Coordination of tissue cell polarity by auxin transport and signaling

As you can see from their comments pasted below, both reviewers are very positive about your manuscript. There are a number of revisions, however, as outlined below, and I would like to ask you to submit a revised manuscript along these lines.

Please pay particular attention to the writing. The way the manuscript is written now it is rather tedious to read and could be much improved by eliminating repetitions, summarising conceptual similar experiments, while streamlining the text at the same time.

We thank the Editors for their constructive feedback. As detailed below in our response to the reviewers’ comments, we have restructured the Results section under new headings and sub-headings to bring out the connection between its different parts. Further, we have shortened and focused all the sections in the manuscript – in particular, we have shortened the Results section by ~40%; some parts were moved, some were ruthlessly deleted. Finally, we have consolidated the original 13 core figures into 10 revised ones.

Reviewer #2:

[…] I am very enthusiastic about publication after the following issues have been addressed.

1) The paper lacks flow and is difficult to read. Part of this is because the experiments involve many somewhat complex genotypes. However, the presence of many repeated phrases and general repetitive style sometimes make for rough going. In other cases, there is a lack of clarity. These are important results and I think the authors should work to make them more accessible.

We thank reviewer 2 for their stimulating feedback, based on which we have made the following changes.

- We have presented the hypothesis and its predictions only once, at the beginning of the Results section, under the heading “Testable Predictions of the Current Hypothesis of Coordination of Tissue Cell Polarity and Vein Formation by Auxin”; this solution eliminates unnecessary repetitions.

- We have revised headings to describe better what the reader will find in each sub-section of the Results; this solution provides structure and moves the text at a steady pace in the direction the text is supposed to go.

- We have focused on results as opposed to data; we have moved information that is peripheral to the main message of the manuscript; and we have deleted whatever was not essential to that message; this solution has enabled us to sharpen the focus of the manuscript.

The combination of these changes has enabled us to shorten the Results section by ~40%; it has also allowed shortening, though to a lesser extent, the Introduction and Discussion sections.

2) I think it would be useful to have a figure that illustrates the dynamic movement of the PINs during vein development.

We agree with reviewer 2 that understanding the expression and localization during vein development of PIN2, PIN3, PIN4, and PIN7 (we assume that is what reviewer 2 had in mind – PIN1 expression and localization have in fact already been the focus of at least four papers: Scarpella et al., 2006; Wenzel et al., 2007; Bayer et al., 2009; Marcos and Berleth, 2014) would be interesting; however, as clarified below, we believe such understanding does not belong in this manuscript.

The goal of our study was to test the prevailing hypothesis of coordination of tissue cell polarity and vein formation by auxin. Simultaneous mutation in all the PIN genes with vein patterning function – including PIN3, PIN4, and PIN7 – fails to lead to defects that fall within the vascular phenotype spectrum of gn. Furthermore, the vein pattern phenotype induced by simultaneous mutation in all the PIN genes with vein patterning function – including PIN3, PIN4, and PIN7 – fails to be epistatic to the gn vascular phenotype. Therefore, abnormal expression and localization of PIN1 – and of all the other PIN proteins with vein patterning function, including PIN3, PIN4, and PIN7 – cannot be the cause of the gn vascular phenotype. As such, the expression and localization of PIN3, PIN4, and PIN7 during vein development – as interesting as they may be in themselves – are irrelevant to the goal of testing the prevailing hypothesis of coordination of tissue cell polarity and vein formation by auxin.

As for PIN2, the normal vein pattern of pin2 (Sawchuk et al., 2013), and the inability of pin2 to enhance the vein pattern phenotype of pin1 (Sawchuk et al., 2013) and pin1,3;4;7 (this manuscript) suggest that PIN2 expression in the leaf epidermis is inconsequential for vein patterning and thus beyond the scope of our study.

Incidentally, a description of the different aspects of the expression and localization of PIN1 during vein development has required at least four manuscripts (Scarpella et al., 2006; Wenzel et al., 2007; Bayer et al., 2009; Marcos and Berleth, 2014); an acceptable study of the expression and localization of PIN2, PIN3, PIN4, and PIN7 during vein development cannot require fewer than four figures. As such, we would like to suggest that the addition of four figures to a manuscript that already has – after reduction – 10 information-rich figures, 1 summary figure, 23 figure supplements, and 19 supplementary files (now consolidated into two supplementary files) would only be justified if the new evidence were essential to the main conclusion of the manuscript, which – as detailed above – in the case of our manuscript we believe it is not.

3) The authors’ interpretation of their results depends on the fact that many of the alleles employed are nulls (particularly the pin alleles).

We thank reviewer 2 for the opportunity to clarify this important concept. The alleles used in our study have been well-characterized elsewhere (references in Supplementary file 1 – Key Resources Table) or in this manuscript, including the newly added Figure 5—figure supplement 3, and are either null or the strongest alleles available (except, of course, for the weak and intermediate alleles of gn and for axr1-3, which are partial loss-of-function alleles and were used precisely for that reason). Most important, however, our conclusions do not depend on complete loss of gene function. For example, were the gn vascular phenotype caused by loss of PIN-mediated polar auxin transport, simultaneous mutation of PIN genes with vein patterning function would lead to defects that fall within the vascular phenotype spectrum of gn. This would be the case irrespective of whether those pin mutations led to complete loss of function or to partial loss of function: in the former case, the vascular phenotype of that multiple pin mutant would resemble that of strong gn alleles; in the latter, it would resemble that of weaker gn alleles. Neither prediction is supported by our results; nevertheless, we realize our original text was misleading. We have now revised it to explain more clearly the testable predictions of the hypothesis and the conclusions derived from the results of the tests of those predictions.

4) In some cases, it would be useful if conclusions were stated more clearly. For example, instead of stating that such and such as combination of alleles do not phenocopy gnom, they could say that the phenotype of this allele combination is much less severe than gn, implying that GN is doing something besides mediating PIN localization.

We thank reviewer 2 for their suggestion, which we took into account as we revised the Results section according to the new structure inspired by the comments of reviewer 2 (see our response to comment no. 1 of reviewer 2).

5) Is it correct to say that GN is "controlling" PIN localization, rather than being required for this process?

We believe both of them are correct. Examples abound in literature of the use of the term “control” in the way we have used it; a recent example is the report that retromer in Drosophila controls the localization of core planar polarity proteins (Strutt et al. 2019. Curr Biol. 29: 484–491). In that study, just as in ours, such control is inferred from abnormal protein localization in loss-of-function mutants.

6) In the subsection “Contribution of the GNOM Gene to Coordination of Tissue Cell Polarity During Arabidopsis Vein Formation”, the authors introduce several different gn alleles. What are we to take from this?

We realize we had not been clear about this, and we appreciate the opportunity reviewer 2 has given us to clarify this important point. As it should now be clear from the revised Results section – and as also mentioned in our response to reviewer 2’s comment no. 3 – the prevailing hypothesis of coordination of tissue cell polarity and vein formation by auxin predicts that, were the vein pattern defects of gn the result of loss of polar auxin transport, auxin transport inhibition would lead to defects that fall within the vascular phenotype spectrum of gn. Therefore, to test this prediction it is essential to understand what the vascular phenotype spectrum of gn is; and to understand what the vascular phenotype spectrum of gn is, it is essential to determine the vascular phenotype of an allelic series in the GN gene, which is what we had presented in Figure 2 and had described in the text reviewer 2 refers to.