Quantitative analysis of nuclear pore complex organization in Schizosaccharomyces pombe
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (ASAPbio crowd review)
- @IratxePuebla's saved articles (IratxePuebla)
Abstract
The number, distribution, and composition of nuclear pore complexes (NPCs) in the nuclear envelope varies between cell types and changes during cellular differentiation and in disease. To understand how NPC density and organization are controlled, we analyzed the NPC number and distribution in the fission yeast Schizosaccharomyces pombe using structured illumination microscopy. The small size of yeast nuclei, genetic features of fungi, and our robust image analysis pipeline allowed us to study NPCs in intact nuclei under multiple conditions. Our data revealed that NPC density is maintained across a wide range of nuclear sizes. Regions of reduced NPC density are observed over the nucleolus and surrounding the spindle pole body (SPB). Lem2-mediated tethering of the centromeres to the SPB is required to maintain NPC exclusion near SPBs. These findings provide a quantitative understanding of NPC number and distribution in S. pombe and show that interactions between the centromere and the nuclear envelope influences local NPC distribution.
Article activity feed
-
-
This review reflects comments and contributions by Ankita Jha, Zara Weinberg, Julia Grzymkowski, Julien Berro, Karen Lange, Sónia Gomes Pereira, Arthur Molines, Jacob Herman, and Manoj Yadav. Review synthesized by Jacob Herman.
The work by Joseph Varberg and colleagues uses super resolution microscopy to better characterize the non-random distribution of nuclear pore complexes within the nuclei of the fission yeast Schizosaccharomyces pombe. This work also confirms findings in other organisms that nuclear pore complexes exist in multiple compositions. In addition to better documenting this phenomenon, this work begins to characterize the mechanisms by which nuclear pore position is regulated. Specifically, the authors show that clustering centromeres at the spindle pole body excludes nuclear pore complexes from the spindle pole body, …
This review reflects comments and contributions by Ankita Jha, Zara Weinberg, Julia Grzymkowski, Julien Berro, Karen Lange, Sónia Gomes Pereira, Arthur Molines, Jacob Herman, and Manoj Yadav. Review synthesized by Jacob Herman.
The work by Joseph Varberg and colleagues uses super resolution microscopy to better characterize the non-random distribution of nuclear pore complexes within the nuclei of the fission yeast Schizosaccharomyces pombe. This work also confirms findings in other organisms that nuclear pore complexes exist in multiple compositions. In addition to better documenting this phenomenon, this work begins to characterize the mechanisms by which nuclear pore position is regulated. Specifically, the authors show that clustering centromeres at the spindle pole body excludes nuclear pore complexes from the spindle pole body, and when these two complexes are forcibly dimerized mitotic defects result in decreased fitness.
The commenters were overall quite impressed with the imaging technique. The major conclusions and message of the preprint were generally well received, the comments or questions below relate to very specific text and experiments.
A few key themes mentioned in specific comments were:
- A desire for more consistent statistical analysis of data.
- Suggestions for additional data for some statements or toning down of the claims. NPC clustering is commonly discussed but there were questions as to how this phenotype was being measured.
Specific comments
Introduction
“perhaps explaining links between changes in NPC density and cancer” - The statement could note whether the correlation between NPC density and cancer is positive or negative.
“for example, emerin is enriched at pore-free regions of the NE in cultured cells (44). In budding yeast, NPC density is increased in the region of the NE near the spindle pole body (SPB),” - Does the SPB contain LEM domain proteins or is this a different possible mechanism for the non-random regulation of NPC density?
“Using S. pombe as a model system”- Why not S. cerevisiae, which is discussed earlier to have significant prior art in this regard? I'm sure there is a good reason, I think it could be outlined a bit more in the intro.
“We quantify NPC number under a range of conditions” - It would be useful to mention briefly at this point what types of conditions this refers to.
“Additionally, NPCs are excluded from the NE region surrounding the SPBs by Lem2 and other factors.” - Could the authors clarify if it is also something that is conserved or it is a new finding?
Results
Subheader “3D-SIM image analysis pipeline for NPC quantitation” - Worth mentioning the conclusion that NPC density is independent of cell cycle stage since that is the major conclusion from this section.
“This approach provides a roughly two-fold increase in resolution as compared to conventional light microscopy” (Figure 1A) - For those who have never imaged NPCs it would be really informative to see a confocal or wide field image to better understand how SIM imaging made this project possible.
Figure 1 legend “E) Mean number of NPCs, nuclear surface area and NPC density measurements from four independent replicates. Significant differences (*) determined using Wilcoxon rank sum tests. ns, not significant.”
- Each one of the coloured dots on the graph appear to represent the mean NPC number for each replicate. If so, then this information should be added to the figure legend, as it is not immediately obvious for a broader audience. If not, please clarify what each dot represents. Same in Figure S1E.
- Were these tests conducted pairwise? Are the reported p values corrected for multiple hypothesis testing? Having a dedicated "Statistics" section in the methods would be helpful for reporting this.
“We observed that the number of NPCs also increases through interphase to maintain a constant NPC density (Table S1, Fig. 1D-E)”- There are no cell-cycle markers used to determine cell-cycle progression but only a visual assessment from cell size and nucleus shape. It would be good to show the three plots in figure 1E as scatter plots with the X axis being cell size for cells that are not yet in mitosis (1 nucleus). Then do a correlation analysis between cell size and NPC number, Surface Area, NPC density.
“We observed occasional differences in NPC density between mother and daughter nuclei produced by the symmetric mitotic division in S. pombe, reminiscent of the elevated NPC density observed for daughter nuclei produced by the asymmetric mitosis in S. cerevisiae”
- It is not clear what the authors mean by "occasional", is it 1%, 5%,10%? It would be better to replace this with a specific number/% of events. Additionally, the question arises as to what happens afterwards in the daughter cells, do they retain the NPC density asymmetry? Or do they eventually achieve similar densities?
- Some of these points are addressed later on in the paragraph and Fig S1A. Some edits to this sentence should address this and provide clarity.
“Despite the improved lateral resolution offered by SIM, clustering of NPCs and the comparatively reduced axial resolution likely leads to undercounting of NPCs using 3D-SIM” - This is useful context for Figure 1C - it would be worth mentioning in that section how the automated counting handles clustered NPCs, or placing the paragraph earlier with a short description of the methods.
“Similarly, a constant NPC density was maintained when nuclear size was reduced in mitotic cells using a temperature-sensitive allele of Wee1 kinase (wee1.50)” - The NPC density does not change with temperature in the wee1 mutant but the NPC density in Fig 2B is lower than in a WT in Fig 1. The Nup tagged in the two figures are different, so this could be an explanation (as shown in Fig S1 E) but it could be good to make sure that the wee1 background does not have a different NPC density. I don't see a quantification of the NPC density using Nup37 in the WT elsewhere. In fact, Nup37 seems to be used only in wee1 background and in Sup Fig. 1 B.
“The increase in NPC number was dependent on NE membrane expansion during arrest, as chemical inhibition of fatty acid synthesis by treatment with cerulenin blocked nuclear growth while NPC density was maintained (36°C + Cerulenin = 6.8 ± 1.5 NPCs/μm2, n=110) (Fig. 2C)” -The effect of the Cerulenin drug on a WT background is not shown, was that control experiment done? It would also be helpful to include statistics in this section.
“Yeast lacking core components of the autophagy machinery (atg8Δ or atg1Δ) (75) that targets NPCs for degradation during nutrient deprivation do not show increased NPC density compared to wild-type cells, suggesting that autophagy is not used to remove NPCs to maintain NPC density (Fig. 2D)” - It is unclear that this experiment alone tells us much about the regulation of NPC density. If atg8 and atg1 are known to regulate NPC removal only in response to nutrient deprivation then consider performing these experiments under that condition or revisiting whether they fit here.
“NPC density is maintained by a mechanism that restricts the assembly of new NPCs in the absence of increased available NE surface area.” - This conclusion is indicative of a mechanism where NPC assembly is maintained by restricting the assembly, however all the data above is indicative of the mechanism where NPC assembly is correlated with NE surface area, for increase there must be an additive mechanism and for a decrease in the NE, there must be a mechanism of removal. This suggests that the NE surface area regulation mechanism could be tied to NPC density. One way to clearly show that could be a correlation plot of NE surface area and NPCS density, color coded for all the different conditions tested.
Figure 2- It appears as though no comparative statistical analysis was done with the quantitative data displayed in Figure 2, yet it is stated that e.g., "treatment with cerulenin blocked nuclear growth while NPC density was maintained" or "yeast lacking the autophagy machinery do not show increased NPC density". These conclusions would be strengthened if statistics were run on the data similar to Figure 1.
“NPC clustering is common phenotype in different cell types and in mutants defective in NPC assembly.” - Does this mean that NPC clustering is higher in mutants defective in NPC assembly? Would suggest including references for this. Also, this paragraph needs an introduction to why NPC clustering matters? Does it have any connection with the NPC distribution?
“3D-SIM images revealed the presence of multiple smaller clusters distributed throughout the NE (Fig. 3A).” - It is unclear (also not mentioned in the Methods section) how clusters are identified. The images show rings but it is hard to tell how many clusters compose that ring structure. It will be beneficial to show how clusters are quantified. Can that be resolved with 3DSIM?
“We frequently observed NPC clusters organized in a ring-like structure with diameters ranging from 250-300 nm (Fig. 3B)” - Is it possible to report what was the frequency of the ring structures in nup132-deleted and wt cells?
“Clustering increased in aged nup132Δ cells grown on plates (Fig. 3C)”- The figure depicts the NPC ring like structure, does this mean that the ring increased or the clustering has increased. Does increase in clustering make the rings more continuous?
“NPC clusters were frequently enriched in the anaphase bridge, along with excess membrane (Fig. 3E)” - Providing a quantification of NPC cluster enrichment in the anaphase bridge would be helpful.
“Following completion of nuclear division, the resulting daughter nuclei had normal NE morphologies and NPC densities equivalent to wild-type nuclei (Fig. S2). This suggests that nem1Δ nuclei can remove excess NE membrane and NPCs during mitosis via the anaphase bridge.”
- This implies that prior to mitosis nem1∆ cells have abnormal morphology and NPC densities but the latter is not measured.
- The NPC density reported in Figure S2 for the WT and the Nem1 mutant are different from the NPC density reported for the WT in figure 1 and figure S1 yet it is done using the same tagged Nup, Nsp1. It would be helpful to have an explanation. If the NPC density is “a constant” in the WT it should not be different from one figure to another. If the nem1 mutant has a density of 4 NPC/micron^2 then it is different from the WT. Also, the NPC density in the nem1 mutant in Figure S2 seems almost bimodal. Increasing the number of nem1-delta cells analyzed could help identify if it is bimodal or if it is due to under-sampling.
- For the nem1 mutant the clustering is not quantified.
“In contrast, NPC clusters in nup132Δ nuclei coalesced into larger clusters that preferentially localized to the SPBs in mitosis (Fig. 3G)”
- An overlay image could be included to support this statement.
- Fig. 3F could be referenced here too because otherwise it is not referenced until the discussion; at which point it is used to reference the data that is referenced here as Fig. 3G.
“We observed a clear reduction in NPC density over the nucleolus” - Is this referring to where the yellow and magenta staining meet in Fig. 4? It is not immediately obvious as to where "over the nucleolus" is in those slices. Can the regions that are being compared (NPC staining at NE vs. NPC staining over nucleolus) be highlighted/specified in some way so as to better understand the quantification method?
Figure 4
- 4C is gorgeous - really conveys the point well!
- In this figure the authors at first show a 3D-SIM image, but perform the intensity analysis on the confocal slice. What is the reason for it? Analysis of the 3D-SIM data could provide more information on the characteristics (number, spatial distribution) of NPS density reduction.
Figure 5
- Very minor comment -- the scale bar is very hard to see in Fig 5A.
- Statistics for Fig. 5B would strengthen the conclusion that the exclusion was cell-cycle independent.
Figure 6
- Figure 6D - Looking at the insets, the exclusion area in the lem2(delta)C-off appears to be the smallest one and closer to the exclusion area shown for the lem2(delta) in panel B. However, this is not represented in the quantification/results. I wonder if there is another image that would more closely represent the quantification outcome? Or if the insets might have been mislabeled, for instance lem2(delta)C-off could indeed represent the lem(delta)N-on and vice-versa?
- Figure 6E - This is listed as F in the legend.
“Tethering did not affect microtubule nucleation at the SPB, including the formation of cytoplasmic microtubules.” - Please provide evidence supporting this statement.
Discussion
“In nem1Δ mutants, both excess nuclear membrane material and NPCs are segregated into the anaphase bridge region during nuclear division” - This would benefit from some analysis - are there too many NPCs? Is it specifically the clustered NPCs? Currently the data supporting this is snapshots from a single movie.
“The ability for 3D-SIM to resolve and quantify individual NPCs labelled with multiple fluorescent proteins at endogenous levels provides tools to begin to interrogate_ how altered NPC compositions may allow for functional specialization of NPC function at distinct regions of the NE.” _- The high resolution images are really beautiful! Great job in showing the power of 3D-SIM to help answer these types of biological questions.
Methods
“Images were acquired overa6 μm volume with 0.3 μm z-spacing for 45 min at 2 min intervals.” - For dynamic measurements a 2-minute interval is big and it would be interesting to see a few time-series imaging with smaller intervals to capture the fast changes.
-
