Next Article in Journal
Green Soybean’s Survival Mechanisms Under Cold Stress: A Transcriptomic Perspective
Previous Article in Journal
A Tillage-Dependent System of Arable Pests: How Soil Condition and Prevailing Climate Influence Pest Occurrence?
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Agronomy
Volume 15
Issue 6
10.3390/agronomy15061455
agronomy-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Integrated Transcriptomic and Metabolomic Analysis Identified Key Transcriptional Factors Involved in Flavonoid and Alkaloid Biosynthesis Among Different Tissues of Sophora flavescens

by
Ake Liu
1,†,
Jingjing Dong
1,†,
Mingyang Wang
2,
Jin Li
1,
Junjie Lu
1,
Lintao Liu
3,
Haiying Lei
1,
Zhen Zeng
4,* and
Huifang Song
1,*
1
Department of Life Sciences, Changzhi University, Changzhi 046011, China
2
School of Life Science, Shanxi Normal University, Taiyuan 030031, China
3
Modern Research Center for Traditional Chinese Medicine, Shanxi University, Taiyuan 030006, China
4
School of Life Sciences, Nanjing University, Nanjing 210023, China
*
Authors to whom correspondence should be addressed.
†
These authors contributed equally to this work.
Agronomy 2025, 15(6), 1455; https://doi.org/10.3390/agronomy15061455
Submission received: 2 May 2025 / Revised: 9 June 2025 / Accepted: 11 June 2025 / Published: 15 June 2025
(This article belongs to the Section Plant-Crop Biology and Biochemistry)
Browse Figures
Versions Notes

Abstract

:
Sophora flavescens has anti-inflammatory, analgesic, antibacterial, antiviral, and anti-tumor properties due to its active compounds, including alkaloids and flavonoids. Exploring the biosynthetic regulation mechanism of these compounds is crucial. Here, we identified 227 flavonoids and 55 alkaloids among five tissues (roots, stems, leaves, flowers, and pods) using wide-targeted metabolomics. Flavonoids were predominantly in roots, whereas alkaloids were primarily in roots and flowers. RNA sequencing revealed 18,488 differentially expressed genes (DEGs) in non-root tissues compared to roots. These DEGs were significantly enriched in pathways such as plant hormone signal transduction, carbon metabolism, flavonoid biosynthesis, and isoflavonoid biosynthesis. Utilizing K-means clustering and WGCNA, we identified ten transcription factors (TFs) potentially regulating the biosynthesis and accumulation of flavonoids (e.g., SfWRKY50 and SfbHLH078) and seven TFs involved in alkaloid biosynthesis (e.g., SfbHLH109 and SfbHLH162) in S. flavescens. These TFs can serve as candidate genes for studying the biosynthetic mechanisms of flavonoids and alkaloids, providing insights into the production of key active compounds and supporting the rational use of S. flavescens tissues.

1. Introduction

The dry roots of Sophora flavescens are commonly used in traditional Chinese medicine, and they have properties such as clearing damp heat, promoting diuresis, eliminating parasites, dispelling wind, and alleviating itching [1]. The root contains alkaloids and flavonoids as its main bioactive components, offering significant clinical and economic value [1]. The flavonoids in S. flavescens predominantly consist of prenylated flavonoids with a pterocarpene structure and side chain attached to the flavonoid skeleton (C6-C3-C6) [2]. These compounds exhibit various pharmacological activities, such as antioxidative, antibacterial, anti-inflammatory, and antitumor effects [3]. Yang et al. successfully extracted 21 prenylated flavonoids from S. flavescens roots, demonstrating their ability to trigger anti-proliferative effects by inducing autophagic cell death [4]. Flavonoids also inhibit indoleamine 2,3-dioxygenase 1, a critical survival factor for tumor cells, suggesting a potential use in anticancer immunotherapy [5]. Kurarinone exhibits promise in treating Parkinson’s-like symptoms in animal models by inhibiting soluble epoxide hydrolase [6]. Currently, plant flavonoids are mainly synthesized through the phenylpropanoid pathway. Genes affecting flavonoid metabolism mainly consist of two categories: structural genes and regulatory genes. Structural genes are responsible for encoding enzymes such as phenylalanine ammonia lyase (PAL), chalcone synthase (CHS), cinnamic acid 4-hydroxylase (C4H), and 4-coumarate CoA ligase (4CL). Regulatory genes mainly consist of MYB, bHLH (basic helix-loop-helix), and WD40, which together create the MBW complex. This complex plays a crucial role in governing various stages of flavonoid biosynthesis [7,8,9].
Alkaloids are important natural products, with physiological activities including anti-tumor, anti-inflammatory, anti-gout, immune regulation, and anti-allergy properties. According to their structures, they can be classified into types such as tetrahydroisoquinoline, indole, pyrrolizidine, piperidine, indolizidine, pyridine, quinoline, and xanthine alkaloids [10,11]. Matrine, a tetracyclic quinoline alkaloid from S. flavescens, is used as an anti-tumor drug (Compound Matrine Injection) [12]. It increases apoptosis, reduces migration, and inhibits proliferation in liver cancer HepG2 cells by activating Mst1-JNK pathway-mediated mitochondrial fission [13]. Oxymatrine can inhibit survival and single-cell proliferation through cell cycle arrest and mitochondrial-mediated apoptosis in breast cancer cell lines [14]. It also suppresses the EGFR-Akt signaling pathway, thereby significantly inhibiting non-small cell lung cancer tumor growth [15]. Quinolizidine alkaloids (QAs) are crucial active components in S. flavescens [1,16], reflecting their quality to some extent. They are lysine-derived alkaloids with a biosynthetic pathway that is not yet fully understood. QA biosynthesis involves multiple steps regulated by enzymes like lysine/ornithine decarboxylase (LDC) and copper amine oxidase (CuAO) [17] and transcription factors (TFs) such as AP2/ERF, WRKY, and MYB members [18].
Omics analysis techniques are widely applied in medicinal plant studies, crucial for elucidating the biosynthetic mechanism of secondary metabolites (SMs) [19]. Metabolomics-based methods have compared active components between Tartary and common buckwheat seeds, highlighting Tartary buckwheat’s superior health benefits [20]. Studies have also leveraged metabolomics for the high-throughput profiling of SMs in medicinal plants [21]. Analyzing differentially expressed genes (DEGs) and regulatory patterns under varied conditions aids in identifying related gene functions. For instance, RNA sequencing (RNA-seq) technology explored the underlying mechanisms involved in flavonoid biosynthesis and the accumulation in S. flavescens root tissues across different growth years [22]. Li et al. identified terpene biosynthetic genes in Dendrobium officinale using transcriptome and volatile terpene compound analysis [23]. Li et al. resolved the monoterpene indole alkaloid biosynthesis pathway at the single-cell level in Catharanthus roseus using single-cell transcriptomics and metabolomics [24].
However, the distribution of bioactive components among different tissues, as well as the biosynthetic pathways and regulatory mechanisms for them, remain largely unclear in S. flavescens. Understanding these processes is crucial for optimizing the production of these valuable compounds and enhancing the medicinal value of the plants. Hence, by integrating gene expression data with metabolite accumulation data, this study aimed to identify key regulatory genes for the biosynthesis of flavonoids and alkaloids. This integrative approach can reveal regulatory networks and provide valuable information for genetic improvement and metabolic engineering to increase the yield and quality of bioactive compounds.

2. Materials and Methods

2.1. Plant Materials and Sampling

To study the content differences of flavonoids and alkaloids among five tissues (roots, stems, leaves, flowers, and young pods), we collected samples from five-year-old S. flavescens with consistent growth conditions (Figure 1). For each tissue, three biological replicates were collected. The S. flavescens plants grow naturally after sowing without any artificial intervention in the Changzhi International Shennong Traditional Chinese Medicine Culture Expo Park, located in Shangdang District, Changzhi City, Shanxi Province. After rinsing them with PBS buffer, the samples were immediately submerged in liquid nitrogen for rapid freezing and subsequently stored at −80 °C for further transcriptome and metabolome detection.

2.2. Metabolite Extraction of Samples

Sufficient quantities of each tissue of S. flavescens shown in Figure 1 were freeze-dried using a freeze-dryer (Scientz-100F, Ningbo Scientz Biotechnology Co., Ltd., Ningbo, China). The dried samples were then pulverized into powder using a grinding instrument (MM400, Retsch, Haan, Germany) for 90 s at a frequency of 30 Hz. In total, 100 mg powder of each sample was weighed using an analytical balance (AS 60/220 R2, RADWAG, Poland) and added to 1.2 mL of 70% methanol solution. The resulting mixture was vortexed six times at 30 min intervals, each time for 30 s, utilizing a multi-tube vortex mixer (MIX-200, Shanghai Jingxin Industrial Development Co., Ltd., Shanghai, China). Afterward, the mixture was stored overnight in a refrigerator set at 4 °C. The next day, the sample underwent centrifugation at 11,304× g for 10 min. The supernatant was harvested, passed through a 0.22 μm micro-porous filter, and transferred to an injection vial for subsequent UPLC-MS/MS analysis.

2.3. UPLC-MS/MS Analysis

UPLC conditions: Ultra-pure water (Milli-Q® ultrapure water system, Millipore Corporation, Burlington, MA, USA); acetonitrile (HPLC grade, Tedia Company, Fairfield, OH, USA); formic acid (LC-MS grade, Aladdin Bio-Chem Technology Co., Ltd., Shanghai, China). UPLC: Agilent SB-C18 column (1.8 µm, 2.1 mm × 100 mm); mobile phase A: ultra-pure water (containing 0.1% formic acid) and mobile phase B: acetonitrile (containing 0.1% formic acid) used as mobile phase; elution gradient: commencing at 5% B, linearly transitioning to 95% within 9.00 min and remaining at 95% for 1 min, then decreasing to 5% from 10.00 to 11.10 min and equilibrating at 5% until 14 min; flow rate: 0.35 mL·min−1, column temperature: 40 °C, injection volume: 4 μL.
Mass spectrometry conditions: Tandem mass spectrometry system (Applied Biosystems 4500 QTRAP, AB SCIEX, Framingham, MA, USA). An electrospray ionization (ESI) source was used, and the instrument was operated in both positive and negative ion modes under the control of Analyst software (v1.6.3, AB Sciex). The ESI source operation parameters were as follows: ion source, turbo spray; source temperature 550 °C; ion spray voltage 5500 V (positive ion mode)/−4500 V (negative ion mode); source gas I (GSI), gas II (GSII), and curtain gas (CUR) were set at 50, 60, and 25 psi, respectively; collision-activated dissociation (CAD) parameters were set to high. Instrument tuning and mass calibration were performed using 10 and 100 μmol·L−1 polypropylene glycol solutions in QQQ and LIT modes, respectively. QQQ scans were conducted in multiple reaction monitoring (MRM) mode, with nitrogen as the collision gas at a medium setting. Further optimization of the declustering potential and collision energy for each MRM transition was carried out by making additional adjustments to ion pairs. A specific set of MRM ion pairs was monitored for each period based on the metabolites eluted. The peak extraction and correction used the MWDB database from Wuhan Metware Biotechnology Co., Ltd. (Wuhan, China). Qualitative analysis relied on secondary spectrum information, while isotope signals; K+, Na+, and NH4+ ions; and fragment ion duplicates were removed. Quantification was accomplished using the MRM mode on a triple quadrupole mass spectrometer. After obtaining mass spectrometry data, peak areas of all the metabolites were integrated and corrected across different samples [25].

2.4. Metabolomics Data Analysis

Unsupervised principal component analysis (PCA) was conducted using Simca-P software (v13.0), and the filtered data were standardized by Z-score. Metabolite clustering analysis among the samples was performed using R software (v4.2.1). To gain deeper insights into the metabolite distribution, orthogonal partial least squares discriminant analysis (OPLS-DA) was employed to identify the differentially accumulated metabolites (DAMs) across various tissues. The selection criteria were defined as VIP (variable importance in projection) ≥ 1 and |log2(FC)| ≥ 1, where FC represents the fold change.

2.5. RNA Sequencing and Analysis

Total RNA was isolated from 15 samples utilizing the Total Mini prep kit (Axygen Biosciences, Union City, CA, USA). The integrity and purity of RNA were assessed using the Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA, USA). Poly(A) mRNA was isolated with the TruSeq mRNA Sample Preparation Kit (Illumina, San Diego, CA, USA). Subsequently, fragmentation was carried out at a high temperature in the proprietary fragmentation buffer of Illumina with divalent cations. Total RNA (1 μg) was converted into cDNA through reverse transcription using the PrimeScript RT Master Mix (TaKaRa, Dalian, China), and then library fragments were amplified by PCR. Following this, the quality and effective concentration of the libraries were assessed using the Agilent 2100 Bioanalyzer. Finally, pair-end (PE) sequencing was performed for these libraries on the Illumina HiSeq™ 2000 platform.
Quality control (QC) and the filtering of raw data were performed using fastp (v0.23.2) [26]. The alignment of high-quality reads to the reference genome of S. flavescens (accessed on 31 May 2023 from https://doi.org/10.5281/zenodo.7750935) [17] was performed using HISAT2 (v2.2.1) [27]. To ensure the comparability of gene expression levels across different genes or samples, normalization was performed using FPKM (Fragments Per Kilobase of transcript per Million fragments mapped) through featureCounts (v2.0.3) [28] with default parameters. DESeq2 (v1.22.1) was utilized to analyze DEGs [29], with the filtering criteria set as |log2(FC)| > 1.0 and the p-adj value < 0.05.

2.6. Functional Enrichment Analysis

The potential function of the genes was annotated by the Gene Ontology (GO) resource and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway database. To predict the potential functions of DEGs, hypergeometric tests were used to calculate the significance of each GO and KEGG pathway in the context of the whole genome (corrected p value < 0.05). The p values were subjected to a false discovery rate (FDR, range: 0–1).

2.7. K-Means Cluster and Gene Co-Expression Network Analysis

The RNA-seq raw data have been uploaded to the NCBI database (PRJNA1136989). The K-means method was used to perform clustering on the DAM expression profiles (five clusters), and correlation tests between each DEG and the content of each metabolite (stats package in R v4.2.0) were conducted. Those DEGs with a |Pearson correlation coefficient (r) value| > 0.9 (p < 0.05) were retained. The DEG-DAM pairs with the highest correlation were selected, and the DEGs were grouped according to their associated DAM clusters. Finally, five gene clusters positively and negatively correlated with the metabolite clusters were obtained, respectively. The R-based weighted correlation network analysis (WGCNA) package (v1.71) [30] was utilized to build the co-expression network of the 15 expression profiles of S. flavescens. A soft threshold value (β) of 18 was chosen to guarantee that the network exhibited an approximately scale-free topology (Figure S1). Subsequently, the neighbor-joining matrix was transformed into a topological overlap matrix (TOM) to measure the associated dissimilarity. Hierarchical clustering modules were established, and a clustering dendrogram was created using TOM (Figure S2). These modules represent groups of genes that are highly interconnected, where genes within the same group exhibit strong correlation coefficients. In this analysis, modules containing at least 50 genes and showing highly correlated eigengenes (with a threshold of 0.25) were combined. Within the co-expression network, the weight value of edge (ranging from 0 to 1) was calculated based on their topological overlap to represent the connection strengths between any two linked genes. The network, filtered for edge weights exceeding 0.3, was visualized using Cytoscape (v3.8.2) [31].

2.8. RT-qPCR and Statistical Analysis

Total RNA was extracted from each tissue of S. flavescens using an RNA extraction kit (GeneBetter, Beijing, China). RNA integrity was confirmed by Nanodrop 2000 (Thermo Fisher, Waltham, MA, USA). Reverse transcription and qPCR were conducted following our previous study [9]. The primers were designed by Primer5 software (Table S1). The EF-1α gene was selected as the internal reference. The levels of relative expression from the three biological replicates were determined using the 2−△△Ct method and statistically evaluated via one-way ANOVA followed by Duncan’s test.

3. Results

3.1. Transcriptional Metabolic Network Built from Wide-Targeted Metabolomics and Transcriptomics

To construct a comprehensive metabolic regulatory network in S. flavescens across five tissues, parallel metabolic profiling and transcriptome analysis from five samples (Figure 1) with three biological replicates were integrated to generate a transcriptional metabolic network dataset. QC analysis of the metabolomes showed a high consistency in retention time and peak intensity across UPLC-MS/MS detections, indicating stable signals for the same samples. In total, 227 flavonoids and 55 alkaloids were identified from the five tissues of S. flavescens (Figure 2A, Table S2). The former included flavones (23.79%), isoflavones (21.15%), flavonols (16.30%), flavanones (14.10%), flavonoid carbonoside (7.49%), chalcones (5.29%), flavanols (2.20%), flavanonols (0.88%), biflavones (0.44%), and other flavonoids (8.37%); the latter included alkaloids (30.91%), isoquinoline alkaloids (1.82%), phenylamine (9.09%), piperidine alkaloids (3.64%), plumeran indole (18.18%), pyridine alkaloids (1.82%), pyrrole alkaloids (1.82%), and QAs (32.73%). This indicates that S. flavescens is rich in flavonoids and alkaloids.
To analyze the differential distribution of metabolites, hierarchical clustering was performed across different tissues. The roots formed distinct clusters separate from the stems, leaves, flowers, and pods. Meanwhile, the stems grouped together with the leaves, and the flowers clustered with the pods (Figure 2A). In S. flavescens, the flavonoid content was highest in the roots, followed by flowers and pods, with lower levels in stems. Alkaloids were more abundant in flowers and roots but less in other tissues (Figure 2A). Based on the K-means clustering method, all 282 metabolites could be classified into five clusters (Figure S3A). Cluster 1 showed a lower metabolite content in roots compared with other tissues (Figure S3B,C, e.g., rutin and sophoranol), while Cluster 4 had a higher root content (e.g., kurarinone and mamanine), especially for all kushenols and kurarinols. Overall, the tissue-specific differences were significant, with flavonoids concentrated in roots and alkaloids in flowers and roots.
PCA was conducted on the metabolite profiles of flavonoids and alkaloids in S. flavescens. The PC1 accounted for 48.57% of the variance, separating roots from other tissues (leaves, stems, flowers, and pods). The PC2 contributed 19.98% to the variance, distinguishing stems and leaves from the remaining three tissues (Figure 2B). This suggests significant variations in metabolite profiles across the five tissues. The root metabolite content significantly differed from other tissues (Figure 2C), with root containing the highest number and concentration of flavonoids and alkaloids (244 types). Stems and leaves had fewer types (228 and 220 types, respectively). A total of 172 metabolites were shared across all the tissues (130 flavonoids and 42 alkaloids), while unique metabolites were detected in roots (21), leaves (1), flowers (2), and pods (1), respectively (Figure 2D, Table 1). In summary, metabolite accumulation patterns varied significantly among tissues, with roots showing distinct differences to other non-root tissues.
After QC, 99.25 Gb high-quality reads were obtained for samples of five S. flavescens tissues, with an average Q20 of 98.11% and Q30 of 94.46%. The read mapping rate to the reference genome exceeded 92.8%, indicating reliable sample collection and sequencing. PCA and correlation analysis showed that biological replicates had higher correlation coefficients than between samples (Figure 2B,C). Among the 15 samples, approximate half of the genes showed either no or very low levels of expression (Figure 2E, FPKM < 1, 44.10–49.25%), while expressed genes that were moderately expressed accounted for 48.16–53.06% (1 ≤ FPKM ≤ 100). Highly expressed genes made up 2.14–2.87% (FPKM > 100). A total of 15,444 genes were detected with FPKM > 1 across all these samples (Figure 2F). Overall, the RNA-seq data exhibits a high quality and is appropriate for subsequent analysis.

3.2. DAM and DEG Analysis Among S. flavescens Tissues

Given the significant differences in metabolites between S. flavescens roots and the other four tissues, DAM analyses were conducted to explore tissue-specific accumulation patterns. Totally, 258 DAMs were detected (VIP ≥ 1 and |log2(FC)| ≥ 1), including 210 flavonoids and 48 alkaloids (Figure S4A). Compared with roots, stems had 173 differentially accumulated flavonoids (67 increased, 106 decreased) and 27 alkaloids (13 increased, 14 decreased). Leaves had 170 flavonoids (75 increased, 95 decreased) and 35 alkaloids (14 increased, 21 decreased). Flowers had 171 flavonoids (83 increased, 88 decreased) and 26 alkaloids (18 increased, 8 decreased). Pods had 160 flavonoids (78 increased, 82 decreased) and 37 alkaloids (22 increased, 15 decreased) (Figure 3A). In total, 131 DAMs were shared across all comparisons (Figure 3B), including 120 flavonoids (22 flavones, 19 isoflavones, 25 flavonols, 18 flavanones, 1 flavanonols, 11 flavonoid carbonosides, 4 chalcones, 1 flavanols, and 19 other flavonoids) and 11 alkaloids (3 alkaloids, 1 phenylamines, 1 plumeran indoles, and 6 QAs) (Figure S4A). The majority of the 120 shared flavonoids were most abundant in roots (54.2%) and least abundant in stems (Figure S4B), while the majority of the 11 shared alkaloids were most abundant in flowers and leaves and least abundant in stems (Figure S4C).
To clarify gene expression differences among tissues in S. flavescens, 18,488 DEGs were identified using a threshold of log2(FC) ≥ 1 and p-adj < 0.05. It showed that 8739 DEGs (3274 upregulated and 5465 downregulated), 11,247 DEGs (4672 upregulated and 6575 downregulated), 11,951 DEGs (4528 upregulated and 7423 downregulated), and 10,614 DEGs (3951 upregulated and 6663 downregulated) were identified in Stem vs. Root, Leaf vs. Root, Flower vs. Root, and Pod vs. Root, respectively (Figure S5A). Among these comparison groups, a total of 2602, 4272, 4184, and 3551 genes showed a ≥8-FC (Figure 3C). Except for Stem vs. Root, over half of the DEGs had an FC greater than 4, indicating their significant changes in gene expression. In total, there were 4160 shared DEGs identified among these four comparisons (Figure 3D). Key enzymes and TFs related to flavonoid or alkaloid biosynthesis were validated by RT-qPCR (Figure 3E), showing a strong correlation with transcriptome data.
To gain insights into the roles of DEGs, functional enrichment analyses were conducted for those identified in the four comparison groups. The results showed significant enrichment in numerous GO and KEGG categories (corrected p value ≤ 0.05, Figure 3F,G). In GO enrichment, 172 biological processes (BPs), 17 cell components (CCs), and 124 molecular functions (MFs) were found to be significantly enriched across all four comparisons (Figure S5B, Table S3). For example, significantly enriched BP terms included the phenylpropanoid and flavonoid metabolic/biosynthetic process, isoprenoid biosynthetic process, and cellular amino acid catabolic process; significantly enriched MF terms included UDP-glucosyltransferase activity, UDP-glycosyltransferase activity, O-methyltransferase activity, and O-acetyltransferase activity. Furthermore, Stem vs. Root, Leaf vs. Root, Flower vs. Root and Pod vs. Root had 22, 24, 36, and 25 significantly enriched pathways, respectively (Figure 3F, Table S4). Fourteen pathways showed significant enrichment across all four comparisons (Figure S5B), including plant hormone signal transduction, carbon metabolism, flavonoid biosynthesis, and isoflavonoid biosynthesis (Figure 3F), implying differential flavonoid accumulations may relate to tissue-specific gene expression.

3.3. Correlation Analysis of DEGs and DAMs Based on K-Means Clustering

To understand the accumulation and distribution of DAMs in different tissues of S. flavescens and the expression changes of DEGs, K-means clustering was applied to DAMs (Figure 4A, Cluster 1–5). Based on r value (|r| > 0.9, p < 0.05), five clusters positively and five negatively correlated with metabolite clusters were identified (Figure 4B,C). Figure 4A shows tissue-specific metabolite abundance: Cluster 1 (99 DAMs) was highly expressed in roots, mainly flavonoids (87) and alkaloids (12); Cluster 2 (46 DAMs) and Cluster 3 (25 DAMs) were most abundant in leaves and stems; Cluster 4 (53 DAMs) had the highest content in flowers, with the largest proportion of alkaloids (accounting for one-third); Cluster 5 (35 DAMs) was most abundant in pods.
To associate gene expression with metabolite accumulation, 10,007 (positive) and 17,224 (negative) DEGs were identified to co-regulate with at least one of the 258 DAMs (Figure 4B,C). It indicates that under normal growth conditions, S. flavescens gene expression patterns are largely paralleled by the dynamics of major metabolic pathways. Among these significantly correlated DEGs, 1106 and 727 encode TFs. A total of 302 TFs related to flavonoid biosynthesis (Figure 4D) and 158 TFs related to alkaloid biosynthesis were identified (Figure 4E). Based on correlations between TF expression levels and DAF content, four candidate core regulatory TFs for the biosynthesis of flavonoids (SfbHLH078, SfWRKY48, SfbZIP30, and SfWRKY50) and three for alkaloids (SfbHLH153, SfbHLH074, and SfERF097) were screened (Figure 4F).

3.4. Regulatory Network of Flavonoids and Alkaloid Biosynthesis in S. flavescens

In total, 28,320 genes were analyzed using WGCNA after normalization and filtering, resulting in the identification of 27 co-expression modules (Figure 5A and Figure S2). The correlation between modules and the relative content of flavonoids and alkaloids was evaluated using correlation analysis (Figures S6 and S7) (|r| > 0.75, p < 0.05). It indicated that flavonoid levels were positively correlated with the turquoise module (5361 genes) and negatively correlated with the darkturquoise (153), royalblue (221), and darkgrey (117) modules (Figure 5B). For instance, kurarinone showed a strong positive correlation with the turquoise module (r = 0.95, p < 0.05) and negative correlations with the darkturquoise, royalblue, and darkgrey modules (r = −0.77, −0.81, and −0.82; p < 0.05). Enrichment analysis indicated that plant hormone signal transduction, MAPK signaling pathway, and starch and sucrose metabolism were significantly enriched in the positively correlated modules, while plant hormone signal transduction, flavonoid biosynthesis, and biosynthesis of SMs were significantly enriched in the negatively correlated modules (Figure 5C and Table S5). Additionally, 1515, 59, 70 and 50 genes in the turquoise, darkturquoise, royalblue, and darkgrey modules, respectively, exhibited differential expression across the four comparisons (Stem vs. Root, Leaf vs. Root, Flower vs. Root, and Pod vs. Root).
Meanwhile, the relative content of alkaloids (especially QAs) was positively correlated with the green (2688 genes) and yellow (2940) modules and negatively correlated with the grey60 (340) module (Figure 5B). For instance, 9α-hydroxysophoramine showed a positive correlation with the green and yellow modules (r = 0.84 and 0.92, respectively) and a negative correlation with the grey60 module (r = −0.92; p < 0.05). It showed that glycolysis/gluconeogenesis, TCA cycle, plant hormone signal transduction, starch and sucrose metabolism, biosynthesis of amino acids, and pentose phosphate pathway were significantly enriched in positively correlated modules, and plant hormone signal transduction and MAPK signaling pathway were significantly enriched in negatively correlated modules (Figure 5C and Table S5). There were 298, 385, and 2 genes in the green, yellow, and grey60 modules that exhibited differential expression across the four comparisons.

3.5. Key Enzymes and TFs Regulating the Flavonoid Biosynthesis in S. flavescens

Accordingly, the expression levels of key enzymes were mapped to the proposed biosynthetic pathways for flavonoids/isoflavonoids (Figure 6A). The relative content of different flavonoids and derivatives are also exhibited (Figure 6B). At least one flavonoid’s detected content showed a significant correlation with the expression levels of these 20 key enzymes in different tissues (Figure S8). These compounds are derived from phenylalanine through enzymatic reactions. For example, a high expression of the CHS gene in roots may lead to the accumulation of isoliquiritigenin in this tissue, and HIDH gene expression was positively correlated with two daidzein derivatives (6-hydroxydaidzein and 2′-hydroxydaidzein). The accumulation of rutin was significantly positively correlated with the expression of the FG2 gene, and the derivative of epicatechin (epicatechin-4′-O-β-D-glucopyranoside) showed a significant positive correlation with the expression of the ANR (anthocyanidin reductase) gene (Figure S8). In addition, the content of those flavonoids and their derivatives was also more likely to be significantly positively correlated (Figure 6C). Therefore, the expression of these key enzyme genes influences the biosynthesis and accumulation of flavonoids, with tissue-specific differences leading to varied flavonoid levels in S. flavescens.
Besides structural enzyme genes, TFs like MYB, bHLH, and WRKY can also independently or in collaboration control multiple enzymatic steps in flavonoid biosynthesis. Among the four network modules related to flavonoids, four differentially expressed TFs were significantly associated with the content of more flavonoids and their derivatives in S. flavescens, including SfWRKY50, SfbZIP32, SfMYB275, and SfERF006 (Figure 6C). Additionally, other important flavonoid components, such as kurarinone, kushenol, and sophoraflavanone and their derivatives, were found to accumulate specifically in roots (Figure S9). A correlation analysis was performed between the metabolites and the differentially expressed TFs from four network modules. The results revealed that five TFs were positioned at the core of the network (Figure 6D). Among them, four TFs (SfMYB275, SfWRKY53, SfbHLH073, and SfERF023) exhibited a significantly positive correlation with the metabolites, whereas SfbHLH014 showed a significant negative correlation with all these important flavonoid components. These observations indicate that these TFs are likely to play essential roles in the regulation of flavonoids’ biosynthesis and accumulation.

3.6. Key Enzymes and TFs Regulating QA Biosynthesis in S. flavescens

The core enzyme in QA biosynthesis, LDC, converts L-lysine to cadaverine through decarboxylation. This gene showed no expression in the roots but exhibited an extremely high expression in stems (Figure 7A, FPKM > 4700), with a relatively low expression in the other three tissues. The CuAO family oxidizes cadaverine to 5-aminopentanal. Among the 11 CuAO members, most showed a higher expression in stems than other tissues, especially CuAO01, CuAO08, CuAO09, and CuAO10. CuAO03 had a relatively high expression across all five tissues, with especially high levels in roots, surpassing any other gene’s expression in any tissue. The content of 18 detected QAs was further compared (Figure S10A) across different tissues, and it was found that mamanine, cytisine, and baptifoline had the highest content in roots, and flowers had the highest proportion of QAs, with a higher content compared with other tissues (Figure 7B, 7/18). Except for sophoranol and lehmannine, the content of other QAs correlated significantly with LDC or CuAO expression levels in different tissues (Figure S10B).
Two co-expression networks were built utilizing the QA content and expression levels of TFs and key enzyme genes (Figure 7C,D). Hub genes identified in these networks include CuAO02, CuAO08, CuAO09, CuAO10, and LDC. Within the networks, CuAO02 and CuAO08 are the most interconnected key enzymes in the alkaloid metabolic pathway. Many TFs, such as WRKY, bHLH, and MYB, were co-expressed with LDC and CuAO genes, indicating their potential roles in regulating enzyme activities in QA biosynthesis. Besides enzyme genes, some TFs like bHLH and WRKY have been documented to be essential for alkaloid biosynthesis. Within the three modules, four differentially expressed TFs showed a significant correlation with the levels of QAs in S. flavescens, including SfWRKY26, SfbHLH037, SfbHLH109, and SfbHLH162 (Figure 7E), suggesting their importance in modulating the biosynthesis and accumulation of QAs.

4. Discussion

4.1. Metabolomics Analysis Revealed Profile Variations in Flavonoids and Alkaloids Among Different Tissues of S. flavescens

S. flavescens, a traditional Chinese medicine herb, contains primarily flavonoids and alkaloids in its roots [32]. Recently, more than 130 flavonoids and 50 alkaloids have been discovered [16,33,34,35]. However, little is known about their types and accumulation characteristics among different tissues. In total, 227 flavonoids and 55 alkaloids were identified in the flower, leaf, root, stem and pod tissues of S. flavescens using wide-targeted metabolomics technology (Figure 2A). Among them, flavonoids were predominantly found in roots, while alkaloids were mainly present in flowers and roots. All the 21 metabolites unique to roots were flavonoids (Table 1), with various types exhibiting physiological activities. For instance, 5,7,4′-Trimethoxyflavone inhibits the proliferation of human SNU-16 gastric cancer cells [36]. Fu et al. demonstrated that glabrone from Glycyrrhiza uralensis showed inhibitory effects on COVID-19 [37]. Kushenol A suppresses the growth of breast cancer cells by targeting the PI3K/AKT/mTOR signaling pathway [38]. The root extract of S. flavescens has traditionally been utilized as a clinical drug for treating tumors [39], which is related to the rich flavonoids and alkaloids in the roots. In total, 131 metabolites showed differential accumulation among the four comparisons, comprising 120 flavonoids and 11 alkaloids (Figure 3B and Figure S4A). K-means clustering analysis showed significant differences in flavonoid and alkaloid profiles across the five tissues. Notably, 99 DAMs were highly accumulated in roots (Figure 4A), including 87 flavonoids and 12 alkaloids. Metabolomics analysis of active components in different parts aids in the rational development and utilization of S. flavescens resources.

4.2. Dynamic Changes in Flavonoid and Alkaloid Biosynthesis in S. flavescens

Currently, the biosynthetic pathways of flavonoids and alkaloids and their regulatory factors in S. flavescens remain largely unclear. Understanding these processes is crucial for optimizing the production of these valuable compounds and enhancing the medicinal value of the plants [20,23]. Comparing roots with other tissues, 18,488 DEGs and 258 DAMs were identified in total. Functional enrichment analysis revealed that four significantly enriched pathways, including plant hormone signal transduction, carbon metabolism, flavonoid biosynthesis, and isoflavonoid biosynthesis, were shared across the four comparisons (Figure 3H). Flavonoid biosynthesis has been extensively studied in model plants. These compounds originate from phenylalanine and undergo a sequence of enzymatic transformations facilitated by crucial enzymes, including PAL, C4H, and 4CL. This process ultimately results in the biosynthesis of various flavonoids and their associated derivatives [40,41,42]. The biosynthesis and accumulation of flavonoids are not only influenced by enzyme genes but also regulated by multiple TFs [43]. In S. flavescens, several typical alkaloids belong to the QAs [1,16], so our study mainly focused on the regulation of QA biosynthesis. The correlation analysis between the expression of curial enzymes and the profile of metabolites indicates that the production of flavonoids and QAs is affected by the expression levels of enzyme-related genes. The differences in their expression levels and the regulatory roles of TFs may cause the differential accumulation of these two main components among different tissues (Figure 6A and Figure 7A).
A co-expression network was constructed by integrating transcriptome and metabolome data, and 27 co-expressed modules were identified in total (Figure 5B). The pathway related to plant hormone signal transduction was found to be enriched in modules that were both positively and negatively correlated with flavonoid and alkaloid content (Figure 5C). Accordingly, plant hormones can indeed regulate the biosynthesis of SMs. For instance, melatonin regulates the accumulation of SMs in grape skins through the induction of VvMYB14’s expression [44]; melatonin increases the content of flavonoids under UV-B stress by enhancing the expression and activity of key flavonoid enzymes in soybeans [45]. The external application of salicylic acid and methyl jasmonate to Capparis spinosa can upregulate the expression of enzyme genes in rutin biosynthesis and increase the rutin content [46]. SlERF.H6 can influence the buildup of steroidal glycoalkaloids in tomatoes by mediating the ethylene and gibberellin signaling pathways [47]. The jasmonic acid signaling cascade is crucial for the biosynthesis of anthocyanins, sesquiterpenes, and alkaloids [48,49]. Additionally, signaling pathways interact to regulate the accumulation of SM. For instance, in soybeans, the interaction between melatonin and ethylene activates isoflavone biosynthesis and antioxidant activity, increases isoflavone content, and mitigates ethephon’s effect on germination [50].

4.3. Key TFs Regulating Flavonoid and Alkaloid Biosynthesis via Integrated Analysis of Transcriptomics and Metabolomics

By analyzing the relationship between TF expression and flavonoid content, the typical TF families responsible for regulating flavonoid biosynthesis were determined in S. flavescens (Figure 4, Figure 6, and Figure 7). Previous studies have shown that many bHLH proteins form complexes with other TFs and play crucial roles in regulating flavonoid biosynthesis [18,48,51]. For example, the VvMYC1 gene, which encodes bHLH proteins, collaborates with MYB to control the expression of three flavonoid biosynthesis genes: ANR, UFGT (UDP glucose: flavonoid-3-O-glucosyltransferase), and Chalcone Isomerase [52]. The CPC (a MYB member) and GL3 (a bHLH member) from Arabidopsis have been shown to influence anthocyanin biosynthesis in tomatoes [53]. The MBW complex can regulate flavonoid biosynthesis [7,8], as well as other TFs such as AP2, ERF, bZIP, and WRKY TF families [54]. Additionally, other TF families, such as bHLH, MYB, AP2/ERF, and WRKY, also contribute significantly to alkaloid biosynthesis (nicotine, steroidal glycoalkaloids, monoterpenoid indole alkaloids, and benzylisoquinoline alkaloids, etc.) [18,51]. In this study, through WGCNA and correlation analysis, the TFs mainly associated with the productions of flavonoids and alkaloids were identified as MYB, WRKY, bHLH, and ERF family members (Figure 6 and Figure 7). Of course, further functional verification is still needed for these potential key regulatory TFs for the biosynthesis of flavonoids and alkaloids. Moreover, due to the limitations of wide-targeted metabolomics technology, the absolute content of typical flavonoids and alkaloids in various tissues remains unknown, and further quantitative and in-depth exploration is still required.

5. Conclusions

This study analyzed the tissue distribution and content of two main active compounds, alkaloids and flavonoids, and gene expression profiles of these among different tissues of S. flavescens based on transcriptome and metabolome methods. K-means clustering and WGCNA were used to analyze the potential mechanisms and key candidate genes and TFs of flavonoid and alkaloid biosynthesis. Our study could not only offer a deeper understanding of the biosynthesis mechanism of flavonoids and alkaloids but also offer reference information for the rational development and utilization of different tissue parts of S. flavescens.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/agronomy15061455/s1, Figure S1: Filter of appropriate soft thresholds; Figure S2: Module clustering and heatmaps of modular gene clustering; Figure S3: K-means clustering of metabolites and their relative content comparison across five tissues. A. K-means clustering based on metabolite content, where light red lines represent individual metabolite levels, and red lines represent median values. Numbers marked at the upper right corner indicate the counts of flavonoids and alkaloids in each cluster. B and C. Distribution of 5 flavonoids and 5 alkaloids across five tissues. CPS denotes content per second. Vertical bars labeled with different letters (a, b, or c) indicate significantly differences as determined by ANOVA followed by Duncan’s test (p < 0.05); Figure S4: Comparison of DAMs. A. Comparison of DAFs and DAAs. B. Distribution of 120 DAFs across five tissues. C. Distribution of 11 DAAs across five tissues; Figure S5: Comparison of DEG numbers and significantly enriched categories among different comparisons; Figure S6: Correlation between co-expressed modules and flavonoids; Figure S7: Correlation between co-expressed modules and alkaloids; Figure S8: Correlation between flavonoid content in different tissues and key enzyme genes in flavonoid synthesis; Figure S9: The relative content of other important flavonoids among five tissues of S. flavescens; Figure S10: The detected QAs in S. flavescens. A. Chemical structural formula of 18 identified QAs in S. flavescens. B. Correlation between QA content in different tissues and expression of LDC and CuAO genes; Table S1: The primer sequences for 9 selected genes used in RT-qPCR; Table S2: Flavonoid and alkaloid components detected in different tissues of Sophora flavescens; Table S3: The results of GO enrichment analysis for DEGs among four comparisons; Table S4: The results of KEGG enrichment analysis for DEGs among four comparisons; Table S5: The significantly enriched KEGG pathways of genes from 7 modules (turquoise, darkturquoise, royalblue, darkgrey, green, yellow, and grey60).

Author Contributions

Conceptualization, Z.Z. and H.S.; methodology, H.S. and Z.Z.; software, A.L., J.D. and M.W.; validation, M.W., H.S., J.D. and L.L.; formal analysis, A.L., M.W., J.L. (Junjie Lu) and J.D.; investigation, Z.Z., L.L. and H.L.; resources, A.L. and H.S.; data curation, M.W. and J.L. (Jin Li); writing—original draft preparation, A.L., M.W., J.D. and J.L. (Jin Li); writing—review and editing, H.L., H.S. and Z.Z.; visualization, A.L., J.L. (Junjie Lu) and J.D.; supervision, H.S. and Z.Z.; project administration, A.L., H.S. and Z.Z.; funding acquisition, A.L. and H.L. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by the Shanxi Basic Research Foundation (Free Exploration) project (202203021221235, 20210302124145), Shanxi Scholarship Council of China (2021-153), and Project of Key Research Base for Humanities and Social Sciences of Universities in Shanxi (2022J030).

Data Availability Statement

The raw data was deposited in NCBI under Bioproject PRJNA1136989.

Acknowledgments

We thank Chenglin Liu (Fudan University) for his technical support and constructive suggestions.

Conflicts of Interest

The authors declare no conflicts of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.

Abbreviations

The following abbreviations are used in this manuscript:
DEGDifferentially expressed gene
TFTranscription factor
PALPhenylalanine ammonia lyase
CHSChalcone synthase
C4HCinnamic acid 4-hydroxylase
4CL4-coumarate CoA ligase
bHLHBasic helix-loop-helix
QAQuinolizidine alkaloid
LDCLysine/ornithine decarboxylase
CuAOCopper amine oxidase
SMSecondary metabolite
RNA-seqRNA sequencing
ESIElectrospray ionization
CURCurtain gas
MRMMultiple reaction monitoring
PCAPrincipal component analysis
OPLS-DAOrthogonal partial least squares discriminant analysis
DAMDifferentially accumulated metabolite
VIPVariable importance in projection
GOGene Ontology
KEGGKyoto Encyclopedia of Genes and Genomes
FDRThe p values were subjected to false discovery rate
WGCNAWeighted correlation network analysis
TOMTopological overlap matrix
BPBiological process
CCCell component
MFMolecular function
UFGTUDP glucose: flavonoid-3-O-glucosyltransferase
CADCollision-activated dissociation

References

  1. Li, J.; Zhang, X.; Shen, X.; Long, Q.; Xu, C.; Tan, C.; Lin, Y. Phytochemistry and biological properties of isoprenoid flavonoids from Sophora flavescens Ait. Fitoterapia 2020, 143, 104556. [Google Scholar] [CrossRef] [PubMed]
  2. Boozari, M.; Soltani, S.; Iranshahi, M. Biologically active prenylated flavonoids from the genus Sophora and their structure–activity relationship—A review. Phytother. Res. 2019, 33, 546–560. [Google Scholar] [CrossRef] [PubMed]
  3. Kong, S.; Liao, Q.; Liu, Y.; Luo, Y.; Fu, S.; Lin, L.; Li, H. Prenylated flavonoids in Sophora flavescens: A systematic review of their phytochemistry and pharmacology. Am. J. Chin. Med. 2024, 52, 1087–1135. [Google Scholar] [CrossRef] [PubMed]
  4. Yang, Y.; Zhu, H.; Yuan, X.; Zhang, X.; Feng, Z.; Jiang, J.; Zhang, P. Seven new prenylated flavanones from the roots of Sophora flavescens and their anti-proliferative activities. Bioorg. Chem. 2021, 109, 104716. [Google Scholar] [CrossRef]
  5. Kwon, M.; Ko, S.-K.; Jang, M.; Kim, G.-H.; Ryoo, I.-J.; Son, S.; Ryu, H.W.; Oh, S.-R.; Lee, W.-K.; Kim, B.Y.; et al. Inhibitory effects of flavonoids isolated from Sophora flavescens on indoleamine 2,3-dioxygenase 1 activity. J. Enzym. Inhib. Med. Chem. 2019, 34, 1481–1488. [Google Scholar] [CrossRef]
  6. Sun, C.-P.; Zhou, J.-J.; Yu, Z.-L.; Huo, X.-K.; Zhang, J.; Morisseau, C.; Hammock, B.D.; Ma, X.-C. Kurarinone alleviated Parkinson’s disease via stabilization of epoxyeicosatrienoic acids in animal model. Proc. Natl. Acad. Sci. USA 2022, 119, e2118818119. [Google Scholar] [CrossRef]
  7. Gonzalez, A.; Zhao, M.; Leavitt, J.M.; Lloyd, A.M. Regulation of the anthocyanin biosynthetic pathway by the TTG1/bHLH/Myb transcriptional complex in Arabidopsis seedlings. Plant J. 2007, 53, 814–827. [Google Scholar] [CrossRef]
  8. Baudry, A.; Caboche, M.; Lepiniec, L. TT8 controls its own expression in a feedback regulation involving TTG1 and homologous MYB and bHLH factors, allowing a strong and cell-specific accumulation of flavonoids in Arabidopsis thaliana. Plant J. 2006, 46, 768–779. [Google Scholar] [CrossRef]
  9. Liu, A.; Lu, J.; Song, H.; Wang, X.; Wang, M.; Lei, Z.; Liu, H.; Lei, H.; Niu, T. Comparative genomics and transcriptomics analysis of the bHLH gene family indicate their roles in regulating flavonoid biosynthesis in Sophora flavescens. Front. Plant Sci. 2024, 15, 1445488. [Google Scholar] [CrossRef]
  10. Mondal, A.; Gandhi, A.; Fimognari, C.; Atanasov, A.G.; Bishayee, A. Alkaloids for cancer prevention and therapy: Current progress and future perspectives. Eur. J. Pharmacol. 2019, 858, 172472. [Google Scholar] [CrossRef]
  11. Bhambhani, S.; Kondhare, K.R.; Giri, A.P. Diversity in chemical structures and biological properties of plant alkaloids. Molecules 2021, 26, 3374. [Google Scholar] [CrossRef] [PubMed]
  12. Wang, W.; You, R.; Qin, W.; Hai, L.; Fang, M.; Huang, G.; Kang, R.; Li, M.; Qiao, Y.; Li, J.; et al. Anti-tumor activities of active ingredients in Compound Kushen Injection. Acta Pharmacol. Sin. 2015, 36, 676–679. [Google Scholar] [CrossRef] [PubMed]
  13. Cao, J.; Wei, R.; Yao, S. Matrine has pro-apoptotic effects on liver cancer by triggering mitochondrial fission and activating Mst1-JNK signalling pathways. J. Physiol. Sci. 2019, 69, 185–198. [Google Scholar] [CrossRef]
  14. Wu, J.; Cai, Y.; Li, M.; Zhang, Y.; Li, H.; Tan, Z. Oxymatrine promotes S-phase arrest and inhibits cell proliferation of human breast cancer cells in vitro through mitochondria-mediated apoptosis. Biol. Pharm. Bull. 2017, 40, 1232–1239. [Google Scholar] [CrossRef]
  15. Li, W.; Yu, X.; Tan, S.; Liu, W.; Zhou, L.; Liu, H. Oxymatrine inhibits non–small cell lung cancer via suppression of EGFR signaling pathway. Cancer Med. 2018, 7, 208–218. [Google Scholar] [CrossRef]
  16. Sun, P.; Zhao, W.; Wang, Q.; Chen, L.; Sun, K.; Zhan, Z.; Wang, J. Chemical diversity, biological activities and Traditional uses of and important Chinese herb Sophora. Phytomedicine 2022, 100, 154054. [Google Scholar] [CrossRef] [PubMed]
  17. Qu, Z.; Wang, W.; Adelson, D.L. Chromosomal level genome assembly of medicinal plant Sophora flavescens. Sci. Data 2023, 10, 572. [Google Scholar] [CrossRef]
  18. Godbole, R.C.; Pable, A.A.; Singh, S.; Barvkar, V.T. Interplay of transcription factors orchestrating the biosynthesis of plant alkaloids. 3 Biotech 2022, 12, 250. [Google Scholar] [CrossRef]
  19. Yang, L.; Yang, Y.; Huang, L.; Cui, X.; Liu, Y. From single- to multi-omics: Future research trends in medicinal plants. Brief. Bioinform. 2023, 24, bbac485. [Google Scholar] [CrossRef]
  20. Li, H.; Lv, Q.; Liu, A.; Wang, J.; Sun, X.; Deng, J.; Chen, Q.; Wu, Q. Comparative metabolomics study of Tartary (Fagopyrum tataricum (L.) Gaertn) and common (Fagopyrum esculentum Moench) buckwheat seeds. Food Chem. 2022, 371, 131125. [Google Scholar] [CrossRef]
  21. Waris, M.; Koçak, E.; Gonulalan, E.M.; Demirezer, L.O.; Kır, S.; Nemutlu, E. Metabolomics analysis insight into medicinal plant science. TrAC Trend Anal. Chem. 2022, 157, 116795. [Google Scholar] [CrossRef]
  22. Lei, H.; Niu, T.; Song, H.; Bai, B.; Han, P.; Wang, Z.; Liu, A. Comparative transcriptome profiling reveals differentially expressed genes involved in flavonoid biosynthesis between biennial and triennial Sophora flavescens. Ind. Crop Prod. 2021, 161, 113217. [Google Scholar] [CrossRef]
  23. Li, N.; Dong, Y.; Lv, M.; Qian, L.; Sun, X.; Liu, L.; Cai, Y.; Fan, H. Combined analysis of volatile terpenoid metabolism and transcriptome reveals transcription factors related to terpene synthase in two cultivars of Dendrobium officinale flowers. Front. Genet. 2021, 12, 661296. [Google Scholar] [CrossRef] [PubMed]
  24. Li, C.; Wood, J.C.; Vu, A.H.; Hamilton, J.P.; Rodriguez Lopez, C.E.; Payne, R.M.E.; Serna Guerrero, D.A.; Gase, K.; Yamamoto, K.; Vaillancourt, B.; et al. Single-cell multi-omics in the medicinal plant Catharanthus roseus. Nat. Chem. Biol. 2023, 19, 1031–1041. [Google Scholar] [CrossRef]
  25. Fraga, C.G.; Clowers, B.H.; Moore, R.J.; Zink, E.M. Signature-discovery approach for sample matching of a nerve-agent precursor using liquid chromatography−mass spectrometry, XCMS, and chemometrics. Anal. Chem. 2010, 82, 4165–4173. [Google Scholar] [CrossRef]
  26. Chen, S.; Zhou, Y.; Chen, Y.; Gu, J. fastp: An ultra-fast all-in-one FASTQ preprocessor. Bioinformatics 2018, 34, i884–i890. [Google Scholar] [CrossRef] [PubMed]
  27. Kim, D.; Paggi, J.M.; Park, C.; Bennett, C.; Salzberg, S.L. Graph-based genome alignment and genotyping with HISAT2 and HISAT-genotype. Nat. Biotechnol. 2019, 37, 907–915. [Google Scholar] [CrossRef]
  28. Liao, Y.; Smyth, G.K.; Shi, W. featureCounts: An efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics 2014, 30, 923–930. [Google Scholar] [CrossRef]
  29. Wang, L.; Feng, Z.; Wang, X.; Wang, X.; Zhang, X. DEGseq: An R package for identifying differentially expressed genes from RNA-seq data. Bioinformatics 2010, 26, 136–138. [Google Scholar] [CrossRef]
  30. Langfelder, P.; Horvath, S. WGCNA: An R package for weighted correlation network analysis. BMC Bioinform. 2008, 9, 559. [Google Scholar] [CrossRef]
  31. Otasek, D.; Morris, J.H.; Bouças, J.; Pico, A.R.; Demchak, B. Cytoscape Automation: Empowering workflow-based network analysis. Genome Biol. 2019, 20, 185. [Google Scholar] [CrossRef] [PubMed]
  32. He, X.; Fang, J.; Huang, L.; Wang, J.; Huang, X. Sophora flavescens Ait.: Traditional usage, phytochemistry and pharmacology of an important traditional Chinese medicine. J. Ethnopharmacol. 2015, 172, 10–29. [Google Scholar] [CrossRef] [PubMed]
  33. Huang, X.B.; Yuan, L.W.; Shao, J.; Yang, Y.; Liu, Y.; Lu, J.J.; Chen, L. Cytotoxic effects of flavonoids from root of Sophora flavescens in cancer cells. Nat. Prod. Res. 2021, 35, 4317–4322. [Google Scholar] [CrossRef]
  34. Kim, C.Y.; Kim, H.J.; Kim, K.M.; Oak, M.H. Vasorelaxant prenylated flavonoids from the roots of Sophora flavescens. Biosci. Biotechnol. Biochem. 2013, 77, 395–397. [Google Scholar] [CrossRef]
  35. Zhang, L.; Xu, L.; Xiao, S.S.; Liao, Q.F.; Li, Q.; Liang, J.; Chen, X.H.; Bi, K.S. Characterization of flavonoids in the extract of Sophora flavescens Ait. by high-performance liquid chromatography coupled with diode-array detector and electrospray ionization mass spectrometry. J. Pharm. Biomed. Anal. 2007, 44, 1019–1028. [Google Scholar] [CrossRef]
  36. Kim, H.; Moon, J.Y.; Burapan, S.; Han, J.; Cho, S.K. Induction of ER stress-mediated apoptosis by the major component 5,7,4′-trimethoxyflavone isolated from kaempferia parviflora tea infusion. Nutr. Cancer 2018, 70, 984–996. [Google Scholar] [CrossRef]
  37. Fu, Y.; Fang, Y.; Gong, S.; Xue, T.; Wang, P.; She, L.; Huang, J. Deep learning-based network pharmacology for exploring the mechanism of licorice for the treatment of COVID-19. Sci. Rep. 2023, 13, 5844. [Google Scholar] [CrossRef]
  38. Liu, T.; Gong, J.; Lai, G.; Yang, Y.; Wu, X.; Wu, X. Flavonoid extract Kushenol a exhibits anti-proliferative activity in breast cancer cells via suppression of PI3K/AKT/mTOR pathway. Cancer Med. 2022, 12, 1643–1654. [Google Scholar] [CrossRef] [PubMed]
  39. Chen, M.H.; Gu, Y.Y.; Zhang, A.L.; Sze, D.M.; Mo, S.L.; May, B.H. Biological effects and mechanisms of matrine and other constituents of Sophora flavescens in colorectal cancer. Pharmacol. Res. 2021, 171, 105778. [Google Scholar] [CrossRef]
  40. Liu, S.; Meng, Z.; Zhang, H.; Chu, Y.; Qiu, Y.; Jin, B.; Wang, L. Identification and characterization of thirteen gene families involved in flavonoid biosynthesis in Ginkgo biloba. Ind. Crop Prod. 2022, 188, 115576. [Google Scholar] [CrossRef]
  41. Daryanavard, H.; Postiglione, A.E.; Mühlemann, J.K.; Muday, G.K. Flavonols modulate plant development, signaling, and stress responses. Curr. Opin. Plant Biol. 2023, 72, 102350. [Google Scholar] [CrossRef] [PubMed]
  42. Shen, N.; Wang, T.; Gan, Q.; Liu, S.; Wang, L.; Jin, B. Plant flavonoids: Classification, distribution, biosynthesis, and antioxidant activity. Food Chem. 2022, 383, 132531. [Google Scholar] [CrossRef]
  43. Nabavi, S.M.; Šamec, D.; Tomczyk, M.; Milella, L.; Russo, D.; Habtemariam, S.; Suntar, I.; Rastrelli, L.; Daglia, M.; Xiao, J.; et al. Flavonoid biosynthetic pathways in plants: Versatile targets for metabolic engineering. Biotechnol. Adv. 2020, 38, 107316. [Google Scholar] [CrossRef]
  44. Ma, W.; Xu, L.; Gao, S.; Lyu, X.; Cao, X.; Yao, Y. Melatonin alters the secondary metabolite profile of grape berry skin by promoting VvMYB14-mediated ethylene biosynthesis. Horticul. Res. 2021, 8, 43. [Google Scholar] [CrossRef] [PubMed]
  45. Yin, Y.; Tian, X.; Yang, J.; Yang, Z.; Tao, J.; Fang, W. Melatonin mediates isoflavone accumulation in germinated soybeans (Glycine max L.) under ultraviolet-B stress. Plant Physiol. Biochem. 2022, 175, 23–32. [Google Scholar] [CrossRef] [PubMed]
  46. Kianersi, F.; Abdollahi, M.R.; Mirzaie-asl, A.; Dastan, D.; Rasheed, F. Identification and tissue-specific expression of rutin biosynthetic pathway genes in Capparis spinosa elicited with salicylic acid and methyl jasmonate. Sci. Rep. 2020, 10, 8884. [Google Scholar] [CrossRef]
  47. Hao, Y.; Xiang, L.; Lai, J.; Li, C.; Zhong, Y.; Ye, W.; Yang, J.; Yang, J.; Wang, S. SlERF.H6 mediates the orchestration of ethylene and gibberellin signaling that suppresses bitter-SGA biosynthesis in tomato. New Phytol. 2023, 239, 1353–1367. [Google Scholar] [CrossRef]
  48. Yamada, Y.; Sato, F. Transcription factors in alkaloid biosynthesis. In International Review of Cell and Molecular Biology; Jeon, K.W., Ed.; Academic Press: San Diego, CA, USA, 2013; pp. 339–382. [Google Scholar]
  49. Todd, A.T.; Liu, E.; Polvi, S.L.; Pammett, R.T.; Page, J.E. A functional genomics screen identifies diverse transcription factors that regulate alkaloid biosynthesis in Nicotiana benthamiana. Plant J. 2010, 62, 589–600. [Google Scholar] [CrossRef]
  50. Tian, X.; Liu, C.; Yang, Z.; Zhu, J.; Fang, W.; Yin, Y. Crosstalk between ethylene and melatonin activates isoflavone biosynthesis and antioxidant systems to produce high-quality soybean sprouts. Plant Sci. 2024, 347, 112197. [Google Scholar] [CrossRef] [PubMed]
  51. Yamada, Y.; Sato, F. Transcription factors in alkaloid engineering. Biomolecules 2021, 11, 1719. [Google Scholar] [CrossRef]
  52. Hichri, I.; Heppel, S.C.; Pillet, J.; Léon, C.; Czemmel, S.; Delrot, S.; Lauvergeat, V.; Bogs, J. The basic helix-loop-helix transcription factor MYC1 is involved in the regulation of the flavonoid biosynthesis pathway in grapevine. Mol. Plant 2010, 3, 509–523. [Google Scholar] [CrossRef] [PubMed]
  53. Wada, T.; Kunihiro, A.; Tominaga-Wada, R. Arabidopsis CAPRICE (MYB) and GLABRA3 (bHLH) control tomato (Solanum lycopersicum) anthocyanin biosynthesis. PLoS ONE 2014, 9, e109093. [Google Scholar] [CrossRef] [PubMed]
  54. Cao, Y.; Mei, Y.; Zhang, R.; Zhong, Z.; Yang, X.; Xu, C.; Chen, K.; Li, X. Transcriptional regulation of flavonol biosynthesis in plants. Horticul. Res. 2024, 11, uhae043. [Google Scholar] [CrossRef] [PubMed]
Agronomy 15 01455 g001
Figure 1. Schematic diagram of S. flavescens and the sampled tissues. (A) Schematic of S. flavescens plant. (B) Images of different S. flavescens tissues.
Figure 1. Schematic diagram of S. flavescens and the sampled tissues. (A) Schematic of S. flavescens plant. (B) Images of different S. flavescens tissues.
Agronomy 15 01455 g001
Agronomy 15 01455 g002
Figure 2. Overview of transcriptome and metabolome. (A) Flavonoid and alkaloid accumulation in S. flavescens. (B) PCA of metabolome and transcriptome. (C) Correlation analysis between any two different samples. The upper right corner represents the transcriptome correlation between samples, and the lower left corner represents the metabolome correlation between samples. (D) The type comparison of metabolites with accumulated content detected among different tissues. (E) Comparative analysis of gene expression levels among the five tissues. (F) Flower Venn diagram of expressed genes (FPKM > 1) among the five tissues.
Figure 2. Overview of transcriptome and metabolome. (A) Flavonoid and alkaloid accumulation in S. flavescens. (B) PCA of metabolome and transcriptome. (C) Correlation analysis between any two different samples. The upper right corner represents the transcriptome correlation between samples, and the lower left corner represents the metabolome correlation between samples. (D) The type comparison of metabolites with accumulated content detected among different tissues. (E) Comparative analysis of gene expression levels among the five tissues. (F) Flower Venn diagram of expressed genes (FPKM > 1) among the five tissues.
Agronomy 15 01455 g002
Agronomy 15 01455 g003
Figure 3. Analysis of DAMs and DEGs among the four comparisons. (A) DAF and DAA numbers in the four comparisons. (B) Comparison of DEG numbers among the four comparisons. (C) DEG numbers categorized by FC: white numbers in the inner circle, black numbers in the outer circle, and numbers outside the circle indicate DEGs with ≥8-FC, ≥4-, <8-FC, and <4-FC in roots compared with other tissues. (D) Comparison of DAM numbers among the four comparisons. (E) Expression profiles of 10 SM biosynthesis-related genes via RT-qPCR technology. The vertical bars followed the different letters (a, b or c) were significantly differences according to ANOVA with Duncan’s test (p < 0.05). (F) The counts of significantly enriched GO terms for DEGs across different comparison groups. (G) The counts of significantly enriched KEGG pathways for DEGs in different comparison groups. (H) The significantly enriched KEGG pathways among the four comparison groups.
Figure 3. Analysis of DAMs and DEGs among the four comparisons. (A) DAF and DAA numbers in the four comparisons. (B) Comparison of DEG numbers among the four comparisons. (C) DEG numbers categorized by FC: white numbers in the inner circle, black numbers in the outer circle, and numbers outside the circle indicate DEGs with ≥8-FC, ≥4-, <8-FC, and <4-FC in roots compared with other tissues. (D) Comparison of DAM numbers among the four comparisons. (E) Expression profiles of 10 SM biosynthesis-related genes via RT-qPCR technology. The vertical bars followed the different letters (a, b or c) were significantly differences according to ANOVA with Duncan’s test (p < 0.05). (F) The counts of significantly enriched GO terms for DEGs across different comparison groups. (G) The counts of significantly enriched KEGG pathways for DEGs in different comparison groups. (H) The significantly enriched KEGG pathways among the four comparison groups.
Agronomy 15 01455 g003
Agronomy 15 01455 g004
Figure 4. K-means clustering of DEGs significantly correlated with DAM content. (A–C) By utilizing the profiles of DAM accumulation and DEG expression, the correlation was calculated and generated a k-means plot. Blue indicates the metabolome k-means, red shows DEGs positively correlated with the metabolome, and green shows DEGs negatively correlated with the metabolome (|r| > 0.9, p < 0.05). (D) TFs significantly correlated with flavonoid metabolites and their classification. (E) TFs significantly correlated with alkaloid metabolites and their classification. (F) Network diagram of the 7 most significant TFs related to flavonoids and alkaloids.
Figure 4. K-means clustering of DEGs significantly correlated with DAM content. (A–C) By utilizing the profiles of DAM accumulation and DEG expression, the correlation was calculated and generated a k-means plot. Blue indicates the metabolome k-means, red shows DEGs positively correlated with the metabolome, and green shows DEGs negatively correlated with the metabolome (|r| > 0.9, p < 0.05). (D) TFs significantly correlated with flavonoid metabolites and their classification. (E) TFs significantly correlated with alkaloid metabolites and their classification. (F) Network diagram of the 7 most significant TFs related to flavonoids and alkaloids.
Agronomy 15 01455 g004
Agronomy 15 01455 g005
Figure 5. Gene co-expressed modules and their metabolite associations. (A) Heatmap of 27 modules showing expression trends across five tissues. (B) Association of some important flavonoids or alkaloids with modules. (C) Enriched KEGG pathways of genes in flavonoid- or alkaloid-correlated modules. (D) Comparison of genes in 7 significantly correlated modules in S. flavescens with DEGs across four comparisons (Stem vs. Root, Leaf vs. Root, Flower vs. Root and Pod vs. Root).
Figure 5. Gene co-expressed modules and their metabolite associations. (A) Heatmap of 27 modules showing expression trends across five tissues. (B) Association of some important flavonoids or alkaloids with modules. (C) Enriched KEGG pathways of genes in flavonoid- or alkaloid-correlated modules. (D) Comparison of genes in 7 significantly correlated modules in S. flavescens with DEGs across four comparisons (Stem vs. Root, Leaf vs. Root, Flower vs. Root and Pod vs. Root).
Agronomy 15 01455 g005
Agronomy 15 01455 g006
Figure 6. Regulation of flavonoid biosynthesis in S. flavescens. (A) Schematic of the flavonoid biosynthesis pathway in S. flavescens. (B) The relative content distribution of flavonoids related to the proposed metabolic pathway. (C) Correlation network diagram of flavonoids and their key regulatory TFs related to the proposed metabolic pathway. *, ** and *** indicate the significantly difference at p < 0.05, 0.01 and 0.001, respectively. (D) Network diagram of other important flavonoids and their key regulatory TFs in S. flavescens.
Figure 6. Regulation of flavonoid biosynthesis in S. flavescens. (A) Schematic of the flavonoid biosynthesis pathway in S. flavescens. (B) The relative content distribution of flavonoids related to the proposed metabolic pathway. (C) Correlation network diagram of flavonoids and their key regulatory TFs related to the proposed metabolic pathway. *, ** and *** indicate the significantly difference at p < 0.05, 0.01 and 0.001, respectively. (D) Network diagram of other important flavonoids and their key regulatory TFs in S. flavescens.
Agronomy 15 01455 g006
Agronomy 15 01455 g007
Figure 7. Biosynthesis and regulation of quinolizidine alkaloids (QAs) in S. flavescens. (A) Illustrative representation of the biosynthetic pathway of QAs in S. flavescens. (B) The relative content distribution of detected QAs. (C) Network diagram of CuAO02, QAs, and related regulatory TFs in S. flavescens. (D) Network diagram of LDC, CuAO08, CuAO09 and CuAO10, QAs, and related regulatory TFs in S. flavescens. (E) Correlation network diagram of QAs and key regulatory TFs related to the metabolic pathway. *, ** and *** indicate the significantly difference at p < 0.05, 0.01 and 0.001, respectively.
Figure 7. Biosynthesis and regulation of quinolizidine alkaloids (QAs) in S. flavescens. (A) Illustrative representation of the biosynthetic pathway of QAs in S. flavescens. (B) The relative content distribution of detected QAs. (C) Network diagram of CuAO02, QAs, and related regulatory TFs in S. flavescens. (D) Network diagram of LDC, CuAO08, CuAO09 and CuAO10, QAs, and related regulatory TFs in S. flavescens. (E) Correlation network diagram of QAs and key regulatory TFs related to the metabolic pathway. *, ** and *** indicate the significantly difference at p < 0.05, 0.01 and 0.001, respectively.
Agronomy 15 01455 g007
Table 1. Tissue-specific metabolites among five tissues in S. flavescens.
Table 1. Tissue-specific metabolites among five tissues in S. flavescens.
IndexMolecular Weight (Da)FormulaIonization ModelCompoundsClassCAS
Root
pmp000106312.1C18H16O5[M+H]+5,7,4′-TrimethoxyflavoneFlavones5631-70-9
pmp000350336.1C20H16O5[M+H]+GlabroneIsoflavones60008-02-8
pmp000636340.131C20H20O5[M+H]+Kushenol SFlavanones254886-72-1
Wmkn004416340.131C20H20O5[M-H]−DesmethylxanthohumolChalcones115063-39-3
pmp000362368.126C21H20O6[M+H]+GlisoflavoneIsoflavones125709-32-2
Wmkn003942370.142C21H22O6[M-H]−2′-HydroxyIsoxanthohumolFlavanones-
MWSHY0070408.194C25H28O5[M+H]+Kushenol AFlavanones99217-63-7
pmp000647424.189C25H28O6[M+H]+Kushenol EFlavanones99119-72-9
pmp000650438.204C26H30O6[M+H]+Leachianone AFlavanones97938-31-3
pmp000652438.204C26H30O6[M+H]+Kushenol DChalcones-
pmp000651438.204C26H30O6[M+H]+Kushenol CaOther Flavonoids-
Wmkn004424440.183C25H28O7[M-H]−Kushenol XFlavanones254886-77-6
pmp000656452.22C27H32O6[M+H]+Kushenol DaOther Flavonoids-
Wmkn004385452.22C27H32O6[M-H]−5-Methylmatrine CFlavonols-
pmp000657454.199C26H30O7[M+H]+KurarinolFlavanones855746-98-4
Wmkn004208454.199C26H30O7[M-H]−Kushenol IFlavanones99119-69-4
Zmkp005880470.231C27H34O7[M+H]+NeokurarinolFlavanones52483-00-8
MWSHY0044438.204C26H30O6[M+H]+KurarinoneFlavanones34981-26-5
MWSmce366452.22C27H32O6[M+H]+2′-MethoxykurarinoneFlavanones270249-38-2
Wmkn004322440.183C25H28O7[M-H]−Kushenol LFlavanonols101236-50-4
MWSHY0065290.079C15H14O6[M+H]+CatechinFlavanols154-23-4
Leaf
Zmdn002049452.132C21H24O11[M-H]−Epicatechin-4′-O-β-D-glucopyranosideFlavanols-
Flower
Zmxp004503654.18C29H34O17[M+H]+Tricin-5,7-O-diglucosideFlavones-
mws0895432.106C21H20O10[M-H]−Genistein-7-O-Glucoside (Genistin)Isoflavones529-59-9
Pod
Zmdp004370458.121C23H22O10[M+H]+6′′-O-AcetyldaidzinIsoflavones71385-83-6
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Liu, A.; Dong, J.; Wang, M.; Li, J.; Lu, J.; Liu, L.; Lei, H.; Zeng, Z.; Song, H. Integrated Transcriptomic and Metabolomic Analysis Identified Key Transcriptional Factors Involved in Flavonoid and Alkaloid Biosynthesis Among Different Tissues of Sophora flavescens. Agronomy 2025, 15, 1455. https://doi.org/10.3390/agronomy15061455

AMA Style

Liu A, Dong J, Wang M, Li J, Lu J, Liu L, Lei H, Zeng Z, Song H. Integrated Transcriptomic and Metabolomic Analysis Identified Key Transcriptional Factors Involved in Flavonoid and Alkaloid Biosynthesis Among Different Tissues of Sophora flavescens. Agronomy. 2025; 15(6):1455. https://doi.org/10.3390/agronomy15061455

Chicago/Turabian Style

Liu, Ake, Jingjing Dong, Mingyang Wang, Jin Li, Junjie Lu, Lintao Liu, Haiying Lei, Zhen Zeng, and Huifang Song. 2025. "Integrated Transcriptomic and Metabolomic Analysis Identified Key Transcriptional Factors Involved in Flavonoid and Alkaloid Biosynthesis Among Different Tissues of Sophora flavescens" Agronomy 15, no. 6: 1455. https://doi.org/10.3390/agronomy15061455

APA Style

Liu, A., Dong, J., Wang, M., Li, J., Lu, J., Liu, L., Lei, H., Zeng, Z., & Song, H. (2025). Integrated Transcriptomic and Metabolomic Analysis Identified Key Transcriptional Factors Involved in Flavonoid and Alkaloid Biosynthesis Among Different Tissues of Sophora flavescens. Agronomy, 15(6), 1455. https://doi.org/10.3390/agronomy15061455

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop