Confined migration induces heterochromatin formation and alters chromatin accessibility

  1. Chieh-Ren Hsia
  2. Jawuanna McAllister
  3. Ovais Hasan
  4. Julius Judd
  5. Seoyeon Lee
  6. Richa Agrawal
  7. Chao-Yuan Chang
  8. Paul Soloway
  9. Jan Lammerding

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

No abstract available

Article activity feed

  1. Version published to 10.1016/j.isci.2022.104978
  2. ASAPbio crowd review

    Review coordinated via ASAPbio’s crowd preprint review

    This review reflects comments and contributions by Richa Arya, Luciana Gallo, Lauren Gonzalez, Sam Lord, Dipika Mishra, Arthur Molines, Mugdha Sathe, Ryman Shoko, Ewa Maria Sitarska. Review synthesized by Ehssan Moglad.

    Study conducted by Chieh-Ren Hsia et al. which looked at nuclear deformation in confined migration and its effect in chromatin organization and function.

    Major comments

    Results ‘To distinguish between true changes in chromatin modifications and effects of physical compression of the nuclear content due to deformation, we normalized the heterochromatin mark intensity to the euchromatin mark intensity in each cell.’ - The results are normalized to H3K9ac, with the assumption that its levels do not change during migration/confinement. Has this assumption been confirmed? For example, by normalizing both H3K27me3 and H3K9ac to total H3 instead - and showing that K27me3 increases with confined migration while H3K9ac doesn't.

    Results ‘Increased heterochromatin formation should result in an increased ratio of heterochromatin marks to euchromatin marks, whereas physical compression of chromatin would increase both marks, and thus not alter their ratio…’ - Can some comments be provided on what the meaning would be for heterochromatin to "increase" and euchromatin to not change? There are two ways in which heterochromatin could "increase" - either the portion of the genome in heterochromatin could increase (which would mean the portion in euchromatin would decrease), or the portion of the genome in heterochromatin could stay the same but K27me3 levels could be higher in those regions (which might not affect euchromatin levels). One way to distinguish between these would be to stain for K36me3 as the "euchromatin" marker instead of K9ac - because K36me3 and K27me3 are mutually exclusive.

    Figure 1

    • Could the effects seen be due to cells spending different amounts of time in the channels? Do all cells migrate at a similar speed?
    • Panels D, F, I: it is unclear if the cells shown in the plot for the change in heterochromatin marks are all that migrated or only those that show the difference. Suggest including a dot plot to also show individual data. Can some clarification be provided for how to interpret that controls "before" in 1D and 1F are statistically different?

    Counts in Fig S2A-D are sometimes very low (same applies to Fig 1I, Fig 2B,C,E.), it may be nice to compare some more cells.

    Results ‘Although the effect was less pronounced than in the ≤2×5 μm2 confined channels (Fig. 1C-F)’ - Can the normal size of these cells be reported ? Also the size of nuclei. is it bigger than the pore size?

    There are concerns about the statistical analysis related to SEM and p-values based on multiple measurements or cells within each sample. The t-test and ANOVA assume that each measurement is independent, and multiple cells within the same sample are not independent. Suggest to either not report p-values or average together the values from each sample and calculate the p-value using those sample-level means. For more information, see https://doi.org/10.1371/journal.pbio.2005282. and https://doi.org/10.1083/jcb.202001064.

    Minor comments

    Results ‘custom-made polydimethylsiloxane (PDMS) microfluidic devices with precisely defined constrictions that mimic interstitial space’ - The manuscript report the size of the channels, and notes that it mimics interstitial spaces, it would be helpful to also report the size range for interstitial spaces in vivo.

    Figure IH: Are these the same cells as in the reference (cells in which vertical confinement is sufficient to induce a nuclear response)? Are 5 um channels squeezing the nucleous?

    “significantly larger increase in heterochromatin than cells migrating through the 10-μm tall channels (Fig. 1H, I), demonstrating that the observed effect is primarily attributed to the confinement and not the migration process per se” - There is a statistical difference between the confined migration and non-confined migration groups, but there is also a statistically significant increase in heterochromatin in the non-confined migration group compared to baseline (and with larger sample sizes than in the confined group), so it may be worth commenting on the possibility of the effect of migration alone.

    “Cells maintained CMiH even after completing at least one round of mitosis, without any trend of reversion in their heterochromatin levels (Fig. 2C; Fig. S4A, B), suggesting that the epigenetic modifications were inheritable through DNA replications” - This is an intriguing concept, however, it is unclear whether the cells that migrated did so before or after dividing. To support the claim about inheriting CMiH, it would be relevant to see heterochromatin levels in a mother cell increase after it squeezes through a channel, then the daughter cell (which doesn't squeeze through a channel) having a higher heterochromatin level than the "before" cells. That's not possible with immunofluorescence, maybe the GFP-HP1a could be useful for such a live-imaging approach? Otherwise, if all these "mitotic cells" divided after squeezing through a channel, that could be stated in the text, legend, and/or methods. Alternatively, the conclusion could be nuanced/toned down.

    Figure 3 - The number of samples analyzed in some cases appears small. Suggest showing the data as dot plots to allow interpretation of the sample sizes for each group and the differences between the groups.

  3. Version published to 10.1101/2021.09.22.461293v2 on bioRxiv
  4. Version published to 10.1101/2021.09.22.461293v1 on bioRxiv