A Magnesium Transport Protein Related to Mammalian SLC41 and Bacterial MgtE Contributes to Circadian Timekeeping in a Unicellular Green Alga
Round 1
Reviewer 1 Report
This reviewer is pleased with most of the corrections the authors have made to the manuscript. There are a few small remaining points that would benefit from improvement, particularly with regard statistical analysis.
1. For consistency, I recommend statistical analysis to be incorporated into Fig. 1C. One wonders whether or not the differences in mRNA levels differ significantly between ZT in the overexpressor line.
2. This reviewer is happy with the incorporation of interesting data from the drawer for the new Fig. 4. However, there is no statistical test for panel C allow conclusions to be reached about the differences in periods, so please can this be added. Also, the period estimation for the parental line at 50 mM Mg2+ (control) is around 22 h for the parental line and 24 h in the overexpressor line in this figure but in Fig. 1B the parental line has a period estimation of 24 and the overexpressor of 26. Why does this variability occur? I would suggest adding a few explanatory lines to the Discussion in case readers are wondering about this.
Very minor
1. In Fig. 3, letter the first panel is incorrectly labelled C rather than A.
Author Response
Thank you for your additional feedback on our manuscript.
We have now provided statistical analyses in Figure 1C: mRNA levels do differ significantly between ZT in the overexpression line, reflective of added values of mRNA from the wild-type and overexpression copies of the gene.
We have also provided the statistical analyses of data in Figure 4C: in both lines, severe magnesium depletion induces long period rhythms compared to the controls whereas intermediate levels do not. We have also added a sentence to the results section describing Figure 4 to explain that we do observe period plasticity between different experiments in Ostreococcus (P7: L211-214), most likely due to batch-to-batch variation in the growth media that we prepare from scratch. Even though total salinity is measured and corrected in each batch, the relative composition of ions might vary. Additionally, the age of the culture also influences period length. It is therefore important to always compare relative effects of treatments versus the control within every experiment. The period lengthening observed in the overexpression line versus the parent is consistent between Figures 1 and 4.
Lastly, we have amended the panel label in Figure 3.
Reviewer 2 Report
It is much clearer what the authors propose with the research but still need to show more data to fully explain the conclusion.
1. The effect magnesium on the circadian period is still debatable because there is only one publication supporting the hypothesis. Therefore, to find the effect of magnesium transporter or inhibitor on the circadian rhythm, both the cellular magnesium concentration and the circadian rhythm should be measured with all cell lines and the inhibitor. The correlation of two variables (magnesium concentration and period) should be explained to clearly conclude the effect.
Author Response
The main conclusions of our study are that overexpression ofOtMgtE increases circadian period and removes the trough of intracellular Mg2+at dawn. While it is plausible that the two effects are mechanistically linked, we do not provide direct evidence that the effects of overexpression on circadian period is due to the effect on intracellular magnesium. For that reason, we do not claim a direct mechanistic relation in the manuscript. The conclusions we do draw on our results are therefore supported by the data, as reviewer 1 agrees.
Round 2
Reviewer 2 Report
The manuscript by Feord et al. touches on the effect of magnesium on the circadian clock. Authors developed a trans-genic alga to overexpress the magnesium transporter. Some of the inhibitors are used to observe the relationship between the magnesium concentration and the circadian rhythm. The conclusion is still unclear.
1. The mechanism of the magnesium effect on the circadian period is not elucidated yet (There is only one publication). Therefore, monitoring circadian rhythm to observe the effect of magnesium transporter or inhibitor on the cellular magnesium concentration is not a good method. Measuring the cellular magnesium concentration directly for the inhibitor experiment will give clear result to explain the effect of the inhibitors. I think the information is not enough to develop a new biological or physiological mechanism. It needs more detailed hypothesis and experiments to lead a novel conclusion.
Final Decision Comments
Dear Gerben,
Thank you for submitting the 2nd revised version of your paper. Comments and suggestions of Reviewer 1 are now fully addressed and I accept your arguments regarding the major concern of Reviewer 2. Thus, we accept your manuscript for publication in Genes, within the special issue on circadian rhythms.
On the other hand, I believe that suggestions of Reviewer 2 are valuable and helpful to design meaningful experiments revealing novel aspects of this emerging topic.
Best regards,
Laszlo
This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.
Round 1
Reviewer 1 Report
In this paper “A magnesium transport protein related to mammalian SLC41 and bacterial MgtE contributes to circadian timekeeping in a unicellular green alga,” the authors describe the effect of the overexpression of a magnesium transporter domain-containing protein upon the circadian rhythm in the algal model system Ostreococcus tauri. Overall, this paper contributes to the field by dissecting potential roles for the magnesium channel in circadian rhythms though genetic and pharmacological approaches. This is an interesting piece of work that needs some revisions to the text and statistical methods to clarify the interpretation of the data. It also seems that some further (computational) additions will improve the confidence in deciding whether or not SLC41 might be a Mg2+ transport protein. Such revisions will help provide a clearer basis for subsequent studies in this area (e.g. drop assays and protein localization to find out about the protein). These revisions do not involve further experimentation but do need some re-examination of data.
MAJOR REVISIONS
1. The mathematical methods used for period estimation are not explained in the paper. Please can the authors add information about the algorithm that was used and the tools used to perform the analysis (e.g. BRASS, biodare, etc). If FFT-NLLS was used, is it worth presenting also the RAE values for the data to allow assessment of whether there were alterations in the rhythmic robustness of the data?
2. Revisions for Fig. 1
a) Individual data points are not apparent within panel A. Are these continuous data or individual points at specific time intervals connected by lines? Please can the SEM bars be made clearer also.
b) OtMgtE overexpression lengthens the period significantly but no mathematical method for period estimation are explained.
c) The use of “relative expression” is not accurate on the y axis. It would be more accurate to say “relative mRNA abundance” or “relative mRNA expression”.
3. Revisions for Fig. 2
Statistical comparisons are not consistent. The OtMgtE ZT12 bars are not compared to the rest of the dataset, therefore an ANOVA comparison would be more informative and would be statistically more appropriate than t-tests.
Line 161-163 states: “Although the relative effective size is variable between the different assays and replicate experiments, overexpression of OtMgtE consistently heightens [Mg2+]i at ZT0 between experiments and assays.” This is not correct since Fig. 2a comparing the overexpressor vs parental line at ZT0 is not significant (see Fig. 2a). The authors should take advantage of analysing the entire dataset with ANOVA to provide more depth and nuance of interpretation of what is happening here. For example, is OtMgtE suppressing Mg oscillations throughout the day?
4. Revisions for Fig. 3
In panels a and b, individual data points are not clear and the plots appear to lack SEM bars for all points (is the variability very small, one wonders?). Panel c needs some form of statistical analysis to compare the parental and the overexpressor lines. In panel c, what are the units on the y axis? Should the y axis be labelled “period lengthening relative to…”?
CHA is a competitive inhibitor of Mg2+ transport (Dalmas et al. 2014) mediated by corA transporters, it blocks the entrance of the transporter, therefore the effect it has would be to decrease [Mg2+]I, how do the authors reconcile this with their findings from OtMgtOX? If overexpressing OtMgt increases Mg2+ influx, presumably that would shorten period if CHA blocks influx. On the other hand, the interpretation might be different if OtMgtE mediates Mg2+ efflux? There are some points that could be expanded upon in the Discussion to help interpret this and provide a basis for future work.
REF: Olivier Dalmas, Walter Sandtner, David Medovoy, Ludivine Frezza, Francisco Bezanilla, Eduardo Perozo. Magnesium selectivity in CorA. Proceedings of the National Academy of Sciences Feb 2014, 111 (8) 3002-3007; DOI: 10.1073/pnas.1319054111
5. Potential revision for Fig. 4
Why does Amphimedon queenslanda appear in two branches? If this is a faithful representation of the phylogenetic relationships between the proteins, should they not generally cluster? I wondered whether it might be helpful to collapse some of the branches in the tree. Does Arabidopsis carry any putative Mgt-domain proteins (an obvious question, given the conservation of the Arabidopsis and Ostreococcus circadian systems)?
6. The authors need to provide some more information about the OtMgtE protein to help readers understand its level of conservation with MgtE domain proteins that are known to be membrane transport proteins. I recommend two additional bioinformatic analysis; (1) an alignment of the protein sequences of the functionally-important conserved domains of OtMgtE with those of several known MgtE domain transport proteins, showing the amino acid alignments and conservation within those important domains; (2) an analysis of the extent to which the protein sequence of OtMgtE is likely to form a protein with membrane spanning domains. This might be conducted using one of the tools that exist to assess how the hydrophobicity/hydrophilicity of the protein might make it fold through a membrane. If suggestion (1) works well, I recommend it is included early in the paper, whereas (2) could go either in the paper or SI depending on the result. Both should be described and interpreted.
MINOR REVISIONS
1. Overexpression of a Mg transporter domain protein indeed lengths the period, however, there is no evidence that this is due to Mg2+ concentration changes or oscillations during the day. The authors must be very careful in their Discussion to not force this link. This is one possible interpretation, but would need further experimental work to be sure. For example, it is alternatively possible that OtMgtE-ox might cause a stress response that alters Mg2+ and the clock, but the three are not linked mechanistically. Please review and revise the text to check for this over-interpretation and add other possible interpretations where it seems approrpriate. Point 2 below is a specific suggestion to help with this.
2. lines 275-276. The conclusions (“…the algal homolog affects cellular timekeeping via Mg2+ transport”) needs to be toned down. I suggest as an alternative, “…overexpression of an algal MgtE-domain containing protein alters circadian period. One interpretation is that this change in timekeeping is due to altered Mg2+ transport.” This could be a good place to suggest other possible interpretations also.
3. If the authors happened to have any extra data derived from effects of general chelators of [Mg2+]i upon the rhythms, this could be interesting to include. I am not asking for this experiment to be done, but saying that it could be a good addition if the data existed already. Specifically, does adding such a chelator to the overexpressor line decrease the period back to that of the WT levels?
4. Add a literature citation to the sentence about mTOR on lines 57-59.
5. Please review the English/editing of the entire paper. There are some sections that might have become over-edited and are difficult to follow. Perhaps ask a colleague to read it and check for readability.
Reviewer 2 Report
The manuscript by Feord et al. touches on the effect of magnesium on the circadian clock. Authors developed a trans-genic alga to overexpress the magnesium transporter. Some of the inhibitors are used to observe the relationship between the magnesium concentration and the circadian rhythm. The conclusion is still unclear.
1. The mechanism of the effect magnesium on the circadian period is not elucidated yet. Therefore, monitoring circadian rhythm to observe the effect of magnesium transporter or inhibitor on the cellular magnesium concentration is not a good method. Measuring the cellular magnesium concentration directly for the inhibitor experiment will give clear result to explain the effect of the inhibitors.
2. There are no novel findings in this article. They found the OX strain was consistent with the previous findings and the mammalian inhibitor did not work for the algae. I think the information is not enough to develop a new biological or physiological mechanism. It needs more detailed hypothesis and experiments to lead a novel conclusion.
