Poplar plants (Populus tomentosa Carr. clone 741) were grown in a greenhouse at 23-25°C, 16 h light/8 h dark, 5000 lux supplementary light, and humidity 60%. The Col-0 seedlings of Arabidopsis thaliana were grown in a incubator at 22°C, 16 h light/8h dark, 10,000 lux supplementary light, and humidity 80%. Arabidopsis seeds were germinated on MS medium (8.5 g/L Agar and 30 g/L sucrose) and placed in a lighted incubator. Arabidopsis seedlings were transferred to nutrient soil and continued to be incubated and growth in a light incubator. The culture conditions: 80% humidity, 20–23°CC, 16/8 h light/dark cycle and 555 mmol m−2 s−1 supplemental light (Murashige and Skoog, 1962). Arabidopsis transformation using 30-day-old plants by floral dip method (Zhang et al., 2006). For poplar transformation, the leaf disk method was used to the pCXSN-PtoJAZ5 vector according to the methods previously (Jia et al., 2010a).

Exogenous jasmonic acid (JA) treatment of wild-type (WT) poplars was performed by continuous watering with a 1 mM solution of methyl Jasmonate (MeJA) for one month, and the plants were cultured in a greenhouse and observed for phenotype.

Total RNA was extracted from poplar stems and reverse transcribed into cDNA using a kit (TAKARA), specific primers were designed and the full-length coding sequence of PtoJAZ5 (831 bp) was amplified by Polymerase Chain Reaction (PCR). The amplified fragments were then cloned into pCXSN vector (Chen et al., 2009). For construction of the PtoJAZ5pro::GUS reporter vector, the 1857-bp promoter fragment was amplified and assembled into the pCAMBIA1305 vector. All plant expression vectors were used to transform plant material using the Agrobacterium-mediated leaf disc method according to previous study (Jia et al., 2010b). The software DNAMAN9 (Lynnon Biosoft, San Ramon, CA, USA) was used for multiple sequence alignments, and the software MEGA7.0 was used for phylogenic analysis using the neighbor-joining (NJ) method.

Plant materials such as roots, stems, xylem, bast, old and young leaves in good growth condition were selected and ground thoroughly using liquid nitrogen, and then total RNA was extracted using the Trizol Reagent (Tiangen, China), followed by reverse transcription of cDNA using the PrimeScript™ RT reagent Kit (Takara, Shiga, Japan). Specific quantitative primers were designed as needed, and the quantitative RT-qPCR was performed using SYBR Premix ExTaq™ (Takara) in a TP800 Real-Time PCR machine (Takara). The UBQ11 gene was used as an internal control in poplar and AtUBC in Arabidopsis. The RT-qPCR reaction conditions were 95°C, 30 s; 40 cycles: 95°C, 5 s; 60°C, 1 min; 95°C, 15 s; 60°C, 30 s; 95°C, 15 s. cycles: 95°C, 5 s; 60°C, 1 min; 95°C, 15 s; 60°C, 30 s; 95°C, 15 s. Three biological replicates were required for each gene of the RT-qPCR, and the primers involved were listed in Table S1. To analyze their expression levels, the 2^-ΔΔCt method (Livak and Schmittgen, 2001) was used in Figure 1F and the 2^-ΔCt method (Chiba-Falek et al., 2006) was used in other figures.

The full-length coding sequence of PtoJAZ5 was amplified via the specific primer and assembled into the pCAMBIA1300 vector. Then the tobacco leaf flesh epidermal cells were transformed using A. tumefaciens mediated transformation method. Empty vector 35S-GFP was used as positive control. 4’,6-diamidino-2-phenylindole (DAPI) was used to stain the nucleus of cells. A confocal laser microscope (Olympus FV1200, Tokyo, Japan) was used to detected the DAPI and GFP fluorescent signals.

Full-length and mutant DNA sequences of PtoJAZ5 were amplified and constructed into the PGBKT7 (BD) vector, while full-length DNA sequences of PtoMYC2, PtoMYB3/2/20/21/74, and PtoWND1A/1B/2A/2B/3A/3B/4A/4B/5A/5B/6A/6B were amplified and constructed into the PGADT7 (AD) vector. Positive controls included AD-T and BD-53 vectors, while negative controls consisted of AD-T and BD-lam, BD and AD empty vectors. Y2H gold strain was used in yeast experiments, yeast transformation was performed using the PEG/LiAC method according to the Yeast Protocol Handbook (Clontech), and transformed yeast colonies were selected on nutrient-deficient medium. Transcriptional self-activation verification was performed using media lacking histidine, adenine, and tryptophan (SD-Ade-His-Trp) or only lacking tryptophan (SD-Trp), while protein interaction verification was performed using media lacking leucine and tryptophan (SD-Leu-Trp) or lacking tryptophan, leucine, histidine, and adenine (SD-Leu-Trp-Ade-His). Transcription activation activity was identified using X-α-gal, a chromogenic substrate for yeast galactosidase (MEL1). The transformed yeast cells were grown in a 30°C constant temperature incubator for 3-5 days.

The ORF of PtoJAZ5 was cloned into the PXY106 vector containing the N-terminal of YFP and formed into a complete fusion expression vector. The ORF of PtoMYB3 and PtoWND6A were cloned into the PXY104 vector containing the C-terminal of YFP and formed into a complete fusion expression vector. The resulting constructs were transfered into tobacco leaf flesh epidermal cells using A. tumefaciens mediated transformation method. A confocal laser microscope (Olympus FV1200, Tokyo, Japan) was used to detected the DAPI and YFP fluorescent signals after 2-3 days of incubation.

Poplar and Arabidopsis stems were cross-sectioned by hand or semi-automatic slicers (FINESSE 325, Thermo, Runcorn, UK). Histochemical staining including phloroglucinol-HCl, toluidine blue and GUS staining of stem cross sections was performed according to previously described (Li et al., 2015). Scanning electron microscope (Phenomtm Pure FEI, Eindhoven, Netherland) was used to observe the stem cross-sections according to the operation manual.

For immunocytochemical analysis of JAs, stem samples were fixed with 4% (w/v) 1-ethyl-3-(3-dimethyl aminopropyl)-carbodiimide hydrochloride (Merck KgaA) in PBS, embedded in polyethylene glycol (PEG) 1500, and immunostained with an anti-JA antibody as reported for immunocytochemical examination of JAs (Mielke et al., 2011). Refer to the previous description for the detailed operation steps (Fu et al., 2021).

Klason and acid-soluble lignin methods as previously described (Li et al., 2015) were used to measure the total lignin content of poplar stems. Cellulose, pectin and hemicellulose were determined using Peng method (Peng et al., 2000; Fan et al., 2020).

The ImageJ (https://imagej.nih.gov/ij/) (Collins, 2007) software was used for image data measurement, the GraphPad Prism 7 software (Mitteer et al., 2018) was used for data analysis and statistics and One-way ANOVA followed by Dunnett’s test analysis was performed to evaluate statistical significance. For pairwise comparisons, Dunnett’s test was utilized in the statistical analysis.

The promoter sequences of the lignin synthesis genes PtoCCR2, PtoCOMT2, PtoCCoAOMT1 and PtoMYB3 were amplified and fused to the GUS reporter gene into the PCXGUS-P vector. Agrobacterium cells carrying the 35S-PtoJAZ5, 35S-PtoMYB3 and 35S-PtoWND6A constructs were used as effectors. Empty vector as negative control. Infiltrating into 2-week-old tobacco leaf flesh epidermal cells with different mixes contain the effectors and reporters of Agrobacterium spp. After three days of cultivation, transgenic tobacco leaves were then collected and GUS activity was quantitatively measured by a spectrophotometry according to previously method (Jefferson et al., 1987).

To determine the distribution of JA in poplar stem, we conducted immunocytochemical detection experiments and observed that JA is predominantly located in the phloem and developing xylem regions (Figure S1), suggesting its potential role in the regulation of xylem development. To further explore the essential role of JA in SCW development, we observed sections of the JA synthesis mutant opdat1 (Zhao et al., 2021) under scanning electron microscope, which revealed significant reduction in SCW thickness of xylem fiber cells, xylem vessel cells and phloem fibers (Figures 1A-D). Furthermore, the lignin and cellulose content of SCW also significantly decreased in opdat1 mutants (Figure 1E), while the xylan content did not differ significantly. RT-qPCR analysis further demonstrated a significant down-regulation in the expression level of key SCW synthesis-related enzyme genes, including PAL1, F5H2, COMT1, CCoAOMT1, CESA3A, and GT43B in the opdat1 mutant poplar (Figure 1F). These findings suggest that the decreased JA content leads to delayed SCW development in poplar.

To investigate the effects of JA on lignin deposition and SCW thickening in poplar, we treated wild-type (WT) poplar with MeJA solution for one month. Methyl jasmonate (MeJA) treatment significantly inhibited the growth of poplar stems (Figure S2). Toluidine blue and phloroglucinol-HCl staining of sections from the sixth internode stem revealed that MeJA treatment significantly increased the depth of xylem staining and SCW thickness (Figure S3A). Scanning electron microscopy also showed that MeJA treatment promoted the thickening of SCW in vessel and fiber cells (Figures S3B-D), which was consistent with the increased lignin content in treated poplar (Figure S3E). Furthermore, RT-qPCR analysis revealed that the expression levels of lignin synthase genes (COMT2, CCoAOMT1, CCR2, CAld5H2, PAL1, CSE1, C4H2, and 4CL5), cellulose synthase gene (CESA3A), xylan synthase gene (GT43B), and some laccase genes (LAC29, LAC30, and LAC41) were significantly induced by MeJA treatment (Figure S3F). The application of MeJA to the opdat1 mutant also significantly restored the phenotype of SCW thinning (Figure 2), suggesting that MeJA treatment promotes SCW thickening by inducing the expression of genes related to secondary wall biosynthesis.

It is well known that the degradation of JAZ proteins is the main mechanism of JA signaling in plants (Thines et al., 2007). To investigate the role of JA signaling in lignin deposition and SCW synthesis, we aimed to reduce JA signaling by overexpressing JAZ genes. We identified 12 JAZ proteins in poplar and constructed a phylogenetic tree of Arabidopsis and poplar JAZ protein sequences (Figure S4A). Poplar JAZ5 has a 58.48% amino acid sequence similarity to Arabidopsis AtJAZ1/2, and contains the conserved Tify and Jas domains with a conserved LPIARRA amino acid sequence in the Jas domain (Figure S4B). Previous studies have demonstrated that AtJAZ1 is involved in plant growth, development, and response to freezing and senescence (Robson et al., 2010; Zhou et al., 2017; Valenzuela-Riffo et al., 2020). Gene expression profile data of poplar JAZ proteins showed that PtoJAZ5 is specifically expressed in the xylem (Figures S4C, D), indicating that PtoJAZ5 may play a role in the transcriptional regulation of JA signaling in the xylem region. Therefore, we cloned the poplar PtoJAZ5 gene for further investigation.

To investigate the subcellular localization of PtoJAZ5, we generated a PtoJAZ5-GFP fusion protein driven by the CaMV 35S promoter and transiently expressed it in tobacco epidermal cells. The localization of the fusion protein was visualized by observing the GFP fluorescence signal. The results showed that PtoJAZ5 is primarily localized in the nucleus and co-localized with the DAPI signal, confirming that PtoJAZ5 is a nuclear-localized protein (Figure 3A). To further explore the transcriptional activity of PtoJAZ5, we fused PtoJAZ5 to the GAL4 DNA binding domain (BD) and transformed the recombinant plasmids (BD-PtoJAZ5) into yeast cells. The transformed yeast cells were able to grow on selective medium lacking tryptophan (SD/-T), but could not grow on medium lacking tryptophan, histidine and adenine (SD/-ATH), indicating that PtoJAZ5 does not have transcriptional activation activity (Figure 3B).

To investigate the spatiotemporal expression pattern of PtoJAZ5, we constructed a PtoJAZ5 promoter-GUS fusion vector. GUS staining revealed that PtoJAZ5 is mainly expressed in the cambium layer and developing xylem region, young leaves, roots, and stems of poplar (Figure 3C). These results suggest that PtoJAZ5 may play a role in the growth and development of xylem.

To determine the role of PtoJAZ5 in SCW formation, we overexpressed PtoJAZ5 in poplar under the control of the CaMV 35S promoter. RT-qPCR assay was performed to select transgenic lines (Line 8 and Line 9) with high expression of PtoJAZ5 for further experiments (Figure S5D). No significant difference in the appearance phenotype was observed between transgenic and WT poplars, except for a slight increase in height of some transgenic plants (Figures S5A-C). RT-qPCR results showed that the expression levels of JA signaling marker genes (PR4, VSP1, VSP2) were significantly down-regulated in PtoJAZ5 overexpression plants (Figure S5E), indicating the suppression of JA signaling.

To further investigate the effect of PtoJAZ5 on SCW formation, we performed section staining and scanning electron microscopy on the sixth internode stem of transgenic poplar. In comparison to WT, a lighter staining color of the xylem region and a significant reduction in the thickness of vessel cells and fiber cells were observed in the PtoJAZ5 overexpression plants compared with WT. However, the area of vessel cells and fiber cells was significantly increased (Figures 4A-E). RT-qPCR results showed that the expression levels of SCW biosynthase genes including lignin synthase genes (C4H2, CCR2, CAD1, CCoAOMT1, and CAld5H2), cellulose synthase genes (CESA3A, CESA2B), and xylan synthase genes (GT8D and GT43B) were significantly down-regulated in the transgenic plants (Figures 4F, G). Additionally, exogenous MeJA treatment significantly restored the phenotype of SCW thinning in transgenic poplars (Figure S8). These results indicate that the SCW thinning in PtoJAZ5 overexpression poplar is mediated by a decrease in the expression level of SCW synthesis-related enzyme genes.

To examine the conservation of JAZ5 function, we also heterologously overexpressed PtoJAZ5 in Arabidopsis. Analysis of stem sections of transgenic Arabidopsis showed that the autofluorescence intensity of lignin was significantly reduced, and the color of toluidine blue staining and phloroglucinol-HCl staining in stem of Arabidopsis overexpressing PtoJAZ5 was lighter compared to WT. Scanning electron microscopy revealed significant thinning of SCW in transgenic Arabidopsis (Figure S6). Consistent with the phenotypic observation, RT-qPCR results revealed that transcript levels of SCW synthase genes in transgenic Arabidopsis were significantly reduced (Figure S7). These results suggest that PtoJAZ5 is able to inhibit lignin deposition and SCW thickening in Arabidopsis, indicating its conserved function.

Intensive studies have shown that JA signaling is mediated by the interaction of JAZ proteins with transcription factors, which results in the suppression or activation of downstream genes depending on JA levels (Chini et al., 2007; Thines et al., 2007; Katsir et al., 2008; Pauwels et al., 2010). Using yeast two-hybrid (Y2H) experiments, we searched for SCW synthesis-related transcription factors that interact with JAZ5. The Y2H results revealed that PtoJAZ5 interacted with several SCW synthesis switch masters, such as MYB2, MYB3, MYB74, WND3A, WND4A, WND6A, and WND6B (Figure 5A). Previous studies have shown that poplar MYB3/2/20/21, which are mainly expressed in the developing xylem region, act as secondary switches in the regulatory network of SCW synthesis (Wilkins et al., 2009; Mccarthy et al., 2010). The WNDs, belonging to the NAC gene family, are first-order switches in the regulatory network of SCW synthesis in poplar (Zhong et al., 2010; Zhong et al., 2011). Based on the gene expression analysis and Y2H experiment, we selected MYB3 and WND6A as representative transcription factors to further investigate the function of PtoJAZ5 (Figure 5A and Figure S9). We constructed PtoJAZ5-PXY104, a C-terminal fusion expression vector of PtoJAZ5 with yellow fluorescent protein (PtoJAZ5-cYFP), and PXY106-MYB3 and PXY106-WND6A, fusion expression vectors of the N-terminal of YFP with MYB3 and WND6A (nYFP-MYB3 and nYFP-WND6A), for fluorescent bimolecular complementation experiments (BiFC). The combination of PtoJAZ5/MYB3 and PtoJAZ5/WND6A showed a yellow fluorescent signal in tobacco leaves, while the other two negative controls showed no signal when observed under laser confocal electron microscopy (Figures 5B, C). These results suggest that PtoJAZ5 interacts with MYB3 and WND6A in plant cells.

To identify the core amino acid sequence fragments of PtoJAZ5 that interact with MYB3 or WND6A, we constructed a yeast fusion expression vector of PGBKT7 by truncating PtoJAZ5 into different structural domains, including JAS domain, Tify domain, and N-terminal domain, based on the amino acid sequence specificity of PtoJAZ5. The Y2H assays showed that proteins containing the JAS domain could interact with MYB3 or WND6A, and yeast transformants grew normally on SD/-ATLH-deficient medium. Conversely, yeast transformants failed to grow when the JAS domain was absent (Figures S10A, B). Therefore, we conclude that the JAS domain is the necessary core sequence for PtoJAZ5 to interact with MYB3 or WND6A.

Previous studies have revealed that the JAS structural domain mediates the degradation of JAZ protein by JA receptor COI1 (Sheard et al., 2010). Based on amino acid sequence analysis of PtoJAZ5, we mutated the core sequence of the JAZ degradation domain, LPIAR, to LPIGA, to construct a PtoJAZ5 mutant protein (PtoJAZ5m). Using a Y2H assay to verify whether PtoJAZ5m protein could still interact with MYB3/WND6A, and found that this sequence mutation did not affect the interaction of PtoJAZ5 with MYB3 or WND6A (Figure S10C). Next, we investigated whether the LPIAR sequence mutation affected the degradation of PtoJAZ5. We constructed PtoJAZ5-GFP and PtoJAZ5m-GFP fusion protein expression vectors and transformed them into Arabidopsis using floral dip method. Transgenic Arabidopsis seedlings were treated with MeJA solution and observed under a fluorescence microscope. The results showed that the green fluorescence signal was diminished or even disappeared in transgenic Arabidopsis roots expressing PtoJAZ5 but not in those expressing PtoJAZ5m (Figures S11A, B). These findings indicate that LPIAR is the core sequence of the degradation domain of PtoJAZ5, which responds to JA signals.

To investigate whether the interaction between PtoJAZ5 and MYB3/WND6A affects downstream SCW synthase gene expression levels, we used the promoters of CCoAOMT1, CCR2, COMT2, and C4H2 to control the expression of GUS reporter genes. The generated reporter vectors were co-transformed into tobacco leaf epidermal cells with overexpression vectors, enabling transactivation analysis. Our findings demonstrate that MYB3 or WND6A alone significantly activated the expression of GUS genes under the control of the aforementioned promoters. However, this activation was inhibited by PtoJAZ5. Upon MeJA treatment, the repressive effect of PtoJAZ5 on MYB3 and WND6A was alleviated (Figure 6). Thus, our results suggest that PtoJAZ5 regulates SCW biosynthesis via interacting with MYB3 or WND6A, and that this regulatory mechanism is modulated by JA signaling.