Remyelination alters the pattern of myelin in the cerebral cortex

Acceptance summary:

The authors used time-series two-photon in vivo imaging to examine how oligodendrocytes myelinate axons in the gray matter after cuprizone-induced demyelination. This topic has been difficult to address with more conventional approaches, yet is very important for understanding remyelination in the CNS. Their analysis revealed that remyelination was more efficient in the upper layers of the cortex, and that some oligodendrocytes appear in new locations, where they form sheaths at positions of the axons that were previously unmyelinated, thereby forming new patterns of myelination. The paper is full of new insights into cortical remyelination that raise many important questions for future research.

Decision letter after peer review:

Thank you for submitting your article "Remyelination alters the pattern of myelin in the cerebral cortex" for consideration by eLife. Your article has been reviewed by three peer reviewers, and the evaluation has been overseen by a Reviewing Editor and Gary Westbrook as the Senior Editor. The following individuals involved in review of your submission have agreed to reveal their identity: Robert H Miller (Reviewer #1); Mikael Simons (Reviewer #2); Ben Emery (Reviewer #3).

The reviewers have discussed the reviews with one another and agree that only modest revisions are necessary. Congratulations on such positive reviews. As the comments are straightforward, the original reviews are included below. Please include a point-by-point response to these points with your revision.

Reviewer #1:

This is a very interesting and well-done manuscript that provides detailed insights into the regenerative capacity of the oligodendrocyte lineage in response to gray matter demyelination. Using the cuprizone model the authors clearly demonstrate that while myelin sheaths are highly stable in the undamaged adult, following demyelination/remyelination the pattern, axonal selection and number of myelin sheaths and oligodendrocytes are altered. There are a number of interesting aspects to the work including the description of local differences in the extent of repair and the notion that there is a residual "axonal mark" of nodes of Ranvier. By its nature this is a descriptive manuscript, however this in no way detracts from its importance as a foundation for further mechanistic studies. Overall the analyses are detailed, rigorous and appropriately applied. I have no major issues with the current manuscript.

Reviewer #2:

This is an outstanding paper on the regeneration of myelinating oligodendrocytes in the cortex of mice. The authors use longitudinal two photon in vivo imaging to examine how oligodendrocytes myelinate axons in the cortex after cuprizone induced demyelination. This analysis led to a number of new exciting discoveries: Remyelination was more efficient in the upper layers of the cortex. Some oligodendrocytes appear in new locations, where they form sheaths at positions of the axons that were previously unmyelinated, thereby forming new pattern of myelination. They also found that oligodendrocytes preferentially formed sheaths on axons that were previously myelinated. Analysis of sheath position revealed spatial specificity, possibly by βIV-spectrin marked landmarks along the axon. Thus, the paper is full of new important insights into cortical remyelination that raise many important questions for future research. The paper is very well-written and technically convincing. I recommend publication with only few suggestions for the revision:

1) The Discussion is well written, but the possibility that axonal damage may have occurred during cuprizone treatment is not mentioned. This may, at least partly, have influenced the layer specific remyelination pattern and whether sheaths were replaced or newly formed at a previously unmyelinated axon.

2) It would be informative to add a Y-projection of an example region from cuprizone-treated animals at a late time point of recovery, similar to Figure 2A.

3) In Figure 3A and B, it is not easy to see where in the 3D space the individual data points are located. Is there a different way to illustrate this?

4) Since new cells are inserted in between stable cells in control animals, I would expect nearest neighbor distance in Figure 3C to be significantly lower in "all cells" vs. "stable cells". Please comment.

5) Referring to Figure 4E and F, please add a description of how volumes for randomly appearing oligodendrocytes were predicted.

6) In the "2 neighbors" subpanel of Figure 6B, it appears like part of the right neighbor at 3 weeks of cuprizone and the left neighbor at 10d of recovery were lost, too. Please comment.

7) Figure 3—figure supplement 1B does not seem to match Figure 3—figure supplement 1A (sheath length is < 40 μm at 0 dpf, curve should be steeper between 0 and 8 dpf). Please verify.

Reviewer #3:

In this manuscript Orthmann-Murphy et al. use in vivo two photon imaging to study the spatial patterns of remyelination in the cortex following cuprizone intoxication, making several interesting findings. These include that oligodendrocytes are replaced in different locations to the developmentally born oligodendrocytes, and that they are relatively inefficiently replaced within deeper cortical layers. Where new oligodendrocytes have overlapping domains with the lost oligodendrocytes, the previously existing myelin internodes can be replaced with a surprising degree of fidelity, including replication of the previous locations of nodes of Ranvier. Maintained expression of some of the cytoskeletal components of the node by demyelinated axons may contribute to this. In spite of the ability of new oligodendrocytes to replicate many previously existing myelin segments, the overall pattern of myelin is substantially disrupted following de/remyelination; the authors provide an excellent discussion of the implications of this for circuit function in the remyelinated cortex.

Although there are some inherent limitations in the study (most notably that the patterns of remyelination seen in the cuprizone model may or may not extrapolate to other rodent models or to MS), the work is highly original and of a very high standard. The findings are novel and highlight the power of in vivo live imaging techniques. I am enthusiastic about the manuscript and have only one request for revised presentation of data.

1) Figure 5E would benefit from a more comprehensive set of images showing a region of cortex at baseline, peak demyelination and then remyelination (in addition to the pseudocolored analysis showing novel and replaced segments at the end of the experiment), thoroughly documenting examples of internodes that are lost and then replaced. Although the authors emphasize the overall disruption of the previously existing patterns of myelin following remyelination, the fact that over half of the myelin internodes that degenerated were subsequently replaced (Figure 5D) is actually fairly remarkable.