Molecular Cloning and Functional Characterization of the DELLA Gene Family in Liriodendron Hybrids

Round 1

Reviewer 1 Report

The manuscript “Molecular cloning and functional characterization of the DELLA gene family in Liriodendron hybrid” by Liu et al. describes the identification of two genes encoding Liriodendron hybrid orthologues of the DELLA proteins – important negative regulators in the gibberellin’s pathway. The authors present phylogenetic and structural analysis of the LhRGL1 and 2 proteins followed by the investigation of the organ-specific expression of the respective genes and overexpression experiments aiming to define the function of the putative LhDELLAs.

Identification of the essential growth regulators in economically important timber tree species makes this report an significant contribution to plant sciences but there are some serious experimental issues which require explanation and revision.

Broad comments

• The names of the species used for phylogenetic analysis presented in Fig. 1 given in the main text are abbreviated which makes them difficult to identify. The full names are explained neither in the figure legend nor in the Materials and methods. The Table S1 in Supplemental data contains the species names but there is no clear indication that these are the same species as in the Fig.1. For example in the Fig.1 the Bc species is mentioned, while in the Table S1 there is no species which name could be abbreviated Bc.
• Line 121 – 122 – The conclusion “the ancient evolutionary status is similar to the phylogenetic position of Liriodendron“ is unclear – LhRGL2 is indeed positioned on the phylogenetic tree at location corresponding to the phylogenetic position of Liriodendron based on genome sequence. However, LhRGL1 is positioned together with Physcomitrella and Selaginella – most ancient plant species. Somehow the authors draw the same conclusion referring to both protein sequences which have very different position on the tree.
• The section 2.4 is entitled ”LhDELLAs are tissue-specifically expressed in Liriodendron hybrid” while in fact organs not tissues are examined for the expression level of the DELLA genes. The tissue/organ confusion continues in the manuscript and it is not well defined in the manuscript which tissues (for example in the case of stamen) were taken for analysis.
• Biological material is not precisely described: in the section 2.4 authors mention “root, stem, bud, leaf, flower, petal and pistil tissue”, in Materials and methods section 4.1. “Plant materials and growth conditions” slightly different biological material is mentioned: “roots, stems, buds, leaves, flowers, petals, stamens and pistils”. Finally, in the section 4.7. “Quantitative RT-PCR” biological material is described in yet different way: “Total RNA was extracted from different tissue parts of the Liriodendron hybrid: mature leaves, buds, petals, stamens and pistils from adult plants on the Nanjing Forestry University campus, and roots, stems, leaves and tips of tissue culture seedlings in the greenhouse.” It is not indicated which of the mentioned samples were in fact used for the real-time PCR analysis: leaves originating from adult plants or from seedlings?, were tips of seedlings used or not (they are not mentioned in the results and in Fig. 4)?; which part of the stamen was used? How old were seedlings? How the single biological replicate is defined – does it include material from one plant or several plants?
• The real-time quantitative PCR analysis of gene expression is very insufficiently described: the thermal profile of the reaction is not provided, the sequences of the used primers are not included and finally the reference gene is not specified. The choice of the right reference gene is the crucial quality aspect of the real-time PCR. In the case of the analysis including samples representing diverse tissues and, even more, different organs, finding of the stably expressed reference gene can be very difficult. Therefore, the qPCR analysis in such situation should be accompanied by the reference gene stability test. The authors mentioned in the legend of the Fig. 4 “Expression was normalized to the expression of the Liriodendron DELLA gene.” It is unclear how the expression of LhDELLAs in Liriodendron hybrid can be normalized to the expression of the Liriodendron DELLA gene.
• The number of the transgenic LhRGL1 tobacco roots is described as reduced, however only the Fig.S1 is shown as a proof without any quantitative analysis like the one performed for the transgenic LhRGL2 Arabidopsis.
• In the lines 181 -182 the phenotype of the transgenic Arabidopsis plants is described: “LhRGL1 overexpression promoted stem development and caused branching and dwarf-like growth in transgenic plant”. While the increased in the transgenic plants number of the side branches was quantified and shown as significant, the dwarf-like growth was not confirmed by the similar quantitative data. Moreover, in Fig. 5c the main inflorescence of the transgenic plant is bent or broken and therefore it is difficult to evaluate its height.

Specific comments

• The sentence in lines 77 and 78 is difficult to understand.
• Lines 79 – 81 – it is not clear how “mutations of GA20ox2 80 and DELLA proteins in rice and wheat” may contribute to “ increased production of (…) raw materials used for clothing”.
• Line 114 “homologues” – since proteins sequences originate from different species should be named “orthologues”.
• Line 118 – is SmDELLA, should be SkDELLA.
• Lines 119 – 120 The sentence starting “Proteins from…” should be corrected (most probably the word “which” should be removed).
• 3a – numbers of the motifs are too small.
• 5b – the chart is described “Bolting time” while in the text the flowering time is mentioned.

Fig. 6b – the chart is described “Plant height” while the figure legend says “root length”. Since the values are in cm and are rather big, I guess the chart presents plant height.

Author Response

Point 1: The names of the species used for phylogenetic analysis presented in Fig. 1 given in the main text are abbreviated which makes them difficult to identify. The full names are explained neither in the figure legend nor in the Materials and methods. The Table S1 in Supplemental data contains the species names but there is no clear indication that these are the same species as in the Fig.1. For example in the Fig.1 the Bc species is mentioned, while in the Table S1 there is no species which name could be abbreviated Bc.

 Response 1: Thank you for your careful work. Bc is the abbreviation of Brassica campestris. But, to prevent similar mistakes, we changed all abbreviations to full names in Figure 1 according to your suggestions (Line 127).

Point 2: Line 121 – 122 – The conclusion “the ancient evolutionary status is similar to the phylogenetic position of Liriodendron” is unclear – LhRGL2 is indeed positioned on the phylogenetic tree at location corresponding to the phylogenetic position of Liriodendron based on genome sequence. However, LhRGL1 is positioned together with Physcomitrella and Selaginella – most ancient plant species. Somehow the authors draw the same conclusion referring to both protein sequences which have very different position on the tree.

 Response 2: Thank you for your suggestions. Referring to the classification of the different DELLA subfamilies shown in Hernandez-Garcia et al (2019), we modify the conclusion in the manuscript at lines 122-126.

Point 3: The section 2.4 is entitled “LhDELLAs are tissue-specifically expressed in Liriodendron hybrid” while in fact organs not tissues are examined for the expression level of the DELLA genes.

 Response 3: Thank you for your careful work. We have modified the title of section 2.4 to "Expression patterns of LhDELLAs in Liriodendron hybrids" (Line 169).

Point 4: The tissue/organ confusion continues in the manuscript and it is not well defined in the manuscript which tissues (for example in the case of stamen) were taken for analysis.

 Response 4: Thanks for your suggestions. We totally agree with your idea. The root, stem, bud, leaf and flower used for qPCR in our experiment should be called organs. Thus, we changed corresponding “tissues” to “organs” when description according to your suggestions (Line 174). Besides, we also separately checked the genes expression in the petal, stamen and pistil of flowers by real-time PCR, these materials we think should be called tissues.

Point 5: Biological material is not precisely described: in the section 2.4 authors mention “root, stem, bud, leaf, flower, petal and pistil tissue”, in Materials and methods section 4.1. “Plant materials and growth conditions” slightly different biological material is mentioned: “roots, stems, buds, leaves, flowers, petals, stamens and pistils”.

Response 5: Thank you for your careful work. We have checked the materials harvested for real-time PCR and added stamen in section 2.4 "root, stem, bud, leaf, flower, petal and pistil organ"(Line 174).

Point 6: Finally, in the section 4.7. “Quantitative RT-PCR” biological material is described in yet different way: “Total RNA was extracted from different tissue parts of the Liriodendron hybrid: mature leaves, buds, petals, stamens and pistils from adult plants on the Nanjing Forestry University campus, and roots, stems, leaves and tips of tissue culture seedlings in the greenhouse.” It is not indicated which of the mentioned samples were in fact used for the real-time PCR analysis: leaves originating from adult plants or from seedlings? How the single biological replicate is defined – does it include material from one plant or several plants?

Response 6: Thank you for your careful work. All the samples used for real-time PCR analysis came from three 25-year-old Liriodendron hybrids planted in Xiashu Forest Farm of Nanjing Forestry University (32.13N, 119.22E). We have changed the description to "Total RNA was extracted from different organs of the adult Liriodendron hybrids: root, stem, bud, leaf, flower, petal, stamen and pistil"(Lines 349-350).

Point 7: were tips of seedlings used or not (they are not mentioned in the results and in Fig. 4)? How old were seedlings?

 Response 7: Thank you for your comment. We did not use Liriodendron hybrids seedlings in this real-time PCR. But the buds of adult Liriodendron hybrids were used to test gene expression.

Point 8: which part of the stamen was used?

 Response 8: Thank you for your comment. In real-time PCR, we used the entire stamen.

Point 9: The real-time quantitative PCR analysis of gene expression is very insufficiently described: the thermal profile of the reaction is not provided, the sequences of the used primers are not included and finally the reference gene is not specified.

 Response 9: Thank you for your careful work. We have added the thermal profile of the reaction (Figure S6) and the sequence of the primers used for real-time quantitative PCR in the supplementary file Table S3. We have added the sentence "The Liriodendron hybrids 18S gene was used as the reference gene." in the manuscript at lines 185-186.

Point 10: The choice of the right reference gene is the crucial quality aspect of the real-time PCR. In the case of the analysis including samples representing diverse tissues and, even more, different organs, finding of the stably expressed reference gene can be very difficult. Therefore, the qPCR analysis in such situation should be accompanied by the reference gene stability test.

 Response 10: Thank you for your careful work. The reference gene we used is the Liriodendron hybrids 18S rRNA gene, which has been identified as a reference gene because of its stable expression during Liriodendron hybrids development by Li Tingting et al. (2012) (DOI: 10.1371/journal.pone.0043451).

Point 11: The authors mentioned in the legend of the Fig. 4 “Expression was normalized to the expression of the Liriodendron DELLA gene.” It is unclear how the expression of LhDELLAs in Liriodendron hybrid can be normalized to the expression of the Liriodendron DELLA gene.

Response 11: Thank you for your careful work. We also found the sentence was useless for the legend, thus we have deleted it (Lines 184-185).

Point 12: The number of the transgenic LhRGL1 tobacco roots is described as reduced, however only the Fig.S1 is shown as a proof without any quantitative analysis like the one performed for the transgenic LhRGL2 Arabidopsis.

Response 12: Thank you for your comment. We have added a quantitative analysis of the number of transgenic LhRGL1 tobacco roots in Figure 6d (Lines 208-213).

Point 13: In the lines 181 -182 the phenotype of the transgenic Arabidopsis plants is described: “LhRGL1 overexpression promoted stem development and caused branching and dwarf-like growth in transgenic plant”. While the increased in the transgenic plants number of the side branches was quantified and shown as significant, the dwarf-like growth was not confirmed by the similar quantitative data.

 Response 13: Thank you for your comment. We added a quantitative analysis of dwarf growth in Figure 5e (lines 201-206).

Point 14: Moreover, in Fig. 5c the main inflorescence of the transgenic plant is bent or broken and therefore it is difficult to evaluate its height.

Response 14: Thank you for your comment. We revised Figure 5c according to your suggestion (line 201).

Point 15: The sentence in lines 77 and 78 is difficult to understand.

 Response 15: Sorry for the unclear description! We have modified the sentence to " The GRAS region at the C-terminus of DELLA proteins is responsible for the interaction with hundreds of transcription factors. This interaction either downregulates or enhances the activity of the transcription factors, thereby regulating thousands of genes in different contexts." (Lines 75-78).

Point 16: Lines 79 – 81 – it is not clear how “mutations of GA20ox2 80 and DELLA proteins in rice and wheat” may contribute to “increased production of (…) raw materials used for clothing”.

 Response 16: Thank you for your careful work. We found it’s not necessary words, so we have deleted "and raw materials used for clothing".

Point 17: Line 114 “homologues” – since proteins sequences originate from different species should be named “orthologues”.

 Response 17: Thank you for your careful work. We have changed "homologues" to "orthologues" (line 114).

Point 18: Line 118 – is SmDELLA, should be SkDELLA.

 Response 18: Thank you for your careful work. We have changed " Interestingly, LhSLRL together with Physcomitrella patens DELLAa, P. patens DELLAb, Selaginella kraussiana DELLA, Amborella trichopoda RHT1, Aquilegia coerulea DELLA3, Nelumbo nucifera GAI-like, and N. nucifera SLR1-like formed a clade, here named outgroup" (lines 118-120).

Point 19: Lines 119 – 120 The sentence starting “Proteins from…” should be corrected (most probably the word “which” should be removed).

 Response 19: Thank you for your comment. We have deleted " Proteins from Selaginella kraussiana and Physcomitrella patens, which belong to the most ancient plant species", because it’s redundant with our new description as response 18 described.

Point 20: 3a – numbers of the motifs are too small.

Response 20: Thank you for your comment. We have enlarged the font and redrawn Figure 3a (line163).

Point 21:5b – the chart is described “Bolting time” while in the text the flowering time is mentioned.

Response 21: Thank you for your comment. We also found “Bolting time” was wrong, thus we have modified it to " Flowering time" (line 201).

Point 22: Fig. 6b – the chart is described “Plant height” while the figure legend says “root length”. Since the values are in cm and are rather big, I guess the chart presents plant height.

Response 22: Thank you for your comment. We also found “root length” in the legend of the Fig. 6b was wrong, thus we have modified it to "plant height" (line 211).

Reviewer 2 Report

This manuscript describes the two DELLA sequences found in Liriodendrom, and interesting species in which GA signaling has not been previously investigated. The authors perform a not very deep phylogenetic analysis which agrees with previous observations, and then overexpress the two full-length clones in Arabidopsis. Here, the results are not as convincing as one would expect, but the authors discuss different angles that might (or might not) explain the strange behavior. Here I comment a few issues (both formal and scientific) that could improve the current version of the manuscript, and suggest one important control and an additional experiment to consolidate the functionality of LhRGL genes.

- Introduction: a large revision of the literature cited is required. Very often, the articles cited are not representative of the statement made in the text:

Ref 2 should be: Daviere and Achard (2013) Development 14:1147-1151-

References 3-6 do not refer to key elements isolated by geneteic analysis. Instead, I suggest:

Ikeda et al. (2001) slender rice, a constitutive gibberellin response mutant, is caused by a null mutation of the SLR1 gene, an ortholog of the height-regulating gene GAI/RGA/RHT/D8. Plant Cell 13: 999–1010.

Peng et al (1997) The Arabidopsis GAI gene defines a signaling pathway that negatively regulates gibberellin responses. Genes Dev 11:3194-3205.

Ueguchi-Tanaka et al (2005) GIBBERELLIN INSENSITIVE DWARF1 encodes a soluble receptor for gibberellin. Nature 437:693-698.

Willige et al (2007) The DELLA Domain of GA INSENSITIVE Mediates the Interaction with the GA INSENSITIVE DWARF1A Gibberellin Receptor of Arabidopsis. Plant Cell 19:1209-1220.

Ref 7 is also not appropriate: the work that shows how GID1 interacts with DELLAs in a GA dependent manner, to regulate DELLA stability is Willige et al (2007) and Ueguichi-Tanaka (2005) mentioned above, as well as Nakajima et al (2006) Plant J 46:880-889.

Ref 10 is also wrong. But if we consider DELLA as GA signaling elements, they have already been cited above, so I would remove "and DELLA[10]" from line 44.

Line 48: the first time that DELLAs are mentioned as part of the GRAS family of transcription factors, the most suitable reference would be Vera-Sirera et al (2015) DELLA proteins, a group of GRAS transcription regularors, mediate gibberellin signaling" in Plant Transcription Factors; Evolutionary, Structural and Functional Aspects, pp313-328. Elsevier.

Ref 12 is very wrong (the cited paper is about the introduction of an Arabidopsis DELLA in rice, not the cloning of the rice DELLA (SLR1, which is described in Ikeda et al [2001]-see above).

Ref 13 is also wrong (the correct paper for wheat DELLAs is Peng et al 1999) 'Green revolution' genes encode mutant gibberellin response modulators. Nature 400:256-261.

Ref 21 and 22 should also be eliminated. Instead of "DELLA and the TVHYNP motif are GAs signal-sensing domains" I would say "DELLA, LEQLE and TVHYNP motifs confer the GA signal-sensing and transcriptional coactivation capacity of this domain, which is functionally conserved across all land plants (Hernández-García et al (2019) Origin of Gibberellin-Dependent Transcriptional Regulation by Molecular Exploitation of a Transactivation Domain in DELLA Proteins. Mol Biol Evol 36:908-918)

Line 74. I would change the whole paragraph to include more relevant and recent information about how the GRAS domain transduces the signal and modulates transcription in response to GAs. I would say: "The GRAS region at the C-terminus of DELLA proteins is responsible for the interaction with hundreds of transcription factors (Marin-de la Rosa et al [2015]; Lanzouni et al [2020]). This interaction either downregulates (de Lucas et al [2008]; Gallego-Bartolomé et al [2012]) or enhances (Marín-de la Rosa [2014]; Fukazawa et al [2014]) the activity of  the transcription factors, thereby regulating thousands of genes in different contexts (Locascio et al [2013])."

Refs are:

Locascio et al (2013) Genomic analysis of della protein activity. Plant Cell Physiol 54:1229-1237

Fukazawa et al (2014) DELLAs Function as Coactivators of GAI-ASSOCIATED FACTOR1 in Regulation of Gibberellin Homeostasis and Signaling in Arabidopsis. Plant Cell 26:2920-2938

Marin-de la Rosa et al (2014) Genome Wide Binding Site Analysis Reveals Transcriptional Coactivation of Cytokinin-Responsive Genes by DELLA Proteins. PLOS Genet. 11, e1005337.

Marin-de la Rosa et al (2015) Large-Scale Identification of Gibberellin-Related Transcription Factors Defines Group VII ETHYLENE RESPONSE FACTORS as Functional DELLA Partners. PLANT Physiol. 166, 1022–1032.

Lantzouni et al (2020) GROWTH-REGULATING FACTORS Interact with DELLAs and Regulate Growth in Cold Stress. Plant Cell 32, 1018–1034.

Gallego-Bartolomé et al (2012) Molecular mechanism for the interaction between gibberellin and brassinosteroid signaling pathways in Arabidopsis. Proc. Natl Acad. Sci. USA 109: 13446–13451.

de Lucas et al (2008) A molecular framework for light and gibberellin control of cell elongation. Nature 451: 480–484.

Refs 28 and 29 are also unrelated to what is mentioned in the text. I would remove those and cite: Thomas (2017) Novel Rht-1 dwarfing genes: tools for wheat breeding and dissecting the function of DELLA proteins. J Exp Bot 68:354-358.

Ref 30 is not related to what is mentioned in the text. Instead, please cite Locascio et al (2013).

- Phylogenetic analysis: I would suggest that the authors refer their nomenclature of the different DELLA subfamilies to the most recent classification shown in Hernandez-Garcia et al (2019) (mentioned above). These authors explain the origin of "DELLA1", "DELLA2", "DELLA1/2" and "DELLA3" subfamilies, and how Arabidopsis only has "DELLA1" and "DELLA2". I would say that LhRGL2 is "DELLA1" or "DELLA2" (hard to differentiate with this tree) and LhRGL1 is a "DELLA3", which is the only one found in early diverging land plants, but also some angiosperms have it (like Aquilegia caerulea and Nelumbo nucifera -please check their genome drafts).

Besides, instead of ref 32 (which is a general mention to OneKP database), I would use again "Hernandez-Garcia et al (2019)", who extracted all DELLA sequences from OneKP and studied their evolutionary origin.

- Overexpression in Arabidopsis. I think that the authors should demonstrate that LhRGL1 is being overexpressed in the transgenic lines. At least they should perform  RT-qPCR with primers specific for the Liriodendron transgene

Moreover, the phenotypic effect is very minor, and one possible explanation is that the Liriodendron RGL1 protein does not accumulate at high level because of the active GA-mediated degradation. I suggest to apply different doses of paclobutrazol to WT and transgenic plants, and check if the transgenic plants are more sensitive to the inhibition of GA synthesis. Otherwise, with the current results is very difficult to confirm that LhRGL1 is truly acting as a DELLA in Arabidopsis.

With respect to LhRGL2, the increase in size is totally counterintuitive and it has never been described for the overexpression of other DELLAs. Again, it is important to show if LhRGL2 is expressed at higher levels in the transgenics.

Both for LhRGL1 and LhRGL2, a much better description of what transgenic lines are being shown is essential. Are these T1 plants? Are they homozygous for the transgene? Is the segregation as expected in the T2?

- Line 238. "LhRGL1 has an evolutionary state similar to SkDELLA, PpGAIa and PpGAIb, which belong to the most ancient plant species." I think this sentence is not accurate. As mentioned above, other angiosperms and monocots also have "DELLA3", like LhRGL1. o there is nothing special in this DELLA, nor can we say that its "evolutionary state" is similar to ancient plants.

Author Response

Point 1: Ref 2 should be: Daviere and Achard (2013) Development 14:1147-1151.

References 3-6 do not refer to key elements isolated by geneteic analysis. Instead, I suggest:

Ikeda et al. (2001) slender rice, a constitutive gibberellin response mutant, is caused by a null mutation of the SLR1 gene, an ortholog of the height-regulating gene GAI/RGA/RHT/D8. Plant Cell 13: 999–1010.

Peng et al (1997) The Arabidopsis GAI gene defines a signaling pathway that negatively regulates gibberellin responses. Genes Dev 11:3194-3205.

Ueguchi-Tanaka et al (2005) GIBBERELLIN INSENSITIVE DWARF1 encodes a soluble receptor for gibberellin. Nature 437:693-698.

Willige et al (2007) The DELLA Domain of GA INSENSITIVE Mediates the Interaction with the GA INSENSITIVE DWARF1A Gibberellin Receptor of Arabidopsis. Plant Cell 19:1209-1220.

Ref 7 is also not appropriate: the work that shows how GID1 interacts with DELLAs in a GA dependent manner, to regulate DELLA stability is Willige et al (2007) and Ueguichi-Tanaka (2005) mentioned above, as well as Nakajima et al (2006) Plant J 46:880-889.

 Response 1: Thank you for your careful work. They are real good suggestions. We have replaced the references (Line 39, 41, 43).

Point 2: Ref 10 is also wrong. But if we consider DELLA as GA signaling elements, they have already been cited above, so I would remove "and DELLA[10]" from line 44.

 Response 2: Thank you for your careful work. We have removed this reference and "and DELLA" (Line 45)

Point 3: Line 48: the first time that DELLAs are mentioned as part of the GRAS family of transcription factors, the most suitable reference would be Vera-Sirera et al (2015) DELLA proteins, a group of GRAS transcription regularors, mediate gibberellin signaling" in Plant Transcription Factors; Evolutionary, Structural and Functional Aspects, pp313-328. Elsevier.

Ref 12 is very wrong (the cited paper is about the introduction of an Arabidopsis DELLA in rice, not the cloning of the rice DELLA (SLR1, which is described in Ikeda et al [2001]-see above).

Ref 13 is also wrong (the correct paper for wheat DELLAs is Peng et al 1999) 'Green revolution' genes encode mutant gibberellin response modulators. Nature 400:256-261.

 Response 3: Thank you for your careful work. They are real good suggestions. We have replaced the references (Line 49, 50).

Point 4: Ref 21 and 22 should also be eliminated. Instead of "DELLA and the TVHYNP motif are GAs signal-sensing domains" I would say "DELLA, LEQLE and TVHYNP motifs confer the GA signal-sensing and transcriptional coactivation capacity of this domain, which is functionally conserved across all land plants (Hernández-García et al (2019) Origin of Gibberellin-Dependent Transcriptional Regulation by Molecular Exploitation of a Transactivation Domain in DELLA Proteins. Mol Biol Evol 36:908-918)

 Response 4: Thank you for your careful work. We have eliminated ref 21 and 22. We have revised this sentence based on your suggestion and quoted the references you provided (Lines 57-58).

Point 5: Line 74. I would change the whole paragraph to include more relevant and recent information about how the GRAS domain transduces the signal and modulates transcription in response to GAs. I would say: "The GRAS region at the C-terminus of DELLA proteins is responsible for the interaction with hundreds of transcription factors (Marin-de la Rosa et al [2015]; Lanzouni et al [2020]). This interaction either downregulates (de Lucas et al [2008]; Gallego-Bartolomé et al [2012]) or enhances (Marín-de la Rosa [2014]; Fukazawa et al [2014]) the activity of the transcription factors, thereby regulating thousands of genes in different contexts (Locascio et al [2013])."

Refs are:

Locascio et al (2013) Genomic analysis of della protein activity. Plant Cell Physiol 54:1229-1237

Fukazawa et al (2014) DELLAs Function as Coactivators of GAI-ASSOCIATED FACTOR1 in Regulation of Gibberellin Homeostasis and Signaling in Arabidopsis. Plant Cell 26:2920-2938

Marin-de la Rosa et al (2014) Genome Wide Binding Site Analysis Reveals Transcriptional Coactivation of Cytokinin-Responsive Genes by DELLA Proteins. PLOS Genet. 11, e1005337.

Marin-de la Rosa et al (2015) Large-Scale Identification of Gibberellin-Related Transcription Factors Defines Group VII ETHYLENE RESPONSE FACTORS as Functional DELLA Partners. PLANT Physiol. 166, 1022–1032.

Lantzouni et al (2020) GROWTH-REGULATING FACTORS Interact with DELLAs and Regulate Growth in Cold Stress. Plant Cell 32, 1018–1034.

Gallego-Bartolomé et al (2012) Molecular mechanism for the interaction between gibberellin and brassinosteroid signaling pathways in Arabidopsis. Proc. Natl Acad. Sci. USA 109: 13446–13451.

de Lucas et al (2008) A molecular framework for light and gibberellin control of cell elongation. Nature 451: 480–484.

 Response 5: Thanks for your careful work and kind comments. We have revised the entire paragraph based on your suggestions and cited these references. (Lines 75-80).

Point 6: Refs 28 and 29 are also unrelated to what is mentioned in the text. I would remove those and cite: Thomas (2017) Novel Rht-1 dwarfing genes: tools for wheat breeding and dissecting the function of DELLA proteins. J Exp Bot 68:354-358.

 Response 6: Thank you for your careful work. We have deleted refs 28-29 and cited the reference you provided (Line 83).

Point 7: Ref 30 is not related to what is mentioned in the text. Instead, please cite Locascio et al (2013).

 Response 7: Thank you for your careful work. We have replaced the reference (Line 96)

Point 8: Phylogenetic analysis: I would suggest that the authors refer their nomenclature of the different DELLA subfamilies to the most recent classification shown in Hernandez-Garcia et al (2019) (mentioned above). These authors explain the origin of "DELLA1", "DELLA2", "DELLA1/2" and "DELLA3" subfamilies, and how Arabidopsis only has "DELLA1" and "DELLA2". I would say that LhRGL2 is "DELLA1" or "DELLA2" (hard to differentiate with this tree) and LhRGL1 is a "DELLA3", which is the only one found in early diverging land plants, but also some angiosperms have it (like Aquilegia coerulea and Nelumbo nucifera -please check their genome drafts).

 Response 8: Thank you for your careful work. We renamed LhRGL1 to LhSLR-like (LhSLRL), and LhRGL2 to LhRGA, referring to the latest classification shown in Hernandez-Garcia et al. (2019). Referring to the most recent classification of the different DELLA subfamilies shown in Hernandez-Garcia et al (2019), we modify the conclusion at lines 122-126. We have added the DELLA proteins of Aquilegia coerulea and Nelumbo nucifera to the phylogenetic tree, redrawn the phylogenetic tree in Figure 1 (Lines 127).

Point 9: Besides, instead of ref 32 (which is a general mention to OneKP database), I would use again "Hernandez-Garcia et al (2019)", who extracted all DELLA sequences from OneKP and studied their evolutionary origin.

 Response 9: Thank you for your careful work. We have deleted ref 30 and cited the reference you provided (Line 126).

Point 10: Overexpression in Arabidopsis. I think that the authors should demonstrate that LhRGL1 is being overexpressed in the transgenic lines. At least they should perform RT-qPCR with primers specific for the Liriodendron transgene

 Response 10: Thanks for your kind suggestions! We totally agree with your idea. For the demonstration of LhRGL1 being overexpressed in the transgenic lines, we can provide the PCR testing for transgenic plants as Figure S3-S5. However, we feel so sorry that we cannot provide the qPCR data for LhRGL1 overexpression under the limited time. Anyway, in our follow-up study, we will perform RT-qPCR with specific primers to demonstrate the genes overexpressing in the transgenic plants. Thank you again for your very valuable comment!

Point 11: Moreover, the phenotypic effect is very minor, and one possible explanation is that the Liriodendron RGL1 protein does not accumulate at high level because of the active GA-mediated degradation. I suggest to apply different doses of paclobutrazol to WT and transgenic plants, and check if the transgenic plants are more sensitive to the inhibition of GA synthesis. Otherwise, with the current results is very difficult to confirm that LhRGL1 is truly acting as a DELLA in Arabidopsis.

 Response 11: Thank you for your careful work. We have not correctly described our observation. About “the minor phenotypic effect”, it indicated the minor difference of flowering time between transgenic lines. Sorry for the not precise description! So, we changed the sentence to "Flowering time showed no significant difference between transgenic lines" (Lines 191-192). About your suggestion on paclobutrazol application, thank you for your very valuable comment! Your suggestions will be incorporated into our follow-up study.

Point 12: With respect to LhRGL2, the increase in size is totally counterintuitive and it has never been described for the overexpression of other DELLAs. Again, it is important to show if LhRGL2 is expressed at higher levels in the transgenics.

 Response 12: Thank you for your careful work. According to the observation of the transgenic lines, we did find the phenomenon of rosette leaves and thus performed a quantitative analysis on it. The transgenic lines we selected have the most obvious phenotype among all transgenic lines. When the T3 plants were treated with 0.1-0.5 µM GA₃, the roots of the transgenic lines were significantly shorter than that of the wild type, however, there was no difference between transgenic seedlings and WT plants under 1-5 µM GA₃ treatment. This indicates that GA₃ should first degrade the overexpressed LhRGL2 protein in the transgenic plants and then promote elongation of roots, showing that there is excessive LhRGL2 protein in the transgenic lines (Figure 8 and S5).

Point 13: Both for LhRGL1 and LhRGL2, a much better description of what transgenic lines are being shown is essential. Are these T1 plants? Are they homozygous for the transgene? Is the segregation as expected in the T2?

 Response 13: Thank you for your careful work. These transgenic lines are T3 homozygous plants. The segregation in the T2 was as expected.

Point 14: Line 238. "LhRGL1 has an evolutionary state similar to SkDELLA, PpGAIa and PpGAIb, which belong to the most ancient plant species." I think this sentence is not accurate. As mentioned above, other angiosperms and monocots also have "DELLA3", like LhRGL1. o there is nothing special in this DELLA, nor can we say that its "evolutionary state" is similar to ancient plants.

Response 14: Thank you for your careful work. We have deleted it.

Reviewer 3 Report

The manuscript by Liu et al. entitled “ Molecular cloning and functional characterization of 2 the DELLA gene family in Liriodendron hybrid” presents some interesting results regarding the function of two DELLA proteins of Liriodendron. The phenotypes of Arabidopsis and tobacco plants overexpressing LhRGL1 and LhRGL2 genes are noteworthy. However, the manuscript does not provide information necassary to evaluate the results and to justify conclusions.

1. 2.1. The first sentence: „DELLA proteins are crucial factors in the GA signaling pathway and have been known to mediate plant development by regulating the expression of their target genes[30]“ is not appropriate here, it only repeats the facts in Introduction.
2. What does it mean „Liriodendron hybrid transcriptome“ (p.3. l.,96) ? Where have been Liriodendron transcriptomic resource published? The sentence “To analyze the function of DELLA genes from Liriodendron hybrid, we cloned their sequence from total Liriodendron RNA“ (p.3, l. 95-96), is incorrect, when the authors searched the (unspecified) transcriptome.
3. What is the correct taxonomic name of Liriodendron hybrid ? Hybrid between what?
4. Plant material. How old were plants used for gene expression studies? Under which environmental conditions were samples for RNA extraction taken?
5. Which reference gene(s) were used for RT qPCR study (Figure 4) ? Was the invariant expression of the reference gene validated? How were the relative expression value calculated?
6. Phylogenetic analysis (Figure 1). How the branch support was calculated? Were there bootstrap values. The abbreviations on the tips shall be explained in the legend.
7. The half of Introduction is spent on the description of the DELLA protein structure. Please, abbreviate it substantially and add the figure showing specific motifs. What is the difference between the DELLA domain and the CTD domain, which are described in two different places ?
8. The transformation of Arabidopsis and tobacco with LhRGL constructs is a very interesting part of the manuscript. However, why only LhRGL1 was used to transform tobacco, why not also LhRGL2? Why different phenotypic traits were followed in LhRGL1 and LhRGL2 transformants? This makes very difficult to compare the impact of the two genes on plant development.
9. Last, but not least, the text has to be substantially rewritten and corrected by a native English speaker. There are many errors in grammar and style.

Author Response

Point 1: 2.1. The first sentence: “DELLA proteins are crucial factors in the GA signaling pathway and have been known to mediate plant development by regulating the expression of their target genes[30]” is not appropriate here, it only repeats the facts in Introduction.

Response 1: Thank you for your comment. We have changed the sentence to " DELLAs mediate plant development by interacting with and regulating the activity of DNA-binding TFs to regulate the expression of downstream target genes [30]" (Lines 95-96).

Point 2: What does it mean “Liriodendron hybrid transcriptome” (p.3. l.,96)? Where have been Liriodendron transcriptomic resource published?

Response 2: Thank you for your comment. We have searched these genes from Liriodendron hybrids transcriptomic resource published in Chen Jinhui et al. (2019) (DOI: 10.1038/s41477-019-0368-1).

Point 3: The sentence “To analyze the function of DELLA genes from Liriodendron hybrid, we cloned their sequence from total Liriodendron RNA” (p.3, l. 95-96), is incorrect, when the authors searched the (unspecified) transcriptome.

Response 3: Thank you for your comment. We have checked the method about gene clone and modified the description to "First, to analyze the function of DELLA genes from Liriodendron hybrids, we used the rice and Arabidopsis DELLA protein sequences downloaded from NCBI to perform a BLASTP search against the Liriodendron hybrids transcriptome" (Lines 96-99).

Point 4: What is the correct taxonomic name of Liriodendron hybrid?

Response 4: Thank you for your comment. The correct taxonomic name of Liriodendron hybrids is Liriodendron chinense (Hemsl.) Sarg.×L.tulipifera L. We have added it at line 17.

Point 5: Hybrid between what?

Response 5: Thank you for your comment. Liriodendron hybrids was from Liriodendron chinense cross with Liriodendron tulipifera.

Point 6: Plant material. How old were plants used for gene expression studies? Under which environmental conditions were samples for RNA extraction taken?

Response 6: Thank you for your comment. Plants used for gene expression studies were 25-year-old. Samples for RNA extraction were taken from Xiashu Forest Farm of Nanjing Forestry University, located at 32.13N, 119.22E.

Point 7: Which reference gene(s) were used for RT qPCR study (Figure 4)?

Response 7: Thank you for your careful work. The Liriodendron hybrids 18S rRNA gene was used as the reference gene for internal standard (Lines 354-355).

Point 8: Was the invariant expression of the reference gene validated?

Response 8: Thank you for your careful work. The reference gene we used is the Liriodendron hybrids 18S rRNA gene, which has been identified as a reference gene because of its stable expression during Liriodendron hybrids development by Li Tingting et al. (2012) (DOI: 10.1371/journal.pone.0043451).

Point 9: How were the relative expression value calculated?

Response 9: Thank you for your careful work. The relative expression value was calculated using 2(-Delta Delta C(T)) Method (Lines 355-356).

Point 10: Phylogenetic analysis (Figure 1). How the branch support was calculated?

Response 10: Thank you for your careful work. The evolutionary history was inferred by using the Maximum Likelihood method and JTT matrix-based model. The percentage of trees in which the associated taxa clustered together is shown next to the branches. Initial tree(s) for the heuristic search were obtained automatically by applying Neighbor-Join and BioNJ algorithms to a matrix of pairwise distances estimated using a JTT model, and then selecting the topology with superior log likelihood value. The tree is drawn to scale, with branch lengths measured in the number of substitutions per site (Lines 322-327).

Point 11: Were there bootstrap values.

Response 11: Thank you for your careful work. Numbers near the nodes are bootstrap support values in Figure 1 (Line 130).

Point 12: The abbreviations on the tips shall be explained in the legend.

Response 12: Thank you for your careful work. We changed all abbreviations to full names and recreated this diagram according to your suggestion and named it Figure 1 (Line 127).

Point 13: The half of Introduction is spent on the description of the DELLA protein structure. Please, abbreviate it substantially and add the figure showing specific motifs. What is the difference between the DELLA domain and the CTD domain, which are described in two different places?

Response 13: Thank you for your comment. To prevent the confusion about these two domains again, “CTD" has been revised to "GRAS". DELLA protein is a branch of the GRAS family. The DELLA domain is the N-terminal domain of the DELLA protein, and the GRAS domain is the C-terminal domain of the DELLA protein. These two domains have different functions (Lines 55-60).

Point 14: The transformation of Arabidopsis and tobacco with LhRGL constructs is a very interesting part of the manuscript. However, why only LhRGL1 was used to transform tobacco, why not also LhRGL2? Why different phenotypic traits were followed in LhRGL1 and LhRGL2 transformants? This makes very difficult to compare the impact of the two genes on plant development.

Response 14: Thank you for your comment. First, we used LhRGL1 and LhRGL2 to transformed Arabidopsis. Then, we found that the phenotype of LhRGL1 was more obvious, so we used LhRGL1 to transform tobacco. Thank you again for your very valuable comment, and your suggestions will be incorporated into our follow-up study.

Point 15: Last, but not least, the text has to be substantially rewritten and corrected by a native English speaker. There are many errors in grammar and style.

Response 15: Thank you for your comment. The text has been substantially rewritten and corrected by Dr. R.A. Mentink.

Round 2

Reviewer 1 Report

Authors sufficiently addressed all my concerns. Methods are much better described and description of experiments is improved. Especially the qPCR and phenotype analysis parts are now more convincing.

Author Response

Point 1: Authors sufficiently addressed all my concerns. Methods are much better described and description of experiments is improved. Especially the qPCR and phenotype analysis parts are now more convincing.

Response 1: Thank you very much for your approval of the article!

Reviewer 2 Report

Thanks for addressing all the issues raised during the previous review. I still consider that the manuscript would increase its potential impact if the level of expression of the transgenes were confirmed (not just the presence of the transgenes, as shown by PCR in Suppl Figs 3-5). But I understand the explanations given by the authors.

Author Response

Point 1: Thanks for addressing all the issues raised during the previous review. I still consider that the manuscript would increase its potential impact if the level of expression of the transgenes were confirmed (not just the presence of the transgenes, as shown by PCR in Suppl Figs 3-5). But I understand the explanations given by the authors.

Response 1: It is really a great suggestion as you pointed out that we conducted q-PCR testing on Arabidopsis and tobacco overexpressing LhDELLAs and put the results in Figure S5.

Reviewer 3 Report

The authors corrected several errors by their revision, but the some problems persists. Most of all, it is difficult to compare  the functions of LhRGL1 and LHRGL2, when the two genes are not transferred to the same plant species (Arabidopsis or tobacco). Please, provide the phenotypes of Arabidopsis and tobacco transformed by both DELLA genes. As for gene expression studies, it is difficult to imagine, which part of stem of root was taken, if 25 year-old trees were analyzed.

Author Response

Point 1: The authors corrected several errors by their revision, but the some problems persists. Most of all, it is difficult to compare the functions of LhRGL1 and LHRGL2, when the two genes are not transferred to the same plant species (Arabidopsis or tobacco). Please, provide the phenotypes of Arabidopsis and tobacco transformed by both DELLA genes.

Response 1: It is really a good idea as you suggested, and we have changed them to meet your thoughts. We put the phenotypes of Arabidopsis overexpressing LhRGL1 and LhRGL2 together for comparison. In order not to cause ambiguity, we put the phenotype of tobacco overexpressing LhRGL1 in the figure S1. However, we feel so sorry that we cannot use LhRGL2 to transform tobacco under the limited time. Anyway, in our follow-up study, we will use LhRGL2 to transform tobacco to further compare the functional differences between LhRGL1 and LhRGL2. Thank you again for your very valuable comment!

Point 2: As for gene expression studies, it is difficult to imagine, which part of stem of root was taken, if 25-year-old trees were analyzed.

Response 2: Thank you for your comment. We built a very high shelf around the tree. Using this shelf, we collected stems from the top of the 25-year-old Liriodendron hybrids, as shown in Figure 1. Liriodendron hybrids is a shallow-rooted tree species, and its lateral roots are mainly distributed in the 0-20cm soil layer. We collected lateral roots of Liriodendron hybrids as the q-PCR material, as shown in Figure 2.

Author Response File: Author Response.docx