Abstract
Introduction:
Canavalia gladiata (C. gladiata) is an important medicinal and edible plant. However, systematic research on the distribution of metabolites in different tissues of C. gladiata and their potential transcriptional regulatory mechanisms remain poorly understood.
Methods:
We performed integrated metabolomic and transcriptomic analyses across five tissues (roots, stems, leaves, seeds, and fruit pericarps) of C. gladiata, combined with antioxidant capacity and bioactive component content assays, to dissect the regulatory networks of flavonoid and terpenoid biosynthesis.
Results:
Seeds exhibited the highest antioxidant activity and total phenolic content, whereas leaves accumulated the highest levels of total flavonoids and terpenic acids. A total of 4,405 DAMs and 25,597 DEGs were identified, revealing pronounced tissue-specific metabolic and transcriptional divergence. Flavonoid and terpenoid biosynthesis pathways were significantly enriched in comparisons between seeds and other tissues. Key structural genes, including 4CL, CHS, FLS, HMGR, and DXS, displayed strong tissue-specific expression patterns. Co-expression network analysis identified candidate regulatory modules, highlighting MYB, bHLH, and MYC2 transcription factors as central regulators of flavonoid and terpenoid metabolism in seeds and fruit pericarps.
Discussion:
This study provides the first comprehensive landscape of tissue-specific flavonoid and terpenoid metabolism in C. gladiata, offering a theoretical foundation and valuable genetic resources for the targeted exploitation of its bioactive components and molecular breeding.
1 Introduction
Canavalia gladiata (sword bean) is an underutilized legume with both edible and medicinal value, cultivated widely across Asia and other tropical regions ().The plant is rich in bioactive compounds, including flavonoids, terpenoids, and phenolic acids, which contribute to its diverse pharmacological properties, such as antioxidant, anti-inflammatory, and antitumor effects (; ; Rokkam et al., 2023; Qian et al., 2025). Notably, the total flavonoid content of C. gladiata is significantly higher than that of soybean, highlighting its strong potential as a high-value functional food source (; ). Pharmacological studies have further demonstrated that different tissues of C. gladiata exhibit distinct pharmacological activities and clinical application potentials (; Pal et al., 2022; Qian et al., 2025; ). For instance, a comparative metabolomic analysis identified 31 metabolites in the leaves and seeds, highlighting both tissues as significant repositories of antioxidant components (More et al., 2022). Extracts from sword bean pods have also been shown to promote osteoblast differentiation via activation of the BMP2/SMAD/RUNX2 signaling pathway, supporting its beneficial role in bone health (). Nevertheless, systematic research on the tissue-specific enrichment of functional metabolites, along with their underlying biosynthetic and regulatory mechanisms, remains severely limited in C. gladiata.
Flavonoids and terpenoids represent two predominant classes of plant secondary metabolites, playing indispensable roles in defense responses and environmental adaptation (; Wu et al., 2023; ).Flavonoids are synthesized via the phenylpropanoid pathway, while terpenoids originate from the mevalonate (MVA) or methylerythritol phosphate (MEP) pathways (Movahedi et al., 2022; Zhuang et al., 2023).Despite their distinct biochemical origins, biosynthesis of both metabolite classes is subject to sophisticated, tightly coordinated transcriptional regulation. R2R3-MYB transcription factors can independently activate early flavonoid biosynthetic genes, such as CHS, CHI, F3H and FLS, thereby controlling the initial flux through the phenylpropanoid and flavonoid pathways (). In contrast, coordinated activation of late flavonoid biosynthetic genes, such as DFR and ANS, typically requires the formation of MYB–bHLH–WD40 (MBW) transcriptional complexes (). Similarly, terpenoid accumulation is modulated by a diverse set of TFs including MYC2, WRKY, and AP2/ERF family members, which directly target and regulate key enzyme-encoding genes in the MVA and MEP pathways (Singh et al., 2024). Importantly, accumulating evidence from medicinal plant studies indicates that the regulatory mechanisms governing flavonoid and terpenoid biosynthesis exhibit pronounced tissue specificity (Zhang et al., 2022; ; ; Wang et al., 2025; Xu et al., 2025). Distinct transcription factor repertoires and co-expression modules have been identified in roots, leaves, stems, and flowers, resulting in organ-specific activation of flavonoid and terpenoid biosynthetic pathways (Zhang et al., 2022; ; Wang et al., 2025). Consequently, the tissue-dependent assembly of transcriptional complexes and spatial expression of structural genes collectively shape differential metabolite accumulation patterns.
While the biosynthetic and regulatory frameworks of flavonoids and terpenoids are well-characterized in many model and medicinal plants, the corresponding regulatory architecture in C. gladiata remains largely unexplored. Integrated transcriptomic and metabolomic analyses offer a systematic approach to identifying key structural genes and regulators, thereby elucidating complex metabolic networks (; Shen et al., 2023; ). In this study, we conducted the first comprehensive investigation of the metabolomic and transcriptomic profiles across five distinct tissues (roots, stems, leaves, seeds, and fruit pericarps) of C. gladiata. Our objectives were to: (1) characterize tissue-specific metabolite accumulation patterns, focusing on flavonoids and terpenoids; (2) identify key differentially expressed genes (DEGs) and transcription factors (TFs) associated with these pathways; and (3) construct integrated regulatory networks linking TFs to structural genes and metabolites. This work provides a theoretical foundation for understanding the molecular basis of metabolite synthesis in C. gladiata and offers valuable genetic resources for its functional optimization and molecular breeding.
2 Materials and methods
2.1 Plant materials and tissue collection
Plant materials of C. gladiata were collected from Shaoshan, Hunan, China (112°31′17.9″E, 27°55′6.0″N; elevation=126 m) during the fruit ripening stage in October 2024. The collected samples were categorized into five groups based on tissue type: roots (R), stems (St), leaves (L), seeds (Se), and fruit pericarps (FP). The flowering period of C. gladiata typically spans two months. Flowers were tagged at 0 days post-anthesis (0 DPA), and seeds and fruit pericarps were harvested at 25 DPA. All samples, selected for being free of decay, mold, and mechanical damage, were immediately snap-frozen in liquid nitrogen and stored at −80 °C. Four biological replicates were used for metabolomic extraction, while three biological replicates were employed for transcriptome analysis and the determination of physicochemical indicators.
2.2 Determination of antioxidant capacity and physicochemical indicators
The antioxidant capacity of C. gladiata extracts was evaluated by three commonly used methods, and each assay was performed using the corresponding commercial kit following the manufacturer’s protocols. Specifically, the 2,2-diphenyl-1-picrylhydrazyl (DPPH) radical scavenging assay was measured using the DPPH assay kit (Macklin, T931095); the 2,2’-azino-bis (3-ethylbenzothiazoline-6-sulfonic acid) (ABTS) radical cation decolorization assay was conducted with the ABTS assay kit (Macklin, T931096); and the ferric reducing antioxidant power (FRAP) assay was determined using the FRAP assay kit (Macklin, T931094).
Total phenolic content (TPC) (Solarbio, BC1330) and total flavonoid content (TFC) (Macklin, T930799) were determined according to the manufacturers’ instructions. The total terpenic acid content was measured in accordance with the Agricultural Industry Standard of the People’s Republic of China (NY/T 3676-2020) (Ministry of Agriculture and Rural Affairs of the People’s Republic of China, 2020).
2.3 Metabolome analysis
The samples were gradually thawed at 4 °C, and an appropriate aliquot (50–100 mg) of each sample was precisely weighed and transferred into a centrifuge tube. Subsequently, 1 mL of a pre-cooled extraction solvent (a mixture of water, acetonitrile, and isopropyl alcohol in a ratio of 1:1:1, v/v/v) was added. The mixture was homogenized for 60 seconds and subjected to low-temperature ultrasonic extraction for 30 minutes. Following extraction, the samples were centrifuged at 12,000 rpm for 10 minutes at 4 °C. To facilitate protein precipitation, the samples were allowed to stand at -20 °C for 1 hour and then centrifuged again under the same conditions. The resulting supernatant was vacuum-dried, reconstituted in 0.2 mL of 30% acetonitrile solution, and homogenized. Finally, the solution was centrifuged at 14,000 rpm for 15 minutes at 4 °C, and the supernatant was collected for subsequent computer-based analysis.
A Vanquish UPLC system (Thermo Fisher Scientific, USA) was used in conjunction a Q Exactive HFX mass spectrometer (Thermo Fisher Scientific, USA).The samples were injected onto a Waters HSS T3 column (100 × 2.1 mm, 1.8μm) maintained at 40°C, with a flow rate of 0.3 mL/min and an injection volume of 2μL.The mobile phase consisted of Milli-Q water containing 0.1% formic acid (eluent A) and acetonitrile containing 0.1% formic acid (eluent B). The gradient elution program was as follows: 0 min phase A/phase B (100:0, v/v), 1 min phase A/phase B (100:0, v/v), 4 min phase A/phase B (40:60, v/v), 6.5 min phase A/phase B (5:95, v/v), 6.6 min phase A/phase B (100:0, v/v), 8.0 min phase A/phase B (100:0, v/v). A positive/negative polarity Q ExactiveTM HF-X mass spectrometers were set as follows: spray voltage +3 kV/-2.8 kV; capillary temperature of 350 °C, sheath gas flow rate of 40 arb, and aux gas flow rate of 10 arb was used. The temperature of the ion transport tube was set at 320°C. The scanning range for the primary mass spectrometry was set from 70 to 1050 Da, with a primary resolution of 70,000 and a secondary resolution of 17,500.
Raw MS data were acquired using Xcalibur 4.1 software (Thermo Fisher Scientific, USA) and processed with Progenesis QI (Waters Corporation, USA) for feature extraction and alignment. Principal Component Analysis (PCA) was performed using SIMCA-P 14.1 (Umetrics, Sweden). Pearson’s Correlation Coefficients (PCC) between samples were calculated with the “cor” function in the R (version 3.5.1) and the results were visualized as heatmaps. Differentially accumulated metabolites (DAMs) were identified based on a fold change (FC) ≥ 2 or ≤ 0.5 and a variable importance in projection (VIP) ≥ 1. Metabolites were annotated using the KEGG Compound database (http://www.kegg.jp/kegg/compound/), followed by mapping of the annotated metabolites to the KEGG Pathway database (http://www.kegg.jp/kegg/pathway.html).
2.4 RNA extraction and transcriptome analysis
Total RNA was extracted from the plant materials using pre-frozen TRIzol reagent (Thermo Scientific, 15596026CN) in accordance with the manufacturer’s instructions. The quantity and quality of total RNA was extracted and assessed using NanoDrop 2000 (Thermo Fisher Scientific, USA). The mRNA is then enriched using Oligo dT magnetic beads and fragmented into small pieces of approximately 300 bp. Double-stranded cDNA is synthesized through reverse transcription and ligated with Y-shaped adapters. Finally, after fragment selection, PCR amplification, and library purification, high-throughput sequencing of mRNA is performed using the DNBSEQ-T7 sequencer (MGI Tech, China). The clean reads were then mapped to the reference genome (GCA_037954105.1) using the STAR software (Dobin and Gingeras 2015). Differential expression analysis of gene expression was performed using DESeq2 () with a threshold of |log2FoldChange|> 1 for DEGs and a significance P-value < 0.05. Genes were aligned with protein sequences in the NR database and SwissProt database via BLAST (evalue ≤ 1e-5), and the alignment results were used as annotations for the corresponding genes. GO and KEGG enrichment analyses of DEGs were performed using the clusterProfiler R package (Xu et al., 2024). Transcription factors (TFs) were predicted from the assembled transcriptome of C. gladiata via the Plant Transcription Factor Database (PlantTFDB, http://planttfdb.gao-lab.org/).
2.5 Integrated analysis of transcriptome and metabolome
The screened DAMs and DEGs were mapped to the KEGG pathway database (www.kegg.jp/kegg/kegg.html), thereby identifying flavonoid- and terpenoid-related KEGG pathways at the transcriptome and metabolome levels. To construct a metabolite-gene expression regulatory network, we first screened genes significantly correlated with each DAMs according to the criteria of absolute Pearson correlation coefficient > 0.8 and p-value < 0.05. Subsequently, K-means analysis was performed on the transcriptome and metabolome data of different tissue samples using R software and its stats package (version 4.2.0). Finally, a regulatory network involving transcription factors, structure genes and metabolites was constructed via the Pearson correlation algorithm.
2.6 Quantitative real-time PCR
To validate the RNA-seq data, ten genes were selected for qRT-PCR. Gene-specific primers were designed using Primer Premier 5 software. cDNA was synthesized from 2 μg of total RNA using a cDNA synthesis kit (Thermo Scientific, M1682). qRT-PCR was performed in technical triplicates using PowerUp SYBR Green Master Mix (Yeasen, 11202ES03) on a QuantStudio 6 Flex Real-Time PCR system (Applied Biosystems, USA). Relative expression levels were calculated using the 2-ΔΔCt method, with actin11 (gene-VNO77_21608) serving as the internal reference gene.
2.7 Statistical analysis
One-way analysis of variance (ANOVA) was performed using IBM SPSS Statistics (Version 25.0) to determine statistical significance. For all analyses, the significance threshold was set at p < 0.05.
3 Results
3.1 Antioxidant capacity and physicochemical properties of different tissues in C. gladiata
The medicinal value of C. gladiata varies across different tissues due to distinct profiles of bioactive components (; Pal et al., 2022). To evaluate these differences, we determined antioxidant indices (DPPH, FRAP, and ABTS) in five tissues (roots, stems, leaves, seeds, and fruit pericarps). The results manifested that seeds exhibited the highest antioxidant activity, followed by leaves, while the fruit pericarps showed the lowest (Figures 1A, B). Consistent with these findings, seeds and leaves contained the highest concentrations of total flavonoids and total phenols. Furthermore, total terpenic acids were relatively enriched in leaves and fruit pericarps, but remained low in roots and seeds (Figure 1B).
Figure 1
3.2 Metabolomic analysis of different tissues in C. gladiata
To further investigate the spatiotemporal distribution of bioactive components, metabolomic profiling of diverse tissue samples was performed using liquid chromatography-mass spectrometry (LC-MS), and a total of 4,645 metabolites were identified (Supplementary Table 1). Among these metabolites, the number of metabolites annotated by the KEGG database was 1,062, accounting for approximately 22.9% of the total identified metabolites (Supplementary Table 1). Based on their skeleton structures, these metabolites were categorized into 11 distinct classes. The most abundant class of detected metabolites was Carboxylic acids and derivatives (accounting for 14.8%), followed by Prenol lipids (11.9%), Flavonoids (11.5%) and Organooxygen compounds (11.3%) (Figure 2A). Subsequently, unsupervised principal component analysis (PCA) was conducted for multivariate statistical analysis to identify the overall metabolic differences among different tissue samples of C. gladiata (Figure 2B). The results showed that principal component 1 (PC1) and principal component 2 (PC2) accounted for 30.65% and 26.85% of the total variance contribution rate, respectively, which could effectively distinguish samples from different tissues. The replicate samples of each tissue exhibited low variability, whereas a distinct differentiation pattern was observed between different tissue types. The metabolite accumulation patterns showed the highest similarity between roots and stems, as well as between seeds and fruit pericarps, while leaves displayed the most significant differences compared with other tissue.
Figure 2
3.3 Analysis of differentially accumulated metabolites
Based on the criteria of fold change (FC) ≥ 2 or ≤ 0.5 and variable importance in projection (VIP) ≥ 1, a total of 4,405 significantly DAMs were identified from pairwise comparisons between groups (Supplementary Table 2). The largest number of DAMs was detected in the comparison between seeds and roots, while the fewest were identified between seeds and fruit pericarps (Figure 2D). Venn diagram analysis revealed 795 DAMs shared across the four comparisons involving seeds, suggesting these metabolites may play unique roles in seed development (Figure 2E). Cluster heatmap analysis further confirmed tissue-specific spatial distribution patterns (Figure 2C).
The top 20 DEMs were identified in each comparison group (Supplementary Figure 3). In the comparison between seeds and roots, 6-Ketoestriol, Cussovantoside B and Myricetin 3-Robinobioside were significantly up-regulated, while the flavonoid metabolites 5, 7-Dihydroxy-3, 6-dimethoxyflavone, 5,6-Dihydroxy-4’-Methoxyflavanone and 3,7-Dihydroxy-3’,4’-Dimethoxyflavone were significantly down-regulated. In the comparison between leaves and seeds, the polyhydroxy unsaturated fatty acids (5r,6z,8e,10e,12s,14z)-5,12,20,20-Tetrahydroxyicosa-6,8,10,14-Tetraenoic Acid and (5r,6z,8e,12r,14z)-5,12,20,20,20-Pentahydroxyicosa-6,8,14-Trienoic Acid were significantly down-regulated, whereas the flavonoid metabolite Phloretin 3’,5’-Di-C-Glucoside was significantly up-regulated. In the comparisons between stems and seeds, as well as between fruit pericarps and seeds, the flavonoid metabolites Sophoronol B and 5,6-Dihydroxy-4’-Methoxyflavanone were both significantly up-regulated.
For the DAMs identified in each comparison group, KEGG enrichment analysis was performed to elucidate their associated metabolic pathways (Supplementary Figure 4). The results showed that the Flavone and flavonol biosynthesis and Isoflavonoid biosynthesis pathways were significantly enriched in the comparisons between stems and seeds, as well as between seeds and roots (P < 0.05). In addition, the Flavone and flavonol biosynthesis and Diterpenoid biosynthesis pathways were significantly enriched in the comparison between fruit pericarps and seeds (P < 0.05). These results suggested that the flavonoid and terpenoid biosynthesis pathways might be the key pathways responsible for the metabolic differences between seeds and other tissues.
3.4 Transcriptomic analysis of different tissues in C. gladiata
To explore the molecular mechanisms underlying the differences in antioxidant activity and metabolome among different tissues of C. gladiata, transcriptomic analysis was performed. A total of five RNA sequencing libraries were constructed from samples of different tissues (with three biological replicates per sample). After removing low-quality reads, we obtained a total of 133.4 Gb of clean data, with an average of 8.89 Gb per sample (Supplementary Table 3). The clean reads of each sample were mapped to the reference genome, and the mapping rates ranged from 84.6% to 96.4% (Supplementary Table 4). The Q30 value exceeded 96.33%, and the GC content was 41% (Supplementary Table 3), indicating that the RNA sequencing results were of extremely high quality. A total of 49,074 genes were detected to be expressed in the samples (Supplementary Table 5). The overall gene expression profiles of different samples are shown in Supplementary Figure 5. The Pearson correlation coefficients among the biological replicates of the five tissue samples were relatively high (0.9–1), suggesting that the transcriptomic data had good reproducibility (Figure 3A). The principal component analysis plot showed clear separation among different tissue types, and the biological replicates within each sample clustered closely together. The gene expression patterns in leaves, stems and roots exhibited certain similarities, whereas the gene expression pattern in seeds was significantly different from that in other tissues (Figure 3B).
Figure 3
3.5 Identification of differentially expressed genes
Based on the criteria of |log2(fold change)| ≥ 1 and false discovery rate (FDR) ≤ 0.05, a total of 25,597 DEGs were identified (Supplementary Table 6). The largest number of DEGs was observed in the comparison between seeds and roots, with 16,289 genes identified in total, among which 7,541 genes were up-regulated and 8,748 genes were down-regulated. This was followed by the comparisons between seeds and stems, as well as between seeds and leaves, where 16,193 and 15,845 DEGs were identified, respectively. On the other hand, the smallest number of DEGs was detected between stems and leaves, with only 7,925 DEGs, including 3,727 up-regulated genes and 4,198 down-regulated genes (Figure 3C). To further investigate the expression patterns of DEGs across different tissues, hierarchical clustering analysis was performed. The clustering results showed that the gene expression patterns among biological replicates within each sample group were highly similar, while some DEGs exhibited tissue-specific expression patterns (Figure 3D).
As core components of the plant gene expression regulatory network, transcription factors play crucial roles in mediating plant growth and development processes as well as regulating secondary metabolite biosynthesis. Transcription factor annotation of the DEGs identified a total of 1,330 transcription factors, which were classified into 53 families. Among these families, the bHLH family was the most abundant (accounting for 11.2%), followed by the ERF (8.4%), NAC (6.2%), WRKY (5.8%), C2H2 (5.3%) and MYB (5.0%) families (Supplementary Table 6; Supplementary Figure 6).
3.6 GO and KEGG enrichment analyses of differentially expressed genes
To analyze the functions of the DEGs, GO enrichment analysis was performed. GO enrichment analysis categorizes gene functions into three categories: biological process, cellular component, and molecular function. The results showed that significant differences were observed between seeds and roots, stems, leaves in terms of transcription regulator activity (GO:0140110), DNA-binding transcription factor activity (GO:0003700), kinase activity (GO:0016301), membrane (GO:0016020), cell periphery (GO:0071944), plasma membrane (GO:0005886), response to stimulus (GO:0050896), response to chemical (GO:0042221) and other functional terms (Figure 4).
Figure 4
To explore the gene functional pathways associated with the DEGs across different tissues, we conducted KEGG Enrichment Analyses. We selected the top 20 most important pathways and visualized them using scatter plots (Figure 4). The results revealed that DEGs were significantly enriched in metabolic pathways including Starch and sucrose metabolism (gmx00500), Glycolysis/Gluconeogenesis (gmx00010), Pyruvate metabolism (gmx00620), Fatty acid degradation (gmx00071), and Alanine, aspartate and glutamate metabolism (gmx00250). In the comparisons between seeds and other tissues, DEGs were enriched in pathways such as Flavonoid biosynthesis (gmx00941), Isoflavonoid biosynthesis (gmx00943), Flavone and flavonol biosynthesis (gmx00944) and Terpenoid backbone biosynthesis (gmx00900), indicating the characteristics of tissue-specific differentiation of secondary metabolites in C. gladiata.
3.7 Correlation analysis of transcriptome and metabolome expression data
Flavonoids and terpenoids are important secondary metabolites in C. gladiata, and they possess significant value in antioxidation, antidiabetic effects and anti-inflammatory activity (Pal et al., 2022). To dissect the molecular regulatory network of flavonoids and terpenoids in C. gladiata, we first performed a correlation analysis between the 348 screened DAMs belonging to flavonoids, isoflavonoids and terpenoids and the 25,597 DEGs. Genes significantly correlated with each DAMs were screened according to the criteria of absolute Pearson correlation coefficient > 0.8 and p-value < 0.05. A total of 22,615 genes co-regulated with at least one metabolite were identified (Supplementary Table 7).
Furthermore, the 348 metabolites and 22,615 genes were divided into 7 co-expression clusters via k-means analysis, which exhibited consistent or opposite expression patterns across different tissues of C. gladiata (Figure 5; Supplementary Table 7). The significantly correlated metabolites and genes within the same co-expression cluster indicated a closer expression regulatory relationship between them. Through the analysis of these 7 clusters, we identified tissue-enriched clusters: for instance, Cluster 1 was mainly enriched in roots, Cluster 2 in stems, Cluster 3 and Cluster 5 in seeds, Cluster 4 in fruit pericarps, and Cluster 6 in leaves. In addition, several flavonoid and terpenoid metabolites in Cluster 7 were simultaneously enriched in both leaves and roots.
Figure 5
3.8 Analysis of the regulatory network of flavonoid metabolites
Among all the differential accumulated metabolites, a total of 201 metabolites were annotated as flavonoids or isoflavonoids, including 85 in Cluster 1, 52 in Cluster 2, 18 in Cluster 3, 9 in Cluster 4, 1 in Cluster 5, 20 in Cluster 6, and 16 in Cluster 7 (Supplementary Figure 7). Meanwhile, based on the KEGG annotation and homology alignment results of the DEGs, we identified a total of 84 genes involved in the biosynthetic pathway of flavonoid metabolites. These included 1 PAL gene, 1 Cinnamate 4-hydroxylase (C4H) gene, 8 4-Coumarate: CoA ligase (4CL) genes, 8 CHS genes, 3 CHI genes, 1 Flavanone 3-hydroxylase (F3H) gene, 3 Flavonoid 3’-hydroxylase (F3’H) genes, 2 Flavonoid 3’,5’-hydroxylase (F3’5’H) genes, 3 Flavonol synthase genes, 4 Flavone 6-hydroxylase genes, 2 Anthocyanidin reductase (ANR) genes, as well as other modification genes involved in flavonoid biosynthesis (Figure 6A). Interestingly, we found that homologous genes exhibited distinct tissue-specific expression patterns. F3’H catalyzes the conversion of apigenin to luteolin. The three F3’H genes displayed different expression patterns: F3’H_1 (gene-VNO77_24489) was highly expressed mainly in leaves, followed by stems; F3’H_2 (gene-VNO77_24490) was highly expressed in stems, followed by leaves and roots; whereas F3’H_3 (gene-VNO77_42569) was highly expressed mainly in roots, followed by stems. The metabolites apigenin 5-(6’’-malonylglucoside) and luteolin 7-glucuronide are the derivatives of apigenin and luteolin, respectively, but showed different tissue-specific expression patterns. Apigenin 5-(6’’-malonylglucoside) was highly expressed mainly in roots, while luteolin 7-glucuronide was highly expressed mainly in stems and leaves (Figure 6B). ANR catalyzes the conversion of cyanidin to (-)-epicatechin. Among the two identified ANR genes, ANR_1 (gene-VNO77_05126) showed the highest expression in seeds and stems, and ANR_2 (gene-VNO77_05127) was most highly expressed in stems. The expression pattern of ANR_1 was consistent with the relatively high accumulation of the metabolite (-)-epicatechin in seeds.
Figure 6
We conducted an integrative analysis of co-expression clusters, encompassing DAMs, differentially expressed structural genes, and transcription factors, to screen for key transcription factors of flavonoid biosynthetic genes. Based on the screening criteria of Pearson correlation coefficient > 0.8 and p-value < 0.05, we identified that the expression levels of 7 transcription factors and 11 structural genes were closely correlated in Cluster 3. In particular, the transcription factors MYB4 (gene-VNO77_16383) and MYB1R1 (gene-VNO77_04235) showed significant correlations with the structural genes 4CL_1 (gene-VNO77_23988), CHS_1 (gene-VNO77_01236), CHS_4 (gene-VNO77_16715), CHI_2 (gene-VNO77_11878), F6H_3 (gene-VNO77_43128), FLS_2 (gene-VNO77_10355), and FLS_3 (gene-VNO77_31263) (Figure 6C). These genes exhibited the highest expression levels in the seeds of C. gladiata, but relatively low expression levels in other tissues (Figure 6B). Cluster heatmap analysis revealed that the metabolites including myricetin-3-O-galactoside, myricetin-3-O-xyloside, quercetin 3-arabinoside 7-glucoside, and laricitrin 3-rutinoside, were more abundant in seeds and showed strong co-expression with MYB4 (gene-VNO77_16383), MYB1R1 (gene-VNO77_04235), bHLH130 (gene-VNO77_15549), bHLH157 (gene-VNO77_13361), bHLH94 (gene-VNO77_02851), MYB11 (gene-VNO77_18081), and MYB61 (gene-VNO77_18224) (Figure 6B). During fruit development, the up-regulation of 4CL, CHS, CHI, FLS and F6H gene expression could contributes to the accumulation of flavonoid metabolites. These results indicated that these transcription factors might participate in the regulation of flavonoid biosynthesis through the relevant structural genes in seeds.
3.9 Analysis of the regulatory network of terpenoid metabolites
Among all the differential accumulated metabolites, a total of 147 metabolites were annotated as terpenoids, including 33 in Cluster 1, 17 in Cluster 2, 47 in Cluster 3, 17 in Cluster 4, 7 in Cluster 5, 25 in Cluster 6, and 1 in Cluster 7 (Supplementary Figure 7). Meanwhile, we analyzed the differential gene expression profiles of the mevalonate (MVA) pathway and the 2-C-methyl-D-erythritol 4-phosphate (MEP) pathway, which are involved in terpenoid backbone biosynthesis, to characterize the terpenoid biosynthetic pathways across different tissues. A total of 52 DEGs involved in the terpenoid biosynthetic pathway were identified (Figure 7A), including 6 AACT genes, 3 HMGS genes, 4 HMGR genes, 1 MK gene, 1 PMK gene, 4 DXS genes, 1 DXR gene, 1 MCT gene, 1 MECS gene, and 1 HDR gene. In the MVA pathway, 3-hydroxy-3-methylglutaryl-CoA reductase (HMGR) is the rate-limiting enzyme, whereas DXS and DXR are the key rate-limiting enzymes in the MEP pathway (Movahedi et al., 2022).The results of the cluster heatmap showed that the four HMGR genes exhibited distinct tissue-specific expression patterns. HMGR_1 (gene-VNO77_02361) had the highest expression in stems, followed by seeds and fruit pericarps. HMGR_2 (gene-VNO77_30043) and HMGR_3 (gene-VNO77_40459) shared a similar tissue expression pattern, with the highest expression in seeds. HMGR_4 (gene-VNO77_44685) showed the highest expression in leaves, followed by stems and fruit pericarps. Among the four identified differentially expressed DXS genes, DXS_1 (gene-VNO77_01613) and DXS_2 (gene-VNO77_05792) had the highest expression in seeds, whereas DXS_3 (gene-VNO77_24015) and DXS_4 (gene-VNO77_40184) were most highly expressed in fruit pericarps. The DXR gene (gene-VNO77_24682) had the highest expression in leaves. Two triterpenoid metabolites (acetylsoyasaponin A3 and soyasapogenol E monoglucuronide) were significantly enriched in the roots of C. gladiata, which was consistent with the tissue expression pattern of the CYP93 (gene-VNO77_17161), suggesting that the CYP93 might be involved in regulating the synthesis of the above-mentioned metabolites.
Figure 7
Based on the criteria of Pearson correlation coefficient > 0.8 and p-value < 0.05, we screened the transcription factors potentially involved in regulating terpenoid biosynthesis in different clusters. In Cluster 4, we found that the transcription factor MYC2 (gene-VNO77_40240) was strongly co-expressed with the TPS11 (gene-VNO77_28593), which was consistent with the findings of previous studies (Song et al., 2022). Further studies revealed that MYC2 was co-expressed with MYB1 (gene-VNO77_11271), a member of the MYB family of transcription factors, as well as with the structural genes 7-DLGT_1(gene-VNO77_05256), UGT74_1 (gene-VNO77_35292), DXS_3 (gene-VNO77_24015), and TPS11 (gene-VNO77_28593) (Figure 7B). In addition, the transcription factor MYB61 (gene-VNO77_18188) was co-expressed with the structural genes 7-DLGT_1 (gene-VNO77_05256) and 7-DLGT_3 (gene-VNO77_22234). Compared with other tissues, these genes showed the highest expression in the fruit pericarps of C. gladiata (Figure 7C). These results indicated that the transcription factors might be involved in the regulation of terpenoid biosynthetic genes.
3.10 Quantitative real-time PCR validation
To further verify the RNA-seq results, 10 candidate genes involved in the biosynthesis of flavonoid and terpenoid metabolites were selected for qRT-PCR analysis. The results showed that the expression trends of different genes in the five tissues of C. gladiata were consistent with those in the RNA-seq data, confirming the accuracy and reliability of the transcriptome results (Figure 8).
Figure 8
4 Discussion
Canavalia gladiata is a traditional leguminous crop that has garnered widespread attention for its high application value in the food, pharmaceutical, and health product industries. Previous studies on this species have primarily focused on its pharmacological activities and the functional characterization of a small subset of metabolites (Qian et al., 2025). However, the spatiotemporal distribution patterns of flavonoid and terpenoid metabolites across different C. gladiata tissues remain poorly characterized, and there is a critical lack of research dissecting the biosynthetic and regulatory networks governing these metabolites from an integrated multi-omics perspective. Accordingly, in this study, we integrated metabolomic and transcriptomic datasets to reveal, for the first time, the tissue-specific accumulation patterns and transcriptional regulatory networks of flavonoids and terpenoids in C. gladiata. Our findings provide a robust theoretical foundation for the targeted exploitation of bioactive components and metabolic engineering of this underutilized legume.
4.1 Medicinal potential of flavonoids and terpenoids in C. gladiata
In this study, we performed a systematic assessment of phenotypic, physiological and biochemical indices across five C. gladiata tissues: stems, leaves, roots, seeds and fruit pericarps. Our analysis revealed that the seeds possessed the strongest antioxidant activity, whereas fruit pericarps displayed the weakest. Meanwhile, the distribution patterns of total flavonoid and total phenolic acid contents were consistent with the antioxidant activity. This finding aligns with the well-documented role of flavonoids in scavenging free radicals and inhibiting oxidative stress, indicating that the seeds might be the tissue richest in antioxidant medicinal components in C. gladiata (Pal et al., 2022). In addition, although the fruit pericarps had relatively low antioxidant activity compared with other tissues, their high terpenoid acid content suggests that this tissue may possess unique medicinal value. Consistent with this, previous studies have shown that extracts from immature C. gladiate pods (ISBP) alleviated inflammation in RAW264.7 cells by modulating the production of pro-inflammatory cytokines ().
4.2 Tissue-specific accumulation of flavonoids and terpenoids in C. gladiata
Metabolomic profiling identified 4,405 DAMs, with 201 annotated as flavonoids and 147 as terpenoids. We observed that four flavonoid glycosides (Myricetin-3-O-Galactoside, Myricetin-3-O-Xyloside, Quercetin 3-Arabinoside 7-Glucoside, Laricitrin 3-Rutinoside) and three flavans((-)-Epigallocatechin Gallate, (-)-Epicatechin, (-)-Epiafzelechin 3-O-Gallate) were significantly enriched in seeds compared to other tissues (Supplementary Table 2). These metabolites have been widely reported to exert potent antioxidant, anti-inflammatory, and antitumor activities (; Pal et al., 2022), suggesting that they are key bioactive constituents in C. gladiata seeds. Previous studies have identified a variety of flavonoid compounds in C. gladiata seeds, including glycosylated derivatives of kaempferol and quercetin such as kaempferol 3-O-rutinoside and quercetin 3-(6-O-acetyl-β-glucoside) (More et al., 2022). Interestingly, we found that these compounds were highly enriched not only in the seeds but also in the leaves of C. gladiata (Supplementary Table 2; Figure 1). The high accumulation of flavonoid compounds in leaves may reflect an important adaptive mechanism of C. gladiata in response to the external environment. It is well established that drought and salt stress induce flavonoid accumulation in plants; these metabolites act as antioxidants to alleviate damage caused by reactive oxygen species (ROS), and also improve osmotic balance via signal regulatory pathways (Nakabayashi et al., 2014). Furthermore, studies indicate that glycosylation modification enhances the water solubility and stability of flavonoid metabolites, and facilitates their storage in vacuolars, thereby enabling rapid mobilization to exert antioxidant and defensive functions under stress (Yonekura-Sakakibara et al., 2019). Similarly, our results showed that terpenoid metabolites displayed distinct tissue-specific accumulation patterns in C. gladiata (Figure 1). We detected a greater number of terpenoid metabolites significantly enriched in seeds compared with all other tissues, predominantly diterpenoids and sesquiterpenoids. Further functional characterization of these tissue-specific terpenoid metabolites will facilitate a deeper understanding of the key bioactive constituents in C. gladiata and their potential roles in plant development and medicinal activity.
4.3 Tissue-specific expression of key genes in flavonoid and terpenoid biosynthesis
Based on the KEGG annotation and homology alignment results of DEGs, we identified 84 structural genes involved in the flavonoid biosynthetic pathway and 52 structural genes involved in the terpenoid biosynthetic pathway. Most enzymes related to flavonoid and terpenoid biosynthesis are encoded by multi-gene families rather than by single genes, consistent with previous reports in model and crop species including Arabidopsis thaliana and soybean (; Yonekura-Sakakibara et al., 2019). These genes often exhibit pronounced organ-specific or cell-type-specific expression patterns and are typically up-regulated under biotic or abiotic stress (). Gene duplication followed by subfunctionalization has been a driving force in legume flavonoid diversification, allowing distinct paralogs to fulfill organ-specific metabolic roles. We observed that the majority of phenylpropanoid pathway genes exhibited higher expression levels in stems of C. gladiata, followed by roots (Figure 6A). Correspondingly, a substantial number of flavonoid metabolites were significantly enriched in these tissues (Figure 2C), supporting coordinated regulation between gene expression and metabolite distribution. Similar stem-biased activation of phenylpropanoid metabolism has been reported in other medicinal plants (Zhu et al., 2021; Zhang et al., 2025). However, total flavonoid content was higher in leaves and seeds than in stems and roots (Figure 1). This discrepancy likely reflects a physiological trade-off, where the high diversity of metabolic intermediates in stems and roots does not necessarily translate to high quantitative accumulation of end products. This finding also suggests the presence of sophisticated, tissue-specific partitioning of flavonoid biosynthesis and storage in C. gladiata. Three flavonol glycosides, myricetin-3-O-xyloside, myricetin-3-O-galactoside and laricitrin 3-O-rutinoside, showed significant positive correlations in their abundance, and all were prominently enriched in C. gladiata seeds (Figure 6B).These metabolites are synthesized via the conserved flavonol branch of the phenylpropanoid pathway and the expression profiles of flavonoid 3’,5’-hydroxylase (F3’5’H) and flavonol synthase (FLS) co-determine the accumulation of myricetin-derived metabolites (Xing et al., 2021). Integrated transcriptomic and metabolomic analysis revealed significant co-expression between 4CL-1, CHS-1, CHS-4, CHI-2, F6H-1, F6H-3, FLS-2 and FLS-3 and these flavonol glycosides, indicating that these genes are key candidates mediating their biosynthesis.
The expression patterns of terpenoid biosynthetic genes showed no obvious global trend, but distinct tissue-specific characteristics were observed (Figure 7A). Genes in the mevalonate (MVA) pathway, including HMGS, HMGR, MK, and PMK, were preferentially expressed in seeds, fruit pericarps, and roots, with low transcript levels in leaves. In contrast, genes in the methylerythritol phosphate (MEP) pathway, such as DXR and MECS, were predominantly expressed in leaves. This spatial divergence underscores the tissue-specific partitioning of terpenoid metabolism, aligning with observations in other medicinal species like Panax ginseng and Camellia sinensis (Xue et al., 2019; ). Soyasaponins, an oleanane-type triterpenoid predominantly found in leguminous plants, has been reported to exhibit lipid-lowering, antioxidant, anti-inflammatory, and antitumor activities (). We found that soyasapogenol E monoglucuronide was mainly enriched in roots, and was significantly co-expressed with the CYP93 gene (Figure 7B). CYP93 belongs to the cytochrome P450 family, and encodes a key rate-limiting enzyme for triterpenoid saponin biosynthesis in legumes. Specifically, in soybean, CYP93 catalyzes the early steps of soyasaponin biosynthesis, and functions in concert with downstream UDP-glycosyltransferases (UGTs) and acetyltransferases (ACTs) to generate diverse saponin derivatives (; ).
4.4 Transcriptional regulatory networks of flavonoid and terpenoid metabolism
Transcription factors (TFs) play pivotal roles in regulating the biosynthesis of specialized secondary metabolites in medicinal plants (Shi et al., 2024). To explore the regulatory networks underlying flavonoid and terpenoid biosynthesis across different tissues of C. gladiata, we constructed a multi-omics co-expression network, which grouped genes and metabolites into seven distinct co-expression clusters. We found that genes and metabolites in cluster 3 exhibited significant co-expression specifically in seeds (Figure 5), and multiple transcription factors were significantly correlated with the flavonoid biosynthetic genes 4CL, CHS, F6H and FLS (Figure 6B). MYB11, a member of the R2R3-MYB TF family, has been well characterized to activate the expression of flavonoid biosynthetic genes including CHS, CHI, F3H, and FLS, thereby promoting flavonoid accumulation in plants (Stracke et al., 2007). Meanwhile, we found that the transcription factors bHLH94, bHLH130, and bHLH157 were significantly co-expressed with MYB11. Studies have shown that R2R3-MYB, bHLH, and WD40 proteins can form the MBW (MYB–bHLH–WD40) complex in plants. This complex directly binds to the promoter regions of structural genes related to flavonoid and anthocyanin biosynthesis, thereby driving the accumulation of secondary metabolites (). We therefore hypothesize that these transcription factors may interact with each other in C. gladiata seeds to coordinately regulate flavonoid biosynthesis. In addition, we found that the transcription factors MYB1R1 and MYB4 were strongly co-expressed. Mao et al. reported that MYB1R1 expression is significantly correlated with the transcript abundance of flavonoid biosynthetic pathway genes and flavonoid metabolite content (). Ectopic overexpression of CstMYB1R1 in saffron petals upregulated the expression of genes involved in the flavonoid and anthocyanin biosynthetic pathways (). In contrast, MYB4 is generally regarded as a negative regulator of flavonoid and phenylpropanoid metabolism. In Arabidopsis thaliana, AtMYB4 can directly repress the transcription of structural genes, thereby reducing the flux of phenylpropanoid metabolism and limiting flavonoid synthesis (). However, in mulberry (Morus alba) seeds, MaMYB4 cooperates with bHLH3 and TTG1 to regulate flavonoid metabolism, preventing excessive enhancement of metabolic flux through a feedback mechanism and maintaining the balance between flavonoids and anthocyanins (). This findings suggest that MYB4 may play a role in maintaining flavonoid homeostasis in C. gladiata seeds. Regarding the regulation of terpenoid biosynthesis, our results showed that MYC2 was co-expressed with terpenoid biosynthetic genes including DXS, TPS11, and multiple UGTs in C. gladiata fruit pericarps, and functioned synergistically with the TFs MYB1 and MYB61 (Figure 7B). Studies have demonstrated that MYC2 is a core regulatory factor in the jasmonic acid (JA) signaling pathway, which can activate genes in the MEP pathway (e.g., DXS, DXR and GGPPS) to promote terpenoid biosynthesis (; Zeng et al., 2025). Moreover, MYC2 functions synergistically with members of the R2R3-MYB family to regulate the expression of terpenoid biosynthetic genes (Yang et al., 2020). This functional synergy suggests that MYC2 may interact with MYB1 or MYB61 to modulate terpenoid accumulation in C. gladiata. Collectively, these findings provide novel insights into the transcriptional regulation of flavonoid and terpenoid metabolites across different C. gladiata tissues, and identify candidate genes for future functional validation and potential metabolic engineering applications. Further studies using dual-luciferase reporter assays, yeast one-hybrid assays, or transgenic approaches will be required to confirm the regulatory roles of these candidate TFs in C. gladiata.
5 Conclusion
C. gladiata is a plant with both nutritional and medicinal values. Various tissues of C. gladiata (roots, stems, leaves, seeds, and fruit pericarps) exhibit extensive biological effects in antioxidation, anti-inflammation and immune regulation. However, the tissue-specific differences in bioactive components and their underlying molecular regulatory mechanisms remain to be elucidated. Antioxidant activity assays showed that the seeds of C. gladiata had the highest antioxidant activity and total phenol content, whereas the leaves had the highest contents of total flavonoids and total terpenic acids. A total of 4,405 DAMs were identified across tissues by LC-MS analysis, revealing pronounced tissue-specific distribution patterns. The flavonoid and terpenoid biosynthesis pathways were significantly enriched in the KEGG analysis, consistent with the notion that they are key regulators underlying the metabolic divergence between seeds and other tissues. RNA-Seq analysis identified 25,597 DEGs, including 1,330 transcription factors. Subsequently, integrated correlation and k-means clustering of these data with metabolite profiles enabled the construction of regulatory networks and the classification of co-expressed metabolites and genes into 7 distinct clusters. Guided by these clusters, we uncovered potential “metabolite-transcription factor-structural gene” regulatory relationships specific to flavonoid and terpenoid metabolism in seeds and fruit pericarps, respectively. This study provides comprehensive insights into the metabolome and transcriptome of different tissues of C. gladiata, and laying a solid foundation for the functional development and resource utilization of this plant.
Statements
Data availability statement
The original contributions presented in the study are publicly available. This data can be found here: PRJNA1449233.
Author contributions
YG: Investigation, Writing – review & editing, Conceptualization, Writing – original draft, Resources, Formal analysis, Project administration, Visualization, Data curation. YY: Methodology, Software, Writing – original draft, Data curation. YMZ: Software, Writing – original draft, Methodology, Data curation. YZ: Data curation, Methodology, Software, Writing – original draft. ZF: Data curation, Software, Methodology, Writing – original draft.
Funding
The author(s) declared that financial support was received for this work and/or its publication. This research was funded by Science and Technology Research Project Plan of Hubei Provincial Department of Education, grant number B2023272, and the Special Research Project of Wuhan Business University, grant number 2023KT05.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fpls.2026.1792177/full#supplementary-material
Supplementary Figure 1OPLS-DA plots of pairwise comparisons between different groups. (A). OPLS-DA plot of the comparison between stems and seeds; (B). OPLS-DA plot of the comparison between seeds and mature roots; (C). OPLS-DA plot of the comparison between leaves and seeds; (D). OPLS-DA plot of the comparison between pod pericarps and seeds. The abscissa represents the predictive principal component; the ordinate represents the orthogonal principal component. The percentage indicates the explanation rate of the component to the dataset. Each point represents a sample, and samples from the same group are marked with the same color.
Supplementary Figure 2Permutation test plots of OPLS-DA for pairwise comparisons between different groups. (A). Permutation test plot of OPLS-DA for the comparison between stems and seeds; (B). Permutation test plot of OPLS-DA for the comparison between seeds and mature roots; (C). Permutation test plot of OPLS-DA for the comparison between leaves and seeds; (D). Permutation test plot of OPLS-DA for the comparison between pod pericarps and seeds. The abscissa represents the model values of R²Y and Q²; the ordinate represents the frequency of the model classification effect in 200 random permutation experiments. In the figure, orange represents R²Y of the random grouping model, purple represents Q² of the random grouping model, and the values indicated by black arrows are R²X, R²Y and Q² of the original model.
Supplementary Figure 3Bar charts of differential accumulated metabolites from pairwise comparisons between different groups. (A). Bar chart of differential accumulated metabolites from the comparison between stems and seeds; (B). Bar chart of differential accumulated metabolites from the comparison between seeds and mature roots; (C). Bar chart of differential accumulated metabolites from the comparison between leaves and seeds; (D). Bar chart of differential accumulated metabolites from the comparison between pod pericarps and seeds. The abscissa represents the Log2FC of differential metabolites; the ordinate represents the top 20 metabolites ranked by fold change. Red indicates up-regulated metabolite content, and green indicates down-regulated metabolite content.
Supplementary Figure 4KEGG pathway analysis of differential accumulated metabolites from pairwise comparisons between different groups. (A). KEGG pathway enrichment plot of differential accumulated metabolites from the comparison between stems and seeds; (B). KEGG pathway enrichment plot of differential accumulated metabolites from the comparison between seeds and mature roots; (C). KEGG pathway enrichment plot of differential accumulated metabolites from the comparison between leaves and seeds; (D). KEGG pathway enrichment plot of differential accumulated metabolites from the comparison between pod pericarps and seeds. The abscissa represents the Rich Factor corresponding to each pathway; the ordinate represents pathway names. The color of the points reflects the magnitude of P-value, and the size of the points represents the number of enriched differential metabolites.
Supplementary Figure 5Boxplots of gene expression levels in C. gladiata samples. The abscissa represents sample names; the ordinate represents FPKM values.
Supplementary Figure 6Statistical chart of the number of transcription factor families. The abscissa represents transcription factor families; the ordinate represents the number of identified members of the corresponding transcription factor families in C. gladiata.
Supplementary Figure 7Statistical chart of the number of flavonoid, isoflavonoid and terpenoid metabolites in different clusters. The abscissa represents different clusters; the ordinate represents the number of flavonoid, isoflavonoid and terpenoid metabolites in different clusters.
References
1
AlfieriM.VaccaroM. C.CappettaE.AmbrosoneA.De TommasiN.LeoneA. (2018). Coactivation of MEP-biosynthetic genes and accumulation of abietane diterpenes in Salvia sclarea by heterologous expression of WRKY and MYC2 transcription factors8, 11009. doi: 10.1038/s41598-018-29389-4
2
BanerjeeS.AgarwalP.ChoudhuryS. R.RoyS. (2024). MYB4, a member of R2R3-subfamily of MYB transcription factor functions as a repressor of key genes involved in flavonoid biosynthesis and repair of UV-B induced DNA double strand breaks in211, 108698. doi: 10.1016/j.plaphy.2024.108698
3
BhatZ. Y.MirJ. A.YadavA. K.SinghD.AshrafN. (2024). CstMYB1R1, a REVEILLE-8-like transcription factor, regulates diurnal clock-specific anthocyanin biosynthesis and response to abiotic stress in Crocus sativus L43, 20. doi: 10.1007/s00299-023-03082-x
4
BiancoG.PascaleR.CarboneC. F.AcquaviaM. A.CataldiT. R. I.Schmitt-KopplinP.et al. (2018). Determination of soyasaponins in Fagioli di Sarconi beans (Phaseolus vulgaris L.) by LC-ESI-FTICR-MS and evaluation of their hypoglycemic activity410, 1561–1569. doi: 10.1007/s00216-017-0806-8
5
BulanovA. N.AndreevaE. A.TsvetkovaN. V.ZykinP. A. (2025). Regulation of flavonoid biosynthesis by the MYB-bHLH-WDR (MBW) complex in plants and its specific features in cereals26, 734. doi: 10.3390/ijms26020734
6
CâmaraJ.-S.PerestreloR.FerreiraR.BerenguerC.-V.PereiraJ.-A.CastilhoP.-C. (2024). Plant-derived terpenoids: A plethora of bioactive compounds with several health functions and industrial applications—a comprehensive overview29, 3861. doi: 10.3390/molecules29163861
7
CaoY.MeiY.ZhangR.ZhongZ.YangX.XuC.et al. (2024). Transcriptional regulation of flavonol biosynthesis in plants11, uhae043. doi: 10.1093/hr/uhae043
8
ChenF.ThollD.BohlmannJ.PicherskyE. (2011). The family of terpene synthases in plants: a mid-size family of genes for specialized metabolism that is highly diversified throughout the kingdom66, 212–229. doi: 10.1111/j.1365-313X.2011.04520.x
9
ChenC.ZhuH.KangJ.WarusawitharanaH. K.ChenS.WangK.et al. (2022). Comparative transcriptome and phytochemical analysis provides insight into triterpene saponin biosynthesis in seeds and flowers of the tea plant (Camellia sinensis)12, 204. doi: 10.3390/metabo12030204
10
ConfalonieriM.CarelliM.GianoglioS.MogliaA.BiazziE.TavaA. (2021). CRISPR/Cas9-mediated targeted mutagenesis of CYP93E2 modulates the triterpene saponin biosynthesis in12, 690231. doi: 10.3389/fpls.2021.690231
11
DixonR. A.AchnineL.KotaP.LiuC. J.ReddyM. S.WangL. (2002). The phenylpropanoid pathway and plant defence-a genomics perspective3, 371–390. doi: 10.1046/j.1364-3703.2002.00131.x
12
DongS.HanX.ManX.DengX.ChenZ. (2025). Tissue-specific biosynthesis and regulation of alkaloids, flavonoids, and terpenoids in fenugreek (Trigonella foenum-graecum L.): insights from integrated metabolomics and transcriptomics analysis16, 1669610. doi: 10.3389/fpls.2025.1669610
13
EkanayakeS.JanszE. R.NairB. M. (2000). Literature review of an underutilized legume: Canavalia gladiata L55, 305–321. doi: 10.1023/A:1008119107738
14
FanY.HussainS.WangX.YangM.ZhongX.TaoL.et al. (2024). Metabolomic and transcriptomic analyses of flavonoid biosynthesis in different colors of soybean seed coats26, 294. doi: 10.3390/ijms26010294
15
GanR.-Y.LuiW.-Y.CorkeH. (2016). Sword bean (Canavalia gladiata) as a source of antioxidant phenolics51, 156–162. doi: 10.1111/ijfs.12979
16
GanR. Y.WangM. F.LuiW. Y.WuK.DaiS. H.SuiZ. Q.et al. (2017). Diversity in antioxidant capacity, phenolic contents, and flavonoid contents of 42 edible beans from China94, 291–297. doi: 10.1094/CCHEM-03-16-0061-R
17
HwangK.-A.HeoW.HwangH.-J.HanB. K.SongM. C.KimY. J. (2020). Anti-inflammatory effect of immature sword bean pod (Canavalia gladiata) in lipopolysaccharide-induced RAW264.7 cells23, 1183–1191. doi: 10.1089/jmf.2020.4733
18
HwangY. J.HwangH.-J.GoH.ParkN.HwangK.-A. (2023). Sword bean (Canavalia gladiata) pods induce differentiation in MC3T3-E1 osteoblast cells by activating the BMP2/SMAD/RUNX2 pathway15, 4372. doi: 10.3390/nu15204372
19
HwangH.-J.YoonH.ChoJ.-H.LeeS.HwangK.-A.KimY. J. (2025). Immature sword bean (Canavalia gladiata) pod alleviates allergic rhinitis (a double-blind trial) through PI3K/Akt/mTOR signaling17, 468. doi: 10.3390/nu17030468
20
JoshiK. R.DevkotaH. P.Al-MutairiK. A.SugimuraK.YaharaS.KhadkaR.et al. (2024). Therapeutic potential of Leea asiatica: chemical isolation and validation of ethnomedicinal claims through in vitro and in silico assessment of antioxidant and anti-inflammatory properties10, e38074. doi: 10.1016/j.heliyon.2024.e38074
21
KimJ.-P.YangY.-S.KimJ.-H.LeeH.-H.KimE.-S.MoonY.-W.et al. (2012). Chemical properties and DPPH radical scavenging ability of sword bean (Canavalia gladiata) extract44, 441–446. doi: 10.9721/KJFST.2012.44.4.441
22
KrishnamurthyP.FujisawaY.TakahashiY.AbeH.YamaneK.MukaiyamaK.et al. (2019). High-throughput screening and characterization of a high-density soybean mutant library elucidate the biosynthesis pathway of triterpenoid saponins60, 1082–1097. doi: 10.1093/pcp/pcz025
23
LaijaS. N.MaheshS. P.SmithaL. S.RemaniP. (2010). Lymphocyte proliferation studies of Canavalia gladiata lectin8, 51–55.
24
LiY.ChenY.ZhouL.YouS.DengH.ChenY.et al. (2020a). MicroTom metabolic network: rewiring tomato metabolic regulatory network throughout the growth cycle13, 1203–1218. doi: 10.1016/j.molp.2020.06.005
25
LiH.YangZ.ZengQ.WangS.LuoY.HuangY.et al. (2020b). Abnormal expression of bHLH3 disrupts a flavonoid homeostasis network, causing differences in pigment composition among mulberry fruits7, 49. doi: 10.1038/s41438-020-0302-8
26
LiW. W.YuJ. Y.XuH. L.BaoJ. K. (2011). Concanavalin A: a potential anti-neoplastic agent targeting apoptosis, autophagy and anti-angiogenesis for cancer therapeutics414, 282–286. doi: 10.1016/j.bbrc.2011.09.072
27
LiC.ZhaW.LiW.WangJ.YouA. (2023). Advances in the biosynthesis of terpenoids and their ecological functions in plant resistance24, 11561. doi: 10.3390/ijms241411561
28
LoveM. I.HuberW.AndersS. (2014). Moderated estimation of fold change and dispersion for RNA-seq data with DESeq215, 550. doi: 10.1186/s13059-014-0550-8
29
MaQ.WangS.TanH.SunZ.LiC.ZhangG. (2024). Tissue-specific transcriptome analyses unveils candidate genes for flavonoid biosynthesis, regulation and transport in the medicinal plant14, 29999. doi: 10.1038/s41598-024-81319-9
30
MaoJ.GaoZ.WangX.YaoD.LinM.ChenL. (2024). Integrated transcriptome and targeted metabolome analyses provide insights into flavonoid biosynthesis in kiwifruit (Actinidia chinensis)14, 19417. doi: 10.1038/s41598-024-70600-6
31
Ministry of Agriculture and Rural Affairs of the People’s Republic of China (2020). Determination of total triterpene in Ganoderma—spectrophotometric method (Beijing: Agriculture Press of China). NY/T 3676-2020.
32
MoreK. C.TayadeS. N.GawandeP. A.ManikS.ShelkeD. (2022). Antioxidant and antimicrobial potential of Canavalia gladiata (Jacq.) DC. leaves and seeds: GC-MS based metabolic profiling13, 163–169. doi: 10.56042/ijnpr.v13i2.47499
33
MovahediA.WeiH.PuckerB.Ghaderi-ZefreheiM.RasouliF.Kiani-PouyaA.et al. (2022). Isoprenoid biosynthesis regulation in poplars by methylerythritol phosphate and mevalonic acid pathways13, 968780. doi: 10.3389/fpls.2022.968780
34
NakabayashiR.Yonekura-SakakibaraK.UranoK.SuzukiM.YamadaY.NishizawaT.et al. (2014). Enhancement of oxidative and drought tolerance in Arabidopsis by overaccumulation of antioxidant flavonoids77, 367–379. doi: 10.1111/tpj.12388
35
PalM.PandeyV. S.DhawalA.UpadhyayP.VartikaSharmaR.et al. (2022). The pharmacological and therapeutic activities of Canavalia gladiata (Jacq.) DC12, 208–216. doi: 10.22270/jddt.v12i5-s.5729
36
QianH.MengD.YueL.XuH.FengK.WangJ. (2025). Ethnobotanical use, phytochemistry, pharmacology, and toxicity of19, 3779–3803. doi: 10.2147/DDDT.S519102
37
RokkamR.PinipayF.PaidiH.TamanamR. (2023). Phytochemical analysis and evaluation of the antioxidant and anti-inflammatory activity of16, 3157–3164. doi: 10.52711/0974-360X.2023.00519
38
ShenS.ZhanC.YangC.FernieA. R.LuoJ. (2023). Metabolomics-centered mining of plant metabolic diversity and function: past decade and future perspectives16, 43–63. doi: 10.1016/j.molp.2022.09.007
39
ShiM.ZhangS.ZhengZ.MaozI.ZhangL.KaiG. (2024). Molecular regulation of the key specialized metabolism pathways in medicinal plants66, 510–531. doi: 10.1111/jipb.13634
40
SinghS.ChhatwalH.PandeyA. (2024). Deciphering the complexity of terpenoid biosynthesis and its multi-level regulatory mechanism in plants. 43, 3320–33336. doi: 10.1007/s00344-024-11347-2
41
SongC.CaoY.DaiJ.LiG.ManzoorM. A.ChenC.et al. (2022). The multifaceted roles of MYC2 in plants: toward transcriptional reprogramming and stress tolerance by jasmonate signaling13, 868874. doi: 10.3389/fpls.2022.868874
42
StrackeR.IshiharaH.HuepG.BarschA.MehrtensF.NiehausK.et al. (2007). Differential regulation of closely related R2R3-MYB transcription factors controls flavonol accumulation in different parts of the Arabidopsis thaliana seedling50, 660–677. doi: 10.1111/j.1365-313X.2007.03078.x
43
WangZ.XuX.ZhangP.HuangY.ZhaoC.El-KassabyY. A.et al. (2025). Multi-omics integration reveals organ-specific biosynthesis of flavonoids and terpenoids in the medicinal plant16, 1675432. doi: 10.3389/fpls.2025.1675432
44
WuJ.LvS.ZhaoL.GaoT.YuC.HuJ.et al. (2023). Advances in the study of the function and mechanism of the action of flavonoids in plants under environmental stresses257, 108. doi: 10.1007/s00425-023-04136-w
45
XingM.CaoY.RenC.LiuY.LiJ.GriersonD.et al. (2021). Elucidation of myricetin biosynthesis in Morella rubra of the Myricaceae108, 411–425. doi: 10.1111/tpj.15449
46
XuS.HuE.CaiY.XieZ.LuoX.ZhanL.et al. (2024). Using clusterProfiler to characterize multiomics data19, 3292–3320. doi: 10.1038/s41596-024-01020-z
47
XuP.LiJ.ChenC.ChenJ.YangM.DengH.et al. (2025). Integrated transcriptomic and metabolomic analyses reveal tissue-specific accumulation and expression patterns of monoterpene glycosides, gallaglycosides, and flavonoids in Paeonia lactiflora Pall Pall. 26, 561. doi: 10.1186/s12864-025-11750-3
48
XueL.HeZ.BiX.XuW.WeiT.WuS.et al. (2019). Transcriptomic profiling reveals MEP pathway contributing to ginsenoside biosynthesis in20, 383. doi: 10.1186/s12864-019-5718-x
49
YangZ.LiY.GaoF.JinW.LiS.KimaniS.et al. (2020). MYB21 interacts with MYC2 to control the expression of terpene synthase genes in flowers of Freesia hybrida and Arabidopsis thaliana71, 4140–4158. doi: 10.1093/jxb/eraa184
50
Yonekura-SakakibaraK.HigashiY.NakabayashiR. (2019). The origin and evolution of plant flavonoid metabolism10, 943. doi: 10.3389/fpls.2019.00943
51
ZengT.SuH.WangM.HeJ.GuL.WangH.et al. (2025). The role of MYC2 transcription factors in plant secondary metabolism and stress response mechanisms14, 1255. doi: 10.3390/plants14081255
52
ZhangK.WangN.GaoX.MaQ. (2022). Integrated metabolite profiling and transcriptome analysis reveals tissue-specific regulation of terpenoid biosynthesis in114, 110388. doi: 10.1016/j.ygeno.2022.110388
53
ZhangM.ZengZ.SunC.WangS.HuR.LiuY.et al. (2025). Identification, evolution, and expression analysis of ten gene families of general phenylpropanoid and monolignol-specific metabolism in cucumber177, e70300. doi: 10.1111/ppl.70300
54
ZhuY.WangQ.WangY.XuY.LiJ.ZhaoS.et al. (2021). Combined transcriptomic and metabolomic analysis reveals the role of phenylpropanoid biosynthesis pathway in the salt tolerance process of22, 2399. doi: 10.3390/ijms22052399
55
ZhuangW. B.LiY. H.ShuX. C.PuY. T.WangX. J.WangT.et al. (2023). The classification, molecular structure and biological biosynthesis of flavonoids, and their roles in biotic and abiotic stresses28, 3599. doi: 10.3390/molecules28083599
Summary
Keywords
Canavalia gladiata, flavonoids, metabolome, terpenoids, transcriptome
Citation
Gao Y, Yang Y, Zhang Y, Zhao Y and Fan Z (2026) Integrative metabolomic and transcriptomic analyses reveal the regulatory mechanisms underlying the biosynthesis of flavonoid and terpenoid metabolites in different tissues of Canavalia gladiata. Front. Plant Sci. 17:1792177. doi: 10.3389/fpls.2026.1792177
Received
20 January 2026
Revised
23 February 2026
Accepted
10 March 2026
Published
15 April 2026
Volume
17 - 2026
Edited by
Rufeng Wang, Shanghai University of Traditional Chinese Medicine, China
Reviewed by
Ni Zhan, Langfang Normal University, China
Longsheng Hu, Northeast Forestry University, China
Updates
Copyright
© 2026 Gao, Yang, Zhang, Zhao and Fan.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Yue Gao, gaoyue@m.scnu.edu.cn
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.