Genome-wide identification and expression analysis of the LBD gene family in Jasminum sambac Authors: Ji Huang, Qun Su, Mingyan Sun, Yiqing Zheng, Yanhua Zeng, Zide Jiang, Chunniu Li, Zhaoyang Bu Published: 2026 ABSTRACT Jasminum sambac (L.) Ait is widely cultivated for its ornamental, medicinal, and economic value. However, infection by Sclerotium rolfsii often causes stem and root rot, severely affecting yield and quality. In this study, transcriptome and hormone profiling analyses, combined with a systematic characterization of the LBD gene family, were performed to investigate the molecular responses of J. sambac during the early stages of S. rolfsii infection. The results showed that four key genes (DjLBD09, DjLBD13, DjLBD17, and DjLBD28) exhibited significant differential expression at the early infection stage, and their expression patterns detected by qPCR were highly consistent with RNA-seq data, confirming the reliability of the transcriptome analysis. Integrated metabolomic and transcriptomic analyses further suggested that these LBD genes may participate in regulating signal transduction and defense-related metabolic pathways, thereby playing central roles in early defense responses. This study not only elucidates the molecular mechanisms underlying early defense against S. rolfsii in J. sambac but also provides potential targets and theoretical support for genetic improvement of disease resistance.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-026-40654-9. Introduction Jasminum sambac (L.) Aiton is a major ornamental species in the family Oleaceae and genus Jasminum. It carries ornamental, medicinal, and economic value: a renowned raw material for scented tea and fragrances, with broad applications in horticulture and wellness, hence the epithet “the first fragrance under heaven”1. China has a long history of jasmine cultivation, with major production areas in Guangxi, Fujian, Sichuan, and Yunnan; among them, Heng County (Guangxi) and Qianwei County (Sichuan) are typical concentrated production regions2. However, as the industry has expanded, jasmine production has come under serious threat from diseases, especially southern blight caused by Sclerotium rolfsii Sacc.3 Southern blight is a destructive, soil-borne fungal disease with a very broad host range, infecting more than 500 plant species including jasmine, peanut, tomato, and sunflower3. In the main jasmine-growing regions of southern China, its incidence has risen year by year and has become a key constraint on yield and quality4. The disease typically initiates at the stem base, where white, web-like mycelia develop and brown sclerotia form; it causes basal/root rot and can lead to whole-plant death2. Because sclerotia can persist for long periods in soil or crop residues and epidemics readily occur under hot, humid conditions, control is challenging5. Field surveys reveal substantial variation in cultivar resistance, offering opportunities for the exploitation of resistant germplasm and resistance breeding6,7. At present, no chemical agents provide highly effective and durable control of southern blight8, and biological control is becoming an important alternative. Trichoderma harzianum shows significant efficacy against jasmine southern blight9, and Trichoderma-based biocontrol formulations not only suppress disease but also improve soil microbial community structure and enhance rhizosphere stability10. In parallel, agronomic measures (e.g., crop rotation and soil management) and integrated management strategies have been proposed to curb epidemics8. Pathogen-secreted virulence factors—such as oxalic acid and polygalacturonases—also play crucial roles in pathogenesis, deepening our understanding of disease mechanisms11,12. Beyond external control, intrinsic plant defenses are critical for resisting infection. Multiple hormone signaling pathways are involved, among which jasmonic acid (JA) and salicylic acid (SA) play central roles in induced resistance. For example, during infection of konjac, JA and SA levels rise markedly alongside increased expression of related genes, highlighting the importance of both pathways13. Similar findings in other crops point to a general role for JA/SA signaling in resistance to soil-borne diseases14. With advances in genomics and transcriptomics, attention has increasingly focused on transcription factors in plant stress responses. The LBD (Lateral Organ Boundaries Domain) gene family comprises plant-specific transcription factors that harbor a highly conserved LOB domain and specifically recognize the “GCGGCG” motif to regulate downstream gene expression15. LBD members participate broadly in the development of lateral organs (lateral roots, branches, leaves, floral organs) and also function in stress responses, secondary growth, regeneration, and metabolic regulation16–19. Notably, one of the first functional studies of LBD genes in plant–fungus interactions demonstrated that AtLBD20 from Arabidopsis acts as a negative regulator of jasmonate-mediated defense responses and increases susceptibility to the necrotrophic root-infecting fungus Fusarium oxysporum20. In the broader context of necrotrophic pathogens, transcriptional reprogramming controlled by transcription factors has been recognized as a key layer of defense because classical R-gene-mediated resistance is often ineffective against necrotrophs21,22. Together these findings underscore the value of targeting transcription factors such as LBDs for improving resistance to necrotrophic pathogens. Based on our systematic analysis of the jasmine LBD family, we hypothesize that certain members act early during southern blight infection by modulating signal transduction and defense-related metabolic pathways. In summary, southern blight is now a major constraint on the sustainable development of the jasmine industry. Although conventional agronomic and chemical control measures can mitigate losses to some extent, durable solutions will rely on breeding for resistance and deeper elucidation of the underlying molecular mechanisms. Focusing on the early defense response of jasmine to southern blight, this study integrates transcriptomic and metabolomic analyses with a systematic characterization of the LBD gene family to delineate regulatory patterns and provide a theoretical basis and potential targets for resistance breeding in jasmine. Results and analysis Identification and basic properties We identified 38, 39, and 36 LBD genes in ‘Hutou’ (Ht), single-petal (Sj), and double-petal (Dj) jasmine, respectively, by combining genome annotation, chromosomal mapping, and BLAST validation. Their general characteristics are summarized in Table 1. Protein lengths spanned ~ 142–452 aa, with theoretical pI values centered around ~ 6.5 and most members showing an instability index > 40, suggesting weakly acidic and largely unstable proteins (full statistics in Table S1). Gene names were assigned in chromosomal order within each genome.Table 1Summary of physicochemical properties and predicted subcellular localization of LBD family proteins in jasmine.SpeciesNo. of LBD genesProtein length (aa)MW (kDa)Theoretical pIInstability indexAliphatic indexGRAVYPredicted localization (counts)Hutou jasmine (HtLBD)38152–316 (mean ≈ 215)16.9–35.2 (mean ≈ 24.6)4.68–9.28 (mean ≈ 6.65)36.3–81.9 (mostly > 40, unstable)58–94–0.80 to 0.00 (hydrophilic)Nucleus (28), Chloroplast (3), Extracellular (4), Mitochondrion (1), Other (2)Single-petal jasmine (SjLBD)39152–452 (mean ≈ 226)16.9–50.9 (mean ≈ 26.7)4.61–9.23 (mean ≈ 6.51)36.3–77.059–94–0.80 to –0.01Nucleus (29), Chloroplast (3), Extracellular (3), Mitochondrion (2), Other (2)Double-petal jasmine (DjLBD)36142–315 (mean ≈ 209)16.9–35.1 (mean ≈ 25.1)4.61–9.23 (mean ≈ 6.54)36.3–72.758–89–0.80 to –0.01Nucleus (27), Chloroplast (3), Extracellular (3), Mitochondrion (2), Other (1)MW, molecular weight; GRAVY, grand average of hydropathicity. Table 1Summary of physicochemical properties and predicted subcellular localization of LBD family proteins in jasmine.SpeciesNo. of LBD genesProtein length (aa)MW (kDa)Theoretical pIInstability indexAliphatic indexGRAVYPredicted localization (counts)Hutou jasmine (HtLBD)38152–316 (mean ≈ 215)16.9–35.2 (mean ≈ 24.6)4.68–9.28 (mean ≈ 6.65)36.3–81.9 (mostly > 40, unstable)58–94–0.80 to 0.00 (hydrophilic)Nucleus (28), Chloroplast (3), Extracellular (4), Mitochondrion (1), Other (2)Single-petal jasmine (SjLBD)39152–452 (mean ≈ 226)16.9–50.9 (mean ≈ 26.7)4.61–9.23 (mean ≈ 6.51)36.3–77.059–94–0.80 to –0.01Nucleus (29), Chloroplast (3), Extracellular (3), Mitochondrion (2), Other (2)Double-petal jasmine (DjLBD)36142–315 (mean ≈ 209)16.9–35.1 (mean ≈ 25.1)4.61–9.23 (mean ≈ 6.54)36.3–72.758–89–0.80 to –0.01Nucleus (27), Chloroplast (3), Extracellular (3), Mitochondrion (2), Other (1)MW, molecular weight; GRAVY, grand average of hydropathicity. Summary of physicochemical properties and predicted subcellular localization of LBD family proteins in jasmine. MW, molecular weight; GRAVY, grand average of hydropathicity. Chromosomal distribution of jasmine LBD genes LBD loci were unevenly distributed across 12 chromosomes in each genome: Ht peaked on chr1 (6 genes), Sj on chr3 (8), and Dj on chr1 (5), with at least one chromosome in each genome harboring a single locus. Because the spatial patterns do not affect downstream conclusions, detailed karyoplots are relegated to Fig. S1. Phylogenetic analysis of the LBD gene family in jasmine A maximum-likelihood tree built from jasmine and Arabidopsis LBD proteins resolved two major classes (Class I, Class II) and seven subfamilies (Fig. 1). Class I dominated (118/156; 75.6%), while subfamilies ranged from 13 (II-a) to 35 (I-c) members. All subfamilies contained representatives from Ht/Sj/Dj and Arabidopsis, but lineage representation was uneven, consistent with differential expansion.Fig. 1Phylogeny of LBD proteins from three jasmine types and Arabidopsis. Phylogeny of LBD proteins from three jasmine types and Arabidopsis. A maximum-likelihood phylogenetic tree was constructed using full-length LBD protein sequences from single-petal jasmine, double-petal jasmine, ‘Hutou’ jasmine, and Arabidopsis thaliana. Different color branches denote distinct LBD subfamilies (Class I and Class II). The tree reveals conserved clustering patterns among jasmine species and their evolutionary relationship with Arabidopsis homologs, indicating a high degree of sequence conservation and diversification within the LBD family. Gene structure and conserved domain analysis of jasmine LBD genes Exon–intron architectures were simple (mostly 1–2 exons; a minority with ≥ 3 in Sj), indicating compact gene models (Fig. 2). All proteins harbored the canonical LBD domain, whereas a few contained additional domains (e.g., COG6, PRK11815), hinting at functional diversification. MEME analysis showed motifs 1 and 2 to be universally present, with diagnostic motifs restricted to specific classes or subfamilies (e.g., Class II–specific motifs; I-c–specific motifs), and most motifs concentrated toward the C terminus, suggesting higher C-terminal conservation and N-terminal divergence. This N-terminal divergence likely reflects relaxed evolutionary constraint on regulatory or interaction regions outside the conserved LOB domain, allowing sequence diversification that may underlie functional specificity among LBD members.Fig. 2Exon–intron maps, domain architecture, and motif composition of jasmine LBDs. (A) Hutou jasmine, (B) single-petal jasmine, and (C) double-petal jasmine. Exon–intron maps, domain architecture, and motif composition of jasmine LBDs. (A) Hutou jasmine, (B) single-petal jasmine, and (C) double-petal jasmine. Cis-acting element and transcription factor binding site (TFBS) analysis of jasmine LBD promoters Across the 2-kb promoter regions of Jasminum sambac LBD genes, we detected 22 representative cis-elements associated with light, hormone, stress, and developmental responses, as well as diverse transcription factor binding sites (TFBSs). Light-responsive motifs were nearly ubiquitous, while approximately 72–77% of genes carried ABA-responsive elements, 71–72% contained defense or stress-related motifs, and about half included drought-responsive MYB sites in each genome. Moreover, 18 TF families were identified in promoter regions, with BBR-BPC, Dof, and MIKC_MADS showing the highest binding frequencies. Although Ht, Sj, and Dj shared most TF families, several lineage-specific patterns were observed—for instance, WRKY binding sites appeared uniquely in Dj. The wide variation in TFBS number per gene suggests member-specific transcriptional regulation. Detailed element-by-gene matrices and TFBS heatmaps are provided in Figs. S2 and S3, respectively. Collinearity (synteny) analysis of the jasmine LBD gene family Intra-genome analyses detected 12 duplicated pairs in Ht, 9 in Sj, and 7 in Dj, indicating modest family expansion mainly via segmental/tandem duplication. Inter-genome synteny yielded 55 (Ht–Sj), 52 (Ht–Dj), and 48 (Sj–Dj) orthologous pairs (Fig. 3), supporting conserved relationships among cultivars(Full intra-/cross-species panels in Fig. S4). Cross-species macrosynteny with Arabidopsis and rice further corroborated conservation (summary edges shown below; full panels in supplement).Fig. 3Synteny summary among the three jasmine genomes (single consolidated chord/links panel). Genome-wide collinearity analysis was performed using TBtools based on MCScanX results. Colored links represent homologous LBD gene pairs among single-petal jasmine, double-petal jasmine, and ‘Hutou’ jasmine. The outer circle indicates chromosomal positions, while the inner chords illustrate intergenomic syntenic relationships. Extensive cross-links reflect strong collinearity among the three genomes, supporting conserved genomic organization and shared evolutionary origins of the LBD family. Synteny summary among the three jasmine genomes (single consolidated chord/links panel). Genome-wide collinearity analysis was performed using TBtools based on MCScanX results. Colored links represent homologous LBD gene pairs among single-petal jasmine, double-petal jasmine, and ‘Hutou’ jasmine. The outer circle indicates chromosomal positions, while the inner chords illustrate intergenomic syntenic relationships. Extensive cross-links reflect strong collinearity among the three genomes, supporting conserved genomic organization and shared evolutionary origins of the LBD family. Ka/Ks analysis of syntenic LBD gene pairs revealed that all calculable ratios were < 1, indicating purifying selection as the predominant evolutionary force acting on the jasmine LBD family. Specifically, HtLBD pairs showed Ka/Ks values ranging from 0.07 to 0.23, SjLBD pairs from 0.09 to 0.29, and DjLBD pairs from 0.15 to 0.28. These results suggest that the expansion of the LBD family in jasmine is largely driven by strong functional constraints rather than positive selection (Table S2). Validation of infection and early-stage symptoms in jasmine stems To verify the establishment and early progression of Sclerotium rolfsii infection, Jasminum sambac stems were photographed at 48 h and 96 h post-inoculation (Fig. 4).Fig. 4Validation of Sclerotium rolfsii infection and early‐stage symptom development in Jasminum sambac stems. Representative images showing the early progression of infection in inoculated jasmine stems. (A-1, A-2) 48 h post-inoculation: no visible mycelial growth was observed on the stem surface, though slight tissue discoloration appeared around the inoculation sites, indicating the onset of internal colonization. (B-1, B-2) 96 h post-inoculation: infected areas exhibited deeper browning beneath the cotton wrap and subtle surface changes, suggesting further pathogen progression and early host responses without prominent lesion formation. Validation of Sclerotium rolfsii infection and early‐stage symptom development in Jasminum sambac stems. Representative images showing the early progression of infection in inoculated jasmine stems. (A-1, A-2) 48 h post-inoculation: no visible mycelial growth was observed on the stem surface, though slight tissue discoloration appeared around the inoculation sites, indicating the onset of internal colonization. (B-1, B-2) 96 h post-inoculation: infected areas exhibited deeper browning beneath the cotton wrap and subtle surface changes, suggesting further pathogen progression and early host responses without prominent lesion formation. At 48 h, no visible mycelial growth was observed on the stem surface, while slight tissue discoloration appeared around the inoculation sites, suggesting the onset of internal colonization. By 96 h, the infected areas exhibited deeper browning beneath the cotton wrap, accompanied by subtle surface changes, indicating further fungal progression and early host responses without prominent lesion formation. These observations suggest that the sampled tissues represented the transition from initial, largely asymptomatic infection (48 h) to an intensified but still early phase of pathogen development (96 h), providing an appropriate window for capturing dynamic defense responses prior to visible disease onset. Having confirmed successful inoculation and the early infection status of Jasminum sambac stems, we next sought to elucidate the molecular basis underlying the host response to Sclerotium rolfsii. Because defense activation involves extensive transcriptional reprogramming and metabolic adjustments, an integrated transcriptomic and metabolomic approach was employed to capture coordinated changes in gene expression and metabolite accumulation. This combined strategy enables a systems-level view of the temporal dynamics of jasmine’s defense network, from early signaling events to downstream metabolic reconfiguration, providing a foundation for subsequent multivariate analyses such as PCA. Endogenous hormone metabolite analysis in jasmine stems after infection with S. rolfsii Principal component and OPLS–DA analyses of metabolomic profiles To capture the global metabolic shifts in Jasminum sambac stems during Sclerotium rolfsii infection, multivariate statistical analyses were performed on quantified hormone metabolites. Principal component analysis (PCA) showed that PC1 (48.85%) and PC2 (14.98%) together explained 63.83% of the total variance (Fig. 5). Samples collected at 0 h were clearly separated from those at 48 h and 96 h, while the latter two clustered more closely, indicating a strong early-stage metabolic response followed by partial stabilization.Fig. 5PCA of endogenous hormone metabolites in jasmine stems at different infection stages. PCA score plot of quantified hormone metabolites from Jasminum sambac stems at 0, 48, and 96 h post-inoculation with Sclerotium rolfsii. Percent variance explained by PC1 and PC2 is indicated. Points represent biological replicates. PCA of endogenous hormone metabolites in jasmine stems at different infection stages. PCA score plot of quantified hormone metabolites from Jasminum sambac stems at 0, 48, and 96 h post-inoculation with Sclerotium rolfsii. Percent variance explained by PC1 and PC2 is indicated. Points represent biological replicates. Consistently, orthogonal partial least squares–discriminant analysis (OPLS–DA) further confirmed distinct metabolic profiles among the three time-point comparisons (Fig. S5). In A-1/A-2 (0 h vs. 48 h), samples were well separated, and permutation validation supported model robustness (R2Y = 0.989, Q2 = 0.856), reflecting pronounced metabolic reprogramming during early infection. In C-1/C-2 (0 h vs. 96 h), clear separation persisted (R2Y = 0.996, Q2 = 0.897), suggesting sustained metabolic alterations. By contrast, in B-1/B-2 (48 h vs. 96 h), the difference was comparatively weaker (R2Y = 0.992, Q2 = 0.802), implying that after the 48 h peak, the metabolome tended toward stabilization. Identification and KEGG pathway analysis of differential metabolites To further characterize the metabolic reprogramming of Jasminum sambac stems following Sclerotium rolfsii infection, differential metabolites were identified and functionally annotated through pathway enrichment analyses. Volcano plots revealed extensive metabolic reprogramming during the early infection phase, with numerous significantly upregulated metabolites detected at 48 h (DJh2) relative to 0 h (DJh1), while fewer were observed at 96 h (DJh3) (Fig. S6). This pattern indicates a burst of metabolite accumulation during the initial defense activation, followed by a gradual decline in overall responsiveness. At 48 h (DJh2), numerous metabolites showed significant accumulation compared with 0 h (DJh1), reflecting strong early metabolic activation. By 96 h (DJh3), many of these changes persisted but with reduced amplitude, and only minor differences were detected between DJh2 and DJh3, indicating that the system tended toward metabolic stabilization. Heatmaps further confirmed time-dependent clustering, with strong induction of hormone-related metabolites such as JA, OPDA, MEJA, and IAA at 48 h (DJh2), followed by low induction at 96 h (DJh3), consistent with the attenuation of early defense responses (Fig. 6A).Fig. 6Cluster heatmap and KEGG pathway enrichment of differentially accumulated metabolites (DAMs) in Jasminum sambac young stems infected by Sclerotium rolfsii. (A) Hierarchical clustering heatmap showing relative abundance patterns of DAMs across three time points (0 h, 48 h, and 96 h) following inoculation. Color intensity indicates metabolite abundance (log₂-scaled and normalized). (B) KEGG pathway enrichment analysis of DAMs, presented as bubble plots: (B-1) 0 h vs. 48 h, (B-2) 0 h vs. 96 h, and (B-3) 48 h vs. 96 h. Bubble size represents the number of enriched metabolites within each pathway, and bubble color denotes enrichment significance (P value). Major pathways include energy metabolism, amino-acid metabolism, α-linolenic acid metabolism, and secondary-metabolite biosynthesis. KEGG pathway enrichment was performed based on KEGG annotations23–25. Cluster heatmap and KEGG pathway enrichment of differentially accumulated metabolites (DAMs) in Jasminum sambac young stems infected by Sclerotium rolfsii. (A) Hierarchical clustering heatmap showing relative abundance patterns of DAMs across three time points (0 h, 48 h, and 96 h) following inoculation. Color intensity indicates metabolite abundance (log₂-scaled and normalized). (B) KEGG pathway enrichment analysis of DAMs, presented as bubble plots: (B-1) 0 h vs. 48 h, (B-2) 0 h vs. 96 h, and (B-3) 48 h vs. 96 h. Bubble size represents the number of enriched metabolites within each pathway, and bubble color denotes enrichment significance (P value). Major pathways include energy metabolism, amino-acid metabolism, α-linolenic acid metabolism, and secondary-metabolite biosynthesis. KEGG pathway enrichment was performed based on KEGG annotations23–25. KEGG pathway enrichment analysis showed that differential metabolites were primarily associated with fatty acid metabolism, phytohormone signaling, and secondary-metabolite biosynthesis. At 48 h, the dominant responses centered on α-linolenic acid metabolism and hormone signaling pathways; at 96 h, enrichment extended to phenylpropanoid and carotenoid biosynthesis; by 48–96 h, fewer enriched pathways were detected, consistent with metabolic convergence during late-stage infection (Fig. 6B). Transcriptomic analysis of jasmine stems after infection by Sclerotium rolfsii Transcriptome profiling and functional enrichment of differentially expressed genes (DEGs) Transcriptome analysis revealed dynamic gene expression changes during Sclerotium rolfsii infection in Jasminum sambac. Across the three pairwise comparisons, 8,222 DEGs were detected in 0 h vs. 48 h (4,546 downregulated and 3,676 upregulated), 8,370 in 0 h vs. 96 h (4,654 downregulated and 3,716 upregulated), and only 620 in 48 h vs. 96 h (Table 2). These results indicate that major transcriptional reprogramming occurred within the first 48 h post-infection, characterized by extensive induction and repression of defense- and metabolism-related genes, whereas expression changes markedly declined thereafter, suggesting transition to a stabilized regulatory phase.Table 2Summary statistics of differentially expressed genes.ComparisonTotalDownregulatedUpregulated0 h vs. 48 h8,2224,5463,6760 h vs. 96 h8,3704,6543,71648 h vs. 96 h620274346 Table 2Summary statistics of differentially expressed genes.ComparisonTotalDownregulatedUpregulated0 h vs. 48 h8,2224,5463,6760 h vs. 96 h8,3704,6543,71648 h vs. 96 h620274346 Summary statistics of differentially expressed genes. Gene Ontology (GO) enrichment analysis revealed stage-specific functional patterns (Fig. 7A). In the early stage (0 h vs. 48 h), DEGs were predominantly enriched in photosynthesis-related terms such as “chloroplast thylakoid membrane” and “photosynthesis.” In the mid-stage (0 h vs. 96 h), enrichment expanded to “secondary metabolic process” and “transmembrane transporter activity,” while in the late stage (48 h vs. 96 h), fewer DEGs were detected, mainly associated with “cell wall biosynthesis” and “phenylpropanoid metabolism,” reflecting a shift toward structural remodeling. The overall GO enrichment bubble plot is shown in the main text (Fig. 7A), and detailed pairwise plots are provided in the Supplementary Materials (Fig. S7).Fig. 7GO and KEGG enrichment analyses of differentially expressed genes (DEGs) in Jasminum sambac young stems infected by Sclerotium rolfsii. (A) Multi-comparison Gene Ontology (GO) enrichment bubble plot showing significantly enriched biological processes, cellular components, and molecular functions among DEGs from pairwise comparisons (0 h vs. 48 h, 0 h vs. 96 h, and 48 h vs. 96 h). The x-axis represents the rich factor, bubble size indicates the number of DEGs, and bubble color denotes enrichment significance (Q-value). (B) Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment of DEGs across the same comparisons. Bubble size represents the number of DEGs assigned to each pathway, and bubble color indicates enrichment significance (Q-value). Prominently enriched pathways include plant hormone signal transduction, phenylpropanoid biosynthesis, and α-linolenic acid metabolism. KEGG pathway enrichment was performed based on KEGG annotations23–25. GO and KEGG enrichment analyses of differentially expressed genes (DEGs) in Jasminum sambac young stems infected by Sclerotium rolfsii. (A) Multi-comparison Gene Ontology (GO) enrichment bubble plot showing significantly enriched biological processes, cellular components, and molecular functions among DEGs from pairwise comparisons (0 h vs. 48 h, 0 h vs. 96 h, and 48 h vs. 96 h). The x-axis represents the rich factor, bubble size indicates the number of DEGs, and bubble color denotes enrichment significance (Q-value). (B) Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment of DEGs across the same comparisons. Bubble size represents the number of DEGs assigned to each pathway, and bubble color indicates enrichment significance (Q-value). Prominently enriched pathways include plant hormone signal transduction, phenylpropanoid biosynthesis, and α-linolenic acid metabolism. KEGG pathway enrichment was performed based on KEGG annotations23–25. KEGG pathway enrichment further supported these temporal trends (Fig. 7B). In the early stage, DEGs were significantly enriched in secondary metabolism, sugar metabolism, hormone signaling, and photosynthesis pathways. In the mid-stage, enrichment extended to stress-related pathways such as glutathione metabolism, while retaining secondary metabolism and hormone signaling. By the late stage, enrichment was limited to phenylpropanoid, diterpenoid, and cuticle biosynthesis pathways, consistent with structural strengthening. Notably, secondary metabolism, hormone signaling, and photosynthesis pathways were consistently enriched throughout all stages, highlighting their central roles in the defensive reprogramming of J. sambac. The overall KEGG enrichment bubble plot is presented in the main text (Fig. 7B), with detailed pairwise plots in the Supplementary Materials (Fig. S8). To further link genome-wide transcriptional responses with the specific transcription factor family under investigation, we next examined the expression dynamics of LBD genes in the RNA-Seq dataset. Expression profiles of LBD genes in RNA-Seq To further associate genome-wide transcriptional reprogramming with the transcription factor family of interest, we examined the expression dynamics of the LBD gene family using RNA-Seq data across the three infection stages (0 h, 48 h, and 96 h). FPKM values for all 37 annotated DjLBD genes were extracted from the transcriptome dataset, and their log₂ fold changes and regulation status were compiled from differential expression analyses (48 h vs. 0 h, 96 h vs. 0 h, and 96 h vs. 48 h). The resulting Table 3—Expression levels and differential expression of LBD genes in Jasminum sambac stems after Sclerotium rolfsii infection—was generated by integrating the FPKM expression matrix with the DESeq2-derived differential expression list, thus providing a comprehensive summary of both quantitative expression levels and statistical regulation trends for all LBD family members.Table 3Expression levels and differential expression of LBD genes in Jasminum sambac stems after Sclerotium rolfsii infection.LBD_nameRaw_IDFPKM_0hFPKM_48hFPKM_96hlog2FC_48h_vs_0hreg_48h _vs_0hlog2FC_96h_vs_0hreg_96h_vs_0hlog2FC_96h_vs_48hreg_96h_vs_48hDjLBD01DJ005920.540.190.34nsnsnsDjLBD02DJ008860.410.130.37nsnsnsDjLBD03DJ009130.124.8726.955.43↑8.0↑nsDjLBD04DJ009147.767.713.65nsnsnsDjLBD05DJ01079nsnsnsDjLBD06DJ241200.060.00.09nsnsnsDjLBD07DJ248254.660.80.23−2.49↓−4.19↓nsDjLBD08DJ254330.430.390.15nsnsnsDjLBD09DJ2934128.482.810.93−3.26↓−4.81↓−1.55↓DjLBD10DJ293810.360.00.04nsnsnsDjLBD11DJ295558.5610.3322.49ns1.53↑nsDjLBD12DJ30834nsnsnsDjLBD13DJ037990.0831.7319.348.96↑8.3↑nsDjLBD14DJ0484810.1313.9720.12nsnsnsDjLBD15DJ053536.021.320.45−2.15↓−3.61↓nsDjLBD16DJ058362.029.4415.872.36↑3.19↑nsDjLBD17DJ0691325.7158.18422.232.67↑4.16↑1.48↑DjLBD18DJ0701439.0521.5722.59nsnsnsDjLBD19DJ070710.480.30.77nsnsnsDjLBD20DJ091379.180.231.02−5.34↓−3.0↓nsDjLBD21DJ094090.00.00.09nsnsnsDjLBD22DJ10816nsnsnsDjLBD23DJ1216350.032.622.33−4.17↓−4.28↓nsDjLBD24DJ130943.260.770.57−2.01↓−2.45↓nsDjLBD25DJ130950.090.070.0nsnsnsDjLBD26DJ1470115.1712.7517.0nsnsnsDjLBD27DJ150540.180.921.69ns3.38↑nsDjLBD28DJ1520847.63.444.09−3.71↓−3.4↓nsDjLBD29DJ174682.362.992.72nsnsnsDjLBD30DJ1753114.694.389.28−1.65↓ns1.15↑DjLBD31DJ178561.070.460.43nsnsnsDjLBD32DJ198768.825.6711.87nsnsnsDjLBD33DJ199590.110.131.79nsnsnsDjLBD34DJ2155423.511.250.82−4.19↓−4.71↓nsDjLBD35DJ217160.030.030.0nsnsnsDjLBD36DJ231590.6710.2115.614.04↑4.68↑nsValues represent mean FPKM across three biological replicates at 0 h, 48 h, and 96 h post-inoculation. Up- and down-regulated genes (adjusted P < 0.05) are indicated by arrows (↑/↓). ‘ns’ denotes genes without significant differential expression. Log₂FC values are derived from pairwise RNA-Seq comparisons (48 h vs. 0 h, 96 h vs. 0 h, and 96 h vs. 48 h). Table 3Expression levels and differential expression of LBD genes in Jasminum sambac stems after Sclerotium rolfsii infection.LBD_nameRaw_IDFPKM_0hFPKM_48hFPKM_96hlog2FC_48h_vs_0hreg_48h _vs_0hlog2FC_96h_vs_0hreg_96h_vs_0hlog2FC_96h_vs_48hreg_96h_vs_48hDjLBD01DJ005920.540.190.34nsnsnsDjLBD02DJ008860.410.130.37nsnsnsDjLBD03DJ009130.124.8726.955.43↑8.0↑nsDjLBD04DJ009147.767.713.65nsnsnsDjLBD05DJ01079nsnsnsDjLBD06DJ241200.060.00.09nsnsnsDjLBD07DJ248254.660.80.23−2.49↓−4.19↓nsDjLBD08DJ254330.430.390.15nsnsnsDjLBD09DJ2934128.482.810.93−3.26↓−4.81↓−1.55↓DjLBD10DJ293810.360.00.04nsnsnsDjLBD11DJ295558.5610.3322.49ns1.53↑nsDjLBD12DJ30834nsnsnsDjLBD13DJ037990.0831.7319.348.96↑8.3↑nsDjLBD14DJ0484810.1313.9720.12nsnsnsDjLBD15DJ053536.021.320.45−2.15↓−3.61↓nsDjLBD16DJ058362.029.4415.872.36↑3.19↑nsDjLBD17DJ0691325.7158.18422.232.67↑4.16↑1.48↑DjLBD18DJ0701439.0521.5722.59nsnsnsDjLBD19DJ070710.480.30.77nsnsnsDjLBD20DJ091379.180.231.02−5.34↓−3.0↓nsDjLBD21DJ094090.00.00.09nsnsnsDjLBD22DJ10816nsnsnsDjLBD23DJ1216350.032.622.33−4.17↓−4.28↓nsDjLBD24DJ130943.260.770.57−2.01↓−2.45↓nsDjLBD25DJ130950.090.070.0nsnsnsDjLBD26DJ1470115.1712.7517.0nsnsnsDjLBD27DJ150540.180.921.69ns3.38↑nsDjLBD28DJ1520847.63.444.09−3.71↓−3.4↓nsDjLBD29DJ174682.362.992.72nsnsnsDjLBD30DJ1753114.694.389.28−1.65↓ns1.15↑DjLBD31DJ178561.070.460.43nsnsnsDjLBD32DJ198768.825.6711.87nsnsnsDjLBD33DJ199590.110.131.79nsnsnsDjLBD34DJ2155423.511.250.82−4.19↓−4.71↓nsDjLBD35DJ217160.030.030.0nsnsnsDjLBD36DJ231590.6710.2115.614.04↑4.68↑nsValues represent mean FPKM across three biological replicates at 0 h, 48 h, and 96 h post-inoculation. Up- and down-regulated genes (adjusted P < 0.05) are indicated by arrows (↑/↓). ‘ns’ denotes genes without significant differential expression. Log₂FC values are derived from pairwise RNA-Seq comparisons (48 h vs. 0 h, 96 h vs. 0 h, and 96 h vs. 48 h). Expression levels and differential expression of LBD genes in Jasminum sambac stems after Sclerotium rolfsii infection. Values represent mean FPKM across three biological replicates at 0 h, 48 h, and 96 h post-inoculation. Up- and down-regulated genes (adjusted P < 0.05) are indicated by arrows (↑/↓). ‘ns’ denotes genes without significant differential expression. Log₂FC values are derived from pairwise RNA-Seq comparisons (48 h vs. 0 h, 96 h vs. 0 h, and 96 h vs. 48 h). Among the 37 DjLBD genes, 14 showed significant differential expression in at least one pairwise comparison. Strongly up-regulated genes included DjLBD03, DjLBD13, DjLBD16, DjLBD17, DjLBD27, and DjLBD36, many of which are associated with hormone signaling (e.g., α-linolenic acid and jasmonate pathways) and proteostasis processes. In contrast, DjLBD07, DjLBD09, DjLBD20, DjLBD23, DjLBD28, and DjLBD34 were markedly down-regulated, suggesting suppression of growth-related LBD programs to prioritize defense responses. Most changes occurred within the first 48 h, consistent with the global DEG patterns showing pronounced early transcriptional reprogramming followed by partial stabilization at 96 h. Overall, these RNA-Seq data delineate time-dependent transcriptional modulation within the LBD family, highlighting its active involvement in early defense activation and later adjustment of hormone-regulated networks. This comprehensive RNA-Seq-based expression overview provides the molecular framework for the subsequent qPCR validation of representative LBD genes (see Sect. 1.9.3). qPCR validation of differentially expressed genes qPCR of four LBD genes (DjLBD09, DjLBD13, DjLBD17, DjLBD28) confirmed RNA-seq reliability, with consistent temporal expression patterns: DjLBD09 and DjLBD13 were highly expressed at 0 h, DjLBD17 increased at later stages, and DjLBD28 decreased after infection, suggesting time-specific regulatory roles in defense responses(Fig. 8).Fig. 8qPCR and RNA-seq validation of four LBD genes (DjLBD09, DjLBD13, DjLBD17, and DjLBD28) in jasmine stems at different infection stages (0 h, 48 h, and 96 h). qPCR and RNA-seq validation of four LBD genes (DjLBD09, DjLBD13, DjLBD17, and DjLBD28) in jasmine stems at different infection stages (0 h, 48 h, and 96 h). Integrated metabolomic and transcriptomic analysis Integrated KEGG and correlation analyses revealed dynamic cross-omics coordination during Sclerotium rolfsii infection in Jasminum sambac. Joint KEGG enrichment identified both shared and time-specific pathways. At 0 h vs 48 h, differentially expressed genes (DEGs) and metabolites were co-enriched in tryptophan metabolism, zeatin biosynthesis, α-linolenic acid metabolism, plant-hormone signal transduction, and secondary-metabolite biosynthesis, reflecting strong early transcription–metabolism coordination. At 0 h vs. 96 h, the enrichment pattern remained largely similar but involved a broader range of pathways, indicating strengthened coupling with prolonged infection. By contrast, 48 h vs. 96 h showed markedly fewer shared pathways—mainly secondary metabolism, α-linolenic acid metabolism, and hormone signaling—suggesting that cross-omics integration narrowed at the late stage (Fig. 9).Fig. 9Joint KEGG enrichment of transcriptomic and metabolomic data. (A) 0 h vs. 48 h, (B) 0 h vs. 96 h, and (C) 48 h vs. 96 h. Circles represent metabolites and triangles represent genes; bubble size indicates the number of differential molecules, and color denotes enrichment significance (P value). KEGG pathway enrichment was performed based on KEGG annotations23–25. Joint KEGG enrichment of transcriptomic and metabolomic data. (A) 0 h vs. 48 h, (B) 0 h vs. 96 h, and (C) 48 h vs. 96 h. Circles represent metabolites and triangles represent genes; bubble size indicates the number of differential molecules, and color denotes enrichment significance (P value). KEGG pathway enrichment was performed based on KEGG annotations23–25. Nine-quadrant correlation plots further highlighted strong concordant gene–metabolite changes at 48 h and 96 h relative to 0 h, while comparisons between 48 and 96 h exhibited reduced significance, consistent with early large-scale reprogramming followed by metabolic stabilization. Correlation heatmaps confirmed this time-dependent trend: early responses were dominated by auxin (IAA), cytokinin (CK), and jasmonate (JA) pathways, whereas by 96 h, cross-talk expanded to include salicylic acid (SA) signaling; later profiles again contracted, maintaining only a limited subset of hormone-related pathways (Fig. S9 and Fig. 10).Fig. 10Correlation clustering heatmaps from the integrated transcriptome–metabolome analysis of jasmine stems following Sclerotium rolfsii infection. (A) 0 h vs. 48 h, (B) 0 h vs. 96 h, and (C) 48 h vs. 96 h. Colors from blue to red indicate low to high relative expression/abundance levels of genes and metabolites. Correlation clustering heatmaps from the integrated transcriptome–metabolome analysis of jasmine stems following Sclerotium rolfsii infection. (A) 0 h vs. 48 h, (B) 0 h vs. 96 h, and (C) 48 h vs. 96 h. Colors from blue to red indicate low to high relative expression/abundance levels of genes and metabolites. O2PLS modeling further validated tight coupling between the transcriptome and metabolome, showing high explanatory power (R2X = 0.9527, R2Y = 0.9584) and strong correlation (R2Xcorr = 0.8836, R2Ycorr = 0.9281). Hormone-related metabolites (OPDA, MEJA, IAA-Asp) and several high-loading DEGs contributed most strongly, underscoring coordinated regulation of hormone signaling and defense responses (Table 4, Fig. 11A, B). Collectively, these results demonstrate that hormone-mediated metabolic reprogramming and LBD-associated transcriptional regulation act in concert to shape the early defense response of J. sambac to S. rolfsii infection.Table 4O2PLS model contribution statistics.R2XR2YR2XcorrR2Ycorr0.95270.95840.88360.9281Fig. 11O2PLS loadings plots from the integrated transcriptome–metabolome analysis of Jasminum sambac young stems infected by Sclerotium rolfsii. (A) Metabolite loadings plot showing contributions of individual metabolites to the O2PLS model. (B) Gene loadings plot showing contributions of individual genes to the O2PLS model. Color intensity indicates contribution magnitude, with higher values representing greater influence on model variation and cross-omics covariance. Table 4O2PLS model contribution statistics.R2XR2YR2XcorrR2Ycorr0.95270.95840.88360.9281 O2PLS model contribution statistics. O2PLS loadings plots from the integrated transcriptome–metabolome analysis of Jasminum sambac young stems infected by Sclerotium rolfsii. (A) Metabolite loadings plot showing contributions of individual metabolites to the O2PLS model. (B) Gene loadings plot showing contributions of individual genes to the O2PLS model. Color intensity indicates contribution magnitude, with higher values representing greater influence on model variation and cross-omics covariance. Discussion This study integrates gene-family characterization, metabolomics, transcriptomics, and cross-omics modeling to delineate the regulatory landscape of jasmine young stems challenged by the necrotrophic pathogen Sclerotium rolfsii. Beyond demonstrating evolutionary and structural diversity of the LBD family and pronounced time-resolved reprogramming of transcripts and metabolites, our data reveal tight and stage-dependent coupling between specific LBD transcription factors, phytohormone networks, and defense pathways. An early (0–48 h) JA-centric reconfiguration coincides with selective repression of auxin-linked LBDs and induction of proteostasis- and defense-associated LBDs. At 48 h post-inoculation—when jasmonate-pathway intermediates (e.g., OPDA/MEJA) peak in the metabolome—multiple LBDs shift coherently. DjLBD15 (ARF19-like; ko04075, Plant hormone signal transduction) is strongly down-regulated, whereas DjLBD16, DjLBD17, and DjLBD36 are up-regulated, and DjLBD03 (Proteasome; ko03050) increases. This pattern fits a mechanistic scenario in which early dampening of auxin-associated LBD activity, together with enhanced ubiquitin–proteasome turnover (DjLBD03 and later DjLBD11 at 96 h), relieves repression by AUX/IAA/JAZ-type inhibitors and favors JA/ET-dominated immunity, a hallmark of effective defense against necrotrophs. Concomitantly, broad down-regulation of additional LBDs (e.g., DjLBD09/20/23/24/28/30/34) suggests that growth-linked LBD programs are curtailed to prioritize defense. Relevance of LBD transcription factors to quantitative necrotroph resistance. Because S. rolfsii is a necrotroph that kills host cells and exploits dead tissue, its infection outcomes are largely quantitative, determined by the extent and coordination of inducible defense rather than by discrete resistance loci. Transcription factors such as LBDs are ideally positioned to shape this quantitative resistance because they integrate multiple hormonal and metabolic signals to calibrate defense strength. During the early infection phase (0–48 h), several LBDs (DjLBD16/17/36) appear to function as positive regulators that rapidly activate JA/ET signaling and secondary metabolism, conferring a timely but graded defense response that restricts pathogen growth without triggering excessive cell death. In contrast, suppression of the auxin-linked DjLBD15 likely prevents activation of growth-associated programs antagonistic to JA/ET-mediated defense. Such context-dependent modulation of transcription factor activity provides a tunable regulatory layer for quantitative resistance—fine-tuning defense amplitude to optimize both protection and metabolic economy. Moreover, early induction of proteasome-related LBDs (DjLBD03/11) may expedite degradation of AUX/IAA and JAZ repressors, enhancing hormone crosstalk and accelerating transcriptional reprogramming. Together, these features illustrate how LBD-mediated transcriptional circuits could underlie quantitative resistance to necrotrophs by controlling the timing and intensity of defense activation. By 96 h, the system shows signs of transcriptional and metabolic stabilization with a controlled re-engagement of auxin nodes. While DjLBD15 remains repressed, DjLBD27 (ARF-related; ko04075) becomes up-regulated, indicating a late-phase rebalancing of auxin signaling as tissues transition from acute defense to adaptation. DjLBD16/17/36 sustain high expression, in line with ongoing α-linolenic acid/JA metabolism and phenylpropanoid/cell-wall routes, which likely maintain structural and chemical barriers. DjLBD03 and DjLBD11 (Proteasome; ko03050) remain elevated, consistent with continued proteostatic control over defense-related repressors during recovery. Cross-omics correlation analyses (nine-quadrant plots, co-expression/metabolite heatmaps, and O2PLS/CCA with high explanatory and cross-covariance metrics) converge on LBD–hormone–defense modules. For example, DjLBD13, DjLBD17, DjLBD28, and DjLBD09 co-vary with defense-related metabolites (e.g., phenylpropanoids/flavonoids) and hormone intermediates despite not always appearing within enriched GO/KEGG terms themselves—an expected limitation because transcription factors often lack direct pathway annotations. Importantly, the differential expression of these LBDs was validated by qPCR, reinforcing their biological relevance. Collectively, these observations support a testable working model:(i)Early (0–48 h): S. rolfsii perception triggers JA/ET-biased signaling, suppresses auxin-linked LBDs (e.g., DjLBD15), and activates proteasome-associated and defense-supportive LBDs (DjLBD03/11, DjLBD16/17/36) to accelerate removal of negative regulators and drive defense gene expression.(ii)Mid-to-late (48–96 h): transcriptional programs partially contract and re-equilibrate, with select auxin-related LBDs (DjLBD27) re-engaging to restore growth-defense balance while JA-linked metabolism remains at a protective set-point. (i)Early (0–48 h): S. rolfsii perception triggers JA/ET-biased signaling, suppresses auxin-linked LBDs (e.g., DjLBD15), and activates proteasome-associated and defense-supportive LBDs (DjLBD03/11, DjLBD16/17/36) to accelerate removal of negative regulators and drive defense gene expression. Early (0–48 h): S. rolfsii perception triggers JA/ET-biased signaling, suppresses auxin-linked LBDs (e.g., DjLBD15), and activates proteasome-associated and defense-supportive LBDs (DjLBD03/11, DjLBD16/17/36) to accelerate removal of negative regulators and drive defense gene expression. (ii)Mid-to-late (48–96 h): transcriptional programs partially contract and re-equilibrate, with select auxin-related LBDs (DjLBD27) re-engaging to restore growth-defense balance while JA-linked metabolism remains at a protective set-point. Mid-to-late (48–96 h): transcriptional programs partially contract and re-equilibrate, with select auxin-related LBDs (DjLBD27) re-engaging to restore growth-defense balance while JA-linked metabolism remains at a protective set-point. While our multi-omics framework pinpoints candidate LBDs as high-leverage regulatory nodes in quantitative necrotroph resistance, functional dissection will be essential to clarify direct targets and hormone crosstalk logic (e.g., genetic perturbation of DjLBD15/17/36/27; chromatin profiling to map downstream defense and hormone-pathway genes; temporal hormone/UPS manipulations). Even so, by aligning LBD dynamics with JA/ET activation, auxin dampening and later re-balancing, proteasome engagement (ko03050), and pathway-level reprogramming (ko04075, α-linolenic acid and phenylpropanoid metabolism), our results provide a coherent regulatory narrative for jasmine’s response to S. rolfsii and a roadmap for trait improvement through TF-guided rewiring of hormone–defense networks. Note on pathway evidence: Gene–set enrichment and gene–metabolite correlation results supporting these inferences—including the specific LBD hits to ko03050/ko04075 and the cross-omics links to hormone and secondary-metabolism pathways—are provided in the Supplementary files (e.g., Supplementary Tables S3–S5: Differentially expressed LBD genes (48 h vs. 0 h, 96 h vs. 0 h, and 96 h vs. 48 h)). Conclusions By integrating gene-family characterization, metabolomic and transcriptomic profiling, and cross-omics modeling, this study delineates the molecular regulatory landscape of Jasminum sambac young stems responding to the necrotrophic pathogen Sclerotium rolfsii. The 0–48 h interval represents a critical window of intense transcriptional and metabolic reprogramming, dominated by activation of jasmonate (JA) and ethylene (ET) signaling, suppression of auxin-responsive modules, and enhanced proteasome turnover. These early adjustments coincide with strong enrichment of α-linolenic acid metabolism, plant hormone signal transduction (ko04075), and proteasome pathways (ko03050), reflecting coordinated reallocation of resources toward defense activation. By 96 h, transcriptional and metabolic shifts moderate, accompanied by partial re-engagement of auxin signaling and sustained activity of phenylpropanoid and cell-wall–associated metabolism, indicating a transition from acute defense to structural and metabolic recovery. Cross-omics analyses further reveal that LBD transcription factors function as pivotal integrators of hormone and defense pathways. DjLBD16, DjLBD17, and DjLBD36 act as early positive regulators in JA/ET-mediated immunity, whereas DjLBD15 repression dampens auxin-linked growth programs antagonistic to defense. Late-stage induction of DjLBD27 suggests a rebalancing of auxin signaling as defense stabilizes. Concomitant elevation of proteasome-related LBDs (DjLBD03, DjLBD11) underscores the importance of protein turnover in fine-tuning hormone crosstalk and removing transcriptional repressors. Together, these results support a coherent model in which LBD-mediated regulation dynamically coordinates hormone signaling, secondary metabolism, and proteostasis to modulate quantitative resistance against necrotrophic infection. Overall, this work establishes a multi-omics framework that links temporal pathway enrichment with transcriptional control and identifies key LBD members as high-leverage regulators within the JA/ET–auxin–proteasome–phenylpropanoid axis. These insights not only deepen our understanding of jasmine’s adaptive defense mechanisms but also provide valuable molecular targets for functional validation and resistance-oriented genetic improvement. Materials and methods Materials Experimental site Field trials were conducted in the open nursery of the Flower Research Institute, Guangxi Academy of Agricultural Sciences (22°48′ N, 108°22′ E), where new black weed-control fabric had been laid on the ground surface. The site is located in a southern subtropical monsoon climate zone, with a mean annual temperature of 21.6 °C, mean annual precipitation of 1,304.2 mm, mean relative humidity of 79%, and a frost-free period of 334 days2. Experimental materials Three-year-old potted plants of the main local cultivar—Hengxian double-petal jasmine—were used. The jasmine southern blight pathogen (Sclerotium rolfsii Sacc.) isolate SRNN1 was obtained from diseased plants collected in Jiaoyi Town, Heng County, Guangxi; after purification and identification, its rDNA-ITS sequence was deposited in GenBank under accession number MT634388. The pathogen was cultured on potato dextrose agar (PDA) composed of: potato 200 g, glucose 20 g, agar 15 g, distilled water 1 L; pH unadjusted (“natural”). Inoculation with S. rolfsii by the stem-adhesion method Materials included PDA plates fully covered with mycelia, sterile cotton, sterile water (ultrapure; stored in Erlenmeyer flasks and spray bottles), sealing tape, sterile toothpicks, sterile pipette tips, large plastic bags, and jasmine plants. All procedures were performed with gloves. Using a sterile pipette tip (large end downward), perforate the PDA from the colony margin and continue until the plate surface is covered; seal the plate and reserve for use within 1 day. Prepare strips of sterile cotton and moisten them with sterile water. With a sterile toothpick, gently score around the intended inoculation site on the stem; remove excess leaves so the wound is exposed without penetrating deeply into the stem. For inoculation, use a new sterile toothpick to pick up an agar plug bearing mycelia and place it onto the moistened cotton with the mycelial surface facing the scored stem. Affix the cotton to the wound, wrap firmly, and secure with sealing tape, keeping the openings at the upper and lower ends as small as possible. In non-air-conditioned rooms or outside an incubator, a high-humidity microenvironment can be created with a large plastic bag: mist the inner surface, then cover the entire plant with the bag inverted so that water droplets coat the interior. Six hours after inoculation, assess moisture loss, randomly inspect several plants, and add water as needed to determine subsequent watering intervals. To replenish moisture, spray an appropriate amount of sterile water through the upper and lower openings at the tape. After approximately 1–3 days, once stem infection is evident, proceed to the next stage. RNA extraction from infected stems Healthy, uniformly growing jasmine plants were selected. Stems inoculated with S. rolfsii served as the treatment group, and samples collected before inoculation (time 0; Day 0) served as the control. Sampling time points were set at pre-inoculation (0 h) and 48 h (Day 2) and 96 h (Day 4) post-inoculation. Three biological replicates were collected per time point, with three stems pooled per replicate. For each sample, 7–8 cm stem segments were cut using sterile pruning shears and divided into two portions: one for RNA extraction and transcriptome analysis, and the other for hormone profiling. Samples were snap-frozen in liquid nitrogen immediately after processing and stored at − 80 °C until further use. Sequence mining Genome assemblies of single-petal jasmine, double-petal jasmine, and ‘Hutou’ jasmine were obtained from collaborative institutions. Coding sequences (CDS) and protein sequences were extracted using TBtools26, and reciprocal searches were performed against the Swiss-Prot database to obtain candidate proteins. In addition, the Arabidopsis thaliana genome was downloaded from NCBI (https://www.ncbi.nlm.nih.gov/), and LBD gene sequences were identified based on annotation information. Identification of jasmine LBD family members and analysis of physicochemical properties In total, 39 LBD genes were identified in single-petal jasmine, 36 in double-petal jasmine, and 38 in ‘Hutou’ jasmine. Physicochemical properties of LBD proteins from the three jasmine types were analyzed using Python (Biopython package)27, and subcellular localization was predicted with Plant-mPLoc (http://www.csbio.sjtu.edu.cn/bioinf/plant-multi/). Gene structure, conserved motifs, and domain prediction for the jasmine LBD family Based on gene annotations, LBD protein sequences from the three jasmine types were obtained. Conserved motifs were predicted using the “Simple MEME Wrapper” in TBtools26 with default parameters and confirmed using MEME Suite28. Gene structure information and motif prediction results were organized and visualized in TBtools. Chromosomal localization, phylogenetic tree construction, and sequence alignment Chromosomal locations of LBD genes were mapped and visualized with TBtools (v2.149)26 for all three jasmine types. Multiple sequence alignments of LBD proteins were generated using the “One Step Build a ML Tree” function in TBtools, and a phylogenetic tree was constructed based on amino-acid sequence similarity. The tree was visually refined using the Chiplot web tool (https://www.chiplot.online/). Cis-acting element prediction in promoters of jasmine LBD genes Promoter regions (2,000 bp upstream of the transcription start site) of LBD genes from the three jasmine types were retrieved and analyzed for cis-acting regulatory elements using PlantCARE (http://bioinformatics.psb.ugent.be/webtools/plantcare/html)29. Predicted elements were categorized and visualized in TBtools. Real-time quantitative PCR (qRT–PCR) of jasmine LBD genes Total RNA extraction and reverse transcription were outsourced to a biotechnology company. The reference gene was UPL730. All primers were designed with Primer Premier 6; sequences are listed in Table 5. qRT–PCR was performed on an ABI StepOne Plus instrument in a 20.0 μL reaction containing 0.8 μL each of forward and reverse primers, 10.0 μL of 2 × qPCR SYBR Green Master Mix, 1.0 μL cDNA template, and nuclease-free water to 20.0 μL. Cycling conditions were: 95 °C for 30 s; then 40 cycles of 95 °C for 5 s and 60 °C for 10 s. Each sample was run with three technical replicates. Relative transcript levels were calculated using the 2^ − ΔΔCt method. Calculations were performed in Excel, and statistical visualization was conducted in GraphPad Prism 8.0.2. For label consistency with DjLBD identifiers, we format it as ‘UPL07’ in primer IDs; this is an alias of UPL7.Table 5Primer sequences for target and reference genes used in qRT–PCR.GenePrimer sequences (5′ → 3′; forward/reverse)DjLBD09F: GCACAAGCCGAAGTGGTACA/R: TGACTGGACGATGGGAAGGCDjLBD13F: GCAGGAGCTTCCGTTACACC/R: AGCACCAACACAGCCGTAGADjLBD17F: GCCGAGCCATGAGTCGTCTT/R: AGTCTGCCCAGAAACCAAGCDjLBD28F: TGCCGAGTTCTCCGAAAGGG/R: ACGGTGGCGTTGCTTTGAGUPL07 (reference)F: CAGAAGAGGCTGGTAGATGTCG/R: GCCCTTTCATTCAACAGCAA‘UPL07’ is a zero-padded label used for layout consistency and refers to the same gene as UPL7. Table 5Primer sequences for target and reference genes used in qRT–PCR.GenePrimer sequences (5′ → 3′; forward/reverse)DjLBD09F: GCACAAGCCGAAGTGGTACA/R: TGACTGGACGATGGGAAGGCDjLBD13F: GCAGGAGCTTCCGTTACACC/R: AGCACCAACACAGCCGTAGADjLBD17F: GCCGAGCCATGAGTCGTCTT/R: AGTCTGCCCAGAAACCAAGCDjLBD28F: TGCCGAGTTCTCCGAAAGGG/R: ACGGTGGCGTTGCTTTGAGUPL07 (reference)F: CAGAAGAGGCTGGTAGATGTCG/R: GCCCTTTCATTCAACAGCAA‘UPL07’ is a zero-padded label used for layout consistency and refers to the same gene as UPL7. Primer sequences for target and reference genes used in qRT–PCR. ‘UPL07’ is a zero-padded label used for layout consistency and refers to the same gene as UPL7. OPLS–DA analysis and model validation To distinguish metabolic profiles among treatments, orthogonal partial least-squares discriminant analysis (OPLS–DA) was applied to the metabolomics data. Prior to modeling, the data matrix was log-transformed and mean-centered to reduce noise and enhance robustness. Modeling and visualization were performed in R (v4.3.0)31using the ropls package32. Model fit and predictive ability were evaluated by R2Y and Q2, respectively, and reliability was assessed by 200-time permutation testing to prevent overfitting. When the original model’s R2Y and Q2 values exceeded those of the permuted models and the regression line intercept was < 0, the model was considered robust with good predictive performance. Transcriptome sequencing and data analysis RNA-seq and initial data processing were performed by Wuhan MetWare Biotechnology Co., Ltd. After total RNA extraction and QC, libraries were sequenced on the Illumina platform. Clean reads were processed with fastp33 and mapped to the reference genome (Jasminum sambac, GWHBFHJ00000000) using HISAT34. Gene expression levels were quantified as FPKM, and read counting was conducted with featureCounts35. Differential expression analysis was performed using DESeq2, with |log₂FC|≥ 1 and FDR < 0.05 as the significance thresholds36. Results (DEG summaries, volcano plots, GO enrichment, and KEGG pathway enrichment) were used for downstream biological interpretation23,37. Metabolomic analysis Targeted plant-hormone metabolomics was conducted by Wuhan MetWare Biotechnology Co., Ltd. using an LC–MS/MS platform (UPLC: ExionLC™ AD; MS/MS: QTRAP® 6500 +). After grinding samples in liquid nitrogen, metabolites were extracted with a methanol/water/formic acid mixture, with internal-standard–based QC following established protocols for plant hormone extraction and preprocessing38–40. Chromatographic separation employed a C18 column; mobile phases were ultrapure water with 0.04% acetic acid and acetonitrile, with a gradient program adapted from published HPLC/UPLC–MS methods38,41,42. Mass spectrometry used an ESI source in both positive and negative modes, with parameters following typical UHPLC–MS/MS hormone assays38,43. Compound identification was based on an in-house library and authentic standards; quantification used MRM with external-standard calibration, following protocols for multi-hormone targeted profiling38–41,43. Differential metabolites were selected using combined criteria of OPLS–DA VIP > 1, fold change ≥ 2 or ≤ 0.5, and t-test p < 0.05. Identified differential metabolites were annotated and enriched against KEGG to infer pathway-level changes23,37. Integrated analysis of metabolome and transcriptome Integrated analyses were performed by Wuhan MetWare Biotechnology Co., Ltd. First, principal component analysis (PCA) of metabolomic and transcriptomic datasets was conducted using the prcomp function in R (v4.2.0) to evaluate global separation among treatments31. KEGG annotation and enrichment of DEGs and differential metabolites followed the same pipelines and parameters as in the transcriptome analysis, ensuring cross-omics comparability23. Pairwise Pearson correlations were computed using the stats package in R to identify significant gene–metabolite associations (|r|≥ 0.8, p < 0.05). Results were visualized as nine-quadrant plots, clustered heatmaps (pheatmap package), and correlation networks constructed in Cytoscape v3.9.1 (Shannon et al., 2003; https://cytoscape.org) to reveal potential regulatory relationships44. At the global level, canonical correlation analysis (CCA) was performed with the CCA package in R (González et al., 2008; https://cran.r-project.org/package=CCA) to characterize canonical variate pairs between the two omics layers45. In parallel, O2PLS modeling using the OmicsPLS package in R (Bouhaddani et al., 2016; https://cran.r-project.org/package=OmicsPLS) decomposed each dataset into joint, orthogonal, and noise components; key contributing variables were identified based on R2 and variable loadings46,47. Supplementary Information Supplementary Information Supplementary Information