PAT mRNA decapping factors function specifically and redundantly during development in Arabidopsis

  1. Zhangli Zuo
  2. Milena Edna Roux
  3. Yasin F. Dagdas
  4. Eleazar Rodriguez
  5. Morten Petersen

Reviewed by

Read the full article

Listed in

Log in to save this article

Abstract

Evolutionarily conserved PAT1 proteins activate mRNA decay through binding mRNA and recruiting decapping enzymes and other factors hence optimize transcriptional reprogramming during development. Here, we generated multiple mutants of pat1 (Protein Associated with Topoisomerase II), path1 and path2 and inspected their growth and leaf morphology phenotype. pat triple mutants exhibit extreme stunted growth and all mutants with pat1 exhibit leaf serration while mutants with pat1 and path1 all display short petioles. All 3 PATs can be found localized to Prossessing Bodies (PBs) upon auxin treatment and RNA-seq analysis indicate that all 3 PATs redundantly regulate auxin responses. Moreover, shade avoidance and NAC genes are misregulated in pat1path1 double and pat triple mutants suggesting PAT1 and PATH1 function in petiole elongation and leaf patterning. In conclusion, PAT proteins exhibit both specific and overlapping functions during different stages of plant growth and our observations underpin the importance of the mRNA decay machinery for proper development.

Article activity feed

  6. Review Commons

    Note: This rebuttal was posted by the corresponding author to Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Reply to the reviewers

    __Reviewer #1 (Evidence, reproducibility and clarity (Required)): __

    The manuscript investigates the role of PAT1 gene family in Arabidopsis thaliana. Though the PAT1 protein has been previously investigated and displayed immune-related and developmental phenotypes, the other two members of the family, PATH1 and PATH2, have not been well studied. The authors set out to understand the role of these proteins in relation to the role of PAT1. They thus generated single, double, and triple mutants of the possible combinations of PAT1 genes and examined their phenotypes. As the study focused on the developmental effects of PAT1, the mutants were generated on the background of the summ2 mutant to avoid phenotypes related to immune response. The authors notice a developmental difference between the pat1 mutant combinations, suggesting that PAT1 acts differently than PATH1 and PATH2 and that the PATH proteins serve a redundant function. They also performed RNA-seq analysis to identify differentially-regulated genes in the mutant combinations. The study is interesting and well-executed, yet I believe some questions should still be addressed:

    __Our response: __We thank the reviewer for acknowledging the significance of our findings. Please see our detailed answers to the reviewer’s suggestions in the following.

    The research mainly focuses on the developmental phenotype of pat mutants but also tests the interaction of PATH proteins with RNA decapping enzymes to check their function and localization during different treatments. I found it a bit confusing since Figure 1 also shows the developmental phenotype of the mutants. I think editing the order of the figures would make the overall story more coherent.

    __Our response: __We agree with the reviewer thus we moved old Fig 1C to new Fig 3A, we believe the new figure orders make the overall story more coherent.

    My main concern is the correlation between the developmental phenotype of the mutants and the gene expression. Leaf samples for RNA extraction were taken when the plants were 6 weeks old, and the developmental phenotype is very evident. It is thus not possible to tell whether the differences in gene expression are a cause or effect of the developmental phenotype. I think performing qPCR of selected candidates at earlier developmental times might help solve this issue, as well as the characterization of younger plants for the developmental phenotypes (such as leaf number).

    __Our response: __We followed the reviewer’s suggestions and performed qRT-PCR on IAA19,* IAA29, SAUR23 and PIL2 *in pats mutants under different developmental stages (Line 162, 169; Fig S4), we also characterized leaf number of pats mutants from younger stages (Line 109; new Fig 3C).

    Overall, the manuscript is missing data regarding replicate numbers in the IP and confocal microscopy experiments.

    __Our response: __We thank the reviewer for pointing this it out, the replicate numbers are provided now in our new figure legends.

    Minor comments:

    Figure 1C - the authors should add a picture of Col0 plants as well as the mutants.

    Our response: To be reader friendly, the picture of Col-0 plant is added in Fig S1A. For the reviewer’s information, plant pictures in FigS1A and old Fig1C (new Fig 3A) were taken at the same time.

    Figure 3 - Calculating the leaf-to-petiole ratio in the different mutants would be good.

    Our response: We now calculate PBR (petiole blade ratio) of all pats mutants in Fig3F (Line 121).

    Figure 4 - the details in the figure are very unclear, especially in the PCA. It would be good to display the data in 2D for PC1 and PC3 and change the colors a bit.

    Our response: We agree with the reviewer; thus, we remade the PCA plot from RNA-seq reads data in a 2D style and also changed the colors for each mutant (Fig 4A). We need to point out that the PCs number also changed because the old PCA plot were made by mistake from expression data.

    Reviewer #1 (Significance (Required)): Both PATH proteins have been less investigated than PAT1, and in that sense, the work is novel. However, it seems that most of the phenotype is attributed to PAT1 rather than the other family members, limiting the interest to the broad plant science community.

    Our response: We appreciate the reviewer think our work is novel. We agree that PAT1 plays the main role during plant development (old Line 171), however the pat triple mutant exhibit the most severe dwarfism as well as the most mis-regulated genes compared to any single or double mutants, indicating all 3 PATs are essential for development.

    __Reviewer #2 (Evidence, reproducibility and clarity (Required)): __

    Zuo et al., characterize the role of three cytoplasmic mRNA-decay activator proteins PAT1, PATH1 and PATH2 in the context of plant development and leaf morphology in Arabidopsis thaliana and Nicotiana benthamiana. The authors show that the triple pat mutant displays the most severe dwarfism of all combinatorial mutants. Through treatment with different stimulants the authors found that only IAA treatment induces the three homologues to form condensates (possibly PBs), while PAT1 forms condensates upon every tested stimulus. An extensive RNA seq experiment revealed miss-regulation of several hundred genes in the higher order mutants, several of which were involved in auxin responsive and leaf morphology determinant genes.

    __Our response: __We thank the reviewer for the peer review. Please see our detailed answers to the reviewer’s suggestions in the following.

    Major points: 1.Title is not meaningful as is and, in my opinion, does not reflect the main findings in the manuscript.

    Our response: We now changed our title into “PAT mRNA decapping factors are required for proper development in Arabidopsis”.

    The results section could benefit from improved flow between the paragraphs and more reasoning for the next steps taken to help readers understand the aims of the authors.

    Our response: We followed the reviewer’s suggestion and modified the wording in our result part(Line 79,81,94,146-151).

    L46: "So far little is known about the functions of these three PATs in plant development.", The authors themselves have studied these proteins in the context of seed germination and ABA control, as well as apical hook formation and auxin responses. Should at least be mentioned and the results discussed in this context.

    Our response: We thank the reviewer for noticing our other work and we now included this information in the new introduction and discussion part (Line56&237).

    What are the expression levels and patterns of PATH1 and PATH2 compared to PAT1? Is anything known about spatial or temporal regulation of these proteins?

    Our response: All three PATs are expressed in roots, stems, leaves, flowers, siliques, and seeds during the whole developmental stages, PAT1 has higher expression level in leaves but lower expression levels in petals. (Klepikova et al., 2016;

    https://www.arabidopsis.org/servlets/TairObject?id=138009&type=locus for PAT1; https://www.arabidopsis.org/servlets/TairObject?id=38646&type=locus for PATH1 and https://www.arabidopsis.org/servlets/TairObject?id=128694&type=locus for PATH2).

    Figure 1:

    o I do not agree that the authors have shown that "PATH1 and PATH2 are also mRNA decapping factors", rather that these proteins can co-localize (and possibly interact) with LSM1. Decapping assays for example with the known PAT1 de-capping targets from their previous work and their extensive mutant collection could be used to test this.

    Our response: We thank the reviewer for pointing it out and reminding us about the characterized mRNA decapping target from our previous work, we now include the decapping assays in new Fig5 (Line 197).

    From the BiFC experiment (Figure 1B) it looks like PATs are mostly soluble in the cytoplasm (like LSM1) and might be stress-induced components of PBs (like LSM1). Do PATs co-localize with other canonical PB markers that are more prone to condensation, like DCPs or VCS? BiFC could be performed after IAA treatment to confirm that the cytoplasmic foci are indeed LSM1-positive PBs.

    Our response: We agree with the reviewer that PATs behave more like LSM1. Given time limit of the project, we unfortunately are not able to check the colocalization of PATs with DCPs or VCS. However, we performed BIFC after IAA treatment, and the cytoplasmic foci are indeed LSM1-positive foci (new Fig1B).

    A: please provide uncropped images of all Western blots in supplemental data.

    Our response: To be reader friendly, we decide to show the original western blots here (see in the file named "RC-Full-revision"), instead of in supplemental data. However, we will leave the final decision to the editor.

    I applaud the authors for establishing this great higher order mutant collection that will be very useful for researchers in the field. However, I am confused about the description of these mutants. If I understood it correctly, these mutants were already used in a previous study by the authors, namely “Zuo, Z., et al., Molecular Plant-Microbe Interactions, 35(2), 125-130.” & Zuo, Z., et al., (2021). FEBS letters, 595(2), 253-263.” In this study the authors refer to a BioRxiv “Zuo, Z., et al., (2019).” As the reference for these Arabidopsis lines. Is this current manuscript a continuation of the BioRxiv? Please elaborate whether these lines have been used and described In previous studies.

    Our response: We truly appreciate the reviewer for acknowledging the significance of our work. These pats mutants have been used in the FEBS letters paper (2021), MPMI paper (2022), and the new published paper in Life Science Alliance (2023, but preprinted in BioRxiv 2019 and 2022). However, they have not been fully described or characterized in any of the mentioned published stories. Characterization of these pats mutants were originally only included in preprint 2019 which was cited in FEBS letters paper (2021) and MPMI paper (2022).

    L72: Is the strong developmental phenotype of the higher order mutants persistent under long day conditions? Considering the strong developmental phenotypes of the mutants, the flowering transition and morphology could be an interesting trait to study. Why did you choose short day conditions for this study?

    Our response: The pat triple mutant also has strong developmental phenotype under long day condition and exhibits early flowering phenotype. We are currently preparing a manuscript regarding mRNA decay and flowering. We did not “choose” short day condition, we just started with short day condition and observed phenotypical differences hence we kept this condition.

    L78: This statement is hard to see in Figure 1C and best described for Figure 3A.

    Our response: We now change this statement for Fig 3.

    L82: Please include a reasoning for testing PATs localization after hormone treatment. Do you have any indication that other PB proteins behave similar to either PAT1 or the PATHs after hormone treatment to substantiate that these foci observed are indeed PBs? What is known about PBs after hormone treatment in planta?

    Our response: We were interested in investigating if all three PAT proteins may also form PBs in Arabidopsis thus we tested PATs localization with/without hormone treatment (old Line 84, new line 81). For the reviewer’s interest we also observe LSM1 localization after hormone treatment (Fig 2). PBs have been published to respond to light, cold treatment, PAMPs, ABA, ACC and auxin (Line 39-42).

    Figure 2:

    o How does the localization of LSM1 change under the same treatments? Does ist behave like PAT1 or the homologues?

    Our response: Please see our new Fig 2 for LSM1 localization, and it behaves more like PAT1.

    Which part of the root was imaged for this experiment? Is it possible that the observed foci are ARF-condensates as reported by Jing et al., 2022? Do you observe a gradual change in numbers or morphology along the root?

    Our response: We use root elongation zone for this experiment. We don’t know if the foci are ARF-condensates, but it’s possible to study in the future. If the reviewer is interested, we are happy to share our materials. We do observe more foci in the cell division zone and less in the mature zone.

    How did the authors decide on the concentrations for the stimulant treatments? Have you tried different doses, and could the responses be dose-dependent?

    Our response: We did not try different doses; we searched for and applied the commonly used concentrations for different hormones.

    A representative image is not sufficient for quantitative responses, like RNA granule condensation. Please provide a quantification of stimulant-induced foci after the different treatments.

    Our response: Please see the quantification in our new Fig 2.

    L91: Does that mean that most co-precipitated signal comes from the soluble fraction and not PB-localized? Would an RNAse treatment step eliminate the co-precipitation (optional)?

    Our response: We believe it means LSM1 and PATs are in the same complex regardless of PB localization.

    L92/93: Or alternatively that PAT1 localizes to PBs independent of the stress, while PATHs are signal-specific PB components?

    Our response: We think PAT1 aggregates upon broad stimuli/stress, while PATHs respond to specific/limited stimuli, for example, auxin.

    Figure 3:

    o I wonder if these results fit better in conjunction with Figure 1, either as a combined figure or move before Figure 2.

    Our response: We agree with the reviewer thus we moved old Fig 1C into Fig 3.

    It is interesting that path2/pat1, while being dwarfed, is less serrated compared to pat1 or path1/pat1. Can you find any indications in your RNAseq set which genes might be involved?

    Our response: ANAC016 might be involved, but more research needs to be done to confirm it and this is not the focus of the current project.

    Indicate statistical test used to determine p-value

    Our response: We now indicate the statistic test in Materials and Methods part (Line 369).

    L116/L117: Doesn't the result in Figure 3E indicate that PATH1 and PATH2 are not fully redundant, but that PATs have specific and narrow roles in leaf development? L116 goes against your statement in L150 & L160. What is known about the expression patterns of PAT1, PATH1 and PHATH2?

    Our response: We agree and thus modified our statement (Line 137). All three PATs are expressed in roots, stems, leaves, flowers, siliques, and seeds during the whole developmental stages. Please also see our answer to major comment #4.

    L123: PC3 only explains 0.55% of the variance, so differences along this axis will be overinflated. In my interpretation the pat1/path2 mutant is clustering apart from the other higher order mutants, which is also reflected in the leaf phenotypes. A 2D PCA would be sufficient to describe most of the variation.

    Our response: We agree and thus we changed the PCA plot into a 2D style, please also see our response to reviewer 1 minor comment #3.

    Figure 4: o A: The 3D-PCA inflates the differences between higher order mutants along PC3, even though this axis explains only 0.55% of the variance, maybe a 2D-PCA would more intuitively cluster the samples together?

    Our response: Please see our new PCA plot in Fig4A.

    B: Please explain the scale in the figure legend and which genes were included? Only DEGs between triple mutant and summ2-8 or DEGs that were different in at least one higher order mutant?

    Our response: We now explained more details in the figure legends. The genes which were included in Fig4B were DEGs that were differently expressed in at least one of the pat mutants.

    C: several comparisons are missing from the upset-plot. Please show the complete plot, also is there a white box laid over the second bar in the upper graph? It would help the reader, if the results section would explain the plots and the comparisons took. Which differences are the authors interested in?

    Our response: We covered all the comparisons we wanted to show, but we thank the reviewer for suggesting a more detailed explanation and we therefore explain Fig4C more in detail in Line 146. There is no white box over the second bar, it’s only 1 gene mis-regulated specifically by PATH1 (mis-regulated in plants with path1 mutation).

    From Figure 4B, the triple mutant has an almost inverted expression of mis-regulated genes. High expression genes are now lowly expressed and vice-versa. Has this been reported for other RNA decay mutants before?

    Our response: Our RNA-seq data indicate the pat tripe mutant has more than 1000 mis-regulated genes and based on microarray data on 2-week-old lsm1alsm1b plants (Perea-Resa et al, 2012), more than 600 genes are misregulated in lsm1alsm1b mutant.

    How do you explain that mutants in RNA decay have a large group of repressed transcripts and a large group of enriched transcripts? Wouldn't you suspect a general higher expression in RNA decay mutants or which kind of feedback loop would you propose is happening here? Also, since both kinds of expression changes are recorded in your RNA seq can you speculate on the specificity? Why are some genes up- and others downregulated? Would you suspect that transcription factors are under PATs control?

    Our response: We assume that the mRNA decapping machinery target genes should accumulate in mRNA decapping mutants, *pat mutants in our case. On the other hand, the down-regulated genes could be target genes of other mRNA degradation pathways such as exosome pathway (Line 257); We agree with the reviewer that the down regulated genes in pat *triple could also be negatively regulated by the mRNA decapping targets which could be transcription factor genes. For example, our previous research indicates the transcription factor gene ASL9/LBD3 is mRNA decapping targets under PATs control.

    Where is the sequencing data deposited? This dataset can be of great value for researchers in the field, but the raw data needs to be made commonly available.

    Our response: We thank the reviewer for acknowledging the significance of our work. The raw data has been submitted to NCBI, accession number is PRJNA1006171(Line 307)

    Minor points:

    Check order and nomenclature for protein / gene names in Abstract and Introduction

    Our response: We now carefully double check the order and nomenclature for protein / gene names in abstract and introduction (Line 8,11,14,18,19,24)

    L26 / L83 "aggregate" implies non-functionality, I would use "concentrate", "condensate" or "accumulate".

    Our response: We thank the reviewer for pointing it out, we now use “concentrate” (Line 29&80)

    L35, L45 & L54 all state the same. Maybe remove at least one mention to reduce redundancy?

    Our response: We modified these statements hopefully in a satisfactory way. (Line 56)

    L211: Did you use the same imaging settings for all lines?

    Our response: We used the same settings for all the lines and treatment (Line 284)

    L217: RNA quality "control" word missing?

    Our response: The word “control” is added in Line 296

    L477: Authors should cite the newest version of their BioRxiv: Zuo, Z., Roux, M. E., Chevalier, J. R., Dagdas, Y. F., Yamashino, T., H�jgaard, S. D., ... & Petersen, M. (2022). The mRNA decapping machinery targets LBD3/ASL9 to mediate apical hook and lateral root development in Arabidopsis. bioRxiv, 2022-07.

    Our response: The latest version is cited in our new manuscript (Line 42)

    Figure 3B-F, Figure 4C: check spelling on the axis titles.

    Our response: We carefully checked the spelling on the axis titles in our new manuscript.

    Reviewer #2 (Significance (Required)):

    This manuscript represents a continuation of the author's characterization of the 3 PAT1s in Arabidopsis development after Zuo et al., 2021; Zuo et al., 2022a; Zuo et al., 2022b. The mutants and the corresponding RNA sequencing experiments are of value to the community working on RNA regulation and degradation or plant development. While the initial findings are interesting, the authors do not explore the stimulus-induced condensation differences between the homologues or try to link the extreme differences in expression profiles mechanistically or functionally. I think the manuscript could greatly benefit from contextualizing their work within the frame of their previous studies and what is known about PBs in terms of plant development. While the RNA seq is a comprehensive data set, a closer examination and a better representation of the results would help readers to access the findings.

    __Our response: __We thank the reviewer for the constructive criticism. We hope the reviewer is satisfied by our modified manuscript.

    Reviewer expertise: RNA granule biology, Arabidopsis, molecular biology

    __Reviewer #3 (Evidence, reproducibility and clarity (Required)): __

    Summary:

    In the study "PAT mRNA decapping factors function specifically and redundantly during development in Arabidopsis" authors investigate potential specific functions of Arabidopsis PAT1 orthologs in plant development. Authors observe differences in rosette phenotypes (leaf size, serration and number) of single and multiple mutants of PAT1 gene family, show variation in translocation of the corresponding PAT1 proteins to processing bodies under a set of stress conditions and perform transcriptomics on the established mutants to elucidate the impact of individual PATs on posttranscriptional regulation of plant gene expression. Authors conclude that PAT1 orthologs have both overlapping and specific roles in plant development.

    __Our response: __We thank the reviewer for the peer review. Please see our detailed answers to the reviewer’s suggestions in the following.

    Major comments:

    The study contains intersting transcriptomics data that will be of use for the scientific community. However, analysis of the transcriptomics results could be discussed a bit more in depth. Authors could express their opinion about what gene expression changes might be caused by direct degradation via PAT1-dependent decapping mechanism and what changes are more likely to have occurred indirectly via other factors.

    __Our response: __We followed the reviewer’s suggestion and thus we analysed and discussed more in depth about the transcriptomic data (Line145, 220 &232)

    The intersting phenotypic observations are currently poorly linked to the transcriptomics/qPCR data provided, resulting in a somewhat fragmented story flow.

    __Our response: __We appreciate the reviewer thought the *pat *mutants’ phenotype are interesting, however we disagre with the reviewer on the statement of “poorly linked to the transcriptomics/ qPCR data”. For instance, downregulation of developmental and auxin responsive genes could explain the stunt growth phenotype in the pat triple mutant. Furthermore, the published petiole elongation regulator genes XTR7/XTH15 and *PIL2/PIF6 *exhibit decreased expression level only in mutants with shorter petioles. Nevertheless, we hope our new data and analysis will satisfy the reviewer.

    The transcriptomics was performed on the 6-weeks old plants. It would be helpful to learn more about authors reasoning for choosing this developmental stage for sampling. Why did authors decide against sampling at the earlier stages, before the observed leaves phenotypes were established?

    __Our response: __The pat mutants growth phenotypes showed bigger difference among each other at the late stage, therefore we performed RNA-seq on these samples. But we agree with the reviewer (also reviewer 1, major comment #2), transcriptomic shift at earlier stage could also be responsible for the observed phenotype, thus we performed qRT-PCR on the *pat *mutants at earlier stages for certain genes to examine this (Line 162 &169)

    Authors obtained intriguing results on specific translocation of PAT1, PATH1 and PATH2 to processing bodies in the root cells upon various stresses. Perhaps root transcriptomics of single PAT1, PATH1 and PATH2 knockouts under control conditions, treatment that translocate all three proteins to PBs(IAA) and selectively translocate only PAT1 (e.g. cytokinin) could shed more light on the redundancy an specificity of these proteins as the mRNA decapping factors.

    __Our response: __We appreciate the reviewer found our findings interesting. The specific translocation of PAT1, PATH1 and PATH2 to PBs in the root cells upon various stimuli indicates functional specificity and redundancy in cellular level which correlates with mutants’ growth phenotype. However, we agree with the reviewer that root transcriptomic data on *pat *mutants are very interesting, we are more than willing to share these mutants with peers who want to persue this in more detail.

    Do authors consider PAT1, PATH1 and PATH2 to be localized to different PBs sub-populations? It could be intersting to check co-localization of PAT1, PATH1 and PATH2 under various stress conditions. Could authors elaborate on their view of PBs composition and fate to which different PAT1s are recruited?

    __Our response: __We agree with the reviewer that it’s interesting to check co-localization of PAT1, PATH1 and PATH2. We observed partial localization of CFP-PATH2(in blue) and Venus-PAT1(in yellow) when transiently expressed in Benthmiana. But for permanent lines, we failed at observing separate CFP-PATH2(Blue) signal due to too much signal leakage from Venus-PAT1(Green). Given the fact that PATs function redundantly, we would assume they are partially co-localized in cellular level.

    Could authors speculate what features in the PAT1 protein might cause it being recruited to PBs more efficiently (or better to say, under a broader range of stresses) in comparison to PATH1 and 2?

    __Our response: __The release of ribosome-free mRNPs induces PB formation (Brengues et al., 2005). We suspect PAT1 could bind broader mRNAs compared to PATH1 and PATH2, therefor PAT1-mRNPs could form PBs more efficiently. Moreover, Sachdev et al found yeast PAT1 enhances the condensation of Dhh1 and RNA and PAT1-DHH1 interaction is essential for PB assembly (Sachdev et al., 2019), we assume PAT1 might have better interaction with DHH1 compared to PATH1 and PATH2 thus promote PB formation more efficiently. Please see our discussion part (Line 252)

    Are all three Arabidopsis PAT paralogs co-expressed in the same tissues /developmental stages?

    __Our response: __Please see our response to reviewer 2 major comment #4.

    Could authors elaborate a bit more why the triple pat1 knockout has a much more severe phenotype in comparison to a single pat1 loss-of-function mutant or any of the double pat1 mutants. Do authors observe complementary changes in the PAT1 genes expression in the mutant lines, e.g. is PATH1 expressed at a higher level in the absence of PAT1 and PATH2?

    __Our response: __We now elaborate more about the reason why triple pat1 knockout has the most severe phenotype in the multiple pat mutants (Line 210). We do see higher transcriptional level of *PAT1 *in path1-4path2-1summ2-8 and also higher transcriptional level of PATH1 in pat1-1path2-1summ2-8 but the same PATH2 transcriptional level in pat1-1path1-4summ2-8 compared to summ2-8 (Fig S1C, Line 104)

    Please provide the name of the used statistical test in all figure legends.

    __Our response: __We now provide the statistical test in “Material and Methods” part (Line 367).

    Minor comments:

    Authors might want to reconsider the title as it is somewhat too vague in its current form.

    __Our response: __We now changed our title into “ PAT mRNA decapping factors are required for proper developmental in Arabidopsis”

    Line 9: explanation of PAT1 and PATH1 and 2 abbreviations is best placed at the first mentioning of the name.

    __Our response: __We carefully followed the reviewer’s suggestion (Line 10)

    Line 10: mRNA degradation is rather a posttranscriptional regulation of gene expression.

    __Our response: __We agree and changed our statement in the new ms (Line 9).

    Lines 11 and 12: path1 and path2 abbreviation are not explained. Please note that on the Figure 1A the same proteins are labelled as PAT1H1 and PAT1H2

    __Our response: __We thank the reviewer for pointing it out, we now have PATH1 and PATH2 abbreviations explained in Line 10 and also correct the labels in Fig 1A.

    Lines 22-25: Would you be so kind to rephrase or elaborate on what yoPBu mean. LSM1-7/PAT1 complex are known to bind oligoadenylated transcripts indeed and even stabilize their 3' ends, it is not clear what "engage transcripts containing deadenylated tails" means in this context.

    __Our response: __We hope we now rephrase the statement in a clear way (Line 25)

    Line 29: for the sake of clarity, it might be beneficial to list the known activators of the decapping DCP2 enzyme, including the VCS. Generally the introduction could benefit from a bit more in depth review of the decapping mechanism.

    __Our response: __We hope the more detailed introduction will satisfy the reviewer (Line 27).

    Line 51:"other 2 PATs" => "other two PATs". Generally the text is quite well written, but might need a bit of polishing.

    __Our response: __The text is corrected now (Line 64).

    Authors are absolutely correct in their attempt to provide full information about mutant backgrounds. However, for the sake of comprehension, it would be great to grant the double and triple mutants in the summ2 background shorter and more legible names. For example, the pat1-1path1-4path2-1summ2-8 mutant could be named as pat1/h1/h2/s.

    __Our response: __We originally used pat1/h1/h2/s for the triple but a colleague pointed out “h1” or “h2” are not proper gene names and suggested us to rename them. But we agree that the double and triple pat names are comprehensive, to compromise we change the triple *pat *mutants into pat triple.

    Figure 1B:

    • it would be intersting to have authors opinion on why PBs are formed in this case under non-stress(?) conditions.

    __Our response: __Forming PBs is a dynamic process, and we assume that even under normal conditions, there is still ongoing mRNA decay and translational repression which should be seen as some background level of PBs (Line 85).

    Please note that expressing only the N-terminal part of CFP is a weak negative control for BiFC. No restoration of CFP can occur in such case and thus it is a given that no fluorescence can be observed in these samples. For example, co-expression of nCFP-PAT1 with cCFP-GUS, would be a more rigorous negative control, better aligned with the coIP experiments.

    __Our response: __We had nCFP-Gus with cCFP-LSM1 as real negative control in old Fig 1B (bottom lane). We also agree with the reviewer that only the N-terminal part of CFP is a weak negative control for BiFC, therefore we removed the weak control and only left the rigorous negative control (new Fig 1B).

    Please note that some arrows point at a structure that seems to be not discernible a signal.

    __Our response: __It’s due to the poor quality of the picture from the PDF file, arrows in the original high-resolution figure do point at discernible foci.

    Figure 1C: It might be helpful to also include a Col-0 WT plant

    __Our response: __Col-WT plant is now included in Fig S1A.

    It is not clear how qPCR data and complementation lines help to characterize the established PATH1 and PATH2 loss-of-function mutants. There is no immunodetection of the corresponding proteins in the knockouts, qPCR shows no dramatic decrease in the transcript level of PATH1 and H2 and the phenotypes of complemented lines presented in the Fig S1E at a glance look quite similar to the phenotypes of the corresponding knockout mutants. Complementation lines are not used for any other experiments in this study and it is not clear why authors decided to include this material into the article.

    __Our response: __To characterize the path1 and path2 mutants, we first did qRT-PCR to check the transcriptional level expression, but like the reviewer mentioned, there was no dramatic decrease indicating the mutations of path1-4 and path2-1 did not change PATH1 and *PATH2 *transcriptional level expression. We also tried to raise antibodies against PATH1 and PATH2, however the antibodies failed to recognize any PAT proteins. Therefore, we used the complementation lines to characterize the mutations in PATH1 and PATH2. Since path1 and path2 single mutants don’t have obvious growth phenotype and the dwarf *pat triple is barely possible to transform, we had to complement the pat1path1 and pat1path2 double mutants. If the reviewer takes a closer look, the growth phenotype of the complementation lines Venus-PATH1/ pat1-1path1-4summ2-8 and Venus-PATH2/ pat1-1path2-1summ2-8 *are similar to pat1-1summ2-8 but not the background *pat *double mutants. The complementation lines were also used to study PATH1 and PATH2 cellular localization.

    Figure S1C misses labels indicating what detection of what gene is shown on what chart.

    __Our response: __We thank the reviewer for pointing it out, the gene names are indicated now in new FigS1C.

    Experiments to visualize PBs under various stress stimuli were conducted on roots for the Figure 2 while coIP was performed on the green tissue. Could authors elaborate on whether PB formation could be expected to be the same in different plant organs? Somewhat related to the same topic, Figure 2 contains micrographs obtained on meristematic, transition and elongation root zones, in which epidermal cells are present at various developmental stages. Since PAT proteins are suggested to impact plant development, it might be prudent to obtain observations for all samples at the same developmental stage. Could authors provide their opinion about how representative the provided micrographs are for all root zones? Furthermore, Venus-PATH2 under ACC treatment shows punctate localization only in a single cell out of the three-ish cells visible on the micrograph, potentially indicating differences in PAT2 recruitment to PBs in trichoblasts and atrichoblasts. This in itself could be an intersting observation helpful for elucidating the specific roles of PAT1 orthologs.

    __Our response: __CoIP results from Benthamiana leaves indicate Arabidopsis PATs and LSM1 are in the same complex, and PB visualization in root area suggests PATs respond to different hormone treatments. flg22 treatment has been published to induce PB formation in Arabidopsis root but dissemble PBs in Arabidopsis protoplasts, indicating a tissue specific manner of PB formation. We randomly chose 1 picture/treatment from 9 (3 plants * bio-triplicates) which showed the same. However, we thank the reviewer for pointing out the confocal pictures we chose were not all from elongation zone, we now carefully checked all our confocal pictures and made sure they are from the same developmental stages. We also discuss more of PATH2 localization in response to ACC (Line 251).

    Figure 4C would greatly benefit from a more detailed description in the main text and figure legend of what authors show/conclude.

    __Our response: __We thank the reviewer for the suggestion hence we describe Fig 4C in more detail in our new manuscript (Line 146).

    Figure 5, please avoid using the same color for the bars for the triple pat knockout and the control summ2-8 line

    __Our response: __We changed the colour scheme for all the mutants (new Fig 4E).

    Figure 5B legend should include the name of the statistical test.

    __Our response: __We now include the name of the statistical test in “Material and Methods” (Line 367).

    Figure S2: The coIP experiment is a bit difficult to interpret due to the extremely low protein quantities in some of the input samples. Perhaps a repetition with more balanced input quantities would be beneficial. The figure legend does not contain information on how normalized intensity values were obtained.

    __Our response: __We used the same amount of total protein for each sample (3mg) for each IP, PATH1 and PATH2 don’t express as high as PAT1. The numbers indicate the comparative ratio between PAT-HA protein signal and LSM1-GFP signal, and PAT1-HA/LSM1-GFP under non-treatment condition is normalized as 1.

    Line 130: Fig S2 is referenced but Fig S3 is meant

    __Our response: __We thank the reviewer for pointing out our mistake, the correct figure is now referenced.

    Reviewer #3 (Significance (Required)):

    Strength:

    Regulation of gene expression by mRNA decay is an extremely intersting topic and is highly relevant in plant stress and developmental biology. This study provides a more in depth view on the potential specific roles of the three PAT1 orthologs in Arabidopsis plants. Authors established loss-of-function mutants of the corresponding genes and performed transcriptomics analysis that will be a valuable source for future studies. Furthermore, microscopy analysis of PATH1 and PATH2 translocation to PBs indicates their potential specific roles in plant stress response.

    Weakness: The current version of this study suffers from vague presentation of the results. Starting from the title and ending with discussion authors provide a "general" view on their results and do not go into detailed interpretations. Thus, no mechanistic insight has been derived or at least suggested from the wealth of the transcriptomics, phenotypic and microscopy data.

    The introduction should provide more detailed information on what is known on the PAT1 role in the mRNA decapping pathway and its relevance for plant stress response and development.

    Please note, that the above mentioned suggestion of different sampling for transcriptomics analysis is not meant as a request for this particular study, but rather as an illustration of an expectation a reader would built while following the current version of the text. A thorough description of the strategy for transcriptomics and a more in depth analysis might significantly strengthen the study's coherence and impact.

    Advance:

    At this stage, the study looks more like an incremental advance of the work from the same laboratory performed for the single PAT1 protein. However, as mentioned in the comments above, the study might be made significantly stronger by elaborating the results analysis and highlighting potential discoveries.

    Audience:

    The topic of this study is of a significant interest to a broad audience performing research in plant stress biology and also developmental plant biology.

    __Our response: __We thank the reviewer for acknowledging the significance of our work and the structural criticism. We hope our detailed answers to the reviewer’s suggestions and the additional data we included in the manuscript will satisfy the reviewer.

    Reviewer's and co-reviewer's fields of expertise:

    Molecular Biology, Plant cell biology, Plants Stress response, Autophagy, Stress granules

    __Reviewer #4 (Evidence, reproducibility and clarity (Required)): __

    PAT1 (Protein Associated with Topoisomerase II) are RNA-binding proteins involved in the control of mRNA decay in the cytoplasm. Plants possess multiple PAT1 family members, three in Arabidopsis, PAT1, PATH1 and PATH2. According to the literature, the pat1 mutant shows dwarfism and de-repressed immunity. In this paper, Zou et al. describe the function of PATH1 and PATH2. Two pieces of evidence are consistent with their role in the control of mRNA decay. First, Co-IP and bimolecular Fluorescence Complementation assays in tobacco indicate physical interaction and co-localization of PAT1, PATH1 or PATH2 with LSM1 (Fig. 1), which is a protein present in decapping complexes that form the cytoplasmic foci involved in mRNA decay. Second, PAT1, PATH1 and PATH2 are present in these cytoplasmic Processing Bodies (Fig. 2). Zou et al. generated path1 and path2 mutants, double mutants with pat1 and the triple mutant using independent alleles and the summ2 background to avoid autoimmunity interference. The mutants show leaf growth (Fig. 3) and gene expression (Fig. 4) phenotypes that are not exactly similar among the different family members, but there is significant redundancy revealed by these phenotypes.

    __Our response: __We thank the reviewer for the peer review. Please see our detailed answers to the reviewer’s suggestions in the following.

    The conclusions are straight forward and, apparently, well supported by the data. However, the authors should confirm that when they provide the number of replicates (n) in the legends to the figures, this actually refers to the number of biological replicates. The statements should be based on true biological replicates (not technical replicates). The statistical tests should also be explicitly indicated (including that used to identify DEG in the RNAseq experiment).

    __Our response: __We carefully went through our figures and made sure the number of replicates (n) were correctly stated in figure legends and the statistical tests were indicated (Line 367)

    Reviewer #4 (Significance (Required)):

    The results are useful but mainly descriptive. Personally, I am interested in the mechanisms involved in the control of growth and the manuscript does not mechanistically link the action of PAT1, PATH1 and PATH2 to the transcriptome and the latter to the growth patterns.

    __Our response: __We thank the reviewer for acknowledging the significance of our work of characterizing PATs and we hope our new data could satisfy the reviewer in regarding to “mechanistical link”.

  7. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #4

    Evidence, reproducibility and clarity

    PAT1 (Protein Associated with Topoisomerase II) are RNA-binding proteins involved in the control of mRNA decay in the cytoplasm. Plants possess multiple PAT1 family members, three in Arabidopsis, PAT1, PATH1 and PATH2. According to the literature, the pat1 mutant shows dwarfism and de-repressed immunity. In this paper, Zou et al. describe the function of PATH1 and PATH2. Two pieces of evidence are consistent with their role in the control of mRNA decay. First, Co-IP and bimolecular Fluorescence Complementation assays in tobacco indicate physical interaction and co-localization of PAT1, PATH1 or PATH2 with LSM1 (Fig. 1), which is a protein present in decapping complexes that form the cytoplasmic foci involved in mRNA decay. Second, PAT1, PATH1 and PATH2 are present in these cytoplasmic Processing Bodies (Fig. 2). Zou et al. generated path1 and path2 mutants, double mutants with pat1 and the triple mutant using independent alleles and the summ2 background to avoid autoimmunity interference. The mutants show leaf growth (Fig. 3) and gene expression (Fig. 4) phenotypes that are not exactly similar among the different family members, but there is significant redundancy revealed by these phenotypes.

    The conclusions are straight forward and, apparently, well supported by the data. However, the authors should confirm that when they provide the number of replicates (n) in the legends to the figures, this actually refers to the number of biological replicates. The statements should be based on true biological replicates (not technical replicates). The statistical tests should also be explicitly indicated (including that used to identify DEG in the RNAseq experiment).

    Significance

    The results are useful but mainly descriptive. Personally, I am interested in the mechanisms involved in the control of growth and the manuscript does not mechanistically link the action of PAT1, PATH1 and PATH2 to the transcriptome and the latter to the growth patterns.

  8. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #3

    Evidence, reproducibility and clarity

    Summary:

    In the study "PAT mRNA decapping factors function specifically and redundantly during development in Arabidopsis" authors investigate potential specific functions of Arabidopsis PAT1 orthologs in plant development. Authors observe differences in rosette phenotypes (leaf size, serration and number) of single and multiple mutants of PAT1 gene family, show variation in translocation of the corresponding PAT1 proteins to processing bodies under a set of stress conditions and perform transcriptomics on the established mutants to elucidate the impact of individual PATs on posttranscriptional regulation of plant gene expression. Authors conclude that PAT1 orthologs have both overlapping and specific roles in plant development.

    Major comments:

    • The study contains intersting transcriptomics data that will be of use for the scientific community. However, analysis of the transcriptomics results could be discussed a bit more in depth. Authors could express their opinion about what gene expression changes might be caused by direct degradation via PAT1-dependent decapping mechanism and what changes are more likely to have occurred indirectly via other factors.
    • The intersting phenotypic observations are currently poorly linked to the transcriptomics/qPCR data provided, resulting in a somewhat fragmented story flow.
    • The transcriptomics was performed on the 6-weeks old plants. It would be helpful to learn more about authors reasoning for choosing this developmental stage for sampling. Why did authors decide against sampling at the earlier stages, before the observed leaves phenotypes were established?
    • Authors obtained intriguing results on specific translocation of PAT1, PATH1 and PATH2 to processing bodies in the root cells upon various stresses. Perhaps root transcriptomics of single PAT1, PATH1 and PATH2 knockouts under control conditions, treatment that translocate all three proteins to PBs(IAA) and selectively translocate only PAT1 (e.g. cytokinin) could shed more light on the redundancy an specificity of these proteins as the mRNA decapping factors.
    • Do authors consider PAT1, PATH1 and PATH2 to be localized to different PBs sub-populations? It could be intersting to check co-localization of PAT1, PATH1 and PATH2 under various stress conditions. Could authors elaborate on their view of PBs composition and fate to which different PAT1s are recruited?
    • Could authors speculate what features in the PAT1 protein might cause it being recruited to PBs more efficiently (or better to say, under a broader range of stresses) in comparison to PATH1 and 2?
    • Are all three Arabidopsis PAT paralogs co-expressed in the same tissues /developmental stages?
    • Could authors elaborate a bit more why the triple pat1 knockout has a much more severe phenotype in comparison to a single pat1 loss-of-function mutant or any of the double pat1 mutants. Do authors observe complementary changes in the PAT1 genes expression in the mutant lines, e.g. is PATH1 expressed at a higher level in the absence of PAT1 and PATH2?
    • Please provide the name of the used statistical test in all figure legends.

    Minor comments:

    • Authors might want to reconsider the title as it is somewhat too vague in its current form.
    • Line 9: explanation of PAT1 and PATH1 and 2 abbreviations is best placed at the first mentioning of the name.
    • Line 10: mRNA degradation is rather a posttranscriptional regulation of gene expression.
    • Lines 11 and 12: path1 and path2 abbreviation are not explained. Please note that on the Figure 1A the same proteins are labelled as PAT1H1 and PAT1H2
    • Lines 22-25: Would you be so kind to rephrase or elaborate on what you mean. LSM1-7/PAT1 complex are known to bind oligoadenylated transcripts indeed and even stabilize their 3' ends, it is not clear what "engage transcripts containing deadenylated tails" means in this context.
    • Line 29: for the sake of clarity, it might be beneficial to list the known activators of the decapping DCP2 enzyme, including the VCS. Generally the introduction could benefit from a bit more in depth review of the decapping mechanism.
    • Line 51:"other 2 PATs" => "other two PATs". Generally the text is quite well written, but might need a bit of polishing.
    • Authors are absolutely correct in their attempt to provide full information about mutant backgrounds. However, for the sake of comprehension, it would be great to grant the double and triple mutants in the summ2 background shorter and more legible names. For example, the pat1-1path1-4path2-1summ2-8 mutant could be named as pat1/h1/h2/s.
    • Figure 1B:
    • it would be intersting to have authors opinion on why PBs are formed in this case under non-stress(?) conditions.
    • Please note that expressing only the N-terminal part of CFP is a weak negative control for BiFC. No restoration of CFP can occur in such case and thus it is a given that no fluorescence can be observed in these samples. For example, co-expression of nCFP-PAT1 with cCFP-GUS, would be a more rigorous negative control, better aligned with the coIP experiments.
    • Please note that some arrows point at a structure that seems to be not discernible a signal.
    • Figure 1C: It might be helpful to also include a Col-0 WT plant
    • It is not clear how qPCR data and complementation lines help to characterize the established PATH1 and PATH2 loss-of-function mutants. There is no immunodetection of the corresponding proteins in the knockouts, qPCR shows no dramatic decrease in the transcript level of PATH1 and H2 and the phenotypes of complemented lines presented in the Fig S1E at a glance look quite similar to the phenotypes of the corresponding knockout mutants. Complementation lines are not used for any other experiments in this study and it is not clear why authors decided to include this material into the article.
    • Figure S1C misses labels indicating what detection of what gene is shown on what chart.
    • Experiments to visualize PBs under various stress stimuli were conducted on roots for the Figure 2 while coIP was performed on the green tissue. Could authors elaborate on whether PB formation could be expected to be the same in different plant organs? Somewhat related to the same topic, Figure 2 contains micrographs obtained on meristematic, transition and elongation root zones, in which epidermal cells are present at various developmental stages. Since PAT proteins are suggested to impact plant development, it might be prudent to obtain observations for all samples at the same developmental stage. Could authors provide their opinion about how representative the provided micrographs are for all root zones? Furthermore, Venus-PATH2 under ACC treatment shows punctate localization only in a single cell out of the three-ish cells visible on the micrograph, potentially indicating differences in PAT2 recruitment to PBs in trichoblasts and atrichoblasts. This in itself could be an intersting observation helpful for elucidating the specific roles of PAT1 orthologs.
    • Figure 4C would greatly benefit from a more detailed description in themain text and figure legend of what authors show/conclude.
    • Figure 5, please avoid using the same color for the bars for the triple pat knockout and the control summ2-8 line
    • Figure 5B legend should include the name of the statistical test.
    • Figure S2: The coIP experiment is a bit difficult to interpret due to the extremely low protein quantities in some of the input samples. Perhaps a repetition with more balanced input quantities would be beneficial. The figure legend does not contain information on how normalized intensity values were obtained.
    • Line 130: Fig S2 is referenced but Fig S3 is meant

    Significance

    Strength: Regulation of gene expression by mRNA decay is an extremely intersting topic and is highly relevant in plant stress and developmental biology. This study provides a more in depth view on the potential specific roles of the three PAT1 orthologs in Arabidopsis plants. Authors established loss-of-function mutants of the corresponding genes and performed transcriptomics analysis that will be a valuable source for future studies. Furthermore, microscopy analysis of PATH1 and PATH2 translocation to PBs indicates their potential specific roles in plant stress response.

    Weakness: The current version of this study suffers from vague presentation of the results. Starting from the title and ending with discussion authors provide a "general" view on their results and do not go into detailed interpretations. Thus, no mechanistic insight has been derived or at least suggested from the wealth of the transcriptomics, phenotypic and microscopy data.
    The introduction should provide more detailed information on what is known on the PAT1 role in the mRNA decapping pathway and its relevance for plant stress response and development. Please note, that the above mentioned suggestion of different sampling for transcriptomics analysis is not meant as a request for this particular study, but rather as an illustration of an expectation a reader would built while following the current version of the text. A thorough description of the strategy for transcriptomics and a more in depth analysis might significantly strengthen the study's coherence and impact.

    Advance: At this stage, the study looks more like an incremental advance of the work from the same laboratory performed for the single PAT1 protein. However, as mentioned in the comments above, the study might be made significantly stronger by elaborating the results analysis and highlighting potential discoveries.

    Audience: The topic of this study is of a significant interest to a broad audience performing research in plant stress biology and also developmental plant biology.

    Reviewer's and co-reviewer's fields of expertise:

    Molecular Biology, Plant cell biology, Plants Stress response, Autophagy, Stress granules

  9. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #2

    Evidence, reproducibility and clarity

    Zuo et al., characterize the role of three cytoplasmic mRNA-decay activator proteins PAT1, PATH1 and PATH2 in the context of plant development and leaf morphology in Arabidopsis thaliana and Nicotiana benthamiana. The authors show that the triple pat mutant displays the most severe dwarfism of all combinatorial mutants. Through treatment with different stimulants the authors found that only IAA treatment induces the three homologues to form condensates (possibly PBs), while PAT1 forms condensates upon every tested stimulus. An extensive RNA seq experiment revealed miss-regulation of several hundred genes in the higher order mutants, several of which were involved in auxin responsive and leaf morphology determinant genes.

    Major points:

    • Title is not meaningful as is and, in my opinion, does not reflect the main findings in the manuscript.

    • The results section could benefit from improved flow between the paragraphs and more reasoning for the next steps taken to help readers understand the aims of the authors.

    • L46: "So far little is known about the functions of these three PATs in plant development.", The authors themselves have studied these proteins in the context of seed germination and ABA control, as well as apical hook formation and auxin responses. Should at least be mentioned and the results discussed in this context.

    • What are the expression levels and patterns of PATH1 and PATH2 compared to PAT1? Is anything known about spatial or temporal regulation of these proteins?

    • Figure 1:

      • I do not agree that the authors have shown that "PATH1 and PATH2 are also mRNA decapping factors", rather that these proteins can co-localize (and possibly interact) with LSM1. Decapping assays for example with the known PAT1 de-capping targets from their previous work and their extensive mutant collection could be used to test this.
      • From the BiFC experiment (Figure 1B) it looks like PATs are mostly soluble in the cytoplasm (like LSM1) and might be stress-induced components of PBs (like LSM1). Do PATs co-localize with other canonical PB markers that are more prone to condensation, like DCPs or VCS? BiFC could be performed after IAA treatment to confirm that the cytoplasmic foci are indeed LSM1-positive PBs.
      • A: please provide uncropped images of all Western blots in supplemental data.

    • I applaud the authors for establishing this great higher order mutant collection that will be very useful for researchers in the field. However, I am confused about the description of these mutants. If I understood it correctly, these mutants were already used in a previous study by the authors, namely "Zuo, Z., et al., Molecular Plant-Microbe Interactions, 35(2), 125-130." & Zuo, Z., et al., (2021). FEBS letters, 595(2), 253-263." In this study the authors refer to a BioRxiv "Zuo, Z., et al., (2019)." as the reference for these Arabidopsis lines. Is this current manuscript a continuation of the BioRxiv? Please elaborate whether these lines have been used and described in previous studies.

    • L72: Is the strong developmental phenotype of the higher order mutants persistent under long day conditions? Considering the strong developmental phenotypes of the mutants, the flowering transition and morphology could be an interesting trait to study. Why did you choose short day conditions for this study?

    • L78: This statement is hard to see in Figure 1C and best described for Figure 3A.

    • L82: Please include a reasoning for testing PATs localization after hormone treatment. Do you have any indication that other PB proteins behave similar to either PAT1 or the PATHs after hormone treatment to substantiate that these foci observed are indeed PBs? What is known about PBs after hormone treatment in planta?

    • Figure 2:

      • How does the localization of LSM1 change under the same treatments? Does ist behave like PAT1 or the homologues?
      • Which part of the root was imaged for this experiment? Is it possible that the observed foci are ARF-condensates as reported by Jing et al., 2022? Do you observe a gradual change in numbers or morphology along the root?
      • How did the authors decide on the concentrations for the stimulant treatments? Have you tried different doses, and could the responses be dose-dependent?
      • A representative image is not sufficient for quantitative responses, like RNA granule condensation. Please provide a quantification of stimulant-induced foci after the different treatments.

    • L91: Does that mean that most co-precipitated signal comes from the soluble fraction and not PB-localized? Would an RNAse treatment step eliminate the co-precipitation (optional)?

    • L92/93: Or alternatively that PAT1 localizes to PBs independent of the stress, while PATHs are signal-specific PB components?

    • Figure 3:

      • I wonder if these results fit better in conjunction with Figure 1, either as a combined figure or move before Figure 2.
      • It is interesting that path2/pat1, while being dwarfed, is less serrated compared to pat1 or path1/pat1. Can you find any indications in your RNAseq set which genes might be involved?
      • Indicate statistical test used to determine p-value

    • L116/L117: Doesn't the result in Figure 3E indicate that PATH1 and PATH2 are not fully redundant, but that PATs have specific and narrow roles in leaf development? L116 goes against your statement in L150 & L160. What is known about the expression patterns of PAT1, PATH1 and PHATH2?

    • L123: PC3 only explains 0.55% of the variance, so differences along this axis will be overinflated. In my interpretation the pat1/path2 mutant is clustering apart from the other higher order mutants, which is also reflected in the leaf phenotypes. A 2D PCA would be sufficient to describe most of the variation.

    • Figure 4:

      • A: The 3D-PCA inflates the differences between higher order mutants along PC3, even though this axis explains only 0.55% of the variance, maybe a 2D-PCA would more intuitively cluster the samples together?
      • B: Please explain the scale in the figure legend and which genes were included? Only DEGs between triple mutant and summ2-8 or DEGs that were different in at least one higher order mutant?
      • C: several comparisons are missing from the upset-plot. Please show the complete plot, also is there a white box laid over the second bar in the upper graph? It would help the reader, if the results section would explain the plots and the comparisons took. Which differences are the authors interested in?
      • From Figure 4B, the triple mutant has an almost inverted expression of mis-regulated genes. High expression genes are now lowly expressed and vice-versa. Has this been reported for other RNA decay mutants before?

    • How do you explain that mutants in RNA decay have a large group of repressed transcripts and a large group of enriched transcripts? Wouldn't you suspect a general higher expression in RNA decay mutants or which kind of feedback loop would you propose is happening here? Also, since both kinds of expression changes are recorded in your RNA seq can you speculate on the specificity? Why are some genes up- and others downregulated? Would you suspect that transcription factors are under PATs control?

    • Where is the sequencing data deposited? This dataset can be of great value for researchers in the field, but the raw data needs to be made commonly available.

    Minor points:

    • Check order and nomenclature for protein / gene names in Abstract and Introduction
    • L26 / L83 "aggregate" implies non-functionality, I would use "concentrate", "condensate" or "accumulate".
    • L35, L45 & L54 all state the same. Maybe remove at least one mention to reduce redundancy?
    • L211: Did you use the same imaging settings for all lines?
    • L17: RNA quality "control" word missing?
    • L477: Authors should cite the newest version of their BioRxiv: Zuo, Z., Roux, M. E., Chevalier, J. R., Dagdas, Y. F., Yamashino, T., Højgaard, S. D., ... & Petersen, M. (2022). The mRNA decapping machinery targets LBD3/ASL9 to mediate apical hook and lateral root development in Arabidopsis. bioRxiv, 2022-07.
    • Figure 3B-F, Figure 4C: check spelling on the axis titles.

    Significance

    This manuscript represents a continuation of the author's characterization of the 3 PAT1s in Arabidopsis development after Zuo et al., 2021; Zuo et al., 2022a; Zuo et al., 2022b. The mutants and the corresponding RNA sequencing experiments are of value to the community working on RNA regulation and degradation or plant development. While the initial findings are interesting, the authors do not explore the stimulus-induced condensation differences between the homologues or try to link the extreme differences in expression profiles mechanistically or functionally. I think the manuscript could greatly benefit from contextualizing their work within the frame of their previous studies and what is known about PBs in terms of plant development. While the RNA seq is a comprehensive data set, a closer examination and a better representation of the results would help readers to access the findings.

    Reviewer expertise: RNA granule biology, Arabidopsis, molecular biology

  10. Review Commons

    Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

    Learn more at Review Commons

    Referee #1

    Evidence, reproducibility and clarity

    The manuscript investigates the role of PAT1 gene family in Arabidopsis thaliana. Though the PAT1 protein has been previously investigated and displayed immune-related and developmental phenotypes, the other two members of the family, PATH1 and PATH2, have not been well studied. The authors set out to understand the role of these proteins in relation to the role of PAT1. They thus generated single, double, and triple mutants of the possible combinations of PAT1 genes and examined their phenotypes. As the study focused on the developmental effects of PAT1, the mutants were generated on the background of the summ2 mutant to avoid phenotypes related to immune response. The authors notice a developmental difference between the pat1 mutant combinations, suggesting that PAT1 acts differently than PATH1 and PATH2 and that the PATH proteins serve a redundant function. They also performed RNA-seq analysis to identify differentially-regulated genes in the mutant combinations. The study is interesting and well-executed, yet I believe some questions should still be addressed:

    • The research mainly focuses on the developmental phenotype of pat mutants but also tests the interaction of PATH proteins with RNA decapping enzymes to check their function and localization during different treatments. I found it a bit confusing since Figure 1 also shows the developmental phenotype of the mutants. I think editing the order of the figures would make the overall story more coherent.
    • My main concern is the correlation between the developmental phenotype of the mutants and the gene expression. Leaf samples for RNA extraction were taken when the plants were 6 weeks old, and the developmental phenotype is very evident. It is thus not possible to tell whether the differences in gene expression are a cause or effect of the developmental phenotype. I think performing qPCR of selected candidates at earlier developmental times might help solve this issue, as well as the characterization of younger plants for the developmental phenotypes (such as leaf number).
    • Overall, the manuscript is missing data regarding replicate numbers in the IP and confocal microscopy experiments.

    Minor comments:

    • Figure 1C - the authors should add a picture of Col0 plants as well as the mutants.
    • Figure 3 - Calculating the leaf-to-petiole ratio in the different mutants would be good.
    • Figure 4 - the details in the figure are very unclear, especially in the PCA. It would be good to display the data in 2D for PC1 and PC3 and change the colors a bit.

    Significance

    Both PATH proteins have been less investigated than PAT1, and in that sense, the work is novel. However, it seems that most of the phenotype is attributed to PAT1 rather than the other family members, limiting the interest to the broad plant science community.

  11. Version published to 10.1101/2022.07.06.498930v1 on bioRxiv