Integrated transcriptomic and proteomic analysis reveals inhibitory effects of naringenin application on soybean roots mainly by interfering with auxin and ROS concentrations

Abstract

Background

Flavonoids play a crucial role as secondary metabolites in plant growth and development. Naringenin is a central intermediate in flavonoid biosynthesis. It is also a pivotal branching point that gives rise to metabolites such as flavonols, anthocyanins, and isoflavonoids. However, the mechanisms underlying plant root development regulated by flavonoids remain fragmentary, especially in crops.

Results

In this study, we observed a significant inhibition of soybean (Glycine max) root growth after exposure to 75 μM naringenin. The treated roots showed reduced auxin content and increased reactive oxygen species (ROS) levels in root tips. Through RNA-sequencing and nanoElute high-performance liquid chromatography-tandem mass spectrometry (UHPLC-MS/MS) analyses, we further investigated molecular responses of soybean roots to naringenin treatment. A total of 6,390 differentially expressed genes (DEGs) were identified. Among them, 3,910 genes were up-regulated and 2,480 were down-regulated. These DEGs notably included genes related to ROS metabolism and the auxin signaling pathway. We also detected 806 differential accumulation proteins (DAPs), including 481 up-regulated and 325 down-regulated proteins. The integration of transcriptomic and proteomic data showed a close association between DEGs and DAPs in roots, especially in pathways related to glutathione metabolism.

Conclusion

Taken together, this comprehensive analysis highlights the intricate and specific responses of soybean roots to naringenin exposure, mainly affecting the levels of auxin and ROS. It provides valuable insights and candidate genes for further investigating root growth regulated by flavonoids.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12870-025-08000-9.

Background

Plant roots function in anchorage, uptake, storage, and transport of minerals and water. They are also the first organs that sense changes in the soil environment. Through their roots, plants can interact with the soil microbiome and communicate with neighboring plants [1–4]. Multiple hormones participate in the control of root growth and development, among which auxin plays a central role [5, 6]. In the tap root (TR) apex, a localized auxin maximum is required to establish and maintain the quiescent center (QC) and the stem cell niche (SCN). This auxin peak is essential for proper TR formation [7, 8]. In Arabidopsis, the generation of this auxin maximum mainly depends on the expression of several pin-formed efflux transporters, including PIN1, PIN2, PIN3, PIN4, and PIN7, in the TR [9]. Local auxin levels in the TR tip are precisely modulated by biosynthesis and transport. This regulation occurs through an auxin response factor (ARF10/16) and indole-3-acetic acid (IAA17/AXR3) signaling pathway. Such control suppresses the expression of wuschel related homeobox5 (WOX5) and confines its activity to the QC, thereby ensuring TR growth and patterning [10]. Auxin has also been shown to regulate successive steps of lateral root (LR) organogenesis, including pericycle cell priming, founder cell specification, initiation, primordium formation, and emergence. The priming process requires ARF7 and its auxin-sensitive repressor IAA18 [11]. The stabilized potent/iaa18 protein disrupts auxin oscillation dynamics and produces more founder cells [11]. The establishment of founder cell identity involves the expression of specific genes, such as the auxin-regulated transcription factor GATA23. This identity formation and subsequent asymmetric division rely on IAA28, solitary-root/IAA14 (SLR/IAA14), and shoot hypocotyl2/IAA3 (SHY2/IAA3) [12–14]. Consequently, lateral root primordium (LRP) initiation is nearly eliminated in the slr/iaa14 mutant and the arf7 arf19 double mutant [12, 15]. Downstream of the ARF7/19 module, lateral organ boundaries-domain (LBD) transcription factors contribute to several steps of LR initiation and emergence [16]. For instance, LBD18 and LBD33 directly enhance expansin14 expression and indirectly influence expansin17, both of which are required for cell wall loosening during LR emergence [5].

Accumulating evidence indicates that redox signaling pathways, particularly involving reactive oxygen species (ROS), play a role in root growth and development [6, 17]. ROS is essential for stem cell development in the TP. For example, in Arabidopsis, prohibitin 3 (PHB3) was involved in maintaining the integrity of the SCN by regulating the expression of ethylene response factors (ERF104/114/115) and controlling ROS distribution [18]. In the phb3 mutant, both WOX5 expression and TR elongation are reduced [18]. Enzymes involved in ROS metabolism are also essential for normal LRP development and emergence. For instance, the transcription factor upbeat1 UPB1 and its downstream peroxidase genes are key regulators of the balance between cell proliferation and elongation in the root apical meristem (RAM) [19]. UPB1 is also active in developing LRPs, where it contributes to the regulation of lateral root emergence [19]. In addition, respiratory burst oxidase homolog (RBOH) proteins modulate ROS accumulation in the root tissues overlying the LRP, thereby promoting its emergence [20]. ROS further influence the activities of enzymes such as peroxidases and those involved in lignin biosynthesis. These effects alter cell wall cross-linking and composition, which in turn reshape cell wall rigidity, plasticity, and extensibility [21–23]. A feed-forward interaction between auxin and ROS signaling has also been identified during lateral root (LR) formation. A recent study showed that ROS promote the formation of specific higher-order IAA3 multimers, enhancing their interaction with the co-repressor topless. This strengthened binding leads to a strong repression of auxin-responsive gene expression and ultimately limits LR development [24].

Flavonoids represent the largest group of polyphenols, encompassing more than 8,000 identified metabolites [25, 26]. These compounds share a characteristic diphenylpropane (C6-C3-C6) backbone. Based on the oxygenation pattern of the heterocyclic C ring, flavonoids can be classified into six major subclasses: flavones, flavonols, flavanones, flavanols, anthocyanidins, and isoflavones [27–29]. Naringenin serves as a central intermediate in flavonoid biosynthesis. It functions as a key branching point that gives rise to diverse flavonoid metabolites, such as flavones, flavonols, anthocyanins, and isoflavonoids [25, 26]. Beyond its biosynthetic role, naringenin has been widely applied in the prevention or treatment of several diseases, including metabolic syndrome, cancer, neurological disorders, and cardiovascular and hepatic diseases [30–32]. Meanwhile, studies have demonstrated that the exogenous application of naringenin inhibits roots growth across various plant species [33–35]. For example, exogenous addition of 100 μM naringenin significantly inhibited roots elongation and gravitropism in Arabidopsis [33]. Conversely, in tt4 mutant seedlings (which accumulate no naringenin), primary root length and lateral root number were increased [33]. In white lupin (Lupinus albus), lateral root density notably decreased with 10 μM naringenin exposure, whereas the lateral root length unchanged [34]. Naringenin is also involved in plant responses to diverse biotic and abiotic stresses. For example, in soybean roots, naringenin levels significantly increased under salt stress and phosphate starvation [36, 37]. Additionally, phosphate starvation stimulated the secretion of naringenin in soybean roots, thereby facilitating the recruitment and colonization of rhizobia [38, 39].

Soybeans are legumes that are major sources of protein and oil for humans [40]. The latest data show that global soybean production and area harvested was 390.7 million metric tons and 146.9 hectares in 2023 respectively and is projected to increase yearly (http://faostat.fao.org/, accessed on 16 September 2025). While naringenin’s roles in Arabidopsis root development are partially characterized, its impact on crop species like soybean where root architecture directly influences yield and resource use efficiency remains unexplored. In this study, we investigated how naringenin application inhibited soybean root growth though meditating auxin and ROS concentration in soybean roots. Next, we performed a comprehensive transcriptome and proteome analysis of naringenin treated soybean roots to identify differentially expressed genes and differential accumulation proteins (DEGs/DAPs). Interestingly, DEGs and DAPs related to the auxin signaling pathway, ROS metabolism and glycolysis/gluconeogenesis were identified. The results of this study will help us understand how naringenin affects soybean roots growth and development, and provide a reference for the rational application of naringenin.

Results

Applied naringenin inhibited soybean root growth and auxin accumulation, but increased ROS accumulation

To ascertain the function of naringenin, wild type (WT) soybean roots were exposed to different concentrations of naringenin (i.e. 10 μM, 50 μM, 75 μM and 100 μM) (Fig. S1). As the concentration of naringin increases, the inhibition of soybean roots growth and development also intensifies. However, at naringenin concentrations of 75 μM or 100 μM, no significant change in the level of inhibition is observed (Fig. S1A-C). Therefore, for following studies, we utilized 75 μM naringenin.

To further investigate the impact of naringenin application on soybean root growth and auxin distribution, WT (Fig. 1A-E) and transgenic soybean plants harboring DR5:GUS (Fig. 1F) were exposed to 75 μM naringenin (+ Nar) for 4 d. Naringenin notably suppressed soybean root growth, as reflected by decreases of root fresh weight, root length and lateral root number (Fig. 1A). Specifically, root fresh weight, tap root length, lateral root number and total root length were significantly decreased by 33.2%, 34.5%, 59.1%, and 41.2%, respectively (Fig. 1B-E). In order to examine auxin accumulation affected by naringenin, we also detected GUS activity in soybean harboring DR5:GUS with or without naringenin application. It was found that GUS activity was decreased in both tap root and lateral root tips supplied with naringenin (Fig. 1F), strongly suggesting application of naringenin could inhibit auxin accumulation in root tips.

Fig. 1.

Effects of naringenin application on soybean root growth. A Phenotypes of soybean root system. Soybean seedlings were supplied with 75 μM naringenin (+ Nar). After 4 d, roots were analyzed. Mock represents that seedlings were supplied with nutrient solution without naringenin. Scale bar is 2 cm. B Root fresh weight. C Tap root length. D Lateral root number. E Total root length. Data are means of 5 replicates ± SE. Asterisks indicate significant differences between Mock and after 75 μM naringenin treatments in Student’s t-test (** P < 0.01; *** P < 0.001). F GUS activity in lateral root tips and tap root tips of transgenic soybean with DR5:GUS. Scale bars are100 μm

In order to test whether ROS is accumulated by the treatment with naringenin, we performed histochemical staining with DAB, NBT and H₂DCF-DA to monitor the production of H₂O₂, O₂^– and ROS, respectively. Results showed that exogenously adding naringenin significantly increased the DAB, NBT and H₂DCF-DA staining in both tap root and lateral root tips, respectively (Fig. 2A-D).

Fig. 2.

Effects of naringenin application on soybean root ROS levels. DAB A, NBT B and H₂DCF-DA C staining of soybean root tips. Bars, 1 mm in DAB/NBT, and 200 μm in H₂DCF-DA. DAB: 3, 3′-diaminobenzidine; NBT: nitro blue tetrazolium; H₂DCF-DA: 2′,7′-dichlorofluorescein diacetate. Quantification of the DAB staining D, NBT staining E and DCF autofluorescence F in soybean root tips. DAB, NBT and DCF intensity were normalized, the DAB, NBT and DCF intensity of Mock treatments were set to 1. Data are means of 3 replicates ± SE. Asterisks indicate significant differences between Mock and after 75 μM naringenin treatments (+ Nar) in Student’s t-test (* P < 0.05; ** P < 0.01; *** P < 0.001)

Identification of DEGs in soybean roots

To elucidate molecular responses underlying naringenin application on root growth, we conducted RNA-seq assays of soybean roots with or without naringenin application. The principal component analysis (PCA) showed a significant separation between Mock and + Nar application group (Fig. S2A). PCA1 and PCA2 explained 52.7% and 8.6% of the variability, respectively (Fig. S2A). A total of 3,910 up-regulated and 2,480 down-regulated DEGs were identified (Fig. S2B, C and Table S1).

GO enrichment analysis indicated that DEGs were primarily enriched in 174 biological processes, 10 cellular components, and 86 molecular functions (Table S2). Notable GO classes included cellular responses to oxygen levels, cellular response to hypoxia. For molecular function, DEGs mainly involved peroxidase activity, glucosyltransferase activity, oxidoreductase activity, glutathione transferase activity, and antioxidant activity (Fig. S3A and Table S2). Additionally, KEGG assays showed that DEGs were primarily involved in a set of pathways, including plant-pathogen interaction, plant hormone signal transduction, and glutathione metabolism (Fig. S3B and Table S3).

DEGs were involved in plant hormone signal transduction

Totally, 306 DEGs were identified to participate in plant hormone signal transduction (Fig. S3B and Table S1). Using the STRING database, a co-expression network was constructed, mapping 50 DEGs to an interaction network, with nodes representing genes and edges indicating gene–gene associations (Fig. 3A). Among them, an indole-3-acetic acid-amido synthetase, GmGH3.4 was the top hub node with tightly interacting with other 20 DEGs (Fig. 3A), which was followed by GmDELLA3 encoding DELLA protein, Glyma.05G119500 encoding brassinosteroid insensitive receptor kinase, Glyma.17G113900 and Glyma.13G166200 encoding leucine-rich repeat domain superfamily (Fig. 3A). Meanwhile, GmAUX28-4 encoding auxin-induced protein, and GmLAX1/3 encoding auxin influx carrier were also included in the network. (Fig. 3A). Furthermore, a group of DEGs related to auxin signaling pathway were identified, including 4 GmGHs, 3 GmLAXs, 4 GmAUXs,4 GmPINs, 2 auxin response factors (GmARFs), 1 auxin-induced protein (GmIAA), 17 small auxin up RNA (GmSAURs) (Fig. 3B). To confirm the reliability of RNA-seq, transcripts of 5 DEGs was investigated by qRT-PCR analysis, including GmGH3.1, GmGH3.4, GmPIN4, GmSAUR8, and GmSAUR11. It was found their expression patterns were consistent with results from transcriptome analysis (Fig. 3C).

Fig. 3.

Analysis of DEGs related to plant hormone signal transduction in soybean roots under 75 μM naringenin treatment (+ Nar). A Co-expression network of DEGs related to plant hormone signal transduction. The color shade and size of the circles represent the degrees of connection with the other genes. B Heatmap analysis of DEGs related to auxin signaling pathways. C qRT-PCR analysis of DEGs. Data are means of 5 replicates ± SE. Asterisks indicate significant differences between Mock and after 75 μM naringenin treatments in Student’s t-test (* P < 0.05; ** P < 0.01; *** P < 0.001)

DEGs related to glutathione metabolism

A co-expression network analysis was conducted on 48 DEGs related to glutathione metabolism, mapping 40 DEGs to an interaction network (Fig. 4A). 2 glutathione hydrolases (GmGAGT1/2), and 3 glutathione peroxidases (GmGPX1/2/3) were the core hub nodes of the co-expression network (Fig. 4A). 31 glutathione transferases (GmGSTs) were included in the network. Additionally, 37 DEGs were GmGSTs with only GmGST17 down-regulated and the other up-regulated (Fig. 4B). Meanwhile, transcripts of 5 DEGs in this group were investigated by qRT-PCR analysis, including GmGAGT1, GmGPX2, GmGST2, GmGST4, and GmGST5. Their consistent expression patterns with the transcriptomic analysis were found (Fig. 4C).

Fig. 4.

Analysis of DEGs related to glutathione metabolism in soybean roots under 75 μM naringenin treatment (+ Nar). A Co-expression network of DEGs related to glutathione metabolism. The color shade and size of the circles represent the degrees of connection with the other genes. B Heatmap analysis of DEGs related to glutathione metabolism. C qRT-PCR analysis of DEGs. Data are means of 5 replicates ± SE. Asterisks indicate significant differences between Mock and after 75 μM naringenin treatments in Student’s t-test (* P < 0.05; ** P < 0.01; *** P < 0.001)

Identification of DAPs in soybean root

Proteomic analysis was subsequently conducted to investigate global protein profiles in soybean roots responsive to naringenin application. PCA showed a significant separation between Mock and + Nar group (Fig. S6A). PCA1 and PCA2 explained 42.4% and 18.1% of the variability, respectively (Fig. S6A). Totally, 12,056 unique proteins were identified in soybean roots. Out of these, 806 proteins were found to exhibit differential accumulation proteins (DAPs) with 481 up-regulated and 325 down-regulated proteins (Fig. S6B, C and Table S4).

GO enrichment analysis indicated that DAPs primarily enriched in 7 biological processes, including glutathione metabolic process, response to chemical, response to oxidative stress, hydrogen peroxide catabolic process, anatomical structure morphogenesis, lipid oxidation, oxylipin biosynthetic process. Molecular functions were mainly associated with extracellular region, plant-type cell wall, and extracellular space. Notable cellular components mainly included glutathione transferase activity, UDP-glycosyltransferase activity, heme binding, peroxidase activity, ABC-type transporter activity, and oxidoreductase activity (Fig. S7A and Table S5). Meanwhile, KEGG analysis revealed significantly enrichment in glutathione metabolism, phenylpropanoid biosynthesis, ABC transporters, glycolysis/gluconeogenesis, flavonoid biosynthesis, and pyruvate metabolism. (Fig. S7B and Table S6).

Protein–protein interaction network in DAPs

A PPI network was constructed via the STRING database to elucidate DAP interactions. Out of 806 DAPs, 440 were mapped to an interaction network (Fig. S8). We obtained 5 individually separated clusters using the MCODE plug-in in Cytoscape to screen all the DAPs (Fig. 5A). Cluster I comprised 22 proteins, including 2 pyruvate decarboxylases (GmPDC1/2), and 4 pyruvate kinases (GmPK1/2/3/4) related to glycolysis/gluconeogenesis were the core hub nodes of the PPI network (Fig. 5A). Cluster Ⅱ consists of 1 phosphopyruvate hydratase (GmPPH1), 1 phosphoglycerate mutase (GmPGAM1), 1 glucose-6-phosphate 1-dehydrogenase (GmG6PDH2), 1 glucose-6-phosphate isomerase (GmG6PI1), 3 fructose-bisphosphate aldolases (GmFBA1/2/3), which were all associated with glycolysis/gluconeogenesis (Fig. 5A). Among the 29 DAPs related to glycolysis/gluconeogenesis, only the accumulation levels of an aldehyde dehydrogenase (GmALDH2), GmG6PDH1 and GmPK1/2 were decreased by naringenin application (Fig. 5B). The glutathione peroxidase (GmGPX1) and glutathione reductase (GmGR1) related to glutathione metabolism were the core hub nodes of Cluster Ⅳ and Cluster Ⅴ, respectively (Fig. 5A). Meanwhile, 1 peroxiredoxin (GmPRX1) and 5 GmGSTs (GmGST25/40/42/43/44) related to glutathione metabolism were the crucial components of Cluster Ⅳ. The accumulation of all 24 DAPs related to glutathione metabolism were up-regulated (Fig. 5C).

Fig. 5.

Analysis of DAPs in soybean roots under 75 μM naringenin treatment (+ Nar). A Co-expression network of DAPs. The color shade and size of the circles represent the degrees of connection with the other proteins. B Heatmap analysis of DAPs related to glycolysis/gluconeogenesis. C Heatmap analysis of DAPs related to glutathione metabolism

qRT-PCR analysis of genes encoding DAPs related to glycolysis/gluconeogenesis and glutathione metabolism

Based on proteomic analysis, transcripts of 10 genes encoding DAPs related to glycolysis/gluconeogenesis and glutathione metabolism in soybean roots were further analyzed. In terms of glycolysis/gluconeogenesis, the expression of GmADH1, GmPDC1, and GmPK3/4/5 increased by 24.9% to 92.9% (Fig. 6). The expression of GmGPX1, GmGR1, GmGST25, GmGST43, and GmPRX1, which is related to glutathione metabolism, also increased by 52.4% to 77.0% (Fig. 6).

Fig. 6.

qRT-PCR analysis of DAPs. Data are means of 5 replicates ± SE. Asterisks indicate significant differences between Mock and after 75 μM naringenin treatments (+ Nar) in Student’s t-test (* P < 0.05; ** P < 0.01; *** P < 0.001)

Integration of transcriptomic and proteomic analysis

In order to fully elucidate molecular responses of soybean roots to naringenin application, identified DEGs and DAPs were integrated for further analysis. It was found that there were 282 genes or proteins were overlapped (Fig. 7A). Among them, accumulation patterns of 184 up-regulated and 85 down-regulated DAPs exhibited the same pattern as their corresponding genes, while the remaining 13 exhibited opposite patterns (Fig. 7B). KEGG analysis of the overlapped 269 genes and proteins were mainly enriched in metabolic pathways, glutathione metabolism, glycolysis/gluconeogenesis, biosynthesis of secondary metabolites, phenylpropanoid biosynthesis, pyruvate metabolism, ascorbate and aldarate metabolism (Fig. 7C). GO enrichment analysis indicated that the overlapped 269 genes and proteins were primarily enriched in 4 biological processes, which cellular components were associated with plasma membrane, nucleosome, and extracellular space. Molecular functions of the overlapped proteins and genes mainly included endopeptidase inhibitor activity, oxidoreductase activity, aldehyde dehydrogenase activity, ATPase-coupled transmembrane transporter activity, naringenin-chalcone synthase activity, chalcone synthase activity, ABC-type transporter activity, peroxidase activity, and glutathione transferase activity (Fig. 7D). Particularly, 29 and 24 overlapped genes and proteins were found to participate in glycolysis/gluconeogenesis, and glutathione metabolism, respectively. Furthermore, 28 out 29 overlapped genes and proteins were up-regulated (Fig. 8A). While, 24 overlapped genes and proteins associated with glutathione metabolism were notably up-regulated (Fig. 8B).

Fig. 7.

Combined analysis of transcriptome and proteome of soybean roots after 75 μM naringenin treatment (+ Nar). A Venn diagrams of co-regulated between DEGs and DAPs. B Quadrant distribution diagram of coexpression between DEGs and DAPs. C KEGG enrichment analysis. D GO enrichment analysis

Fig. 8.

Glycolysis/gluconeogenesis A and glutathione metabolism B pathway showing differentially co-expressed genes and proteins under 75 μM naringenin treatment (+ Nar)

Applied naringenin increased soybean roots GPX, SOD and POD activities

To further investigate the impact of naringin addition on glutathione metabolism in soybean roots, we assessed the enzyme activities of glutathione peroxidase (GPX), superoxide dismutase (SOD) and peroxidase (POD) (Fig S9). Following treatment with naringenin, the activities of GPX, SOD and POD at the root tips increased significantly by 1.2-fold, 13.4% and 19.4%, respectively (Fig S9).

Analysis of transcription factors

To identify the core regulatory transcription factors (TFs) of the glycolysis/gluconeogenesis and glutathione metabolism, Pearson correlation coefficients (PCC) between DEGs/DAPs detected of the pathway and TFs were calculated. TFs with high connectivity (0.95 <|PCC|< 1) and potential regulation of the pathway were selected, and an association network was constructed between TFs and DEGs/DAPs (Fig. 9A). 27 GmAP2/ERFs, 22 GmWRKYs, 14 GmNACs, and 10 GmMYBs TF family members responded to naringenin application (Fig. 9A-D). Among them, 2 GmAP2/ERFs (i.e. Glyma.09G072000 and Glyma.10G239400) had the highest connectivity, and were both up-regulated (Fig. 9B).

Fig. 9.

Analysis of connection between transcription factors and DEGs/DAPs related to glycolysis/gluconeogenesis and glutathione metabolism. A Co-expression network connecting DEGs/DAPs in glycolysis/gluconeogenesis and glutathione metabolism with transcription factors (TFs). Expression correlations between TFs (colored solid circles) and glycolysis/gluconeogenesis and glutathione metabolism-related genes/proteins (green solid squares), colored lines that show a positive (red) or negative (green) correlation. The size of the solid circles and squares in the graph indicates the degree of connectivity in the network. Heatmap analysis of GmAP2/ERFs B, GmWRKYs C, and GmMYBs/MACs D expression

Discussion

Naringenin exhibits a diverse range of effects on plant growth and development

Naringenin is a plant-derived flavonoid produced through the phenylpropanoid biosynthetic pathway [41]. Naringenin, like other flavonoids contributes to multiple physiological processes, including the regulation of antioxidant activity and auxin transport. It also enhances environmental responsiveness and resistance in plants [33, 34, 42–44]. In agreement with earlier studies, our findings indicate that exogenous naringenin triggers diverse physiological and biochemical changes in plants. These changes include reductions in root fresh weight, tap root length, lateral root number, total root length, and auxin content in the root tips (Fig. 1A-F). Naringenin treatment also increases H₂O₂, O₂⁻, and overall ROS levels in roots (Fig. 2A-D) [33, 43, 45]. The incorporation of naringenin, while inhibiting the growth and development of plant roots, simultaneously improved the plant's capacity to withstand abiotic and biotic stressors [34, 44, 46–48]. For example, exogenous naringenin can effectively alleviate the damage of pigeon pea caused by salt stress, increase the salt tolerance of pigeon pea and the content of flavonoids [34]. Naringenin induced tolerance to salt/osmotic stress through the regulation of nitrogen metabolism, cellular redox and ROS scavenging capacity in common bean [46]. Moreover, naringenin activated plant resistance responses to pathogenic bacteria, inhibit the growth of blight mold, Magnaporthe oryzae, and Pseudomonas syringae, among others [47, 48]. Despite the significance of naringenin regulates plant growth and development, its molecular mechanisms remain elusive. Therefore, this study investigated the molecular responses of soybean seedlings to naringenin for the first time by employing integrated transcriptome and proteome, with the aim of improving the molecular mechanism by which naringenin regulate soybean roots growth and development.

Naringenin regulates the growth and development of soybean roots via the auxin signaling pathway

Auxin transport is tightly regulated by PIN, LAX, and ABCB transporters [49]. Naringenin functions as a natural auxin transport inhibitor. Its effects are particularly evident in the root elongation zone and the root-shoot junction, where it modulates auxin movement, influences plant architecture, and alters root gravitropism [33, 50]. Previous studies have shown that naringenin treatment affects the expression of PIN1 and PIN2, as well as the localization of PIN2 and PIN4 proteins. These changes reduce auxin transport from the shoot apex to the root, thereby inhibiting root elongation and weakening geotropism in Arabidopsis [50]. Consistently, pin2/3/4 mutants exhibit markedly reduced root growth rates [51]. In contrast, tt4 seedlings display longer primary roots and increased lateral root numbers [33]. Our results further support these observations. Several genes associated with auxin transport were either up-regulated or down-regulated following naringenin exposure in soybean roots. Notably, GmLAX1/2/3/4 and GmPIN4 were significantly down-regulated (Fig. 3B, Table S1).Downregulation of GmPIN4 may reduce auxin transport to root tips, aligning with decreased of soybean roots growth and DR5:GUS activity after application of naringenin (Fig. 1). SAUR, Aux/IAA, and GH3 genes comprise three large gene families that are rapidly induced in response to auxin [52]. Among them, GH3 family has been reported to negatively regulate root growth in multiple species [53, 54] For example, gh3.1/2/3/4/5/6 hextuple mutant attenuated auxin inactivation, modulated meristem activities that in turn promoted lateral root initiation and growth in Arabidopsis [54]. Meanwhile, contrasting evidence indicates that overexpression of SlGH3.15 led to decrease in endogenous free IAA and severe growth defects in tap and lateral roots in tomato [53]. In our research, we observed a co-expression network centered around GmGH3.4 (Fig. 3A). Transcriptome analysis revealed upregulation of GmGH3.1/3.2/3.3/3.4 transcription (Fig. 3B-C), suggesting that increased GmGH3s transcription reduce the accumulation of auxin and suppress root growth and development (Fig. 1). SAUR facilitates auxin-mediated root development by activating H⁺-ATPases to lower apoplastic pH and activating expansins and other cell wall-modifying proteins [55]. Previous study has reported AtSAUR19 promote cell expansion and root growth [52]. A recent study on soybean indicated that overexpression of GmEXPA7 significantly increases root length and the number of root tips [56]. Similarly, in the present study, the expression of 8 GmSAURs, 20 expansins and 6 extensins genes were significantly decreased (Fig. 3B-C and S4), strongly hinting that decreased GmSAURs and expansins transcription suppress root growth and development, which merits further functional analysis.

Naringenin induces ROS outbreak in soybean roots, which inhibits root growth

ROS were recognized as critical signaling molecules that regulate plant growth and development [6, 57]. However, uncontrolled accumulation of ROS causes several damages including membrane and protein modifications in cells [17, 58, 59]. It is noteworthy that, in contrast to typical flavonoids like kaempferol, quercetin, and genistein, naringenin induced the accumulation of ROS in plant tissues [44, 60] Consistently, our findings demonstrated that the exogenous application of naringenin increases the expression of 6 GmRBOHs genes, which are the major producers of ROS in plants (Fig. S5A, Table S2); and enhances H₂O₂, O₂^– and ROS levels in roots (Fig. 2A-D). GST/GPX activity is involved in the alteration of GSH and ascorbate metabolism that leads to reduced oxidative damage [61]. Research showed abiotic stress induced root growth inhibition is strongly correlated with increased GST and GPX activity [62]. We found that glutathione metabolism was significantly enriched after naringenin induction in both the transcriptome and proteome. Most genes and proteins associated with this pathway were up-regulated under naringenin treatment, including GmGPXs and GmGSTs (Fig. 5). Among these components, GmGPX1 showed increased expression following naringenin application and acted as the central hub within two co-expressed networks (Fig. 5A). Although the specific roles of the DEGs and DAPs related to glutathione metabolism were not functionally characterized in this study, their differential responses to naringenin indicate that they warrant further investigation.

Naringenin influences soybean roots growth and development by modulating sugar metabolism

Glycolysis is the main pathway for organisms to obtain energy [63]. The regulatory mechanism of carbon source utilization was shown in Fig. 8A. In the glycolysis, the up-regulated aldose 1-epimerase (AEP), ATP-dependent 6-phosphofructokinase (PFKA), fructose-bisphosphate aldolase (FBA) and 2-phosphoglycerate dehydratase (PHGDH) indicated that naringenin treatment accelerated the conversion of β-D-Glucose to phosphoenolpyruvate in soybean roots, which provides sufficient substrates for the subsequent pyruvate pathways (Fig. 8A, Table S4). Pyruvate Kinase (PK) is a key regulatory enzyme in the glycolysis pathway. It catalyzes the last step of glycolysis by transferring the high-energy phosphate group of phosphoenolpyruvates to adenosine diphosphate (ADP), and then produces ATP and pyruvate [63, 64]. PKs have been reported to negatively regulate root growth in multiple species [65–67]. For example, overexpression ZmPK2 significantly decreased the number of shoot-borne roots; Arabidopsis heterologous overexpression GmPK21 inhibited tap root elongation under salt tolerance [66, 67]. Pyruvate is one of the main substrates for gluconeogenesis [63]. Meanwhile, naringenin treatment up-regulated key enzymes such as pyruvate decarboxylase (GmPDCs) and alcohol dehydrogenase (GmADHs) in gluconeogenesis (Fig. 8A). This phenomenon indicated that naringenin treatment promoted the conversion of pyruvate to ethanol (Fig. 8A) instead of converting to acetyl-CoA for energy metabolism, resulting in a considerable reduction in the energy supply in plant cells [68, 69]. Based on these findings, we hypothesize that naringenin treatment up-regulated GmPKs, GmPDCs, and GmADHs among others in glycolysis/gluconeogenesis, promted the excessive synthesis and accumulation of ethanol, and thereby inhibited root growth [70].

Conclusions

The phenotypical, transcriptomics and proteomics analysis of soybean roots exposed to naringenin revealed the potential regulatory molecular mechanisms of naringenin in controlling soybean root growth and development. Naringin mainly regulates soybean root growth and development through auxin and ROS content. This is reflected in the reduction of auxin content and the increase of H₂O₂, O₂⁻ and ROS at the root tips. Numerous DEGs, such as GmGH3.4, GmLAX1/2/3, and GmPIN4, were identified through transcriptomics analysis were enriched in plant hormone signal transduction. The integrated transcriptomics and proteomics analysis identified important naringenin responsive pathways, including glutathione metabolism and glycolysis/gluconeogenesis that plausibly, in large part, contribute to naringenin in controlling soybean root growth and development. Overall, this study contributes to understanding the responses and adaptations of plants to naringenin. Furthermore, the identified DEGs and DAPs will provide potential target areas for future development of soybean varieties through marker-assisted breeding and genetic engineering.

Materials and methods

Plant materials and growth conditions

A soybean genotype, YC03-3 was used in this study, which was provided by Root Biology Center, South China Agricultural University, China. Transgenic soybean seeds with DR5:GUS were provided by Dr. Xu Chen, Fujian Agriculture and Forestry University, China. Hydroponic experiments were conducted following previously described protocols [71]. In brief, seeds were germinated for 4 days. Uniform seedlings were then transferred into a nutrient solution containing 1900 μM KNO₃, 1200 μM Ca(NO₃)₂, 500 μM (NH₄)₂SO₄, 25 μM MgCl₂, 100 μM K₂SO₄, 500 μM MgSO₄, 0.5 μM CuSO₄, 1.5 μM MnSO₄, 1.5 μM ZnSO₄, 0.16 μM (NH₄)₆Mo₇O₂₄, 40 μM Fe-Na-EDTA, 2.5 μM NaB₄O₇, and 250 μM KH₂PO₄, with or without naringenin. The pH of the nutrient solution was maintained between 5.8 and 6.0. Roots were harvested to determine fresh weight, tap root length, lateral root number, and total root length at 4 d. Total root length was evaluated by the computer image analysis software (WinRhizo Pro, Canada) after acquiring root images from a scanner (Epson, Japan). Each treatment had at least five biological replicates. Roots were also harvested for further transcriptomic, proteomic, and quantitative RT-PCR (qRT-PCR) analysis.

Histochemical GUS staining assay

Histochemical GUS staining was performed as described previously [71]. In brief, samples were incubated in a GUS staining buffer containing 2 mM 5-bromo-4-chloro-3-indolyl-β-D-glucuronic acid, 0.1% Triton X-100, 2 mM K₄[Fe(CN)₆]·3H₂O, 50 mM Na₂HPO₄-NaH₂PO₄ (pH 7.2), 2 mM K₃Fe(CN)₆ and 10 mM EDTA-2Na at 37 °C. After staining, samples were washed with 70% ethanol. Sections of tap root and lateral root tips, 35 μm thick, were prepared using a Leica vibratome VT1200S (Leica, Germany). The GUS-stained roots were embedded in 7% agarose prior to sectioning. Root sections were then examined under a light microscope (LEICA DM5000B, Germany).

DAB, NBT and H2DCF-DA staining

3,3′-diaminobenzidine (DAB), nitro blue tetrazolium (NBT) and 2′,7′-dichlorofluorescein diacetate (H₂DCF-DA) imaging were performed as described previously with minor modifications [72, 73]. To detect hydrogen peroxide, superoxide anion and reactive oxygen species (ROS), seedlings were submerged into 0.1 mg.mL⁻¹ DAB (Dingguo, China), 0.1 mg.mL⁻¹ NBT (Dingguo, China) and 50 μM H₂DCF-DA (Aladdin, China) solution for 10 min, respectively. The pictures of root tips were captured as mentioned above. DAB, NBT and DCF signals were measured by Image J software. The average signal intensity measured by Image J in the Mock group was set to 1.

Transcriptomic analysis

Soybean seedlings were grown in nutrient solution containing 0 μM (Mock) or 75 μM naringenin (+ Nar) as described above. After 4 days, roots were harvested separately for total RNA extraction and mRNA library construction. RNA sequencing (RNA-seq) was performed following previously described methods [74]. Each treatment included four biological replicates. Total RNA from roots was isolated using Trizol reagent (Invitrogen, USA) according to the manufacturer’s instructions. The RNA samples were sent to Metware Biotechnology Co., Ltd (Wuhan, China) for sequencing on an Illumina NovaSeq 6000 platform. Transcript abundance was quantified as fragments per kilobase of transcript per million reads mapped (FPKM) using the RSEM program. DEGs in specific biological pathways between Mock and + Nar groups were distinguished based on the parameters (a fold-change ≥ 2.0 or < 0.5 and adjusted p-value ≤ 0.05) using DESeq 2 program for multiple comparisons. Benjamini–Hochberg procedure was used to p-value correction. Enrichment and annotation were conducted through Gene Ontology (GO) and Kyoto Encyclopedia of Gene and Genomes (KEGG) analysis. The heatmap was generated with TBtools (https://github.com/CJChen/TBtools). All raw RNA-seq data is available and retrieved on National Center for Biotechnology Information (https://www.ncbi.nlm.nih.gov) with the accession number PRJNA1347270.

RNA extraction and qRT-PCR analysis

Total RNA was extracted from soybean roots using RNA-solve reagent (OMEGA Bio-Tek, USA) according to the manufacturer’s instructions. To remove genomic DNA, RNA samples were treated with RNase-free DNase I (Invitrogen, USA). First-strand complementary DNA (cDNA) was synthesized from 1 µg of RNA using GoScript (Promega, USA) following the manufacturer’s guidelines. qRT-PCR was performed on an ABI7500 real-time PCR system (Thermo Fisher Scientific, USA). Each 20 μL reaction contained 2 μL of 1:10 diluted cDNA, 0.5 μL of each primer (final concentration 0.2 μM), 7 μL of DNase-free ddH₂O, and 10 μL of 2 × SYBR™ Green PCR master mix (Thermo Fisher Scientific, USA). The PCR cycling program included an initial denaturation at 95 °C for 1 min, followed by 40 cycles of 95 °C for 15 s, 60 °C for 60 s, and 72 °C for 30 s. A dissociation curve was generated after the final cycle. Relative gene expression levels were calculated by normalizing the expression of candidate genes to the housekeeping gene GmEF-1a (Glyma.17G186600) as described previously [75]. The primers used for qRT-PCR are listed in Table S7.

Quantitative proteomics

The protein extraction, digestion and identification were performed by Metware Biotechnology Co., Ltd Company (Wuhan, China) based on the method in a previous study [76] with a little modification. Each sample had three biological replicates. Briefly, the frozen roots were ground into powder and sonicated three times in a lysis buffer containing 4% SDS, 100 mM Tris–HCL, 10 mM DTT, 1 mM PMSF, and 2 mM EDTA. The debris was removed by centrifugation at 15,000 × g, and the protein concentration was determined using a bicinchoninic acid (BCA) assay kit (Beyotime, China). Equal amounts of protein samples were digested with trypsin, and the supernatants were mixed with 8 M urea, reduced with 5 mM DTT, and alkylated with 11 mM iodoacetamide. The protein samples were precipitated with chilled acetone, air-dried, and digested overnight at 37 °C in 25 mM ammonium bicarbonate solution with trypsin. The pH of the digested peptides was adjusted to 2–3 using 20% TFA, followed by desalting with C18 (Millipore, Billerica, MA) resin. Finally, the peptide concentration was determined by using a Pierce™ Quantitative Peptide Assay Kit with standards (Thermo Fisher, USA).

Liquid chromatography (LC) was performed on a nanoElute UHPLC (Bruker Daltonics, Germany). About 200 ng peptides were separated within 20 min at a flow rate of 0.5 μL/min on a commercially available reverse phase C18 column with an integrated CaptiveSpray Emitter (25 cm × 75 μm ID, 1.6 μm, Aurora Series with CSI, IonOpticks, Australia). The separation temperature was kept by an integrated Toaster column oven at 50 °C. Mobile phases A and B were produced with 0.1% formic acid in water and 0.1% formic acid in acetonitrile. Mobile phase B was increased from 5 to 25% over the first 17 min, increased to 40% over the next 1 min, further increased to 95% over the next 1 min, and then held at 95% for 1 min. The LC was coupled online to a hybrid timsTOF Pro2 (Bruker Daltonics, Germany) via a CaptiveSpray nano-electrospray ion source (CSI). To establish the applicable acquisition windws for diaPASEF mode, the timsTOF Pro2 was operated in Data-Independent Parallel Accumulation-Serial Fragmentation (PASEF) mode with 4 PASEF MS/MS frames in 1 complete frame. The capillary voltage was set to 1500 V, and the MS and MS/MS spectra were acquired from 100 to 1700 m/z. As for ion mobility range (1/K0), 0.85 to 1.3 Vs/cm² was used. The “target value” of 10,000 was applied to a repeated schedule, and the intensity threshold was set at 1500. The range of charge state was set from 0 to 5. The collision energy was ramped linearly as a function of mobility from 45 eV at 1/K0 = 1.3 Vs/cm² to 27 eV at 1/K0 = 0.85 Vs/cm². The quadrupole isolation width was set to 2Th for m/z < 700 and 3Th for m/z > 800.

MS raw data were analyzed using DIA-NN (v1.8.1) with library-free method. The uniprotkb_proteome_UP000008827_Glycinemax_dadou_2024_09_25.fasta database (A total of 74,861 sequences) was used to create a spectra library with deep learning algorithms of neural networks. The option of MBR (Match Between Runs) was employed to create a spectral library from DIA data and then reanalyzed using this library. FDR (false discovery rate) of search results was adjusted to < 1% at both protein and precursor ion levels. Differential accumulation proteins (DAPs) between comparisons of two groups were selected by t-test with the criteria of fold change > 1.5 or < 0.67 (p-value ≤ 0.05). The MS data have been deposited to the ProteomeXchange Consortium via the iProX (www.iprox.org) repository with the data set identifier IPX0013904000.

Determination of GPX, SOD and POD activity

The activities of GPX, SOD, and POD were measured using assay kits (Keming, China). Specifically, 40 µL of enzyme extract was added for the GPX reaction, while 50 µL was used for both the SOD and POD reactions. Absorbance changes were recorded at wavelengths of 412 nm, 470 nm, and 450 nm, respectively, using a microplate reader (Thermo Fisher Scientific, USA).

Statistical analysis

Statistical analysis of all data was performed using Microsoft Excel 2019 (Microsoft Company, Redmond, WA, USA). Significant differences between treatments were evaluated using Student’s t-test.

Abbreviations

DAB 3

3′-Diaminobenzidine

DAP

Differential accumulation protein

DEG

Differentially expressed gene

GPX

Glutathione peroxidase

GO

Gene Ontology

H₂DCF-DA

2′,7′-Dichlorofluorescein diacetate

KEGG

Kyoto Encyclopedia of Gene and Genomes

LR

Lateral root

LRP

Lateral root primordium

NBT

Nitro blue tetrazolium

Nar

Naringenin

PCC

Pearson correlation coefficients

PCA

Principal component analysis

POD

Peroxidase

QC

Quiescent center

RAM

Root apical meristem

ROS

Reactive oxygen specie

qRT-PCR

Quantitative real-time polymerase chain reaction

SOD

Superoxide dismutase

TF

Transcription factors

TR

Tap root

UHPLC-MS/MS

High-performance liquid chromatography-tandem mass spectrometry

WT

Wild type

Authors’ contributions

ZZ: Writing-review & editing, Writing-original draft, Data curation, Conceptualization. CB: Formal analysis, Conceptualization. ZY: Formal analysis. MX: Formal analysis. LX: Project administration. WT: Project administration. LC Project administration. TJ: Writing-review & editing, Project administration. All authors reviewed the manuscript.

Funding

This work was supported by the Key Areas Research and Development Programs of Guangdong Province (2022B0202060005), the National Key Research and Development Program of China (2021YFF1000500, 2023YFD1901300), National Natural Science Foundation of China (32172658, 32172659), the Natural Science Foundation of Guangdong Province (2025A1515012159).

Data availability

The RNA-seq produced in this study can be found in the NCBI SRA database under project number PRJNA1347270. The proteomics data have been deposited to the ProteomeXchange Consortium via the iProX (www.iprox.org) repository with the data set identifier IPX0013904000.

Declarations

Ethics approval and consent to participate

Not applicable.

Consnt for publication

Not applicable.

Competing interests

The authors declare no competing interests.