Manipulating microRNA miR408 enhances both biomass yield and saccharification efficiency in poplar

The conversion of lignocellulosic feedstocks to fermentable sugar for biofuel production is inefficient, and most strategies to enhance efficiency directly target lignin biosynthesis, with associated negative growth impacts. Here we demonstrate, for both laboratory- and field-grown plants, that expression of Pag-miR408 in poplar (Populus alba × P. glandulosa) significantly enhances saccharification, with no requirement for acid-pretreatment, while promoting plant growth. The overexpression plants show increased accessibility of cell walls to cellulase and scaffoldin cellulose-binding modules. Conversely, Pag-miR408 loss-of-function poplar shows decreased cell wall accessibility. Overexpression of Pag-miR408 targets three Pag-LACCASES, delays lignification, and modestly reduces lignin content, S/G ratio and degree of lignin polymerization. Meanwhile, the LACCASE loss of function mutants exhibit significantly increased growth and cell wall accessibility in xylem. Our study shows how Pag-miR408 regulates lignification and secondary growth, and suggest an effective approach towards enhancing biomass yield and saccharification efficiency in a major bioenergy crop.

Overall, the study was well designed and the experiments were well performed, especially those phenotypic analysis and localization investigation of LACs and Pag-miR408 using fluorescence in situ hybridization; the writing is also easy to follow and understand. I only have a few comments as follows, these issues need to be addressed before the manuscript can be accepted: 1. Title "Manipulating non-coding RNA…….." better changes to "Manipulating an microRNA miR408…" or "Manipulating Pag-miR408….".
2.In the Introduction section, the authors only generally described lignocellulosic biomass and the approaches to overcoming the growth defects in lignin-modified plants, the authors should introduc why choosing and focusing miR408.
3. Figure 1d: MIR408-OX plants clearly showed enlarged vascular cambium zones, suggesting that overexpression of miR408 not only enhances biomass yield and saccharification efficiency, but also promotes cambium division, can authors discuss what are the underlined mechanism for miR408 to regulate cambium? 4. Figure 3a. The basal stem xylem width, also the stem width of MIR408-OX plants indeed showed significantly increased compared to WT, while no significant change occurred in knockout poplars (Figure 3a,Supplemental Figure 9a). Generally, knockout mutants should show an opposite phenotype with that of OX plants, can authors explain this? 5. Figure 5. Double lac25 lac32 and triple lac19 lac25 lac32 mutants showed lighter lignin staining, and loose cell arrangement with a degree of vessel collapse(figure 5b and c) ,and the cell wall morphology of the single gene mutants of lac19 was not as obvious as that of the triple and double mutants (Figure 5d), however, Figure 5d showed that the lac19 plants grow well as those of double and triple mutants, can authors explain these phenomenon?
In addition, LAC19-OX, LAC25-OX and LAC32-OX (Figure 5f-h) exhibited neatly arranged xylem cells, with similar morphology to WT. If these laccases indeed the targets of miR408, the LACs OX plants should show an opposite phenotype with those of miR408 OX plants, and the staining of xylem cells should be stronger than that of WT? 6. In the discussion section, could authors discuss more the function of laccases?
Reviewer #3: Remarks to the Author: This manuscript describes the phenotypes of transgenic poplar trees expressing a microRNA (miR408) that represses expression of genes encoding laccases. A triple mutant of three laccase-encoding genes phenocopies the miR408-expressing lines. Lignin content and composition are not substantially affected in these lines. However, the yield of glucose in saccharification assays is increased when the materials are treated with cellulase enzymes. There is an increased abundance of binding sites in transverse sections of these materials for probes that bind to cellulose microfibrils compared to wild type.
These results are significant in a biotechnological context of the deconstruction of lignocellulosic biomass as a source of sugars for conversion to fuels or other products. As the authors point out, reduction of lignin content is generally associated with reduced growth, an undesirable phenotype in a bioenergy crop. There are some examples in the literature, however, where changing lignin composition does not impact growth, for example, by expressing ferulate-5-hydroxylase in poplar trees to synthesize predominantly syringyl (S)-lignin. S-lignin is a linear molecule with a single type of linkage that is more labile that the multiple kinds of linkages between aromatic subunits found in wild type lignin, comprising guaiacyl (G)-and S-lignin monomers. By contrast, the miR408-expressing lines and the triple laccase mutant described in this manuscript show slightly reduced S-lignin composition (Supplemental Table 1) compared to wild type, and about a 10% reduction in lignin content by AcBr assays.
Of more fundamental scientific interest, but very briefly noted in a single sentence in the Discussion, is the function of laccases in plant cell walls, and more specifically in lignin cross-linking. Laccases have previously been hypothesized to be involved in cross-linking monomeric subunits of lignin, and so, one interpretation of the authors' results is that reduced cross-linking accounts for the observed phenotypes.
This manuscript is rich in experimental results and there are few studies of such depth in transgenic tree species. However, the authors should address a number of issues in the writing of the manuscript before it is suitable for publication.
1. The authors should correct grammatical and typographical errors throughout. 2. Some terms are not precisely defined. For example, line 103, "strongly observed", line 107, 115 and Supplemental Figure 7, "loosely arranged xylem". In particular, the authors use the term "accessibility" to describe both macroscopic properties (glucose yields in saccharification assays) and microscopic properties (binding sites in transverse sections for cellulose-binding probes). I assume that the authors mean the accessibility of cellulose microfibrils to cellulase enzymes. It is reasonable to infer that both of these phenotypes are correlated with cellulase accessibility to its substrate but these are proxy measurements of accessibility rather than direct measurements. 3. Loss-of-function phenotypes from knockout of the miRNA-encoding gene are not described or discussed beyond a cursory mention in the text. However, Supplemental Figure 9b and c shows interesting phenotypes of reduced cell wall thickness and an increased intensity of phloroglucinol staining in successive internodes. Discussion of these phenotypes could enrich understanding of the function of miR408. 4. Throughout the results, the authors should clarify the nature of the samples measured. For example, in Figure 1i, Table 1, Supplemental Figure 4, Figure 4d, the legend states that n = 3, without specifying whether these represent 3 transgenic lines, 3 trees sampled from one transgenic line, or three replicate samples from a single tree. 5. Figure 4 also indicates that Lac 47 and Lac 55 are potential targets of miR408 but no data are presented -are there changes in the expression levels of these two laccases? Supplemental Figure  12c shows altered expression ratios of Laccases 1, 3, 4, 10, 11 and 17, but not 19, 25 or 32 in the miR408-expressing lines. Is this a typographical error? Or is the expression of these other laccase genes also impacted in the laccase triple mutant? 6. Line 267 refers to Figure 4e to g, but these panels are not part of Figure 4. 7. The authors should cite relevant literature and discuss their own findings in the context of literature with respect to the function of laccases in lignification. For example, a dirigent protein (Dirigent protein 23) is also a predicted target gene of miR408 (Supplemental Table 2), and this class of protein has also been implicated in lignin cross-linking. How is its expression affected in the triple laccase mutant and does this impact (or not) the interpretation of the triple mutant phenotype? 8. Supplemental Figure 8a -please clarify if "cell wall residues" refers to cell walls after cellulase treatment, or if this is the total sugar content of the starting cell wall materials. 9. Supplemental Figure 14b legend, where are the "lower panels" showing bright field photographs?

Comments of Reviewer 1:
This is a resubmission of a previously reviewed manuscript . In this new submission, the authors have made several important revisions with substantial new data to address the comments raised during the previous review. I particularly appreciate the efforts generating and characterizing overexpression and knockout out poplar lines for LAC genes to provide genetic evidence supporting that these LAC genes are targets of miR408 and work in the same pathway to regulate plant growth and saccharification efficiency.
Overall, the quality of the present manuscript has been much improved over the previous version. There are several concerns that the authors need to address.
Point 1: MiR408 targets: In the previous version, LAC19/20 were shown to be the targets of MiR408 whereas in the present version, LAC19/25/32 were shown to be the targets. I am just curious about the discrepancy. In vitro transactivation assays were used in both versions.
Some explanations about the potential cause of discrepancy would be helpful. I did comment in the previous version that the examination was not comprehensive since only LAC19/20 were selected for the in vitro transactivation assays and other LACs (LAC16/25/32) were not. I am glad that the authors addressed my previous comments which has led discovery of new targets (LAC25/32), but now I wonder how LAC20 is no longer a target.
Response 1: Sorry for the unclear explanation here. After receiving your comments on the previous version, we performed detailed new experiments to re-address the targets. In the last two years since the first submission, we have established the 5' RACE technique to confirm the true targets of miR408 in 84K poplar.
 As for LAC16, we found from the first-time RNA-Seq data that LAC16 was upregulated.
Since microRNAs play a negative regulatory role after transcription, LAC16 could not be the target gene of miR408. On the other hand, using psRNAtarget prediction, we also found that LAC16 was not present in the potential targets list of miR408 (Supplemental   table 2). Therefore, we can safely conclude LAC16 is not the target of miR408.
 In the second version of the paper, based on the fact that we have obtained the miR408cr lines, we re-performed the RNA-seq using WT, miR408_OX and also miR408_cr. The transcript data showed the same trend in duplicate analyses; the expression of LAC19, 25 and 32 was all down-regulated in the over-expression lines. Furthermore, from psRNAtarget prediction, LAC19, LAC25, and LAC32 were predicted as the highest potential targets of miR408 (Supplemental table 2). This is the main reason why we selected LAC19, 25 and 32 to verify our analyses. We did not find that LAC20 was present in the potential targets list of miR408 and there is no potential cleavage site of LAC20 that can be targeted by miR408. Moreover, we did not find differences in the expression level of LAC20 in the second RNA-seq data. Based on these results, we are confident that LAC20 is not the target of miR408 and therefore deleted it in the revised version. Thanks for your comments which led to discovery of new targets (LAC25/32) in from our additional experiments.
Point 2: Determination of miR408 targets: the authors used RNA-Seq data from miR408_OX plants to determine DEGs and subsequently used ontology clustering to narrow down downregulated genes/pathways. The authors then used psRNAtarget prediction to identify LAC19/25/32 as the highest potential targets of miR408. Since LAC47/55 were also predicted to be targets of MiR408 (Table S2) and given that they are in the same LAC family, I wonder why LAC47/55 (again, for comprehensiveness) were not included in the RT-PCR analysis, in vitro transactivation assays and 5' RACE. It is nice to see all positive data on LAC19/25/32, but it is equally important to see negative data on LAC47/55 to demonstrate specificity.

Response 2:
After reading your comments, we realized this problem in our previous version.
To solve this, we carried out qRT-PCR and 5´ RACE for the identification of target genes. We first analyzed the down-regulated laccases in transcriptome data, and also checked whether the down-regulated laccases exist in the list of predicted target genes. If down-regulated laccases do not exist in the predicted target gene list, it means that these laccases have no target cleavage site for miR408. We believe these down-regulated laccases are not the targets of miR408. Therefore, we selected LAC19, 25 and 32 to do 5´ RACE to find the cleavage site of miR408.
The results showed that LAC47 and 55 were not in the list of down-regulated genes in both transcriptomic analyses ( Figure S12d). Furthermore, in our qRT-PCR analysis, we found that there were no significant differences in the expression level of LAC47 and 55 between WT and miR408_OX ( Figure S12e). Although LAC47 and 55 were found in the predicted target gene list, the predicted score is rather low, indicating that the interaction between miR408 and LAC47 and 55 is likely very weak. Nevertheless, we carried out 5´ RACE to find whether miR408 can cleave LAC47 and 55. Although bands of the predicted size could be amplified ( Figure SS1), we found from sequencing data that there was no cleavage site in LAC47 and 55 by miR408. Taken together, we are confident that LAC47 and 55 are not real targets of miR408. Accordingly, we added these results in Figure    We also clearly defined in the figure legend that "5 nucleotides were mutated at the predicted binding site of target LACs and miR408, in order to disrupt the miR408 recognition site, while at the same time guaranteeing that the amino acid sequences were unchanged. The mutated CDSs of LAC19, LAC25 and LAC32 were named ΔLAC19, ΔLAC25, and ΔLAC32. The original predicted binding sites of LAC19, LAC25 and LAC32 were TCCAGTGAAGAGGCTGTGCAA, TCCAGTGAAGAGGCTGTGCAA and ACCAGTGAAGAGGCTGTGCAG, respectively, and the mutated binding sites of ΔLAC19, ΔLAC25, and ΔLAC32 were TCCGGTAAAAAGACTGTGTAA, TCCGGTGAAAAGACTCTGTAA and ACCGGTAAAAAGACTGTGTAG, respectively." We annotated this on Figure S14a.
 We apologize for missing the bright field photographs. They have now been added as Point 5: Consistence in nomenclatures: MIR408 or miR408? It shows up in many places.
Response 5: Sorry for some inconsistencies in nomenclature. After referring to the paper in Plant Cell (Jiang et al, 2021), we changed all "MIR408" in the paper including in the manuscript and figures to "miR408" in order to ensure consistency, in which "MIR408" and "miR408" refer to the precursor sequence of miRNA, while "miR408" refers to the mature sequence of 21bp that can bind the cleavage site of the target genes. In the text, it states that two CRISPR lines of miR408 (#8 and #20) were selected for further study but in Fig 1, there are 3 lines (#8/18/20). It is not a major issue, but consistency would be nice.
Response 6: Yes, we agree. In fact, we selected CRISPR lines of miR408 (#8 and #20) for further analysis, after statistical analysis of plant height and stem diameter in three lines. Now we reworded this sentence as "After statistical analysis of plant height and stem diameter in three independent miR408_OX lines (1, 5, and 6) (Supplemental Figure 1c), two independent homozygous lines which had 218bp genomic deletions (miR408_cr #8 and #20) were selected for further study".

Comments of Reviewer 2:
Woody biomass is one of the most important sources of renewable energy around the word.
In the study, the authors showed that overexpression of an miRNA, Pag-miR408 in hybrid poplar can significantly improve plant growth and also saccharification efficiency without acid pretreatment. They further validated that three laccases genes, Pag-LAC19, Pag-LAC25 and Pag-LAC32, are the directly targets of Pag-miR408 in hybrid poplar. LACCASES loss of function mutants also showed significantly enhanced growth and increased saccharification efficiency, and the cell wall deconstruction phenotypes observed are likely the result of laccases down-regulation. The findings revealed the mechanism of Pag-miR408 in lignification and secondary growth, and represent an effective approach towards enhancing biomass and producing lignocellulosic bioenergy for sustainable development.
Overall, the study was well designed and the experiments were well performed, especially those phenotypic analysis and localization investigation of LACs and Pag-miR408 using fluorescence in situ hybridization; the writing is also easy to follow and understand. I only have a few comments as follows, these issues need to be addressed before the manuscript can be accepted: Point 1. Title "Manipulating non-coding RNA…….." better changes to "Manipulating an microRNA miR408…" or "Manipulating Pag-miR408….".
Response 1: Thanks for the helpful suggestions. We have replaced the title with "Manipulating microRNA miR408 enhances both biomass yield and saccharification efficiency in poplar" as suggested.
Point 2. In the Introduction section, the authors only generally described lignocellulosic biomass and the approaches to overcoming the growth defects in lignin-modified plants, the authors should introduce why choosing and focusing miR408.
Response 2: As suggested, we have added several sentences to explain why we selected and focused on miR408 in the "Introduction" section. suggesting that overexpression of miR408 not only enhances biomass yield and saccharification efficiency, but also promotes cambium division, can authors discuss what are the underlined mechanism for miR408 to regulate cambium?
Response 3: Thanks for your suggestion. We have added the discussion about miR408 in regulating cambium division in the "Discussion" section.
"Overexpressing miR408 in poplar contributed the enhanced plant growth associated with a significant increase in net photosynthetic rate, somewhat similar to the phenotype when overexpressing miR408 in Arabidopsis (Pan et al., 2018). In addition, we found miR408_OX poplar showed wider cambium zone and increased xylem area (with enlarged xylem cells). We also showed that miR408 was mainly expressed in the vascular cambium and developing xylem. These results suggest a specific function for miR408 in wood formation, which is characterized by sequential differentiation of vascular cambial cells into xylem cells, cell expansion, massive deposition of secondary cell walls, programmed cell death, and ultimately, the formation of heartwood (Zhang et al., 2014;Wang et al., 2020).
Although these anatomical features were clearly linked to miR408 overexpression or knockout, more work is necessary to understand the coordination of the developmental and were also predicted to be targeted by other microRNAs such as miR475, miR396, miR1447, miR397, miR169, miR7826, miR7466 and miR7817 (Figure 4c)." This result indicates that LAC19, LAC25 and LAC32 can also be targeted by other microRNAs. In the miR408 mutant, to compensate for the lack of function of miR408, the plants may be able to generate other microRNAs to target and down-regulate LAC19, LAC25 and LAC32, so as to ensure that copper ions can be preferentially distributed to key copper-containing proteins (Zhang et al., 2014). Therefore, it is reasonable to believe that lac knockout mutants may not always show an opposite phenotype to that of overexpression poplars in plant height and stem diameter.  respectively. We also repeated the phloroglucinol staining assay of these lines and the new results were similar to the chemical data, showing stronger staining than that of WT.
Based on the new experimental results, we used the new staining photos in Figure 5.   Contents of acid insoluble lignin (AIL), acid soluble lignin (ASL), total lignin. Values are means ± SE (n = 3, n represents 3 trees sampled respectively from each transgenic line). Values are expressed as weight percent based on vacuum-dried extractive free wood weight (%, w/w).

Point 6.
In the discussion section, could authors discuss more the function of laccases?
Response 6: Thanks for the helpful suggestion. We have added discussion about laccase function in the "Discussion" section.
Overexpression of miR408 results in a large increase in saccharification efficiency with no requirement for acid-pretreatment for both laboratory-and field-grown poplar plants. We

Comments of Reviewer 3:
This manuscript describes the phenotypes of transgenic poplar trees expressing a microRNA (miR408) that represses expression of genes encoding laccases. A triple mutant of three laccase-encoding genes phenocopies the miR408-expressing lines. Lignin content and composition are not substantially affected in these lines. However, the yield of glucose in saccharification assays is increased when the materials are treated with cellulase enzymes.
There is an increased abundance of binding sites in transverse sections of these materials for probes that bind to cellulose microfibrils compared to wild type.  Table 1) compared to wild type, and about a 10% reduction in lignin content by AcBr assays.
Of more fundamental scientific interest, but very briefly noted in a single sentence in the Discussion, is the function of laccases in plant cell walls, and more specifically in lignin crosslinking. Laccases have previously been hypothesized to be involved in cross-linking monomeric subunits of lignin, and so, one interpretation of the authors' results is that reduced cross-linking accounts for the observed phenotypes.
This manuscript is rich in experimental results and there are few studies of such depth in transgenic tree species. However, the authors should address a number of issues in the writing of the manuscript before it is suitable for publication.

Point 1. The authors should correct grammatical and typographical errors throughout.
Response 1: Done.
Point 2. Some terms are not precisely defined. For example, line 103, "strongly observed", line 107, 115 and Supplemental Figure 7, "loosely arranged xylem". In particular, the authors use the term "accessibility" to describe both macroscopic properties (glucose yields in saccharification assays) and microscopic properties (binding sites in transverse sections for cellulose-binding probes). I assume that the authors mean the accessibility of cellulose microfibrils to cellulase enzymes. It is reasonable to infer that both of these phenotypes are correlated with cellulase accessibility to its substrate but these are proxy measurements of accessibility rather than direct measurements.

Response 2:
We have made revisions to more precisely define the terms we are using. For the sentence on line 103-105, "GUS signal was detected in leaf veins (Supplemental Figure   3b-c) and was strongly observed in root vascular tissue (Supplemental Figure 3d), strong in the vascular cambium that will differentiate into xylem, but weak in mature xylem (Supplemental Figure 3e-g)". The sentence has been changed into "GUS signal was detected in leaf veins (Supplemental Figure 3b-c) and in root vascular tissue (Supplemental Figure 3d).
Promoter activity of miR408 was mainly detected in the vascular cambium that will differentiate into xylem, but weak in mature xylem (Supplemental Figure 3e-g)."  Agreed, for the sentence in line 107, 115 and Supplemental Figure 7, the term of "loosely arranged xylem" is inappropriate.  For the sentence in line 115, "The loosely arranged xylem" has been changed into "To test whether the enlarged xylem cells might possess more loosely-organized cell walls, we utilized green fluorescent protein………".
 For the sentence in Supplemental Figure 7, "were also more loosely arranged" has been changed into "were also enlarged".
 For the accessibility, we used the term following Ding et al (2012, Science). We agree that the term "accessibility" in the context of our work means "accessibility of cellulose microfibrils to cellulase enzymes", thus we changed the title in line 113 into "Overexpression of miR408 increases secondary cell wall accessibility to cellulase enzymes and enhances saccharification efficiency."  For the sentence in line 115, "The loosely arranged xylem might increase the accessibility of secondary cell walls to cellulase." has been changed into "The enlarged xylem cells might increase secondary cell wall accessibility of cellulose microfibrils to cellulase enzymes".
 For the sentence in line 124-125, "Similarly increased cell wall accessibility" has been changed into "Similarly increased cell wall accessibility of cellulose microfibrils to cellulase enzymes……………….".
 For the sentence in line 164-165, the title "Overexpression of miR408 enhances biomass yield and saccharification efficiency in field-grown plants" has been changed into "Overexpression of miR408 enhances biomass yield and accessibility of cellulose microfibrils to cellulase enzymes in secondary cell walls of field-grown plants"  For the sentence on line 171, "increased cell wall accessibility to hydrolytic enzymes" has been changed into "increased cell wall accessibility of cellulose microfibrils to cellulase enzymes".
 For the sentence in line 325, "cell wall accessibility" has been changed into "cell wall accessibility of cellulose microfibrils to cellulase enzymes".
 For the sentence in line 337, "showed increased cellulase accessibility" has been changed into "showed increased cell wall accessibility of cellulose microfibrils to cellulase enzymes".
 For the sentence in line 342, "was" has been changed into "were".
 We also agree that the accessibility of cellulose microfibrils to cellulase enzymes is a proxy measurement of accessibility rather than a direct measurement. Accordingly, we carried out additional chemical analysis of saccharification efficiency of lac mutants (Table S5). The results of chemical saccharification assays accorded with the microscopic results. Point 3. Loss-of-function phenotypes from knockout of the miRNA-encoding gene are not described or discussed beyond a cursory mention in the text. However, Supplemental Figure   9b and c shows interesting phenotypes of reduced cell wall thickness and an increased intensity of phloroglucinol staining in successive internodes. Discussion of these phenotypes could enrich understanding of the function of miR408.
Response 3: After reading your comments, we realized that loss-of-function phenotypes of miR408 were not well discussed in detail. Therefore, we added the discussion of miR408 in cell wall thickening in the revised manuscript as follows: "Previous studies showed that over-expression miR408 can promote vegetative growth, while the impaired growth was observed in miR408 T-DNA insertion Arabidopsis mutant lines (Zhang et al., 2013). In addition, cell wall thickening was associated with the deposition of lignin and cellulose (Watanabe et al., 2015;Liu et al., 2021). In our study, the knock-out of miR408 in poplar resulted in enhanced cell wall lignification but reduced cell wall thickness.
Although lignin related genes were up-regulated in the miR408 knock-out plants from the RNA-seq data, the genes encoding activators (VND7 and SND1) (Li et al., 2012;Yang et al., 2013;Takata et al., 2019) in the transcriptional regulatory network pathway of secondary wall synthesis were down-regulated. Moreover, the expression level of LBD15, a key TF that can down-regulate the expression of cellulose synthesis genes (Zhu et al., 2014), was increased to a large extent in miR408 knock-out poplars. The genetic evidence and gene expression analysis together suggested that miR408 may possess an additional role of regulating cell We also added some discussion on the increased intensity of phloroglucinol staining of miR408_cr poplars in the Results part "miR408 targets LAC19，LAC25 and LAC32 " as Supplemental Figure 12c shows altered expression ratios of Laccases 1, 3, 4, 10, 11 and 17, but not 19, 25 or 32 in the miR408-expressing lines. Is this a typographical error? Or is the expression of these other laccase genes also impacted in the laccase triple mutant?
Response 5: To address this issue, we carried out qRT-PCR and 5´ RACE for the identification of target genes.
 See Response 2 to Reviewer 1.
 As for the question about other laccases in Supplemental Figure 12c, we confirm that the altered expression ratios are not a typographical error.  To confirm the transcript levels of LAC1, LAC3, LAC4, LAC10, LAC11 and LAC17, we searched in the RNA-seq data between the triple laccase mutants and WT (Table SS2).
The results showed that the transcript levels of LAC1, LAC4 and LAC11 were downregulated, while LAC10 was upregulated. In addition, LAC3 and LAC17 were almost unexpressed in stem.    term 7/20 means that seven of twenty clones from the PCR products contained an miR408guided cleavage 5´ end that mapped precisely to exon 2. Based on 5´ RACE assays of LAC25 ( Figure 4f) and LAC32 (Figure 4g), the cleavage sites were all located at exon 2, and six and seven, respectively from the twenty PCR products mapped precisely to the cleavage sites".
Point 7. The authors should cite relevant literature and discuss their own findings in the context of literature with respect to the function of laccases in lignification. For example, a dirigent protein (Dirigent protein 23) is also a predicted target gene of miR408 (Supplemental Table 2), and this class of protein has also been implicated in lignin cross-linking. How is its expression affected in the triple laccase mutant and does this impact (or not) the interpretation of the triple mutant phenotype?
Response 7: Thanks for the helpful suggestions. As suggested, we added several sentences to the Discussion section; see Response 6 to Reviewer 2.
To clarify the function of the predicted target gene, dirigent 23 (Potri.001G214600) as suggested, we searched its transcript levels from the transcriptome data between MIR408_OX and WT ( Table 2 in our reply), and between the triple laccase mutants and WT (Table 3 in our reply; Data not show in this paper). The results showed that dirigent 23 (Potri.001G214600) is almost not expressed in WT, miR408_OX and triple laccase mutants, as shown in the following tables, suggesting that the expression level of this gene is very low in stem and therefore unlikely to play a crucial role in stem development, implying dirigent 23 (Potri.001G214600) is not a key gene in lignin cross-linking. This is consistent with only very few reports suggesting roles for dirigent proteins in lignin (as opposed to lignan) biosynthesis, although very recent data suggest involvement of DPs in lignification in the Casparian strip.   Response 8: Sorry for the unclear definition of "Cell wall residues". Here we use the term to refer to the cell wall material extracted by methanol and chloroform after grinding through a 40-mesh sieve. The specific extraction process is as follows: cell wall residues were