PREreview of A novel nuclease is the executing part of a bacterial plasmid defense system

Chauvin Gauthier, Coupet Cédric, Dewailly Marie, Garcia Chloé, Lorendeau Marius, Orlando Mathias, Ramos-Hue Marvin, Rose Virgile, Pouzet Esther, Rose Virgile, Tassoni Marion, Vicente Sylvain, Jousselin Ambre, Redder Peter

Université Toulouse Paul Sabatier, Master Microbiologie Moléculaire

LMGM-CBI, Toulouse

This Pre Peer Review is an exercise performed by Master 2 students from University Toulouse Paul Sabatier (Master Microbiologie Moléculaire). Writing a pre-review from a bioRxiv article is part of their module “Scientific analysis”.

It tooks us 3 sessions to finalize the peer-review:

Session 1: presentation of the scientific edition landscape + article reading

Session 2: critical reading of each figure

Session 3: pooling students peer-reviews

The final text was slightly edited for grammar and syntax but the peer-review reflects the work of the students.

In the manuscript entitled “A novel nuclease is the executing part of a bacterial plasmid defense system”, Weiß et al describe that MksBEFG is a complex functioning as a plasmid defense system, with MksG acting as a nuclease and MksF being the central hub of the system.  MksBEFG is a conserved system among different bacterial species. They determine the structure of the MksG protein by combining low resolution X-ray data at 4.6 Å resolution with high confidence Alphafold2 models. Then they show that MksG possesses a nuclease activity in an in vitro cleaving assay on plasmid DNA. After that, they confirm that MksB has ATPase activity, reinforcing their hypothesis that the complex MskBEFG is a condensine-like protein involved in this plasmid defense system. Then, interactions between the proteins composing the MksBEFG complex are proved by bio-layer interferometry: the interactions tested are those between MksE and MksF, MksB and MksF and between MksG and MksF. Finally, they strengthen their previous results about the interaction between MksB, MksE, MksF, by showing similar localisation of those three proteins within the cell in absence and presence of substrate plasmid DNA.

Major issues

• (1)  As a general concern for all figures, quality was globally too poor and pixelized to a point where some of the key data are unreadable. This impeded the interpretation of some results. 

• (2)  A statistical analysis would be globally appreciated. In particular it could be added in figures 1B, 2C, 4C, 6 and 7, where conclusions have been drawn without mentioning significancy. We noticed several occurrences of the word "statistical" without this analysis.

• (3)  Regarding Figure 3:

(3.1) According to the molecular scale, plasmid size is not the same in gels for mutants E236 and ED279 (CCC DNA migrates until marker band 3 from the top) when compared to wild types (CCC DNA migrates until marker band 4 from the top). Is it a different gel?

(3.2) We would appreciate to see pJC1 in the same conditions using the WT enzyme as it could be important as a control and to gain clues to know whether pJC1 is cut or not, since the Mks system seems “inactive” on pJC1.

(3.3) The wording about the interpretation of this figure needs to be more consistent. In particular, precising when it is supposed to be double-stranded or single-stranded cleaving will help the reader to understand and evaluate the proposed model.

• (4)  Regarding Figure 6:

In our opinion there’s a lack controls. This figure lacks the condition where fluorescent MksB is in presence of pBHK18. It also lacks conditions with pJC1 (the medium copy plasmid) in order to see if there is a different localisation in presence of low or medium copy plasmid. It also lacks the localisation of MksG in cells delta MksE or MksF. 

Finally it lack statistics in order to properly compare the different conditions (D and E) and the similars conditions (A,B and C).

•  (5)  Regarding Figure 7:

Panel C is so pixelated that it is impossible to read what the sub-panel shows. The reader cannot therefore evaluate the data.

Minor issues

• (1)  Regarding Figure 1:

(1.1)      The different loci should have the genes in the same orientation so that the alignment of the sequences will be easier for the reader to understand (eg : mskG pink arrow always on the left hand side in the alignment, instead of right and left).  

(1.2)      For the Fig. 1B, the dots for delta MksG in the pBHK18 condition are missing, perhaps an error during the construction of the figure. Moreover, even though the changes of plasmid copy number is clearly visible, it would be better to add stats significancy between WT and delta in both pBHK18 and pJC1 conditions.  

(1.3)      Finally the Fig. 1C is not very informative for the main story, perhaps it would be better to keep it in the supplementary figures. Moreover it lacks information such as the percentages mentionned in the text.

• (2)  Regarding Figure 2:

Please mention what method were used to obtain those structures in the legends and/or in the text (which parts were obtained by Alpha-Fold and which one were obtains by crystallography). Moreover, in the Fig. 2C, it may be good to keep the color code for the representation of the MskG monomers (in green). Before the superposition figure, it will be clearer for the reader to have the representation of the TopoVI structure separated from MskG. Finally, the data for modeling must be credited in the legend, no references given here. In Figure 2D, please indicate the location of the two putative active sites (the AA highlighted in Figure 2B). Are the two active sites close to the DNA that has been modelled into Figure 2D?

• (3)  Regarding Figure 3:

(3.1) We do not agree to a certain extend to the conclusion of the author in relation with these data. In particular, no emphasis was given to the hypothesis of a nickase activity of the enzyme. The only way to form the OC is via nicking, and MksG must therefore have nickase activity (since a lot of OC is observed). It is even possible that a succession of individual nicks along the sequence is responsible for the linear forms. We suggest that this hypothesis should at least be taken into consideration in the interpretation. A kinetic analysis of the appearance of the OC and linear forms over time would probably resolve this issue.

(3.2) On the first gel, the legend indicates 10mM of ions but not in the other gel. We felt that this was a problem of coherence and should be indicated on each gel or only in the legend. The concentration of the protein could also only be indicated in the legend.

• (4)  Regarding Figure 4:

(4.1) Initially for the whole figure an improvement in quality/readability of the legends and representation will facilitate their understanding by the reader.

For the SDS page of the co-purified proteins in figure 4A, we think it is necessary to explain the method by which this purification was performed in the Mat&Met section. The main information to be added should explain which proteins have been expressed with the His-Tag and its position, which would allow it to reproduce the experiment.

(4.2) Moreover, for the titration assay in figure 4C, we think it could be interesting to test the activity of MksB in presence of MksE since it was only shown with MksB and MksF together. Even if the MksB activity is not increased along with MksE, that would allow us to see the activity of MksB in every possible combination. 

(4.3) Furthermore, in order to qualify the differences observed in figure 4C, a simple Student test should be added.

It is interesting to analyze the activity of the MskB in presence of DNA, as it is done in the second panel, nevertheless, in order to see all the informations, we think it could me more interesting to compare it to the activity of MskB in presence of DNA from pJC1 which is a medium copy plasmid, as the negative control. 

• (5)  Regarding Figure 5:

Firstly, bio-layer interferometry experiments that do not show interactions could be incorporated into the publication, to improve the understanding of the readers, for example, in supplementary data. In addition, it would be interesting to add the dissociation constants in figure as well as in table, in order to promote the interpretation of the data by the reader. At last, it could be interesting to try if MksB and MksF are dimers.

• (6)  Regarding Figure 7:

(6.1) It is not stated anywhere in the manuscript which SMLM was used, assumed STED?

(6.2) We felt the clustering method used to separate into 3 types of movement population is not easily understandable to non specialists

• (7)  Regarding Figure 8:

MksG is not drawn as a dimer as it should

Positive feedbacks

The first figure is a great overview on the MksBEFG complex (how it is organized, and what is it's effect). It helps the readers to understand it better before diving into the article core. More specifically, the results of the 1C figure is convincing and the controls are overall correct. Moreover, the proposed alignments are already known and well referenced. It allow to the reader to make there own opinion.

As for the figure 2, it represents the catalytic site of the MskG protein well and its related function to the topoisomerase VI. The different orientation of the protein MskG are appreciated. It gives a good idea of the protein structure . 

For figure 3 the results are convincing and the controls are adequate. It is clear that we identify on gel the shift from super-coiled plasmidic DNA to open circular and linear reporting the nuclease activity of MksG that is notably stronger with Mn2+. Mutations in the amino acids supposed to be important to the catalytic site indeed decrease the ratio of linear DNA in favour of open circular and super-coiled DNA.

The results of figure 4 are overall satisfactory and convincing, the experiments are well done and the interpretation associated with the panels is clearly written and accessible even to a non-specialist audience. According to me/us, no major modification is to be realized on this figure.

Concerning figure 5, we are not experts on the "bio-layer interferometry" technique but we find this approach to be very relevant. Furthermore, the results of this experiment complement quite well the results described in the figure 4. The data in figure 5 appear to be correctly interpretable and correctly interpreted.

In regard to the figure 6, we found that the demograph is a good plot in order to represents the cellular localisation and density of focci. The epifluorescence is also a good mean to show those results especially by showing on the first panel the phase contrast by comparison and by putting arrows on the foci. 

Competing interests

The author declares that they have no competing interests.