Nanoscale organization of betaII-spectrin within segments of the membrane-associated periodic skeleton in mouse sciatic nerve axons

Reviewer #1 (Public review):

Summary:

The article "Nanoscale organization of beta-II spectrin within segments of the membrane-associated periodic skeleton in mouse sciatic nerve axons" by Gazal et al. looks into the organization of the spectrin scaffold in mouse sciatic nerves using super-resolution microscopy. It is now well established that axons, across species, contain a membrane-associated periodic scaffold mainly composed of circumferential actin filaments and longitudinally arranged spectrin tetramers. While super-resolution imaging of neurons in cell culture is relatively easy, exploring the ultrastructure of myelinated axons in intact nerve fibers is a daunting task. Nevertheless, the authors have attempted this by fixing and preparing cross-sections of sciatic nerves. They have then tried to quantify the fluorescence intensity patterns of specific components, especially that of labeled beta-II spectrin and have analysed its distribution.

One of the main findings is that spectrin is distributed along the axonal periphery and along the outer part of the myelin sheath. By labelling multiple cellular components and using intensity analysis, the authors show the sequence of structural organization of a few key components. They see that, unlike in the case of axons in culture, the axonal cross-sections within the sciatic nerve deviate significantly from a circular shape. They then use 3D-dSTORM to investigate the distribution of beta-II spectrin along the axonal circumference. They see that this distribution is very heterogeneous, both in the sizes of spectrin puncta and their arrangement along the periphery. The amount of spectrin scales linearly with axonal circumference.

Strengths:

Super-resolution imaging of axons of intact nerve fibers to investigate the organization of beta-II spectrin.

Weaknesses:

While most of the findings, like the spatial distribution of spectrin and related components, are reasonably well supported by data, I have concerns regarding the subsequent claims made in the article. The detection of axial periodicity based on the observation of a peak in the inter-tetramer spacing distribution is not very convincing, and a 3D representation (or a video of 3D reconstruction) would have been better. And so are the claims on characteristic spectrin spacing of 200 nm along the axonal circumference. A peak in the distribution does not imply a periodic arrangement.

Reviewer #2 (Public review):

Summary:

This is an interesting paper by the Unsain lab looking at the nanoscale organization of the membrane-associated periodic cytoskeleton in mouse sciatic nerve axons. The precise organization of the structure remains unclear, especially in vivo, and this manuscript significantly adds to our knowledge of this important structure. While some of the findings in the study are somewhat expected (though still valuable to see in an in vivo setting), an interesting observation is the presence of discrete nanoscale clusters that scale up with the size of the axon, which challenges previous assumptions.

Strengths:

Strong, convincing data; clever combination of imaging and analytical tools to make novel points; well written; excellent composition of figures.

Weaknesses:

(1) Figure 2A/3A: The large and small clusters of spectrin, as seen in cross sections, are unexpected and novel. The authors have done a clever job of combining imaging and analyses, but some things are still unclear. First, the authors should be consistent in their language when they talk about the spectrin clusters. Recommend precise language to define the small and large clusters when they first appear in the text, and then use the definitions consistently throughout the text. Second, based on the data shown, one does not get a clear idea of how the small and large clusters are organized along the longitudinal axis of the axon. In that context, are Figure 2B and C from imaging along the longitudinal axis? If not, it's unclear how the authors can conclude that the spectrin assemblies have a distance of ~170 nm along the linear axis. In general, a perceived limitation of this study is that while the authors have done a good job looking at cross sections, there is no information on the longitudinal distribution of spectrin in these axons. Looking at both cross- and longitudinal sections would also clarify details about the large spectrin clusters. For instance, are they small sausage-like structures, or long rods of spectrin running along the length of the axon? One assumes that all the analyses in Figures 3 and 4 are from the small clusters. Can the authors do a similar analyses of the large clusters? Finally, a schematic model showing both cross- and longitudinal- sections would make things clearer, but the authors would need to show the longitudinal data for that.

(2) It is interesting to think that the larger spectrin accumulations may be similar to the condensate-like structures seen by Boyer et al., as the authors mention in the discussion. In that context, it is possible that these focal accumulations are local reservoirs of spectrin that are also seen in mature axons (indeed, these accumulations were also seen in mature axons in the Boyer et al. paper, and they also speculated that these accumulations may be local reservoirs). Can the authors check if actin/adducin is also present in these larger spectrin accumulations?

(3) While talking about the nanoscale clusters, it is important to specify that the authors are talking about circumferential clusters. Though the writing is excellent, one still does not get the precise definition of "clusters" from just reading the abstract, and it would be good if the authors could work on that more (I recognize that this is not easy to do).

Reviewer #3 (Public review):

Summary:

In the presented work, the authors investigate spectral staining in axons of the sciatic nerve, where the MPS has been detected before using STED microscopy. They employ 3D-dSTORM in tissue sections and analyze the data, measuring localization of clusters on the axon perimeter and the relative distribution of those. From these data the conclude that large gaps in spectrum localizations exist and that clusters around the axon exist that are spaced at 200nm.

Major Comments:

(1) The presented data are at times overinterpreted, and the discussion lacks a critical view of the data. For example, the statement "...Unlike previous suggestions from qualitative evidence in cultured neurons (REfs), βII‑spectrin distribution in MPS segments of peripheral nerves is discontinuous, with extensive stretches of the perimeter lacking βII‑spectrin." is quite strong, given it is based on immunofluorescence staining and dSTORM microscopy in tissue. Absence of evidence of staining is not evidence of absence.

(2) The authors claim in the abstract that "The number of these clusters scales linearly with the axonal perimeter, maintaining a constant membrane occupancy of ~20% across varying axon diameters." Again, this is from a cut through an axon, while measuring the density of clusters on the perimeter. If they claim area occupancy, an area should be imaged, and the dots (clusters) should be measured in surface coverage in a 2D projection of the axonal surface.

(3) In general, this reviewer suggests being a bit more moderate in statements such as: "These findings challenge simplified models of the MPS based on cultured systems and demonstrate that the MPS in peripheral nerves is composed of discrete structural units." These statements are bold from the relatively few measurements in a single method and a single viewpoint. Especially when considering that techniques such as dSTORM depend extremely highly on labeling density, and apparent clustering of localization is highly prone to misinterpretation. If the authors desire to make such statements, working with endogenously labeled protein would be warranted. The authors should at least hedge such statements.

(4) If the authors want to make statements about general organization, why do they not compare adjacent cuts through the axon? If there are continuous spectrin filaments, the clusters should appear at the same site across repeated cuts through the axon.

Besides this, this reviewer welcomes the effort that has been made to establish dSTORM in tissue sections and to investigate the MPS in native tissue.

Author response:

We sincerely thank the editors and reviewers for their overall positive assessment and constructive feedback on our manuscript detailing the nanoscale organisation of βII-spectrin of the membrane-associated periodic skeleton (MPS) in mouse sciatic nerve axons. Their perspective and comments will help refining the manuscript.

A common comment by the reviewers relates to the description of the characteristic longitudinal periodicity of the MPS. We value these comments, which we believe are motivated by the fact that the longitudinal periodicity of the MPS is undoubtedly the most studied and prominent feature of the MPS in cultured neurons. However, the main goal of the present project was to describe how βII-spectrin is organised in the transverse axis of individual segments of the MPS in nerve tissue. This is why we utilised cross-sections of the sciatic nerve, hence achieving the best resolution possible in that plane, at the expense of the resolution in the axial axis. Furthermore, this study clearly shows that the transverse morphology of axons, and thus of the MPS, of neurons in the tissue is highly irregular, in comparison to cultured neurons. This imposes an extra challenge to observe correlated longitudinal structures when the observation length is limited, as in our studies. Nonetheless, to improve this aspect of the manuscript, we will revise our data and previous evidence, clarify the methodological trade-offs made, and make our interpretations more accurate.

Additionally, we will clarify several imaging- and definition-related inquiries, including tests for insufficient staining, the interpretation of βIII-tubulin staining, the assessment of axon–glia boundaries, and consistency in the use of terms like ‘clusters’ and ‘periodicity’, among others.

We believe these and other revisions will substantially strengthen the manuscript and comprehensively address the reviewers' feedback.