Avalilação PREreview de Structural basis of lenacapavir-induced HIV-1 capsid disruption during virion maturation

The manuscript by Tanaka et al. explores the effects of long-acting lenacapavir (LEN) on the structure of the HIV-1 capsid. To understand how LEN incorporation alters capsid architecture and viral infectivity, the authors used cryo-electron microscopy to determine the structure of mature HIV-1 capsid lattice assembled within virus-like particles formed in presence of LEN. The structural analysis suggests that LEN enforces flattened lattice geometry, leading to an altered capsid and larger viral particles. Finally, the authors confirmed the loss of reverse transcriptase activity due to a compromised capsid integrity through biochemical analysis. Overall, the data in this manuscript is well presented and supports the authors' claims. We outline major and minor adjustments to improve clarity in reporting and presentation, as well as providing additional context for a broader audience.

Major comments

-While we found the data in Figure 1 (and associated Figures S4 and S7) compelling, we were concerned that when demonstrating the flattening effect of the LEN binding the capsid, all calculated angles for the LEN-negative structures come from models in the electron microscopy data bank (EMDB) rather than your own data. We understand that this may have been done to decrease time and costs. However, to ensure that flattening is not an artifact of the way the capsid was imaged, please consider including a comparison of your own atomic model of capsid without LEN. Alternatively, please elaborate in the discussion why it may not be necessary to include your own LEN(-) 3D atomic structure.

-The structures in Figure 1 demonstrate the arrangement of the capsid hexamers in the presence of LEN. However, without a structure of wildtype capsid produced in the absence of LEN, it is difficult for the reader to visualize the structural changes. Please include a wildtype capsid model obtained from your own images or from the EMDB for Figure 1d,e to strengthen your argument that the structure of the hexamers is altered in the presence of LEN.

-Figure 3 indicates a loss of reverse transcriptase (RT) activity, but a retention of integrase protein when wildtype capsid is treated with LEN. Given that RT presence was measured indirectly through activity rather than through a western blot or another quantitative protein method, this leaves two possible conclusions; either RT itself is being lost, or the LEN is interacting with RT in some way to reduce its efficacy. It seems unlikely that integrase would be retained with the capsid while reverse transcriptase is lost, especially since reverse transcriptase is a larger protein than integrase. Including a western blot similar to Figure 3b, e (or another quantitative protein analysis method) would allow to investigate changes in RT association with the capsid. We also recommend changing the figure caption to “LEN-containing cores in VLPs exhibit reduced integrity and lose reverse transcriptase activity” since in the current state, the figure does not explore changes in RT protein levels.

-In the discussion section: “Recent preprint further reported curvature-dependent destabilization of mature HIV-1 capsids by LEN in post entry cores” references another recent preprint that demonstrates a similar loss of capsid curvature in the presence of LEN (dos Santos et al. bioRxiv 2026). We appreciated referencing a similar preprint, and including such a reference strengthened the findings of your manuscript. We recommend including a few lines in the discussion detailing how your study compares to or differentiates from the dos Santos study.

- Please include your rationale for using 293T cells for VLP production and capsid structural analysis since these are not the native cell type HIV-1 infects. Do you believe the capsid structure in the presence of LEN varies if the cell type is produced in 293Ts versus a more native cell type such as MDMs or Jurkats?

Minor comments

-Figure 1 a,b,c and Figure 3 i,j,k contains cryoEM and cryoET images, respectively, of VLPs produced in the presence or absence of LEN. Consider grouping or contrasting them in the text to explain the advantage of doing both analyses.

-Figure S4 and Figure S7 legends contain description of which structures are LEN (-) and their origin (native virions, VLPs, etc.), but Figure 1h is lacking this information. Please state in the figure what type of structure the LEN (-) map originated from.

- On Figure 2b,c,d, showing a color map of the subtraction of the +/-LEN structures would help improve understanding of the structural differences when LEN is bound.

-Figure 3h uses a benzonase assay to indirectly measure how open the capsids are; if the capsids are fully closed, the benzonase cannot access the RNA and degrade it, while if the capsids are open the benzonase degrades the RNA. The y axis is titled “relative RNA to benzonase -”, but we did not understand how the RNA levels were normalized between the LEN (-), the LEN (+), and the M66I cores. Please include the raw data without normalization as the M66I virus seems to be less infectious, and the lack of difference upon benzonase-treatment could be due to less RNA overall.

- Figure 4a,e demonstrates that the M66I capsid mutant is less infectious than the WT virus, but this is never addressed in the text. Stating in the text that M66I is less infectious and providing a brief explanation or hypothesis as to why, would improve reader comprehension.

-The p values and standard deviation bars appear to be missing on Figure 4. Adding these in, along displaying all data points, would strengthen data presentation and improve assessment of data variability.

-To increase comprehension in readers not familiar with HIV-1 research, please define Jurkat cells and VSV-G pseudotyped HIV-GFP for Figure 4, as well as a sentence about the effects of RAL and NVP on HIV-1 infection in the discussion of Figure 4.

-For the readability of a broader audience please consider including an explanation for the Gag polyprotein in the introduction.

-In the discussion, the unfortunate alignment of reference 2 with the pharmacologically relevant concentration of LEN leads to what looks like “5 nM²”, or five nanomolar squared. We recommend shifting the location of the reference in the sentence to prevent any potential misunderstandings.

-In the methods for the nuclease sensitivity assay, the MgCl₂ has a box in place of the 2.

Camy Guenther and Josy Joseph (Indiana University Bloomington) - not prompted by a journal; this review was written within a Peer Review in Life Sciences graduate course led by Alizée Malnoë with input from group discussion including Carter Collins, Lily Pumphrey and Tahreem Zaheer. None of the authors of this review have extensive expertise in cryo-EM methodology, and our comments are made with respect to that. We are part of the Dept. of Biology where Adarsh Dharan’s group is located, and Camy Guenther is a graduate student from his group. Adarsh is a collaborator at the Pittsburg Center for HIV Protein Interactions with Barbie Ganser-Pornillos (corresponding author of ref. 42); this interaction did not influence the choice of this preprint for our class.

Competing interests

The authors declare that they have no competing interests.

Use of Artificial Intelligence (AI)

The authors declare that they did not use generative AI to come up with new ideas for their review.