كتابة تعليق

Overall assessment:

The research article entitled “Mechanoresponsive modulation of nuclear pore complex structure and function by O-GlcNAc,” authored by Chandra et al., proposed how post-translational modification of Phenylalanine-Glycine repeats-nucleoporins (FG-Nups) can modulate the diffusion barrier of the nuclear pore complex (NPC) in vivo. This modification in turn alters NPC permeability and changes the rate of macromolecule exchange between the cytoplasm and the nucleus. Based on these findings, the authors propose that O-GlcNAcylation of Nups acts as a counteracting mechanism that partially reverses NPC constriction under osmotic stress.

The authors use established fluorescent reporters carrying multiple Protein A fusions to measure diffusion rates in different cell-types, including HeLa cells and neuron-like SH-SY5Y cells, and then extend the analysis to iPSC-derived motor neurons. Quantification of O-GlcNAcylation on Nups with the RL2 antibody and metabolic labeling shows that a higher diffusion rate correlates with a higher O-GlcNAcylation level. The authors then manipulate Nup O-GlcNAcylation in two ways: by introducing engineered OGT/OGA tethered to the NPC, and by adding OGT/OGA inhibitors. Lastly, they measure NPC ring diameters by pan-expansion microscopy and find that the diameters vary with substrate stiffness and osmotic stress.

In general, the data supports their claims in the paper. Although extensive in vitro studies have shown that O-GlcNAcylation changes the properties and permeability of the FG network, physiologically relevant studies have been lacking. This paper sheds lights on the field as the first paper to study the model in vivo. Overall, the paper writes clearly and is easy for the readers to follow. The authors show a collection of amazing results. However, an in-depth mechanism on how the FG-Nups gets O-GlcNAcylated under mechano-stress condition and how this modification further alters the diffusion barrier property can be further addressed to fully strengthen the conclusion.

Major comments:

● As shown by the authors, different cell types have varied O-GlcNAcylation levels and diffusion rates. The authors could expand the discussion more about how and why different cells modulate diffusion barrier in different physiologocal conditions. In particular, why do neuron cells exhibit higher FG-Nups O-GlcNAcylation than other cell types, and what functional advantages might this confer on nucleocytoplasmic transport? Given the authors’ proposal that increased FG-Nup O-GlcNAcylation serves as a protective response to mechano-stresses, it would also be interested to discuss whether the intrinsically higher O-GlcNAcylation levels in neurons influence their susceptibility to acute mechanical stress. Addressing these questions would help establish the physiological significance of cell-type-specific FG-Nup O-GlcNAcylation and clarify whether it represents an adaptive mechanism for tuning the BPC diffusion barrier in response to distinct cellular demands and environmental stresses.

● As mentioned in the introduction section, mechanical forces regulate macromolecule shuttling in both passive and active transport through the NPC (Andreu et al., 2022; DOI: 10.1038/s41556-022-00927-7). Because many neurodegenerative diseases are caused by failures of active transport, it would be worthwhile to discuss the active transport machinery. One possible experiment is a digitonin-permeabilized nuclear import assay, which would allow the authors to compare active nuclear import kinetics between cells with different Nup O-GlcNAcylation levels. Probing mRNA export would also help readers understand whether this PTM globally alters bidirectional nucleocytoplasmic transport.

● In a broader context, the article does not fully describe the upstream mechanosensing pathway. The authors show that OGT is recruited to the NPC upon osmotic stress and that this recruitment further elevates the O-GlcNAcylation level of FG-Nups. However, it remains unclear which upstream components or signaling pathways trigger OGT recruitment, and whether the O-GlcNAcylation level decreases during recovery from osmotic stress. The authors also demonstrate that OGA is not recruited to the nuclear rim (Figure 7A). What, then, is a possible mechanism by which the cell modulates O-GlcNAcylation at the NPC after the stress? An additional discussion would help readers see the complete picture of mechanosensing at the NPC.

Minor comments:

● In Figures 1E and 2E, the unit for the y-axis (rate of initial nuclear fluorescence decrease) is missing. Based on the other figure panels, it is possibly a.u. per second. Adding the unit would help readers interpret the results more clearly.

● In multiple plots, including Figures 1C, 1D, 1E, 2B, 2C, and 2D, the molecular weight (kDa) label is missing for the various reporters. Consistent and proper labeling would enhance clarity.

● Figure 2E shows the variation in diffusion rate among different cell types. Because the rest of Figure 2 focuses on iPSCs and iPSC-derived i³LMNs, it would be more coherent to move this panel to Figure 1 and remove the iPSC and i³LMN data from it. The text can then explain that different cell types have different diffusion rates, which is more relevant to Figure 1.

● In Figure 3D, the Nup62 (or other FG-Nup) blot is not shown. The authors cleverly used Nup93 as a reference to show a global reduction in Nup content in i³LMNs compared with their parental iPSCs. The authors could add the Nup62 blot, as shown in Figure 3A, to indicate that FG-Nups are reduced by a similar ratio to other Nups. Similarly, in Figures 6B and 6C, the authors could add Nup62 blots to show comparable FG-Nup levels in cells grown on matrices of different stiffness.These modifications would help readers see that the stronger reduction in the RL2 signal is not caused by a global reduction of Nups in different cell types or the various matrices.

● In Figures 4D, 4E, and 4F, the authors show that introducing engineered OGA (the inactive mutant) results in higher RL2 fluorescence at the nuclear rim. Including untransfected (UT) cells as a negative control in these panels would make them more consistent with Figures 4A, 4B, and 4C.

● In Figures 5C and 5D, the diameter of each ring is not labeled. Although readers can find the numbers in the quantification, it would be more intuitive to compare the diameters directly on the diagram.

● In Figures 6B, 6C, S2A, S2B, and S2D, the quantification of the RL2 signal is not labeled. The authors could add the intensity ratio, as in Figures 3A and 3D, to make the figure set more consistent.

● In Figures 7E, 7F, and 7G, the authors show that TMG treatment under hyperosmotic stress alters nuclear volume and excess surface area but not the IR diameters. Additional evidence that higher FG-Nup O-GlcNAcylation alters the diffusion rate under osmotic stress would strengthen the claim.

● In Figures S5E and S5F, the authors claim that NUP O-GlcNAcylation during hyperosmotic stress has little effect on mitochondrial cristae distance or centriole diameter. However, the sample size (n = 4–5 cells) does not seem comparable to that of other measurements, such as those in Figures S5C and S5D. A larger sample would make the data more convincing.

● Does the O-GlcNAcylation level of FG-Nups vary across cell cycle stages? Cell cyle stages may influence nucleocytoplasmic transport rate and possibly the O-GlcNAcylation levels of FG-Nups. To ensure consistency and minimize potential confounding effects, the authors could consider synchronizing the HeLa, U2OS, and the differentiated SH-SY5Y cells to determine whether the observed changes in diffusion rate and FG-Nups O-GlcNAcylation are not attributable to difference in cell-cycle stages.

I am one of 2 participants from the same research lab that reviewed this preprint as part of the Transparent and Accountable Peer Review training program. This report was written with the support of my research mentor.

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.