This manuscript by Husser and colleagues describes a new tool kit created by the authors to allow better analysis of cytokinesis in the context of mammalian tissue culture cell lines that are frequently used for cell biology studies: HEK293, HeLa, MDCK, HCT116, and HepG2. The goal was to tag three proteins involved in cytokinesis at their endogenous locus with mNG using CRISPR/Cas9. Additional fluorophores were also added. Anillin and Ect2 were successfully tagged at their N-terminus, and could be grown as homozygous strains. RhoA tagging at the 5’end created a line that had to be kept as a heterozygote.

In HeLa cells, the expression of mNG-Anillin and mNG-Ect-2 showed recruitment to the plasma membrane, in particular to the equatorial cortex, and in the case of Ect-2 also to the chromosomes, at specific times in the cell cycle. RhoA shows more diffuse pattern, and I was a bit unclear if we are viewing only the heterozygous strain.

Comparing the expression pattern of Anillin in different cell lines shows differences in many aspects of cytokinesis, and differences regarding when in the cell cycle Anillin is recruited to the cortex. Anillin accumulates at the cortex before anaphase in some cell lines, but not others, like HepG2 cells, which undergo slow cytokinesis. Anillin enrichment at the equatorial cortex is evident by anaphase in all the cell lines, with differences in the breadth (width) of the furrow accumulation. Cells that divide quickly, like MDCK and HCTT6, seem to have a larger breadth (broader width) of cortical Anillin than cells that divide slowly like HepG2 and HEK293, though it is harder to see this in the HEK293 example shown.

One underlying purpose for the study is to allow comparison of cytokinesis regulation in different types of mammalian cell lines. The authors identify some intriguing differences in how Anillin is recruited to the cortex, and in cytokinesis timing. MDCK cells are more likely to set up a highly asymmetric furrow, even when grown under non-confluent conditions. The implications of this asymmetry are not fully explained, but seem worth following up on.

This study, which creates useful resources for the cytokinesis community, raises intriguing questions that can be further explored. A model is proposed for how cell size, and ploidy, may regulate the recruitment of Anillin to the cortex. In the future, it will be interesting to view localization of importins, to see if their expression supports the model. It will also be interesting to view other RhoA effectors, relative to these proteins, especially non-muscle myosin 2.

This study may be of broader appeal beyond the cytokinesis community if a few issues could be addressed. The first suggestion to strengthen this story and broaden its appeal is to better emphasize the context regarding why variations in cytokinetic patterns could offer important insights in general for cytokinesis studies. Specific suggestions to strengthen or clarify the manuscript are given below.

Abstract: Seems quite broad. To focus it, is there a specific example that could be cited for why basing our models on over-expressed proteins in HeLa cells has been bad for the field?

p.9 “The most popular tool for gene editing is CRISPR….” the latest? currently more useful?

p. 13 line 203, first mention of AAVSI locus, please define

p.13 endogenous tagged expression is generally lower than over-expressed transgenes. Is there an example where a result based on over-expression was found to be wrong or misleading?

Fig. 2: Comment: Over-expression also seems to cause cell rounding, less adherent appearance. Is what is shown the typical level used when these proteins are studied? i.e. are you showing us how over-expression alters cell appearance? It may be worth commenting on this, and even quantifying the difference.

Fig. 3.C.; RhoA seems over-exposed throughout, making it harder to appreciate the subcellular distribution. You do not need to use same exposure for three different proteins here, since the goal here is showing the distribution, not the levels.
Also: are we looking at a heterozygous line for RhoA?

Fig. 4. Clearly shows big differences in the timing, that correlate with peaks of Anillin at cortex. In some cells, even if anaphase has started, and Anillin is at cortex, a long time is needed before cytokinesis occurs: MDCK: 4 hours; HCT116: < 8 hours; HepG2, 20 hours; HEK293, 20 hours.

Fig. 5: The point being made is that levels do not explain “efficiency” but Fig. 5D seems to show that levels prior to anaphase correlate with speed; high cortical levels in HCT116 result in faster cytokinesis. Any comment on this?
The main point did not jump out from the figure, since none of the graphs matches 5B: duration of ingression. I could follow the argument from the text, but the figure could be interpreted differently. I don’t understand why MDCK cells ingress quickly, if they have
a narrow furrow breadth. Could this be due to how “breadth” was measured? Maybe MDCK cells are more broad, if the asymmetry is taken into account?

General suggestion: add rationale for why diversity in cytokinesis may be biologically meaningful. Example, p. 19, line 322 “To reveal cytokinetic diversity…”
First introduce why cytokinetic diversity is meaningful or important, then give the data.
Are there arguments or conflicting models in the field that variation in protein localization would illuminate? Is there confusion in terms of disease studies, where a better understanding of cytokinesis could better explain phenotypes, or suggest treatments?

Discussion

p. 24 line 409 missing word: for the first….(time)

The issue of ploidy is raised several times, in part to explain differences in expression in the endogenous lines. It would be helpful to give some more context here. What is the known ploidy of these cell lines? Is it highly variable, or are some consistently of much higher ploidy?

NOTE: tagged RhoA is referred to as “active RhoA” however, that is not correct.
Tagging RhoA reveals total RhoA. One possible way to compare the pattern of RhoA and active RhoA would be to compare RhoA to an endogenously tagged target of active RhoA. Based on Figure 1, that would include Anillin. Figure 3 does this, except for the comment above about over-exposed RhoA.