Natural variation in PtoCPK3 governs drought tolerance by orchestrating xylem remodeling and lignin metabolism in Populus.

Natural variation in PtoCPK3 governs drought tolerance by orchestrating xylem remodeling and lignin metabolism in Populus - PMC Skip to main content An official website of the United States government Here's how you know Here's how you know Search Log in Primary site navigation Search Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice Sci Adv . 2026 Jun 3;12(23):eaee2172. doi: 10.1126/sciadv.aee2172 Natural variation in PtoCPK3 governs drought tolerance by orchestrating xylem remodeling and lignin metabolism in Populus Dan Wang 1, 2, 3, † , Yongsen Jiang 1, 2, 3, † , Mingyang Quan 1, 2, 3, † , Borong Liao 1, 2, 3 , Yongming Chen 2, 3 , Mingyue Gu 2, 3 , Yicen Guan 2, 3 , Donghai Zhang 2, 3 , Shitong Qin 2, 3 , Yuanyuan Fang 2, 3 , Weina Qi 2, 3 , Liang Xiao 1, 2, 3 , Tailin Ren 2, 3 , Tingxuan Zhao 2, 3 , Qingzhang Du 1, 2, 3 , Pär K Ingvarsson 4 , Yousry A El-Kassaby 5 , Deqiang Zhang 1, 2, 3, * Author information Article notes Copyright and License information 1 State Key Laboratory of Tree Genetics and Breeding, College of Biological Sciences and Technology, Beijing Forestry University, Beijing 100083, China. 2 National Engineering Research Center of Tree Breeding and Ecological Restoration, College of Biological Sciences and Technology, Beijing Forestry University, Beijing 100083, China. 3 Key Laboratory of Genetics and Breeding in Forest Trees and Ornamental Plants, Ministry of Education, College of Biological Sciences and Technology, Beijing Forestry University, Beijing 100083, China. 4 Linnean Center for Plant Biology, Department of Plant Biology, Swedish University of Agricultural Sciences, SE-750 07 Uppsala, Sweden. 5 Department of Forest and Conservation Sciences, Faculty of Forestry, Forest Sciences Centre, University of British Columbia, Vancouver, BC V6T 1Z4, Canada. * Corresponding author. Email: deqiangzhang@bjfu.edu.cn † These authors contributed equally to this work. Roles Dan Wang : Investigation, Validation, Visualization, Writing - original draft, Writing - review & editing Yongsen Jiang : Formal analysis, Investigation, Resources, Validation, Visualization Mingyang Quan : Conceptualization, Formal analysis, Investigation, Methodology, Resources, Supervision, Validation, Visualization, Writing - original draft, Writing - review & editing Borong Liao : Formal analysis, Investigation, Resources, Visualization Yongming Chen : Investigation Mingyue Gu : Investigation, Validation Yicen Guan : Methodology, Validation, Visualization Donghai Zhang : Investigation, Validation Shitong Qin : Resources Yuanyuan Fang : Resources Weina Qi : Investigation Liang Xiao : Writing - review & editing Tailin Ren : Formal analysis, Investigation Tingxuan Zhao : Investigation, Resources Qingzhang Du : Formal analysis, Validation, Writing - review & editing Pär K Ingvarsson : Supervision, Writing - review & editing Yousry A El-Kassaby : Writing - review & editing Deqiang Zhang : Conceptualization, Data curation, Funding acquisition, Methodology, Project administration, Resources, Supervision, Validation, Visualization, Writing - review & editing Received 2025 Nov 26; Accepted 2026 Apr 23; Collection date 2026 Jun 5. Copyright © 2026 The Authors, some rights reserved; exclusive licensee American Association for the Advancement of Science. No claim to original U.S. Government Works. Distributed under a Creative Commons Attribution NonCommercial License 4.0 (CC BY-NC). This is an open-access article distributed under the terms of the Creative Commons Attribution-NonCommercial license , which permits use, distribution, and reproduction in any medium, so long as the resultant use is not for commercial advantage and provided the original work is properly cited. PMC Copyright notice PMCID: PMC13232617  PMID: 42234763 Abstract Drought stress severely limits the growth of perennial trees. Xylem structure is central to water conduction; however, lignin monomer composition driving xylem remodeling for drought adaptation remains enigmatic. By integrating multi-omics analyses, genotype-environment association analysis, and metabolite-based genome-wide association studies, we identified PtoCPK3 as a key gene associated with precipitation-related traits and the aridity index in Populus tomentosa . This variation in association was linked to coniferyl alcohol and ferulic acid, two metabolites related to guaiacyl (G)-lignin monomer biosynthesis. Overexpressing PtoCPK3 enhanced drought tolerance in transgenic poplar by promoting xylem remodeling associated with a decreased lignin S/G ratio. Mechanistically, drought-induced Ca 2+ signaling activates PtoCPK3 to phosphorylate PtoERF72 at Ser 96 , enhancing its activation of PtoWOX13b and direct repression of PtoUGT72AZ2 . Synergistic repression of PtoUGT72AZ2 by PtoWOX13b reduces glycosylation of G-monomer precursors, favoring ferulic acid and coniferyl alcohol accumulation. Furthermore, natural variants in PtoCPK3 and PtoERF72 drive geographic divergence, with the PtoCPK3 II _ PtoERF72 CC genotype conferring superior drought resilience via elevated phosphorylation efficiency. This module links drought signaling to xylem remodeling, providing genetic targets for breeding drought-resilient trees. The protein kinase PtoCPK3 regulates drought adaptation by lignin metabolism in Populus . INTRODUCTION Forests, covering approximately 30% of terrestrial land, play a crucial role in regulating regional climate and global biogeochemical cycles ( 1 , 2 ). However, climate change and human disturbance are intensifying the drought events, leading to widespread tree mortality and compromised ecosystem resilience ( 3 – 5 ). Because trees are long-lived and sessile, their survival depends on structural and physiological adjustments that maintain water transport under abiotic stress. A key strategy is xylem structural adaptation, which involves coordinated changes in vessel structure and cell wall composition ( 6 ). Drought-resistant species typically undergo vessel lumen reduction and cell wall thickening, together with concomitant adjustments in pit architecture, both of which enhance xylem strength and mitigate the risk of cavitation, thereby maintaining water transport ( 7 – 9 ). Lignin, a major component of the secondary cell wall (SCW), not only provides rigidity and hydrophobicity but also determines the mechanical properties of vessels through its monomeric composition. Variations in syringyl-to-guaiacyl (S/G) ratio, coupled with increased lignin deposition, have been consistently observed as adaptive responses to drought stress ( 10 – 12 ). Beyond structural adaptations, calcium signaling serves as a central regulator of drought responses, where transient calcium ion (Ca 2+ ) fluctuations are decoded by calcium-dependent protein kinases (CPKs) ( 13 , 14 ). Ca 2+ binding to the CPK activation domain activates kinase activity, phosphorylating downstream targets such as transcription factors, facilitating their cytoplasm-to-nucleus translocation, which mediates gene expression and stress adaptation ( 15 – 17 ). The best-characterized function of CPKs in drought adaptation is their role in regulating stomatal closure. For example, CPK4 and CPK11 modulate stomatal responses by phosphorylating ABF1 and ABF4 involved in drought resistance, contributing to water retention during drought stress ( 18 ). Although the role of CPKs in guard cell regulation is well established, their involvement in xylem structural adaptations, such as hydraulic acclimation and lignin biosynthesis, remains poorly characterized. Elucidating these integrated pathways will provide mechanistic insights into how signaling and structural adaptations converge to enhance drought resilience in trees. Xylem architecture remodeling and lignin metabolism are widely recognized as central to drought resistance in trees; however, the cascading molecular regulatory networks that coordinate these processes remain largely unknown. Transcriptomic and metabolomic analyses have provided valuable insights into drought-responsive genes and metabolic pathways, yet they primarily reflect steady-state changes and overlook posttranscriptional regulation ( 19 , 20 ). Thus, posttranscriptional regulatory layers, particularly proteomic and phosphoproteomic modifications, are essential for elucidating the dynamic phosphorylation events that control enzyme activity, signal transduction, and transcription factor function. Integrating these datasets within a multi-omics framework enables a systems-level understanding of signaling hierarchies and cross-layer coordination, providing a comprehensive view of how molecular signaling and metabolism jointly shape xylem plasticity under drought stress ( 21 ). At the population level, genomic analyses reveal heritable variation through genome-wide polymorphisms such as single-nucleotide polymorphisms (SNPs) and insertion and/or deletion (InDel). Metabolite-based genome-wide association studies (mGWASs) directly link genetic variation to lignin metabolic remodeling, but often overlook environmental effects, which can confound genotype-phenotype relationships ( 22 – 24 ). Genotype-environment association (GEA) addresses this limitation by detecting alleles whose frequencies covary with climatic gradients (temperature, precipitation, and aridity), thereby uncovering adaptive differentiation across landscapes in species such as Populus trichocarpa and Quercus robur ( 25 – 27 ). Integration of mGWAS and GEA reveals the mechanistic basis of lignin metabolic regulation and identifies the selective pressures shaping these alleles in nature. Together, they bridge molecular regulation and adaptive evolution, illuminating how natural genetic variation orchestrates drought resilience and resource allocation in trees. Here, we used a multi-omics and association genetics framework to identify PtoCPK3 as a central regulator of transcriptional and posttranslational changes, with allelic variation linked to ferulic acid and cinnamyl alcohol levels and to drought-related environmental indices . PtoCPK3 links xylem structural remodeling with lignin S/G ratio modulation, improving drought tolerance, acting via autophosphorylation and phosphorylation of PtoERF72 to enhance its transcriptional regulation efficiency. Furthermore, natural allelic variation in PtoCPK3 [1–base pair (bp) InDel] and PtoERF72 (nonsynonymous variants) underlies geographic divergence, with the PtoCPK3 II _PtoERF72 CC genotype exhibiting the greatest drought tolerance due to enhanced phosphorylation efficiency. Collectively, these findings uncover a regulatory module for lignin metabolites, PtoCPK3-ERF72-WOX13b-UGT72AZ2, providing a mechanistic basis and genetic targets for engineering drought-resilient tree varieties. RESULTS Multi-omics analyses reveal the phenylpropanoid pathway as the pivotal drought-response axis in Populus tomentosa To investigate the metabolic responses in the xylem tissue to drought stress, we subjected Populus tomentosa cultivar “1316” to a 15-day water-withholding treatment. Drought resulted in severe wilting, decreased xylem width, and thickened SCWs, indicating that the plants had perceived water-deficit cues and undergone structural remodeling ( Fig. 1, A and B ). Subsequently, broad-targeted xylem metabolomic profiling identified 395 differentially accumulated metabolites (fig. S1A and data S1). Metabolite set enrichment analysis (MSEA) revealed notable enrichment of metabolites associated with phenylpropanoid biosynthesis under drought stress, whereas metabolites related to flavonoid biosynthesis were predominantly enriched in well-watered conditions ( Fig. 1C ). These findings suggest that drought stress preferentially redirects carbon flux toward lignin biosynthesis, thereby enhancing SCW deposition. We further analyzed changes in all metabolites within the phenylpropanoid pathway and found that three general precursors ( l -phenylalanine, cinnamic acid, and p -coumaric acid) of the phenylpropanoid pathway were up-regulated, most flavonoids (53 of 67, 79.10%) were down-regulated, whereas nearly all lignin-related metabolites (9 of 10, 90.00%) were up-regulated under drought stress ( Fig. 1D ). Furthermore, lignin-related metabolites exhibited significantly greater changes than flavonoid metabolites in response to drought ( P = 2.69 × 10 −9 ; Fig. 1E ), suggesting that lignin metabolism plays a predominant role in the drought response. Fig. 1. Integrated transcriptomic, proteomic, phosphoproteomic, and metabolomic profiling of the P. tomentosa cultivar genotype “1316” under drought stress. Open in a new tab ( A ) Four-month-old soil-grown P. tomentosa “1316” seedlings were used as the experimental materials for 15-day drought treatment until the soil relative water content (RWC) decreased to 19%. Well-watered poplars were used as controls. Scale bar, 15 cm. ( B ) Xylem width (left) and cell wall thickness (right) of P. tomentosa “1316” under drought treatment and well-watered conditions. ( C ) Metabolite set enrichment analysis (MSEA) of the broad-target metabolome data was determined based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways. The colors of the statistics in the inset table correspond to the colors of the enrichment curves and pathway labels. ( D ) Changes in metabolites in the phenylpropanoid pathway and its downstream lignin and flavonoid pathways under drought stress treatment, with deeper red indicating higher content. ( E ) Comparison of the |log 2 FC| values of differentially expressed lignin and flavonoid metabolites. Box plots show the distribution quantiles. Two-sided Wilcoxon rank sum test, P < 0.001. ( F ) Scatterplot of merged P values from directional analysis (DPM, y axis) and nondirectional analysis (Brown, x axis). Prioritized genes (blue) show consistent multi-omics directions, while penalized genes (red) show inconsistencies. Unadjusted P values are shown. ( G ) Heatmap of significantly penalized or prioritized top genes [Brown, false discovery rate (FDR) < 0.001]. Prioritized genes were often characterized by coordinated changes in mRNA and protein expression, as well as phosphorylation levels, while penalized genes often showed inconsistencies in the direction of all three. Known stress response genes are listed and colored as directionally penalized or prioritized. ( H ) Enrichment map of pathways and processes in response to drought stress of poplar. The network shows pathways as nodes that are connected by edges if the corresponding pathways share many genes. To further elucidate the regulatory networks driving these metabolic shifts, we integrated the transcriptome, proteome, and phosphoproteome profiles of xylem tissue under both well-watered and drought conditions. In total, three types of differentially expressed molecules (DEMs), including 4383 genes, 1702 proteins, and 3612 phosphosites from 1854 phosphoproteins, were identified, collectively indicating extensive transcriptional and posttranslational reprogramming under drought stress (fig. S1B and data S2 to S4). Pathway enrichment analysis revealed that a substantial proportion of the enriched pathways corresponded to metabolic processes for the DEMs (fig. S1C). Among all the identified molecules, 1206 genes (corresponding to 4441 phosphorylation sites) were shared across the three omics datasets (data S5). However, only 687 phosphosites (corresponding to 230 proteins) exhibited notable changes across all datasets. Of these, 149 were consistently up-regulated and 417 were consistently down-regulated (fig. S1D). The results showed that synergistic changes among the three regulatory layers (87.03%) played a central role in the drought response of P. tomentosa . We further integrated transcriptome, proteome, and phosphoproteome by modeling directional P value merging (DPM method) associations between the three omics layers. DPM analysis identified 2770 drought-response phosphorylation sites (DRPSs; P < 0.05) from 928 genes, among which 297 DRPSs were excluded due to directional conflicts relative to the reference nondirectional analysis ( P < 0.05; Fig. 1F ). Although phosphorylation modification can positively or negatively regulate protein activity, the transcriptional, proteomic, and phosphoproteomic levels of directional priority DRPSs typically exhibit concordant changes (1716 of 2770, 61.95%), with a predominant down-regulation (1033 of 2770, 37.29%). In contrast, DRPSs excluded by the DPM method frequently exhibited opposing trends between protein abundance and phosphorylation levels, as well as discordant transcriptional changes ( Fig. 1G ). Pathway enrichment analysis of DPM-prioritized targets revealed 320 biological processes, including stress response, signal transduction, and secondary metabolism. Twenty-six pathways were supported by all three omics layers, notably “protein phosphatases and associated proteins,” “lignin metabolic process,” and “flavonoid metabolic process” ( Fig. 1H and fig. S1E), suggesting drought-activated transcriptional and posttranslational regulation of phenylpropanoid biosynthesis. The DPM analysis additionally captured unique drought-associated programs, such as “cell wall biogenesis” (family-wise error rate, FWER = 0.010; fig. S1E), which were absent from the nondirectional results. In line with the metabolite profiles of the phenylpropanoid pathway, directional priority genes involved in flavonoid synthesis were consistently down-regulated at the transcriptional, protein, and phosphorylation levels, whereas lignin biosynthetic genes showed synergistic changes across multi-omics layers (fig. S1F). In summary, the coordinated regulation of metabolic flux at the transcriptional, proteomic, and phosphoproteomic levels preferentially redirects carbon flux toward lignin biosynthesis, thereby enhancing the lignification of SCWs and reducing water loss. Environment drives adaptive selection of lignin metabolites in P. tomentosa population To further explore the drought adaptability of P. tomentosa , we analyzed environmental differentiation among three genetically and geographically defined subpopulations, south (S; n = 102), northeast (NE; n = 92), and northwest (NW; n = 108) ( 28 ). Analysis of variance (ANOVA) revealed significant differences between the S and NE/NW subpopulations across 24 environmental variables, including 15 precipitation-related traits and the aridity index ( Fig. 2A , fig. S2A, and data S6). GEA identified 260 drought-adaptive SNPs annotated to 215 genes significantly associated with 18 environmental variables ( P < 1.0 × 10 −5 ; Fig. 2B and data S7). Notably, 72 SNPs exhibited pleiotropic effects, with the greatest overlap detected between the drought index and precipitation-related variables (fig. S2B). In addition, drought-adaptive SNPs exhibited significantly higher F st values than neutral variants ( P < 0.0001; Fig. 2C ), indicating spatially varying selection. Mantel and partial Mantel tests further showed that neutral loci displayed isolation by distance (IBD) only within the S subpopulation, whereas adaptive loci displayed both IBD and isolation by environment (IBE) across populations ( Fig. 2, D and E , and fig. S2, C to H), supporting the view that geographic and environmental factors jointly shape adaptive divergence. Partial redundancy analysis (pRDA) showed that precipitation-related variables and the aridity index explained 45.3% of the genetic variance (fig. S2I). Gradient forest modeling identified five key environmental predictors: annual precipitation (Bio12), precipitation in October and November (Prec10, Prec11), precipitation seasonality (Bio15), and aridity index, which highlights contrasting adaptive responses between the humid S and the relatively drier N (NE and NW) subpopulations (fig. S2, J and K). Fig. 2. Genomic and metabolic mechanisms underlying drought adaptation in P. tomentosa . Open in a new tab ( A ) Distribution of 302 P. tomentosa accessions along the aridity index gradient, where pie charts represent ancestral component proportions inferred from ADMIXTURE. ( B ) Identification of SNPs related to precipitation and aridity index, and their pleiotropy using environmental association analysis (GEA). Each colored circle represents a distinct trait; dots correspond to trait-associated loci. Larger dots and radial lines highlight pleiotropic loci, with exhibiting genes highlighted on the outer circle. ( C ) Comparison of genetic differentiation ( F st) among south (S), northeast (NE), and northwest (NW) groups using neutral SNPs versus 260 drought-adaptive variants detected by latent factor mixed models (LFMMs). Box plots display the median (center line), quartiles (box limits), and range (whiskers); data beyond whiskers are outliers. Significance was determined via two-tailed Wilcoxon tests. ( D and E ) Isolation-by-distance (D) and isolation-by-environment (E) analyses (two-sided partial Mantel tests) for the S subpopulation ( n = 102) using neutral (blue dots/black line) and adaptive (yellow dots/red line) variants. Shaded areas of linear regression denote 95% confidence intervals. ( F ) Divergent patterns of lignin metabolites detected through targeted metabolomics in S versus NE and S versus NW comparisons. Bars represent 99% CIs of Q st. Symbols denote observed Q st: Red dots indicate no significant difference between Q st and F st in S versus NE, while blue dots indicate the same for S versus NW. Upward and downward triangles indicate significantly higher and lower metabolite content in NE/NW relative to S, respectively. Dashed lines indicate genome-wide neutral F st thresholds. ( G ) Identification of SNPs with metabolite-based genome-wide association (mGWAS). Each point corresponds to a trait, with multiple points indicating SNPs associated with a single gene ( y axis). ( H ) Venn diagram showing PtoCPK3 as the shared gene identified by multi-omics, GEA, and mGWAS analyses. To establish the link between genomic adaptation and lignin metabolic response, we further evaluated the divergence of 13 key metabolites (3 general precursors and 10 lignin metabolites) among the three geographic subpopulations based on Q st- F st comparisons. Specifically, we identified significant metabolic divergence in pairwise comparisons: Caffeic acid, ferulic acid, coniferyl alcohol, sinapinaldehyde, sinapyl alcohol, and p -coumaraldehyde differed in S versus NE; l -phenylalanine, caffeic acid, ferulic acid, coniferyl alcohol, sinapinaldehyde, and sinapyl alcohol differed in S versus NW; whereas only ferulic acid, sinapyl alcohol, and p -coumaraldehyde showed divergence in NE versus NW ( Fig. 2F and data S8). Strikingly, the greatest divergence was observed between the S and N (NE/NW) subpopulations, consistent with regional aridity index gradients. Seven lignin metabolites ( l -phenylalanine, caffeic acid, ferulic acid, coniferyl alcohol, sinapinaldehyde, sinapyl alcohol, and p -coumaraldehyde) showed clear selection signatures, suggesting that these lignin metabolites concurrently contribute to drought adaptation and physiological drought response in poplar. Subsequently, an mGWAS was performed using the same SNP dataset used in GEA analysis to dissect the genetic basis of the seven metabolites. A total of 62 SNPs were significantly associated with seven lignin metabolites ( Fig. 2G , fig. S3, and data S9). The mean phenotypic variance explained (PVE) was 9.57%, with individual SNP accounting for a PVE of 8.85% to 11.95%. Mapping of these substantial SNPs identified 56 candidate genes. By integrating DPM, GEA, and mGWAS results, we identified Ptom.003G.001130 , encoding calcium-dependent protein kinase 3 ( PtoCPK3 ), as a candidate regulator potentially involved in drought responses in the xylem ( Fig. 2H ). Both transcript and protein abundances of PtoCPK3 were significantly up-regulated under drought stress, and its autophosphorylation levels at Ser 235 , Thr 517 , and Thr 35 were markedly elevated, indicating that PtoCPK3 functions as a positive regulator of drought tolerance in poplar. Furthermore, PtoCPK3 was significantly associated with precipitation traits and the aridity index, as well as with environmentally selected metabolites such as ferulic acid and coniferyl alcohol. Collectively, these findings suggest that PtoCPK3 regulates lignin metabolite accumulation through phosphorylation-mediated signaling, thereby enhancing drought resistance and adaptive capacity in poplar. PtoCPK3 improves drought tolerance in transgenic poplar by xylem structural remodeling Phylogenetic analysis and multiple sequence alignment demonstrated that PtoCPK3 clusters tightly with its “84K” orthologs and shares high amino acid sequence identity (>85%) (figs. S4 and S5A). Consequently, to further characterize drought responses of PtoCPK3 , we generated transgenic poplar lines overexpressing PtoCPK3 (designated PtoCPK3-OE ) and knockdown lines (designated ptocpk3 -KD) in “84K” poplar ( Fig. 3A and fig. S5, B and C). Four-month-old wild-type (WT), ptocpk3 -KD, and PtoCPK3 -OE plants were exposed to progressive drought stress by gradually reducing soil relative water content (RWC) from 70% to 20% over a 15-day period. Under these conditions, ptocpk3 -KD plants exhibited severe wilting and partial mortality, WT plants displayed pronounced leaf drooping, whereas PtoCPK3-OE lines showed only mild wilting ( Fig. 3A and fig. S5, D and E). We further measured several physiological indicators associated with drought stress in the treated poplars. The results showed that stomatal conductance, abscisic acid (ABA), and Ca 2+ content were higher in the PtoCPK3 -OE lines than in WT, while those in the ptocpk3 -KD lines were markedly lower ( Fig. 3, B and C ). A similar pattern was observed for chlorophyll and proline (PRO) contents in WT, PtoCPK3 -OE, and ptocpk3 -KD plants (fig. S5, F and H). Fig. 3. Xylem morphology of PtoCPK3 transgenic poplars under drought stress. Open in a new tab ( A ) One-month-old poplar tissue-cultured seedlings were transplanted into soil and grown for 4 months under well-watered conditions, followed by a 2-week drought treatment until the soil relative water content (RWC) decreased to 19%. Well-watered poplars were used as controls. Scale bar, 15 cm. WT, wild type; PtoCPK3 -OE, PtoCPK3 overexpressed line; ptocpk3 -KD, PtoCPK3 knockdown line. ( B to E ) Determination of abscisic acid (ABA) (B) and Ca 2+ content (C), stem water potential (D), and xylem hydraulic conductivity (E) of all experimental poplars. The poplars are indicated by colored bars: gray for WT, orange for PtoCPK3 -OE, and green for ptocpk3 -KD. ( F ) Histochemical staining of stem internode 15 (IN15) transverse sections of all experimental poplars. Sections were observed at magnifications of ×10 (scale bar, 50 μm) and ×40 (scale bar, 60 μm); red dots represent vessel cells. Xy, xylem. ( G and H ) Vessel density (G) and vessel lumen area (H) were quantified accordingly ( n = 100). In each violin, the central black dashed line represents the median, while the upper and lower dashed lines indicate the 25th and 75th percentiles. Significant differences among groups were tested using one-way ANOVA followed by Tukey’s test: Different letters represent significant differences. Values represent the mean ± SD of three biological replicates. Given the involvement of PtoCPK3 in lignin metabolism, we first identified it through quantitative reverse transcription polymerase chain reaction (qRT-PCR). It was found that the expression level of PtoCPK3 in the xylem of poplar was substantially higher than that in the leaves and roots (fig. S5J). To precisely determine its intracellular localization, RNA in situ hybridization experiments showed that PtoCPK3 transcripts were abundant in the phloem, developing xylem, and mature xylem (fig. S5K). Consequently, we hypothesized that it enhances drought tolerance by regulating xylem structural traits. To test this, stem water status and hydraulic properties were assessed. Under drought conditions, PtoCPK3-OE lines maintained higher stem RWC and water potential, as well as xylem hydraulic conductivities, which decreased only by 10.66%, 47.33%, and 26.65%, respectively, compared to well-watered controls ( Fig. 3, D and E , and fig. S5L). In contrast, ptocpk3 -KD stems showed drastic reductions, including a 3.19-fold decline in water potential, whereas PtoCPK3-OE stems displayed a 2.69-fold increase relative to WT ( Fig. 3D ). To examine xylem vessel structure, vessel cell morphology was first analyzed. Safranin O–Fast Blue staining in stem internode 15 (IN15) revealed that, under well-watered conditions, PtoCPK3-OE lines exhibited higher vessel density and smaller vessel lumen area compared with WT, whereas ptocpk3 -KD lines displayed the opposite phenotype ( Fig. 3, F to H ). Upon drought treatment, vessel density increased in all genotypes, with the highest increase in PtoCPK3 -OE lines (49.67%), followed by WT (25.39%) and ptocpk3 -KD (20.21%) ( Fig. 3G ). Meanwhile, vascular lumen area decreased in all genotypes under drought, with PtoCPK3-OE lines exhibiting significantly greater reductions, 57.91% and 71.23%, compared with WT and ptocpk3 -KD, respectively ( Fig. 3H ). Furthermore, scanning electron microscopy (SEM) revealed that drought stress markedly altered secondary wall thickening and pit architecture in xylem vessels ( Fig. 4, A and B , and fig. S5, M and N). Notably, under drought stress, vessel wall thickness increased by 31.70% in PtoCPK3 -OE lines relative to WT, whereas ptocpk3 -KD lines exhibited a 9.95% decrease ( Fig. 4C ). In parallel, compared with WT, PtoCPK3 -OE lines displayed coordinated reductions in pit aperture diameter (22.67%) and pit membrane area (30.55%), whereas ptocpk3 -KD lines showed corresponding increases of 14.54% and 21.53%, respectively ( Fig. 4C and fig. S5N). Confocal Raman microspectroscopy (CRM) analysis revealed that drought stress increased lignin accumulation in all genotypes, with the strongest signal observed in PtoCPK3-OE and the weakest in ptocpk3 -KD lines ( Fig. 4, C and D ). These results were consistent with bulk lignin and metabolite content analyses ( Table 1 ). Under drought conditions, PtoCPK3 -OE lines exhibited a lower relative abundance of S-lignin units (including sinapyl alcohol and its precursor sinapic acid) but a higher abundance of G-lignin units (coniferyl alcohol and its precursor ferulic acid), whereas ptocpk3 -KD plants displayed the opposite trend relative to WT ( Fig. 4E ). Two-dimensional heteronuclear single-quantum coherence nuclear magnetic resonance (2D-HSQC NMR) further showed that the lignin S/G ratio in PtoCPK3 -OE plants was 14.35% lower than that in WT, while ptocpk3 -KD plants exhibited a 7.87% higher S/G ratio as soil RWC declined from 70% to 20% ( Fig. 4F and Table 1 ). Collectively, these findings demonstrate that PtoCPK3 enhances drought tolerance in P. tomentosa by promoting water retention, modifying xylem structure, and shifting lignin composition to favor G-lignin biosynthesis, thereby improving xylem integrity under water-deficit conditions. Fig. 4. Lignin deposition of PtoCPK3 transgenic poplars under drought stress. Open in a new tab ( A ) Scanning electron microscopy (SEM) of transverse sections from stem IN15 across all experimental poplars. Sections were observed at magnifications of ×1000 (scale bar, 30 μm) and ×3000 (scale bar, 5 μm). Ve, vessel. ( B ) SEM of longitudinal sections from stem IN15 across all experimental poplars. Sections were observed at magnifications ×3000 (scale bar, 10 μm) and ×10,000 (scale bar, 5 μm). Ve, vessel; PA, pit aperture. ( C ) Vessel cell wall thickness and pit aperture diameter were quantified accordingly ( n = 100). In each violin, the central black dashed line represents the median, while the upper and lower dashed lines indicate the 25th and 75th percentiles. Significant differences among groups were tested using one-way ANOVA followed by Tukey’s test: Different letters represent significant differences. Values represent the mean ± SD of three biological replicates. ( D ) Confocal Raman microscopy (CRM) images revealed lignin deposition in vessel cell walls of all experimental poplars, with arrows indicating the vessel cell corners analyzed for spectral differences shown in (E). ( E ) Stem IN15 was analyzed by integrating over the 1600 cm −1 band (1500 to 1700 cm −1 ). Differences in overall raw Raman signal intensity (CCD counts) reflect relative lignin abundance among poplar transgenic lines, which is consistent with enhanced lignin deposition and secondary wall thickening. ( F ) The lignin metabolites were identified using the ultrahigh-performance liquid chromatography–electrospray ionization–tandem mass spectrometry (UPLC-ESI-MS/MS) system. ( G ) 2D heteronuclear single-quantum coherence nuclear magnetic resonance (2D-HSQC-NMR) was used to detect S, G, and H lignin units in all experimental poplars. Table 1. Chemical composition of the poplar woods during well-watered and drought conditions. Values represent the mean ± SD of three biological replicates. Statistical comparisons were conducted separately for the “Total lignin” and “S/G” columns. Within each column, different superscript lowercase letters indicate statistically significant differences among the indicated material-treatment groups according to one-way ANOVA followed by Tukey’s test ( P < 0.05). Values sharing the same letter within the same column are not statistically different. See source data for this table in data S36. ASL, acid-soluble lignin (the measurements were performed three times, and the results are average values); AIL, acid-insoluble lignin (the measurements were performed three times, and the results are average values). Condition Materials ASL (%) AIL (%) Total lignin (%) S (%) G (%) H (%) PB (%) S/G Well-watered WT 1.94 ± 0.13 19.78 ± 1.50 21.72 ± 1.40 c 70.30 28.50 1.20 31.70 2.47 b PtoCPK3-OE 1.62 ± 0.12 23.87 ± 1.15 25.49 ± 1.21 b 67.80 30.40 1.80 34.20 2.23 d ptocpk3 -KD 3.07 ± 0.12 16.02 ± 1.32 19.09 ± 1.27 d 71.30 28.30 0.40 15.10 2.52 a Drought WT 1.98 ± 0.10 24.01 ± 1.38 25.99 ± 1.41 b 67.30 31.10 1.50 19.50 2.16 e PtoCPK3-OE 1.73 ± 0.09 27.43 ± 0.97 29.17 ± 0.95 a 64.10 34.70 1.10 24.50 1.85 f ptocpk3 -KD 2.96 ± 0.13 19.10 ± 0.86 22.06 ± 0.85 c 69.30 29.70 1.00 25.10 2.33 c Open in a new tab PtoCPK3-mediated interaction and phosphorylation of PtoERF72 at Ser 96 To elucidate the molecular cascade regulated by PtoCPK3, we conducted quantitative phosphoproteomic analysis on xylem segments from PtoCPK3 -OE lines under drought conditions. A total of 1052 PtoCPK3-induced phosphosites, corresponding to 744 phosphoproteins, were identified as differentially phosphoproteins (DPs) between WT and PtoCPK3 -OE xylem tissue [fold change (FC) > 1.50 or < 0.67, P < 0.05; fig. S6A and data S10). Among these, 648 phosphosites from 491 DPs were identified as drought-response phosphoproteins (DRPs) and consequently defined as PtoCPK3-induced DRPs (fig. S6B and data S11). Motif analysis identified a conserved CPK recognition motif […K/RxxS], highlighting the serine-preferential phosphorylation signature of PtoCPK3 ( Fig. 5A ). Notably, phosphorylation levels were elevated at residues of proteins involved in drought response and lignin biosynthesis, including SnRK2.4 (Ser 158 ), PPC4 (Ser 438 ), NCED3 (Ser 73 ), NST1 (Ser 183 ), and PAL1 (Thr 107 ) (FC > 1.50, P < 0.05; fig. S6C). Transcriptional analysis of 491 PtoCPK3-induced DRPs under drought stress revealed 86 proteins (141 phosphosites) with altered transcript abundance, including 11 transcription factors such as WRKY33, AREB1, NST1, ZIP1, and ERF72 (data S12). These findings underscore the pivotal role of PtoCPK3 in transcriptional regulation via specific phosphorylation of key transcription factors (TFs) under drought stress. Fig. 5. PtoCPK3 mediates the phosphorylation of Ser 96 of PtoERF72 depending on Ca 2+ . Open in a new tab ( A ) PtoCPK3-induced drought-responsive phosphopeptide motif analysis. ( B ) Domain analysis of PtoCPK3 showing the protein kinase (orange), calmodulin-like calcium-binding (green), and four EF-hand (gray) domains. ( C ) Subcellular localization of 35S: PtoCPK3-GFP, with 35S: GFP serving as the control. For colocalization analysis, PtoCPK3-GFP was coexpressed with the nuclear stain DAPI, the cytoplasmic marker HSBP-mCherry, and the plasma membrane marker PIP2A-mCherry. Scale bar, 100 μm. For Western blot analysis of subcellular fractionation, the 35S: PtoCPK3-GFP fusion protein was detected using an anti-GFP antibody, while specific antibodies against histone H3 (nuclear marker), GAPDH (cytoplasmic marker), and H + -ATPase (plasma membrane marker) were used as internal controls. ( D ) Overlapping targets of PtoCPK3-induced drought-responsive phosphorylated proteins and interacting proteins. ( E ) Different tissues’ relative expression levels of 7 candidate PtoCPK3 target proteins overlapping PtoCPK3-induced DRP and PtoCPK3-interacting proteins. ( F and G ) Yeast two-hybrid (Y2H) (F) and pull-down (G) assays of PtoCPK3-PtoERF72 protein interaction. ( H ) In vitro phosphorylation assay of PtoCPK3 and PtoERF72, with Ca 2+ concentrations of 1 to 15 mM. ( I ) LC-MS/MS identification of PtoERF72 Ser 96 phosphorylation by PtoCPK3. ( J ) In vitro phosphorylation assay after mutation of the phosphorylation site Ser 96 (wild-type, PtoERF72 S96 ) to Ala 96 (mutant-type, PtoERF72 A96 ), with a Ca 2+ concentration of 10 mM. In (H) and (J), Coomassie brilliant blue (CBB) staining was used as a control. The red asterisk indicates the target band. ( K ) Surface plasmon resonance (SPR) sensograms measuring interaction affinities of PtoCPK3 with PtoERF72 S96 and PtoERF72 A96 . Yellow lines represent the 1:1 binding model fit. The resultant kinetic parameters are summarized ( n = 3), and tested concentrations are noted in blue. RU, resonance units. ( L ) Biomolecular fluorescence complementation assay (BiFC) measured the protein interaction localization between PtoCPK3 and PtoERF72 S96 , or PtoERF72 A96 ; H2B-mCherry was coexpressed as the nuclear maker. Scale bar, 25 μm. Consistent with the prediction of transmembrane helices and domains of PtoCPK3 protein ( Fig. 5B and fig. S6D), subcellular localization analysis revealed that the PtoCPK3–green fluorescent protein (GFP) fusion protein localized predominantly to the cytoplasm and nucleus ( Fig. 5C ). This observation was further corroborated by immunoblot analysis of subcellular fractions, which detected the PtoCPK3-GFP fusion protein in the nuclear and cytoplasmic fractions but not in the plasma membrane fraction ( Fig. 5C ), suggesting that PtoCPK3 functions primarily in the nucleocytoplasmic compartments. A nuclear stem cDNA library of P. tomentosa “1316” was screened via yeast two-hybrid sequencing assay (Y2H-seq), identifying 93 PtoCPK3-interacting proteins (fig. S6E). Among these, seven candidates were also identified as PtoCPK3-induced DRPs: PtoSnRK1.1, PtoSKM1, PtoC/VIF1, PtoGRF8, PtoERF72, PtoZIP1, and PtoNAC028. Of these, PtoERF72 (Ptom.008G.001767, ethylene-responsive factor 72) showed high expression in vascular tissue ( Fig. 5, D and E , and data S13). These observations are consistent with our qRT-PCR and RNA in situ hybridization results, which revealed abundant PtoERF72 expression in the phloem and developing xylem (fig. S6, F and G). Meanwhile, the interaction between PtoCPK3 and PtoERF72 was validated by Y2H and pull-down assays ( Fig. 5, F and G ). In vitro kinase assays demonstrated that PtoCPK3 phosphorylates PtoERF72 in a Ca 2+ -dependent manner, with both substrate phosphorylation and autophosphorylation levels enhanced by increasing Ca 2+ concentrations ( Fig. 5H and fig. S6H). Consistently, drought-treated plants exhibited elevated Ca 2+ levels ( Fig. 3C ). Liquid chromatography–tandem mass spectrometry (LC-MS/MS) analysis identified Ser 96 as the primary phosphorylation site ( Fig. 5I ), which showed significantly higher phosphorylation levels of WT PtoERF72 S96 in PtoCPK3 -OE plants under drought stress (fig. S6I). Substitution of Ser 96 with alanine (Ala, mutant-type, PtoERF72 A96 ) nearly abolished PtoCPK3-mediated phosphorylation ( Fig. 5J ), as evidenced by a marked reduction in luciferase complementation imaging (LCI) signals (fig. S6, J and K). Surface plasmon resonance (SPR) analysis further revealed that PtoCPK3 interacted with PtoERF72 S96 ( K D = 614 nM) with a 13.49-fold higher affinity than PtoERF72 A96 ( K D = 8.28 μM; Fig. 5K ). Additionally, biomolecular fluorescence complementation assay (BiFC) showed that substitution of Ser 96 with Ala 96 abolished detectable nuclear colocalization between PtoCPK3 and PtoERF72 A96 , despite both proteins localizing to the nucleus ( Fig. 5L and fig. S6L). Collectively, these findings demonstrate that Ser 96 is indispensable for maintaining the nuclear interaction between PtoCPK3 and PtoERF72, which is a prerequisite for PtoCPK3-mediated phosphorylation. PtoCPK3 enhances the transcriptional activity of PtoERF72 via phosphorylation at Ser 96 Phylogenetic and alignment analyses confirmed high conservation between PtoERF72 and its “84K” orthologs (figs. S7 and S8A). To explore the regulatory role of PtoERF72 in xylem development and lignin biosynthesis, we identified 2089 PtoERF72-induced differentially expressed genes (DEGs) in transgenic poplar lines ( PtoERF72-OE and ptoerf72 -KD) ( Fig. 6A ; fig. S8, B and C; and data S14). RNA-sequencing (RNA-seq) analysis demonstrated that PtoERF72 positively regulated genes involved in phenylpropanoid biosynthesis, signaling pathways, and environmental adaptation, including LAC4, VND4, HB7, and RD21 (fig. S8D). This finding was consistent with enhanced lignin deposition fluorescence observed in PtoERF72 -transgenic poplars ( Fig. 6B and fig. S8E). Moreover, gas chromatography–MS (GC-MS) showed that the concentration of coniferyl alcohol was significantly higher in PtoERF72-OE plants compared to WT and ptoerf72 -KD lines ( Fig. 6C ). Putative PtoERF72-binding regions across the P. tomentosa genome were identified through DNA affinity purification sequencing (DAP-seq), yielding 483 candidate binding sites, of which 9.67% were localized within the 2 kb upstream promoter regions (fig. S8F and data S15). The core binding motif enriched in these regions was the GCC-box “GCCGCC” ( P = 1.2 × 10 −810 ) ( Fig. 6D ). Functional analysis of these high-confidence PtoERF72 target genes showed substantial enrichment in biological processes such as osmotic stress, defense response, and cell differentiation, in addition to phenylalanine pathways (fig. S9, A and B). Fig. 6. PtoCPK3 mediates transcriptional regulation of PtoWOX13b and PtoUGT72AZ2 by PtoERF72 via phosphorylation at Ser 96 . Open in a new tab ( A ) Phenotypes of 4-month-old WT, PtoERF72-OE11 , PtoERF72-OE12 , ptoerf72- KD4, and ptoerf72- KD9 plants. Scale bar, 15 cm. ( B ) Lignin UV fluorescence emission in the stem IN15 of WT, PtoERF72-OE , and ptoerf72 -KD poplars, observed under different magnifications: 500 μm (×4), 250 μm (×40), and 50 μm (×100). ( C ) Coniferyl alcohol content of WT, PtoERF72-OE, and ptoerf72 -KD poplars detected by gas chromatography–mass spectrometry (GC-MS). ( D ) The enriched motifs in the PtoERF72-binding peaks, with significance represented by the P value calculated by MEME. A pie chart representing the PtoERF72-binding regions identified by DAP-seq across the genome. ( E and F ) Expression patterns of PtoWOX13b (E) and PtoUGT72AZ2 (F) in PtoERF72-OE and ptoerf72 -KD transgenic lines. ( G ) Dual-luciferase reporter assay (DLRA) was used to assess the transcriptional regulation of PtoERF72, as well as the coexpression of PtoCPK3 and PtoERF72, on the PtoUGT72AZ2 and PtoWOX13b promoters, and to analyze the LUC/REN ratio. Values represent the mean ± SD of three biological replicates. Significant differences among groups were tested using one-way ANOVA followed by Tukey’s test: Different letters represent significant differences. ( H ) Electrophoretic mobility shift assay (EMSA) evaluated the effect of PtoCPK3-mediated phosphorylation on the binding affinity of PtoERF72 S96 and PtoERF72 A96 for the promoters of PtoUGT72AZ2 and PtoWOX13b . The probe sequences, derived from PtoUGT72AZ2 and PtoWOX13b promoters, contained the GCC-box motif. “+” and “-” indicate the presence or absence of reagents during protein electrophoresis, respectively. The bound probe represents the binding affinity of PtoERF72 S96 and PtoERF72 A96 for the promoters of PtoUGT72AZ2 and PtoWOX13b , as well as the impact of PtoCPK3-GST protein (×50 and ×100) phosphorylation modification on transcriptional regulation. Particularly, integrated RNA-seq, DAP-seq, and GEA analyses identified two downstream candidate DEGs, Ptom.007G.00290 (glycosyltransferase, PtoUGT72AZ2 ) and Ptom.005G.02376 (WUSCHEL-related homeobox 13, PtoWOX13b ), that were strongly associated with precipitation (fig. S9, C and D). In PtoERF72-OE plants, the expression levels of PtoUGT72AZ2 and PtoWOX13b were inversely regulated, whereas the opposite pattern was observed in ptoerf72 -KD plants ( Fig. 6, E and F ). Meanwhile, qRT-PCR and RNA in situ hybridization results showed that PtoWOX13b and PtoUGT72AZ2 were universally expressed across vascular tissue (fig. S9, E to H). Yeast one-hybrid (Y1H) and dual-luciferase reporter assay (DLRA) further confirmed that PtoERF72 directly binds to and represses the PtoUGT72AZ2 promoter while activating the PtoWOX13b promoter ( Fig. 6G and fig. S9, I to K). These regulatory effects were enhanced by coexpression with PtoCPK3 ( Fig. 6G ). To determine whether PtoCPK3-mediated phosphorylation at Ser 96 is required for PtoERF72 transcriptional activity, electrophoretic mobility shift assays (EMSAs) revealed that PtoERF72 phosphorylated by PtoCPK3 exhibited substantially enhanced DNA binding activity to the PtoUGT72AZ2 and PtoWOX13b promoters compared to the PtoERF72 A96 mutant ( Fig. 6H ). Consistently, DLRA demonstrated that the Ser 96 -to-Ala 96 mutation reduced PtoERF72 binding to the PtoUGT72AZ2 and PtoWOX13b promoters by 37.85% and 52.85%, respectively (fig. S9, J and K). These results indicate that PtoCPK3-mediated phosphorylation at Ser 96 enhances PtoERF72 target binding and transcriptional regulation. Synergistic regulation of PtoUGT72AZ2 by PtoWOX13b and PtoERF72 on lignin G-monomer biosynthesis in poplar To investigate the enzymatic function of PtoUGT72AZ2 in lignin-related metabolism, in vitro assays coupled with high-performance liquid chromatography (HPLC)–MS/MS analysis demonstrated that PtoUGT72AZ2 catalyzes the glycosylation of ferulic acid and coniferyl alcohol using uridine diphosphate (UDP)–glucose as the donor substrate, yielding ferulic acid glucoside and coniferyl glucoside, respectively ( Fig. 7, A and B ). Phylogenetic and alignment analyses confirmed high conservation between PtoUGT72AZ2 and its “84K” orthologs (figs. S10 and S11A). In vivo, the concentrations of ferulic acid and coniferyl alcohol in PtoUGT72AZ2 -OE lines were decreased by 13.88% and 11.61%, respectively, whereas ptougt72az2 -KD lines exhibited increases of 25.47% and 13.23%, respectively, compared with WT ( Fig. 7C and fig. S11, B and C). However, no significant differences in total lignin content (TLC) were detected among the different poplar genotypes ( Fig. 7C ). Additionally, the expression levels of genes involved in the biosynthesis of ferulic acid and coniferyl alcohol, such as Pto4CL , PtoCCR , PtoCAD , and PtoLAC , were down-regulated in PtoUGT72AZ2 -OE lines, while these genes were up-regulated in ptougt72az2 -KD lines (fig. S11D). These results suggest that PtoUGT72AZ2 promotes a metabolic shift toward glycosylation and sequestration of lignin intermediates, which is consistent with the reduced accumulation of monolignols. Fig. 7. PtoWOX13b negatively regulates the expression of PtoUGT72AZ2 . Open in a new tab ( A and B ) HPLC-MS/MS identification of PtoUGT72AZ2 enzymatic products from ferulic acid and coniferyl alcohol in vitro. ( C ) Measurement of ferulic acid, coniferyl alcohol, and lignin content in PtoUGT72AZ2 transgenic poplars. Values represent the mean ± SD of three biological replicates. Significant differences among groups were tested using one-way ANOVA followed by Tukey’s test: Different letters represent significant differences. ( D ) Correlation analysis of the expression patterns of PtoWOX13b and PtoUGT72AZ2 in the mature xylem of 302 P. tomentosa accessions. ( E ) DLRA showing the repression of the PtoUGT72AZ2 promoter activity (LUC/REN ratio) by PtoWOX13b. Values represent the mean ± SD of three biological replicates. Significant differences were determined using a t test, * P < 0.05, ** P < 0.01, *** P < 0.001. ( F ) 2D-HSQC NMR spectra detected S, G, and H lignin units in stem IN15 of 3-month-old WT, PtoWOX13b -OE, and ptowox13b -KD poplars. ( G ) EMSA to assess the binding ability of PtoWOX13b to the common WBE motif (CAATCAA/C). “+” and “-” indicate the presence or absence of reagents during protein electrophoresis, respectively. Biotin-labeled probes and unlabeled competitor probes (×50 and ×100) were used to evaluate binding affinity. ( H ) Yeast one-hybrid (Y1H) verified direct binding of PtoWOX13b protein to the PtoUGT72AZ2 promoter and genes involved in vessel development. pGADT7-p53 + p53-pABAi was used as a positive control, and pGADT7 + pABAi was used as a negative control. The transcription level of PtoWOX13b was negatively correlated with PtoUGT72AZ2 in mature xylem ( r = −0.6317; Fig. 7D ). Moreover, a WOX-binding cis-acting element (WBE; CAATCAA/C) was identified in the promoter of PtoUGT72AZ2 ( Fig. 7E ), which is also a target of PtoERF72 ( Fig. 6, G and H ). Assays confirmed that PtoWOX13b binds to the promoter of PtoUGT72AZ2 and represses its expression ( Fig. 7, E, G, and H , and fig. S11E). Given the high conservation between PtoWOX13b and its “84K” orthologs (figs. S12 and S13A), we further investigated its function in “84K” transgenic lines. The results showed that the lignin S/G ratio of PtoWOX13b -OE poplars was significantly decreased by 4.89%, whereas ptowox13b -KD poplars showed an 11.56% increase, compared to WT ( Fig. 7F and fig. S13, B and C). Safranin O–Fast Green staining of the IN15 revealed no significant difference width among WT, PtoWOX13b -OE, and ptowox13b -KD plants (fig. S13, D and E). In contrast, the number of xylem layers was significantly altered: PtoWOX13b -OE plants showed an approximately 20.44% increase, whereas ptowox13b -KD poplars exhibited a 34.31% decrease compared with WT (fig. S13F). SEM analysis further demonstrated that PtoWOX13b overexpression led to thicker cell walls in xylem vessels and fibers, accompanied by increased vessel density and reduced vessel and fiber cell areas (fig. S13, G to I). Previous studies have highlighted the pivotal roles of ERF15 , MYC2b , NAC005 , NAC006 , NAC007 , NAC118 , NAC120 , and LAC14 in conferring drought tolerance and regulating vessel development in Populus ( 29 – 31 ). Promoter sequence analysis revealed one to three tandem repeats of the common WOX-binding element (CAATCAA/C) within the promoters of PtoERF15 , PtoNAC005 , PtoNAC006 , PtoNAC118 , PtoNAC120 , and PtoLAC14 , whereas no such motifs were detected in the promoters of PtoNAC007 or PtoMYC2b (fig. S13J). Subsequent EMSA and Y1H assays demonstrated that PtoWOX13b directly binds to the promoters of PtoERF15 , PtoNAC005 , PtoNAC006 , PtoNAC118 , PtoNAC120 , and PtoLAC14 ( Fig. 7, G and H ). Collectively, PtoWOX13b and PtoERF72 cooperatively mediate the negative regulation via the GCC-box and WBE cis-elements within the PtoUGT72AZ2 promoter, thereby attenuating the glycosylation activity of PtoUGT72AZ2, impairing the biosynthesis of lignin G-monomer, and ultimately modifying xylem structure. Natural variations in PtoCPK3-ERF72 underpinning drought adaptation divergence To dissect causal variants underlying lignin biosynthesis, mGWAS analysis identified a synonymous SNP226 (Chr3_6384016) in PtoCPK3 that was significantly associated with ferulic acid and coniferyl alcohol concentrations (data S9). In a natural population of 302 P. tomentosa individuals, the AA genotype exhibited markedly higher metabolite concentrations than the CC genotype of PtoCPK3 (fig. S14A). SNP226 was in complete linkage disequilibrium (LD) with InDel217, a 1-bp deletion at the 23rd nucleotide of the coding sequence (CDS), which caused a frameshift and premature termination ( Fig. 8, A and B , and fig. S14B). InDel217 defined two alleles, PtoCPK3 II (WT, functional) and PtoCPK3 DD (mutant, nonfunctional). The PtoCPK3 II genotype occurred in 53.97% of individuals, while PtoCPK3 DD was present in 34.11% (excluding heterozygotes). Poplars carrying PtoCPK3 II showed a 16.00% higher LC than those carrying PtoCPK3 DD (fig. S14C). Genotypic distribution analysis revealed that PtoCPK3 DD was enriched in the warm, humid S region, whereas PtoCPK3 II predominated in the colder and drier NE and NW regions (fig. S15, A and B). Functional assays confirmed that PtoCPK3 DD was restricted to the cytoplasm and failed to interact with PtoERF72, displaying a complete loss of kinase activity ( Fig. 8, C and E ). Fig. 8. PtoCPK3 - PtoERF72 allelic module analysis of drought tolerance in P. tomentosa populations. Open in a new tab ( A ) SNP226 and InDel217 loci linkage disequilibrium (LD) pattern of PtoCPK3 . ( B ) Protein sequence alignments of PtoCPK3 II and PtoCPK3 DD , where the PtoCPK3 DD gene exhibits a frameshift mutation leading to a premature translation termination. ( C ) Subcellular localization of 35S: PtoCPK3 DD -GFP, with 35S: GFP serving as the control. For colocalization analysis, 35S: PtoCPK3 DD -GFP was coexpressed with the nuclear stain DAPI and the cytoplasmic marker HSBP-mCherry. Scale bar, 100 μm. For Western blot analysis of subcellular fractionation, the 35S :PtoCPK3 DD -GFP fusion protein was detected using an anti-GFP antibody, while specific antibodies against histone H3 (nuclear marker) and GAPDH (cytoplasmic marker) were used as internal controls. ( D ) Kinase activity analysis of different PtoCPK3 genotypes and different PtoERF72 genotypes. ( E ) Y2H assays of different PtoCPK3 genotypes and different PtoERF72 genotypes. ( F ) EMSA to evaluate the effect of phosphorylation modification on the binding affinity of different PtoERF72 genotypes to the promoters of PtoUGT72AZ2 and PtoWOX13b . “+” and “−” indicate the presence or absence of proteins during electrophoresis, respectively. ( G ) Geographic distribution of PtoCPK3 _ PtoERF72 genotype combinations along the aridity index gradient, where the pie sections represent the proportion of genotypes, and the pie area is proportional to the number of accessions at each sampling site. ( H ) Evaluation of drought tolerance in the different PtoCPK3 _ PtoERF72 AA genotype populations. TLC, total lignin content; VCD, vessel cell density; SS, soluble sugar; SOD, superoxide dismutase; PRO, proline; MDA, malondialdehyde; ABA, abscisic acid; REL, relative electrolyte leakage; WUE, water use efficiency; VLA, vessel lumen area; Gs, stomatal conductance; Tr, transpiration rate; Pn, net photosynthetic rate. Values represent the mean ± SD of three biological replicates. Significant differences among groups were tested using one-way ANOVA followed by Tukey’s test: Different letters represent significant differences. Furthermore, to identify functional variants in PtoERF72 , a 4.5-kb genomic region was sequenced, revealing 96 SNPs and 65 indels, including five nonsynonymous SNPs (fig. S15C). Among these, Chr8_13497045 (SNP605, A-C, Lys 126 -Thr 126 ) and Chr8_13497213 (SNP773, A-G, Asn 182 -Ser 182 ) were significantly associated with lignin-related metabolites (data S9). Individuals carrying PtoERF72 CC (SNP605) and PtoERF72 GG (SNP773) exhibited 26.83% and 45.27% higher ferulic acid and coniferyl alcohol concentrations, respectively, compared with the corresponding PtoERF72 AA (fig. S15, D and E). However, the indistinct geographic distribution of PtoERF72 suggests that it is less influenced by environmental selection (fig. S15, F and G), supporting the dominant role of PtoCPK3 in drought adaptation. PtoCPK3 II , localized to both the cytoplasm and nucleus, interacted with and phosphorylated PtoERF72 AA (Lys 126 ) and PtoERF72 CC (Thr 126 ) in a Ca 2+ -dependent manner, exhibiting higher efficiency toward PtoERF72 CC owing to an additional phosphorylation site (Thr 126 ) introduced by SNP605 ( Fig. 8, D and E , and fig. S15, H and I). Mutation of Ser 96 to Ala 96 abolished phosphorylation ( Fig. 8D ), and Y2H assays confirmed Ser as essential for the PtoCPK3 II _PtoERF72 AA/CC interaction ( Fig. 8E ). Substitution of Ser and Thr with aspartate (Asp, D) to mimic constitutive phosphorylation enhanced binding of PtoERF72(D 96 D 126 ) CC to downstream targets PtoUGT72AZ2 and PtoWOX13b ( Fig. 8F ). However, the gained Thr 126 site in PtoERF72 CC improved phosphorylation efficiency, but it failed to compensate for the functional loss of Ser 96 ( Fig. 8, D and E ). Geographical distribution analysis of combined PtoCPK3 - PtoERF72 genotypes revealed that the drought-sensitive genotype PtoCPK3 DD _PtoERF72 AA was predominantly distributed in the warm and humid S region ( n = 73). In contrast, the moderately drought-tolerant PtoCPK3 II _ PtoERF72 AA and highly tolerant PtoCPK3 II _ PtoERF72 CC genotypes were mainly found in the colder and drier NE and NW regions ( n = 143) ( Fig. 8G and fig. S15J). To further validate drought tolerance, 15 individuals from each genotypic group were randomly selected from the S, NE, and NW regions. Under drought stress, PtoCPK3 II _PtoERF72 CC individuals exhibited significantly higher water use efficiency (WUE), ABA content, PRO accumulation, LC, and vessel cell density (VCD) than did PtoCPK3 II _ PtoERF72 AA and PtoCPK3 DD _ PtoERF72 AA individuals ( Fig. 8H ). Conversely, PtoCPK3 II _ PtoERF72 CC individuals showed reduced stomatal conductance (Gs), photosynthetic rate (Pn), and vessel lumen area (VLA) ( Fig. 8H ). Expression analysis of 45 individuals confirmed that PtoWOX13b was up-regulated, whereas PtoUGT72AZ2 was down-regulated in the highly drought-tolerant genotype, consistent with the enhanced transcriptional activity of phosphorylated PtoERF72 CC (fig. S16, A and B). Collectively, these findings identify PtoCPK3 II _ PtoERF72 CC as a key drought-resilient allelic module that enhances lignin biosynthesis and drought adaptation in P. tomentosa , providing a valuable genetic target for the molecular breeding of drought-tolerant cultivars ( Fig. 9 ). Fig. 9. Proposed model of PtoCPK3 -mediated drought adaptation in P. tomentosa. Open in a new tab Under drought stress, PtoCPK3 is activated by Ca 2+ and translocates to the nucleus, where it phosphorylates PtoERF72 at the Ser 96 residue, essential for nuclear interaction, thereby enhancing its transcriptional regulatory ability. Subsequently, PtoERF72 binds to the promoters (GCC-box) of target genes, activating the xylem developmental regulator ( PtoWOX13b ), while suppressing the G-lignin monomer glycosylation gene ( PtoUGT72AZ2 ). In parallel, PtoWOX13b further reinforces PtoUGT72AZ2 repression through the WBE motif, synergistically reducing the lignin S/G ratio and altering vessel structure to enhance drought tolerance in P. tomentosa . In the northern region, natural variations within the PtoCPK3 (1-bp insertion, PtoCPK3 II ) and PtoERF72 (a nonsynonymous substitution introducing an additional phosphorylation site, Thr 126 ) lead to differential phosphorylation efficiency and transcriptional binding ability, resulting in higher drought tolerance in PtoCPK3 II _PtoERF72 CC genotypes compared with PtoCPK3 II _PtoERF72 AA genotypes. However, PtoCPK3 (1-bp deletion, PtoCPK3 DD ) completely loses kinase activity and phosphorylation capacity, rendering the PtoCPK3 DD _PtoERF72 AA genotypes prevalent in the southern region highly susceptible to drought stress. DISCUSSION Plant drought adaptation involves multilayered regulation networks. Among these, CPKs play a pivotal role by integrating Ca 2+ sensing with kinase activity, functioning as direct “sensor-responders” that transduce intracellular oscillations into specific phosphorylation events. In Arabidopsis , CPK3 acts as a key positive regulator of ABA- and Ca 2+ -dependent stomatal closure by phosphorylating target proteins such as TPK1 (Ser 42 ) and SnRK2s (Ser 175 ), and its loss in the cpk3cpk6 double mutant severely impairs stomatal closure and drought survival ( 32 , 33 ). These findings establish CPK3 as a central component of guard cell signaling ( 34 , 35 ), yet stomatal regulation constitutes only one aspect of the multifaceted drought response. In woody plants, long-term survival relies less on transient stomatal movements and more on sustaining hydraulic stability over time. Here, we address this gap using P. tomentosa as a model, demonstrating that PtoCPK3 enhances drought tolerance by modulating xylem remodeling and lignin metabolism, which are shaped by both intrinsic molecular regulation and environmentally driven selection across natural populations. PtoCPK3 enhances drought adaptation in poplar by regulating xylem structure and lignin metabolism Here, overexpression of PtoCPK3 resulted in a coordinated remodeling of xylem hydraulic architecture, characterized by narrower and more numerous vessels together with reduced pit aperture diameter and pit membrane area under drought stress, traits that mitigate xylem embolism and maintain stem water potential under drought conditions ( Figs. 3 and 4 ). This integrated anatomical pattern aligns with the well-established hydraulic safety-efficiency trade-off observed in woody species, whereby both vessel- and pit-level traits shift toward enhanced resistance to cavitation and embolism spread ( 8 , 36 , 37 ). Conversely, ptocpk3 -KD lines developed wider and less dense vessels, impairing drought resistance ( Figs. 3 and 4 ). This effect appears to be mediated by PtoWOX13b , which is an indirect downstream target of PtoCPK3. Overexpression of PtoWOX13b in P. tomentosa promoted the development of smaller and more numerous vessels (fig. S13), consistent with the conserved function of PtrWOX13 orthologs in P. trichocarpa . For instance, the CLE41-PXY-WOX4 signaling cascade and WOX13a-VCS2-GCN5.1-ADA2b.3-WOX4a complex coordinate genetic and epigenetic regulation to maintain vascular cambium activity during wood formation in poplar ( 38 , 39 ). Specifically, PtrVCS2 overexpression increases vascular cell number, reduces xylem embolism, and enhances drought tolerance in poplar by improving water transport efficiency ( 40 ). Furthermore, fusion of PtrWOX13a or PtrWOX13c with the dominant ethylene responsive element binding factor-associated amphiphilic repression (EAR) repression domain markedly reduced the number of secondary xylem cells across the wood ring ( 41 ). Strong expression of PtrWOX13a/b/c in differentiating xylem cells supports our finding that PtoWOX13b overexpression increased xylem layers (fig. S13). In addition, overexpression of PtoCPK3 resulted in increased LC and thicker SCWs, providing a clear structural foundation for enhanced drought resistance ( Fig. 4 ). Such anatomical reinforcement represents a hallmark of drought adaptation in woody plants, as lignin deposition enhances mechanical strength, prevents xylem collapse, and maintains hydraulic conductivity under water deficit. Drought stress commonly redirects the phenylpropanoid metabolic flux toward lignin for vascular reinforcement and toward flavonoids for antioxidant defense ( 42 ). However, this metabolic balance in P. tomentosa “1316” shifted markedly toward lignification, with most flavonoids (79.10%) down-regulated and lignin metabolites accumulating, indicating a preferential carbon allocation to structural rather than chemical defense ( Fig. 1 ). This lignin-biased metabolic adjustment aligns with recent findings in tea ( 43 ). At the population scale, gradients of precipitation and aridity shaped adaptive divergence in both drought-associated SNPs and lignin metabolites, and mGWAS linked PtoCPK3 variants to ferulic acid and coniferyl alcohol contents ( Fig. 2 and fig. S3). Within this broader metabolic context, PtoCPK3 acts as a central regulatory hub that directs metabolic flux toward lignin biosynthesis, particularly enhancing the accumulation of ferulic acid and coniferyl alcohol under drought stress ( Fig. 4 ). PtoCPK3-OE lines increased coniferyl alcohol content and reduced the lignin S/G ratio ( Fig. 4 ), a modification known to substantially affect cell wall hydrophobicity and mechanical rigidity ( 44 – 46 ). Monolignol glucosylation, which modulates the lignification process, has been widely investigated in relation to the UGT72 family of glucosyltransferases. These enzymes catalyze the glycosylation of lignin monomers and their precursors in Arabidopsis and Populus ( 47 , 48 ). Unlike AtUGT72E1 to AtUGT72E3 (homologs of PtoUGT72AZ2), which glycosylate ferulic acid and sinapyl alcohol in Arabidopsis ( 49 ), PtoUGT72AZ2 catalyzes the glycosylation of coniferyl alcohol to produce coniferyl glucoside in P. tomentosa ( Fig. 7 ). Notably, the down-regulation of PtoUGT72AZ2 observed under drought stress reflects a survival-first strategy ( 50 – 51 ). While coniferin serves as a stable transportable storage form of lignin precursors for future growth, its accumulation represents a metabolic luxury during severe water deficit. By repressing PtoUGT72AZ2 , P. tomentosa effectively unlocks the pool of monolignols, redirecting the carbon flux away from metabolic storage and toward immediate structural lignification ( Fig. 1 ). Such a strategy enhances xylem mechanical stability, increases the availability of G-type precursors, and alters the lignin S/G ratio without substantially affecting TLC ( Fig. 7 ), consistent with previous observations ( 52 ). PtoCPK3-PtoERF72 cascade coordinates transcriptional regulation of PtoWOX13b and PtoUGT72AZ2 promoters Protein phosphorylation is one of the most ubiquitous posttranslational modifications in plants, enabling the rapid and reversible regulation of protein activity, subcellular localization, and intermolecular interactions during stress responses ( Fig. 1 ) ( 53 , 54 ). Y2H-seq and phosphoproteomic analyses identified PtoERF72 as a key phosphorylation target of PtoCPK3 ( Fig. 5 ). Subcellular localization analysis revealed that PtoCPK3 localizes to both the cytoplasm and the nucleus ( 55 ), a dual distribution also observed for other CPKs and regarded as essential for coupling Ca 2+ oscillations with nuclear signaling ( 16 ). Further assays demonstrated that PtoCPK3 physically interacts with PtoERF72 at Ser 96 in a Ca 2+ -dependent manner, and substitution of this residue completely abolished their nuclear colocalization and phosphorylation ( Fig. 6 ). These phosphorylation-dependent nuclear interactions are consistent with observations in other plant systems, where phosphosite mutations similarly disrupt nuclear complex formation, such as SnRK2.6-ABF2 (Ser 126 ) and MPK6-WRKY53 (Ser 229 ) in Arabidopsis , and SAPK2-bZIP46 (Thr 180 ) in rice ( 56 , 57 , 58 ). Functionally, phosphorylation of PtoERF72 at Ser 96 by PtoCPK3 enhances its transcriptional activity, establishing a regulatory cascade that controls two key downstream target genes: PtoWOX13b (xylem development) and PtoUGT72AZ2 (lignin metabolism). This dual regulatory mode—activation of one target and repression of another—reflects the well-established bidirectional transcriptional activity of ethylene response factor (ERF) family members. For example, ERF72 acts as both a transcriptional activator and repressor, and it suppresses SUS1 and HA2 to regulate iron uptake and carbon allocation ( 59 , 60 ), while activating stress-responsive genes such as RD29A , PAD3 , and CYP71A13 ( 61 ). Similarly, our RNA-seq and DAP-seq analyses revealed that PtoERF72 binds the GCC-box motif in target promoters. DLRA demonstrated that PtoERF72 represses PtoUGT72AZ2 and activates PtoWOX13b , and the regulatory effects were further enhanced by PtoCPK3 coexpression, while mutation of Ser 96 markedly reduced DNA binding activity to both promoters ( Fig. 6 ). This phosphorylation-dependent enhancement of transcriptional activity of PtoERF72 parallels observations in other systems, such as WRKY33 and ERF72, where MPK3/6-mediated phosphorylation at Ser 220 and Ser 151 , respectively, enhanced transactivation of downstream target genes ( 61 , 62 ). PtoWOX13b functions as a context-dependent transcriptional regulator, reflecting the evolutionary versatility and functional diversification of the WOX13 family. In Arabidopsis , WOX13 acts as a repressor of shoot apical meristem regeneration by suppressing WUS and CLV3 while modulating cell wall–related genes such as EXP and MAN7 ( 63 , 64 ). By contrast, woody homologs have functionally specialized toward regulating vascular reinforcement. In P. trichocarpa , PtrWOX13c interacts with KNAT7 through their shared capacity to bind to a previously uncharacterized domain of unknown function (DUF3755) within PtrDRIF1 ( 65 – 67 ). Additionally, PtrWOX13c is transcriptionally induced by PtrWNDs and directly binds to the promoter of PtrPAL4 ( 67 – 69 ). Consistent with these findings, our results demonstrated that PtoWOX13b simultaneously activates PtoNAC005/006/118/120 and PtoLAC14 while repressing PtoUGT72AZ2 ( Fig. 7 ). Together, these findings illustrate that the PtoCPK3-PtoERF72-PtoWOX13b-PtoUGT72AZ2 module integrates Ca 2+ -dependent phosphorylation with transcriptional regulation in P. tomentosa ( Fig. 9 ). Natural variation of PtoCPK3-PtoERF72 underpins allelic divergence in drought adaptation Natural allelic variation constitutes the fundamental basis of local adaptation and serves as an essential genetic resource for molecular breeding ( 70 ). For example, natural allelic variation of ZmVPP1 and ZmAbh4 in maize modulates ABA signaling, while variation of Ghd7.1 in rice regulates phenology, collectively influencing WUE and drought tolerance ( 71 – 73 ). In P. tomentosa , natural allelic variation of PtoWRKY68 and PtoRSZ21 has been shown to affect drought responses by modulating stomatal morphology and ABA-responsive gene expression ( 74 , 75 ). Here, functional polymorphisms in PtoCPK3 and PtoERF72 were identified as key determinants of drought adaptation through combined GEA and mGWAS analyses in P. tomentosa ( Fig. 2 ). A frameshift allele of PtoCPK3 (PtoCPK3 DD ) abolished kinase activity and disrupted the nuclear interaction with PtoERF72 AA and PtoERF72 CC , whereas the functional allele (PtoCPK3 II ) retained phosphorylation capacity and promoted xylem reinforcement ( Fig. 8 ). Similarly, a nonsynonymous variant of PtoERF72 (PtoERF72 CC ) introduced an additional phosphorylation site at Thr 126 compared with the PtoERF72 AA allele, enhancing phosphorylation efficiency and strengthening downstream regulation of PtoWOX13b and PtoUGT72AZ2 ( Fig. 8 ). The combined genotype PtoCPK3 II _PtoERF72 CC conferred the greatest drought tolerance in natural populations, characterized by higher lignin accumulation, increased vessel density, and elevated WUE compared with other allelic combinations. Comparable mechanisms have been reported in other species, where allelic variation in kinases or their substrates reshapes phosphorylation dynamics and modulates stress responses. For example, a promoter deletion in ZmSnRK2.10 (Del-356) diminishes phosphorylation of ZmHAK4 at Ser 5 and consequently alters stress tolerance in maize ( 21 ). Geographic analysis of PtoCPK3-PtoERF72 genotypes revealed pronounced regional differentiation: Drought-tolerant genotypes were enriched in the arid NE and NW regions, whereas drought-sensitive alleles predominated in the humid S subpopulation. This geographic partitioning highlights a genomic signature of local adaptation, echoing precipitation-linked allele frequency shifts reported in P. trichocarpa ( Fig. 2 ) ( 76 , 77 ). Collectively, these findings demonstrate that allelic variation in natural tree populations underlies adaptive divergence along environmental gradients, reflecting the evolutionary imprint of local selection. In summary, based on integrative evidence from multi-omics analyses and genetic association approaches (GEA and mGWAS), we propose an allele-specific regulatory cascade comprising PtoCPK3-PtoERF72-PtoWOX13b-PtoUGT72AZ2 that governs lignin metabolism by modulating coniferyl alcohol content and xylem structure ( Fig. 9 ). In the NE/NW regions, PtoCPK3 II interacts with and phosphorylates PtoERF72 AA at Ser 96 , or PtoERF72 CC at both Ser 96 and Thr 126 in the nucleus, enhancing its transcriptional activation of PtoWOX13b and repression of PtoUGT72AZ2 . Notably, PtoWOX13b also represses PtoUGT72AZ2 expression, reducing its glycosylation of ferulic acid and coniferyl alcohol (G-monomer precursors). This redirection of phenolic precursors toward polymerization lowers the lignin S/G ratio and promotes vessel wall thickening. In contrast, in the S region, the PtoCPK3 DD protein lacks kinase activity and interaction capacity, leading to weakened lignification and drought sensitivity. Collectively, this mechanism provides a molecular framework for understanding drought-induced lignin plasticity and offers potential genetic targets for developing drought-resilient poplar cultivars. MATERIALS AND METHODS Drought tolerance assay Four-month-old soil-grown P. tomentosa “1316” seedlings were used as the experimental materials for drought treatment. Initially, sterile plantlets were subcultured and multiplied on 1/2 Murashige-Skoog basal medium for 2 months at 23° to 25°C, under a 16-hour/8-hour light/dark cycle with a light intensity of 50 μmol m −2 s −1 provided by cool white fluorescent tubes. Subsequently, uniform seedlings were transplanted into plastic pots filled with a mixture of perlite, vermiculite, and nutrient soil (1:1:3) and grown in a controlled greenhouse for an additional 4 months for acclimation and further development. The greenhouse environment was maintained at 25° to 28°C during the day and 20° to 22°C at night, with a 16-hour photoperiod, and a relative humidity of 60% ± 5%. Upon reaching 6 months of age, uniform seedlings (based on height and ground diameter) were selected for drought-stress treatments. Drought stress was imposed by withholding irrigation over a 15-day period, during which the soil RWC was allowed to decrease gradually from 70% ± 5% to 20% ± 5% to simulate progressive drought conditions in the field. Control poplars were maintained under well-watered conditions, with soil RWC at 70% ± 5%. Soil RWC was determined using the standard oven-drying method. Through this daily monitoring system, the water gradient fluctuation was strictly controlled within an error margin of ±5%. At the end of the 15-day treatment, tissues including leaves and stem segments were harvested for physiological measurements, ultrastructural observations, and molecular analyses. Multi-omics dataset analysis Xylem tissue collection and processing For 4-month-old soil-grown poplars, stem segments were collected from the basal and middle parts of the trunk. The bark (including phloem and cambium) was manually stripped to expose the woody core. Xylem tissues adjacent to the cambium were gently scraped using a sterile razor blade, immediately frozen in liquid nitrogen, and stored at −80°C until further use. RNA-seq analysis Transcriptomic profiling of xylem tissues from 4-month-old poplars under well-watered and drought conditions was performed by Novogene (Beijing, China). Briefly, mRNA was enriched from liquid nitrogen–ground samples using oligo(dT) beads and fragmented. First-strand cDNA was synthesized using random hexamers, followed by second-strand synthesis utilizing DNA polymerase I and ribonuclease (RNase) H. After purification with AMPure XP beads, double-stranded cDNA underwent end-repair, A-tailing, and adapter ligation. Final libraries were generated via size selection and PCR amplification, then sequenced on the DNBSEQ-T7 platform. RNA-seq data were normalized as fragments per kilobase of transcript per million mapped reads, and DEGs were identified using Cuffdiff ( P < 0.05, |FC| > 2). PtoCPK3 -OE and ptocpk3 -KD lines were subjected to drought treatment for 15 days as described above, and phenotypic detection was conducted. Proteomic and phosphoproteomic analyses Proteomic and phosphoproteomic analyses of xylem tissues from the same material and treatments were performed by Qinglian Bio (Beijing, China). For proteomic analysis, proteins were reduced with 5 mM dithiothreitol (DTT) (1 hour, 37°C) and alkylated with 10 mM iodoacetamide (45 min, room temperature). Following dilution, samples were digested overnight with trypsin at a 50:1 protein-to-enzyme ratio. Resulting peptides were desalted using C18 columns, eluted with 70% acetonitrile, and lyophilized. LC-MS/MS analysis was conducted on a timsTOF HT mass spectrometer equipped with a CaptiveSpray source. Chromatographic separation was performed at a flow rate of 500 nl/min. Data were acquired in data-dependent acquisition–parallel accumulation serial fragmentation (DDA-PASEF) mode with a mass range of 100 to 1700 mass/charge ratio ( m / z ). The ion mobility range was set to 0.6 to 1.6 V·s/cm 2 , with a total cycle time of 1.1 s containing 10 PASEF ramps. Finally, differentially expressed proteins were identified using FragPipe via BioLadder (bioladder.cn) ( P < 0.05, |FC| > 1.5) ( 78 ). Phosphoproteomic analysis followed a specific enrichment workflow. Proteins were reduced with 10 mM tris(2-carboxyethyl)phosphine (TCEP) and alkylated with 25 mM CAA before overnight trypsin digestion (50:1 ratio). After C18 desalting, phosphopeptides were enriched using TiO 2 beads and subsequently fractionated into three combined pools using a stepwise high-pH acetonitrile gradient (6% to 50%). Fractions were lyophilized and analyzed on a timsTOF HT mass spectrometer at a flow rate of 300 nl/min. Data were acquired in DDA-PASEF mode with a scan range of 100 to 1700 m / z . Specific acquisition settings included a customized ion mobility range of 0.7 to 1.4 V·s/cm 2 and a cycle time of 1.17 s containing 10 PASEF ramps. Phosphopeptide motif discovery was performed using Motif-X. Widely targeted metabolomic analysis Widely targeted metabolomic profiling was performed by Metware Biotechnology Co. Ltd. (Wuhan, China) using an ultrahigh-performance liquid chromatography–electrospray ionization–MS/MS (UPLC-ESI-MS/MS) system (Shimadzu UFLC coupled with an AB Sciex QTRAP 6500 system). Chromatographic separation utilized an Agilent SB-C18 column (1.8 μm, 2.1 mm by 100 mm) at 40°C. The mobile phase consisted of 0.1% formic acid in water (A) and acetonitrile (B), eluting with a gradient from 5% to 95% B over 9 min at a flow rate of 0.35 ml/min. MS detection was conducted on a triple quadrupole system operating in multiple reaction monitoring (MRM) mode with an ESI source temperature of 500°C. Metabolites were identified against the MWDB database (Metware) based on MS2 spectra and quantified via peak area integration using MultiQuant software, following the removal of isotopic and redundant fragment signals ( 79 ). Differential metabolites were defined based on variable importance in projection (VIP) ≥ 1, |Log2FC| ≥ 1, and P < 0.05. DPM analysis To integrate multi-omics datasets through gene P value and directional information, we performed DPM analysis ( 80 ). Initially, we integrated proteins and their corresponding phosphorylation sites to generate DRPS, thereby encompassing the effects of distinct phosphorylation sites within a single gene. Using the transcriptional and protein expression levels of the corresponding genes as representative of those of the DRPS, we retained 4441 DRPS present across all three datasets. Subsequently, based on the observed consistent variation trends across the three omics layers, we defined constraint directions as [1,1,1] and [−1,−1,−1], corresponding to DRPS that are consistently up-regulated or down-regulated across all three layers. We performed DPM analysis and nondirectional P value aggregation (Brown method) using the ActivePathways R package, utilizing the P values of the 4441 shared DRPS across the three omics datasets and their fold changes [fragments per kilobase of transcript per million mapped reads (FPKM)/protein abundance/phosphorylated proteins] under drought conditions relative to normal conditions, with all parameters set to defaults. For pathway enrichment analysis of the DRPS, we integrated the biological process (BP) entries from Gene Ontology (GO) annotations and the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway information for the genes. By mapping DRPS to their corresponding genes, we substituted the genes in the background gene set with the corresponding DRPS, thereby creating a custom background DRPS set. Pathway enrichment analysis was subsequently conducted using the “ActivePathways” function in the ActivePathways R package, with all parameters set to defaults. Association population and phenotypic data The association population comprised 1047 accessions established in a clonal plantation of P. tomentosa located in Guanxian County, Shandong Province, representing the species’ natural distribution range across China (30°N to 40°N, 105°E to 125°E) ( 81 ). From this population, 302 genetically distinct genotypes were selected for metabolomic profiling. Xylem tissues were sampled from all individuals between 8:00 a.m. and 11:00 a.m. to minimize diurnal fluctuations. For each accession, the bark was peeled, and the vascular cambium was carefully removed to expose the xylem. Sampling was performed in triplicate for each individual. Pure xylem tissues were collected by scraping the exposed surface with a steel file, immediately snap-frozen in liquid nitrogen, and stored at −80°C. Lignin metabolites were subsequently quantified by Metware Biotechnology Co. Ltd. (Wuhan, China) using a targeted metabolomics approach. Association study Genome resequencing of the association population and population genetic analyses The whole-genome resequencing dataset comprising 302 P. tomentosa accessions was obtained, and after quality filtering [minor allele frequency (MAF) > 0.05; missing rate < 0.2; heterozygosity < 0.9], 8,686,835 high-confidence SNPs were retained ( 82 , 83 ). Population structure was inferred using ADMIXTURE v1.3.0 using SNPs pruned for LD < 0.2. Cross-validation error indicated K = 3 as the optimal number of subpopulations, consistent with simple sequence repeat (SSR)-based clustering ( 81 ). Kinship was estimated using EMMAX with independent SNPs (LD < 0.2) ( 84 ). Identification of drought-associated genetic variants Environmental data (1970–2000) were compiled across the natural distribution range of P. tomentosa , including 20 precipitation-related and 11 temperature-related bioclimatic variables from WorldClim ( https://www.worldclim.org/data/bioclim.html ) ( 85 ), along with the aridity index from CGIAR-CSI ( https://cgiarcsi.community ) ( 86 ). The aridity index, calculated according to the FAO-56 Penman-Monteith equation, represents moisture availability, where smaller values indicate greater drought intensity. Environmental differences among subpopulations were evaluated using one-way ANOVA followed by least significant difference (LSD) post hoc tests ( P < 0.05). GEA analysis was conducted with the LFMM v1.1 linear mixed model ( 87 ), incorporating three latent factors to correct for population structure. Associations between 8,686,835 SNPs and 21 environmental variables were tested, with a significance threshold of P < 1 × 10 −5 . Geographical and environmental selection patterns of drought-adaptive SNPs F st was estimated for neutral SNPs (LD < 0.2, n = 980,165) and drought-adaptive SNPs among subpopulations (S versus NE, S versus NW, NE versus NW) using VCFtools v0.1.16, with differences evaluated by the Wilcoxon rank sum test ( 88 ). IBD and IBE patterns were examined using VCF2Dis v1.5.5 and partial Mantel tests with 999 permutations, implemented in the vegan R package ( 89 – 91 ). Environmental distance matrices were calculated from 20 precipitation-related variables and the aridity index using Euclidean distance via the vegdist function in vegan. Geographical distance matrices were computed based on the distVincentyEllipsoid model using the distm function. RDA was applied to partition the independent effects of precipitation and aridity to the distribution to adaptive SNPs, excluding temperature variables ( 92 , 93 ). To determine the variables that have a greater impact on drought-adaptive SNPs, a gradient forest model based on the relationship between environmental variables and drought-adaptive SNPs was constructed using the gradientForest package in R ( 94 ). Comparative analysis of Q st- F st of environmentally selected metabolites A Q st- F st comparison framework was used to identify metabolites subject to divergent selection ( 95 ). Q st was calculated as Q st = (1 + F IS ) σ2 B/[(1 + F IS ) σ2 B + 2σ2 w], where σ2 B and σ2 w denote between- and within-population variance components, respectively. F IS , the inbreeding coefficient, was estimated using GCTA ( 96 , 97 ). Confidence intervals (95%) for Q st were generated through 1000 bootstrap re-samplings ( 76 ). Metabolites whose Q st confidence intervals exceeded the neutral F st estimates were regarded as being under divergent selection. Neutral F st estimates among subpopulations were obtained using putatively neutral SNPs ( 82 ). Metabolite-based genome-wide association study mGWAS was performed with EMMAX, applying a linear mixed model to associate 8,686,835 high-quality SNPs with the concentrations of seven lignin metabolites ( l -phenylalanine, caffeic acid, ferulic acid, coniferyl alcohol, sinapaldehyde, sinapyl alcohol, and p -coumaraldehyde). Metabolite data were Z score normalized, and values with | Z | ≥ 3 were treated as outliers and replaced with NA. To control for population stratification and reduce false-positive associations, both population structure ( Q ) and kinship ( K ) were included as covariates in the model. The genome-wide significance threshold was determined using GEC v1.0 ( 98 ), yielding P < 3.68 × 10 −7 . Significant SNPs and putative candidate genes were identified within 2-kb flanking regions, and LD blocks surrounding significant SNPs were visualized using the R package LDheatmap ( 99 ). Phylogenetic analysis To construct the phylogenetic relationships of PtoCPK3 , PtoERF72 , PtoWOX13b , and PtoUGT72AZ2 , homologous protein sequences were identified using BLASTP v2.12.0 searches against the genomes of hybrid poplar “84K” ( Populus alba × Populus glandulosa ), P. trichocarpa , Populus euphratica , Populus lasiocarpa , Populus wilsonii , Populus adenopoda , Populus davidiana , and Salix purpurea . Homologs were filtered based on an E -value cutoff <1 × 10 −50 and a sequence identity threshold of 80%. Multiple sequence alignments were performed using the ClustalW algorithm implemented in MEGA v7.0.26 software. Maximum likelihood (ML) phylogenetic trees were constructed using IQ-TREE with the best-fit substitution models automatically selected by ModelFinder ( -m MFP ). Branch support values were assessed using the SH-aLRT test (100 replicates) and the Ultrafast Bootstrap approximation with 1000 replicates ( -bb 1000 ). The resulting phylogenetic trees were visualized and annotated using FigTree v1.4.4. Overexpression and RNAi vector construction and transformation of poplar Full-length CDS of PtoCPK3 ( PtoCPK3 II ), PtoERF72 [also named PtoERF72 AA or PtoERF72 S96 , or PtoERF72 ( S 96 ) AA ], PtoWOX13b , and PtoUGT72AZ2 without stop codons was amplified from a P. tomentosa “1316” clone and were then individually cloned into the pBI121-GFP vector at the Xba I restriction site under the control of the CaMV35S promoter to generate OE constructs. For KD constructs, 200-bp antisense fragments of PtoCPK3 , PtoERF72 , PtoWOX13b , and PtoUGT72AZ2 were also cloned into the pBI121 vector to induce posttranscriptional gene silencing through sequence-specific RNA interference, thereby reducing the accumulation of the corresponding endogenous transcripts. Vectors, clone sequences, and corresponding restriction sites for generating OE and KD of PtoCPK3 , PtoERF72 , PtoWOX13b , and PtoUGT72AZ2 poplar are listed in data S16. Constructs were subsequently transformed into poplar 84K ( P. alba × P. glandulosa ) via leaf disc–mediated transformation ( 100 ). The plants were maintained in a controlled growth chamber with an ambient temperature of 23° to 25°C, under a 16-hour/8-hour light/dark cycle with a light intensity of 50 μmol m −2 s −1 provided by cool white fluorescent tubes. 84K leaf discs were infected with a culture [optical density at 600 nm (OD 600 ) = 0.8] of Agrobacterium tumefaciens strain GV3101 harboring the construct. Post-infection, the leaf discs were transferred onto the shoot induction medium (SIM), comprising Murashige-Skoog basal medium supplemented with 6-benzyl amino purine (0.5 mg/liter) and naphthalene acetic acid (0.05 mg/liter), for a period of 5 days at 22 ± 2°C in the dark. The selection of transgenic poplar leaf discs occurred on SIM, supplemented with kanamycin (50 mg/liter) and timentin (300 mg/liter), under a 16-hour/8-hour light/dark regime. Following a month, individual regenerated shoots were detached and transferred to a root induction medium (RIM), consisting of 1/2 Murashige-Skoog medium supplemented with indole-3-butyric acid (0.05 mg/liter), naphthalene acetic acid (0.05 mg/liter), kanamycin (50 mg/liter), and timentin (300 mg/liter), for adventitious root induction. All plants underwent propagation via subculturing. Vectors, clone sequences, and corresponding restriction sites for generating OE and KD of PtoCPK3 , PtoERF72 , PtoWOX13b , and PtoUGT72AZ2 poplars are listed in data S16. Quantitative reverse transcription PCR Total RNA was extracted from transgenic lines, and RNA quality and integrity were assessed by agarose gel electrophoresis. First-strand cDNA was synthesized using HiScript IV All-in-one Ultra RT SuperMix for qPCR (#R433; Vazyme Biotech, Nanjing, China) according to the manufacturer’s instructions. The concentration of the resulting cDNA was quantified using a NanoDrop spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA) and subsequently used as a template for qRT-PCR analysis. Primers were designed using the RealTime qPCR Assay Entry tool ( https://sg.idtdna.com/Scitools/Applications/RealTimePCR/default.aspx ; data S16), with the poplar Actin gene (accession no. EF145577 ) serving as an internal control. All qRT-PCRs were performed in technical and biological triplicate, and primer specificity was validated by melting-curve analysis. Relative gene expression levels were calculated using the 2 −ΔΔCt method ( 101 ). Furthermore, to characterize the tissue-specific expression profiles of PtoCPK3 , PtoERF72 , PtoWOX13b, and PtoUGT72AZ2 , total RNA was isolated from the root, developing xylem, mature xylem, cambium, phloem, and leaf of P. tomentosa clone “1316.” The expression levels were then determined via qRT-PCR, following the protocols described above. Measurement of drought-related indicators WT, PtoCPK3-OE (L4 and L7), and ptocpk3 -KD (L2 and L3) poplars were subjected to drought treatment for 15 days, as described above. Basal leaves and stem tissues from well-watered and drought-treated poplars were harvested for subsequent physiological, biochemical, ionic, and hydraulic analyses. (i) Growth traits (plant height and stem diameter). Plant height was measured from the soil surface to the apical meristem, and basal stem diameter was measured at 1 cm above the soil surface using a digital caliper. (ii) Physiological indexes [stomatal conductance, chlorophyll, ABA, malondialdehyde (MDA), and PRO contents]. Stomatal conductance was measured using a LI-6400 system (LI-COR, Lincoln, NE, USA) ( 102 ). Measurements were conducted between 09:00 a.m. and 11:00 a.m. under ambient light and CO 2 conditions to minimize diurnal variation. Chlorophyll, ABA, MDA, and PRO contents were quantified using specific enzyme-linked immunosorbent assay (ELISA) kits, according to the manufacturer’s instructions. Briefly, 10 μl of the prepared sample and 40 μl of sample diluent were added to microtiter plates precoated with specific antibodies and incubated at 37°C for 30 min. After five wash cycles with a 30-fold diluted washing buffer, 50 μl of horseradish peroxidase (HRP)-conjugated reagent was added to each well (except blank wells) and incubated at 37°C for another 30 min. Following a second wash cycle, 50 μl each of chromogen solutions A and B were added for color development in the dark at 37°C for 10 min. The reaction was terminated by adding 50 μl of stop solution, causing a color shift from blue to yellow, and the absorbance was measured at 450 nm using a microplate reader within 15 min. Target concentrations or activities were calculated using linear regression equations derived from standard curves of purified standards and adjusted by the initial dilution factor. (iii) Ca 2+ content. Stem samples were oven-dried, ground to fine powder, and digested with nitric acid before analysis. Ca 2+ content was determined by inductively coupled plasma MS (ICP-MS) monitoring the 42 Ca isotope ( 103 ). (iv) Hydraulic properties (stem RWC, stem water potential, and xylem hydraulic conductivity). Stem RWC was determined using fresh weight (FW) and dry weight (DW) according to the following formula: Stem RWC (%) = FW/DW × 100. Stem water potential was measured with a Model 1515D Pressure Chamber (PMS Instruments, Albany, OR, USA). Stem segments were excised and immediately sealed in plastic bags to prevent dehydration. Samples were pressurized gradually until xylem sap appeared at the cut surface, and the balancing pressure was recorded as the stem water potential ( 104 , 105 ). Xylem hydraulic conductivity was assessed using the HPFM-Gen3 system (Dynamax, Austin, TX, USA) under constant pressure. Stem segments were flushed to remove potential embolisms before measurement. Hydraulic conductivity was calculated based on the steady-state flow rate, applied pressure gradient, and the cross-sectional area of the stem IN15. RNA in situ hybridization Gene-specific fragments were amplified and cloned into the pGEM-T Easy vector (Promega). Digoxigenin (DIG)-labeled sense and antisense RNA probes were synthesized by in vitro transcription using the DIG RNA Labeling Kit (Roche) according to the manufacturer’s instructions. Paraffin-embedded stem sections were deparaffinized in xylene (two changes, 15 min each) and rehydrated through a graded ethanol series (100%, 95%, 85%, and 75%, 5 min each), followed by rinsing in distilled water. For permeabilization, sections were treated with 20 μg ml −1 proteinase K in phosphate-buffered saline–Tween 20 (PBST) at 37°C for 20 min and washed three times in PBS. Pre-hybridization was performed at 37°C for 1 hour in hybridization buffer containing salmon sperm DNA (1:100). Hybridization was carried out overnight at 42°C with DIG-labeled RNA probes (approximately 1 μg per slide) in a humidified chamber sealed with rubber cement. After hybridization, sections were washed sequentially in 2×, 1×, and 0.2× SSC at 37°C. Slides were then blocked and incubated with alkaline phosphatase–conjugated anti-DIG antibody at 37°C for 1 hour. Signal detection was performed using bromochloroindolyl phosphate–nitro blue tetrazolium (BCIP-NBT) substrate until blue or blue-violet staining was observed. Sense probes were used as negative controls. Slides were dehydrated through an ethanol series, cleared in xylene, and mounted with neutral balsam. Images were captured using a Nikon light microscope under bright-field illumination. The probe sequences are listed in data S16. Determination and distribution of chemical composition in stems Cell wall composition analysis The chemical composition of cell walls from 4-month-old poplar stems was analyzed according to modified National Renewable Energy Laboratory (NREL) protocols ( 106 ). Stem samples were harvested, oven-dried to constant weight, and ground into fine powder before analysis. For removal of soluble compounds, 100 mg of stem powder was extracted with 80% (v/v) methanol under continuous shaking, followed by centrifugation. The insoluble residue was collected, air-dried, and used for subsequent lignin analysis. Cell wall polymers were hydrolyzed using a two-step sulfuric acid hydrolysis procedure. Briefly, the residue was incubated with 72% (w/w) H 2 SO 4 at 30°C for 1 hour with occasional mixing to allow complete impregnation. The reaction mixture was then diluted to a final sulfuric acid concentration of 4% (w/w) and autoclaved at 121°C for 1 hour to complete hydrolysis of structural carbohydrates. After hydrolysis, the reaction mixture was filtered to separate the acid-insoluble residue from the hydrolysate. Acid-soluble lignin (ASL) content was quantified by measuring absorbance at 205 nm using a UV-visible spectrophotometer, with appropriate acid blanks used for baseline correction. Acid-insoluble lignin (AIL) content was determined using the Klason method ( 107 ). TLC was calculated as the sum of ASL and AIL content. All assays were performed in triplicate, and mean values were calculated. Raman analysis Raman spectra and chemical mapping were acquired using an inVia-Reflex 532/XYZ confocal Raman microscope (Renishaw, Great Britain). The stem IN15 (10 μm thick) of well-watered and drought-treated poplars (4-month-old WT, PtoCPK3-OE , and ptocpk3 -KD) were prepared with a Leica SM2010 R sliding microtome (Leica, Wetzlar, Germany). Cross-sections were illuminated with 532-nm excitation laser, and Raman spectra were collected using a 50× objective lens. The laser power incident on the sample surface was maintained at approximately 5 mW. Raman scattering signals were detected with an air-cooled charge-coupled device (CCD) detector, and the spectrometer center was set to 1600 cm −1 , corresponding to the aromatic ring stretching vibration of lignin. Spectra were acquired with an integration time of 1 s per scan. For Raman mapping, a spatial step size of 5 μm was applied across a 10 × 10 μm region, with each pixel representing one measurement point. Measurements were performed on vessel SCW regions at comparable developmental positions within the xylem. The total Raman intensity of 1600 cm −1 band was used as a semiquantitative measure of lignin concentration. No normalization was performed between poplar transgenic lines to maintain the inherent differences in Raman scattering cross-sections as a function of lignin density. NMR analysis Double enzymatic lignin (DEL) was isolated from stems of 4-month-old poplars (WT, PtoCPK3 -OE, and ptocpk3 -KD) cultivated under well-watered and drought conditions. Stem samples were oven-dried at 60°C, ground to 20-40 mesh, and ball-milled at 500 rpm for 5 hours. The resulting powder was suspended in acetate buffer (pH 4.8) containing cellulase [2.5 ml, 100 filter paper units (FPU)/ml] and incubated at 50°C with shaking at 150 rpm for 48 hours. After centrifugation, washing, and freeze-drying, the residue was subjected to an additional 2 hours of ball-milling and enzymatic hydrolysis to obtain DELs for NMR analysis. For 2D-HSQC analysis, 50 mg of lignin was dissolved in 0.5 ml of dimethyl sulfoxide (DMSO)–d 6 , and spectra were recorded on a Bruker AV-HD-800 MHZ spectrometer (Bruker, Billerica, MA, USA) with the hsqcetgpsisp.2 pulse sequence ( 108 ). Spectral processing was performed using TopSpin 3 software (Bruker), including zero-filling, phase correction, and baseline correction. Volume integration of well-resolved cross peaks corresponding to major lignin interunit linkages (β-O-4, β-β, β-5) and aromatic units (S, G, and H) was conducted as described previously ( 108 ). Relative abundances were calculated based on normalized cross peak volumes and used for comparative analysis among genotypes and treatments. Xylem anatomical structure analysis The stem IN15 of well-watered and drought-treated poplars was collected for anatomical and ultrastructural analyses. Samples were immediately fixed in FAA (formalin:acetic acid:70% ethanol = 5:5:90) solution for 24 hours, followed by softening in a 70% ethanol:glycerol mixture (1:1) for 1 week to facilitate sectioning. Transverse and longitudinal sections with a thickness of 20 μm were obtained using a sliding microtome (Leica SM2010 R, Germany). Sections were dehydrated through a graded ethanol series (70%, 80%, 90%, 95%, and 100%). The sections were sequentially stained with aniline blue and safranin, and mounted in neutral balsam for microscopic observation. To specifically examine lignin distribution without staining artifacts, unstained sections were examined using an EVOS FL digital inverted fluorescence microscope (Thermo Fisher Scientific, USA). Lignin autofluorescence was captured under ultraviolet (UV) excitation at 355 to 400 nm, providing a direct visualization of SCW deposition patterns in vessel elements and fibers. For SEM, samples were fixed in 2.5% glutaraldehyde for 24 hours, followed by three rinses in the same buffer. The fixed tissues were dehydrated through a graded ethanol series (30%, 50%, 70%, 80%, 90%, 95%, and 100%). Ethanol was then replaced with tert-butanol in a series of substitutions (2:1, 1:1, 1:2, and pure tert-butanol), and the samples were subsequently dried using a vacuum freeze-dryer (Hitachi, Tokyo, Japan). To characterize different anatomical traits, the dried samples were prepared in two orientations. Transverse sections were obtained by cleanly fracturing the stem segments to expose the xylem cross-sections for measuring vessel wall thickness and lumen area. For analysis of pit patterns, including pit aperture diameter and pit membrane area, samples were longitudinally split along the grain with a fresh razor blade to expose the vessel luminal surfaces. After mounting on aluminum stubs, the samples were sputter-coated with gold-palladium and imaged using a Regulus 8100 SEM (Hitachi, Tokyo, Japan). Xylem anatomical traits were quantified using ImageJ software. For each plant, at least 100 vessel elements and pits were randomly selected, and the corresponding vessel wall thicknesses were measured for statistical analysis. As the pit apertures were mostly elliptical, diameter was measured as the widest part of the opening ( 109 ). Subcellular localization The full-length CDS of PtoCPK3 ( PtoCPK3 II ), PtoCPK3 DD , PtoERF72 ( PtoERF72 S96 ), and PtoERF72 A96 without the stop codons was cloned into the Xba I restriction site of pBI121 vector in frame with GFP (#P0275, MiaoLingBio, China), generating CaMV35S-driven fusion constructs (35S: PtoCPK3-GFP, PtoCPK3 DD -GFP, PtoERF72 S96 -GFP, and PtoERF72 A96 -GFP). To explore their subcellular localization, these constructs were infiltrated into Nicotiana benthamiana leaves via A. tumefaciens –mediated infiltration. Co-infiltration was performed with A. tumefaciens strains carrying 35S: PIP2A-mCherry (plasma membrane localization signal marker) or 35S: HSBP-mCherry (cytoplasm localization signal marker). To visualize the nuclei, infiltrated leaf tissues were incubated with DAPI (4′,6-diamidino-2-phenylindole; 1 μg/ml) for 15 min before imaging. At 48 hours post-infiltration, fluorescence signals in transiently expressing leaf epidermal cells were visualized and captured using a TCS SP8 confocal laser scanning microscope (Leica Microsystems, Wetzlar, Germany). GFP was excited at 488 nm (intensity: 3%; collection bandwidth: 500 to 560 nm; gain: 710 V), and mCherry was excited at 561 nm (intensity: 5%; collection bandwidth: 590 to 630 nm; gain: 750 V). DAPI was excited at 405 nm (intensity: 2%; collection bandwidth: 415 to 480 nm; gain: 650 V). Then, to biochemically verify the subcellular localization, subcellular fractionation and immunoblot analyses were performed. N. benthamiana leaves transiently expressing the fusion proteins with GFP ( 35S: PtoCPK3-GFP) or the free GFP control (pBI121-GFP empty vector, 35S: GFP) were harvested 48 hours post-infiltration. Nuclear, cytoplasmic, and membrane fractions were separated using plant nucleoprotein/cytoplasmic protein (#EX2232) and plasma membrane protein extraction kits (#EX2080) (Solarbio, Beijing, China). The isolated protein fractions were resolved by SDS–polyacrylamide gel electrophoresis (PAGE) and transferred onto polyvinylidene difluoride (PVDF) membranes. Immunoblotting was carried out using an anti-GFP antibody (1:2000 dilution) to detect the target fusion proteins. To validate the purity of the fractions and serve as loading controls, we used specific antibodies against histone H3 (nuclear marker), glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (cytoplasmic marker), and H + -dependent adenosine triphosphatase (H + -ATPase) (plasma membrane marker). Protein bands were visualized using an enhanced chemiluminescence (ECL) detection system. Protein interaction Yeast two-hybrid sequencing For Y2H-seq, a yeast cDNA library was constructed from stem tissues of P. tomentosa “1316” by OEbiotech Co. Ltd. (Shanghai, China). Full-length cDNAs were cloned into pDONR222 and pGADT7-DEST and transformed into Saccharomyces cerevisiae strain Y187. In the meantime, the full-length CDS of PtoCPK3 was inserted into the Eco RI and Bam HI restriction sites of pGBKT7 (BD) vector to generate bait PtoCPK3-BD protein. The bait PtoCPK3-BD protein was tested for self-activation and screened for interacting proteins using the Y2H mating protocol by OEbiotech Co. Ltd. (Shanghai, China). Positive colonies were collected for downstream identification. To maximize the diversity of interacting fragments, 10 positive colonies were pooled in a single PCR containing 50 μl of ddH 2 O, and 1 μl of the pooled suspension was used as the PCR template. PCR amplification was performed with an extended elongation time of 3 min per cycle to ensure efficient amplification of long inserts (assuming an extension rate of approximately 1 kb min −1 ). Finally, we used the purified products to perform high-throughput sequencing. The raw data were converted to fasta format using fq2fa and then aligned to the P. tomentosa reference genome using BLASTn v2.12.0 with the parameters outfmt 6 and evalue 1.0E-5. For Y2H validation, the full-length CDS of PtoERF72 ( PtoERF72 AA or PtoERF72 CC ) was inserted into the Eco RI restriction site of pGADT7 (AD) to generate the PtoERF72 AA/CC -AD construct. In addition, PtoERF72 ( A 96 ) CC CDS was cloned into the Eco RI restriction site of pGADT7 through site-directed mutagenesis of Ser 96 to Ala 96 . The full-length CDS of PtoCPK3 DD was constructed in the Eco RI and Bam HI restriction sites of the pGBKT7 vector. Different combinations of PtoERF72-AD and PtoCPK3-BD were cotransformed into Y2H yeast strains. Different transformants were selected on synthetic dropout medium lacking leucine and tryptophan (SD/-Leu/-Trp) and tested for interaction on SD/-Leu/-Trp/-His/-Ade/X-α-gal plates. Pairs of empty vectors pGBKT7 and pGADT7 were used as negative controls, while the pGBKT7-p53 and pGADT7-T pairs served as positive controls for protein-protein interaction. The vector for Y2H and the corresponding primers are listed in data S16. GST pull-down assay The full-length CDS of PtoCPK3 and PtoERF72 was amplified and cloned into the Bam HI and Sal I restriction sites of the pGEX-4T-1 [glutathione S-transferase (GST)–tag] and pET32a (His-tag) vectors to generate a GST-tagged fusion protein (PtoCPK3-GST) and His-tagged fusion protein (PtoERF72-His), respectively. Recombinant PtoCPK3-GST and PtoERF72-His proteins were expressed in Escherichia coli BL21 at 16°C [isopropyl-β- d -thiogalactopyranoside (IPTG): 0.5 mM] and purified using glutathione-Sepharose 4B (GE Healthcare, Chicago, IL, USA) and Ni-NTA agarose (Qiagen, Hilden, Germany), respectively, following the manufacturers’ protocols. For binding assays, 20 μg of GST or PtoCPK3-GST was incubated with 10 μg of PtoERF72-His in 500 μl of binding buffer (50 mM tris-HCl, pH 7.5, 150 mM NaCl, 1 mM DTT, 1 mM EDTA, and 0.5% NP-40) at 4°C for 2 hours with gentle rotation. After five washes with binding buffer, bound proteins were eluted by boiling in SDS loading buffer and analyzed by SDS-PAGE followed by immunoblotting. His-tagged proteins were detected using an anti-His-Tag mouse monoclonal antibody (1:5000; #CW0286M, CWBIOTECH, China), and GST-tagged proteins were detected using an anti–GST-Tag mouse monoclonal antibody (1:5000; #CW0084M, CWBIOTECH, China). GST protein alone was used as a negative control. The vectors for GST pull-down and the corresponding primers are listed in data S16. In vitro phosphorylation Recombinant PtoERF72-His protein was phosphorylated in vitro by activated PtoCPK3-GST in kinase reaction buffer [50 mM tris-HCl pH 7.5; 10 mM MgCl 2 ; 1, 10, or 15 mM CaCl 2 ; 1 mM DTT; 1 mM adenosine triphosphate (ATP)] at 30°C for 2 hours. Reaction products were separated by SDS-PAGE and either subsequently subjected to immunoblotting with anti-phosphoserine/threonine antibodies (ECM, PP2551; 1:1000 dilution) or CBB-stained for LC-MS/MS analysis (Beijing BiotechPack Scientific, Beijing, China). Bands corresponding to PtoERF72-His were excised and subjected to in-gel digestion, and the resulting peptides were analyzed with a Q-Exactive Orbitrap mass spectrometer (Thermo Fisher Scientific, Waltham, MA, USA) coupled to an EASY-nLC 1200 system equipped with a 150 μm × 150 mm in-house–packed column (1.9 μm, 100 Å, Dr. Maisch GmbH, Germany) at a flow rate of 600 nl/min while running a 6% to 95% acetonitrile gradient and sprayed directly into the mass spectrometer. Phosphopeptides were identified and quantified using MaxQuant (1.6.2.10) based on the extracted ion chromatogram peak areas, with carbamidomethylation as a fixed modification and phosphorylation (Ser/Thr) and oxidation as variable modifications. Then, site-directed mutagenesis was performed to substitute Ser 96 with alanine (PtoERF72 A96 ). The mutant PtoERF72 A96 -His protein was expressed and purified using the same procedure as the WT protein (PtoERF72 S96 ) and subsequently subjected to in vitro phosphorylation assays with PtoCPK3-GST under identical reaction conditions. Phosphorylation of the mutant protein was analyzed by immunoblotting and compared with that of the PtoERF72 S96 -His protein. Surface plasmon resonance SPR was performed using a Biacore T200 (Cytiva, Chicago, IL, USA). Purified PtoCPK3-GST protein was immobilized on a Series S CM5 sensor chip via standard amine coupling. Briefly, the surface was activated with 0.4 M 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide (EDC) and 0.1 M N-hydroxysuccinimide (NHS), followed by ligand injection at 10 μg/ml in 10 mM sodium acetate buffer (pH 5.0), yielding ~4000 response units (RUs). Unreacted sites were blocked with 1 M ethanolamine-HCl (pH 8.5). Kinetic assays were conducted in a specialized running buffer tailored for kinase-substrate interactions, consisting of 30 mM tris-HCl, pH 7.8, 100 mM NaCl, 2.5 mM CaCl 2 , 2.5 mM MgCl 2 , 0.5 mM TCEP, 2 mM reduced glutathione, 0.4 mM ATP, and 0.05% Tween 20. Analytes PtoERF72 S96 and PtoERF72 A96 were prepared in a twofold serial dilution series ranging from 0.31 to 40 μM in running buffer and injected using a single-cycle kinetics program at a flow rate of 30 μl/min. Each injection cycle included a 90-s association phase and a final 600-s dissociation phase to allow the signal to return to baseline. Sensorgrams were double-referenced by subtracting the response from both the reference flow cell (no ligand) and the buffer-only injections. The resulting data were globally fitted to a 1:1 binding model using Biacore T200 Evaluation Software ( 110 ) to calculate the association rate constant ( k a ), dissociation rate constant ( k d ), and the equilibrium dissociation constant ( K D ). Luciferase complementation imaging The full-length CDS of PtoCPK3 was cloned into the Bam HI and Sal I restriction sites of pCambia1300-NLuc vector to generate NLuc-PtoCPK3, while full-length PtoERF72 S96 or PtoERF72 A96 was cloned into the same restriction sites of pCambia1300-CLuc vector to generate PtoERF72 S96 -CLuc and PtoERF72 A96 -CLuc, respectively. All constructs were transformed into A. tumefaciens , adjusted to an OD 600 of 0.4, and mixed at a 1:1 ratio for infiltration into N. benthamiana leaves. Plants were kept in the dark for 24 hours, followed by 24 hours under normal growth conditions. For calcium treatment, infiltrated leaves were sprayed with 10 mM CaCl 2 solution at the indicated time point before imaging. Control leaves were treated with an equal volume of distilled water. For luminescence imaging, leaves were sprayed with 100 mM luciferin, incubated in the dark for 10 min, and imaged using a Lumazone_1300B CCD system (Roper Bioscience, Trenton, NJ, USA). To quantify luciferase activity, three 1-cm leaf discs were collected from each infiltrated area and luciferase activity was measured using the Dual-Luciferase Reporter Assay System (Vazyme, China). Each experiment was performed in triplicate. The vectors for LCI and the corresponding primers are listed in data S16. Biomolecular fluorescence complementation assay The full-length CDS of PtoCPK3 was cloned into the Bam HI and Sal I restriction sites of pSPYNE(R)173 vector to generate YFPn-PtoCPK3, and the full-length CDS of PtoERF72 S96 or PtoERF72 A96 was cloned into the same restriction sites of pSPYCE(MR) vector to generate PtoERF72 S96 -YFPc or PtoERF72 A96 -YFPc, respectively. The resulting constructs were introduced into A. tumefaciens GV3101 competent cells. Pairs of PtoCPK3 and PtoERF72 constructs were co-infiltrated into N. benthamiana leaves. After 48 hours, the YFP signals were observed under a TCS SP8 confocal laser scanning microscope (Leica Microsystems, Wetzlar, Germany). YFP fluorescence was detected using the following properties: laser, 514 nm; intensity, 3%; collection bandwidth, 520 to 570 nm; and gain, 723 V. H2B-mCherry (nuclear localization signal marker) was detected with the following properties: laser, 561 nm; intensity: 5%; collection bandwidth: 590 to 630 nm; gain: 750 V. A YFP signal was detected in leaves coexpressing YFPn-PtoCPK3 and PtoERF72 S96 -YFPc plasmids. H2B-mCherry fluorescence was used as a control. The vectors for BiFC and the corresponding primers are listed in data S16. Lignin metabolite detection Lignin-related metabolites of all tested poplars (WT, PtoCPK3 -OE, and ptocpk3 -KD) were quantified through broad-targeted metabolomics analysis conducted by MetWare Biotechnology Co. Ltd. (Wuhan, China) following their standardized UPLC-MS/MS platform. The detected lignin-related metabolites included cinnamic acid, p -coumaric acid, p -coumaraldehyde, p -coumaryl alcohol, caffeic acid, caffealdehyde, ferulic acid, 4-hydroxy-3-methoxycinnamaldehyde (coniferaldehyde), coniferyl alcohol, sinapic acid, sinapaldehyde, and sinapyl alcohol. In addition, for transgenic poplars expressing PtoERF72 or PtoUGT72AZ2 , targeted analysis of monolignol-related metabolites was performed using GC-MS. Freeze-dried xylem powder (50 to 100 mg) was extracted with 80% methanol containing 0.1% formic acid, sonicated, and centrifuged at 12,000 g for 10 min. Supernatants were pooled, evaporated to dryness under nitrogen, and derivatized with 100 μl of N-methyl-N-(tert-butyldimethylsilyl)trifluoroacetamide (MTBSTFA) and 25 μl of N , N ′-dimethylformamide (DMF) at 80°C for 30 min. After cooling, the derivatives were diluted in hexane and filtered through a 0.22-μm membrane. GC-MS analysis was performed on an HP-5MS column (30 m × 0.25 mm, 0.25 μm) with helium as the carrier gas at 1.0 ml/min. The oven program was set as follows: 80°C (1 min), ramped to 300°C at 10°C/min, and held for 5 min. Spectra were acquired in electron ionization (EI) mode (70 eV), scanning from 50 to 650 m / z . Quantification was performed using calibration curves generated from authentic standards of ferulic acid and coniferyl alcohol processed under identical derivatization conditions ( 111 , 112 ). DAP-seq sampling and data analysis DAP-seq was performed as described ( 113 ) using genomic DNA extracted from stem tissue of P. tomentosa clone “1316.” A DNA library was prepared using the NGS0602-MICH TLX DNA-Seq Kit (MICH, Baoding, Hebei, China). The CDS of PtoERF72 was cloned into pFN19K HaloTag T7 SP6 Flexi vectors and expressed in the TNT SP6 Wheat Germ Extract System (Promega, Madison, WI, USA). Subsequently, the synthesized PtoERF72 protein was immobilized onto Magne HaloTag Beads (Promega) and incubated with 500 ng of the prepared gDNA library in a binding buffer for 1 hour at room temperature with gentle rotation. After incubation, the protein-DNA complexes were washed five times with PBS buffer containing 0.05% NP-40 to remove nonspecifically bound DNA fragments. The specifically bound DNA was eluted by heating the beads at 98°C for 10 min and subsequently purified. The eluted DNA fragments were sequenced on an Illumina NovaSeq6000 (Illumina, San Diego, CA, USA) with two technical replicates. Raw sequencing reads were preprocessed using Trimmomatic to remove adapter sequences and low-quality reads. The cleaned reads were aligned to the P. tomentosa reference genome using Bowtie2 ( 114 ) with default parameters. Peak calling was performed using MACS2 (v2.1.1) to identify regions of substantial enrichment (peaks) compared to the negative control (empty HaloTag protein), with a significance threshold of q < 0.05. Conserved binding motifs were identified with Multiple EM for Motif Elicitation–Chromatin Immunoprecipitation (MEME-ChIP). Target genes were defined as peaks within 2 kb upstream of the first ATG within the CDS. Functional enrichment analysis of the target genes was conducted using TBtools to identify substantially overrepresented KEGG pathways and GO terms. Transcriptional regulation analysis Electrophoretic mobility shift assay For PtoERF72 DNA binding assays, the full-length CDS of PtoERF72 AA ( PtoERF72 S96 ) and PtoERF72 CC ( PtoERF72 S96T126 ) was used as the WT background for DNA binding assays. To evaluate the effect of phosphorylation on PtoERF72 DNA binding activity, site-directed mutagenesis was performed to generate phosphorylation-related variants, including PtoERF72 D96 -His and PtoERF72 D96D126 -His. Biotin-labeled double-stranded oligonucleotide probes corresponding to the ERF binding sites (GCC-box motifs present in the PtoUGT72AZ2 and PtoWOX13b promoters) were synthesized (Beyotime, China). For PtoWOX13b DNA binding assays, the full-length CDS of PtoWOX13b was cloned into the Bam HI and Sal I restriction sites of the pET32a expression vector to produce recombinant PtoWOX13b-His in E. coli BL21 and purified as described above. Biotin-labeled probes corresponding to the conserved WOX-binding element (WBE; CAATCAA/C) present in the promoters of PtoUGT72AZ2 , PtoERF15 , PtoNAC005 , PtoNAC006 , PtoNAC118 , PtoNAC120 , and PtoLAC14 were synthesized, with unlabeled probes used as competitors. Unlabeled probes with identical sequences were used as competitors. Binding reactions contained 500 ng of purified recombinant protein and biotin-labeled probes in binding buffer, and were incubated at 25°C for 20 min. For each probe, 50× and 100× excess competitor was added. Complexes were separated on 12% PAGE, transferred to nylon membranes (Amersham, Woburn, MA, USA), and cross-linked by UV irradiation. The results of EMSA were visualized using HRP-conjugated streptavidin with an ECL Kit (Beyotime, China), and a ChemiDoc Imaging System (Bio-Rad, Hercules, CA, USA). The vectors for EMSA and the corresponding primers are listed in data S16. Yeast one-hybrid Promoter fragments containing ERF binding sites (GCC-box) from PtoUGT72AZ2 and PtoWOX13b promoters, as well as the WOX-binding site (CAATCAA/C) within the PtoUGT72AZ2, PtoERF15 , PtoNAC005 , PtoNAC006 , PtoNAC118 , PtoNAC120 , and PtoLAC14 promoters, were cloned into the Sac I and Xho I restriction sites of the pBait-AbAi vector. To eliminate self-activation, each bait construct was linearized and integrated into the yeast Y1HGold strain according to the manufacturer’s instructions. The minimal inhibitory concentration of Aureobasidin A (AbA) for each bait strain was determined before interaction screening. The full-length CDS of PtoERF72 and PtoWOX13b was individually amplified and inserted into the Eco RI restriction site of the pGADT7 vector to generate prey constructs (PtoERF72-AD, PtoWOX13b-AD). Transformed yeast cells were selected on SD/-Leu medium supplemented with AbA (200 ng/μl) and incubated at 30°C for 3 to 5 days. Yeast cells harboring the empty AD vector served as negative controls. Growth on selective medium indicated a positive interaction between the transcription factor and the corresponding promoter fragment. The vectors for Y1H and the corresponding primers are listed in data S16. Dual-luciferase reporter assay The CDS of PtoERF72 ( PtoERF72 S96 ) and PtoERF72 A96 was cloned into the Bam HI and Sal I restriction sites of the pGreenII 62-SK vector to generate effector plasmids. Promoter fragments (2000 bp upstream of the first ATG in CDS) of PtoUGT72AZ2 and PtoWOX13b were amplified and inserted into a modified pGreenII 0800-LUC vector to obtain the reporter plasmids, in which the target promoter drives the firefly luciferase (LUC) gene, while Renilla luciferase (REN) under the control of the constitutive promoter serves as an internal control. Each pair of effector and reporter plasmids was introduced into A. tumefaciens strain GV3101, and bacterial cultures were adjusted to the same optical density and mixed at a 1:1 ratio before infiltration. The mixed cultures were transiently co-infiltrated into N. benthamiana leaves. After 48 hours of infiltration, transient expression was confirmed by monitoring luciferase luminescence using the Lumazone_1300B CCD imaging system (Roper Bioscience, Trenton, NJ, USA). Subsequently, three leaf discs (1 cm in diameter) were harvested, and the DLRA was performed using the Dual-Luciferase Reporter Assay System kit (Vazyme, China). Binding activity was evaluated by calculating the ratio of firefly luciferase (LUC) to Renilla luciferase (REN) between treatments and controls. Each experiment was performed in triplicate. The vectors for DLRA and the corresponding primers are listed in data S16. In vitro glycosylation reaction of PtoUGT72AZ2 The full-length CDS of PtoUGT72AZ2 was cloned into the pGEX-4T-1 vector to produce PtoUGT72AZ2-GST protein. Recombinant protein was expressed in E. coli BL21 at 16°C (IPTG: 0.5 mM), and cells were lysed in PBS (pH 7.4) by ultrasonication. The corresponding vectors and primers are listed in data S16. The clarified supernatant was applied to Glutathione Sepharose 4B (GE Healthcare, Chicago, IL, USA), extensively washed, and eluted with 10 mM reduced glutathione in 50 mM tris-HCl (pH 8.0). Eluted proteins were subsequently desalted and concentrated using centrifugal ultrafiltration. Enzyme activity assays were performed in 200 μl of reaction buffer containing 100 mM tris-HCl (pH 7.5), 5 mM UDP-glucose, 2 μg of purified PtoUGT72AZ2-GST protein, and 1 mM substrate at 30°C for 2 hours. Reactions were terminated by adding 100 μl of methanol, and products were analyzed by UPLC and LC-MS/MS to identify glycosylated metabolites. Authentic standards, including ferulic acid, coniferyl alcohol, ferulic acid glucoside, and coniferyl glucoside, were analyzed in parallel for comparison. Chromatographic separation was performed on an Agilent SB-C18 column (1.8 μm, 2.1 mm by 100 mm) with water (0.1% formic acid) and acetonitrile (0.1% formic acid) as mobile phases. The flow rate was 0.35 ml min −1 , column temperature was 40°C, and injection volume was 2 μl. MS detection was conducted in MRM mode with optimized declustering potential and collision energy for each analyte. Kinase activity detection of different genotypes with PtoCPK3-PtoERF72 In vitro phosphorylation assays were performed using purified recombinant proteins to evaluate the kinase activity of PtoCPK3 and its ability to phosphorylate PtoERF72 variants. A total of 11 recombinant reaction groups were included: (i) PtoCPK3(PtoCPK3 II )-GST, (ii) PtoCPK3 DD -GST, (iii) PtoERF72(S 96 ) AA -His, (iv) PtoERF72(S 96 ) CC -His, (v) PtoERF72(A 96 ) CC -His, (vi) PtoCPK3 II -GST + PtoERF72(S 96 ) AA -His, (vii) PtoCPK3 II -GST + PtoERF72(S 96 ) CC -His, (viii) PtoCPK3 II -GST + PtoERF72(A 96 ) CC -His, (ix) PtoCPK3 DD -GST + PtoERF72(S 96 ) AA -His, (x) PtoCPK3 DD -GST + PtoERF72(S 96 ) CC -His, and (xi) PtoCPK3 DD -GST + PtoERF72(A 96 ) CC -His. Kinase activity was assessed using the Promega Kinase-Glo reagent (Promega). Each reaction (100 μl total volume) consisted of 50 μl of kinase reaction mixture and 50 μl of Kinase-Glo reagent, which contains a thermostable luciferase that generates a luminescent signal proportional to the amount of ATP remaining in the reaction. Luminescence was measured using a microplate reader. Because luminescence intensity is inversely proportional to ATP consumption, a decrease in relative luminescence units (RLUs) reflects increased kinase activity. RLU values were recorded, and the rate of RLU decline was used to evaluate and compare kinase activities among different reaction groups. Drought tolerance assessment A total of 45 P. tomentosa individuals were selected, including 15 PtoCPK3 II _PtoERF72 AA and 15 PtoCPK3 II _PtoERF72 CC genotypes from the NE and NW regions, and 15 PtoCPK3 DD _PtoERF72 AA genotypes selected from the S region, respectively. To ensure genetic uniformity and eliminate environmental maternal effects, explants were obtained from accessions asexually propagated via grafting. These explants underwent surface disinfection and were established as in vitro cultures on 1/2 Murashige-Skoog basal medium. Clonal plantlets were then multiplied through subculturing to produce three identical replicates per accession. Four-month-old seedlings with uniform height and ground diameter were assigned to either a well-watered control group or a drought-stress treatment group. During the initial growth stage, soil RWC was kept at 70% ± 5%. Four-month-old poplars were assigned to well-watered controls and drought treatment (soil RWC reduced to 20% ± 5% for 15 days) before phenotyping. Four photosynthetic traits were measured as described above, including Pn, transpiration rate (Tr), Gs, and WUE. To reduce the influence of circadian rhythms on photosynthesis, these traits were measured in functional leaves from 9:00 a.m. to 11:00 a.m. on a sunny day. WUE was defined as the ratio between CO 2 assimilation and Tr. Six physiological traits were measured, including relative electrolyte leakage (REL), ABA, MDA, PRO, superoxide dismutase (SOD), and soluble sugar (SS) contents. REL was assessed by measuring the initial electrical conductivity ( C 1 ) of fresh leaf discs after 24 hours of incubation in deionized water at room temperature, followed by measuring the final conductivity ( C 2 ) after boiling the samples for 20 min; the REL was calculated as ( C 1 / C 2 ) × 100%. The contents of ABA, MDA, PRO, SOD, and SS were determined using specific ELISA kits based on a double-antibody sandwich principle. In addition, three xylem characteristics at the stem IN15 were assessed, including VCD, vessel lumen area (VLA), and TLC. Statistical analysis All statistical analyses were conducted using Student’s t test unless otherwise specified. For comparisons involving multiple genotypes, one-way ANOVA with Tukey’s test was performed to assess pairwise differences. All data are presented as mean ± SD unless otherwise noted. Significance thresholds of * P < 0.05, ** P < 0.01, and *** P < 0.001 were used throughout. Acknowledgments We thank H. Nie and C. Zhang for helpful discussions. Funding: This work was supported by the State “14.5” Key Research Program of China (no. 2021YFD2200101), the National Natural Science Foundation of China (nos. 32370396, 32570425, and 32271907), the China Postdoctoral Science Foundation (no. 2025 M781844), and the 111 Project (no. B20050). Author contributions: Conceptualization: M.Q. and D.Z. Methodology: M.Q., Y.G., and D.Z. Formal analysis: Y.J., M.Q., B.L., T.R., and Q.D. Investigation: D.W., Y.J., M.Q., B.L., Y.C., M.G., D.Z., W.Q., T.R., and T.Z. Resources: Y.J., M.Q., B.L., S.Q., Y.F., T.Z., and D.Z.. Validation: D.W., Y.J., M.Q., M.G., Y.G., D.Z., Q.D., and D.Z. Visualization: D.W., Y.J., M.Q., B.L., Y.G., and D.Z. Writing—original draft: D.W. and M.Q. Writing—review and editing: D.W., M.Q., L.X., Q.D., P.K.I., Y.A.E., and D.Z. Supervision: M.Q., P.K.I., and D.Z. Data curation: D.Z. Project administration: D.Z. Funding acquisition: D.Z. Competing interests: The authors declare that they have no competing interests. Data, code, and materials availability: The raw data of genome resequencing of 302 P. tomentosa accessions, Y2H-seq data of PtoCPK3, DAP-seq data of PtoERF72, and all RNA-seq data of P. tomentosa “1316,” PtoCPK3 -OE, ptocpk3 -KD, PtoERF2 -OE, and ptoerf72 -KD lines have been submitted to the Genome Sequence Archive (GSA) in the BIG Data Center under accession numbers CRA000903 and PRJCA050561 (accessible at https://bigd.big.ac.cn/gsa ). The mass spectrometry proteomics and phosphorylated proteome data have been deposited to the ProteomeXchange Consortium via the iProX partner repository with the dataset identifier PXD070526 (accessible at https://proteomecentral.proteomexchange.org ). The CDS data generated in this study are available in the GenBank databases under the accession numbers PX597595, PX597596, PX597593, PX597594, PX597591, and PX961919 for PtoCPK3 ( PtoCPK3 II ), PtoCPK3 DD , PtoERF72 ( PtoERF72 AA ), PtoERF72 CC , PtoWOX13b , and PtoUGT72AZ2, respectively (accessible at https://www.ncbi.nlm.nih.gov/ ). All data and code needed to evaluate and reproduce the results in the paper are present in the paper and/or the Supplementary Materials. The source codes for this study have been deposited in the Zenodo repository (accessible at https://zenodo.org/records/18886819 ). Alternatively, our GitHub repository can be reached at https://github.com/jiangyongsen121/analysis . Requests for PtoCPK3 , PtoERF72 , PtoWOX13b , or PtoUGT72AZ2 constructs and plasmids generated in this study can be made available by contacting the corresponding author ( DeqiangZhang@bjfu.edu.cn ). Supplementary Materials The PDF file includes: Figs. S1 to S16 Legends for data S1 to S36 sciadv.aee2172_sm.pdf (7.5MB, pdf) Other Supplementary Material for this manuscript includes the following: Data S1 to S36 sciadv.aee2172_data_s1_to_s36.zip (5.9MB, zip) REFERENCES 1. Li Y., Zhao M. S., Motesharrei S., Mu Q. Z., Kalnay E., Li S. C., Local cooling and warming effects of forests based on satellite observations. Nat. Commun. 6, 6603 (2015). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 2. Ellison D., Morris C. E., Locatelli B., Sheil D., Cohen J., Murdiyarso D., Gutierrez V., van Noordwijk M., Creed I. F., Pokorny J., Gaveau D., Spracklen D. V., Tobella A. B., Ilstedt U., Teuling A. J., Gebrehiwot S. G., Sands D. C., Muys B., Verbist B., Springgay E., Sugandi Y., Sullivan C. A., Trees, forests and water: Cool insights for a hot world. Global Environ. Chang. 43, 51–61 (2017). [ Google Scholar ] 3. Reyer C. P. O., Brouwers N., Rammig A., Brook B. W., Epila J., Grant R. F., Holmgren M., Langerwisch F., Leuzinger S., Lucht W., Medlyn B., Pfeifer M., Steinkamp J., Vanderwel M. C., Verbeeck H., Villela D. M., Forest resilience and tipping points at different spatio-temporal scales: Approaches and challenges. J. Ecol. 103, 5–15 (2015). [ Google Scholar ] 4. Allen C. D., Breshears D. D., McDowell N. G., On underestimation of global vulnerability to tree mortality and forest die-off from hotter drought in the Anthropocene. Ecosphere 6, 1–55 (2015). [ Google Scholar ] 5. Allen C. D., Macalady A. K., Chenchouni H., Bachelet D., McDowell N., Vennetier M., Kitzberger T., Rigling A., Breshears D. D., Hogg E. H., Gonzalez P., Fensham R., Zhang Z., Castro J., Demidova N., Lim J. H., Allard G., Running S. W., Semerci A., Cobb N., A global overview of drought and heat-induced tree mortality reveals emerging climate change risks for forests. For. Ecol. Manage. 259, 660–684 (2010). [ Google Scholar ] 6. Ménard D., Blaschek L., Kriechbaum K., Lee C. C., Serk H., Zhu C. T., Lyubartsev A., Nuoendagula Z., Bacsik L., Bergström A., Mathew S., Kajita E. P., Plant biomechanics and resilience to environmental changes are controlled by specific lignin chemistries in each vascular cell type and morphotype. Plant Cell 34, 4877–4896 (2022). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 7. Hacke U. G., Sperry J. S., Pockman W. T., Davis S. D., McCulloh K. A., Trends in wood density and structure are linked to prevention of xylem implosion by negative pressure. Oecologia 126, 457–461 (2001). [ DOI ] [ PubMed ] [ Google Scholar ] 8. Hacke U. G., Stiller V., Sperry J. S., Pittermann J., McCulloh K. A., Cavitation fatigue. Embolism and refilling cycles can weaken the cavitation resistance of xylem. Plant Physiol. 125, 779–786 (2001). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 9. Hori C., Yu X., Mortimer J. C., Sano R., Matsumoto T., Kikuchi J., Demura T., Ohtani M., Impact of abiotic stress on the regulation of cell wall biosynthesis in Populus trichocarpa . Plant Biotechnol. ( Tokyo ) 37, 273–283 (2020). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 10. Vanholme R., Demedts B., Morreel K., Ralph J., Boerjan W., Lignin biosynthesis and structure. Plant Physiol. 153, 895–905 (2010). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 11. Weng J. K., Chapple C., The origin and evolution of lignin biosynthesis. New Phytol. 187, 273–285 (2010). [ DOI ] [ PubMed ] [ Google Scholar ] 12. Renault H., Alber A., Horst N. A., Lopes A. B., Fich E. A., Kriegshauser L., Wiedemann G., Ullmann P., Herrgott L., Erhardt M., Pineau E., Ehlting J., Schmitt M., Rose J. K. C., Reski R., Werck-Reichhart D., A phenol-enriched cuticle is ancestral to lignin evolution in land plants. Nat. Commun. 8, 14713 (2017). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 13. Costa A., Kieffer Y., Scholer-Dahirel A., Pelon F., Bourachot B., Cardon M., Sirven P., Magagna I., Fuhrmann L., Bernard C., Bonneau C., Kondratova M., Kuperstein I., Zinovyev A., Givel A. M., Parrini M. C., Soumelis V., Vincent-Salomon A., Mechta-Grigoriou F., Fibroblast heterogeneity and immunosuppressive environment in human breast cancer. Cancer Cell 33, 463–479.e10 (2018). [ DOI ] [ PubMed ] [ Google Scholar ] 14. Atif R. M., Shahid L., Waqas M., Ali B., Rashid M. A. R., Azeem F., Nawaz M. A., Wani S. H., Chung G., Insights on calcium-dependent protein kinases (CPKs) signaling for abiotic stress tolerance in plants. Int. J. Mol. Sci. 20, 5298 (2019). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 15. Cheng S.-H., Willmann M. R., Chen H.-C., Sheen J., Calcium signaling through protein kinases. The Arabidopsis calcium-dependent protein kinase gene family. Plant Physiol. 129, 469–485 (2002). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 16. Boudsocq M., Sheen J., CDPKs in immune and stress signaling. Trends Plant Sci. 18, 30–40 (2013). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 17. Fan B., Liao K., Wang L. N., Shi L. L., Zhang Y., Xu L. J., Zhou Y., Li J. F., Chen Y. Q., Chen Q. F., Xiao S., Calcium-dependent activation of CPK12 facilitates its cytoplasm-to-nucleus translocation to potentiate plant hypoxia sensing by phosphorylating ERF-VII transcription factors. Mol. Plant 16, 979–998 (2023). [ DOI ] [ PubMed ] [ Google Scholar ] 18. Zhu S. Y., Yu X. C., Wang X. J., Zhao R., Li Y., Fan R. C., Shang Y., Du S. Y., Wang X. F., Wu F. Q., Xu Y. H., Zhang X. Y., Zhang D. P., Two calcium-dependent protein kinases, CPK4 and CPK11, regulate abscisic acid signal transduction in Arabidopsis . Plant Cell 19, 3019–3036 (2007). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 19. Sun F. L., Wang J. H., Geng S. W., Liang Y. J., Gong Z. L., Yang N., Qian S. S., Zhang N. L., Li X. Y., Wang J. D., Zheng J. Y., Comprehensive transcriptomic and metabolomic analysis revealed drought tolerance regulatory pathways in upland cotton. Front. Plant Sci. 16, 1571944 (2025). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 20. Zhang X. M., Tian X. Q., Gao X. K., Sun G. F., Peng Z., Geng X. L., Peng J., Dai P. H., Wang X., Li H. G., Wang Z. Z., Pan Z., Jia Y. H., Chen B. J., Hu D. W., Wang L. R., Pang B. Y., Xing A. S., Lu S., Gao X., Dai C. Y., Wei X., Zhang H. B., Cui J. J., He S. P., Ma L., Du X. M., Integrated metabolomic and transcriptomic analyses identify MYB genes regulating key metabolites and agronomic traits in upland cotton Gossypium hirsutum . Nat. Genet. 57, 2819–2830 (2025). [ DOI ] [ PubMed ] [ Google Scholar ] 21. Zhang M., Zhou X. Y., Wang L. M., Liang X. Y., Liu X., Jiang C. F., A SnRK2-HAK regulatory module confers natural variation of salt tolerance in maize. Nat. Commun. 16, 4026 (2025). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 22. Castellana S., Triozzi P. M., Dell’Acqua M., Loreti E., Perata P., Environmental genome-wide association studies across precipitation regimes reveal that the E3 ubiquitin ligase MBR1 regulates plant adaptation to rainy environments. Plant Commun. 5, 101074 (2024). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 23. Lazic D., Gessner C., Liepe K. J., Lesur-Kupin I., Mader M., Blanc-Jolivet C., Gömöry D., Liesebach M., González-Martínez S. C., Fladung M., Degen B., Müller N. A., Genomic variation of European beech reveals signals of local adaptation despite high levels of phenotypic plasticity. Nat. Commun. 15, 8553 (2024). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 24. Song Y. P., Chen P. F., Xuan A. R., Bu C. H., Liu P., Ingvarsson P. K., El-Kassaby Y. A., Zhang D. Q., Integration of genome wide association studies and co-expression networks reveal roles of PtoWRKY 42-PtoUGT76C1-1 in trans -zeatin metabolism and cytokinin sensitivity in poplar. New Phytol. 231, 1462–1477 (2021). [ DOI ] [ PubMed ] [ Google Scholar ] 25. Lobo A., Hansen J. K., Hansen L. N., Kjær E. D., Differences among six woody perennials native to Northern Europe in their level of genetic differentiation and adaptive potential at fine local scale. Ecol. Evol. 8, 2231–2239 (2018). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 26. Rellstab C., Keller S. R., Can we use genomic data to predict maladaptation to environmental change? Mol. Ecol. Resour. 25, e14059 (2025). [ DOI ] [ PubMed ] [ Google Scholar ] 27. Zhang M., Suren H., Holliday J. A., Phenotypic and genomic local adaptation across latitude and altitude in Populus trichocarpa . Genome Biol. Evol. 11, 2256–2272 (2019). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 28. Li L. Z., Jin Z. Y., Huang R., Zhou J. X., Song F. Y., Yao L. C., Li P., Lu W. J., Xiao L., Quan M. Y., Zhang D. Q., Du Q. Z., Leaf physiology variations are modulated by natural variations that underlie stomatal morphology in Populus . Plant Cell Environ. 46, 150–170 (2023). [ DOI ] [ PubMed ] [ Google Scholar ] 29. Li S., Lin Y. C. J., Wang P. Y., Zhang B. F., Li M., Chen S., Shi R., Tunlaya-Anukit S., Liu X. Y., Wang Z. F., Dai X. F., Yu J., Zhou C. G., Liu B. G., Wang J. P., Chiang V. L., Li W., The AREB1 transcription factor influences histone acetylation to regulate drought responses and tolerance in Populus trichocarpa . Plant Cell 31, 663–686 (2019). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 30. Kong L. F., Song Q., Wei H. B., Wang Y. H., Lin M. H., Sun K., Zhang Y. Q., Yang J. R., Li C. F., Luo K. M., The AP2/ERF transcription factor PtoERF15 confers drought tolerance via JA-mediated signaling in Populus . New Phytol. 240, 1848–1867 (2023). [ DOI ] [ PubMed ] [ Google Scholar ] 31. Niu Z. M., Li G. T., Hu H. Y., Lv J. J., Zheng Q. W., Liu J. Q., Wan D. S., A gene that underwent adaptive evolution, LAC2 (LACCASE), in Populus euphratica improves drought tolerance by improving water transport capacity. Hortic Res. 8, 88 (2021). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 32. Li Q. Z., Hu T., Lu T. J., Yu B., Zhao Y., Calcium-dependent protein kinases CPK3/4/6/11 and 27 respond to osmotic stress and activate SnRK2s in Arabidopsis . Dev. Cell 60, 1423–1438.e8 (2025). [ DOI ] [ PubMed ] [ Google Scholar ] 33. Mori I. C., Murata Y., Yang Y. Z., Munemasa S., Wang Y. F., Andreoli S., Tiriac H., Alonso J. M., Harper J. F., Ecker J. R., Kwak J. M., Schroeder J. I., CDPKs CPK6 and CPK3 function in ABA regulation of guard cell S-type anion- and Ca 2+ -permeable channels and stomatal closure. PLOS Biol. 4, 1749–1762 (2006). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 34. Latz A., Mehlmer N., Zapf S., Mueller T. D., Wurzinger B., Pfister B., Csaszar E., Hedrich R., Teige M., Becker D., Salt stress triggers phosphorylation of the Arabidopsis vacuolar K + channel TPK1 by calcium-dependent protein kinases (CDPKs). Mol. Plant 6, 1274–1289 (2013). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 35. Wu Y. X., Zhao S. J., Tian H., He Y. Q., Xiong W., Guo L., Wu Y., CPK3-phosphorylated RhoGDI1 is essential in the development of Arabidopsis seedlings and leaf epidermal cells. J. Exp. Bot. 64, 3327–3338 (2013). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 36. Kitin P., Voelker S. L., Meinzer F. C., Beeckman H., Strauss S. H., Lachenbruch B., Tyloses and phenolic deposits in xylem vessels impede water transport in low-lignin transgenic poplars: A study by cryo-fluorescence microscopy. Plant Physiol. 154, 887–898 (2010). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 37. Hacke U. G., Spicer R., Schreiber S. G., Plavcová L., An ecophysiological and developmental perspective on variation in vessel diameter. Plant Cell Environ. 40, 831–845 (2017). [ DOI ] [ PubMed ] [ Google Scholar ] 38. Haghighat M., Zhong R. Q., Ye Z. H., WUSCHEL-RELATED HOMEOBOX genes are crucial for normal vascular organization and wood formation in poplar. Plant Sci. 346, 112138 (2024). [ DOI ] [ PubMed ] [ Google Scholar ] 39. Dai X. F., Zhai R., Lin J. J., Wang Z. F., Meng D. K., Li M., Mao Y. L., Gao B. Y., Ma H. Y., Zhang B. F., Sun Y., Li S., Zhou C. G., Lin Y. C. J., Wang J. P., Chiang V. L., Li W., Cell-type-specific PtrWOX4a and PtrVCS2 form a regulatory nexus with a histone modification system for stem cambium development in Populus trichocarpa . Nat. Plants 9, 96–111 (2023). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 40. Li M., Dong H., Li J. Y., Dai X. F., Lin J. J., Li S., Zhou C. G., Chiang V. L., Li W., PtrVCS2 regulates drought resistance by changing vessel morphology and stomatal closure in Populus trichocarpa . Int. J. Mol. Sci. 24, 4458 (2023). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 41. Zhang Y., Liu Y. Y., Wang X. Y., Wang R. Q., Chen X. B., Wang S., Wei H. R., Wei Z. G., PtrWOX13A promotes wood formation and bioactive gibberellins biosynthesis in Populus trichocarpa . Front. Plant Sci. 13, 835035 (2022). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 42. Dong N. Q., Lin H. X., Contribution of phenylpropanoid metabolism to plant development and plant-environment interactions. J. Integr. Plant Biol. 63, 180–209 (2021). [ DOI ] [ PubMed ] [ Google Scholar ] 43. Xu R., Shao C. Y., Luo Y. Q., Zhou B., Zhu Q., Qiu S. Q., Liu Z. H., Liu S. Q., Shen C. W., Tea polyphenol mediated CsMYB77 regulation of CsPOD44 to promote tea plant ( Camellia sinensis ) root drought resistance. Hortic Res. 12, uhaf048 (2025). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 44. Chen X. H., Su W. L., Zhang H., Zhan Y. G., Zeng F. S., Fraxinus mandshurica 4-coumarate-CoA ligase 2 enhances drought and osmotic stress tolerance of tobacco by increasing coniferyl alcohol content. Plant Physiol. Biochem. 155, 697–708 (2020). [ DOI ] [ PubMed ] [ Google Scholar ] 45. Moura J. C. M. S., Bonine C. A. V., Viana J. D. F., Dornelas M. C., Mazzafera P., Abiotic and biotic stresses and changes in the lignin content and composition in plants. J. Integr. Plant Biol. 52, 360–376 (2010). [ DOI ] [ PubMed ] [ Google Scholar ] 46. Tenhaken R., Cell wall remodeling under abiotic stress. Front. Plant Sci. 5, 771 (2015). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 47. Perkins M., Smith R. A., Samuels L., The transport of monomers during lignification in plants: Anything goes but how? Curr. Opin. Biotechnol. 56, 69–74 (2019). [ DOI ] [ PubMed ] [ Google Scholar ] 48. Miyagawa Y., Tobimatsu Y., Lam P. Y., Mizukami T., Sakurai S., Kamitakahara H., Takano T., Possible mechanisms for the generation of phenyl glycoside-type lignin-carbohydrate linkages in lignification with monolignol glucosides. Plant J. 104, 156–170 (2020). [ DOI ] [ PubMed ] [ Google Scholar ] 49. Lim E.-K., Li Y., Parr A., Jackson R., Ashford D. A., Bowles D. J., Identification of glucosyltransferase genes involved in sinapate metabolism and lignin synthesis in Arabidopsis . J. Biol. Chem. 276, 4344–4349 (2001). [ DOI ] [ PubMed ] [ Google Scholar ] 50. Lanot A., Hodge D., Jackson R. G., George G. L., Elias L., Lim E. K., Vaistij F. E., Bowles D. J., The glucosyltransferase UGT72E2 is responsible for monolignol 4- O -glucoside production in Arabidopsis thaliana . Plant J. 48, 286–295 (2006). [ DOI ] [ PubMed ] [ Google Scholar ] 51. Lanot A., Hodge D., Lim E. K., Vaistij F. E., Bowles D. J., Redirection of flux through the phenylpropanoid pathway by increased glucosylation of soluble intermediates. Planta 228, 609–616 (2008). [ DOI ] [ PubMed ] [ Google Scholar ] 52. Vermaas J. V., Dixon R. A., Chen F., Mansfield S. D., Boerjan W., Ralph J., Crowley M. F., Beckham G. T., Passive membrane transport of lignin-related compounds. Proc. Natl. Acad. Sci. U.S.A. 116, 23117–23123 (2019). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 53. Hubbard M. J., Cohen P., On target with a new mechanism for the regulation of protein phosphorylation. Trends Biochem. Sci. 18, 172–177 (1993). [ DOI ] [ PubMed ] [ Google Scholar ] 54. Pawson T., Scott J. D., Signaling through scaffold, anchoring, and adaptor proteins. Science 278, 2075–2080 (1997). [ DOI ] [ PubMed ] [ Google Scholar ] 55. Dammann C., Ichida A., Hong B., Romanowsky S. M., Hrabak E. M., Harmon A. C., Pickard B. G., Harper J. F., Subcellular targeting of nine calcium-dependent protein kinase isoforms from Arabidopsis . Plant Physiol. 132, 1840–1848 (2003). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 56. Miao Y., Laun T., Zimmermann P., Zentgraf U., Targets of the WRKY53 transcription factor and its role during leaf senescence in Arabidopsis . Plant Mol. Biol. 55, 853–867 (2004). [ DOI ] [ PubMed ] [ Google Scholar ] 57. Furihata T., Maruyama K., Fujita Y., Umezawa T., Yoshida R., Shinozaki K., Yamaguchi-Shinozaki K., Abscisic acid-dependent multisite phosphorylation regulates the activity of a transcription activator AREB1. Proc. Natl. Acad. Sci. U.S.A. 103, 1988–1993 (2006). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 58. Tang N., Zhang H., Li X. H., Xiao J. H., Xiong L. Z., Constitutive activation of transcription factor OsbZIP46 improves drought tolerance in rice. Plant Physiol. 158, 1755–1768 (2012). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 59. Liu C., Chen X., Ma P. A., Zhang S. K., Zeng C. Y., Jiang X. Y., Wang W. Q., Ethylene responsive factor MeERF72 negatively regulates Sucrose synthase 1 gene in Cassava. Int. J. Mol. Sci. 19, 1281 (2018). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 60. Zhang G. F., Liu W., Feng Y., Li D. Y., Li K. T., Sun Q. R., Zhai L. M., Wu T., Zhang X. Z., Xu X. F., Wang Y., Han Z. H., Ethylene response factors MbERF4 and MbERF72 suppress iron uptake in woody apple plants by modulating rhizosphere pH. Plant Cell Physiol. 61, 699–711 (2020). [ DOI ] [ PubMed ] [ Google Scholar ] 61. Li Y. H., Liu K., Tong G. L., Xi C., Liu J., Zhao H. P., Wang Y. D., Ren D. T., Han S. C., MPK3/MPK6-mediated phosphorylation of ERF72 positively regulates resistance to Botrytis cinerea through directly and indirectly activating the transcription of camalexin biosynthesis enzymes. J. Exp. Bot. 73, 413–428 (2022). [ DOI ] [ PubMed ] [ Google Scholar ] 62. Mao G. H., Meng X. Z., Liu Y. D., Zheng Z. Y., Chen Z. X., Zhang S. Q., Phosphorylation of a WRKY transcription factor by two pathogen-responsive MAPKs drives phytoalexin biosynthesis in Arabidopsis . Plant Cell 23, 1639–1653 (2011). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 63. Ogura N., Sasagawa Y., Ito T., Tameshige T., Kawai S., Sano M., Doll Y., Iwase A., Kawamura A., Suzuki T., Nikaido I., Sugimoto K., Ikeuchi M., WUSCHEL-RELATED HOMEOBOX 13 suppresses de novo shoot regeneration via cell fate control of pluripotent callus. Sci. Adv. 9, eadg6983 (2023). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 64. Ikeuchi M., Iwase A., Ito T., Tanaka H., Favero D. S., Kawamura A., Sakamoto S., Wakazaki M., Tameshige T., Fujii H., Hashimoto N., Suzuki T., Hotta K., Toyooka K., Mitsuda N., Sugimoto K., Wound-inducible WUSCHEL-RELATED HOMEOBOX 13 is required for callus growth and organ reconnection. Plant Physiol. 188, 425–441 (2022). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 65. Li E. Y., Bhargava A., Qiang W. Y., Friedmann M. C., Forneris N., Savidge R. A., Johnson L. A., Mansfield S. D., Ellis B. E., Douglas C. J., The class II KNOX gene KNAT7 negatively regulates secondary wall formation in Arabidopsis and is functionally conserved in Populus . New Phytol. 194, 102–115 (2012). [ DOI ] [ PubMed ] [ Google Scholar ] 66. Sakakibara K., Reisewitz P., Aoyama T., Friedrich T., Ando S., Sato Y., Tamada Y., Nishiyama T., Hiwatashi Y., Kurata T., Ishikawa M., Deguchi H., Rensing S. A., Werr W., Murata T., Hasebe M., Laux T., WOX13-like genes are required for reprogramming of leaf and protoplast cells into stem cells in the moss Physcomitrella patens . Development 141, 1660–1670 (2014). [ DOI ] [ PubMed ] [ Google Scholar ] 67. Petzold H. E., Rigoulot S. B., Zhao C. S., Chanda B., Sheng X. Y., Zhao M. Z., Jia X. Y., Dickerman A. W., Beers E. P., Brunner A. M., Identification of new protein-protein and protein-DNA interactions linked with wood formation in Populus trichocarpa . Tree Physiol. 38, 362–377 (2018). [ DOI ] [ PubMed ] [ Google Scholar ] 68. Zhong R. Q., McCarthy R. L., Lee C., Ye Z. H., Dissection of the transcriptional program regulating secondary wall biosynthesis during wood formation in poplar. Plant Physiol. 157, 1452–1468 (2011). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 69. Yoon J., Choi H., An G., Roles of lignin biosynthesis and regulatory genes in plant development. J. Integr. Plant Biol. 57, 902–912 (2015). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 70. Neale D. B., Kremer A., Forest tree genomics: Growing resources and applications. Nat. Rev. Genet. 12, 111–122 (2011). [ DOI ] [ PubMed ] [ Google Scholar ] 71. Wang X. L., Wang H. W., Liu S. X., Ferjani A., Li J. S., Yan J. B., Yang X. H., Qin F., Genetic variation in ZmVPP1 contributes to drought tolerance in maize seedlings. Nat. Genet. 48, 1233–1241 (2016). [ DOI ] [ PubMed ] [ Google Scholar ] 72. Blankenagel S., Eggels S., Frey M., Grill E., Bauer E., Dawid C., Fernie A. R., Haberer G., Hammerl R., Medeiros D. B., Ouzunova M., Presterl T., Russ V., Schäufele R., Schlüter U., Tardieu F., Urbany C., Urzinger S., Weber A. P. M., Schön C. C., Avramova V., Natural alleles of the abscisic acid catabolism gene ZmAbh4 modulate water use efficiency and carbon isotope discrimination in maize. Plant Cell 34, 3860–3872 (2022). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 73. Yan W. H., Liu H. Y., Zhou X. C., Li Q. P., Zhang J., Lu L., Liu T. M., Liu H. J., Zhang C. J., Zhang Z. Y., Shen G. J., Yao W., Chen H. X., Yu S. B., Xie W. B., Xing Y. Z., Natural variation in Ghd7.1 plays an important role in grain yield and adaptation in rice. Cell Res. 23, 969–971 (2013). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 74. Fang Y. Y., Wang D., Xiao L., Quan M. Y., Qi W. A., Song F. Y., Zhou J. X., Liu X., Qin S. T., Du Q. Z., Liu Q., El-Kassaby Y. A., Zhang D. Q., Allelic variation in transcription factor PtoWRKY68 contributes to drought tolerance in Populus . Plant Physiol. 193, 736–755 (2023). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 75. Huang R., Jin Z. Y., Zhang D. H., Li L. Z., Zhou J. X., Xiao L., Li P., Zhang M. J., Tian C. D., Zhang W. K., Zhong L. S., Quan M. Y., Zhao R., Du L., Liu L. J., Li Z. H., Zhang D. Q., Du Q. Z., Rare variations within the serine/arginine-rich splicing factor PtoRSZ21 modulate stomatal size to determine drought tolerance in Populus . New Phytol. 243, 1776–1794 (2024). [ DOI ] [ PubMed ] [ Google Scholar ] 76. Evans L. M., Slavov G. T., Rodgers-Melnick E., Martin J., Ranjan P., Muchero W., Brunner A. M., Schackwitz W., Gunter L., Chen J. G., Tuskan G. A., DiFazio S. P., Population genomics of Populus trichocarpa identifies signatures of selection and adaptive trait associations. Nat. Genet. 46, 1089–1096 (2014). [ DOI ] [ PubMed ] [ Google Scholar ] 77. Wang J., Ding J. H., Tan B. Y., Robinson K. M., Michelson I. H., Johansson A., Nystedt B., Scofield D. G., Nilsson O., Jansson S., Street N. R., Ingvarsson P. K., A major locus controls local adaptation and adaptive life history variation in a perennial plant. Genome Biol. 19, 72 (2018). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 78. Zhang Y., Yang C., Wang J., Wang L., Zhao Y., Zhu L., Sun L., Sun W., Zhu Y., Wu S., BioLadder: A bioinformatic platform primarily focused on proteomic data analysis. iMeta 3, e215 (2024). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 79. Xu W., Huang J., Wang P. L., Yang Y. P., Fu S. B., Ying Z., Zhou Z., Using widely targeted metabolomics profiling to explore differences in constituents of three Bletilla species. Sci. Rep. 14, 23873 (2024). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 80. Slobodyanyuk M., Bahcheli A. T., Klein Z. P., Bayati M., Strug L. J., Reimand J., Directional integration and pathway enrichment analysis for multi-omics data. Nat. Commun. 15, 5690 (2024). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 81. Du Q. Z., Wang B. W., Wei Z. Z., Zhang D. Q., Li B. L., Genetic diversity and population structure of Chinese White poplar ( Populus tomentosa ) revealed by SSR markers. J. Hered. 103, 853–862 (2012). [ DOI ] [ PubMed ] [ Google Scholar ] 82. Li P., Xiao L., Du Q. Z., Quan M. Y., Song Y. P., He Y. L., Huang W. X., Xie J. B., Lv C. F., Wang D., Zhou J. X., Li L. Z., Liu Q., El-Kassaby Y. A., Zhang D. Q., Genomic insights into selection for heterozygous alleles and woody traits in Populus tomentosa . Plant Biotechnol. J. 21, 2002–2018 (2023). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 83. Jiang Y., Wang D., Yang L., Cheng Y., Yang L., Zhao T., Zhang D., Zhou J., Li Z., Guan Y., Ren T., He Y., Du Q., Quan M., Allelic variants of IRKI contribute to photosynthetic efficiency by regulating rubisco activase in Populus . Plant Physiol. 199, kiaf465 (2025). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 84. Kang H. M., Sul J. H., S. K. Service, Zaitlen N. A., Kong S. Y., Freimer N. B., Sabatti C., Eskin E., Variance component model to account for sample structure in genome-wide association studies. Nat. Genet. 42, 348–354 (2010). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 85. Fick S. E., Hijmans R. J., WorldClim 2: New 1-km spatial resolution climate surfaces for global land areas. Int. J. Climatol. 37, 4302–4315 (2017). [ Google Scholar ] 86. Zomer R. J., Xu J. C., Trabucco A., Version 3 of the global aridity index and potential evapotranspiration database. Sci. Data 9, 409 (2022). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 87. Frichot E., Schoville S. D., Bouchard G., François O., Testing for associations between loci and environmental gradients using latent factor mixed models. Mol. Biol. Evol. 30, 1687–1699 (2013). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 88. Danecek P., Auton A., Abecasis G., Albers C. A., Banks E., DePristo M. A., Handsaker R. E., Lunter G., Marth G. T., Sherry S. T., McVean G., Durbin R., Grp G. P. A., The variant call format and VCFtools. Bioinformatics 27, 2156–2158 (2011). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 89. Mantel N., The detection of disease clustering and a generalized regression approach. Cancer Res. 27, 209–220 (1967). [ PubMed ] [ Google Scholar ] 90. Smonse P. E., Long J. C., Sokal R. R., Multiple regression and correlation extensions of the Mantel test of matrix correspondence. Syst. Zool. 35, 627–632 (1986). [ Google Scholar ] 91. J. Oksanen, F. G. Blanchet, M. Friendly, R. Kindt, P. Legendre, D. McGlinn, P. Minchin, R. B. O’Hara, G. Simpson, P. Solymos, M. H. H. Stevens, E. Szöcs, H. Wagner, Vegan community ecology package version 2.5-7 (2020). 92. P. Legendre, L. Legendre, Numerical Ecology (Elsevier, 2012), vol. 853. [ Google Scholar ] 93. Forester B. R., Lasky J. R., Wagner H. H., Urban D. L., Comparing methods for detecting multilocus adaptation with multivariate genotype-environment associations. Mol. Ecol. 27, 2215–2233 (2018). [ DOI ] [ PubMed ] [ Google Scholar ] 94. Ellis N., Smith S. J., Pitcher C. R., Gradient forests: Calculating importance gradients on physical predictors. Ecology 93, 156–168 (2012). [ DOI ] [ PubMed ] [ Google Scholar ] 95. Leinonen T., McCairns R. J. S., O’Hara R. B., Merilä J., Q(ST)-F(ST) comparisons: Evolutionary and ecological insights from genomic heterogeneity. Nat. Rev. Genet. 14, 179–190 (2013). [ DOI ] [ PubMed ] [ Google Scholar ] 96. Xu Z., Sun H., Zhang Z., Zhao Q. B., Olasege B. S., Li Q. M., Yue Y., Ma P. P., Zhang X. Z., Wang Q. S., Pan Y. C., Assessment of autozygosity derived from runs of homozygosity in jinhua pigs disclosed by sequencing data. Front. Genet. 10, 274 (2019). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 97. Yang J. A., Lee S. H., Goddard M. E., Visscher P. M., GCTA: A tool for genome-wide complex trait analysis. Am. J. Hum. Genet. 88, 76–82 (2011). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 98. Li M. X., Yeung J. M. Y., Cherny S. S., Sham P. C., Evaluating the effective numbers of independent tests and significant p-value thresholds in commercial genotyping arrays and public imputation reference datasets. Hum. Genet. 131, 747–756 (2012). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 99. Shin J.-H., Blay S., McNeney B., Graham J., LDheatmap: An R function for graphical display of pairwise linkage disequilibria between single nucleotide polymorphisms. J. Stat. Softw. 16, 1–9 (2006). [ Google Scholar ] 100. Wang D., Quan M. Y., Qin S. T., Fang Y. Y., Xiao L., Qi W., Jiang Y. S., Zhou J. X., Gu M. Y., Guan Y. C., Du Q. Z., Liu Q., El-Kassaby Y. A., Zhang D. Q., Allelic variations of WAK106-E2Fa-DPb1-UGT74E2 module regulate fibre properties in Populus tomentosa . Plant Biotechnol. J. 22, 970–986 (2024). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 101. Quan M. Y., Liu X., Du Q. Z., Xiao L., Lu W. J., Fang Y. Y., Li P., Ji L., Zhang D. Q., Genome-wide association studies reveal the coordinated regulatory networks underlying photosynthesis and wood formation in Populus . J. Exp. Bot. 72, 5372–5389 (2021). [ DOI ] [ PubMed ] [ Google Scholar ] 102. Flexas J., Ribas-Carbó M., Bota J., Galmés J., Henkle M., Martínez-Cañellas S., Medrano H., Decreased Rubisco activity during water stress is not induced by decreased relative water content but related to conditions of low stomatal conductance and chloroplast CO 2 concentration. New Phytol. 172, 73–82 (2006). [ DOI ] [ PubMed ] [ Google Scholar ] 103. Zhao F. J., Ma Y. B., Zhu Y. G., Tang Z., McGrath S. P., Soil contamination in China: Current status and mitigation strategies. Environ. Sci. Technol. 49, 750–759 (2015). [ DOI ] [ PubMed ] [ Google Scholar ] 104. Brodribb T. J., Holbrook N. M., Stomatal closure during leaf dehydration, correlation with other leaf physiological traits. Plant Physiol. 132, 2166–2173 (2003). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 105. Haberstroh S., Scarpa F., Seeger S., Christen A., Werner C., Continuous stem water potential measurements of a diffuse-porous tree species offer new insights into tree water relations. Ecohydrology 18, e2761 (2025). [ Google Scholar ] 106. Sluiter A., Hames B., Ruiz R., Scarlata C., Sluiter J., Templeton D., Crocker D., Determination of structural carbohydrates and lignin in biomass. Lab. Anal. Proced. 1617, 1–15 (2008). [ Google Scholar ] 107. Lu S. F., Li Q. Z., Wei H. R., Chang M. J., Tunlaya-Anukit S., Kim H., Liu J., Song J. Y., Sun Y. H., Yuan L. C., Yeh T. F., Peszlen I., Ralph J., Sederoff R. R., Chiang V. L., Ptr-miR397a is a negative regulator of laccase genes affecting lignin content in Populus trichocarpa . Proc. Natl. Acad. Sci. U.S.A. 110, 10848–10853 (2013). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 108. Wen J. L., Sun S. L., Xue B. L., Sun R. C., Recent advances in characterization of lignin polymer by solution-state nuclear magnetic resonance (NMR) methodology. Materials 6, 359–391 (2013). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 109. Scholz A., Klepsch M., Karimi Z., Jansen S., How to quantify conduits in wood? Front. Plant Sci. 4, 56 (2013). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 110. Bentham A. R., De la Concepcion J. C., Benjumea J. V., Kourelis J., Jones S., Mendel M., Stubbs J., Stevenson C. E. M., Maidment J. H. R., Youles M., Zdrzalek R., Kamoun S., Banfield M. J., Allelic compatibility in plant immune receptors facilitates engineering of new effector recognition specificities. Plant Cell 35, 3809–3827 (2023). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 111. Deng C. H., Deng Y. H., Wang B., Yang X., Gas chromatography–mass spectrometry method for determination of phenylalanine and tyrosine in neonatal blood spots. J. Chromatogr. B 780, 407–413 (2002). [ DOI ] [ PubMed ] [ Google Scholar ] 112. Reyt G., Ramakrishna P., Salas-González I., Fujita S., Love A., Tiemessen D., Lapierre C., Morreel K., Calvo-Polanco M., Flis P., Geldner N., Boursiac Y., Boerjan W., George M. W., Castrillo G., Salt D. E., Two chemically distinct root lignin barriers control solute and water balance. Nat. Commun. 12, 2320 (2021). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 113. Bartlett A., O’Malley R. C., Huang S. S. C., Galli M., Nery J. R., Gallavotti A., Ecker J. R., Mapping genome-wide transcription-factor binding sites using DAP-seq. Nat. Protoc. 12, 1659–1672 (2017). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 114. Langmead B., Salzberg S. L., Fast gapped-read alignment with Bowtie 2. Nat. Methods 9, 357–359 (2012). [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials Figs. S1 to S16 Legends for data S1 to S36 sciadv.aee2172_sm.pdf (7.5MB, pdf) Data S1 to S36 sciadv.aee2172_data_s1_to_s36.zip (5.9MB, zip) Data Availability Statement The raw data of genome resequencing of 302 P. tomentosa accessions, Y2H-seq data of PtoCPK3, DAP-seq data of PtoERF72, and all RNA-seq data of P. tomentosa “1316,” PtoCPK3 -OE, ptocpk3 -KD, PtoERF2 -OE, and ptoerf72 -KD lines have been submitted to the Genome Sequence Archive (GSA) in the BIG Data Center under accession numbers CRA000903 and PRJCA050561 (accessible at https://bigd.big.ac.cn/gsa ). The mass spectrometry proteomics and phosphorylated proteome data have been deposited to the ProteomeXchange Consortium via the iProX partner repository with the dataset identifier PXD070526 (accessible at https://proteomecentral.proteomexchange.org ). The CDS data generated in this study are available in the GenBank databases under the accession numbers PX597595, PX597596, PX597593, PX597594, PX597591, and PX961919 for PtoCPK3 ( PtoCPK3 II ), PtoCPK3 DD , PtoERF72 ( PtoERF72 AA ), PtoERF72 CC , PtoWOX13b , and PtoUGT72AZ2, respectively (accessible at https://www.ncbi.nlm.nih.gov/ ). All data and code needed to evaluate and reproduce the results in the paper are present in the paper and/or the Supplementary Materials. The source codes for this study have been deposited in the Zenodo repository (accessible at https://zenodo.org/records/18886819 ). Alternatively, our GitHub repository can be reached at https://github.com/jiangyongsen121/analysis . Requests for PtoCPK3 , PtoERF72 , PtoWOX13b , or PtoUGT72AZ2 constructs and plasmids generated in this study can be made available by contacting the corresponding author ( DeqiangZhang@bjfu.edu.cn ). Articles from Science Advances are provided here courtesy of American Association for the Advancement of Science ACTIONS View on publisher site PDF (5.6 MB) Cite Collections Permalink RESOURCES Similar articles Cited by other articles Links to NCBI Databases Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top

Related documents

• Natural variation in &lt;i&gt;PtoCPK3&lt;/i&gt; governs drought tolerance by orchestrating xylem remodeling and lignin metabolism in &lt;i&gt;Populus&lt;/i&gt;#266471
• GCN5 restricts PRX-dependent lignin deposition during salt stress in Arabidopsis#138357
• GCN5 restricts PRX-dependent lignin deposition during salt stress in Arabidopsis#134296
• Natural variation in PtoCPK3 governs drought tolerance by orchestrating xylem remodeling and lignin metabolism in Populus#257195
• Advances in Lignin Accumulation and Regulation in Horticultural Crops#70493