Abstract
Stem secondary xylem produced by cambial division and differentiation is the main source of tree biomass. Secondary xylem formation involves a complex transcriptional regulatory network; however, the underlying mechanism is still being explored. Here, we report that PagHAM4a and PagHAM4b are positive regulators of cambial differentiation into secondary xylem in hybrid poplar (Populus alba × Populus glandulosa clone 84K). Overexpression of PagHAM4a and PagHAM4b enhanced cambial activity and increased the number of secondary xylem cells in the stems of poplar. By contrast, single or double mutations of PagHAM4a and PagHAM4b generated by CRISPR/Cas9 decreased cambial activity, leading to a significant reduction of secondary xylem. Neither overexpression nor mutation of the two genes affected the size of vessels and fibers in xylem. Both PagHAM4a- and PagHAM4b-regulated gene networks were mainly centered at the stage when cambium had just initiated secondary growth, but the molecular networks regulated by the two genes were distinct. Further analysis revealed that PagSCL21 and PagTCP20 are direct targets of PagHAM4a and PagHAM4b, respectively, and their overexpression also promoted cambial differentiation into secondary xylem. Taken together, we identified two novel key regulatory modules in poplar, PagHAM4a–PagSCL21 and PagHAM4b–PagTCP20, which provide new insights into the mechanism of secondary xylem formation in trees.
Introduction
In trees, the shoot tip which is derived from the stem cells in the shoot apical meristem (SAM) first elongates longitudinally, also known as primary growth, (Groover and Robischon, 2006). After primary growth, trees expand their girth by radially thickening, called secondary growth. Secondary growth is driven by vascular cambium, which continuously divides and differentiates to produce secondary phloem outwards and secondary xylem inwards (Spicer and Groover, 2010; Pierre-Jerome et al., 2018; Fischer et al., 2019). Tracheary elements (i.e. tracheids in gymnosperms and vessels in dicots) and fibers are the main cell types of secondary xylem, which enables water and nutrient transport as well as providing physical or biomechanical support necessary for tree growth (Chai et al., 2014). The secondary xylem in the stems, often referred to as wood, is the main source of tree biomass, and has important economic and ecological value (Ye and Zhong, 2015; Brackmann et al., 2018; Sugimoto et al., 2022). Therefore, research on the developmental regulation of secondary xylem has received increasing attention in recent years. The formation of secondary xylem involves a series of developmental events including the differentiation of cambium into secondary xylem, cell expansion, secondary cell wall formation, and programmed cell death (Wessels et al., 2019; Zhang et al., 2022), constituting a complex molecular regulatory network; however, the underlying regulatory mechanism is still being explored.
As a dynamic and continuous process, the formation of secondary xylem involves the orderly regulation of a large number of transcription factors (Demura and Fukuda, 2007; Nakano et al., 2015; Heo et al., 2017; Sun et al., 2022). Previous studies have revealed that the WUS-related HOMEOBOX (WOX), TEOSINTE BRANCHED1/CYCLOIDEA/PCF (TCP), AUXIN RESPONSE FACTOR (ARF), class III HOMEODOMAIN-LEUCINE ZIPPER (HD-ZIP III), NAM/ATAF/CUC (NAC), and MYB transcription factor family members play important roles in regulating the formation of secondary xylem (Ohtani et al., 2011; Kucukoglu et al., 2017; Xu et al., 2019; Hou et al., 2020; Xiao et al., 2021; Hu et al., 2022). WOX4, a member of the WOX family, is considered to be a key transcription factor in regulating the formation of secondary xylem in poplar, and the down-regulation of PttWOX4 expression can inhibit cambial cell proliferation and the differentiation of secondary xylem in hybrid poplar (Populus tremula × Populus tremuloides, clone T89) (Kucukoglu et al., 2017). In Populus trichocarpa, the expression of PtrWOX4 is regulated by its upstream HD-ZIP III transcription factor PtrHB4 (Zhu et al., 2018), while in Populus tomentosa, PtoWOX4 is found to be activated by an ARF family member PtoARF7 to promote the differentiation of cambium into secondary xylem (Hu et al., 2022). HD-ZIP III genes are also found to be regulated by the ARF family members in P. tomentosa, and PtoARF5 can directly induce the expression of HD-ZIP III genes PtoHB7 and PtoHB8, thereby promoting the formation of secondary xylem (Xu et al., 2019). Also, in P. tomentosa, PtoTCP20, a member of the TCP family, is found to interact with PtoWOX4, and also activates the transcription of PtoWND6, a homolog of Arabidopsis NAC gene family member VND7, to promote the cambial activity and its differentiation into secondary xylem (Hou et al., 2020). In Arabidopsis, VND7 and VND6 initiate xylem vessel differentiation (Kubo et al., 2005), while NST1 and NST3 of the NAC gene family induce xylem fiber differentiation (Zhong et al., 2006; Mitsuda et al., 2007); the direct targets of NST3 are MYB46 and MYB83, which can regulate the synthesis of secondary cell walls in xylem (Mitsuda et al., 2005; Zhong et al., 2007; McCarthy et al., 2009). In P. trichocarpa, PtrMYB3 and PtrMYB20, which are homologs of Arabidopsis MYB46, positively regulate the development of secondary xylem in stems (McCarthy et al., 2010; Zhong and Ye, 2012). Collectively, the synergistic regulation of different transcription factor family members constitutes a complex transcriptional regulatory network for secondary xylem formation in plants.
The GRAS family is a class of plant-specific transcription factors which consist of multiple subfamilies (Tian et al., 2004; Cenci and Rouard, 2017; Zhang et al., 2018). HAIRY MERISTEM (HAM) is a subfamily of the GRAS family, and the HAM proteins have typical structural characteristics of the GRAS family, with a conserved C-terminus and a highly variable N-terminus (Guo et al., 2017; Geng et al., 2021). In Arabidopsis, the HAM subfamily has four members, namely AtHAM1, AtHAM2, AtHAM3, and AtHAM4, which are found to be involved in regulating the proliferation and differentiation of stem cells in the SAM and the root apical meristem (RAM), thereby modulating the primary growth of stems and roots (Engstrom et al., 2011; Zhou et al., 2015; Han et al., 2020). It is worth noting that AtHAM4 is found to interact with AtWOX4 in vitro (Zhou et al., 2015), and the homolog of AtWOX4, PttWOX4, is involved in the regulation of stem cambial activity and secondary xylem formation in poplar (Kucukoglu et al., 2017). Therefore, we speculate that the homolog of AtHAM4 in poplar may be involved in regulating the formation of secondary xylem, but its regulatory mechanism and regulatory relationship with other transcription factor family members are unclear.
To address the above-mentioned questions, we identified two homologs of AtHAM4, PagHAM4a and PagHAM4b, as two novel key transcription factors that regulate secondary xylem formation in hybrid poplar (Populus alba × Populus glandulosa, clone 84K). Functional characterizations of PagHAM4a and PagHAM4b uncovered that the two transcription factors positively regulate cambial activity, thereby promoting the differentiation of cambium into secondary xylem in ‘84K’ poplar stems. Transcriptome analysis revealed that the PagHAM4a-regulated gene network is overwhelmingly different from that regulated by PagHAM4b. We further provide evidence that PagHAM4a and PagHAM4b participate in secondary xylem formation in different pathways by targeting the SCARECROW-LIKE (SCL) gene PagSCL21 and the TCP gene PagTCP20, respectively. Our results provide new insights into complex transcriptional regulatory networks for secondary xylem formation in trees.
Materials and methods
Plant materials and growth conditions
Populus alba × Populus glandulosa (clone 84K) was used for all experiments. Wild-type (WT) and transgenic seedlings were propagated via in vitro microcutting. The microcutting-propagated plantlets were cultured in glass bottles (6 cm diameter/13 cm height) containing 1/2 Murashige and Skoog (MS) medium under 16 h/8 h (light/dark) and 23–25 °C. For phenotype analyses, 2-month-old seedlings were transferred to pots (135 mm upper diameter/85 mm bottom diameter/128 mm height) containing humus soil, perlite, and vermiculite (3:2:1, v/v/v), and all were cultivated together in the same growth chamber under the same growth condition as described above. The cultivated plants were watered every 5 d and fertilized with 1/2 MS solution every 30 d. After 3 months of cultivation, the phenotypes of cultivated plants including plant height, internode number and length, stem diameter, and total area of fully expanded leaves were measured.
Gene cloning and plasmid construction
The full-length coding region of PagHAM4a, PagHAM4b, PagSCL21, and PagTCP20 was amplified from cDNA prepared from ‘84K’ poplar, and cloned individually into the vector pBI121 driven by the cauliflower mosaic virus (CaMV) 35S promoter. The promoter sequences of PagHAM4a (1823 bp upstream of ATG) and PagHAM4b (1904 bp upstream of ATG) were amplified from genomic DNA of ‘84K’ poplar. The CaMV 35S promoter was replaced by the PagHAM4a or PagHAM4b promoter to drive the β-glucuronidase (GUS) reporter gene in the pBI121 vector. The CRISPR/Cas9 [clustered regularly interspaced palindromic repeats (CRISPR)/CRISPR-associated protein 9] system was used to generate single- and double knockout mutants of PagHAM4a and PagHAM4b. Suitable targets within the PagHAM4a and PagHAM4b locus were identified using CRISPRscan (https://www.crisprscan.org/?page=sequence), and the three highest scoring sequences were applied (Supplementary Table S1). Each construct for CRISPR/Cas9-mediated mutagenesis was designed to contain expression cassettes with three different single guide RNAs (sgRNAs) that were placed under the control of the Arabidopsis U6-26 promoter and then inserted into the pCambia2301 vector for targeting PagHAM4a or PagHAM4b. The primer sequences are listed in Supplementary Table S2.
Phylogenetic analysis
The nucleotide and amino acid sequences of HAM transcription factors in Arabidopsis and P. trichocarpa were obtained by a BLAST search and aligned by DNAMAN. The amino acid sequences of HAM homologs from Arabidopsis and poplar were aligned with ClustalW to build a phylogenetic tree based on the Neighbor–Joining method in MEGA version 11. The parameters used were complete deletion and bootstrap (1000 replicates).
RNA extraction and real-time quantitative PCR assay
For the expression pattern of PagHAM4a and PagHAM4b, total RNA was extracted from various tissues of 2-month-old ‘84K’ poplar, using the easy-spin Plant RNA Kit (Aidlab Biotech, RN0902) following the manufacturer’s instructions. Tissue samples were collected from young stems (first to second internode), mature stems (eighth to ninth internode), young leaves (from the second to third internode), mature leaves (from the eighth to ninth internode), young roots (0.5–1 cm from the root tip), and mature roots (2–3 cm from the root tip). To detect gene expression in transgenic plants, samples were collected from the fourth to the eight internode of stems in each line. The extracted total RNAs were reverse transcribed into first-strand cDNAs using the FastKing RT Kit (TIANGEN, KR116). All qPCRs were performed using SuperReal PreMix Plus (SYBR Green) (TIANGEN, FP205) in a LineGene 9600 PCR machine (BIOER). The poplar UBQ (Potri.014G115100) and 18S (AY652861.1) genes were used as the internal reference genes. The relative expression levels of the genes were analyzed using the 2–ΔΔCt method (Pfaffl, 2001) and normalized using the mean of the two reference genes (UBQ and 18S). Three biological replicates and three technical replicates were performed for each sample. The primer sequences used for qPCRs are listed in Supplementary Table S2.
Plant transformation
‘84K’ poplar was transformed using an Agrobacterium-mediated leaf disk method (Song et al., 2019), and then selected on medium containing 30 mg l–1 kanamycin. After regeneration, positive transgenic plants overexpressing target genes were identified by genomic DNA PCR and real-time quantitative-PCR (RT-qPCR). To identify the knockout mutants, the genome sequences containing three sgRNA target sites were amplified from the genome DNA prepared from mutant plants, and sequenced. The primer sequences are listed in Supplementary Table S2.
Histochemical staining of β-glucuronidase
For GUS staining, cross-sections of the eighth internode of PagHAM4apro::GUS and PagHAM4bpro::GUS lines as well as leaves and roots were soaked in 90% acetone solution for 30 min at –20 °C, followed by washing three times in double-distilled H2O. Then the samples were placed in GUS staining buffer containing 0.5 mg ml–1 5-bromo-4-chloro-3-indolyl glucuronide for 30 min in a vacuum before incubation overnight at 37 °C. After staining, the samples were sequentially decolorized with 75, 85, and 100% ethanol for microscopic observation.
RNA sequencing analysis
Total RNA was extracted from the stems that had just initiated secondary growth (third to fourth internode) and those with significant secondary growth (eight to ninth internode) from 2-month-old WT, PagHAM4a-overexpressing (OE), and PagHAM4b-OE lines using the easy-spin Plant RNA Kit (Aidlab Biotech, RN0902). For each stage, three biological replicates were prepared. The transcriptome library was prepared from the high-quality total RNA per sample using a NEB Next Ultra™ RNA Library Prep Kit for Illumina (NEB, USA). All transcriptome libraries were sequenced on an Illumina HiSeq 2500 platform (Biomarker technologies, Beijing, China). Raw reads containing adapter, poly-N, and low-quality reads were removed, and then mapped with the P. trichocarpa genome (Pop_tri_v3). The gene expression level was calculated using the FPKM (fragments per kilobase of script per million fragments mapped) method (Trapnell et al., 2010), and the gene expression pattern was examined using the DESeq2 method (Love et al., 2014). Differentially expressed genes (DEGs) were identified in pairwise comparison of samples based on criteria of false discovery rate (FDR) <0.01 and |log2 fold change| ≥1. Functional enrichment analysis of the DEGs was carried out by KOBAS software (Xie et al., 2011).
Yeast one-hybrid assay
The coding sequences (CDSS) of PagHAM4a and PagHAM4b were each inserted into the pGADT7 vector. Two ~2.0 kb upstream sequences of PagSCL21 and PagTCP20 were amplified from the genomic DNA of ‘84K’ poplar by PCR and ligated into the pAbAi vector. At the same time, the PagSCL21 and PagTCP20 promoter sequences with binding site mutations were ligated into the pAbAi vector. Then, the pGADT7-PagHAM4a vector and pGADT7 empty vector were individually transferred into the Y1HGold yeast strain containing pAbAi-PagSCL21pro or pAbAi-PagSCL21pro-mt vectors. Similarly, pGADT7-PagHAM4b vector and pGADT7 empty vector were individually transferred into the Y1HGold yeast strain containing pAbAi-PagTCP20pro or pAbAi-PagTCP20pro-mt vectors. Positive yeast cells were used to evaluate the interaction between pGADT7-PagHAM4a and pAbAi-PagSCL21pro and between pGADT7-PagHAM4b and pAbAi-PagTCP20pro on SD medium to which Aureobasidin A (AbA) was added, but lacking leucine and uracil (SD-Leu/Ura). The primer sequences are listed in Supplementary Table S2.
Dual-luciferase reporter assay
The CDSs of PagHAM4a and PagHAM4b were each inserted into the pGreenII 62-SK vector and used as effectorS. The promoter sequences of PagSCL21 and PagTCP20 were each fused with the firefly luciferase (LUC) reporter in the pGreenII 0800-LUC vector and used as a reporter. All constructs were transformed into Agrobacterium GV3101 (Weidi Biotechnology), and then injected into Nicotiana benthamiana leaves. The infected N. benthamiana was cultured for 24 h in the dark and then transferred in a growth chamber under normal conditions for a further 48 h. The LUC and Renilla luciferase (REN) activities were measured on a GloMax 20/20 Luminometer (Promega, USA) using the dual-luciferase reporter assay system (Promega, USA). The LUC fluorescence signal was detected using a chemiluminescence imaging system (Gelview 6000Pro II, BLT, China). Activation of the promoter by transcription factors was assessed by the ratio of the LUC activity driven by the PagSCL21 and PagTCP20 promoter to that of REN activity driven by the CaMV 35S promoter. The primer sequences used for the constructions are listed in Supplementary Table S2.
Chromatin immunoprecipitation coupled with quantitative PCR analysis
The CDSs of PagHAM4a and PagHAM4b were each inserted into pBI121 with a 3×Flag tag, to generate the pBI121-PagHAM4a-3×Flag and pBI121-PagHAM4b-3×Flag constructs, respectively. The two constructs were transformed into 84K poplar, and transgenic plants overexpressing PagHAM4a-3×Flag or PagHAM4b-3×Flag were obtained. Four-week-old transgenic poplars and WT (control) were used for ChIP-qPCR assay using an EpiQuik™ Plant ChIP Kit (Epigentek). The plants were submerged in 1% formaldehyde to cross-link genomic DNA and protein under a vacuum, and then the chromatin complex was sonicated to fragment genomic DNA. The DNA–protein mixture was incubated with anti-Flag antibodies and then used for immunoprecipitation. The immunoprecipitated genomic DNA fragments were used for qPCR. The primer sequences are listed in Supplementary Table S2.
Anatomical and histological analysis
The basal stems of 3-month-old WT and transgenic poplars were collected and cut into segments of ~0.5 cm in length, and fixed in FAA solution comprising 5 ml of 4% paraformaldehyde, 90 ml of 70% ethanol, and 5 ml of acetic acid at 4 °C for 24 h. Subsequently, the samples were dehydrated with a series of ethanol concentrations and then the alcohol was replaced with xylene, and the samples embedded in paraffin. The embedded samples were cut into 10 μm sections using a Leica microtome (RM2235, Leica, Germany). The sections were deparaffinized with xylene solution, followed by dehydration with graded concentrations of ethanol, stained with 0.1% (w/v) toluidine blue O (Sigma) for 2 h, and then observed under an optical microscope (Leica, Germany). Cell wall thickness was measured under a transmission electron microscope. The basal stems of 3-month-old WT and transgenic poplars were sectioned, and fixed with 2.5% (w/v) glutaraldehyde and 1% (v/v) osmic acid. After washing with 0.1 M phosphate buffer, samples were dehydrated with a graded ethanol series, and then ethanol was replaced with acetone. The samples were embedded in Spurr’s resin. Sections of 70 nm thickness were cut with an ultramicrotome (Leica, EM UC7, Germany) and observed with a transmission electron microscope (HITACHI HT7700, Japan). The number and size of xylem cells, xylem width, and secondary wall thickness were measured using ImageJ software.
Statistical analysis
The values shown in the figures are expressed as means ±SD. Statistical analysis was performed using SPSS version 26 software by a one-way ANOVA to determine significance, which was defined as P<0.05.
Results
Identification of PagHAM4a/b and their expression patterns in ‘84K’ poplar
Using the Arabidopsis AtHAM4 sequence as a query, two genes (Potri.005G125800 and Potri.007G029200) were identified in the P. trichocarpa genome as putative orthologs of AtHAM4 by BLAST in the Phytozome database (https://phytozome.jgi.doe.gov/pz/portal.html). The two PtrHAM4 orthologs in ‘84K’ poplar were obtained by homology-based cloning, and denoted as PagHAM4a and PagHAM4b. The CDS of PagHAM4a is 1650 bp encoding a deduced protein of 549 amino acids, and the PagHAM4b CDS is 1641 bp encoding a deduced protein of 546 amino acids. The deduced PagHAM4a and PagHAM4b proteins both contain a conservative GRAS domain of the GRAS family that is composed of the leucine heptad repeat I (LHR I), VHIID, LHR II, PFYRE, and SAW motifs (Supplementary Fig. S1A). The two deduced proteins shared 84.88% amino acid similarity (Supplementary Fig. S1B), indicating that they belong to a gene pair arising from a duplication in the Populus lineage. Further analysis showed that the N-terminal regions of PagHAM4a and PagHAM4b proteins had an unknown motif of 40 and 18 acids in length, respectively (Supplementary Fig. S1C). We compared the amino acid sequences of PagHAM4a, PagHAM4b, and Arabidopsis HAM1–HAM4 proteins, and found that all HAM proteins contain a highly variable N-terminal domain and a conserved C-terminal domain (Supplementary Fig. S2). In addition, we conducted a phylogenetic analysis of PagHAM4a, PagHAM4b, and their homologs HAM1–HAM4 from the model plant Arabidopsis and P. trichocarpa using the Neighbor–Joining method. The HAM1, HAM2, and HAM3 homologs from Arabidopsis and P. trichocarpa were clustered in group I, while PagHAM4a and PagHAM4b were clustered in group II with the HAM4 homologs from Arabidopsis and P. trichocarpa (Fig. 1A). In group II, PagHAM4a was clustered together with PtrHAM4a, and PagHAM4b was clustered together with PtrHAM4b (Fig. 1A).
The HAM family in Arabidopsis and poplar, and expression patterns of PagHAM4a/b in ‘84K’ poplar. (A) A phylogenetic analysis of the HAM family from Arabidopsis (At), Populus trichocarpa (Ptr), and Populus alba × Populus glandulosa ‘84K’ (Pag). (B) Real-time quantitative PCR (RT-qPCR) analysis of the expression patterns of PagHAM4a and PagHAM4b genes in different tissues of ‘84K’ poplar. YS, young stems; MS, mature stems; YL, young leaves; ML, mature leaves; YR, young roots; MR, mature roots. Values are means ±SD of three biological and technical replicates. (C) Histological staining of the GUS reporter driven by the promoters of PagHAM4a and PagHAM4b in poplar stems. The eighth internodes of ‘84K’ poplars were cross-sectioned for GUS staining. Bars, 200 µm (left), 100 µm (right). (D) Histological staining of the GUS reporter driven by the promoters of PagHAM4a and PagHAM4b in poplar leaves and roots. Bars, 1 cm.
To study the expression patterns of the PagHAM4a and PagHAM4b, we used RT-qPCR to analyze the relative transcript levels of the two genes in different tissues of ‘84K’ poplar. PagHAM4a and PagHAM4b transcripts were detected in all tissues examined, with the highest abundance in stems, followed by that in leaves, and the lowest abundance in roots (Fig. 1B; Supplementary Fig. S3B). It was also found that the transcript levels of PagHAM4a and PagHAM4b in mature stems were much higher than those in young stems (Fig. 1B; Supplementary Fig. S3B). To further examine the specific spatial expression patterns of PagHAM4a and PagHAM4b, the construct containing the promoter of PagHAM4a or PagHAM4b upstream of the GUS reporter gene was transformed into the ‘84K’ poplar. GUS activity was detected in stems, leaves, and roots in the transgenic poplars (Fig. 1C, D), which was consistent with the result from RT-qPCR (Fig. 1B). Notably, the GUS activity is highest in the cambium and adjacent vascular cells of the secondary stems (Fig. 1C).
Overexpression and CRISPR/Cas9 knockout of PagHAM4a and PagHAM4b in ‘84K’ poplar
To investigate the role of PagHAM4a and PagHAM4b genes in the formation of secondary vascular tissues, we overexpressed PagHAM4a and PagHAM4b in ‘84K’ poplar (PagHAM4a-OE and PagHAM4b-OE) under the control of the CaMV 35S promoter. For the phenotypic analysis, representative PagHAM4a-OE and PagHAM4b-OE lines that displayed significantly increased growth compared with the WT were selected and planted in a glasshouse for 3 months (Fig. 2A). The plant height and stem diameter of PagHAM4a-OE transgenic plants were increased by 55% and 30%, and those of PagHAM4b-OE transgenic plants were increased by 40% and 20% (Fig. 2E, F). The number and length of internodes as well as leaf size were also increased to varying degrees in both PagHAM4a-OE and PagHAM4b-OE transgenic plants (Supplementary Fig. S4C, D, H). The phenotypic effects were positively correlated with the expression levels of PagHAM4a and PagHAM4b in transgenic plants; that is, the higher the expression level of these two genes, the more obvious the positive phenotypic effect (Fig. 2; Supplementary Fig. S4A, B); moreover, the positive phenotypic effects were more significant in PagHAM4a-OE lines than in PagHAM4b-OE lines (Fig. 2; Supplementary Fig. S4). We also found that overexpression of PagHAM4b did not affect PagHAM4a expression (Supplementary Fig. S5A), and overexpression of PagHAM4a did not affect PagHAM4b expression (Supplementary Fig. S5B).
Phenotypes resulting from PagHAM4a- and PagHAM4b-overexpressing poplars. (A) Phenotypes of 3-month-old wild-type (WT), and PagHAM4a- and PagHAM4b-overexpressing (PagHAM4a-OE and PagHAM4b-OE) plants. Bar, 10 cm. (B) Cross-sections of the basal stems of 3-month-old WT, PagHAM4a-OE (#9), and PagHAM4b-OE (#4) plants. Xy, xylem. Bars, 100 µm. (C) Detailed observation of the cambial zone in WT, PagHAM4a-OE, and PagHAM4b-OE plants. Ph, phloem, Xy, xylem, Ca, cambium. Bars, 100 µm. (D) TEM of xylem vessels and fibers in the stems of 3-month-old WT, PagHAM4a-OE (#9), and PagHAM4b-OE (#4) plants. (E–I) Measurements of plant height, stem diameter, xylem width, secondary xylem cell layers, and cambial cell layers of WT, PagHAM4a-OE, and PagHAM4b-OE poplar lines. (J, K) Measurements of secondary wall thickness of vessels and fibers in WT, PagHAM4a-OE, and PagHAM4b-OE poplar lines. Data represent means ±SD from at least five plants for each line, and different lowercase letters indicate significant differences at P<0.05 by ANOVA with Tukey’s test.
Secondary xylem development was promoted in transgenic plants (Fig. 2B, G). The width of secondary xylem at the base of stems was increased by 65% and 49% in PagHAM4a-OE and PagHAM4b-OE transgenic plants, respectively, compared with the WT (Fig. 2G); however, there was no significant change in the width of secondary phloem (Supplementary Fig. S4G). To investigate whether the increased secondary xylem width in transgenic plants was caused by an increase in cell number or cell expansion, we counted and measured the number and size of secondary xylem cells in transgenic and WT plants. Compared with the WT, both PagHAM4a-OE and PagHAM4b-OE transgenic plants produced more secondary xylem cells (Fig. 2H), but did not display any significant difference in the secondary wall thickness and the cell size of vessels and fibers (Fig. 2D, J, K; Supplementary Fig. S4E, F). For this reason, we further examined the cambium zone of transgenic and WT plants, and found that the width of the cambium zone was increased, with more cell layers in transgenic plants compared with the WT (Fig. 2C, I). These results indicate that the enhanced secondary xylem development in transgenic plants is due to the increased cambial activity, which promotes the differentiation towards the secondary xylem. The above findings are consistent with the results of our previous research on PagHAM4a-OE and PagHAM4b-OE transgenic plants cultured in medium (Supplementary Fig. S6).
To further investigate the functions of PagHAM4a and PagHAM4b and their functional redundancy, we generated pagham4a and pagham4b single mutants as well as a pagham4a;b double mutant using CRISPR/Cas9 technology. DNA sequencing showed that there were mainly base deletions, additions, and occasionally substitutions at the targets (Supplementary Fig. S7), leading to the loss of function of PagHAM4a, PagHAM4b, or both of them simultaneously. All mutants displayed a significant reduction in plant height, stem diameter, internode number, and length, as well as leaf size compared with the WT (Fig. 3A, B, E, F; Supplementary Fig. S8B, C, G). The plant height and stem diameter of the pagham4a mutant were decreased by 47.60% and 26.51%, respectively (Fig. 3E, F), and those of the pagham4b mutant were decreased by 34.73% and 21.95%, respectively (Fig. 3E, F), which indicates that the pagham4a mutant may suffer more severe growth inhibition than the pagham4b mutant. Compared with the WT, the plant height and stem diameter of the pagham4a;b double mutant were decreased by 62.85% and 35.57%, respectively (Fig. 3E, F), suffering more severe growth inhibition than pagham4a and pagham4b single mutants (Fig. 3E, F). Further anatomical observation showed that compared with the WT, the width of secondary xylem in pagham4a, pagham4b, and pagham4a;b was decreased by 31.84, 26.95, and 41.67%, respectively, with the decreased cell layers of xylem (Fig. 3G; Supplementary Fig. S8A), but the width of secondary phloem did not significantly change (Supplementary Fig. S8F); fewer cell layers were also observed in the cambium zone of mutants, giving rise to reduced width of the cambium (Fig. 3C, H). However, there were no significant differences in the secondary wall thickness and cell size of vessels and fibers in the secondary xylem of mutants (Fig. 3D, I, J; Supplementary Fig. S8D, E).
Phenotypes of single and double mutants of PagHAM4a and PagHAM4b in ‘84K’ poplar. (A) Dwarf phenotypes of 3-month-old PagHAM4a and PagHAM4b knockout mutants (pagham4a, pagham4b, and pagham4a;b). Bar, 10 cm. (B) Cross-sections of the basal stems of 3-month-old WT and pagham4a (#9), pagham4b (#8), and pagham4a;b (#4), Xy, xylem. Bars, 100 µm. (C) Detailed observation of the cambial zone of the basal stems in WT, pagham4a, pagham4b, and pagham4a;b. Ph, phloem, Xy, xylem, Ca, cambium. Bars, 50 µm. (D) TEM of xylem vessels and fibers in the stems of 3-month-old WT and pagham4a (#9), pagham4b (#8), and pagham4a;b (#4) mutant lines. (E–H) Measurements of plant height, stem diameter, xylem width, and cambial cell layers of pagham4a, pagham4b, and pagham4a;b mutant lines. (I, J) Measurements of secondary wall thickness of vessels and fibers in WT, pagham4a (#9), pagham4b (#8), and pagham4a;b (#4) mutant lines. Data represent means ±SD from at least five plants for each line, and different lowercase letters indicate significant differences at P<0.05 by ANOVA with Tukey’s test.
PagHAM4a and PagHAM4b activate PagSCL21 and PagTCP20 expression
To identify the target genes of PagHAM4a and PagHAM4b, we analyzed the transcriptional profiles of stem samples from transgenic plants that had just initiated secondary growth (stage 1) and those with significant secondary growth (stage 2), using WT plants as control. A total of 3819 and 3120 DEGs were identified in the stem samples of PagHAM4a-OE and PagHAM4b-OE transgenic plants, respectively (Supplementary Table S3). Among them, 3684 (2204 up-regulated/1480 down-regulated) and 135 (43 up-regulated/92 down-regulated) DEGs were identified in developmental stage 1 and stage 2 of PagHAM4a-OE transgenic plants, respectively. A total of 1696 (1023 up-regulated/673 down-regulated) and 1424 (571 up-regulated/853 down-regulated) DEGs were identified in developmental stage 1 and stage 2 of PagHAM4b-OE transgenic plants, respectively (Supplementary Table S3). Our analyses revealed that these DEGs are mainly present in developmental stage 1 of PagHAM4a-OE and PagHAM4b-OE transgenic plants—at this stage the cambium has just initiated secondary growth—and among them there were more up-regulated genes than down-regulated genes. We also found that 4908 out of a total of 5380 DEGs were distinct in stage 1 of PagHAM4a-OE and PagHAM4b-OE transgenic plants, and 1532 out of a total of 1559 DEGs were distinct in stage 2 of transgenic plants (Supplementary Fig. S9). These findings indicate that PagHAM4a and PagHAM4b may regulate secondary growth via different pathways in ‘84K’ poplar stems.
Based on the above analysis, we focused on the up-regulated transcription factors present in stage 1 of PagHAM4a-OE and PagHAM4b-OE transgenic plants, and investigated the transcriptional regulatory networks mediated by PagHAM4a and PagHAM4b. The JASPAR online program (https://jaspar.genereg.net/) was used to predict the binding sites with PagHAM4a or PagHAM4b in the promoter sequences of up-regulated transcription factors present in stage 1. Based on the scoring of the binding sites, the top 10 transcription factors that may be targeted by PagHAM4a and PagHAM4b were selected (Supplementary Tables S4, S5). Among the top 10 candidate target genes, PagSCL21 and PagTCP20 were predicted to be the most likely target genes of PagHAM4a and PagHAM4b, respectively, based on comprehensive considerations including scoring of binding sites, abundance, and fold up-regulation of gene expression, and previous functional studies of candidate target genes or their family members. We found that the expression levels of PagSCL21 and PagTCP20 were significantly up-regulated in stage 1 of PagHAM4a-OE and PagHAM4b-OE transgenic plants compared with the WT, whereas no significant change was observed in stage 2 of the two transgenic plants (Supplementary Table S6).
To verify the above prediction, we conducted a yeast one-hybrid (Y1H) assay using the CDS of PagHAM4a and PagHAM4b as bait, and the upstream promoter sequences of PagSCL21 and PagTCP20 as prey, respectively. All of the yeast transformants grew well on synthetic dropout (SD) (-Leu/Ura) medium lacking AbA; however, after the addition of 400 ng ml–1 AbA, only the yeast co-transformed with pAbAi-PagSCL21pro and pGADT7-PagHAM4a, and those co-transformed with pAbAi-PagTCP20pro and pGADT7-PagHAM4b grew normally (Fig. 4A, B). These results indicated that PagHAM4a could bind to the PagSCL21 promoter and PagHAM4b to the PagTCP20 promoter. A dual-luciferase reporter assay was used to further verify whether PagHAM4a and PagHAM4b could activate the expression of PagSCL21 and PagTCP20, respectively. The PagSCL21 and PagTCP20 promoters were each fused to a LUC reporter and co-transformed into N. benthamiana leaves with either 35S::PagHAM4a, 35S::PagHAM4b, or empty vector. Compared with the control, the LUC activity was significantly higher in N. benthamiana leaves co-transformed with 35S::PagHAM4a and PagSCL21 pro::LUC, and in those co-transformed with 35S::PagHAM4b and PagTCP20 pro::LUC (Fig. 4C, D). These findings were confirmed by a LUC complementation imaging assay (Fig. 4E, F). Further ChIP-qPCR revealed that the promoter fragments of PagSCL21 and PagTCP20 in the ChIP were significantly enriched in PagHAM4a-3×Flag and PagHAM4b-3×Flag plants, respectively, compared with the WT (Fig. 4G, H). These results confirmed that PagHAM4a and PagHAM4b can bind to the PagSCL21 and PagTCP20 promoters in vivo, respectively. In addition, RT-qPCR analysis showed that in the PagHAM4a-OE transgenic plants, the expression of PagSCL21 was significantly up-regulated but the expression of PagTCP20 remained unchanged (Supplementary Fig. S5C, D), and in the PagHAM4b-OE transgenic plants, PagTCP20 expression was significantly up-regulated but PagSCL21 expression remained unchanged (Supplementary Fig. S5C, D). Moreover, the mutation of PagHAM4a significantly decreased the expression of PagSCL21 but did not change the expression of PagTCP20 (Supplementary Fig. S5E, F), and the mutation of PagHAM4b significantly decreased the expression of PagTCP20 but did not change the expression of PagSCL21 (Supplementary Fig. S5E, F). These results confirmed that PagHAM4a and PagHAM4b specifically activate the expression of PagSCL21 and PagTCP20, respectively.
PagHAM4a and PagHAM4b activate PagSCL21 and PagTCP20 in vitro and in vivo, respectively. (A, B) Activation of PagSCL21 and PagTCP20 by PagHAM4a and PagHAM4b, respectively, was detected in vitro by yeast one-hybrid assays. The pGADT7 and pAbAi empty vectors were used as negative controls, and pAbAi-p53 and pGADT7-rec53 were used as positive controls. (C–F) Activation of PagSCL21 and PagTCP20 by PagHAM4a and PagHAM4b, respectively, in vivo was detected by dual-luciferase reporter gene assays. Data represent means ±SD from six independent biological replicates. Different lowercase letters indicate significant differences at P<0.05 by ANOVA. (G, H) ChIP-qPCR to detect the binding of PagHAM4a to the PagSCL21 promoter and PagHAM4b to the PagTCP20 promoter in vivo. Values are the means ±SD of three biological replicates, and different lowercase letters indicate a significant difference with respect to WT plants, P<0.05 by ANOVA with Tukey’s test.
Overexpression of PagSCL21 and PagTCP20 in ‘84K’ poplar
To investigate the function of the PagSCL21 and PagTCP20, we cloned the two genes from the cDNA library of ‘84K’ poplar stems. The CDSs of PagSCL21 and PagTCP20 are 1752 bp and 909 bp in length, encoding 583 and 302 amino acids, respectively. The PagSCL21 and PagTCP20 proteins contain conservative domains of GRAS and TCP family, respectively (Supplementary Fig. S10A, B). We overexpressed PagSCL21 and PagTCP20 in ‘84K’ poplar (PagSCL21-OE and PagTCP20-OE), and observed the anatomical structure of the 12th internode collected from transgenic plants. Compared with the WT, the stem diameter and secondary xylem width significantly increased in PagSCL21-OE and PagTCP20-OE transgenic plants, with more cell layers of secondary xylem and cambium (Fig. 5A, D–F), which is similar to the phenotype of PagHAM4a-OE and PagHAM4b-OE transgenic plants. We further found that compared with the WT, there was no significant difference in the secondary wall thickness and cell size of vessels and fibers in the PagSCL21-OE transgenic plants (Fig. 5H, I; Supplementary Fig. S11C, D). However, in the PagTCP20-OE transgenic plants, the size of fibers decreased but the secondary cell wall width of fibers increased, while the vessels did not show any significant difference (Fig. 5H, I; Supplementary Fig. S11C, D). In addition, we detected a significant increase in the expression level of PagWND6 in the PagTCP20-OE transgenic plants (Supplementary Fig. S4H).
Phenotypes of PagSCL21- and PagTCP20-overexpressing poplars. (A) Cross-sections of the 12th internodes of WT, PagSCL21-OE (#11), and PagTCP20-OE (#9) plants. Bars, 20 µm. (B) Detailed observation of the cambial zone in WT, PagSCL21-OE (#11), and PagTCP20-OE (#9) plants. Bars, 20 µm. (C) TEM of xylem vessels and fibers in the stems of 3-month-old WT, PagSCL21-OE (#11), and PagTCP20-OE (#9) plants. (D–G) Measurements of stem diameter, xylem width, and number of xylem and cambial cell layers in WT, PagSCL21-OE, and PagTCP20-OE poplar lines. (H, I) Measurements of secondary wall thickness of vessels and fibers in WT, PagSCL21-OE (#11), and PagTCP20-OE (#9) plants. Data represent means ±SD from at least five independent plants for each line, and different lowercase letters indicate significant differences at P<0.05 by ANOVA with Tukey’s test.
Discussion
Functional conservation and differentiation of HAM family members
The HAM subfamily has four members, AtHAM1–AtHAM4, in Arabidopsis, but has expanded to eight members in poplar (Fig. 1A), which indicates that the HAM genes have differentiated functionally in herbaceous plants and perennial woody plants. A previous study pointed out that the N-terminal regions of HAM proteins can affect their functions (Geng and Zhou, 2021). In this study, we found that there were differences in the length and sequence of the N-terminal regions between HAM1–HAM3 and HAM4 proteins (Supplementary Fig. S2), suggesting that the HAM1–HAM3 homologs have similar biological functions whereas the HAM4 homologs may have new functions that differ from those of the HAM1–HAM3 homologs, which is also supported by the phylogenetic analysis (Fig. 1A). In Arabidopsis, AtHAM1–AtHAM4 are required for maintaining the activity of the SAM (Engstrom et al., 2011; Zhou et al., 2015; Han et al., 2020). Single and double mutants of AtHAM genes do not show any abnormality in shoot tip development (Engstrom et al., 2011), indicating that the HAM homologs have functional redundancy in maintaining SAM activity. Although the simultaneous mutation of AtHAM1–AtHAM3 can lead to dwarfing, abnormal leaf order, and altered morphology of rosette and inflorescence meristems, it does not affect the development of vascular bundles in various organs (Engstrom et al., 2011). However, simultaneous mutation of AtHAM1–AtHAM4 can not only severely inhibit the growth of seedlings, but can also lead to reduction in the number of xylem vessels and fibers in hypocotyl (Zhou et al., 2015). These pieces of evidence suggest that the HAM4 homologs, which are more distant from the HAM1–HAM3 homologs in terms of genetic relationship, are functionally differentiated and may play an important role in regulating the development of primary vascular tissues, especially xylem.
PagHAM4a and PagHAM4b positively regulate cambial differentiation into secondary xylem
In this study, we found that the overexpression of both PagHAM4a and PagHAM4b led to an increase in plant height and stem diameter (Fig. 2E, F), while single or double mutations of the two genes exhibited the opposite phenotype (Fig. 3E, F), which indicates that PagHAM4a and PagHAM4b not only regulate the longitudinal elongation of ‘84K’ poplar stems, but also their radial thickening. Considering that tree biomass is mainly derived from radial thickening of stems, we further focused on the regulatory role of PagHAM4a and PagHAM4b in the radial thickening of ‘84K’ poplar stems. We revealed that overexpression of PagHAM4a and PagHAM4b resulted in an increase in the number of secondary xylem cells and cambial cell layers in ‘84K’ poplar stems (Fig. 2H, I), whereas the single and double mutations of the two genes showed the opposite phenotype (Fig. 3H; Supplementary Fig. S8A), and neither overexpression nor mutation of PagHAM4a and PagHAM4b in ‘84K’ poplars affected the size of the two main xylem cell types, vessels and fibers (Supplementary Figs S4E, F, S8D, E). These findings provide strong evidence that PagHAM4a and PagHAM4b enhance secondary xylem development by positively regulating cambial cell proliferation and its differentiation into secondary xylem cells, rather than by regulating the expansion of secondary xylem cells.
In view of the fact that the mutation of PagHAM4a had a more severe inhibitory effect on cambial activity and secondary xylem formation in ‘84K’ poplar stems than the mutation of PagHAM4b (Fig. 3G, H), and that the overexpression of PagHAM4a produces a greater positive effect than that of PagHAM4b (Fig. 2H, I), this suggests that the PagHAM4a may play a more important role than PagHAM4b in regulating cambial activity and its differentiation into secondary xylem. A previous study revealed that the N-terminal domains of GRAS proteins contain many molecular recognition features, which can identify their interaction partners and may participate in molecular recognition during plant development (Sun et al., 2011), indicating that the N-terminal sequences of GRAS proteins are of great importance for their function. For this reason, we analyzed the N-terminal sequences of PagHAM4a and PagHAM4b proteins with the conservative GRAS domain, and found that the PagHAM4a and PagHAM4b proteins have different unknown motifs at the N-terminal region (Supplementary Fig. S1B, C). Although there is no report on the two motifs so far, it could not be ruled out that the difference between the N-terminal sequence of PagHAM4a and PagHAM4b proteins may be one of the reasons why PagHAM4a may play a more important role than PagHAM4b in regulating cambial activity and its differentiation into secondary xylem in ‘84K’ poplar stems.
PagHAM4a/b and their target-mediated genetic regulatory network of secondary xylem formation
Overexpression of the PagHAM4a and PagHAM4b caused differential expression of a large number of downstream genes, and these DEGs were mainly centered on the stage when cambium had just initiated secondary growth (Supplementary Table S3). These results indicate that the molecular networks regulated by PagHAM4a and PagHAM4b play an important role mainly at the initial stage of secondary growth of ‘84K’ poplar stems. We further found that the DEGs in PagHAM4a-OE transgenic plants were distinctly different from those in PagHAM4b-OE transgenic plants (Supplementary Fig. S9), indicating that the PagHAM4a and PagHAM4b genes are likely to regulate secondary xylem formation by modulating different downstream target genes. Based on transcriptomic analysis, PagSCL21 and PagTCP20 were identified as direct targets of PagHAM4a and PagHAM4b, respectively, by target gene prediction, Y1H assay, dual-luciferase assays, and ChIP-qPCR (Fig. 4; Supplementary Tables S4, S5). These results reveal that PagHAM4a and PagHAM4b regulate the cambial activity and its differentiation into secondary xylem via different pathways in ‘84K’ poplar stems by targeting PagSCL21 and PagTCP20, respectively.
Like PagHAM4b, PagHAM4b-targeted PagTCP20 also regulated cambial activity and its differentiation into secondary xylem (Fig. 5). However, we also found that PagTCP20 modulated the secondary wall formation of secondary xylem fibers (Fig. 5I). In P. tomentosa, PtoTCP20 activates the transcription of PtoWND6 (Hou et al., 2020), and Arabidopsis VND7, a homolog of PtoWND6, is found to regulate secondary cell wall formation in xylem (Sun et al., 2017). We thus speculate that PagTCP20 may activate the expression of PagWND6, thereby promoting the secondary cell wall formation of fibers. We further confirmed that PagTCP20 overexpression indeed significantly up-regulated the expression level of PagWND6 (Supplementary Fig. S5H). PtoTCP20 is also found to regulate cambial activity through interaction with PtoWOX4 in P. tomentosa (Hou et al., 2020). The regulatory effect of WOX4 homologs on secondary xylem formation has been widely discovered in Populus species. Recently, PagWOX4 knockdown generated by RNAi was shown to inhibit the cambial activity and then reduce the production of secondary xylem in ‘84K’ poplar stems (Tang et al., 2022). PtrWOX4 has also been confirmed as a positive regulator of cambial activity in P. trichocarpa (Dai et al., 2023). It is interesting that the function of poplar WOX4 homologs described above is similar to that of PagHAM4b and PagTCP20 in regulating cambial activity and promoting secondary xylem development in this study (Figs 2, 5), suggesting that PagHAM4b and PagTCP20 may be associated with PagWOX4 during secondary xylem formation in ‘84K’ poplar stems. In addition, PtrWOX4 and PtoWOX4 are found to be activated by the upstream transcription factors PtrHB4 and PtoARF7 in P. tomentosa and P. trichocarpa, respectively (Zhu et al., 2018; Hu et al., 2022); however, it remains to be verified whether a similar regulatory mechanism exists upstream of the PagWOX4 gene in ‘84K’ poplar.
SCL21 protein is a member of the PAT1 (PHYTOCHROME A SIGNAL TRANSACTION 1) subfamily of the GRAS transcription factor family (Bolle et al., 2000; Bolle, 2004). In Arabidopsis, the down-regulation of the AtSCL21 expression level promotes the hypocotyl elongation of seedlings, indicating that AtSCL21 is a negative regulator of hypocotyl elongation (Torres-Galea et al., 2013). A subsequent study reveals that overexpression of AtSCL21 can also inhibit proliferation of root meristem cells and root elongation in Arabidopsis (Bisht et al., 2023). These findings indicate that AtSCL21 is involved in seedling growth and development as a negative regulator in Arabidopsis. However, in this study, we found that PagSCL21 positively regulated the cambial activity and its differentiation into secondary xylem (Fig. 5), showing a completely different function from its homolog Arabidopsis AtSCL21. This functional divergence between Arabidopsis AtSCL21 and its homolog poplar PagSCL21 may be attributed to the different regulatory mechanisms between the growth and development of herbaceous and woody plants, or the different regulatory mechanisms between primary and secondary growth in plants. More importantly, we found for the first time that PagSCL21 was activated by PagHAM4a and involved in regulating cambial activity and its differentiation into secondary xylem in ‘84K’ poplar stems (Figs 4, 5). However, nothing is known about the upstream and downstream regulation of the PagHAM4a–PagSCL21 module at present. In the future, more research on the synergistic regulation of the PagHAM4a–PagSCL21 module with other transcription factors will be helpful for a better understanding of the regulatory mechanism of the module in secondary xylem formation.
In conclusion, our study identified PagHAM4a and PagHAM4b as positive regulators of cambial differentiation into secondary xylem in ‘84K’ poplar stems. Up-regulation of PagHAM4a and PagHAM4b expression can significantly promote the radial thickening of poplar stems, suggesting the two genes as potential targets for increasing tree biomass. Our study also elucidates the mechanism of PagHAM4a and PagHAM4b and their targets PagSCL21 and PagTCP20 on the cambial activity and its differentiation into secondary xylem in ‘84K’ poplar stems, which provides new insights into the complex regulatory network of secondary xylem formation, and these new insights provide more alternatives for improving tree biomass by molecular design breeding. In addition, we provided a molecular network in which the PagHAM4a–PagSCL21 and PagHAM4b–PagTCP20 modules synergistically regulate secondary xylem formation with members of other transcription factor families (Fig. 6).
Proposed model for HAM4a- and HAM4b-mediated secondary xylem formation in poplar. HAM4a directly activates SCL21 expression to promote the differentiation of cambium into secondary xylem. However, HAM4b directly activates TCP20 expression, and then TCP20 activates WND6 transcription and also interacts with WOX4, which is activated by ARF7 and the HD-ZIP III transcription factor HB4, to promote cambial activity and differentiation of cambium into secondary xylem. In the model, the roles of transcription factors with names in white are supported by this study, and those with names in black have been confirmed by other groups (Zhu et al., 2018; Hou et al., 2020; Hu et al., 2022).
Supplementary data
The following supplementary data are available at JXB online.
Fig. S1. Sequence alignment of PagHAM4a and PagHAM4b proteins.
Fig. S2. Sequence alignment of PagHAM4a, PagHAM4b, and Arabidopsis HAM1–HAM4 proteins.
Fig. S3. Measurement of the Cq values of UBQ and 18S rRNA and the expression patterns of PagHAM4a and PagHAM4b in different tissues of ‘84K’ poplar.
Fig. S4. Expression level of PagHAM4a and PagHAM4b in their corresponding overexpression plants and the effects on internodes, vessel cells, and fiber cells.
Fig. S5. RT-qPCR analysis of target genes in WT and transgenic poplars.
Fig. S6. Cross-sections of the 12th internode of WT, PagHAM4a-, and PagHAM4b-overexpressing poplars cultured in 1/2 MS medium.
Fig. S7. Sequencing confirmation of the pagham4a, pagham4b, and pagham4a;b mutants by CRISPR/Cas9.
Fig. S8. Effects of single- and double mutants of PagHAM4a and PagHAM4b on xylem cell layers, internodes, vessel cells, and fiber cells.
Fig. S9. Venn diagram showing pairwise comparisons of differentially expressed genes among stage 1 and stage 2 of wild type, PagHAM4a-OE, and PagHAM4b-OE plants.
Fig. S10. Sequence analysis of PagSCL21 and PagTCP20 proteins.
Fig. S11. Expression level of PagSCL21 and PagTCP20 in their corresponding overexpression plants and the effect on vessel cells and fiber cells.
Table S1. The scores and sequences of the targets of PagHAM4a and PagHAM4b.
Table S2. List of primers used in this study.
Table S3. Number of differentially expressed genes from pairwise comparisons among stage 1 and stage 2 of wild-type, PagHAM4a-OE, and PagHAM4b-OE plants.
Table S4. Information on the downstream candidate target genes for PagHAM4a.
Table S5. Information on the downstream candidate target genes for PagHAM4b.
Table S6. The RNA-seq data of PagSCL21 and PagTCP20 in PagHAM4a-OE and PagHAM4b-OE transgenic plants.
Abbreviations:
Acknowledgements
We thank Dr Yuan Cao (State Key Laboratory of Tree Genetics and Breeding, Chinese Academy of Forestry) for her guidance on the TEM experiments.
Author contributions
HG: conceptualization of the project and design of the experiments. HG and PZ: wrote the manuscript; PZ: conducted most of the experiments; PZ, QY, YH, PS, and HW: data analysis; XZ and YS: contributed the materials. All authors agreed on the final version of the manuscript.
Conflict of interest
The authors declare no conflict of interest.
Funding
This work was supported by the National Natural Science Foundation of China (31870650).
Data availability
All data supporting the findings of this study are available within the paper and its supplementary data. The RNA-seq data of PagHAM4a-OE and PagHAM4b-OE transgenic poplars have been deposited in the NCBI database (BioProject: PRJNA1023758 and PRJNA1023838).
Comments