Multi-omics integration reveals tissue-specific biosynthesis of sesquiterpenoids in the medicinal plant Eupatorium lindleyanum

Yingzhe Wang; Pan Jiang; Jiaqiu Yuan; Jinghan Wu; Yue Zhang; Kun Guo

doi:10.1371/journal.pone.0359566

. 2026 Oct 1;21(10):e0359566. doi: 10.1371/journal.pone.0359566

Multi-omics integration reveals tissue-specific biosynthesis of sesquiterpenoids in the medicinal plant Eupatorium lindleyanum

Yingzhe Wang ^1,^#, Pan Jiang ^1,^#, Jiaqiu Yuan ¹, Jinghan Wu ², Yue Zhang ^3,^*, Kun Guo ^3,^*

Editor: Rongchun Han⁴

PMCID: PMC13630220 PMID: 42821537

Abstract

Background

Eupatorium lindleyanum is a medicinal plant rich in bioactive sesquiterpenoids, yet the genetic and regulatory basis underlying their biosynthesis remains completely unexplored. In particular, the marked tissue-specific accumulation of these compounds—abundant in aerial organs but scarce in roots—lacks a molecular explanation.

Results

Here, we integrated UPLC-MS/MS-based metabolomics with RNA-Seq transcriptomics to comprehensively profile sesquiterpenoid metabolites and their associated gene expression signatures across root, stem, leaf, and flower tissues of Eupatorium lindleyanum. A total of 321 sesquiterpenoids were identified, exhibiting distinct tissue-specific accumulation patterns with highest abundance in flower, followed by leaf. Comparative analysis revealed 299 differentially accumulated sesquiterpenoid metabolites (DSMs), which were clustered into three major groups with preferential accumulation in leaf (Cluster 1), flower (Cluster 2), and root (Cluster 3). Transcriptome sequencing generated 99.43 Gb of clean data, yielding 147743 unigenes. Integration of metabolomic and transcriptomic datasets screened 20 structural genes (88 unigenes) involved in sesquiterpenoid biosynthesis, including MVA pathway genes (ACAT, HMGCS, HMGCR, MVK, PMK, MVD), MEP pathway genes (dxs, dxr, ispD, ispE, ispF, ispG, ispH), and downstream modification genes (FDPS, IDI, GAS, GDS, NES1, GAO, FLDH). Weighted gene co-expression network analysis (WGCNA) further prioritized hub genes within modules significantly correlated with key sesquiterpenoids. Additionally, five transcription factor families (AP2/ERF, NAC, WRKY, MYB, and bHLH) were identified as potential regulators based on strong correlations with DSMs accumulation.

Conclusions

This study provides the first comprehensive view of tissue-specific sesquiterpenoid biosynthesis in Eupatorium lindleyanum, establishing a prioritized set of candidate genes and regulators for functional validation. Our findings offer a foundational resource for metabolic engineering and molecular breeding aimed at enhancing the production of pharmacologically active sesquiterpenoids in this medicinal species.

1. Background

Eupatorium lindleyanum, a perennial herbaceous plant belonging to the Asteraceae family, is indigenous to East Asia [1]. In China, it is particularly abundant in the Xuyi County area of Jiangsu Province [2]. According to the theories of Traditional Chinese Medicine (TCM), this plant serves as the botanical origin of the crude drug “Yema Zhui,” which is traditionally ascribed with therapeutic effects including resolving phlegm, alleviating cough, and relieving dyspnea [3]. Clinically, it has been employed in the management of chronic tracheitis and bronchitis [4]. Pharmacognostical and phytochemical investigations into the medicinal properties of Eupatorium lindleyanum commenced in the 1970s [5]. Through extensive research spanning several decades, over one hundred distinct compounds have been isolated and characterized from this species. These isolates encompass diverse classes of secondary metabolites, including sesquiterpenoids, flavonoids, and alkaloids [6]. Among these, the sesquiterpenoids constitute the class exhibiting the most potent bioactivities. Previous studies have successfully identified numerous bioactive sesquiterpenoids from Eupatorium lindleyanum, such as eupalinolides and eupalinilides, among others. These compounds have demonstrated significant pharmacological effects, including the prevention of lung injury, reduction of blood lipid levels, antitumor activity, attenuation of inflammatory responses, and free radical scavenging capacity ^[7–9^]. Collectively, these findings provide a scientific basis that substantiates the plant’s traditional medicinal applications. The significant pharmaceutical potential of these bioactive sesquiterpenoids positions them as key targets for metabolic engineering and as lead compounds for further development [10].

Despite progress in other species, how sesquiterpenoids are biosynthesized in Eupatorium lindleyanum remains largely unknown. Although the general framework of sesquiterpenoid biosynthesis in plants has been well established, recent investigations in the Asteraceae family have uncovered unique mechanistic innovations and variations within this pathway [11]. For instance, a functionally distinct cytochrome P450 enzyme (CYP71DD1) identified in Xanthium strumarium catalyzes the skeletal rearrangement of the germacrane backbone into a xanthanolide-type framework, a critical step in the formation of 12,8-cis-lactones [12]. Furthermore, the sesquiterpene synthase IlTPS1, cloned from Inula linearifolia, was found to direct the biosynthesis of rare bridged tricyclic presilphiperfolane-type sesquiterpenoids, with its products exhibiting significant antifungal activity [13]. Similarly, comparative genomic analyses of Silybum marianum and Carthamus tinctorius have revealed the presence of biosynthetic gene clusters responsible for caryophyllene-type sesquiterpenoids, whose formation is closely associated with whole-genome duplication and tandem duplication events [14]. These findings highlight the unique and diverse nature of sesquiterpenoid biosynthesis in Asteraceae, characteristics that are not fully captured by the traditional view of a highly conserved universal metabolic pathway. Despite this progress in related species, the specific genetic and regulatory basis for sesquiterpenoid biosynthesis in Eupatorium lindleyanum remains largely unexplored. To date, none of the genes encoding key enzymes in this pathway have been functionally characterized, and the tissue-specific regulatory mechanisms governing their expression are yet to be investigated. Addressing this knowledge gap is essential, as it would not only clarify whether Eupatorium lindleyanum employs unique regulatory strategies compared to other Asteraceae plants but also unlock the significant medicinal potential of its bioactive sesquiterpenoids. Elucidating this pathway holds substantial application value: it would enable the metabolic engineering of key sesquiterpenoid compounds, support the breeding of high-yielding Eupatorium lindleyanum varieties, and inform cultivation practices to enhance the accumulation of these bioactive compounds. Ultimately, a comprehensive understanding of this pathway will contribute to the development of more effective herbal medicines and strengthen the pharmacological application of Eupatorium lindleyanum [15].

Integrated metabolomic and transcriptomic analysis has emerged as a pivotal approach for elucidating the biosynthesis and regulation of plant secondary metabolites. This strategy employs LC-MS/MS to comprehensively profile metabolite accumulation patterns, coupled with RNA-seq to identify differentially expressed genes [16]. Through correlation analysis and pathway enrichment, it enables the construction of gene–metabolite regulatory networks and the identification of key regulatory nodes in critical metabolic pathways [17]. This integrated approach has been successfully applied in diverse plant species, including Cyclocarya paliurus [18], Camphora officinarum [19], Opisthopappus taihangensis [20], and Bidens alba [21], providing support for deciphering important secondary metabolic pathways such as terpenoids biosynthesis and for mining key functional genes.

In this study, we collected samples from four distinct tissues—root, stem, leaf, and flower—of Eupatorium lindleyanum and performed comprehensive metabolomic profiling via UPLC-MS/MS alongside RNA-seq-based transcriptomic sequencing. Using integrated metabolomic and transcriptomic analyses, we systematically investigated the tissue-specific accumulation patterns of sesquiterpenoids and identified key structural genes and transcription factors involved in their biosynthesis. Through weighted gene co-expression network analysis (WGCNA) and correlation analysis between differentially accumulated sesquiterpenoid metabolites and differentially expressed genes, we constructed a sesquiterpenoid biosynthetic regulatory network in Eupatorium lindleyanum. These findings provide a valuable resource for understanding the molecular mechanisms governing sesquiterpenoid biosynthesis and offer potential targets for genetic engineering and breeding programs aimed at enhancing the production of bioactive compounds in this medicinal plant.

2. Materials and methods

2.1. Plant materials

On 14 September 2025, the Eupatorium lindleyanum specimens were harvested from a cultivated plantation in Xuyi County, Jiangsu province, China (32°55′37.831″N, 118°25′46.878″E), all of which were 17-month-old plants at the full-flowering stage. Taxonomic identification was performed by Professor Meng Zhu (Jiangsu Food and Pharmaceutical Science College) following the morphological descriptors in the Flora of China. Supplementary Figure 1 in S1 File provides a detailed view of the plant tissues. A voucher specimen (accession no. NAS00718179) has been archived in the publicly accessible herbarium of the Nanjing Botanical Garden, Chinese Academy of Sciences, and is available for examination upon formal request to the curator (bgs@cnbg.net). For each tissue type, three biological replicates were prepared by separately collecting samples from three distinct individual plants, ensuring that each replicate was independent and representative of the tissue type [22]. After collection, specimens were rinsed with ultrapure water, snap-frozen in liquid nitrogen within 1 minute, and maintained at −80°C pending subsequent analyses [23].

2.2. Plant sample pretreatment and metabolite extraction procedures

A previously described extraction protocol was used with minor modifications [24]. Eupatorium lindleyanum tissues were ground to a fine powder using a ball mill (30 Hz, 1.5 min; Scientz-100F, Ningbo, China). Each 50 mg of powder was mixed with 1.2 mL of 70% methanol pre-chilled to −20°C. This solvent-to-sample ratio was chosen to ensure full submersion of the tissue and efficient release of metabolites across a wide polarity range. The mixture underwent six rounds of vortexing (30 seconds each) at 30-minute intervals. After centrifugation, the resulting supernatant was passed through a 0.22 μm membrane filter before UPLC‒MS/MS measurement. Given that many sesquiterpenoids exist as thermally labile lactones or polar glycosides that are poorly amenable to gas chromatography, the UPLC‒MS/MS platform was selected to enable comprehensive detection without derivatization or thermal degradation.

2.3. UPLC‒MS/MS conditions for metabolite profiling

UPLC-MS/MS analysis was conducted using an ExionLC™ AD chromatograph coupled to an ESI-QTRAP mass spectrometer (https://sciex.com.cn/). A reversed-phase separation was performed on an Agilent SB-C18 column (2.1 × 100 mm, 1.8 μm particle size). The mobile phase consisted of (A) 0.1% formic acid in water and (B) 0.1% formic acid in acetonitrile. The elution gradient was programmed as follows: 0–9 min, 5%–95% B; 9–10 min, hold at 95% B; 10–11.1 min, return to 5% B; and 11.1–14 min, re-equilibration with 5% B. Operational parameters included a flow rate of 0.35 mL/min, a column oven temperature of 40°C, and an injection volume of 2 μL. The column outflow was directed into the mass spectrometer for data acquisition [25].

ESI source settings were configured as follows: temperature, 500°C; ion spray voltage, + 5500 V (positive) and −4500 V (negative); nebulizer gas (GSI), heater gas (GSII), and curtain gas (CUR) at 50, 60, and 25 psi, respectively; with collision-activated dissociation (CAD) set to high. Mass spectral data were collected in multiple reaction monitoring (MRM) mode using nitrogen as the collision gas at a medium setting [26].

2.4. Qualitative and quantitative analysis of metabolites

Metabolite identification, quantification, and multivariate analysis were performed as described previously [2]. Briefly, identification used the MetWare database, quantification used triple quadrupole MS in MRM mode, and data were processed with Analyst 1.6.3. PCA and OPLS-DA were conducted in R (prcomp) and MetaboAnalyst 6.0, respectively. For the OPLS-DA models, the six pairwise comparison groups (Flower vs Leaf, Flower vs Root, Flower vs Stem, Leaf vs Root, Leaf vs Stem, and Stem vs Root) yielded Q² values ranging from 0.992 to 0.995, R²X from 0.804 to 0.860, and R²Y of 1 for all comparisons, indicating excellent predictive performance and reliable group separation (see Supplementary Table 3 in S1 File). DAMs were defined by VIP > 1, |log₂FC| ≥ 1, and FDR < 0.05, from which DSMs were selected. KEGG annotation was then applied to the DSMs. To investigate the dynamic accumulation patterns of DSMs, fuzzy c-means clustering was performed using the Mfuzz package in R. Metabolites were clustered based on their relative abundance profiles, with the optimal cluster number determined by minimizing the cluster centroid distance [27].

2.5. RNA extraction, library construction, and sequencing

RNA isolation was carried out using the TianGen Total RNA Kit (Beijing, China). Briefly, liquid nitrogen-ground tissues were homogenized, and total RNA was eluted in 50 µL of DEPC-treated water. RNA concentration and integrity were measured with a Qubit fluorometer and a Qsep400 bioanalyzer (Thermo Fisher Scientific, USA; Phonton Biotech, Taiwan, China), respectively. For library preparation, poly(A)-containing mRNA was captured using oligo(dT)-coated magnetic beads and reverse-transcribed into cDNA with the PrimeScript™ RT Kit (TaKaRa, China). Sequencing adapters were ligated to both ends of the cDNA, and the resulting libraries were barcoded for multiplexing. The libraries were constructed as non-stranded (unstranded) using the standard double-stranded cDNA synthesis protocol with random hexamer priming. After adapter ligation, library concentration and size distribution were re-assessed using the Qubit and Qsep400 systems. Qualified libraries were then subjected to cluster generation on a cBot instrument with the TruSeq PE Cluster Kit v3 (Illumina), followed by paired-end sequencing (125 or 150 bp) on an Illumina HiSeq platform [28]. Raw sequence data are accessible at NCBI SRA under accession number PRJNA1242508.

2.6. Data analysis of RNA-seq

Raw reads were processed with fastp (v0.23.2) to remove low-quality bases, adapters, and ambiguous reads [29]. Clean reads from all samples were then assembled together using Trinity (v2.15.1) in a merged assembly mode, which improves detection of both highly and weakly expressed transcripts and yields a more complete transcriptome reference. After assembly, Corset (v1.09) was used to cluster and remove redundant transcripts. Functional annotation of the non-redundant transcripts was performed by aligning them against KEGG, NR, Swiss-Prot, GO, KOG, and TrEMBL databases using DIAMOND (v2.1.6). Transcript abundance was estimated with RSEM (v1.3.3), producing FPKM values normalized by transcript length and total mapped reads. Differential expression analysis between tissue pairs was conducted with DESeq2 (v1.47.1), applying thresholds of |log₂FC| ≥ 1 and FDR < 0.05 to identify DEGs [30]. Putative transcription factors among DEGs were predicted using iTAK (v2.0.2). Additionally, Mfuzz clustering was applied to DEGs following the same procedure used for DSMs [31].

2.7. Integrated analysis of transcriptomic and metabolomic data

Based on a comparison of the Mfuzz analysis results of DEGs and DSMs, we discovered DEGs clusters exhibiting expression trends across root, stem, leaf, and flower of Eupatorium lindleyanum that were consistent with the accumulation patterns of the corresponding metabolites. These selected genes were subsequently integrated into a weighted gene co-expression network analysis (WGCNA). The gene co-expression network was constructed in R using the WGCNA package. Co-expression modules were generated with the blockwise Modules function, adhering to default parameters except for three modifications: the soft-thresholding power was set to 18, the mergeCutHeight to 0.25, and the minModuleSize to 50. Following normalization of the FPKM values, an adjacency matrix was created and subsequently converted into a topological overlap matrix (TOM) [32]. Co-expression modules were then identified by grouping genes with similar expression patterns into discrete clusters. We then identified the characteristic genes within these modules, focusing on their expression level and their associations with the synthesis of sesquiterpenoid metabolites. Following WGCNA, ten sesquiterpenoid metabolites (Isoalantolactone, Shizukanolide A, Britannilactone, Eupalinolide B, Dehydrosaussurea lactone, Yomogin, Shizukanolide H, Magnograndiolide, Costunolide, and Dihydrocostunolide) were selected for subsequent analysis based on their significant differential accumulation across Eupatorium lindleyanum tissues and their well-documented bioactivities [33–41]. Pearson correlation analysis was performed between the abundance of each metabolite and the module eigengene values from WGCNA. Modules exhibiting an absolute correlation coefficient > 0.8 with a statistical significance of P < 0.05 were retained for further investigation. Genes within the selected modules were then subjected to KEGG enrichment and cluster analysis. Finally, to pinpoint key regulatory candidates, genes annotated under the terpenoid backbone biosynthesis (ko00900) and sesquiterpenoid biosynthesis (ko00909) pathways were specifically retained and identified [42].

2.8. Quantitative real-time PCR validation of candidate genes

The specific primers for the 12 candidate genes and a reference gene were designed (Primer Premier 5.0), and their amplification efficiencies were validated (see Supplementary Table 1 in S1 File). Quantitative real-time PCR (qRT-PCR) was performed using the QuantiNova SYBR Green PCR Kit (QIAGEN, Hilden, Germany) on an ABI 7500 real-time PCR system (Thermo Fisher, USA). The 10 μL reaction mixture contained 5 μL of 2 × SYBR Green PCR Master Mix, 0.05 μL of QN ROX Reference Dye, 0.7 μL each of forward and reverse primers (0.7 μM final concentration), 1 μL of cDNA template (≤100 ng), and 2.55 μL of RNase-free water. The thermal cycling protocol was as follows: initial denaturation at 95°C for 2 min, followed by 40 cycles of denaturation at 95°C for 5 s and combined annealing/extension at 60°C for 30 s. A melting curve analysis was performed after amplification to verify product specificity. The β-actin gene served as the internal reference, and relative expression levels were calculated using the 2^−ΔΔCt method. Each stage of the experiment involved three biological replicates and three technical replicates. The β-actin gene was subjected to stability assessment using three software tools: geNorm version 3.5, NormFinder version 0.953, and BestKeeper version 1.0. The M value calculated by geNorm was 0.527, the stability value determined by NormFinder was 0.082, and the standard deviation value computed by BestKeeper was 0.453. The analysis results indicate that β-actin exhibits good expression stability and can be used as a reference gene.

2.9. Statistical analysis and data visualization

The metabolomic and transcriptomic datasets were initially stored and processed using Microsoft Excel 2021. Correlation analyses were conducted with SPSS software (version 26.0, IBM, USA). Principal component analysis, hierarchical clustering, and Mfuzz temporal expression analysis were performed, and corresponding visualizations were generated, utilizing the corrplot and base packages in R (version 4.2.1). Bar charts and additional figures were prepared with OriginPro software (version 2022b, OriginLab Corporation, USA).

3. Result

3.1. Profiling of sesquiterpenoid metabolites in Eupatorium lindleyanum derived from four different tissues

Using UPLC-MS/MS, a total of 321 sesquiterpenoid metabolites were identified in Eupatorium lindleyanum (Supplementary Table 2 in S1 File). The comparative total relative content of these metabolites in four distinct tissues is presented in Fig 1A. Sesquiterpenoid metabolites predominantly accumulated in the flower of Eupatorium lindleyanum, followed by the leaf. In contrast, their contents in the root and stem were significantly lower. This pattern indicates a gradual increase in sesquiterpenoid accumulation from root and stem to leaf and flower. Principal component analysis (PCA) of twelve samples (including three replicates) revealed that samples from different tissue parts were distinctly separated into independent principal components (PC1, 50.58%; PC2, 31.17%), indicating significant differences in sesquiterpenoid metabolites among various tissues of Eupatorium lindleyanum (Fig 1B). The correlation heatmap demonstrated strong intra-group consistency, with biological replicates of each tissue type exhibiting high correlation coefficients (typically >0.9). In contrast, inter-group comparisons revealed that root exhibited relatively weak correlations with stem, leaf, and flower, whereas correlations among the above-ground tissues (stem, leaf, and flower) were relatively strong. This suggests a clear metabolic distinction between root and above-ground tissues, while the metabolic profiles of stem, leaf, and flower are more similar to one another. (Fig 1C). To assess metabolite-specific variation, hierarchical clustering analysis (HCA) was performed using the relative abundances of sesquiterpenoids across distinct plant tissues (Fig 1D). Hierarchical clustering analysis further substantiated the above findings, revealing distinct separation among tissue types. Root, stem, leaf, and flower each contained differentially abundant sesquiterpenoid metabolites. In summary, the profiling of sesquiterpenoid metabolites revealed a clear tissue-specific pattern in Eupatorium lindleyanum, providing a basis for subsequent differential metabolites analysis.

Fig 1 — (A) Total sesquiterpenoid content from root, stem, leaf and flower, different letters denote statistically significant differences among tissues (P < 0.05). (B)Principal component analysis of different tissues from *Eupatorium lindleyanum*. (C) Pearson correlation analysis between 12 samples. (D) Hierarchical clustering analysis of 321 sesquiterpenoid metabolites in 12 samples: color intensity reflects relative abundance.

3.2. Differentially accumulated sesquiterpenoid metabolites in four tissues of Eupatorium lindleyanum

299 DSMs were identified by screening DAMs across all pairwise tissue comparisons: flower vs. root, flower vs. stem, flower vs. leaf, stem vs. root, leaf vs. root, and leaf vs. stem (Supplementary Table 3 in S1 File). Among the six tissues pairwise comparisons in Eupatorium lindleyanum, the flower vs. root comparison yielded the highest number of DSMs, with 185 upregulated and 48 downregulated metabolites. A similar pattern was observed in the leaf vs. root comparison, which showed 146 upregulated and 76 downregulated metabolites. The stem vs. root and flower vs. stem comparisons exhibited comparable DSM counts, with 122 upregulated and 88 downregulated and 185 upregulated and 23 downregulated, respectively. In the flower vs. leaf comparison, 146 metabolites were upregulated and 51 were downregulated. The leaf vs. stem comparison contained the fewest DSMs, with only 124 upregulated and 16 downregulated metabolites. Among these differentially accumulated metabolites, 28 DSMs overlapped across all six pairwise comparisons (Fig 2A). These 28 common DSMs represent a conserved metabolic core, likely involved in essential functions across tissues (Supplementary Table 4 in S1 File). Crucially, the vast majority of DSMs were tissue-specific, revealing profound metabolic divergence between the root and aerial organs. This high degree of specialization reflects distinct physiological roles governed by organ-specific gene expression during development. Mfuzz analysis classified the 299 DSMs into three clusters, each exhibiting a distinct tissue-specific accumulation pattern in Eupatorium lindleyanum. In sub cluster 1, DSMs demonstrated the highest relative abundance in leaf compared to other tissues. In sub cluster 2, the highest abundance was observed in flower, while sub cluster 3 was characterized by a predominant accumulation of DSMs in root (Fig 2B). The clustering results of metabolites within each sub-cluster consistently reinforce this tissue-specific partitioning, corroborating the organ-specific metabolic profiles observed in the overall analysis.

Fig 2 — (A) Visualization of differentially accumulated sesquiterpenoid metabolites: a bar chart (left) and a corresponding node graph (right). (B) Cluster analysis of sesquiterpenoid accumulation patterns. (Top) Fuzzy c-means clustering (Mfuzz) identified three distinct metabolite clusters. The cluster centroid is shown as a black line, with remaining lines representing individual compounds. (Bottom) Accumulation levels of sesquiterpenoids belonging to each cluster.

3.3. Transcriptome profiling analysis of Eupatorium lindleyanum in four distinct tissues

A total of 12 cDNA libraries were prepared from four tissues (root, stem, leaf, flower) of Eupatorium lindleyanum, each with three biological replicates. Sequencing generated 47983328–69716676 raw reads per library, and after quality filtering, 47182646–66154460 clean reads were obtained per sample, totaling 99.43 Gb of clean data. The Q20 and Q30 values across all samples were ≥98.38% and ≥95.35%, respectively, with GC content ranging from 42.79% to 44.1%(Supplementary Table 5 in S1 File). De novo assembly produced 147743 unigenes, of which 66.54% were annotated in at least one public database (NR: 90758; TrEMBL: 91058; GO: 80518; KEGG: 70028; SwissProt: 65385; KOG: 60040) (Supplementary Table 6 in S1 File). PCA of FPKM-normalized expression values showed that PC1 and PC2 explained 31.60% and 23.12% of the variance, respectively, with clear separation among tissue groups and high intra-group consistency (Fig 3A). Correlation analysis further confirmed strong repeatability within replicates, and a notable correlation was observed between stem and leaf transcript profiles (Fig 3B). The high-quality transcriptome profile, characterized by clear tissue segregation and strong reproducibility, provides a robust foundation for subsequent differential and integrated analyses.

Fig 3 — (A) PCA score plot. (B) Correlation heatmap of samples. All analyses included biological replicates (n = 3) for root, stem, leaf, and flower tissues.

3.4. Analysis of differentially expressed genes

Based on the threshold of |log2FC| ≥ 1 and FDR < 0.05, a total of 53610 differentially expressed genes (DEGs) were identified through pairwise comparisons across the six tissue combinations, with detailed information provided in Supplementary Table 7 in S1 File. Fig 4A shows that 519 DEGs were common among all six pairs of comparisons. Fig 4B displays the quantities of differentially expressed genes (DEGs) detected in pairwise tissue comparisons. These analyses identified 24654 DEGs between flower and leaf tissues (14761 upregulated, 9893 downregulated), 37090 between flower and root (17943 upregulated, 19147 downregulated), 24617 between stem and leaf (17992 upregulated, 6625 downregulated), 26984 between stem and root (11021 upregulated, 15963 downregulated), 32461 between leaf and root (13661 upregulated, 18800 downregulated), and 21331 between flower and stem (12448 upregulated, 8883 downregulated). To explore potential genes involved in secondary metabolism, DEGs were analyzed via GO enrichment. Terms related to the mevalonate pathway and ribosome/cofactor activities were enriched. This suggests that DEGs may support the cellular infrastructure for sesquiterpenoid biosynthesis, though direct roles require further validation (Fig 4C). Based on expression levels, the 53610 DEGs were classified into six clusters (Fig 4D). Genes in Cluster 1 showed relatively high expression in stem, those in Cluster 3 exhibited a gradually increasing expression trend across root, stem, leaf, and flower, and those in Cluster 6 displayed relatively high expression in root. Notably, the expression patterns of Clusters 2, 4, and 5 resembled those of DSM clusters 1, 2, and 3. Specifically, DEGs in Cluster 2 exhibited the highest expression in leaf, Cluster 4 in flower, and Cluster 5 showed a gradual decrease in expression from root to flower. This parallel expression pattern suggests that DEGs sharing similar tissue-specific trends with DSMs are likely key candidates involved in sesquiterpenoid biosynthesis and its regulation. Subsequently, KEGG functional enrichment analysis was performed on DEGs from Clusters 2, 4, and 5. Significantly, DEGs within each cluster were enriched in pathways related to sesquiterpenoid biosynthesis and terpenoid backbone biosynthesis(Fig 4E).

Fig 4 — (A) Venn diagram of DEGs in the six pairs of comparisons. (B) Number of DEGs in the six pairs of comparisons. (C) GO enrichment analysis of DEGs. (D) Cluster analysis of differentially expressed genes (DEGs). All DEGs were categorized into six distinct clusters based on their expression patterns using the Mfuzz R package. (E) KEGG enrichment analysis of DEGs in *Eupatorium lindleyanum* from Clusters 2、Clusters 4 and Clusters 5.

3.5. Identification of sesquiterpenoid biosynthesis-related genes through WGCNA-based co-expression network analysis

To further explore genes related to sesquiterpenoid biosynthesis, 30971 DEGs exhibiting expression patterns consistent with DSMs across different tissues were consolidated and subjected to WGCNA. These genes were grouped into 13 distinct color-coded modules (with different colors representing different modules; the gray module contains the remaining unrelated genes) (Fig 5A). Module–metabolites correlation analysis showed that seven modules—black, green, magenta, yellow, pink, red, and turquoise—were significantly correlated with the 10 DSMs (correlation coefficient > 0.8). Accordingly, these seven modules were selected for further analysis (Fig 5B). The expression levels of genes contained in each module are presented in Supplementary Figure 2 in S1 File. Genes highly expressed in flower were predominantly clustered in the black, green, magenta, and turquoise modules. Those with high expression in leaf were mainly enriched in the pink and yellow modules. Genes showing relatively high expression in stem, leaf, and flower were primarily assigned to the red module. Furthermore, the genes included in the seven selected modules were subjected to KEGG enrichment analysis respectively (Fig 5C). All modules except the pink one were annotated to the KEGG pathways: metabolic pathways(ko01100), biosynthesis of secondary metabolitess(ko01110), sesquiterpenoid and triterpenoid biosynthesiss(ko00909), and terpenoid backbone biosynthesiss(ko00900). Moreover, based on the WGCNA results, Correlation network diagrams were constructed for the genes within each module. Genes strongly correlated with the hub genes (edge weight ≥ 0.3) are shown in Supplementary Figure 3 in S1 File. Within the black, green, magenta, yellow, pink, red, and turquoise modules, a total of 8, 8, 8, 5, 8, 8, and 5 genes, respectively, were identified as hub genes with high topological importance and exhibited strong correlations with other genes in the network. These results indicate that the hub genes are key players involved in the biosynthesis of sesquiterpenoid metabolites in Eupatorium lindleyanum. Within the seven modules, genes mapped to ko00909 and ko00900 were selected and merged, yielding a total of 20 genes (88 unigenes) associated with sesquiterpenoid and terpenoid backbone biosynthesis, namely, GAO, GAS, PMK, ACAT, HMGCS, HMGCR, MVK, MVD, FDPS, IDI, dxs, dxr, ispD, ispE, ispF, ispG, ispH, FLDH, NES1, GDS, (Supplementary Table 8 in S1 File). These genes have been proposed as key players in regulating sesquiterpenoid biosynthesis in Eupatorium lindleyanum; however, their functional characterization still requires further investigation.

Fig 5 — (A) Cluster dendrogram of co-expressed genes. Each leaf represents a single gene, and the main branches correspond to 13 modules labeled with different colors. (B) Correlation heatmap between 13 module traits and 10 DSMs. Each cell at the intersection of a row and column represents the correlation coefficient. It is displayed based on the color scale on he right. The P values are displayed in the parentheses of each cell. (C) KEGG enrichment analysis of the genes in each module shows only the top 20 pathways with the most significant enrichment.

3.6. Construction of the sesquiterpenoid biosynthesis regulatory network in Eupatorium lindleyanum

Based on the results of WGCNA and KEGG enrichment analyses, the predicted sesquiterpenoid biosynthetic pathway is illustrated in Fig 6. The diagram illustrates the proposed biosynthetic route for sesquiterpenoids in Eupatorium lindleyanum, integrating both the mevalonate (MVA) and the 2-C-methyl-D-erythritol 4-phosphate (MEP) pathways, along with the subsequent enzymatic steps leading to the formation of sesquiterpenoid skeletons and their derivatives. In the MVA pathway, the sequential conversion of acetyl-CoA to isopentenyl diphosphate (IPP) is mediated by acetoacetyl-CoA thiolase (ACAT), hydroxymethylglutaryl-CoA synthase (HMGCS), hydroxymethylglutaryl-CoA reductase (HMGCR), mevalonate kinase (MVK), phosphomevalonate kinase (PMK), and diphosphomevalonate decarboxylase (MVD). Isopentenyl-diphosphate Delta-isomerase (IDI) subsequently catalyzes the reversible conversion of IPP to dimethylallyl diphosphate (DMAPP). In the MEP pathway, the initial steps involving pyruvate and glyceraldehyde-3-phosphate are catalyzed by 1-deoxy-D-xylulose-5-phosphate synthase (dxs) and 1-deoxy-D-xylulose-5-phosphate reductoisomerase (dxr), followed by a series of enzymatic reactions mediated by ispD, ispE, ispF, ispG, and ispH to ultimately generate IPP and DMAPP. The condensation of IPP and DMAPP into farnesyl diphosphate (FPP), the universal precursor of sesquiterpenoid, is catalyzed by farnesyl diphosphate synthase (FDPS). Following the formation of FPP, a diverse array of sesquiterpene synthases drive the structural diversification of terpenoid skeletons. These include germacrene A synthase (GAS), germacrene D synthase (GDS), and (3S,6E)-nerolidol synthase (NES1). Further modifications, such as oxidation and hydroxylation, are catalyzed by cytochrome P450s and dehydrogenases, including germacrene A oxidase (GAO), and NAD + -dependent farnesol dehydrogenase (FLDH). Taken together, Fig 6 presents a data-driven reconstruction of the sesquiterpenoid biosynthetic pathway in Eupatorium lindleyanum, in which each enzymatic step is supported by WGCNA-prioritized unigenes belonging to trait-associated modules. This approach not only consolidates fragmented transcriptome annotations into a coherent metabolic map, but also establishes a prioritized candidate list for subsequent functional characterization and genetic manipulation.

3.7. Analysis of transcription factors associated with sesquiterpenoid biosynthesis

To investigate the regulatory mechanisms underlying tissue-specific accumulation of sesquiterpenoid metabolites in Eupatorium lindleyanum, we characterized transcription factors exhibiting differential expression (|log₂ fold change| ≥ 1, FDR < 0.05) across four tissues and assessed their association with the abundance of ten distinct sesquiterpenoids. Significantly correlated transcription factor–metabolite pairs were identified based on stringent thresholds (|R| > 0.8, P < 0.05) (Fig 7). Among the five transcription factor families regulating sesquiterpenoid biosynthesis, all exhibited substantial associations with sesquiterpenoids, with cluster numbers ranging from 15 (NAC) to 23 (MYB). Specifically, the MYB family showed the highest number of significantly correlated clusters (23 clusters with 8 metabolites), followed by bHLH (21 clusters with 9 metabolites), AP2/ERF (20 clusters with 10 metabolites), WRKY (19 clusters with 9 metabolites), and NAC (15 clusters with 8 metabolites). The differences in the expression patterns of the transcription factors among the four tissues are presented in Suplementary Figure 4 in S1 File. These results suggest that these transcription factors play important roles in regulating sesquiterpenoid biosynthesis in different tissues of Eupatorium lindleyanum. While these correlation-based findings provide a robust framework for prioritizing candidate regulators, experimental validation—such as gene silencing or overexpression combined with metabolite profiling—is warranted to establish causal relationships between specific transcription factors and sesquiterpenoid biosynthesis.

Fig 7 — The color intensity of each grid cell reflects the strength of the correlation coefficient.

3.8. qRT‒PCR validation of candidate genes in the sesquiterpenoid biosynthesis pathway

To assess the reliability of the transcriptome data, the expression patterns of 12 candidate genes were examined using qRT-PCR. Their relative expression across the four examined tissues of Eupatorium lindleyanum showed a consistent trend with the FPKM values obtained from RNA-Seq. (Fig 8). Pearson correlation analysis revealed that eight of these genes exhibited significant positive correlations between the qRT-PCR and RNA-Seq datasets: GAS (r = 0.998, P = 0.002), GAO (r = 0.996, P = 0.004), MVD (r = 0.997, P = 0.003), dxs (r = 0.991, P = 0.009), ACAT (r = 0.984, P = 0.016), HMGCS (r = 0.978, P = 0.022), MVK (r = 0.967, P = 0.033), and PMK (r = 0.963, P = 0.037). By contrast, IDI exhibited a significant negative correlation (r = –0.963, P = 0.037), while HMGCR, FDPS, and dxr did not show statistically significant correlations (P > 0.05). This high concordance for the majority of genes (66.67%) validates the robustness of the RNA-Seq dataset and confirms its suitability for downstream expression analyses. The consistency between the two platforms further supports the fidelity of the transcriptomic profiling and the reliability of the bioinformatic inferences derived from it.

Fig 8 — The horizontal axis represents the four tissues of *Eupatorium lindleyanum*, the right vertical axis represents the relative expression levels of genes determined via qRT‒PCR, and the left vertical axis represents the FPKM values of the genes.

4. Discussion

The medicinal value of Eupatorium lindleyanum derives largely from its repertoire of bioactive sesquiterpenoids. While these compounds have been extensively studied for their pharmacological properties, the spatial organization of their biosynthesis within the plant—and the genes that govern it—has received far less attention [43]. This study aimed to fill that gap by integrating tissue-specific metabolomic and transcriptomic profiles. We integrated UPLC–MS/MS-based metabolomics with RNA-Seq transcriptomics to comprehensively profile sesquiterpenoid metabolites and their associated gene expression signatures across root, stem, leaf, and flower tissues. Through weighted gene co-expression network analysis (WGCNA) coupled with Mfuzz trend clustering, we identified structural genes and transcription factors highly correlated with the accumulation of key sesquiterpenoids and constructed a regulatory network model for sesquiterpenoid biosynthesis in this medicinal plant.

4.1. Tissue-specific accumulation of sesquiterpenoids and its transcriptional basis

In this study, we identified 321 sesquiterpenoid metabolites in Eupatorium lindleyanum, including various sesquiterpene lactones, sesquiterpene hydrocarbons, and their oxygenated derivatives. Notably, compounds such as eupalinolide B, zaluzanin C, isoalantolactone, dehydrocostus lactone, costunolide, spathulenol, and atractylenolide I/II have been previously characterized in other Asteraceae species, confirming the consistency of our findings with earlier reports [44]. Principal component analysis (PCA) and hierarchical clustering analysis revealed distinct tissue-specific signatures of these metabolites in root, stem, leaf, and flower. In root and stem, the relative abundance of sesquiterpenoids was generally low, with only a few metabolites showing high abundance. For example, Agalleudesmanol D (Flower vs Root Log₂FC = −7.98; Leaf vs Root Log₂FC = −7.96; Stem vs Root Log₂FC = −7.66), Calacorene oxide (Flower vs Root Log₂FC = −7.22; Leaf vs Root Log₂FC = −7.09; Stem vs Root Log₂FC = −6.96), and Curcumafuranol (Flower vs Root Log₂FC = −3.27; Leaf vs Root Log₂FC = −3.76; Stem vs Root Log₂FC = −5.71) were highly abundant in root. Similarly, 13-Nor-desacylcynaropicrin (Flower vs Stem Log₂FC = −3.45; Leaf vs Stem Log₂FC = −0.29; Stem vs Root Log₂FC=5.88) and Oxyphyllenone A (Flower vs Stem Log₂FC = −2.59; Leaf vs Stem Log₂FC = −4.59; Stem vs Root Log₂FC=2.04) were elevated in stem. In contrast, the leaf and flower contained relatively high levels of most sesquiterpenoids, with particularly elevated concentrations observed in the flower. This pattern suggests that the accumulation of sesquiterpenoids is tissue-specific and may be regulated by specific genes, as further explored in our integrative metabolomic and transcriptomic analysis below [45]. The tissue-specific variation in sesquiterpenoid metabolic profiles reflects their distinct physiological functions and adaptation to differential environmental exposures [46]. Root is involved primarily in nutrient absorption and belowground defense, whereas leaf and flower are exposed to relatively high levels of environmental stressors, such as ultraviolet (UV) radiation, pathogen attack, and herbivory. The elevated accumulation of sesquiterpenoids in leaf and flower suggests their involvement in both ecological and physiological functions. These compounds may contribute to chemical defense against biotic and abiotic stresses, while certain volatile sesquiterpenoids could also act as attractants for pollinators. Among these, compounds such as costunolide, dehydrocostus lactone, isoalantolactone, eupalinolide B, and zaluzanin C have been previously documented as defense-related sesquiterpenoids that accumulate in response to biotic or abiotic elicitation in Asteraceae species. Meanwhile, volatile sesquiterpenoids such as spathulenol may serve dual functions in both direct defense and pollinator attraction [47,48]. Therefore, we speculate that the tissue-specific distribution of sesquiterpenoids in Eupatorium lindleyanum is regulated by specific genes [49]. Furthermore, we systematically compared the sesquiterpenoid profiles across four tissues of Eupatorium lindleyanum and identified 299 differentially accumulated sesquiterpenoid metabolites (DSMs). The highest number of DSMs was observed in the flower vs. root comparison, followed by leaf vs. root, indicating a pronounced metabolic divergence between belowground and aboveground organs. This suggests that the metabolic flux of sesquiterpenoids is differentially channeled in root compared to aerial tissues [50]. Notably, 28 DSMs were consistently differentially accumulated across all six pairwise comparisons. These common DSMs may represent a core set of sesquiterpenoids involved in essential physiological processes that are modulated across tissues. Further clustering analysis using Mfuzz categorized the DSMs into three major clusters with distinct tissue-preferred accumulation patterns. Cluster 1 comprised metabolites with highest abundance in leaf, suggesting a role in foliar defense against herbivores and environmental stress. Cluster 2 included metabolites predominantly accumulated in flower, which may function as volatile attractants for pollinators or as protectants for reproductive tissues. Cluster 3 was characterized by root-enriched metabolites, likely involved in belowground defense, allelopathy, or symbiosis. This tissue-specific partitioning of sesquiterpenoids aligns with the concept of metabolic specialization, wherein organs optimize their chemical profiles according to their unique physiological and ecological challenges [51]. The striking accumulation of DSMs in flower and leaf compared to root and stem further supports the hypothesis that aerial tissues are more metabolically active in sesquiterpenoid biosynthesis, possibly due to their direct exposure to biotic and abiotic pressures [52]. Importantly, the tissue-specific expression patterns of DEGs recapitulated those of DSMs. Specifically, DEGs from Clusters 2, 4, and 5 exhibited leaf-predominant, flower-predominant, and root-to-flower decreasing trends, respectively, mirroring the accumulation trajectories of DSM Clusters 1, 2, and 3. This concordant transcript–metabolite partitioning provided the rationale for selecting these three DEG clusters as the input for WGCNA, thereby enriching for genes most likely to be functionally associated with sesquiterpenoid biosynthesis.

4.2. Integration of transcriptomic and metabolomic data: structural genes underlying the biosynthetic pathway

In plants, the biosynthesis of sesquiterpenoids is primarily governed by structural genes within the corresponding metabolic pathway [53]. To uncover the molecular basis of this metabolic specialization, the integration of transcriptomic and metabolomic analyses revealed that a total of 20 structural genes encoding 88 unigenes were involved in the biosynthesis of sesquiterpenoids in Eupatorium lindleyanum. The genes were classified into three groups on the basis of their involvement in distinct biosynthetic pathway branches. Specifically, Group I genes are involved in the mevalonate (MVA) pathway, including ACAT, HMGCS, HMGCR, MVK, PMK, and MVD. The Group II genes are associated with the 2-C-methyl-D-erythritol 4-phosphate (MEP) pathway and include dxs, dxr, ispD, ispE, ispF, ispG, and ispH. Group III genes participate in the downstream biosynthesis and modification of sesquiterpenoids, including FDPS, IDI, GAS, GDS, NES1, GAO, and FLDH. In previous studies, most of these genes have been characterized and are well documented to play crucial roles in the biosynthesis of plant sesquiterpenoids. ACAT, HMGCS, and HMGCR are key enzyme genes in the MVA pathway that catalyze the conversion of acetyl-CoA to mevalonate, which provides the precursor for sesquiterpenoid biosynthesis [54]. Dxs and dxr participate in the MEP pathway, further contributing to the pool of isopentenyl diphosphate (IPP) and dimethylallyl diphosphate (DMAPP) [55,56]. FDPS is a core enzyme in the sesquiterpenoid biosynthesis pathway, catalyzing the condensation of IPP and DMAPP to form farnesyl diphosphate (FPP), the universal precursor for all sesquiterpenoids [57,58]. GAS and GDS are sesquiterpene synthases that generate germacrene A and germacrene D skeletons, respectively, providing key intermediates for downstream diversification [59]. NES1 catalyzes the formation of nerolidol, a volatile sesquiterpene alcohol with ecological functions [60]. GAO is cytochrome P450 monooxygenases that introduce oxygen-containing functional groups through oxidation and hydroxylation reactions, enhancing structural complexity and bioactivity [61]. FLDH catalyzes the oxidation of farnesol, further diversifying sesquiterpenoid profiles [62]. The coordinated action of these genes within the sesquiterpenoid biosynthesis pathway dictates the tissue-specific distribution of sesquiterpenoids in Eupatorium lindleyanum. Interestingly, the core structural genes appear to be functionally conserved across medicinal Asteraceae. In Atractylodes macrocephala, transcriptome sequencing of five tissues revealed that multiple unigenes involved in sesquiterpenoid biosynthesis and 287 CYP450s were annotated, with AmFDPS experimentally verified to catalyze sesquiterpenoid biosynthesis. Similarly, in Atractylodes lancea, integrated NGS and SMRT sequencing identified nine full-length terpene synthase genes, among which AlTPS1 and AlTPS2 were functionally characterized: AlTPS1 catalyzes the biosynthesis of elemol, while AlTPS2 exhibits β-farnesene synthase activity, using FPP as a common substrate [63]. Furthermore, genome editing in Cichorium intybus demonstrated that inactivation of the four CiGAS genes, encoding germacrene A synthase, almost completely eliminated sesquiterpene lactone accumulation, confirming the essential role of GAS in the first committed step of STL biosynthesis [64]. Collectively, these studies illustrate that while Eupatorium lindleyanum shares the conserved terpenoid biosynthetic “toolkit” with these other Asteraceae species, the family as a whole deploys lineage-specific structural genes and enzymatic modifications on a conserved sesquiterpenoid scaffold, leading to significant metabolic diversification. Of note, the conversion of germacrene D to eupalinolide B, although chemically plausible, lacks direct metabolomic and WGCNA co-expression support in our data, and is therefore included in Fig 6 as a homology-based inference from other Asteraceae species [65]. Beyond the pathway-level inferences discussed above, the absence of root-dominant modules among the WGCNA-identified candidates merits a methodological clarification.It is noteworthy that although the WGCNA input included DEG Cluster 5 (root-to-flower decreasing trend), none of the seven modules significantly correlated with the selected DSMs displayed a root-predominant pattern. This can be explained by two methodological factors. First, our stepwise strategy employed Mfuzz as a coarse filter to reduce the candidate DEG pool based on trend similarity, followed by WGCNA as a more stringent tool to pinpoint genes with tight module–trait correlations, progressively narrowing the candidate scope. Second, and more critically, among the ten representative sesquiterpenoids used for module–trait correlation analysis, only three exhibited root-predominant accumulation, while the remaining seven were preferentially accumulated in flower. This numerical imbalance inherently biases WGCNA toward detecting modules associated with the dominant aerial-tissue metabolites, diminishing the statistical leverage of root-enriched genes in module identification. Thus, the absence of a root-dominant module does not indicate neglect of root-relevant candidates, but rather reflects the quantitative composition of the metabolite panel and the logic of our stepwise prioritization strategy.

4.3. Transcription factors as candidate regulators of sesquiterpenoid biosynthesis

The involvement of transcription factors (TFs) in the regulation of sesquiterpenoid biosynthesis is well documented. In this study, we identified five major TF families—AP2/ERF, NAC, WRKY, MYB, and bHLH—whose expression levels were significantly correlated with the differential accumulation of ten representative sesquiterpenoids in Eupatorium lindleyanum. Among these, the MYB family exhibited the highest number of significantly correlated clusters, followed by bHLH, WRKY, AP2/ERF, and NAC, suggesting a coordinated, tissue-specific regulatory network. These correlations are consistent with known functions of these families in terpenoid metabolism. For instance, MYB and bHLH TFs, often forming MBW complexes, are known to activate mevalonate pathway genes (e.g., HMGCR, FDPS) and terpene synthase genes [66]. AP2/ERF TFs have been implicated in directly activating sesquiterpene synthase genes like GAS by binding to cis-elements in their promoters, which aligns with the positive correlation observed between specific AP2/ERFs and germacrene-type sesquiterpenoids such as eupalinolide B [67]. Similarly, WRKY TFs, frequently associated with defense responses, showed strong correlations with compounds like isoalantolactone and shizukanolide H, potentially through the regulation of downstream cytochrome P450 monooxygenases that diversify sesquiterpene skeletons [68]. NAC TFs have also been implicated in terpenoid biosynthesis; for instance, SaNAC30 directly activates the MVA pathway gene SaMVK in sandalwood, and AaNACs regulate monoterpene synthase genes in kiwifrui [69]. It must be emphasized, however, that this analysis is based on correlation, not causation. While these findings provide a robust framework for prioritizing candidate regulators, experimental validation—such as gene silencing or overexpression assays—is essential to establish direct regulatory relationships [70].

4.4. Limitations

A key limitation is the correlative nature of our candidate gene assignments. Although WGCNA has prioritized genes strongly associated with sesquiterpenoid accumulation, these associations do not establish causality. The proposed pathway and regulatory network remain hypotheses requiring functional validation through gene silencing, overexpression, or CRISPR/Cas9 knockout. Likewise, the regulatory roles of transcription factors need direct evidence from promoter-binding assays. Additionally, our tissue-specific profiling provides only a spatial snapshot; temporal and environmental dynamics remain unexplored. Despite these limitations, this study offers a robust foundation and prioritized gene list for future functional characterization toward metabolic engineering of high-yielding Eupatorium lindleyanum.

5. Conclusion

This study presents a comprehensive metabolomic and transcriptomic analysis of Eupatorium lindleyanum, revealing tissue-specific accumulation patterns of 321 sesquiterpenoids across root, stem, leaf, and flower. Integration of metabolite profiling with transcriptome data identified 20 structural genes (88 unigenes) involved in sesquiterpenoid biosynthesis, along with five transcription factor families (AP2/ERF, NAC, WRKY, MYB, and bHLH) significantly correlated with sesquiterpenoid metabolites accumulation. WGCNA further prioritized key candidate genes and hub regulators within co-expression modules associated with specific sesquiterpenoids. These findings provide a valuable resource for understanding the molecular mechanisms governing tissue-specific sesquiterpenoid biosynthesis in Eupatorium lindleyanum and offer potential targets for genetic engineering and breeding programs aimed at enhancing production of bioactive compounds. Future work should focus on functional validation of these candidate genes and transcription factors.

Supporting information

S1 File. Supplementary Figure 1: photographs of different tissues of Eupatorium lindleyanum, including root, stem, leaf, and flower.

Supplementary Figure 2: cluster heatmap of genes contained in the WGCNA module with strong correlation to sesquiterpene metabolites. Supplementary Figure 3: correlation network diagram of hub genes in the seven WGCNA modules. Supplementary Figure 4: the expression levels of sesquiterpene biosynthesis-related transcription factors in root, stem, leaf, and flower. Supplementary Table 1: list of primers for candidate and reference genes. Supplementary Table 2: sesquiterpenoid metabolites identified in Eupatorium lindleyanum by UPLC-MS/MS. Supplementary Table 3: 299 differentially accumulated sesquiterpenoid metabolites (DSMs) identified across pairwise tissue comparisons in Eupatorium lindleyanum. Supplementary Table 4: DSMs overlapping across all six pairwise comparisons. Supplementary Table 5: transcriptome data quality control parameters for 12 samples. Supplementary Table 6: all genes annotated based on KEGG, NR, Swissprot, TrEMBL, KOG and GO databases. Supplementary Table 7: differentially expressed genes (DEGs) identified in the six pairwise tissue comparisons of Eupatorium lindleyanum. Supplementary Table 8: genes related to sesquiterpenoid biosynthesis in Eupatorium lindleyanum.

(ZIP)

pone.0359566.s001.zip^{(40.1MB, zip)}

Abbreviations

UPLC‒MS/MS: Ultra performance liquid chromatography‒tandem mass spectrometry
DSMs: Differentially accumulated sesquiterpenoid metabolites
DEGs: Differentially expressed genes
TFs: Transcription factors
MVA: Mevalonate pathway
MEP: 2-C-methyl-D-erythritol 4-phosphate pathway
GAO: Germacrene A oxidase
GAS: Germacrene A synthase
PMK: Phosphomevalonate kinase
ACAT: Acetyl-CoA acetyltransferase
HMGCS: Hydroxymethylglutaryl-CoA synthase
HMGCR: Hydroxymethylglutaryl-CoA reductase
MVK: Mevalonate kinase
MVD: Diphosphomevalonate decarboxylase
FDPS: Farnesyl diphosphate synthase
IDI: Isopentenyl-diphosphate Delta-isomerase
dxs: 1-deoxy-D-xylulose-5-phosphate synthase
dxr: 1-deoxy-D-xylulose-5-phosphate reductoisomerase
ispD: 2-C-methyl-D-erythritol 4-phosphate cytidylyltransferase
ispE: 4-(cytidine 5’-diphospho)-2-C-methyl-D-erythritol kinase
ispF: 2-C-methyl-D-erythritol 2,4-cyclodiphosphate synthase
ispG: (E)-4-hydroxy-3-methylbut-2-enyl-diphosphate synthase
ispH: 4-hydroxy-3-methylbut-2-en-1-yl diphosphate reductase
FLDH: NAD+-dependent farnesol dehydrogenase
NES1: (3S,6E)-nerolidol synthase
GDS: Germacrene D synthase
TCM: Traditional chinese medicine
RNA-seq: RNA sequencing
FPKM: Fragments per kilobase of transcript per million mapped reads
PCA: Principal component analysis
OPLS-DA: Orthogonal projections to latent structures-discriminant analysis
VIP: Variable importance in projection
FDR: False discovery rate
GO: Gene ontology
KEGG: Kyoto encyclopedia of genes and genomes
NR: Non-redundant protein sequence database
KOG: euKaryotic Orthologous Groups
qRT-PCR: Quantitative real-time polymerase chain reaction
WGCNA: Weighted Gene Co-Expression Network Analysis

Data Availability

The transcriptome sequencing data was deposited in NCBI database under accession number: PRJNA1242508.https://www.ncbi.nlm.nih.gov/sra/PRJNA1242508. All data tables in the supplementary materials are deposited in SCIENCE DB: https://doi.org/10.57760/sciencedb.010bi.

Funding Statement

This work was financially supported by the Natural Science Foundation of Jiangsu Food and Pharmaceutical Science College under Grant No. JSFPC2025007.

References

1.Antonio CNS, Selene M de M, Elnatan B de S, Raquel O dos SF. The genus Eupatorium L. (Asteraceae): A review of their antimicrobial activity. J Med Plants Res. 2017;11(3):43–57. doi: 10.5897/jmpr2016.6313 [DOI] [Google Scholar]
2.Wang Y, Wu J, Zhou M, Yang X, Niu Y, Guo K. Integrative metabolomic and transcriptomic analyses reveal flavonoid biosynthesis pathway in Eupatorium lindleyanum. Sci Rep. 2025;15(1):43151. doi: 10.1038/s41598-025-27287-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Yang J, Wang Y, Zhuo L, Lu G, Zhang M, Huang J, et al. Eupatorium lindleyanum DC Ameliorates Carbon Tetrachloride-Induced Hepatic Inflammation and Fibrotic Response in Mice. Pharmaceuticals (Basel). 2025;18(8):1228. doi: 10.3390/ph18081228 [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Lu J, Zheng C, Xue S, Gao Y, Chen G, Shan C, et al. Comprehensive Comparison of Three Different Medicinal Parts of Eupatorium lindleyanum DC. Using the RRLC-Q-TOF-MS-Based Metabolic Profile and In Vitro Anti-Inflammatory Activity. Molecules. 2024;29(15):3551. doi: 10.3390/molecules29153551 [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Geng H. Chemical constituents and their bioactivities of plants from the genus Eupatorium (2015-present). Biology (Basel). 2024;13(5):288. doi: 10.3390/biology13050288 [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Huang L, Xu D-Q, Chen Y-Y, Fu R-J, Yue S-J, Yin J-F, et al. Qualitative and quantitative analysis of chemical components in Eupatorium lindleyanum DC. by ultra-performance liquid chromatography-mass spectrometry integrated with anti-inflammatory activity research. J Sep Sci. 2021;44(17):3174–87. doi: 10.1002/jssc.202100208 [DOI] [PubMed] [Google Scholar]
7.Chu C, Ren H, Xu N, Xia L, Chen D, Zhang J. Eupatorium lindleyanum DC. sesquiterpenes fraction attenuates lipopolysaccharide-induced acute lung injury in mice. J Ethnopharmacol. 2016;185:263–71. doi: 10.1016/j.jep.2016.03.022 [DOI] [PubMed] [Google Scholar]
8.Jiang L, Zhang L, Zhang X. Eupalinilide B as a novel anti-cancer agent that inhibits proliferation and epithelial-mesenchymal transition in laryngeal cancer cells. J Int Med Res. 2022;50(1):3000605211067921. doi: 10.1177/03000605211067921 [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Wang X, Ma S, Lai F, Wang Y, Lou C. Traditional Applications, Phytochemistry, and Pharmacological Activities of Eupatorium lindleyanum DC.: A Comprehensive Review. Frontiers in Pharmacology. 2020:11. doi: 10.3389/FPHAR.2020.577124 [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Li L, Luo W, Wang Q, Wang D, Li Z, Tang R, et al. Multi-omic analysis highlights new candidate genes involved in triterpenoid and sesquiterpenoid biosynthesis in Codonopsis pilosula (Franch.) Nannf. Industrial Crops and Products. 2025;226:120685. doi: 10.1016/j.indcrop.2025.120685 [DOI] [Google Scholar]
11.Zhang Z, Tian Y, Qiao X, Li H, Ouyang L, Li X, et al. Integrated analysis of terpenoid profiles and full-length transcriptome reveals the central pathways of sesquiterpene biosynthesis in Atractylodes chinensis (DC.) Koidz. Int J Mol Sci. 2025;26(3):1074. doi: 10.3390/ijms26031074 [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Li C, Li Y, Wang J, Lu F, Zheng L, Yang L, et al. An independent biosynthetic route to frame a xanthanolide-type sesquiterpene lactone in Asteraceae. Plant J. 2025;121(2):e17199. doi: 10.1111/tpj.17199 [DOI] [PubMed] [Google Scholar]
13.Wu Y, Chen X, Hao F, Liu Y, Luo W, Zhu Y, et al. Biosynthesis of bridged tricyclic sesquiterpenes in Inula lineariifolia. Plant J. 2024;120(2):658–73. doi: 10.1111/tpj.17008 [DOI] [PubMed] [Google Scholar]
14.Yang X, Meng F, Cheng Q, Feng P, Song X, Chen W. Evolution of terpene synthases in the sesquiterpene biosynthesis pathway and analysis of their transcriptional regulatory network in Asteraceae. Hortic Res. 2025;12(12):uhaf229. doi: 10.1093/hr/uhaf229 [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Bogdanović M, Cankar K, Todorović S, Dragicević M, Simonović A, van Houwelingen A, et al. Tissue specific expression and genomic organization of bitter sesquiterpene lactone biosynthesis in Cichorium intybus L. (Asteraceae). Industrial Crops and Products. 2019;129:253–60. doi: 10.1016/j.indcrop.2018.12.011 [DOI] [Google Scholar]
16.Zhu F, Wen W, Cheng Y, Alseekh S, Fernie AR. Integrating multiomics data accelerates elucidation of plant primary and secondary metabolic pathways. aBIOTECH. 2023;4(1):47–56. doi: 10.1007/s42994-022-00091-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Gao H, Li X, Wang C, Li Y, Liu T, Chang N, et al. Integrated multiomics profiling elucidates the spatiotemporal metabolic dynamics and regulatory networks of the bioactive components of Trichosanthes kirilowii. Front Plant Sci. 2026;17:1735703. doi: 10.3389/fpls.2026.1735703 [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Chen D, Chen X, Zheng X, Zhu J, Xue T. Combined metabolomic and transcriptomic analysis reveals the key genes for triterpenoid biosynthesis in Cyclocarya paliurus. BMC Genomics. 2024;25(1):1197. doi: 10.1186/s12864-024-11125-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Hou J, Zhang Q, Wang R, Yao J, Zhong Y, Zhong Y, et al. Integration of transcriptomics and metabolomics provides insights into secondary cell wall deposition and terpenoid biosynthesis during stem development in Camphora officinarum. BMC Genomics. 2025;27(1):6. doi: 10.1186/s12864-025-12266-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Liu M, Li Y, Chen H, He C, Sun L, Zhang X, et al. Integrated omics profiles for exploring the potential mechanism underlying aroma formation in the terpenoid-rich aromatic plant Opisthopappus taihangensis and the bioactivity of its leaf essential oil. Agriculture Communications. 2024;2(4):100061. doi: 10.1016/j.agrcom.2024.100061 [DOI] [Google Scholar]
21.Wang Z, Xu X, Zhang P, Huang Y, Zhao C, El-Kassaby YA, et al. Multi-omics integration reveals organ-specific biosynthesis of flavonoids and terpenoids in the medicinal plant Bidens alba. Front Plant Sci. 2025;16:1675432. doi: 10.3389/fpls.2025.1675432 [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Hazrati S, Mousavi Z, Nicola S. Harvest time optimization for medicinal and aromatic plant secondary metabolites. Plant Physiol Biochem. 2024;212:108735. doi: 10.1016/j.plaphy.2024.108735 [DOI] [PubMed] [Google Scholar]
23.Li J, Liao Q, Zhou H, Hu R, Li Y, Hu Z, et al. Multi-omics analyses reveal regulatory networks underpinning metabolite biosynthesis in Nicotiana tabacum. Nat Commun. 2025;16(1):10339. doi: 10.1038/s41467-025-65299-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Zhong Y, Xie Y, Zhang D, Li G, Yu J. Integrated metabolomic and transcriptomic analysis of metabolic diversity and biosynthesis of glucosinolates and flavonoids in various cultivars of radish microgreens. Food Bioscience. 2024;59:104055. doi: 10.1016/j.fbio.2024.104055 [DOI] [Google Scholar]
25.Zhang X, Chen X, Wang Y, Liu M, Wang C, Zhou T. Integrative transcriptomic and metabolomic analysis of Drynaria roosii reveals genes involved in the biosynthesis of medicinal compounds. Sci Rep. 2026;16(1):9047. doi: 10.1038/s41598-026-39037-x [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Li Y, Sun Z, Lu J, Jin Z, Li J. Integrated transcriptomics and metabolomics analysis provide insight into anthocyanin biosynthesis for sepal color formation in Heptacodium miconioides. Front Plant Sci. 2023;14:1044581. doi: 10.3389/fpls.2023.1044581 [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Jiang Y, Wang J, Tu X, Zhong M, Wan B. Integrative transcriptomic and metabolomic analysis reveals the molecular basis of leaf variegation in Cymbidium ensifolium. Front Plant Sci. 2026;17:1712811. doi: 10.3389/fpls.2026.1712811 [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Lu X, Wei Y, Sun J, Xiao B, Zhang X, Li W, et al. A comparative study of three nucleic acid integrity assay systems. Biopreserv Biobank. 2023;21(6):624–30. doi: 10.1089/bio.2022.0171 [DOI] [PubMed] [Google Scholar]
29.Chen S, Zhou Y, Chen Y, Gu J. fastp: an ultra-fast all-in-one FASTQ preprocessor. Bioinformatics. 2018;34(17):i884–90. doi: 10.1093/bioinformatics/bty560 [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Xu W, Ayyaz A, Hannan F, Rehman KM, Qin T, Song W, et al. Integrative metabolomic and transcriptomic analyses reveal mechanisms of hexavalent chromium toxicity in contrasting rapeseed cultivars. Agronomy,2025,15(12):2892–2892. doi: 10.3390/AGRONOMY15122892 [DOI] [Google Scholar]
31.Zheng Y, Jiao C, Sun H, Rosli H, Pombo M, Zhang P, et al. iTAK: A program for genome-wide prediction and classification of plant transcription factors, transcriptional regulators, and protein kinases. Molecular Plant,2016,9(12):1667–70. doi: 10.1016/j.molp.2016.09.014 [DOI] [PubMed] [Google Scholar]
32.Langfelder P, Horvath S. WGCNA: An R package for weighted correlation network analysis. BMC Bioinformatics. 2008;9:559. doi: 10.1186/1471-2105-9-559 [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Yang G, Yang L, Xu F. Isoalantolactone: A review on its pharmacological effects. Frontiers in Pharmacology. 2024;151:453205–1453205. doi: 10.3389/FPHAR.2024.1453205 [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Zhang J, Zhang H-L, Xu X-R, Feng Y-L, Zhu Q-M, Morisseau C, et al. Targeting PBK with small-molecule 1-O-acetyl-4R,6S-britannilactone for the treatment of neuroinflammation. Proc Natl Acad Sci U S A. 2025;122(29):e2502593122. doi: 10.1073/pnas.2502593122 [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Gu S-L, Liu X-S, Xu Z-S, Li L-L, Wu X-J, Li F-L, et al. Eupalinolide B alleviates rheumatoid arthritis through the promotion of apoptosis and autophagy via regulating the AMPK/mTOR/ULK-1 signaling axis. Int Immunopharmacol. 2025;148:114179. doi: 10.1016/j.intimp.2025.114179 [DOI] [PubMed] [Google Scholar]
36.Jin M, Lee HJ, Ryu JH, Chung KS. Inhibition of LPS-induced NO production and NF-kappaB activation by a sesquiterpene from Saussurea lappa. Arch Pharm Res. 2000;23(1):54–8. doi: 10.1007/BF02976467 [DOI] [PubMed] [Google Scholar]
37.Jeong S-H, Koo S-J, Ha J-H, Ryu S-Y, Park H-J, Lee K-T. Induction of apoptosis by yomogin in human promyelocytic leukemic HL-60 cells. Biol Pharm Bull. 2004;27(7):1106–11. doi: 10.1248/bpb.27.1106 [DOI] [PubMed] [Google Scholar]
38.Liang S, Chu Chen, Li R, Jiang G, Du L. Costunolide and dehydrocostus lactone alleviate ulcerative colitis via regulating TLR4, NF-κB and PI3K expression. Sci Rep. 2024;14(1):29777. doi: 10.1038/s41598-024-77021-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Trentin A, Santos A, Guedes A, Pizzolatti M, Yunes R, Calixto J. Antinociception caused by the extract of Hedyosmum brasiliense and its active principle, the sesquiterpene lactone 13-hydroxy-8,9-dehydroshizukanolide. Planta Med. 1999;65(6):517–21. doi: 10.1055/s-1999-14007 [DOI] [PubMed] [Google Scholar]
40.Lin X, Peng Z, Su C. Potential anti-cancer activities and mechanisms of costunolide and dehydrocostuslactone. Int J Mol Sci. 2015;16(5):10888–906. doi: 10.3390/ijms160510888 [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Yang S, Gao Z, Wu Y, Hu G, Yue J. ChemInform abstract: Chlorahololides C—F: A new class of potent and selective potassium channel blockers from Chloranthus holostegius. ChemInform. 2008;39(37). doi: 10.1002/chin.200837167 [DOI] [Google Scholar]
42.Geng X, Tang R, Zhang A, Du Z, Yang L, Xu Y, et al. Mining, expression, and phylogenetic analysis of volatile terpenoid biosynthesis-related genes in different tissues of ten Elsholtzia species based on transcriptomic analysis. Phytochemistry. 2022;203:113419. doi: 10.1016/j.phytochem.2022.113419 [DOI] [PubMed] [Google Scholar]
43.Jiang Y, Liu Y, Song H, He J. Integrated transcriptomics and metabolomics uncover the molecular basis of flavonoid accumulation in the rhizomes of two Cymbidium tortisepalum var. longibracteatum cultivars. Scientia Horticulturae. 2022;293:110738. doi: 10.1016/j.scienta.2021.110738 [DOI] [Google Scholar]
44.Liu PY, Liu D, Li WH, Zhao T, Sauriol F, Gu YC, et al. Chemical constituents of plants from the genus Eupatorium (1904–2014). ChemInform. 2015;46(49). doi: 10.1002/chin.201549256 [DOI] [PubMed] [Google Scholar]
45.Chen F, Wei Y-X, Zhang J-M, Sang X-M, Dai C-C. Transcriptomics analysis investigates sesquiterpenoids accumulation pattern in different tissues of Atractylodes lancea (Thunb.) DC. plantlet. Plant Cell Tiss Organ Cult. 2017;130(1):73–90. doi: 10.1007/s11240-017-1205-8 [DOI] [Google Scholar]
46.Bai Y, Liu X, Baldwin IT. Using synthetic biology to understand the function of plant specialized metabolites. Annu Rev Plant Biol. 2024;75(1):629–53. doi: 10.1146/annurev-arplant-060223-013842 [DOI] [PubMed] [Google Scholar]
47.Kour S, Singh S, Wani TA, Kaloo ZA. Chemical characterization and the intrusion through elicitation and Agrobacterium rhizogenes mediated hairy root transformation in Saussurea costus C.B. Clarke. Physiol Mol Biol Plants. 2023;29(1):87–91. doi: 10.1007/s12298-022-01270-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
48.Kumar A, Agnihotri VK. NMR based profiling of sesquiterpene lactones in Saussurea lappa roots collected from different location of Western Himalaya. Nat Prod Res. 2022;36(2):621–4. doi: 10.1080/14786419.2020.1789635 [DOI] [PubMed] [Google Scholar]
49.Liao H, Quan H, Huang B, Ji H, Zhang T, Chen J, et al. Integrated transcriptomic and metabolomic analysis reveals the molecular basis of tissue-specific accumulation of bioactive steroidal alkaloids in Fritillaria unibracteata. Phytochemistry. 2023;214:113831. doi: 10.1016/j.phytochem.2023.113831 [DOI] [PubMed] [Google Scholar]
50.Fu X, Zhang Y, Yu M, Zheng H, Pan Y, Liu P, et al. Specialized structures and developmental mechanisms of secondary metabolite-synthesizing organs and tissues in medicinal plants. J Integr Plant Biol. 2026;68(4):926–41. doi: 10.1111/jipb.70133 [DOI] [PubMed] [Google Scholar]
51.Schneider GF, Carlson CA, Jos EM, Beckman NG. Compartmentalization of specialized metabolites across vegetative and reproductive tissues in two sympatric Psychotria species. Am J Bot. 2023;110(7):e16191. doi: 10.1002/ajb2.16191 [DOI] [PubMed] [Google Scholar]
52.He S, Wang J, Yang N, Li H, Li K, Li L, et al. Regulatory mechanisms underlying stress-induced accumulation of plant secondary metabolites. Journal of Applied Research on Medicinal and Aromatic Plants. 2026;50:100689. doi: 10.1016/j.jarmap.2025.100689 [DOI] [Google Scholar]
53.Zhang Z-A, Xue X-Y, Wu J-Y, Zhao X, Zhang H, Sun K, et al. CYP71P2 regulated by MYB44 modulates oxygenated sesquiterpenoids biosynthesis in Atractylodes lancea. Plant Physiol Biochem. 2025;229(Pt D):110700. doi: 10.1016/j.plaphy.2025.110700 [DOI] [PubMed] [Google Scholar]
54.Pütter KM, van Deenen N, Unland K, Prüfer D, Schulze Gronover C. Isoprenoid biosynthesis in dandelion latex is enhanced by the overexpression of three key enzymes involved in the mevalonate pathway. BMC Plant Biol. 2017;17(1):88. doi: 10.1186/s12870-017-1036-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
55.You MK, Lee YJ, Kim JK, Baek SA, Jeon YA, Lim SH, et al. The organ-specific differential roles of rice DXS and DXR, the first two enzymes of the MEP pathway, in carotenoid metabolism in Oryza sativa leaves and seeds. BMC Plant Biol. 2020;20(1):167. doi: 10.1186/s12870-020-02357-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
56.Banerjee A, Wu Y, Banerjee R, Li Y, Yan H, Sharkey TD. Feedback inhibition of deoxy-D-xylulose-5-phosphate synthase regulates the methylerythritol 4-phosphate pathway. J Biol Chem. 2013;288(23):16926–36. doi: 10.1074/jbc.M113.464636 [DOI] [PMC free article] [PubMed] [Google Scholar]
57.Liu X-M, Tao T-T, Meng X-X, Zhang W-W, Chang J, Xu F. Cloning and Expression Analysis of a Farnesyl Diphosphate Synthase (FPPS) Gene from Chamaemelum nobile. Not Bot Horti Agrobo. 2017;45(2):358–64. doi: 10.15835/nbha45210858 [DOI] [Google Scholar]
58.Zhang D, Tang X, Chen L, Qiu X, Song C, Wang H, et al. Functional characterization and transcriptional activity analysis of Dryopteris fragrans farnesyl diphosphate synthase genes. Front Plant Sci. 2023;14:1105240. doi: 10.3389/fpls.2023.1105240 [DOI] [PMC free article] [PubMed] [Google Scholar]
59.Cheng W, Zhi Y, Chen F, Xiao X, Lu H, Li R, et al. Characterization and functional reconstruction of a highly productive germacrene A synthase from Liriodendron chinense. Plant Biotechnol J. 2025;23(6):1927–37. doi: 10.1111/pbi.70023 [DOI] [PMC free article] [PubMed] [Google Scholar]
60.Aharoni A, Giri AP, Deuerlein S, Griepink F, de Kogel W-J, Verstappen FWA, et al. Terpenoid metabolism in wild-type and transgenic Arabidopsis plants. Plant Cell. 2003;15(12):2866–84. doi: 10.1105/tpc.016253 [DOI] [PMC free article] [PubMed] [Google Scholar]
61.Takahashi S, Yeo Y-S, Zhao Y, O’Maille PE, Greenhagen BT, Noel JP, et al. Functional characterization of premnaspirodiene oxygenase, a cytochrome P450 catalyzing regio- and stereo-specific hydroxylations of diverse sesquiterpene substrates. J Biol Chem. 2007;282(43):31744–54. doi: 10.1074/jbc.M703378200 [DOI] [PMC free article] [PubMed] [Google Scholar]
62.Bhandari J, Fitzpatrick AH, Crowell DN. Identification of a novel abscisic acid-regulated farnesol dehydrogenase from Arabidopsis. Plant Physiol. 2010;154(3):1116–27. doi: 10.1104/pp.110.157784 [DOI] [PMC free article] [PubMed] [Google Scholar]
63.Wu J, Hu J, Yu H, Lu J, Jiang L, Liu W, et al. Full-length transcriptome analysis of two chemotype and functional characterization of genes related to sesquiterpene biosynthesis in Atractylodes lancea. Int J Biol Macromol. 2023;225:1543–54. doi: 10.1016/j.ijbiomac.2022.11.210 [DOI] [PubMed] [Google Scholar]
64.Cankar K, Bundock P, Sevenier R, Häkkinen ST, Hakkert JC, Beekwilder J, et al. Inactivation of the germacrene A synthase genes by CRISPR/Cas9 eliminates the biosynthesis of sesquiterpene lactones in Cichorium intybus L. Plant Biotechnol J. 2021;19(12):2442–53. doi: 10.1111/pbi.13670 [DOI] [PMC free article] [PubMed] [Google Scholar]
65.Steliopoulos P, Wüst M, Adam K-P, Mosandl A. Biosynthesis of the sesquiterpene germacrene D in Solidago canadensis: 13C and (2)H labeling studies. Phytochemistry. 2002;60(1):13–20. doi: 10.1016/s0031-9422(02)00068-7 [DOI] [PubMed] [Google Scholar]
66.Singh S, Chhatwal H, Pandey A. Deciphering the complexity of terpenoid biosynthesis and its multi-level regulatory mechanism in plants. J Plant Growth Regul. 2024;43(10):3320–36. doi: 10.1007/s00344-024-11347-2 [DOI] [Google Scholar]
67.Shen S-L, Yin X-R, Zhang B, Xie X-L, Jiang Q, Grierson D, et al. CitAP2.10 activation of the terpene synthase CsTPS1 is associated with the synthesis of (+)-valencene in “Newhall” orange. J Exp Bot. 2016;67(14):4105–15. doi: 10.1093/jxb/erw189 [DOI] [PMC free article] [PubMed] [Google Scholar]
68.Liang H, Liu W, Zhao Z, Li Y, Zha L. Genome-wide identification and expression analysis of the WRKY transcription factors related to sesquiterpenes biosynthesis in Atractylodes lancea. Front Genet. 2025;16:1551991. doi: 10.3389/fgene.2025.1551991 [DOI] [PMC free article] [PubMed] [Google Scholar]
69.Li X, Meng S, Zhou Y, Wang D, Bian Z, Hu L, et al. Genome-wide analysis of NAC gene family and function exploration of SaNAC30 in Santalum album L. Industrial Crops and Products. 2025;227:120827. doi: 10.1016/j.indcrop.2025.120827 [DOI] [Google Scholar]
70.Zeng X, Zhang S, Zhang Z, Chen Y, Wang S, Tang Z, et al. Integration of transcriptomics and metabolomics reveals the mechanism of aroma formation during apricot fruit development. Scientia Horticulturae. 2025;350:114351. doi: 10.1016/j.scienta.2025.114351 [DOI] [Google Scholar]

PLoS One. doi: 10.1371/journal.pone.0359566.r001

Decision Letter 0

Rongchun Han

5 Aug 2026

Dear Dr. Wang,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by Oct 04 2026 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

A letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

As the corresponding author, your ORCID iD is verified in the submission system and will appear in the published article. PLOS supports the use of ORCID, and we encourage all coauthors to register for an ORCID iD and use it as well. Please encourage your coauthors to verify their ORCID iD within the submission system before final acceptance, as unverified ORCID iDs will not appear in the published article. Only the individual author can complete the verification step; PLOS staff cannot verify ORCID iDs on behalf of authors.

We look forward to receiving your revised manuscript.

Kind regards,

Rongchun Han

Academic Editor

PLOS One

Journal Requirements:

1.Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Please note that PLOS One has specific guidelines on code sharing for submissions in which author-generated code underpins the findings in the manuscript. In these cases, we expect all author-generated code to be made available without restrictions upon publication of the work. Please review our guidelines at https://journals.plos.org/plosone/s/materials-and-software-sharing#loc-sharing-code and ensure that your code is shared in a way that follows best practice and facilitates reproducibility and reuse.

3. Please note that funding information should not appear in any section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submission form. Please remove any funding-related text from the manuscript.

4. We note that the grant information you provided in the ‘Funding Information’ and ‘Financial Disclosure’ sections do not match.

When you resubmit, please ensure that you provide the correct grant numbers for the awards you received for your study in the ‘Funding Information’ section.

5. When completing the data availability statement of the submission form, you indicated that you will make your data available on acceptance. We strongly recommend all authors decide on a data sharing plan before acceptance, as the process can be lengthy and hold up publication timelines. Please note that, though access restrictions are acceptable now, your entire data will need to be made freely accessible if your manuscript is accepted for publication. This policy applies to all data except where public deposition would breach compliance with the protocol approved by your research ethics board. If you are unable to adhere to our open data policy, please kindly revise your statement to explain your reasoning and we will seek the editor's input on an exemption. Please be assured that, once you have provided your new statement, the assessment of your exemption will not hold up the peer review process.

6. PLOS requires an ORCID iD for the corresponding author in Editorial Manager on papers submitted after December 6th, 2016. Please ensure that you have an ORCID iD and that it is validated in Editorial Manager. To do this, go to ‘Update my Information’ (in the upper left-hand corner of the main menu), and click on the Fetch/Validate link next to the ORCID field. This will take you to the ORCID site and allow you to create a new iD or authenticate a pre-existing iD in Editorial Manager.

7. Please include captions for your Supporting Information files at the end of your manuscript, and update any in-text citations to match accordingly. Please see our Supporting Information guidelines for more information: http://journals.plos.org/plosone/s/supporting-information.

8. If the reviewer comments include a recommendation to cite specific previously published works, please review and evaluate these publications to determine whether they are relevant and should be cited. There is no requirement to cite these works unless the editor has indicated otherwise.

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

Reviewer #1: Yes

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously? -->?>

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available??>

The PLOS Data policy

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English??>

Reviewer #1: Yes

Reviewer #2: Yes

**********

Reviewer #1: The paper is written well, acceptable with some minor revise.

1. Just refer to the type of your cDNA sequencing library, sranded or unstranded?

2. Explain in the text that why you didn't quantified and qualified volatile sesquiterpenes with GC-FID and GC-MS.

3. Add your metabolite dataset, tables and identified compounds in a open platforms like figshare.

Reviewer #2: I included an attached Word document titled "Comments for article" where I explain all my comments about the article.

This manuscript addresses a genuine, previously unexplored question: the genetic basis of tissue-specific sesquiterpenoid biosynthesis in Eupatorium lindleyanum. Its main strength is the scale and internal consistency of the multi-omics dataset — 321 sesquiterpenoids profiled across four tissues, 99.43 Gb of RNA-Seq data, and 22 candidate biosynthetic genes — figures we cross-checked throughout this review and found to be traceable and coherent across the Abstract, Results, and Methods. The integrative pipeline (Mfuzz clustering, WGCNA, and independent qRT-PCR validation) is methodologically sound, grounded in well-established terpenoid biochemistry, and appropriately careful to frame gene-metabolite associations as correlative rather than causal. The resulting candidate gene and transcription factor list is a genuinely useful, prioritized resource for future functional validation in this species. Our detailed comments below aim to strengthen the manuscript's rigor and clarity, not to question the value of its underlying dataset or research question

**********

what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy

Reviewer #1: No

Reviewer #2: Yes: Olmedo Jesus Cuaspud Caliz

**********

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

To ensure your figures meet our technical requirements, please review our figure guidelines: https://journals.plos.org/plosone/s/figures

You may also use PLOS’s free figure tool, NAAS, to help you prepare publication quality figures: https://journals.plos.org/plosone/s/figures#loc-tools-for-figure-preparation.

NAAS will assess whether your figures meet our technical requirements by comparing each figure against our figure specifications.

Attachment

Submitted filename: Comments for article.docx

pone.0359566.s002.docx^{(16.2KB, docx)}

PLoS One. 2026 Oct 1;21(10):e0359566. doi: 10.1371/journal.pone.0359566.r002

Author response to Decision Letter 1

27 Aug 2026

Point-by-Point Response to Reviewer 1

1. Just refer to the type of your cDNA sequencing library, sranded or unstranded?

Response: We thank the reviewer for this highly professional question. We acknowledge that stranded RNA-seq offers certain advantages in specific applications. In our study, the cDNA libraries were constructed using the standard non-stranded (unstranded) double-stranded cDNA synthesis protocol. We have now explicitly stated this information in the Methods 2.5 section of the revised manuscript (Lines 175-177). We recognize that this is a technical limitation.

We would like to clarify that the use of unstranded libraries does not materially affect the main conclusions of our study. All our key analyses—including gene-level expression quantification, WGCNA, and metabolome-transcriptome integration—are based on gene-level abundance, which does not require strand orientation information. While strand-specific data can offer higher resolution for distinguishing overlapping transcripts, our study does not involve analyses that depend on strand specificity, such as antisense transcript detection, non-coding RNA annotation, or promoter-associated RNA analysis. Therefore, we are confident that the unstranded library type is appropriate for the objectives of this study and does not compromise the validity of our findings.

2. Explain in the text that why you didn't quantified and qualified volatile sesquiterpenes with GC-FID and GC-MS.

Response: We sincerely thank the reviewer for this valuable and professional comment. In response, we have added a brief explanation at the end of Section 2.2 (Plant sample pretreatment and metabolite extraction procedures, lines 134-137) to clarify our methodological choice. Although most sesquiterpenes are volatile in nature, their detection by UPLC-MS does not rely on vaporization but on solubility, which is equally valid. Importantly, many of our target compounds are thermally labile (e.g., sesquiterpene lactones) and exist as high-polarity glycosides or acids, which are unsuitable for GC analysis due to decomposition or high boiling points. Therefore, UPLC-MS was deliberately chosen to detect both the volatile aglycones and their polar derivatives under a single, mild chromatographic condition, providing more comprehensive and degradation-free quantitative data than GC-based methods.

3. Add your metabolite dataset, tables and identified compounds in a open platforms like figshare.

Response: We appreciate the reviewer's constructive suggestion. As recommended, we have now deposited our metabolite dataset, including all tables and identified compounds, in the open-access platform SCIENCE DB. The data can be accessed at https://www.scidb.cn/anonymous/ZVlKbmFp. Correspondingly, we have added a Data Availability Statement at the end of the manuscript to indicate this information (lines 712-713).

Point-by-Point Response to Reviewer 2

Minor comments

1. Supplementary Figure 1: Add a scale bar (or a ruler in the photo) so tissue size is interpretable.

Response: We thank the reviewer for this suggestion. We have now added scale bars to each individual panel in Supplementary Figure 1. Because different tissue sections of Eupatorium lindleyanum were imaged at different magnifications, we have included separate scale bars for each subpanel to ensure accurate size interpretation.

2. Supplementary Table 1: Add amplicon size (bp) and gene annotation.

Response: We greatly appreciate the reviewer's insightful and constructive suggestions. In response, we have supplemented Supplementary Table 1 with two additional columns, namely "Gene Annotation" and "Amplicon Size (bp)" (inserted as Columns 2 and 6, respectively), to ensure that all essential information regarding the qPCR primers is fully presented.

3. Supplementary Table 2: Add mass error (ppm), annotation level, database evidence, and retention time.

We thank the reviewer for the constructive suggestion. In response, we have added mass error (ppm), annotation level, and database evidence in columns C, D, and E of Supplementary Table 2, respectively. Regarding the retention time, we regret that we are unable to provide this information. The metabolomics analysis was conducted in collaboration with a third-party service provider, and despite our sincere efforts to obtain the retention time data, they declined to disclose it due to commercial confidentiality. We hope the reviewer can kindly understand this constraint. We appreciate your understanding and valuable input.

4. Figure 1B: The legend is unnecessary, since every point is already labeled; remove it and enlarge the panel. Also correct the typo “Stam-3” to “Stem-3.”

Response: We thank the reviewer for pointing out these oversights in our work. In response, we have removed the legend from Figure 1B, as each data point is already clearly labeled, and have appropriately enlarged the panel to improve readability. The typographical error "Stam-3" has also been corrected to "Stem-3" in the figure (lines 273).

5. Figure 3A: Use the same color scheme as Figure 2B and remove the legend.

Response:We thank the reviewer for the suggestion. We have revised Figure 3A by adopting the same color scheme as used in Figure 2B and have removed the legend accordingly (lines 327).

6. Figures 4E and 5C: Sort the Y-axis categories by Rich Factor, from highest to lowest, to make the most significant terms easier to identify.

Response:We thank the reviewer for this helpful suggestion. As recommended, we have revised Figure 4E by sorting the Y-axis categories in descending order of Rich Factor to highlight the most significant terms. Regarding Figure 5C, we attempted to apply the same sorting procedure; however, we found that after reordering, several key metabolic pathways—namely, metabolic pathways (ko01100), biosynthesis of secondary metabolites (ko01110), sesquiterpenoid and triterpenoid biosynthesis (ko00909), and terpenoid backbone biosynthesis (ko00900)—could no longer be fully displayed in the figure due to space constraints. Given that these pathways are critically relevant to our study, we respectfully request to retain the original layout of Figure 5C. We hope the reviewer can kindly understand this practical limitation. We greatly appreciate your understanding and valuable feedback.

7. Figure 6 / lines 421 and 614: CYP71D55 (“premnaspirodiene,” not “prenaspirodiene,” per Takahashi et al.) is characterized in Hyoscyamus muticus (Solanaceae), and CDN1 specifically initiates gossypol biosynthesis in cotton (Gossypium, Malvaceae) — neither pathway is established in Asteraceae. Please confirm whether solavetivone or gossypol were actually detected among the 321 profiled metabolites; if not, qualify or remove these branches from Figure 6, and correct the enzyme name spelling throughout.

Response: We sincerely thank the reviewer for pointing out this critical issue and for providing the relevant literature reference. Upon careful verification, we fully agree with the reviewer's assessment that neither CYP71D55 nor CDN1 has been functionally characterized in Asteraceae, nor have solavetivone or gossypol been confirmed as biosynthetic products in this family. Accordingly, we have removed both genes and their associated pathway branches from Figure 6. In addition, we have revised the corresponding description throughout the manuscript: the statement "22 structural genes (91 unigenes)" has been corrected to "20 structural genes (88 unigenes)" to reflect this change. We greatly appreciate the reviewer's valuable and constructive comments, which have helped us rectify an oversight in our work. We are grateful for your guidance and understanding.

8. Section 3.7: The stated ranking “MYB, followed by bHLH, WRKY, AP2/ERF, and NAC” does not match the reported cluster counts (23, 21, 19, 20, 15) — the correct descending order would place AP2/ERF (20) before WRKY (19). Please correct the order (or the counts, if the order is what is accurate).

Response: We thank the reviewer for carefully identifying this inconsistency. Upon rechecking our data, we confirm that the reviewer is correct. We have now revised the order in Section 3.7 to reflect the correct descending ranking based on the cluster counts, which should be: MYB (23), followed by bHLH (21), AP2/ERF (20), WRKY (19), and NAC (15). The corresponding text has been amended accordingly (line 452). We appreciate the reviewer's meticulous attention to detail and constructive suggestion.

9. Sections 2.8/3.8: The Methods and the Results section title both describe the 12 qPCR-validated genes as “candidate genes,” but the Results text calls them “randomly selected genes.” Please clarify which description is correct.

Response:We thank the reviewer for identifying this inconsistency. Initially, we intended to randomly select genes for qPCR validation of the transcriptome data. However, we subsequently chose 12 genes from the 20 candidate genes for validation. Due to our oversight, the wording in the Methods and Results sections was not aligned. We have now carefully revised the manuscript and uniformly changed the description to "candidate genes" throughout both sections. We apologize for this error and appreciate the reviewer's careful reading and constructive suggestion.

10. Figure 5A: The seven modules are listed in two different orders within the same section (“…yellow, pink, red, turquoise” vs. “…pink, red, turquoise, yellow”), risking miscounted hub-gene attribution. Please use a consistent order throughout.

Response:We thank the reviewer for pointing out this inconsistency. We have carefully revised Section 3.5 to ensure that the seven modules are consistently listed in the same order throughout—namely, yellow, pink, red, and turquoise (line 385). We have also thoroughly double-checked the hub-gene counts corresponding to each module to ensure accuracy. We appreciate the reviewer's meticulous attention to detail and constructive suggestion.

11. Line 121: Specify the age of the plant/tissue sampled, as this factor can significantly affect tissue-specific differential expression.

Response: We thank the reviewer for this valuable suggestion. In response, we have now specified the age and developmental stage of the sampled plants in Section 2.1 (line 115) of the revised manuscript. Specifically, we have added that the plants were 17-month-old biennials and at the full-flowering stage at the time of tissue collection.

12. Line 51 (ref. 6): This reference (“Structural characterization of the plastid genome and comparative mitochondrial genomics of Eupatorium lindleyanum”) concerns organellar genomics, not phytochemistry or compound isolation. Please correct the citation.

Response:We thank the reviewer for carefully identifying this citation error. We sincerely apologize for our negligence during the literature retrieval process, which led to the inappropriate citation of an organellar genomics reference in the context of phytochemical isolation. We have now corrected this mistake by replacing the original reference with the appropriate one that accurately supports the statement regarding compound isolation from Eupatorium lindleyanum. The revised reference has been updated as reference [6] in the manuscript.

13. Line 73 (ref. 15): This reference (Wu et al. 2021, safflower genome) focuses on linoleic acid and flavonoid biosynthesis, not sesquiterpenoids or caryophyllene gene clusters. Please confirm whether it supports this specific claim or should be cited in another context.

Response: We thank the reviewer for carefully identifying this citation error. We sincerely apologize for our oversight during the reference compilation process, which led to the inappropriate citation of Wu et al. (2021) in the context of sesquiterpenoid biosynthesis. The claim regarding comparative genomic analyses of Silybum marianum and Carthamus tinctorius is in fact supported by reference 14. Reference 15 was inadvertently added alongside it and has now been removed, with all subsequent references renumbered accordingly in the revised manuscript.

14. Line 270: The text reports “146 upregulated and 46 downregulated” metabolites for leaf vs. root, but Figure 2A’s own data show “146, 76” for that comparison. Please verify and correct.

Response: We thank the reviewer for catching this error. We apologize for the writing mistake. The correct numbers are 146 upregulated and 76 downregulated metabolites for the leaf vs. root comparison, and this has now been corrected in the revised manuscript (line 286).

15. Line 299: The phrase “averaged ≥X%” is confusing, since 98.38%/95.35% are actually the minimum values across samples, not the averages. Please reword for clarity.

Response:We thank the reviewer for the careful observation. We agree that the original wording was imprecise. The sentence has now been revised to: "The Q20 and Q30 values across all samples were ≥98.38% and ≥95.35%, respectively, with GC content ranging from 42.79% to 44.1%." This correction has been made in the revised manuscript (lines 315-316).

16. Line 317: The leaf-root and flower-stem DEG comparisons are not reported in the text; please add them or note that they appear only in the supplementary table.

Response: We thank the reviewer for pointing out this omission. We have now supplemented the numbers for the leaf-root and flower-stem DEG comparisons in Section 3.4 of the revised manuscript, including both the upregulated and downregulated counts. The corresponding data have also been added to the text accordingly (lines 340-342).

17. Line 304: “31.6% and 23.12%” — inconsistent decimal precision between PC1 and PC2; please standardize.

Response:We thank the reviewer for noticing this inconsistency. We have now standardized the decimal precision by changing "31.6%" to "31.60%" in the revised manuscript to ensure consistency with "23.12%." (line 320)

18. Line 432: The number of metabolites associated with AP2/ERF is not reported (only “20 clusters”), unlike the other four TF families.

Response: We apologize for this writing oversight. The number of metabolites associated with the AP2/ERF transcription factor family was inadvertently omitted. We have now revised the text to "AP2/ERF (20 clusters with 10 metabolites)" to maintain consistency with the other four TF families. The correction has been made in the revised manuscript (line 452).

19. Line 469: Specify which sesquiterpenoids found in other plant systems overlap with those identified in this study.

Response: We thank the reviewer for this valuable suggestion. In response, we have revised the description in Section 4.1 of the revised manuscript to specify the sesquiterpenoids that overlap with those previously reported in other plant systems (lines 498-501). Specifically, we have listed compounds including eupalinolide B, zaluzanin C, isoalantolactone, dehydrocostus lactone, costunolide, spathulenol, and atractylenolide I/II, all of which have been previously characterized in other Asteraceae species and show consistent profiles with our findings in Eupatorium lindleyanum.

20. Line 475: Report the log2FC values in parentheses for the differentially accumulated metabolites mentioned.

Response: We thank the reviewer for this valuable suggestion. In response, we have now added the Log2FC values in parentheses for all the differentially accumulated metabolites mentioned in the text (Agalleudesmanol D, Calacorene oxide, Curcumafuranol, 13-Nor-desacylcynaropicrin, and Oxyphyllenone A) in the revised manuscript (lines 504-511).

21. Lines 479–480: Specify which genes may regulate sesquiterpenoid biosynthesis, rather than referring to them generically.

Response: We thank the reviewer for this insightful comment. We agree that the original statement was indeed vague. The reason is that this paragraph focuses on the differential accumulation of metabolites, and we only briefly mention the possibility of gene regulation here, with a more detailed discussion of the specific regulatory genes presented in the subsequent section. To address this concern, we have revised the sentence to clearly indicate that this question is further investigated in the following integrative analysis. See the lines 513-515 of the revised manuscript for details.

22. Line 486: Report which of the sesquiterpenoids found in this study are known to be biosynthesized in response to biotic or abiotic elicitation.

Response: We thank the reviewer for this suggestion. We have revised lines 523-527 to specify several sesquiterpenoidsthat are known to accumulate in response to biotic or abiotic elicitation in Asteraceae species, with relevant citations added.

23. Line 783 (ref. 41): This reference concerns Coptidis Rhizoma, a plant known for berberine-type alkaloids — an unrelated genus and chemical class. Please replace it with a more relevant reference.

Response: We thank the reviewer for this suggestion. We have replaced the original reference 41 (on Coptidis Rhizoma and berberine-type alkaloids) with a more relevant citation that accurately corresponds to the chemical class discussed in our study. The citation in the text has been updated accordingly, and the reference list has been renumbered. This reference is now listed as reference 40 in the revised manuscript.

Major comments

1. Supplementary Table 1 contains gene-identity inconsistencies relative to Figure 8. (a) The primer table lists GGPS (geranylgeranyl diphosphate synthase, Cluster-28107.0) in the position where Figure 8 and the rest of the manuscript instead validate GAO (Cluster-66246.0), a gene that is part of the 22 candidate structural genes; GGPS itself appears nowhere else in the manuscript and is not obviously part of the same pathway (it is typically associated with diterpenoid/carotenoid, not sesquiterpenoid, biosynthesis). (b) ACAT is labeled Cluster-13166.0 in Figure 8 but Cluster-56578.0 in Supplementary Table 1. Please verify which unigene IDs and gene identities are correct and correct Supplementary Table 1 accordingly.

Response:We thank the reviewer for their careful and thorough review of our data. We have carefully re-examined the relevant materials and acknowledge that errors were introduced during the preparation of Supplementary Table 1 due to inadvertent copy-paste mistakes.

Specifically, we have made the following corrections:

(a) The entry previously listed as GGPS (geranylgeranyl diphosphate synthase, Cluster-28107.0) has been replaced with GAO (Cluster-66246.0), which is the correct gene validated in Figure 8 and included among the 20 candidate structural genes. As the reviewer correctly noted, GGPS is not part of the sesquiterpenoid biosynthesis pathway and does not appear elsewhere in the manuscript.

(b) The unigene ID for ACAT has been corrected from Cluster-56578.0 to Cluster-13166.0, aligning Supplementary Table 1 with Figure 8.

All corrections have been made in the revised Supplementary Table 1, which now consistently matches Figure 8 and the main text. We apologize for these oversights and are grateful to the reviewer for bringing them to our attention.

2. Section 2.4 states that DAMs — and, by extension, the 299 DSMs underlying the rest of the manuscript — were defined partly by VIP scores from an OPLS-DA model, but no validation of that model (R²Y, Q², or a permutation test) is reported anywhere. Given the supervised, overfitting-prone nature of OPLS-DA, especially with n = 3 replicates per group, this validation is necessary to confirm that the 299 DSMs reflect genuine biological signal rather than model overfitting. Please report the relevant OPLS-DA validation statistics (R²Y, Q², and/or permutation test results).

Response: We thank the reviewer for this important suggestion. In the revised manuscript, we have added the OPLS-DA validation statistics in Section 2.4 and Supplementary Table 3. For all six pairwise comparisons, the Q² values ranged from 0.992 to 0.995, R²X from 0.804 to 0.860, and R²Y was 1.0 for all models. These high Q² values and the minimal gap between R²Y and Q² confirm that the models are robust and not overfitted (lines 157-161).

3. The claim that qRT-PCR “showed a consistent trend” with RNA-Seq FPKM values (Section 3.8) is not supported by a quantitative statistic. Please report a correlation coefficient (e.g., Pearson’s r) between qPCR and FPKM values for each of the 12 validated genes individually.

Response: We thank the reviewer for this rigorous and constructive suggestion, which has helped us improve the robustness of our manuscript. In the revised manuscript, we have performed Pearson correlation analysis between the qRT-PCR expression levels and the corresponding RNA-Seq FPKM values for each of the 12 candidate genes across the four examined tissues (root, stem, leaf, and flower). The correlation coefficients (r) and associated p-values for all 12 genes have now been incorporated into Section 3.8 of the revised manuscript (lines 469-476).

4. A single, non-independently-validated reference gene (β-actin) was used for qPCR normalization across tissues as divergent as root and flower. Per MIQE guidelines, I recommend validating the stability of at least two reference genes (e.g., via geNorm or NormFinder) across this tissue range, or explicitly justifying the use of a single reference gene.

Response: We sincerely thank the reviewer for this valuable and professional suggestion. We completely agree that validating the stability of reference genes is essential for accurate qRT-PCR normalization, especially across distinct tissues such as root and flower. In our previous studies, we had already evaluated the expression stability of β-actin across different tissues of Eupatorium lindleyanum, and the results were published elsewhere. However, we regret that this information was inadvertently omitted from the present manuscript. In the revised version, we have now incorporated the stability parameters of β-actin in Section 2.8 (lines 234-239). Specifically, we assessed its stability using geNorm (M = 0.527), NormFinder (stability value = 0.082), and BestKeeper (SD = 0.453), all of which fall within the accepted thresholds, confirming that β-actin exhibits good expression stability and is suitable as a reference gene for qRT-PCR normalization in this study. We are grateful for the reviewer's careful reading and constructive comments, which have helped improve the rigor of our work.

5. Line 241: The claim of a “gradual increase” in sesquiterpenoid accumulation across tissues is not supported by a statistical test. Given that three biological replicates are available, I recommend a formal statistical comparison (e.g., ANOVA or Kruskal-Wallis with post-hoc tests) to support this claim.

Response: We thank the reviewer for this suggestion. In fact, we had already performed a one-way ANOVA on the data presented in Figure 1A, and the significant differences among tissues are indicated by different lowercase letters above the bars (P < 0.05). We apologize for not making this clear in the figure legend. We have now revised the legend to explicitly state: "Different letters above the bars indicate significant differences between tissues (P < 0.05)." We hope this clarification addresses the reviewer's concern.

6. Line 249: The statement that “inter-group comparisons revealed relatively weak correlations” (Figure 1C) accurately describes the root-vs-rest comparisons but not the stem-leaf, stem-flower, or leaf-flower comparisons, which are in fact very high. Please qualify this sentence so that it does not generalize to all four tissues a pattern that, in the data, applies only to root.

Response: We thank the reviewer for this valuable suggestion. We have revised the description in Section 3.1 to more accurately reflect the data, specifying that weak correlations were observed between root and the above-ground tissues, while correlations among stem, leaf, and flower were relatively strong (lines 261-265). The revised sentence now reads: "In contrast, inter-group comparisons revealed that root exhibited relatively weak correlations with stem, leaf, and flower, whereas correlations among the above-ground tissues (stem, leaf, and flower) were relatively strong. This suggests a clear metabolic distinction between root and above-ground tissues, while the metabolic profiles of stem, leaf, and flower are more similar to one another." We hope this revision adequately addresses the reviewer's concern.

7. Clusters 1, 3, and 6 from the DEG expression clustering (Figure 4D) — together ~42% of all DEGs (22,638 of 53,610 genes) — are never characterized in the text or figure legend. Please briefly describe the expression pattern each cluster represents, even if functional analysis is pursued only for the three clusters that correlate with the DSMs.

Response: We thank the reviewer for this suggestion. We have now added a brief description of the expression patterns of Clusters 1, 3, and 6 in Section 3.4 to provide a more comprehensive characterization of all six clusters (lines 347-352). Specifically, we have described that genes in Cluster 1 showed relatively high expression in stem, Cluster 3 exhibited a gradually increasing expression trend across root, stem, leaf, and flower, and Cluster 6 displayed relatively high expression in root. We believe this addition makes the description of the clustering results more complete and addresses the reviewer's concern.

8. The Discussion should be restructured into labeled subsections to match the scale and diversity of the datasets generated (metabolomics, transcriptomics, WGCNA, transcription factor correlations, and qRT-PCR validation), ensuring that each data stream receives proportionate interpretive treatment and that the study’s limitations are addressed explicitly. I suggest:

4.1 Tissue-specific accumulation of sesquiterpenoids and its transcriptional basis — consolidating the discussion of the 321 metabolites, 299 DSMs, and their associated DEG clusters.

4.2 Integration of transcriptomic and metabolomic data: structural genes underlying the biosynthetic pathway — consolidating the discussion of the 22 candidate structural genes and their comparison with other Asteraceae. This subsection should specifically address: (a) why none of the seven WGCNA modules significantly correlated with the selected DSMs shows a root-predominant expression pattern, despite root being one of the three principal tissue-specific patterns established earlier in the manuscript (this negative result is currently not discussed anywhere); and (b) which branches of the proposed pathway (Figure 6) are supported by the study’s own metabolomic evidence versus inferred by homology to pathways characterized in taxonomically distant species (e.g., the solavetivone branch from Hyoscyamus muticus and the gossypol branch from Gossypium; see comment 7).

4.3 Transcription factors as candidate regulators of sesquiterpenoid biosynthesis — consolidating the discussion of the five TF families and their correlation with the ten selected metabolites.

4.4 Limitations — a new subsection addressing, at minimum: (i) the modest sample size (n = 3 per tissue) and the absence of OPLS-DA validation (see comment 25); (ii) reliance on a single, non-validated qPCR reference gene (see comment 27); and (iii) the correlative, non-causal nature of all candidate gene and transcription factor assignments, pending functional validation (e.g., gene silencing/overexpression).

Response: We thank the reviewer for this constructive suggestion. In accordance with your advice, we have restructured the Discussion section into four labeled subsections to better reflect the scale and diversity of the datasets generated in this study.

Specifically, Section 4.1 has been reorganized as "Tissue-specific accumulation of sesquiterpenoids and its transcriptional basis," which consolidates the discussion of the 321 identified metabolites, 299 DSMs, and their associated DEG clusters. Section 4.2 is now titled "Integration of transcriptomic and metabolomic data: structural genes underlying the biosynthetic pathway." In this subsection, we have specifically addressed: (a) the absence of root-predominant expression patterns among the seven WGCNA modules significantly correlated with the selected DSMs, and (b) which branches of the proposed pathway (Figure 6) are supported by our own metabolomic evidence versus those inferred by homology from pathways characterized in taxonomically distant species. Section 4.3 has been revised as "Transcription factors as candidate regulators of sesquiterpenoid biosynthesis," covering the five TF families and their correlations with the ten selected metabolites. Finally, Section 4.4 has been added to discuss the limitations of the study.

Regarding the limitations, we have focused our discussion on the correlative and non-causal nature of our candidate gene and transcription factor assignments, which await functional validation (e.g., via gene silencing or overexpression). We would like to note that the other two issues raised by the reviewer—OPLS-DA validation and qPCR reference gene stability—have already been addressed in Section 2.4 and Section 2.8, respectively; therefore, we have not reiterated them as limitations in Section 4.4.

We believe these revisions have made the Discussion more structured and comprehensive, and we are grateful for the reviewer's valuable guidance.

Attachment

Submitted filename: Response to Reviewers.pdf

pone.0359566.s004.pdf^{(206.4KB, pdf)}

PLoS One. doi: 10.1371/journal.pone.0359566.r003

Decision Letter 1

Rongchun Han

15 Sep 2026

<p>Multi-omics integration reveals tissue-specific biosynthesis of sesquiterpenoids in the medicinal plant Eupatorium lindleyanum

PONE-D-26-17903R1

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice will be generated when your article is formally accepted. Please note, if your institution has a publishing partnership with PLOS and your article meets the relevant criteria, all or part of your publication costs will be covered. Please make sure your user information is up-to-date by logging into Editorial Manager at Editorial Manager® and clicking the ‘Update My Information' link at the top of the page. For questions related to billing, please contact billing support.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Rongchun Han

Academic Editor

PLOS One

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions??>

Reviewer #1: (No Response)

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously? -->?>

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available??>

The PLOS Data policy

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English??>

Reviewer #1: Yes

Reviewer #2: Yes

**********

Reviewer #1: The authors have carried out a meaningful work. All amendments are done and now it can be accepted in this form.

Reviewer #2: Overall, the authors have substantially and rigorously addressed the concerns raised in the initial review. The revised manuscript now includes appropriate statistical validation of the multivariate models, a properly validated qPCR reference gene, quantitative correlation statistics between transcriptomic and qPCR data, and a corrected, taxonomically consistent biosynthetic pathway. The Discussion has also been reorganized to better reflect the scope and depth of the datasets generated. I consider the manuscript methodologically sound and suitable for publication in this journal

**********

what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy

Reviewer #1: No

Reviewer #2: Yes: Olmedo Jesus Cuaspud Caliz

**********

PLoS One. doi: 10.1371/journal.pone.0359566.r004

Acceptance letter

Rongchun Han

PONE-D-26-17903R1

PLOS One

Dear Dr. Wang,

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS One. Congratulations! Your manuscript is now being handed over to our production team.

At this stage, our production department will prepare your paper for publication. This includes ensuring the following:

* All references, tables, and figures are properly cited

* All relevant supporting information is included in the manuscript submission,

* There are no issues that prevent the paper from being properly typeset

You will receive further instructions from the production team, including instructions on how to review your proof when it is ready. Please keep in mind that we are working through a large volume of accepted articles, so please give us a few days to review your paper and let you know the next and final steps.

Lastly, if your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

You will receive an invoice from PLOS for your publication fee after your manuscript has reached the completed accept phase. If you receive an email requesting payment before acceptance or for any other service, this may be a phishing scheme. Learn how to identify phishing emails and protect your accounts at https://explore.plos.org/phishing.

If we can help with anything else, please email us at customercare@plos.org.

Thank you for submitting your work to PLOS One and supporting open access.

Kind regards,

PLOS One Editorial Office Staff

on behalf of

Prof Dr Rongchun Han

Academic Editor

PLOS One

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

S1 File. Supplementary Figure 1: photographs of different tissues of Eupatorium lindleyanum, including root, stem, leaf, and flower.

(ZIP)

pone.0359566.s001.zip^{(40.1MB, zip)}

Attachment

Submitted filename: Comments for article.docx

pone.0359566.s002.docx^{(16.2KB, docx)}

Attachment

Submitted filename: Response to Reviewers.pdf

pone.0359566.s004.pdf^{(206.4KB, pdf)}

Data Availability Statement

[pone.0359566.ref001] 1.Antonio CNS, Selene M de M, Elnatan B de S, Raquel O dos SF. The genus Eupatorium L. (Asteraceae): A review of their antimicrobial activity. J Med Plants Res. 2017;11(3):43–57. doi: 10.5897/jmpr2016.6313 [DOI] [Google Scholar]

[pone.0359566.ref002] 2.Wang Y, Wu J, Zhou M, Yang X, Niu Y, Guo K. Integrative metabolomic and transcriptomic analyses reveal flavonoid biosynthesis pathway in Eupatorium lindleyanum. Sci Rep. 2025;15(1):43151. doi: 10.1038/s41598-025-27287-0 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref003] 3.Yang J, Wang Y, Zhuo L, Lu G, Zhang M, Huang J, et al. Eupatorium lindleyanum DC Ameliorates Carbon Tetrachloride-Induced Hepatic Inflammation and Fibrotic Response in Mice. Pharmaceuticals (Basel). 2025;18(8):1228. doi: 10.3390/ph18081228 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref004] 4.Lu J, Zheng C, Xue S, Gao Y, Chen G, Shan C, et al. Comprehensive Comparison of Three Different Medicinal Parts of Eupatorium lindleyanum DC. Using the RRLC-Q-TOF-MS-Based Metabolic Profile and In Vitro Anti-Inflammatory Activity. Molecules. 2024;29(15):3551. doi: 10.3390/molecules29153551 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref005] 5.Geng H. Chemical constituents and their bioactivities of plants from the genus Eupatorium (2015-present). Biology (Basel). 2024;13(5):288. doi: 10.3390/biology13050288 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref006] 6.Huang L, Xu D-Q, Chen Y-Y, Fu R-J, Yue S-J, Yin J-F, et al. Qualitative and quantitative analysis of chemical components in Eupatorium lindleyanum DC. by ultra-performance liquid chromatography-mass spectrometry integrated with anti-inflammatory activity research. J Sep Sci. 2021;44(17):3174–87. doi: 10.1002/jssc.202100208 [DOI] [PubMed] [Google Scholar]

[pone.0359566.ref007] 7.Chu C, Ren H, Xu N, Xia L, Chen D, Zhang J. Eupatorium lindleyanum DC. sesquiterpenes fraction attenuates lipopolysaccharide-induced acute lung injury in mice. J Ethnopharmacol. 2016;185:263–71. doi: 10.1016/j.jep.2016.03.022 [DOI] [PubMed] [Google Scholar]

[pone.0359566.ref008] 8.Jiang L, Zhang L, Zhang X. Eupalinilide B as a novel anti-cancer agent that inhibits proliferation and epithelial-mesenchymal transition in laryngeal cancer cells. J Int Med Res. 2022;50(1):3000605211067921. doi: 10.1177/03000605211067921 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref009] 9.Wang X, Ma S, Lai F, Wang Y, Lou C. Traditional Applications, Phytochemistry, and Pharmacological Activities of Eupatorium lindleyanum DC.: A Comprehensive Review. Frontiers in Pharmacology. 2020:11. doi: 10.3389/FPHAR.2020.577124 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref010] 10.Li L, Luo W, Wang Q, Wang D, Li Z, Tang R, et al. Multi-omic analysis highlights new candidate genes involved in triterpenoid and sesquiterpenoid biosynthesis in Codonopsis pilosula (Franch.) Nannf. Industrial Crops and Products. 2025;226:120685. doi: 10.1016/j.indcrop.2025.120685 [DOI] [Google Scholar]

[pone.0359566.ref011] 11.Zhang Z, Tian Y, Qiao X, Li H, Ouyang L, Li X, et al. Integrated analysis of terpenoid profiles and full-length transcriptome reveals the central pathways of sesquiterpene biosynthesis in Atractylodes chinensis (DC.) Koidz. Int J Mol Sci. 2025;26(3):1074. doi: 10.3390/ijms26031074 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref012] 12.Li C, Li Y, Wang J, Lu F, Zheng L, Yang L, et al. An independent biosynthetic route to frame a xanthanolide-type sesquiterpene lactone in Asteraceae. Plant J. 2025;121(2):e17199. doi: 10.1111/tpj.17199 [DOI] [PubMed] [Google Scholar]

[pone.0359566.ref013] 13.Wu Y, Chen X, Hao F, Liu Y, Luo W, Zhu Y, et al. Biosynthesis of bridged tricyclic sesquiterpenes in Inula lineariifolia. Plant J. 2024;120(2):658–73. doi: 10.1111/tpj.17008 [DOI] [PubMed] [Google Scholar]

[pone.0359566.ref014] 14.Yang X, Meng F, Cheng Q, Feng P, Song X, Chen W. Evolution of terpene synthases in the sesquiterpene biosynthesis pathway and analysis of their transcriptional regulatory network in Asteraceae. Hortic Res. 2025;12(12):uhaf229. doi: 10.1093/hr/uhaf229 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref015] 15.Bogdanović M, Cankar K, Todorović S, Dragicević M, Simonović A, van Houwelingen A, et al. Tissue specific expression and genomic organization of bitter sesquiterpene lactone biosynthesis in Cichorium intybus L. (Asteraceae). Industrial Crops and Products. 2019;129:253–60. doi: 10.1016/j.indcrop.2018.12.011 [DOI] [Google Scholar]

[pone.0359566.ref016] 16.Zhu F, Wen W, Cheng Y, Alseekh S, Fernie AR. Integrating multiomics data accelerates elucidation of plant primary and secondary metabolic pathways. aBIOTECH. 2023;4(1):47–56. doi: 10.1007/s42994-022-00091-4 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref017] 17.Gao H, Li X, Wang C, Li Y, Liu T, Chang N, et al. Integrated multiomics profiling elucidates the spatiotemporal metabolic dynamics and regulatory networks of the bioactive components of Trichosanthes kirilowii. Front Plant Sci. 2026;17:1735703. doi: 10.3389/fpls.2026.1735703 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref018] 18.Chen D, Chen X, Zheng X, Zhu J, Xue T. Combined metabolomic and transcriptomic analysis reveals the key genes for triterpenoid biosynthesis in Cyclocarya paliurus. BMC Genomics. 2024;25(1):1197. doi: 10.1186/s12864-024-11125-0 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref019] 19.Hou J, Zhang Q, Wang R, Yao J, Zhong Y, Zhong Y, et al. Integration of transcriptomics and metabolomics provides insights into secondary cell wall deposition and terpenoid biosynthesis during stem development in Camphora officinarum. BMC Genomics. 2025;27(1):6. doi: 10.1186/s12864-025-12266-6 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref020] 20.Liu M, Li Y, Chen H, He C, Sun L, Zhang X, et al. Integrated omics profiles for exploring the potential mechanism underlying aroma formation in the terpenoid-rich aromatic plant Opisthopappus taihangensis and the bioactivity of its leaf essential oil. Agriculture Communications. 2024;2(4):100061. doi: 10.1016/j.agrcom.2024.100061 [DOI] [Google Scholar]

[pone.0359566.ref021] 21.Wang Z, Xu X, Zhang P, Huang Y, Zhao C, El-Kassaby YA, et al. Multi-omics integration reveals organ-specific biosynthesis of flavonoids and terpenoids in the medicinal plant Bidens alba. Front Plant Sci. 2025;16:1675432. doi: 10.3389/fpls.2025.1675432 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref022] 22.Hazrati S, Mousavi Z, Nicola S. Harvest time optimization for medicinal and aromatic plant secondary metabolites. Plant Physiol Biochem. 2024;212:108735. doi: 10.1016/j.plaphy.2024.108735 [DOI] [PubMed] [Google Scholar]

[pone.0359566.ref023] 23.Li J, Liao Q, Zhou H, Hu R, Li Y, Hu Z, et al. Multi-omics analyses reveal regulatory networks underpinning metabolite biosynthesis in Nicotiana tabacum. Nat Commun. 2025;16(1):10339. doi: 10.1038/s41467-025-65299-6 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref024] 24.Zhong Y, Xie Y, Zhang D, Li G, Yu J. Integrated metabolomic and transcriptomic analysis of metabolic diversity and biosynthesis of glucosinolates and flavonoids in various cultivars of radish microgreens. Food Bioscience. 2024;59:104055. doi: 10.1016/j.fbio.2024.104055 [DOI] [Google Scholar]

[pone.0359566.ref025] 25.Zhang X, Chen X, Wang Y, Liu M, Wang C, Zhou T. Integrative transcriptomic and metabolomic analysis of Drynaria roosii reveals genes involved in the biosynthesis of medicinal compounds. Sci Rep. 2026;16(1):9047. doi: 10.1038/s41598-026-39037-x [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref026] 26.Li Y, Sun Z, Lu J, Jin Z, Li J. Integrated transcriptomics and metabolomics analysis provide insight into anthocyanin biosynthesis for sepal color formation in Heptacodium miconioides. Front Plant Sci. 2023;14:1044581. doi: 10.3389/fpls.2023.1044581 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref027] 27.Jiang Y, Wang J, Tu X, Zhong M, Wan B. Integrative transcriptomic and metabolomic analysis reveals the molecular basis of leaf variegation in Cymbidium ensifolium. Front Plant Sci. 2026;17:1712811. doi: 10.3389/fpls.2026.1712811 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref028] 28.Lu X, Wei Y, Sun J, Xiao B, Zhang X, Li W, et al. A comparative study of three nucleic acid integrity assay systems. Biopreserv Biobank. 2023;21(6):624–30. doi: 10.1089/bio.2022.0171 [DOI] [PubMed] [Google Scholar]

[pone.0359566.ref029] 29.Chen S, Zhou Y, Chen Y, Gu J. fastp: an ultra-fast all-in-one FASTQ preprocessor. Bioinformatics. 2018;34(17):i884–90. doi: 10.1093/bioinformatics/bty560 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref030] 30.Xu W, Ayyaz A, Hannan F, Rehman KM, Qin T, Song W, et al. Integrative metabolomic and transcriptomic analyses reveal mechanisms of hexavalent chromium toxicity in contrasting rapeseed cultivars. Agronomy,2025,15(12):2892–2892. doi: 10.3390/AGRONOMY15122892 [DOI] [Google Scholar]

[pone.0359566.ref031] 31.Zheng Y, Jiao C, Sun H, Rosli H, Pombo M, Zhang P, et al. iTAK: A program for genome-wide prediction and classification of plant transcription factors, transcriptional regulators, and protein kinases. Molecular Plant,2016,9(12):1667–70. doi: 10.1016/j.molp.2016.09.014 [DOI] [PubMed] [Google Scholar]

[pone.0359566.ref032] 32.Langfelder P, Horvath S. WGCNA: An R package for weighted correlation network analysis. BMC Bioinformatics. 2008;9:559. doi: 10.1186/1471-2105-9-559 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref033] 33.Yang G, Yang L, Xu F. Isoalantolactone: A review on its pharmacological effects. Frontiers in Pharmacology. 2024;151:453205–1453205. doi: 10.3389/FPHAR.2024.1453205 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref034] 34.Zhang J, Zhang H-L, Xu X-R, Feng Y-L, Zhu Q-M, Morisseau C, et al. Targeting PBK with small-molecule 1-O-acetyl-4R,6S-britannilactone for the treatment of neuroinflammation. Proc Natl Acad Sci U S A. 2025;122(29):e2502593122. doi: 10.1073/pnas.2502593122 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref035] 35.Gu S-L, Liu X-S, Xu Z-S, Li L-L, Wu X-J, Li F-L, et al. Eupalinolide B alleviates rheumatoid arthritis through the promotion of apoptosis and autophagy via regulating the AMPK/mTOR/ULK-1 signaling axis. Int Immunopharmacol. 2025;148:114179. doi: 10.1016/j.intimp.2025.114179 [DOI] [PubMed] [Google Scholar]

[pone.0359566.ref036] 36.Jin M, Lee HJ, Ryu JH, Chung KS. Inhibition of LPS-induced NO production and NF-kappaB activation by a sesquiterpene from Saussurea lappa. Arch Pharm Res. 2000;23(1):54–8. doi: 10.1007/BF02976467 [DOI] [PubMed] [Google Scholar]

[pone.0359566.ref037] 37.Jeong S-H, Koo S-J, Ha J-H, Ryu S-Y, Park H-J, Lee K-T. Induction of apoptosis by yomogin in human promyelocytic leukemic HL-60 cells. Biol Pharm Bull. 2004;27(7):1106–11. doi: 10.1248/bpb.27.1106 [DOI] [PubMed] [Google Scholar]

[pone.0359566.ref038] 38.Liang S, Chu Chen, Li R, Jiang G, Du L. Costunolide and dehydrocostus lactone alleviate ulcerative colitis via regulating TLR4, NF-κB and PI3K expression. Sci Rep. 2024;14(1):29777. doi: 10.1038/s41598-024-77021-5 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref039] 39.Trentin A, Santos A, Guedes A, Pizzolatti M, Yunes R, Calixto J. Antinociception caused by the extract of Hedyosmum brasiliense and its active principle, the sesquiterpene lactone 13-hydroxy-8,9-dehydroshizukanolide. Planta Med. 1999;65(6):517–21. doi: 10.1055/s-1999-14007 [DOI] [PubMed] [Google Scholar]

[pone.0359566.ref040] 40.Lin X, Peng Z, Su C. Potential anti-cancer activities and mechanisms of costunolide and dehydrocostuslactone. Int J Mol Sci. 2015;16(5):10888–906. doi: 10.3390/ijms160510888 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref041] 41.Yang S, Gao Z, Wu Y, Hu G, Yue J. ChemInform abstract: Chlorahololides C—F: A new class of potent and selective potassium channel blockers from Chloranthus holostegius. ChemInform. 2008;39(37). doi: 10.1002/chin.200837167 [DOI] [Google Scholar]

[pone.0359566.ref042] 42.Geng X, Tang R, Zhang A, Du Z, Yang L, Xu Y, et al. Mining, expression, and phylogenetic analysis of volatile terpenoid biosynthesis-related genes in different tissues of ten Elsholtzia species based on transcriptomic analysis. Phytochemistry. 2022;203:113419. doi: 10.1016/j.phytochem.2022.113419 [DOI] [PubMed] [Google Scholar]

[pone.0359566.ref043] 43.Jiang Y, Liu Y, Song H, He J. Integrated transcriptomics and metabolomics uncover the molecular basis of flavonoid accumulation in the rhizomes of two Cymbidium tortisepalum var. longibracteatum cultivars. Scientia Horticulturae. 2022;293:110738. doi: 10.1016/j.scienta.2021.110738 [DOI] [Google Scholar]

[pone.0359566.ref044] 44.Liu PY, Liu D, Li WH, Zhao T, Sauriol F, Gu YC, et al. Chemical constituents of plants from the genus Eupatorium (1904–2014). ChemInform. 2015;46(49). doi: 10.1002/chin.201549256 [DOI] [PubMed] [Google Scholar]

[pone.0359566.ref045] 45.Chen F, Wei Y-X, Zhang J-M, Sang X-M, Dai C-C. Transcriptomics analysis investigates sesquiterpenoids accumulation pattern in different tissues of Atractylodes lancea (Thunb.) DC. plantlet. Plant Cell Tiss Organ Cult. 2017;130(1):73–90. doi: 10.1007/s11240-017-1205-8 [DOI] [Google Scholar]

[pone.0359566.ref046] 46.Bai Y, Liu X, Baldwin IT. Using synthetic biology to understand the function of plant specialized metabolites. Annu Rev Plant Biol. 2024;75(1):629–53. doi: 10.1146/annurev-arplant-060223-013842 [DOI] [PubMed] [Google Scholar]

[pone.0359566.ref047] 47.Kour S, Singh S, Wani TA, Kaloo ZA. Chemical characterization and the intrusion through elicitation and Agrobacterium rhizogenes mediated hairy root transformation in Saussurea costus C.B. Clarke. Physiol Mol Biol Plants. 2023;29(1):87–91. doi: 10.1007/s12298-022-01270-9 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref048] 48.Kumar A, Agnihotri VK. NMR based profiling of sesquiterpene lactones in Saussurea lappa roots collected from different location of Western Himalaya. Nat Prod Res. 2022;36(2):621–4. doi: 10.1080/14786419.2020.1789635 [DOI] [PubMed] [Google Scholar]

[pone.0359566.ref049] 49.Liao H, Quan H, Huang B, Ji H, Zhang T, Chen J, et al. Integrated transcriptomic and metabolomic analysis reveals the molecular basis of tissue-specific accumulation of bioactive steroidal alkaloids in Fritillaria unibracteata. Phytochemistry. 2023;214:113831. doi: 10.1016/j.phytochem.2023.113831 [DOI] [PubMed] [Google Scholar]

[pone.0359566.ref050] 50.Fu X, Zhang Y, Yu M, Zheng H, Pan Y, Liu P, et al. Specialized structures and developmental mechanisms of secondary metabolite-synthesizing organs and tissues in medicinal plants. J Integr Plant Biol. 2026;68(4):926–41. doi: 10.1111/jipb.70133 [DOI] [PubMed] [Google Scholar]

[pone.0359566.ref051] 51.Schneider GF, Carlson CA, Jos EM, Beckman NG. Compartmentalization of specialized metabolites across vegetative and reproductive tissues in two sympatric Psychotria species. Am J Bot. 2023;110(7):e16191. doi: 10.1002/ajb2.16191 [DOI] [PubMed] [Google Scholar]

[pone.0359566.ref052] 52.He S, Wang J, Yang N, Li H, Li K, Li L, et al. Regulatory mechanisms underlying stress-induced accumulation of plant secondary metabolites. Journal of Applied Research on Medicinal and Aromatic Plants. 2026;50:100689. doi: 10.1016/j.jarmap.2025.100689 [DOI] [Google Scholar]

[pone.0359566.ref053] 53.Zhang Z-A, Xue X-Y, Wu J-Y, Zhao X, Zhang H, Sun K, et al. CYP71P2 regulated by MYB44 modulates oxygenated sesquiterpenoids biosynthesis in Atractylodes lancea. Plant Physiol Biochem. 2025;229(Pt D):110700. doi: 10.1016/j.plaphy.2025.110700 [DOI] [PubMed] [Google Scholar]

[pone.0359566.ref054] 54.Pütter KM, van Deenen N, Unland K, Prüfer D, Schulze Gronover C. Isoprenoid biosynthesis in dandelion latex is enhanced by the overexpression of three key enzymes involved in the mevalonate pathway. BMC Plant Biol. 2017;17(1):88. doi: 10.1186/s12870-017-1036-0 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref055] 55.You MK, Lee YJ, Kim JK, Baek SA, Jeon YA, Lim SH, et al. The organ-specific differential roles of rice DXS and DXR, the first two enzymes of the MEP pathway, in carotenoid metabolism in Oryza sativa leaves and seeds. BMC Plant Biol. 2020;20(1):167. doi: 10.1186/s12870-020-02357-9 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref056] 56.Banerjee A, Wu Y, Banerjee R, Li Y, Yan H, Sharkey TD. Feedback inhibition of deoxy-D-xylulose-5-phosphate synthase regulates the methylerythritol 4-phosphate pathway. J Biol Chem. 2013;288(23):16926–36. doi: 10.1074/jbc.M113.464636 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref057] 57.Liu X-M, Tao T-T, Meng X-X, Zhang W-W, Chang J, Xu F. Cloning and Expression Analysis of a Farnesyl Diphosphate Synthase (FPPS) Gene from Chamaemelum nobile. Not Bot Horti Agrobo. 2017;45(2):358–64. doi: 10.15835/nbha45210858 [DOI] [Google Scholar]

[pone.0359566.ref058] 58.Zhang D, Tang X, Chen L, Qiu X, Song C, Wang H, et al. Functional characterization and transcriptional activity analysis of Dryopteris fragrans farnesyl diphosphate synthase genes. Front Plant Sci. 2023;14:1105240. doi: 10.3389/fpls.2023.1105240 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref059] 59.Cheng W, Zhi Y, Chen F, Xiao X, Lu H, Li R, et al. Characterization and functional reconstruction of a highly productive germacrene A synthase from Liriodendron chinense. Plant Biotechnol J. 2025;23(6):1927–37. doi: 10.1111/pbi.70023 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref060] 60.Aharoni A, Giri AP, Deuerlein S, Griepink F, de Kogel W-J, Verstappen FWA, et al. Terpenoid metabolism in wild-type and transgenic Arabidopsis plants. Plant Cell. 2003;15(12):2866–84. doi: 10.1105/tpc.016253 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref061] 61.Takahashi S, Yeo Y-S, Zhao Y, O’Maille PE, Greenhagen BT, Noel JP, et al. Functional characterization of premnaspirodiene oxygenase, a cytochrome P450 catalyzing regio- and stereo-specific hydroxylations of diverse sesquiterpene substrates. J Biol Chem. 2007;282(43):31744–54. doi: 10.1074/jbc.M703378200 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref062] 62.Bhandari J, Fitzpatrick AH, Crowell DN. Identification of a novel abscisic acid-regulated farnesol dehydrogenase from Arabidopsis. Plant Physiol. 2010;154(3):1116–27. doi: 10.1104/pp.110.157784 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref063] 63.Wu J, Hu J, Yu H, Lu J, Jiang L, Liu W, et al. Full-length transcriptome analysis of two chemotype and functional characterization of genes related to sesquiterpene biosynthesis in Atractylodes lancea. Int J Biol Macromol. 2023;225:1543–54. doi: 10.1016/j.ijbiomac.2022.11.210 [DOI] [PubMed] [Google Scholar]

[pone.0359566.ref064] 64.Cankar K, Bundock P, Sevenier R, Häkkinen ST, Hakkert JC, Beekwilder J, et al. Inactivation of the germacrene A synthase genes by CRISPR/Cas9 eliminates the biosynthesis of sesquiterpene lactones in Cichorium intybus L. Plant Biotechnol J. 2021;19(12):2442–53. doi: 10.1111/pbi.13670 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref065] 65.Steliopoulos P, Wüst M, Adam K-P, Mosandl A. Biosynthesis of the sesquiterpene germacrene D in Solidago canadensis: 13C and (2)H labeling studies. Phytochemistry. 2002;60(1):13–20. doi: 10.1016/s0031-9422(02)00068-7 [DOI] [PubMed] [Google Scholar]

[pone.0359566.ref066] 66.Singh S, Chhatwal H, Pandey A. Deciphering the complexity of terpenoid biosynthesis and its multi-level regulatory mechanism in plants. J Plant Growth Regul. 2024;43(10):3320–36. doi: 10.1007/s00344-024-11347-2 [DOI] [Google Scholar]

[pone.0359566.ref067] 67.Shen S-L, Yin X-R, Zhang B, Xie X-L, Jiang Q, Grierson D, et al. CitAP2.10 activation of the terpene synthase CsTPS1 is associated with the synthesis of (+)-valencene in “Newhall” orange. J Exp Bot. 2016;67(14):4105–15. doi: 10.1093/jxb/erw189 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref068] 68.Liang H, Liu W, Zhao Z, Li Y, Zha L. Genome-wide identification and expression analysis of the WRKY transcription factors related to sesquiterpenes biosynthesis in Atractylodes lancea. Front Genet. 2025;16:1551991. doi: 10.3389/fgene.2025.1551991 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pone.0359566.ref069] 69.Li X, Meng S, Zhou Y, Wang D, Bian Z, Hu L, et al. Genome-wide analysis of NAC gene family and function exploration of SaNAC30 in Santalum album L. Industrial Crops and Products. 2025;227:120827. doi: 10.1016/j.indcrop.2025.120827 [DOI] [Google Scholar]

[pone.0359566.ref070] 70.Zeng X, Zhang S, Zhang Z, Chen Y, Wang S, Tang Z, et al. Integration of transcriptomics and metabolomics reveals the mechanism of aroma formation during apricot fruit development. Scientia Horticulturae. 2025;350:114351. doi: 10.1016/j.scienta.2025.114351 [DOI] [Google Scholar]

PERMALINK

Multi-omics integration reveals tissue-specific biosynthesis of sesquiterpenoids in the medicinal plant Eupatorium lindleyanum

Yingzhe Wang

Pan Jiang

Jiaqiu Yuan

Jinghan Wu

Yue Zhang

Kun Guo

Roles

Abstract

Background

Results

Conclusions

1. Background

2. Materials and methods

2.1. Plant materials

2.2. Plant sample pretreatment and metabolite extraction procedures

2.3. UPLC‒MS/MS conditions for metabolite profiling

2.4. Qualitative and quantitative analysis of metabolites

2.5. RNA extraction, library construction, and sequencing

2.6. Data analysis of RNA-seq

2.7. Integrated analysis of transcriptomic and metabolomic data

2.8. Quantitative real-time PCR validation of candidate genes

2.9. Statistical analysis and data visualization

3. Result

3.1. Profiling of sesquiterpenoid metabolites in Eupatorium lindleyanum derived from four different tissues

Fig 1. Analysis of sesquiterpenoid metabolites in different tissues of Eupatorium lindleyanum.

3.2. Differentially accumulated sesquiterpenoid metabolites in four tissues of Eupatorium lindleyanum

Fig 2. Differential analysis of sesquiterpenoids across tissues in Eupatorium lindleyanum.

3.3. Transcriptome profiling analysis of Eupatorium lindleyanum in four distinct tissues

Fig 3. Principal component and correlation analyses of the Eupatorium lindleyanum transcriptome across four tissues.

3.4. Analysis of differentially expressed genes

Fig 4. Analysis of differentially expressed genes (DEGs) in Eupatorium lindleyanum across different tissues.

3.5. Identification of sesquiterpenoid biosynthesis-related genes through WGCNA-based co-expression network analysis

Fig 5. Weighted gene co-expression network analysis (WGCNA) of DEGs in Clusters 2, 4, and 5.

3.6. Construction of the sesquiterpenoid biosynthesis regulatory network in Eupatorium lindleyanum

Fig 6. Proposed pathway for sesquiterpenoid biosynthesis in Eupatorium lindleyanum.

3.7. Analysis of transcription factors associated with sesquiterpenoid biosynthesis

Fig 7. Correlation analysis between transcription factors and DSMs in Eupatorium lindleyanum (|R| > 0.8, P < 0.05); A: bHLH, B: MYB, C: AP2/ERF, D: NAC; E: WRKY.

3.8. qRT‒PCR validation of candidate genes in the sesquiterpenoid biosynthesis pathway

Fig 8. qRT‒PCR validation of 12 candidate genes.

4. Discussion

4.1. Tissue-specific accumulation of sesquiterpenoids and its transcriptional basis

4.2. Integration of transcriptomic and metabolomic data: structural genes underlying the biosynthetic pathway

4.3. Transcription factors as candidate regulators of sesquiterpenoid biosynthesis

4.4. Limitations

5. Conclusion

Supporting information

Abbreviations

Data Availability

Funding Statement

References

Decision Letter 0

Rongchun Han

Roles

Author response to Decision Letter 1

Decision Letter 1

Rongchun Han

Roles

Acceptance letter

Rongchun Han

Roles

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases