Ectopic Expression of PtrLBD39 Retarded Primary and Secondary Growth in Populus trichocarpa

Primary and secondary growth of trees are needed for increments in plant height and stem diameter, respectively, affecting the production of woody biomass for applications in timber, pulp/paper, and related biomaterials. These two types of growth are believed to be both regulated by distinct transcription factor (TF)-mediated regulatory pathways. Notably, we identified PtrLBD39, a highly stem phloem-specific TF in Populus trichocarpa and found that the ectopic expression of PtrLBD39 in P. trichocarpa markedly retarded both primary and secondary growth. In these overexpressing plants, the RNA-seq, ChIP-seq, and weighted gene co-expression network analysis (WGCNA) revealed that PtrLBD39 directly or indirectly regulates TFs governing vascular tissue development, wood formation, hormonal signaling pathways, and enzymes responsible for wood components. This regulation led to growth inhibition, decreased fibrocyte secondary cell wall thickness, and reduced wood production. Therefore, our study indicates that, following ectopic expression in P. trichocarpa, PtrLBD39 functions as a repressor influencing both primary and secondary growth.


Introduction
Trees, vital to the forest ecosystem, serve as valuable renewable energy sources.Tree wood, its primary product, is used for energy combustion, pulping, paper production, and potential lignocellulosic biofuel production [1].Wood, also known as secondary xylem, develops through the cyclic vascular cambium activity.This process encompasses vascular cambium cell differentiation into secondary xylem mother cells, cell expansion, mass deposition of secondary walls and programmed cell death, and final heartwood formation [2,3].The secondary xylem includes vessel cells, fibrocytes, and axial parenchyma cells [4], influencing wood properties based on their proportions [5].
We analyzed the expression of LBD genes in stem-cambium, stem-xylem, stemphloem, shoot, leaf, and root of Poulus trichocarpa and found two highly stem-phloemspecific members, PtrLBD39 and PtrLBD22.In this work, we focused on the function of PtrLBD39.While previous studies showed that LBDs exhibit different roles in plant growth, we found that transgenic P. trichocarpa plants expressing PtrLBD39 under a cauliflower mosaic virus 35S promoter resulted in severely stunted phenotypes.Therefore, we hypothesized that ectopic expression of PtrLBD39 disrupted the regulation of plant growth and development-related genes.In this study, we conducted an analysis encompassing growth, anatomical phenotypes, transcriptomics combined with ChIP-seq [16], and weighted gene co-expression network analysis (WGCNA) of PtrLBD39 overexpression plants.We found that ectopic overexpression of PtrLBD39 inhibits both primary and secondary growth in transgenic plants by regulating genes across multiple biological pathways.

Tissue-Specific Expression of PtrLBD39 in P. trichocarpa
We previously used laser capture microdissection (LCM) to obtain the gene expression levels of different stem vascular tissues (stem-cambium, stem-xylem, and stemphloem) [32].The LCM RNA-seq data were normalized with the gene expression levels of leaves, roots, and shoots of P. trichocarpa [33], and transcription factors specifically expressed in these six tissues were screened.Among them, we found two highly stemphloem-specific members, namely PtrLBD39 and PtrLBD22 (Figure 1).In this study, we selected PtrLBD39 for functional analysis in transgenic P. trichocarpa.

Ectopic Expression of PtrLBD39 in Transgenic P. trichocarpa Resulted in Stunted Phenotypes
Seven independent transgenic lines (OE-PtrLBD39-L1~L7) exhibited elevated PtrLBD39 transcript abundances in xylem (Figure S1).The three lines (L1, L2, L3) demonstrating the highest expression were chosen for subsequent analysis.In comparison to the wild type (WT), overexpressed plants displayed markedly reduced height (Figure 2A,B) and smaller, more curled leaves (Figure S2), indicating that ectopic expression of PtrLBD39 transgenic significantly inhibited the growth of P. trichocarpa.There was no significant growth improvement in transgenic plants over time.OE-PtrLBD39 significantly increased the internode number while reducing their length compared to WT (Figures 2C and S2), thereby indicating that extreme internode shortening, prompted by PtrLBD39 overexpression, contributed to the dwarf phenotype.Furthermore, stem basal diameter analysis revealed a significant decrease in OE-PtrLBD39 compared to WT (Figure 2D).To investigate the impact of PtrLBD39 on wood formation, we paraffin-sliced and stained the 4th, 6th, 8th, 10th, and 20th stem segments of the greenhouse-grown 4-monthold WT and OE-PtrLBD39 plants with Safranin O and Fast Green dye for observation (Figure 3).Comparative analysis revealed that OE-PtrLBD39 exhibited non-lignified fiber cell walls in the xylem and phloem regions, distinctly differing from WT (Figure 3A,B).This demonstrates that PtrLBD39 overexpression strongly suppresses xylem and phloem fiber cell wall lignification, suggesting its crucial regulatory role in fiber cell development.Additionally, these transgenic plants displayed a significant reduction in single vessel area while the fiber cell area increased compared to the wild type (Figure 3C).Furthermore, OE-PtrLBD39 exhibited an increased vessel count per unit area and a reduced number of fiber cells (Figure 3D).However, the morphological structure of cambium cells in OE-PtrLBD39 remained largely unchanged from that of WT (Figure S3).(C) Statistics analysis of the mean lumen area of individual fiber and vessel cell (mm 2 ) within the 20th internode.(D) Statistics analysis of the number of fiber and vessel cells per cross-sectional area (mm 2 ) within the 20th internode.Error bars represent SE values of three independent replicates with at least 150 vessel cells and 3000 fiber cells for each genotype in each replicate, and asterisks indicate significant differences between the transgenics and wild-type plants found using Student's t-test (* p < 0.05, ** p < 0.01).
Since the ectopic expression of PtrLBD39 affects the fiber cell wall lignification, both scanning electron microscopy (SEM) and transmission electron microscopy (TEM) were used to observe the cell walls of OE-PtrLBD39 transgenic plants.SEM results demonstrated significant thinning of the fiber cell wall compared to WT, while the vessel cell wall remained largely unchanged (Figure 4A).Moreover, TEM analysis results revealed a significantly thinner fiber cell wall in OE-PtrLBD39 plants, with the absence of the S3 layer and partial S2 layer of the secondary cell wall (SCW), as compared to the WT (Figure 4B).These results indicated that the ectopic expression of PtrLBD39 directly affected fibrocyte cell wall development in the vascular tissue of P. trichocarpa.Since the ectopic expression of PtrLBD39 affects the fiber cell wall lignification, both scanning electron microscopy (SEM) and transmission electron microscopy (TEM) were used to observe the cell walls of OE-PtrLBD39 transgenic plants.SEM results demonstrated significant thinning of the fiber cell wall compared to WT, while the vessel cell wall remained largely unchanged (Figure 4A).Moreover, TEM analysis results revealed a significantly thinner fiber cell wall in OE-PtrLBD39 plants, with the absence of the S3 layer and partial S2 layer of the secondary cell wall (SCW), as compared to the WT (Figure 4B).These results indicated that the ectopic expression of PtrLBD39 directly affected fibrocyte cell wall development in the vascular tissue of P. trichocarpa.

The Ectopic Expression of PtrLBD39 Significantly Affects the Lignin Structure and Polysaccharide Contents
In light of the inhibited fiber cell wall development in transgenic plants, we quantified the three wood cell wall components.While total lignin content in stem wood of OE-PtrLBD39 transgenics remained unchanged compared to WT (Table 1), there was a significant alteration in lignin composition.This alteration showed a ~14% reduction in S-mon-

The Ectopic Expression of PtrLBD39 Significantly Affects the Lignin Structure and Polysaccharide Contents
In light of the inhibited fiber cell wall development in transgenic plants, we quantified the three wood cell wall components.While total lignin content in stem wood of OE-PtrLBD39 transgenics remained unchanged compared to WT (Table 1), there was a significant alteration in lignin composition.This alteration showed a ~14% reduction in S-monomers and ~34% increase in G-monomers, leading to S:G ratios ranging from 0.8 to 1.63 in the transgenics as compared with 2.0 in the wild type (Table 2), thereby aligning with anatomical phenotypes observed in paraffin sections of OE-PtrLBD39 transgenics (Figure 3A).Furthermore, significant reductions were evident in the total wood glucose (~12% reduction) and xylose (~21% reduction) (Table 1), thus suggesting that cellulose and hemicellulose biosynthesis were strongly inhibited in the presence of a high level of PtrLBD39 transcripts in xylem.We then examined the effects of PtrLBD39 overexpression on the transcript levels of downstream genes.
In OE-PtrLBD39, the known monolignol biosynthetic pathway genes and SCW cellulose biosynthetic genes (PtrCesA4, PtrCesA7, PtrCesA17, PtrCesA8, and PtrCesA18) exhibited significant reduction in transcript levels [34,35] (Figure 5A and Table S2).Similarly, genes associated with hemicellulose biosynthesis showed a considerable decrease in expression (Figure 5C and Table S2).These changes in cell wall component genes corresponded to alterations in wood components (Tables 1 and 2).After ectopically expressing PtrLBD39, we observed stunted plant growth, indicating its influence on primary and secondary growth in trees.This overexpression notably altered the transcription levels of STM and LBD15, both crucial for plant primary and secondary growth.PtrLBD39 overexpression upregulated the transcription factor STM, encoding the KNOX1 protein (Table S1) [36].Conversely, the expression of LBD15, involved in stem apical meristem development and regulation of secondary xylem cell differentiation, decreased (Table S1) [37,38].Moreover, several hormone response factors governing tree growth and development were upregulated post-PtrLBD39 ectopic expression.Notably, cytokinin response factors CRF2, CRF4, and CRF11 showed elevated expression levels.Additionally, ERF1, a direct target of EIN3 in the ethylene signaling pathway, exhibited increased expression (Table S1) [39].Moreover, the expression levels of GA2OX6, GA2OX8, GA3OX1 and GA20OX2 in gibberellin (GA) metabolism were altered (Table S1).Significantly, distinct alterations in the expression levels of key TFs associated with secondary growth were observed.For instance, ANT was primarily expressed in the vascular cambium, and, along with APL genes, was essential in regulating Arabidopsis thaliana phloem properties also being upregulated (Table S1) [1,40].The differential expression of these TFs, governing primary growth, secondary growth, and hormonal responses in OE-PtrLBD39, suggests the potential influence of PtrLBD39 on the wood formation process through the regulation of these TFs.
Upon analyzing ChIP-seq data from OE-PtrLBD39-FLAG transgenic plants [16], we identified 162 upregulated DEGs and 11 downregulated DEGs, associate targets of PtrLBD39 (Table S3).The binding motifs within the promoters of PtrLBD39 target genes were analyzed using STREME [41].The results showed that four top-ranked motifs in binding peaks of targets' promoter for TF binding (p value < 2.2 × 10 −8 ; Figure S5), indicated that PtrLBD39 may regulate the target genes by binding to these potential motifs.Among the TF genes bound by PtrLBD39, HD-ZIP family TFs HB6 and ARF family TFs MP, linked to hormone response, exhibited upregulation (Table S3) [42,43].The STY1 gene, involved in auxin biosynthesis, showed a modest increase in transcription levels [44,45].Additionally, the expression level of the growth-related NERD increased by 2-3 times (Table S3) [46].PtrMYB128 (MYB103), a member of the transcriptional network regulating secondary cell wall biosynthesis, was directly downregulated by PtrLBD39 (Table S3) [47][48][49].These changes in the expression levels of TFs involved in hormone signaling response and secondary growth confirmed PtrLBD39's regulation of these genes through promoter binding, thereby regulating tree growth.

WGCNA Reveals Gene Modules Associated with Phenotypes
To identify genes linked to the OE-PtrLBD39 trait, we performed WGCNA analysis on DEGs in OE-PtrLBD39 RNA-seq data (Table S1) and correlated phenotypic data encompassing primary (plant height and stem segment count), secondary growth (stem basal diameter), and wood properties (wood composition and lignin content) (Table S4).Unexpressed or under-expressed genes were excluded from analysis, with WGCNA detection being performed for genes with an average FPKM expression >1.Utilizing the soft threshold calculation (Figure S6), β = 20 was chosen to construct a network.The dynamic shear tree method combined similar expression modules, yielding 10 co-expression modules (differentiated by different colors) comprising gene counts in different modules ranging from 215 to 6405.Each highly correlated gene set corresponds to a branch in the gene dendrogram (Figure S7), with module-phenotype correlations demonstrating significant overlap between genes in the same module (Figure 6).Notably, three modules (yellow, black, and brown) exhibit the strongest correlations with the OE-PtrLBD39 trait data (q < 0.05, |R| > 0.85) (Figure 6), suggesting that these genes primarily impact tree growth and development.Consequently, we further scrutinized these three modules.; Y-axis represents the co-expression modules (Genes with similar expression patterns are grouped into one module.One color represents one module of genes).The table is color-coded to represent correlation strength according to the color legend.The intensity and direction of each correlation are indicated on the right side of the heatmap (red, positively correlated; blue, negatively correlated).Numbers on the table represent the correlation coefficient between the corresponding module and the trait, with q-values listed below the correlations in parentheses.Modules within boxes with black lines: q < 0.05, |R| > 0.85.

Functional and Pathway Enrichment Analysis of the Eigengenes in Modules
There were 771, 107, and 790 DEGs in the yellow module, black module, and brown module, respectively (Table S5).Correlation analysis showed that the yellow module exhibited positive correlations with plant height, basal diameter, H-type lignin, and glucose content, while being negatively correlated with the internode number.The black module primarily relates to lignin components: S-type lignin content and S/G ratio were significantly negatively correlated, while G-type lignin content showed a positive correlation.Conversely, the brown module displayed an opposing pattern to the yellow module concerning plant growth traits; for instance, plant height and number of basal diameters exhibited significant negative correlations (Figure 6).This suggests a mutual inhibiting effect on tree growth between genes in the yellow and brown modules.
To explore the role of DEGs in the yellow, black, and brown modules, we performed GO functional enrichment analyses on their respective genes (Figure S8).Data analysis revealed significant enrichment of genes in these modules linked to various GO terms encompassing biological processes, cellular components, and molecular functions.; Y-axis represents the co-expression modules (Genes with similar expression patterns are grouped into one module.One color represents one module of genes).The table is color-coded to represent correlation strength according to the color legend.The intensity and direction of each correlation are indicated on the right side of the heatmap (red, positively correlated; blue, negatively correlated).Numbers on the table represent the correlation coefficient between the corresponding module and the trait, with q-values listed below the correlations in parentheses.Modules within boxes with black lines: q < 0.05, |R| > 0.85.

Functional and Pathway Enrichment Analysis of the Eigengenes in Modules
There were 771, 107, and 790 DEGs in the yellow module, black module, and brown module, respectively (Table S5).Correlation analysis showed that the yellow module exhibited positive correlations with plant height, basal diameter, H-type lignin, and glucose content, while being negatively correlated with the internode number.The black module primarily relates to lignin components: S-type lignin content and S/G ratio were significantly negatively correlated, while G-type lignin content showed a positive correlation.Conversely, the brown module displayed an opposing pattern to the yellow module concerning plant growth traits; for instance, plant height and number of basal diameters exhibited significant negative correlations (Figure 6).This suggests a mutual inhibiting effect on tree growth between genes in the yellow and brown modules.
To explore the role of DEGs in the yellow, black, and brown modules, we performed GO functional enrichment analyses on their respective genes (Figure S8).Data analysis revealed significant enrichment of genes in these modules linked to various GO terms encompassing biological processes, cellular components, and molecular functions.Primarily, these genes are associated with metabolic and cellular processes, cell components, as well as binding and catalytic activities (Figure S8).Notably, the molecular function category, particularly the binding term, contains the largest number of genes.Binding is implicated in diverse biological processes such as signal transduction, metabolic pathways, gene regulation, and protein interactions, suggesting its involvement in regulating plant growth and development.MYBs and NACs, specifically known for their role in secondary cell wall synthesis and xylem differentiation, are notably abundant within the 'binding' term [8,12,50].Furthermore, this term encompasses genes associated with hormone signaling pathways, including auxin response factor ARFs and ERFs involved in ethylene synthesis.These findings strongly suggest that these genes, particularly the MYB and NAC family genes along with those related to hormone signaling pathways, may contribute to the emergence of the OE-PtrLBD39 phenotype.They potentially play a crucial role downstream of PtrLBD39 in tree growth and morphogenesis.

WGCNA Reveals the Genes Expression Directly Related to Phenotype in P. trihcocarpa
In the examination of genes regulated by PtrLBD39 within the modules and their impact on phenotypes, the integration and analysis of OE-PtrLBD39 ChIP-seq data [16] with DEGs in three modules revealed that 27 target genes were regulated by PtrLBD39 in the yellow module.This set includes six TFs (Table S6).Notably, PtrMYB128, (poplar fiber cell wall thickness regulation) and PtrCAD1 (cinnamyl alcohol dehydrogenase; a monolignol biosynthesis enzyme) were significantly downregulated in OE-PtrLBD39 (Table S6) [7,[47][48][49]51,52].Moreover, the TF MP/ARF5, known for mediating Arabidopsis morphogenesis and vascular development, displayed significant upregulation in OE-PtrLBD39 (Table S6) [43].
In the black module, PtrLBD39 directly regulates three target genes, notably including the upregulation of the EOL1 gene associated with Arabidopsis ethylene biosynthesis in OE-PtrLBD39 (Table S6) [53].The brown module reveals PtrLBD39's direct regulation of 38 target genes, encompassing six TFs.These genes are linked to plant short stem, lateral root development, leaf, and cell morphogenesis [42,[54][55][56][57].For instance, Arabidopsis thaliana's homologous genes-LRL3, HB6, WRKY28, SRS5, and OBP1-were all upregulated in OE-PtrLBD39 (Table S6).These analyses indicate that PtrLBD39-mediated direct regulation of target genes within the modules potentially influenced both the primary and secondary tree growth as well as xylem secondary cell wall development.

Discussion
As per the transcriptome data of six tissues (stem-cambium, stem-xylem, stemphloem, shoot, leaf, and root) in P. trichocarpa, PtrLBD39 is highly specific in the stemphloem (Figure 1).Upon overexpressing PtrLBD39 in P. trichocarpa, its expression in the xylem of different transgenic lines was 1000-7500 times higher (Figure S1).However, its ectopic expression of PtrLBD39 severely reduced the plant height, leaf size, and stem basal diameter (Figures 2 and S2), thus profoundly impacting the primary and secondary growth.Overexpression of PtrLBD39 resulted in diverse alterations in cell size and number within xylem fiber and vessel cells (Figure 3); however, it did not impact the stem vascular cambium cells (Figure S3).This suggests that ectopic expression of PtrLBD39 affected the expansion and proliferation of specific cell types.In the transgenic plants, the secondary cell wall of the xylem vessels resembled that of the wild type, but the xylem fiber cell wall was significantly thinner (Figure 4).Consistent with the phenotype of fiber cell wall, glucose and xylan, the primary components of cellulose and hemicellulose, respectively, were significantly reduced in OE-PtrLBD39 (Table 1).Lignin in dicotyledonous plants mainly comprises S-type and G-type lignin, which mainly comes from fiber and vessel cells, and usually exists in a ratio of 2:1, respectively [58][59][60][61].We found that ectopic expression of PtrLBD39 altered lignin structures in the cells.The lignin alteration reduced lignin S:G ratios by 20-60% due to PtrLBD39 ectopic expression, explaining the non-lignification of the fiber cell wall but normal lignification of the vessel cell wall (Table 2 and Figure 3).These phenotypes of cell wall and component changes were consistent with the reduction in abundance of cell wall component gene transcripts (Figure 5).
In the transcriptome data of PtrLBD39 overexpression, we observed upregulation of STM and LBD15, both known for their roles in regulating primary and secondary plant growth (Table S1) [62].STM, a KNOX1 protein class, reportedly inhibited cell differentiation by inducing IPT7 and A-type ARR5 expression [36].ARK1, a poplar homolog of STM, reportedly altered cell wall biochemistry and lignin composition when overexpressed in plants [62], likely through inhibiting cell differentiation.The upregulation of STM in OE-PtrLBD39 plants likely leads to reduced height, altered cell wall thickness, and wood composition.Conversely, OE-PtrLBD39 resulted in a decreased transcriptional abundance of LBD15.Studies have shown that AtLBD15 regulates stem apical meristem development in Arabidopsis and regulates WUS expression [37], potentially inhibiting the primary growth in PtrLBD39 overexpressed plants.LBD15 also plays a role in secondary cell wall synthesis during xylem cell differentiation [7,38,63].Increased LBD15 expression in Arabidopsis was found to downregulate CesA4, CesA7, and CesA8, which are vital for mediating secondary cell wall cellulose synthesis, thereby disrupting secondary cell wall formation [38].Although a decrease in LBD15 expression would theoretically upregulate these cellulose synthase genes, the cellulose synthase gene and secondary cell wall formation are regulated by multiple factors.
Plant hormones regulate plant development.Ethylene influences secondary stem growth by stimulating cambium activity, xylem development, and fibre and vessel formation.Ethylene binds to ethylene receptors (ETRs) at the endoplasmic reticulum, initiating downstream signaling pathways involving phosphorylation, ubiquitination, and translocation, activating a set of TFs, namely ETHYLENE INSENSITIVE 3/ETHYLENE INSENSITIVE3-LIKE1 (EIN3/EIL1) and Ethylene Response Factor (ERF) TFs [64].ERF1, a direct target of EIN3, is ethylene-sensitive [39].Therefore, the upregulated ERF1 expression likely contributes to the unique phenotype of transgenic PtrLBD39.Cytokinin comprises a class of N6-substituted adenine derivatives that can actively regulate various cellular and developmental processes in plants [65].During the cell proliferation phase, cytokinin is necessary to maintain cell proliferation and is a key factor in mediating leaf cells from proliferation to expansion by blocking the cell transition to expansion and the initiation of photosynthesis [66].Cytokinin Response Factors (CRFs), such as CRF2, CRF4, and CRF11, members of the APETALA2 TF family, are upregulated by cytokinin transcription.They regulate gene expression involved in cytokinin biosynthesis (IPT and LOG) and degradation (cytokinin oxidase/dehydrogenase, CKX) and positively regulate type-A ARRs [36].It has also been shown that CKG, a cytokinin response factor, mediates CK-dependent regulation of Arabidopsis cell expansion and cell cycle progression.Overexpression of CKG can increase cell size and promote cell cycle progression.The mutant CKG in Arabidopsis shows smaller cotyledon [67].The elevated expression levels of CRF2, CRF4, and CRF11 in OE-PtrLBD39 may be one of the main reasons for the alterations in the number of xylem fiber and vessel cells (Figure 3).Bioactive gibberellin also plays a variety of roles in plant development [68][69][70].Gibberellin can promote plant growth, including promoting plant cell division and cell elongation [71].In higher plants, the content of endogenous active gibberellin is balanced by its biosynthesis and catabolism, which determine the regulatory behavior of gibberellin during plant tissue development.In our transcriptome data, the expression level of seven genes related to GA biosynthesis (GA3OX and GA20OX) and GA catabolism (GA2OX) were altered (Table S1), which may cause changes in the content and distribution of GA in transgenic plants and is responsible for affecting plant height and leaf size of transgenic plants.
An increasing number of studies have proven that crosstalk among different plant hormones has an effect on plant growth [72][73][74][75].Auxin, cytokinin, ethylene, and GA affect the activity of the stem vascular cambium [74][75][76].Changes in the content and distribution of these plant hormones in different types of cells are responsible for cell division and differentiation, as well as fiber and vessel SCW formation, including SCW deposition, wall thickening, and cellulose biosynthesis [73,77].In this study, the ectopic expression of PtrLBD39 resulted in the alteration in expression levels of numerous genes related to plant hormone metabolism and regulation, stem apical tissue and secondary cell wall composition (Table S1, Figure S4), which may have changed the hormone homeostasis in transgenic plants, thus seriously affecting the primary and secondary growth of the plants.
Following an integrated RNA-seq and ChIP-seq analysis [16], we identified 162 genes upregulated and bound to PtrLBD39, among which 28 are TFs.Notably, these TFs encompass MP and NERD and are crucial in primary growth processes.MONOPTEROS (MP) acts as auxin response TFs which are vital for root formation and vascular development [43].NERD interacts with the exocyst, impacting root growth and cell expansion.The exocyst, a conserved octamer protein complex, mediates plasma membrane secretion and regulates developmental processes like root meristem size, cell elongation, and tip growth.NERD1 indirectly interacts with the exocyst, potentially influencing cell wall polysaccharides and developmental traits [46].The variations in plant height, cell wall thickness, and wood composition due to PtrLBD39 ectopic expression likely result from changes in MP and NERD transcription levels.PtrLBD39 also directly upregulates other TFs like HB6 and STY1, which are involved in hormone regulation.Studies have shown that HB6 (ARA-BIDOPSIS THALIANA HOMEBOX 6) regulates leaf edge development.HB6 participates in cell division and organ differentiation while negatively regulating ABA signaling during abiotic stress in vitro [42].STY1, a transcriptional activator of the auxin biosynthesis gene, and its related genes promote the style's development by regulating auxin homeostasis and the pistil's apical pattern [44,45].The alteration in leaf morphology due to PtrLBD39 ectopic expression could be influenced by these two genes.Moreover, PtrLBD39 was found to downregulate 11 genes.Notably, the TF PtrMYB128 acts as a master switch in P. trichocarpa's secondary wall biosynthesis, previously identified as a member of the transcriptional network regulating secondary wall biosynthesis in Arabidopsis xylem tissue.It is a direct transcriptional target of SECONDARY WALL ASSOCIATED NAC DOMAIN PROTEIN 1 (SND1), which has been proposed to influence cellulose biosynthesis [47][48][49].Consequently, PtrMYB128 downregulation might account for reduced secondary cell wall thickness in transgenic plants.
To explore the association more accurately between abnormal phenotypes and DEGs in plants overexpressing PtrLBD39, we conducted a WGCNA.Out of the 10 co-expression modules, we obtained three target modules for in-depth analysis.Notably, genes within the yellow and brown modules exhibited antagonistic roles in primary plant growth regulation, while the yellow, black, and brown modules collectively regulated both primary and secondary tree growth.It is noteworthy that the downregulation of MYB85, a key regulator of phenylalanine and lignin biosynthesis during secondary cell wall formation [78], and its dominant inhibition significantly reduces secondary wall thickening in fiber cells [79], is downregulated in the yellow module (Table S5), may result in the defective phenotype of OE-PtrLBD39.KNAT1 is an important apical meristem maintenance regulator, with dual effects of promoting cell proliferation and inhibiting differentiation [80].It was upregulated in the brown module (Table S5), which may have resulted in reduced plant height.Furthermore, the upregulated expression of NTL9, a member of the NAC TF family known to negatively regulate vascular cambium development during secondary growth of Arabidopsis stems [81], inhibited secondary growth in OE-PtrLBD39 plants (Table S5).
In conclusion, PtrLBD39 ectopic expression impacts the expression of gene sets crucial for primary and secondary plant growth, leading to reduced plant height and radial growth.This suggests that PtrLBD39 can regulate the growth of trees as an inhibiting factor.

Plant Materials and Growth Conditions
All experiments were performed with Populus trichocarpa genotype Nisqually-1.Wildtype and transgenic plants were grown in a greenhouse as described previously [82].Four-month-old stems of P. trichocarpa plants were used for RNA extraction, growth index measurement, histological analysis, and scanning electron microscopy.Six-month-old plants were used for wood composition analysis and transmission electron microscopy.

Generation of Gene Overexpression Transgenic P. trichocarpa
The coding sequence of PtrLBD39 (Potri.005G097800)was amplified from P. trichocarpa plants and cloned into the pBI121 vector under the control of the CaMV 35S promoter for generating overexpression construct.The plasmid was introduced into Agrobacterium tumefaciens strain GV3101 for P. trichocarpa transformation as described in detail previously [83].The expression level of PtrLBD39 gene in the xylem of transgenic plants was detected by RT-qPCR.Primers are listed in Table S7.

RT-qPCR
Total RNA was isolated from the SDX tissues of P. trichocarpa by using the Qiagen RNeasy Mini Kit.One microgram of total RNA was reverse-transcribed to cDNA using the PrimeScript RT Reagent Kit with gDNA Eraser (Takara, Dalian, China).All RT-qPCR reactions were performed on an Agilent Mx3000P QPCR System with FastStart Universal SYBR Green Master (Roche, Basel, Switzerland).PtrActin was used as a housekeeping gene to calculate the relative expression level in wild-type and transgenic plants.Three biological replicates were used for each RT-qPCR experiment.The relative expression levels were calculated by the 2 −∆∆Ct method [84].The primers are listed in Table S7.

Scanning Electron Micrograph Analysis
Fresh stem segments of the 20th internode of OE-PtrLBD39 transgenic and wild-type plants were collected and coated with gold at 10 mA for 60 s.The samples were imaged using a scanning electron microscope JCM-5000 NeoScope (NIKON, Tokyo, Japan).

Transmission Electron Microscopy
Samples were taken from the 20th stem segments of the wild-type and OE-PtrLBD39 transgenic plants (6-month-old) and cut into samples of 3 mm length and 1 mm width, then fixed in a fixative solution containing 2.5% glutaraldehyde (stored at 4 • C) by vacuum osmosis.The samples were pumped using a vacuum pump for 30 min to two hours.They were then rinsed with 0.1 M phosphate buffer (pH 6.8), fixed with 1% osmic acid fixing solution, and then rinsed with 0.1 M phosphate buffer (pH 6.8).After multi-stage ethanol dehydration at 4 • C, the samples were put into 100% acetone.Soaked with acetone and embedding solution at room temperature and embedding the stem segments in the next day.After seven days of polymerization, the samples were cut into 50~60 nm sections with an ultra-thin microtome, and then stained with uranium-lead citrate double.Finally, transmission electron microscopy was used to observe and take the pictures.

Wood Chemistry
Stem segments of six-month-old OE-PtrLBD39 transgenic and wild-type plants were extracted with 90% (v/v) acetone for 48 h and then transferred into 100% (v/v) acetone for 14 days with fresh acetone replaced every two days.The air-dried wood was used to quantify the wood composition (acid-insoluble lignin, acid-soluble lignin, and sugars) and lignin composition (S-lignin, G-lignin, and H-lignin) following established procedures with three replicates for each assay [52,85].

RNA-Seq and Data Analysis
RNA-seq was carried out with total RNA of SDX tissues isolated from wild-type and OE-PtrLBD39 transgenic plants.The RNA-seq libraries of overexpression transgenics (OE-PtrLBD39-L1, L2 and L3) and wild-type plants were generated with three biological replicates per sample using the NEBNext Ultra II RNA library preparation kit.A total of 12 libraries were sequenced using an Illumina HiSeq4000 to generate paired-end reads.For all libraries, average read lengths of 150 base pairs (bp) were generated by sequencing.After removing the library index sequences from each read, the clean reads were mapped to the P. trichocarpa genome v4.0 (https://phytozome-next.jgi.doe.gov/,accessed on 14 January 2021) using HISAT2 [86].The raw counts were determined and normalized following the established analysis pipeline.DEGs were characterized by FDR < 0.05 and |LogFC| > 1 by using DESeq2 [87].

Motif Analysis
We used the website https://biasaway.uio.no(accessed on 24 January 2024) to generate a background set of sequences, and then used https://biasaway.uio.no/biasaway/g(accessed on accessed on 24 January 2024) to optimize the G/C content.A custom background option was selected through the STREME tool (https://meme-suite.org/meme/ tools/streme, accessed on 26 January 2024) [41].A genomic option was applied to search for de novo motifs.

Weighted Gene Co-Expression Network Analysis
Weighted Gene Co-Expression Network Analysis (WGCNA) was performed using the WGCNA R software package (v1.7.1) [88] inand was detected for FPKM with average expression values greater than 1.The p-values of the WGCNA results were identified using the multiple corrections tests (Benjamini-Hochberg FDR).In WGCNA analysis, the weighted value of correlation coefficient is used, that is, the gene correlation coefficient is taken to the NTH power, so that the connections between genes in the network follow the scale-free network distribution.First, the soft threshold was determined.The 'pickSoftThreshold' algorithm was used for soft threshold calculation [88].According to the correlation coefficient between genes, a hierarchical clustering tree was constructed to cluster genes with similar patterns into one module [88].Combined with the correlation between modules and samples, the correlation between modules and phenotypes was analyzed.The weighted network was visualized, phenotypic data were added, and correlations between modules and phenotypes were analyzed.

Statistical Analysis
Two-tailed Student's t-tests, ANOVA, and Duncan's new multiple range test (MRT) were carried out for statistical analysis to determine data significance.Significance levels were defined as * p < 0.05, ** p < 0.01, or different lowercase letters.

19 Figure 1 .
Figure 1.Transcript abundance of PtrLBD39 in different tissues of P. trichocarpa.Normalized FPKM indicates normalized transcript abundances expressed as Fragments Per Kilobase of exon model per Million mapped fragments.The differences analysis was conducted using the IBM SPSS Statistics v27.0 software with Duncan's new multiple range test (MRT); error bars indicate the SE (standard error) of three biological replicates from independent pools, and bars with different lowercase letters were found to be significantly different (p < 0.05).

Figure 1 .
Figure 1.Transcript abundance of PtrLBD39 in different tissues of P. trichocarpa.Normalized FPKM indicates normalized transcript abundances expressed as Fragments Per Kilobase of exon model per

19 Figure 2 .
Figure 2. Effects of overexpressing PtrLBD39 on P. trichocarpa growth.(A) Phenotypic observations of P. trichocarpa wild-type and PtrLBD39 overexpression plants (OE-PtrLBD39-L1, L2, L3) at 60, 90 and 120 days (Bars = 10 cm).(B-D) Growth indicators determination of wild-type and PtrLBD39 overexpression plants at different growth periods.Error bars represent SE values of three independent experiments, and asterisks indicate significant differences between the transgenics and wildtype plants found using Student's t-test (** p < 0.01).

Figure 3 .
Figure 3. Overexpression of PtrLBD39 affects the number, lignification of fiber cells and the number and lumen area of vessels.(A,B) Stem cross sections and magnified images of 4th, 6th, 8th, 10th, and 20th stem segments of 4-month-old OE-PtrLBD39-L1 and wild-type plants (The cross sections were stained with Safranin O and Fast Green.(A) Bars = 500 µm.(B) Bars = 200 µm.IN: internode).(C) Statistics analysis of the mean lumen area of individual fiber and vessel cell (mm 2 ) within the 20th internode.(D) Statistics analysis of the number of fiber and vessel cells per cross-sectional area (mm 2 ) within the 20th internode.Error bars represent SE values of three independent replicates with at least 150 vessel cells and 3000 fiber cells for each genotype in each replicate, and asterisks indicate

Figure 3 .
Figure 3. Overexpression of PtrLBD39 affects the number, lignification of fiber cells and the number and lumen area of vessels.(A,B) Stem cross sections and magnified images of 4th, 6th, 8th, 10th, and

19 Figure 5 .
Figure 5. RNA-seq analysis of the transcript abundances of cell wall component genes.(A) Monolignol genes.(B) Cellulose genes.(C) Hemicellulose genes.Normalized FPKM indicates normalized transcript abundances expressed as Fragments Per Kilobase of exon model per Million mapped fragments.Error bars indicate the SE of three biological replicates from independent pools.* p < 0.05, ** p < 0.01, ns, not significant.

Figure 5 .
Figure 5. RNA-seq analysis of the transcript abundances of cell wall component genes.(A) Monolignol genes.(B) Cellulose genes.(C) Hemicellulose genes.Normalized FPKM indicates normalized transcript abundances expressed as Fragments Per Kilobase of exon model per Million mapped fragments.Error bars indicate the SE of three biological replicates from independent pools.* p < 0.05, ** p < 0.01, ns, not significant.

Figure 6 .
Figure 6.Co-expression, the correlation between gene modules and traits.X-axis represents the 23 traits (1-3: plant height at different growth stages, 4-6: internode number at different growth stages, 7-9: basal diameter at different growth stages, 10: S-type lignin content, 11: G-lignin content, 12: Hlignin content, 13: S/G ratio, 14: glucose, 15: xylose, 16: acid insoluble lignin, 17: acid soluble lignin, 18: total lignin, 19: lumen area of individual vessel cell, 20: lumen area of vessel cell per area, 21: lumen area of fiber cell per area, 22: number of fiber cell per area, 23: number of vessel cell per area successively); Y-axis represents the co-expression modules (Genes with similar expression patterns are grouped into one module.One color represents one module of genes).The table is color-coded to represent correlation strength according to the color legend.The intensity and direction of each correlation are indicated on the right side of the heatmap (red, positively correlated; blue, negatively correlated).Numbers on the table represent the correlation coefficient between the corresponding module and the trait, with q-values listed below the correlations in parentheses.Modules within boxes with black lines: q < 0.05, |R| > 0.85.

Figure 6 .
Figure 6.Co-expression, the correlation between gene modules and traits.X-axis represents the 23 traits (1-3: plant height at different growth stages, 4-6: internode number at different growth stages, 7-9: basal diameter at different growth stages, 10: S-type lignin content, 11: G-lignin content, 12: H-lignin content, 13: S/G ratio, 14: glucose, 15: xylose, 16: acid insoluble lignin, 17: acid soluble lignin, 18: total lignin, 19: lumen area of individual vessel cell, 20: lumen area of vessel cell per area, 21: lumen area of fiber cell per area, 22: number of fiber cell per area, 23: number of vessel cell per area successively); Y-axis represents the co-expression modules (Genes with similar expression patterns are grouped into one module.One color represents one module of genes).The table is color-coded to represent correlation strength according to the color legend.The intensity and direction of each correlation are indicated on the right side of the heatmap (red, positively correlated; blue, negatively correlated).Numbers on the table represent the correlation coefficient between the corresponding module and the trait, with q-values listed below the correlations in parentheses.Modules within boxes with black lines: q < 0.05, |R| > 0.85.

Author
Contributions: J.Y. and C.Z. designed the experiments and supervised this study.B.G. wrote the manuscript and analyzed RNA-Seq data.D.L. generated and analyzed the transgenic plants, W.L., V.L.C. and S.L. reviewed and edited the manuscript.All authors have read and agreed to the published version of the manuscript.Funding: This work was supported by the National Natural Science Foundation of China (32001332), the Fundamental Research Funds for the Central Universities of China (2572022DQ01).

Table 1 .
Wood composition of OE-PtrLBD39 transgenic and wild-type P. trichocarpa.
Six-month-old plants were tested.Three biological replicates from independent pools of OE-PtrLBD39 and wild-type (WT) stems were carried out.Data are the means of three independent assays with SE values.Asterisks indicate significant differences between OE-PtrLBD39 transgenics and wild-type plants by Student's t-test and ANOVA (* p < 0.05, ** p < 0.01).Units are g per 100 g of dry extractive-free wood.
Six-month-old plants were tested.Three biological replicates from independent pools of OE-PtrLBD39 and wild-type (WT) stems were carried out.Data are the means of three independent assays with SE values.Asterisks indicate significant differences between OE-PtrLBD39 transgenics and wild-type plants by Student's t-test (* p < 0.05, ** p < 0.01).S, S subunits; G, G subunits; H, H subunits.Values indicate the percentage weight of total lignin.