cell imaging of dual-labeled viral cores in HeLa cells stably expressing POM121-HALO. (I) HeLa cells stably expressing POM121-HALO were stained with JF549 dye for 30 min, followed by washing to remove excess dye. (II) HeLa:POM121-HALO (JF549) cells were spin-infected with dual-labeled virus. Confocal z-stacks were acquired every minute for 15 min beginning ~2 hours post-infection. Complete media containing DMSO or 2× LEN was added between the first and second frames, yielding a final LEN concentration of 100 nM.

Figure 5 with 6 supplements
Molecular view of defects and Lenacapavir (LEN)-induced rupture of free capsids.

The color scheme of the capsid and ribonucleoprotein (RNP) is the same as in Figure 3. CACTD domains of the hexamers that constitute the defect site are highlighted in orange spheres. The snapshots in the inset show a zoomed-in view of representative defects. For the uncondensed RNP, snapshot 1 shows partial dissociation of pentamers at the narrow end, and RNP chains extrude out of the capsid interior. Snapshot 2 shows a ruptured narrow end. For the condensed RNP, snapshots 3 and 4 show defects arising from the partial dissociation of pentamers at the wide end and the narrow end. Condensed RNP localizes at the capsid wide end. In snapshot 3, the RNP extrudes out of the defect at the wide end. Finally, snapshot 5 shows rupture of the narrow end.

Figure 5—figure supplement 1
Time series of the appearance of defects at the capsid lattice in unbiased coarse-grained (CG) molecular dynamics (MD) simulations of capsid-Lenacapavir (LEN) complexes in the cytoplasm.

The left panel shows the time series of the undercoordinated CA monomers at the hexamer-pentamer interface. The right panel shows the time series of the undercoordinated CA monomers at the hexamer-pentamer interface.

Figure 5—figure supplement 2
Time series of the collective variable (CV) of well-tempered metadynamics (WTMetaD) simulations of capsid-Lenacapavir (LEN) complexes in the cytoplasm.

Four replica simulations (from top to bottom) are shown. The CV values fluctuate with time, indicating continuous partial dissociation (which leads to defects) and reformation of contact (healing of defects). In at least one replica (each for both uncondensed ribonucleoprotein (RNP) and condensed RNP), these defects lead to irreversible rupture of the narrow end within our simulation timescale (CV values less than 0.6).

Figure 5—figure supplement 3
Time series of the appearance of defects at the capsid lattice in well-tempered metadynamics (WTMetaD) simulations of capsid-Lenacapavir (LEN) complexes in the cytoplasm.

Four replica simulations (from top to bottom) are shown. The upper panel (A) shows the time series of the undercoordinated CA monomers at the hexamer-pentamer interface. The lower panel (B) shows the time series of the undercoordinated CA monomers at the hexamer-pentamer interface.

Figure 5—figure supplement 4
Free energy landscape of capsid disassembly of capsid-Lenacapavir (LEN) complexes in the cytoplasm.

The units of the free energy are in kcal/mol. The free energy landscape is from cumulative data for all four replicas in Figure 5—figure supplement 2.

Figure 5—video 1
Dynamics of Lenacapavir (LEN)-induced rupture of free capsids with internal ribonucleoprotein (RNP) in condensed form.

The color scheme is the same in other movies. LEN molecules that are not bound to the capsid are not shown for clarity. For ease of viewing, we aligned the capsid coordinates in the movie to the same as the initial frame, which removes the translation degree of freedom.

Figure 5—video 2
Dynamics of ribonucleoprotein (LEN)-induced rupture of free capsids with internal ribonucleoprotein (RNP) in uncondensed form.

The color scheme is the same in other movies. LEN molecules that are not bound to the capsid are not shown for clarity. For ease of viewing, we aligned the capsid coordinates in the movie to the same as the initial frame, which removes the translation degree of freedom.

Figure 6 with 1 supplement
Molecular details of Lenacapavir (LEN) binding and alteration of capsid microstructure.

(A) The left panel shows the time series of the number of LEN molecules (normalized by the number of CA hexamer rings) bound to a CA monomer which is either part of the ordered lattice (LENo) or distorted lattice (LENd). The snapshot in the left-center panel shows LEN molecules bound to regions of the lattice that are ordered. The CTD domain of the CA monomers with q6neigh < 0.4 (distorted lattice) is colored in magenta. The CTD domain of the rest of the CA monomers (ordered lattice) is colored in white. The NTD domain of the CA monomer of all the hexamers is represented as cyan spheres. The highlighted region (orange box) shows five LEN molecules bound to two adjoining CA hexamers that are classified as distorted lattice. The snapshot in the right-center panel shows LEN bound to CA that are classified as ordered lattice. The right panel shows the schematic of LEN molecule (represented as orange circle) binding to ordered and distorted CA hexamer ring. (B) The probability distribution of the CA-LEN potential energy calculated for all LEN molecules bound to CA sites that are part of the ordered (LENo) and distorted (LENd) lattice. The CA-LEN potential energy calculations were performed for the final 250×106 τCG. (C) Deviation of the kCACA(i) for each CA monomer i relative to the value is calculated for the free capsid. The CANTD domain of all CA monomers for which the kCACA(i) increase and decrease relative to the free capsid is shown in red and blue, respectively. The red patches indicate effective stabilization relative to free capsid.

Figure 6—figure supplement 1
Probability distribution of the number of Lenacapavir (LEN) molecules bound to a CA hexamer.

The error bars are calculated from four replica simulations. The occupancy probability distributions were performed for the final 250×106 τCG. The adjoining snapshot shows 1, 2, and 3 LEN molecules bound to CA hexamer labeled as LEN1, LEN2, and LEN3, respectively.

Additional files

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Arpa Hudait
  2. Ryan C Burdick
  3. Ellie K Bare
  4. Vinay K Pathak
  5. Gregory A Voth
(2026)
Lenacapavir-induced lattice hyperstabilization is central to HIV-1 capsid failure at the nuclear pore complex and in the cytoplasm
eLife 14:RP109282.
https://doi.org/10.7554/eLife.109282.4