SmbHLH60 and SmMYC2 antagonistically regulate phenolic acids and anthocyanins biosynthesis in Salvia miltiorrhiza

Graphical abstract

We discovered that a new transcription factor SmbHLH60 repressed the biosynthesis of phenolic acids and anthocyanins in S. miltiorrhiza through transcriptional inhibition of SmTAT1 and SmDFR. In addition, SmbHLH60 and SmMYC2 formed a heterodimer to antagonistically regulate phenolic acids and anthocyanins biosynthesis.

Highlights

• •SmbHLH60 is one of the most significantly down-regulated bHLH genes induced by MeJA.
• •Phenolic acids and anthocyanins were obviously reduced in SmbHLH60-OE hairy roots whereas phenolic acids and anthocyanins were significantly increased in SmbHLH60-CRISPR hairy roots.
• •SmbHLH60 negatively regulates phenolic acid biosynthesis mainly via SmTAT1 gene.
• •SmbHLH60 negatively regulates anthocyanin biosynthesis mainly via SmDRF gene.
• •SmbHLH60 works with SmMYC2 in an antagonistic manner in phenolic acid biosynthesis.

Abstract

Introduction

Salvia miltiorrhiza is a renowned traditional Chinese medicinal plant with extremely high medicinal value, especially for cardiovascular and cerebrovascular diseases. The jasmonic acid (JA) signaling pathway plays an important role in the improved biosynthesis of secondary metabolites, which is mediated by a major transcriptional regulator, MYC2. However, the JA regulatory mechanism of secondary metabolites biosynthesis in S. miltiorrhiza is still largely unknown.

Objectives

Our work focuses on the dissection of the molecular mechanism of transcriptional regulation in MeJA-mediated biosynthesis of medicinal components of S. miltiorrhiza. We examined the role of MeJA-responsive bHLH transcription factors (TFs) in improving bioactive secondary metabolites accumulation in S. miltiorrhiza.

Methods

Hairy root transformation based on CRISPR/Cas9 technique was used to decipher gene function(s). Changes in the content of phenolic acids were evaluated by HPLC. Y1H, EMSA and dual-LUC assays were employed to analyze the molecular mechanism of SmbHLH60 in the regulation on the biosynthesis of phenolic acids and anthocyanins. Y2H, BiFC and pull-down affinity assays were used to corroborate the interaction between SmbHLH60 and SmMYC2.

Results

Being one of the most significantly negatively regulated bHLH genes by MeJA, a new transcription factor SmbHLH60 was discovered and characterized. Over-expression of SmbHLH60 resulted in significant inhibition of phenolic acid and anthocyanin biosynthesis in S. miltiorrhiza by transcriptionally repressing of target genes such as SmTAT1 and SmDFR, whereas CRISPR/Cas9-generated knockout of SmbHLH60 resulted in the opposite effect. In addition, SmbHLH60 and SmMYC2 formed a heterodimer to antagonistically regulate phenolic acid and anthocyanin biosynthesis.

Conclusion

Our results clarified that SmbHLH60 is a negative regulator on the biosynthesis of phenolic acids and anthocyanins. SmbHLH60 competed with SmMYC2 in an antagonistic manner, providing new insights for the molecular mechanism of MeJA-mediated regulation on the biosynthesis of secondary metabolites in S. miltiorrhiza.

Introduction

Salvia miltiorrhiza, Lamiaceae family, is a medicinal herb widely used in Chinese medicine for treatments of cardiovascular and cerebrovascular diseases [1], [2], [3], [4], [5]. The composition and concentration of phenolic acids and liposoluble tanshinones are associated with their health-promoting properties [1], [3], [6]. Caffeic acid (CA), salvianolic acid B (SAB), salvianolic acid A (SAA) and rosmarinic acid (RA) are the main bioactive phenolic acids identified and reported to have anti-oxidant, anti-inflammatory, anti-atherosclerosis, anti-tumor, and anti-diabetic activities [7], [8], [9]. High economic value and increasing market demands for enhanced content of phenolic acids derive current researches focus on the regulation of phenolic acid biosynthesis pathways [3], [10], [11], [12].

In plants, the phenolic acid biosynthesis pathway has been extensively studied and in S. miltiorrhiza two upstream pathways have been characterized: the phenylpropane metabolic pathway and the tyrosine-derived metabolic pathway (Fig. S1) [3], [12], [13], [14]. In the phenylpropane metabolic pathway, L-phenylalanine is sequentially metabolized by phenylalanine ammonia-lyase (PAL), cinnamic acid 4-hydroxylase (C4H) and 4-coumarate: CoA ligase (4CL) to generate one of the precursors, 4-coumaroyl-CoA [15], [16], [17]. In parallel, tyrosine conversion is catalyzed by tyrosine aminotransferase (TAT), 4-hydroxyphenylpyruvate reductase (HPPR) and additional uncharacterized enzymatic steps to form the other precursor, 3,4-dihydroxyphenyllactic acid [15], [18]. Subsequently, the two precursors will be used to generate rosmarinic acid catalyzed by rosmarinic acid synthase (RAS) and cytochrome P450-dependent monooxygenase (CYP98A14). The formation of additional phenolic acids by these two precursors may occur by yet unknown step(s) [10], [13], [19].

4-coumaroyl-CoA is an important precursor in the phenylpropanoid pathway taking part in the production of numerous phenylpropanoids including anthocyanins [3]. The latter is extremely important secondary metabolites, widely distributed among higher plants, serve as important natural pigments and have strong antioxidant activity, and are therefore beneficial for the human diet [20], [21]. So far, a major part of the anthocyanin biosynthesis pathway in S. miltiorrhiza has been characterized including chalcone synthase (SmCHS), chalcone isomerase (SmCHI), flavone synthase (SmFNS), flavanone 3-hydroxylase (SmF3H), flavonoid 3′-hydroxylase (SmF3′H), flavonoid 3′,5′-hydroxylase (SmF3′5′H), flavonol synthase (SmFLS), dihydroflavonol 4-reductase (SmDFR), and anthocyanidin synthase (SmANS) [12], [22]. However, only a few works studied the regulation mechanism of anthocyanins in S. miltiorrhiza.

Many elicitors including gibberellins (GAs), methyl jasmonate (MeJA), abscisic acid (ABA), salicylic acid (SA), yeast extract (YE), Ag⁺, nitric oxide (NO), and hydrogen peroxide (H₂O₂) have been proven to increase the phenolic acids accumulation [16], [23], [24], [25]. In addition, MeJA has been used to increase the content of anthocyanins [3], [20], [26]. However, the mechanism in which MeJA-induced phenolic acid and anthocyanin regulation in S. miltiorrhiza is still largely unknown.

The basic helix-loop-helix (bHLH) TF family plays an extremely important role in secondary metabolism regulation [27], [28], [29]. It was also shown that bHLHs can bind to specific promoter regions in their targeted genes, such as E-box (CANNTG) or G-box (CACGTG) [8], [30], [31], [32]. Currently, 127 bHLH TFs were found in S. miltiorrhiza by genome mining [4], [33], some of which were discovered to be involved in the regulation of secondary metabolism. For example, SmbHLH3 negatively regulates phenolic acid biosynthesis by modulating SmTAT1 and SmHPPR1 [34]. RA, SAB and CA levels were increased by 2.87, 4.00 and 5.99 times, respectively, as compared to the control in SmbHLH148 [35]. SmbHLH10 increased the accumulation of tanshinone in the roots of S. miltiorrhiza [36]. SmbHLH92 and SmbHLH37 have been confirmed to negatively regulate phenolic acids biosynthesis pathways while SmbHLH51 functions as a positive regulator [8], [37], [38]. SAB biosynthesis, in S. miltiorrhiza, triggered by MeJA is also regulated by SmbHLH53 presumably with a dual-role [14]. MYELOCYTOMATOSISs (MYCs), another type of bHLH TFs, were also shown to take a central role in secondary metabolites regulation. For example, AaMYC2, a jasmonate-responsive TF, positively regulates artemisinin biosynthesis in Artemisia annua [39]. In Catharanthus roseus, CrMYC2 has been proved to be essential for the accumulation of alkaloids [40]. In S. miltiorrhiza, SmMYC2 plays a role as a core transcription factor in the MeJA-mediated phenolic acid secondary metabolism signaling pathway by actively binding to SmPAL1, SmTAT1 and SmCYP98A14 promoters altering their expression, leading to the accumulation of phenolic acids [37], [41]. Still, in S. miltiorrhiza, the mechanism in which SmMYC2 regulates MeJA-mediated phenolic acid and anthocyanin biosynthesis is not fully understood. Our data suggest that by forming a heterodimer, SmbHLH60 and SmMYC2 antagonistically regulate anthocyanin and phenolic acid biosynthesis. Our new findings reveal the molecular regulation mechanism of SmbHLH60 and SmMYC2 and elucidate the MeJA-mediated regulation of secondary metabolites regulation in S. miltiorrhiza.

Materials and methods

Plant materials

S. miltiorrhiza seedlings were grown in the greenhouse of Zhejiang Chinese Medical University. S. miltiorrhiza seedlings were cultured on Murashige and Skoog (MS) medium at 25 °C, which were lighted for 16 h and dark for 8 h. Nicotiana benthamiana was cultivated in a greenhouse at 25 °C, under the same conditions as S. miltiorrhiza. S. miltiorrhiza hairy roots were grown on 1/2 MS solid medium and cultured in the dark in a greenhouse at 25 °C. For culture in a flask, the hairy roots were cultured in 100 mL 1/2 MS liquid medium at 120 rpm for 50 days in the dark.

Isolation and characterization of SmbHLH60

SmbHLH60 was found to be one of the most significantly down-regulated bHLH genes in RNA-Seq data, and the full-length cDNA was amplified using specific primers (Table S1). ClustalX and MEGA 6.0 software were used for sequence alignment and phylogenetic tree analysis. The phylogenetic tree was constructed based on the amino acid sequence with the neighbor-joining method and then optimized with the iTOL tool. All protein sequences including SibHLH130 (XP_011087339.1), SibHLH130-like (XP_011093536.1), PjbHLH130 (GFQ08151.1), OebHLH130-like (XP_022896111.1) used in multiple sequence alignment were download from NCBI database.

Gene expression profile assay

For RNA isolation, different tissues of one-year-old S. miltiorrhiza plants, including roots, lateral roots, stems, xylems, phloems, leaves and young leaves were collected and performed using the Tiangen Plant RNA Extraction Kit. Followed by reverse transcription and qRT-PCR for SmbHLH60 gene detection, SmActin was used as an internal reference. Exogenous plant hormones including 100 μM methyl jasmonate (MeJA) was used to spray one-month-old S. miltiorrhiza seedlings and sampled at 0, 1, 2, 4, 6, 8, 12, 24 h after treatment for RNA extraction, followed by reverse transcription and qRT-PCR to detect SmbHLH60 gene expression. The relative quantitative analysis method (2^−ΔΔCT) was used to calculate the relative gene expression, and SmActin was used as the internal reference gene and all experiments were repeated with more than three biological replicates.

Subcellular localization of SmbHLH60

To determine the subcellular localization of SmbHLH60, the open reading frame (ORF) of SmbHLH60 was cloned and constructed into the pHB-YFP vector driven by the CaMV 35S promoter to form the SmbHLH60-YFP fusion protein (Fig. S2B). The negative control was performed by an empty pHB-YFP vector. The fusion vector was transformed into Agrobacterium tumefaciens GV3101 strain. Then the transformed GV3101 was injected into five-week-old N. benthamiana leaves while pHB-YFP was used as a negative control. To confirm the position of the nucleus, 10 mg/mL 4′,6-diamidino-2-phenylindole dihydrochloride (DAPI) solution was injected into N. benthamiana leaves 4 h before observing the fluorescence signal. Then the fluorescence signal was observed by laser confocal microscope (Zeiss, Germany) under the excitation of 405 nm and 488 nm laser. All experiments were repeated with more than three biological replicates.

Generating S. miltiorrhiza hairy roots

The complete ORF of SmbHLH60 was cloned into the pCAMBIA2300⁺ vector containing the CaMV 35S promoter to form the pCAMBIA2300⁺-SmbHLH60 recombinant vector (Fig. S2A). For the knockout vector, sgRNA was designed according to the 5′-G-19base-NGG-3' principle, which was constructed on the pCAMBIA2300⁺ vector (Fig. S2A). SmbHLH60 sgRNA was driven by the AtU6 promoter, and hSpCas9 was by the CaMV 35S promoter with pCAMBIA2300⁺ serving as a control vector. The above vectors were transformed into C58C1 and then used to infect S. miltiorrhiza for hairy roots. The hairy root DNA was extracted by the CTAB method and the positive clones were identified with specific primers (Table S1). All experiments were repeated with three biological replicates.

Quantitative real-time PCR (qRT-PCR)

Tissues including taproots, lateral roots, xylems, phloems, stems, leaves, and tender leaves of S. miltiorrhiza were collected and ground into a powder with liquid nitrogen to extract total RNA and follow the steps in the Tiangen plant total RNA extraction kit. The qRT-PCR experiments were used to detect the results with Thermo Fisher quantitative master mix and the Applied Biosystem Step One Real-Time PCR System (Applied Biosystems, USA) and the SmActin was used as an internal control [42]. All experiments had three biological replicates.

Measuring the accumulation of phenolic acids and anthocyanins in hairy roots

0.1 g of dried hairy root powder was weighed for phenolic acid extraction with 80% ethanol solution (4:1, v/v), and ultrasonic extraction was performed for 30 min. The extract was dissolved in 2 mL of distilled water after vacuum rotary steaming. HPLC was used to detect each component in the phenolic acid extract, and the content was calculated by substituting it into the standard curve. In addition, 0.02 g of dried hairy root powder was weighed and mixed in 1 mL of 1% (v/v) hydrochloric acid methanol solution (hydrochloric acid: methanol = 1:99), at 100 rpm, 20 °C overnight for total anthocyanin extraction. Then samples were centrifugated at 12,000 rpm and the supernatant was taken for further analysis. Samples were mixed similar volume of chloroform and the absorbance of the extract at 530 nm and 657 nm wavelengths were detected by spectrophotometry. Anthocyanin content Q (anthocyanin) = (A530-0.25 × A657) × M⁻¹, and M is the dry weight of plant tissue. All measurements were carried out in three biological replicates.

Yeast one-hybrid assay (Y1H)

The complete SmbHLH60 ORF sequence was constructed on the pB42AD vector to form a recombinant vector, and the E/G-box in the promoters of key genes of phenolic acid and anthocyanin biosynthesis were constructed on the pLacZ2u vector. The recombinant plasmid was co-transformed into yeast strain EGY48, followed cultured on SD/-Ura/-Trp medium for 48 h. Then the positive colony was picked and tested for growth of monoclonal strains on SD/-Ura/-Trp/Raf/Gal medium with X-gal for 48 h. Empty vectors pB42AD and pLacZ2u were used as negative controls. We performed the Y1H experiment more than three biological replicates.

Electrophoretic mobility shift assay (EMSA)

The ORF of SmbHLH60 was inserted into the vector pCold-His. Recombinant protein SmbHLH60-His was expressed in E. coli BL21 (DE3) and purified using a protein purification kit with a His-tag, which was purchased from Shanghai Sangon Co., Ltd. Probe fragments from the promoter of SmTAT1 and SmDFR were synthesized by Shanghai Sangon Co., Ltd. The 5 × EMSA binding buffer was purchased from Beyotime Biotechnology Co., Ltd. (Shanghai, China), and the chemiluminescence detection kit was purchased from Thermo Fisher Scientific (Shanghai, China), and the fluorescence was detected using a C300 image scanner (Azure Biosystems, USA). SmbHLH60-His purified protein was combined with probes of SmTAT1pro and SmDFRpro by adding 5 × binding buffer (Beyotime Biotechnology Co., Ltd.) at 25 °C for 25 min. Then follow the previously reported method [12], [42]. EMSA experiment was repeated with three biological replicates.

Dual-luciferase assay (dual-LUC)

Promoters about 2000 bp of all key enzyme genes involved in phenolic acid and anthocyanin biosynthesis were constructed to pGreen0800-LUC vector. The recombinant vector was then transferred to A. tumefaciens GV3101. The N. benthamiana was then transformed instantaneously by infiltrating the Agrobacterium mixture into the back of the leaves. After 48 h, samples were taken from the infected site, and the extract was detected using a dual-luciferase reporter analysis system (Promega, Madison, USA), and the fluorescence values were detected. All dual-LUC experiments were repeated for more than three biological replicates.

Bimolecular fluorescent complementary assay (BiFC)

To verify the interaction between SmbHLH60 and SmMYC2, the full-length ORF of SmbHLH60 and SmMYC2 were constructed on pXY106-nYFP and pXY104-cYFP, respectively, and then transformed into A. tumefaciens GV3101. The resuspended Agrobacteria were mixed and injected into N. benthamiana leaves for transient expression. Then N. benthamiana plants were placed in a greenhouse at 26 °C for 48 h, and the yellow fluorescence was observed by laser confocal microscope (Zeiss, Germany) under the excitation of 488 nm lasers previously described [3], [12]. This experiment was repeated with three replicates.

GST pull-down assay

The complete ORFs of SmbHLH60 and SmMYC2 were respectively constructed on the pCold-His vector and pGEX-4T-1 vector, which were transformed into E. coli BL21 (DE3) for protein expression, using His-tag and GST-tag protein purification kit (Shanghai Sangon Co., Ltd) for purification, respectively. SmMYC2-GST was incubated with GST magnetic beads (Shanghai Sangon Co., Ltd) at 26 °C for 30 min to form a GST-target protein complex. Subsequently, SmbHLH60-His recombinant protein was added for the binding reaction at 4 °C overnight. After 3000 rpm centrifugation, the supernatant was aspirated and used for input detection by anti-GST and anti-His. Then, the magnetic beads were washed 3 times with wash buffer, the supernatant was discarded. Finally, the elution buffer was added to elute the protein, which was taken for electrophoresis and western blot detection by anti-His. This experiment had three biological replicates.

Results

SmbHLH60 expression pattern analysis

SmbHLH60 was found to be one of the most significantly down-regulated bHLH genes in MeJA-mediated RNA-Seq data (Fig. 1A). To examine the involvement of SmbHLH60 in the MeJA signaling pathway, exogenous MeJA was applied to the whole plant, and the expression of SmbHLH60 was examined by qRT-PCR. SmbHLH60 expression was significantly decreased in MeJA-treated samples as compared to control (0 h) from 1 h, reaching the lowest levels at 6 h (Fig. 1B). These results showed good accordance with our previous RNA-seq data. In addition, SmbHLH60 showed an opposite expression pattern with SmMYC2 and genes involved in phenolic acid biosynthesis such as SmPAL1, SmC4H, SmTAT1, SmHPPR, SmRAS and SmCYP98A14 after MeJA treatment (Fig. 1A). Meanwhile, we analyzed the expression patterns of SmbHLH60 and the previously characterized SmbHLHs under MeJA treatment. We found that the expression of SmbHLH92, a negative regulator in the biosynthesis of phenolic acid, decreased in MeJA-mediated RNA-seq data (Fig. S5), which showed a similar expression pattern to SmbHLH60. Therefore, we speculated that SmbHLH60 may participate in the negative regulation of phenolic acid. Total phenolic acid accumulation in the leaf of S. miltiorrhiza was the highest among the three tissues (root, leaf and stem) (Fig. S6). Tissue expression profile showed that SmbHLH60 expressed the highest in leaf, especially in young leaf (Fig. 1C). Meanwhile, SmMYC2 was found to have the highest expression in the leaf [43]. These results support that SmbHLH60 might participate in the regulation of phenolic acids. Therefore, we focused on the research of SmbHLH60.

Fig. 1.

Expression profiles analysis of SmbHLH60. (A). The analysis of SmbHLH60 expression pattern in MeJA-induced transcriptome library at 1 h and 6 h. The heat map was produced using TBtools. The yellow color indicated high expression levels, and the blue color indicated low expression levels. A total of 89 bHLH TFs showed decreased expression after MeJA treatment revealed in the transcriptome library. The expression of SmbHLH60 decrease obviously after MeJA treatment at 1 h (B). The effect of exogenous MeJA on the expression of SmbHLH60. The expression of SmbHLH60 was detected under MJ treatment for 0 h, 1 h, 2 h, 4 h, 6 h, 8 h, 12 h, 24 h. SmbHLH60 was drastically decreased under the regulation of MeJA in the whole plant. (C). SmbHLH60 gene expressions in different tissues of S. miltiorrhiza. The transcription level of SmbHLH60 was detected in the xylem, phloem, taproot, lateral root, mature leaf, steam and young leaf respectively. All experiments were repeated three times, and the error bars represented the standard deviation of the three replicates. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Isolation and characterization of SmbHLH60

SmbHLH60 sequence contained an 1185 bp of an open reading frame, encoding a 394 amino acids protein with a size prediction of 43.58 kDa. To explore the evolutionary relationship between SmbHLH60 and 167 bHLH TFs in Arabidopsis thaliana, the phylogenetic tree was constructed, which showed that SmbHLH60 was high homology with AtbHLH130 (Fig. S3). The latter is known as FLOWERING BHLH4 (FBH4) and it was reported to activate CONSTANS (CO) with FBH1, FBH2, and FBH3 redundantly as part of the flowering regulatory mechanism in A. thaliana [43]. The result of BLAST-Protein (BLASTP) analysis showed that SmbHLH60 has the highest identity (60.58%) to Sesamum indicum bHLH130. All proteins, including SibHLH130-like, PjbHLH130, OebHLH130-like contained the basic helix-loop-helix conserved domain at the C-terminus (Fig. 2A). Subcellular localization assay was used to explore the location of SmbHLH60 in cells. SmbHLH60-YFP fluorescence was observed in the nucleus using a laser confocal microscope, while the YFP signal of pHB-YFP was distributed in the cell nucleus and cytoplasm, and DAPI appeared in the nucleus (Fig. 2B).

Fig. 2.

Characterization and subcellular localization of SmbHLH60. (A). ClustalX was performed for multiple sequence alignment. Multiple amino acid sequence alignments of bHLH TFs including SibHLH130 (XP_011087339.1), PjbHLH130 (GFQ08151.1), OsbHLH130-like (XP_022896111.1) and SibHLH130-like (XP_011093536.1). (B). Subcellular localization of SmbHLH60 in N. benthamiana leaf epidermal cells. YFP fluorescence and DAPI signals were observed at 488 nm and 405 nm.

Generation of SmbHLH60-overexpression and SmbHLH60-CRISPR hairy roots

To decipher the role of SmbHLH60 in phenolic acid and anthocyanin accumulation, transgenic lines were generated. Recombinant vectors pCAMBIA2300⁺-SmbHLH60 and pCAMBIA2300⁺-SmbHLH60-CRISPR/Cas9 were separately transfected into the modified A. tumefaciens C58C1 harboring A. rhizogenes A4 Ri plasmid. For infecting explants of S. miltiorrhiza to obtain genetically modified hairy roots. Fig. S4 showed the results that the positive SmbHLH60-OE lines were identified with primer in Table S1. EV means hairy roots which derive from infecting with A. rhizogenes C58C1 containing empty vector (pCAMBIA2300⁺) plasmid, which was used as a negative control. The CRISPR/Cas9 knockout lines used target sequence-specific primers to amplify about 400 bp sequence for sequence determination. The sequencing results and sequencing diagrams were shown in Fig. 3C-D, and the TGG was the PAM sequence. Four SmbHLH60 overexpression (SmbHLH60-OE) and four CRISPR/Cas9 knockout lines (SmbHLH60-CRISPR) were used for further analysis (Fig. 3A). In SmbHLH60-OE lines, the expression levels of SmbHLH60 were significantly higher as compared to the control. Conversely, expression levels of SmbHLH60 were much lower in the knockout lines as compared to the control (Fig. 3B).

Fig. 3.

Generation of transgenic hairy roots. (A). Hairy roots were cultured for 50 days in 1/2MS. Four knockout lines and four overexpression lines were used for cultivation and subsequent experiments. (B). Transcript-level expression of SmbHLH60 in SmbHLH60-CRISPR lines and SmbHLH60-OE lines were tested by qRT-PCR and error bars represented the standard deviation of three biological replicates (*, p < 0.05; **, p < 0.01). (C-D). CRISPR/Cas9 system knocks out hairy root. TGG was the PAM sequence. After being edited with CRISPR/Cas9 system, line 5 and line 36 lacked 2 bases, line 78 lacked 3 bases, and line 83 lacked 10 bases.

SmbHLH60 reduces the phenolic acid accumulation

Total polyphenol content was determined using Folin-Phenol reagent was the highest in the knockout lines, then in the control, and the lowest levels were detected in the SmbHLH60 overexpression hairy root lines (Fig. 4A). HPLC analysis confirmed our results with total phenolic acids (TS) of 32.77 ± 1.75 mg/g in the knockout lines, significantly higher than the control line with 22.07 ± 2.55 mg/g, while the TS content in SmbHLH60 overexpression hairy roots was 12.88 ± 3.21 mg/g significantly lower than the control line (Fig. 4B). In accordance with our chemical profiling, the expression level of genes involved in the biosynthesis of phenolic acid SmPAL1, SmC4H1, Sm4CL2, SmTAT1, SmHPPR1, SmRAS1 and SmCYP98A14 were significantly higher in the SmbHLH60-CRISPR lines as compared to the control. The expression levels in the SmbHLH60-OE lines were significantly lower as compared to the control line (Fig. 4C).

Fig. 4.

SmbHLH60 negatively regulate biosynthetic genes and the production of phenolic acids (A). Color of total phenolic acids (TS) extracts of transgenic hairy roots. Folin-Phenol was used for staining of total phenolic acid extract. The shade of color represented the total content of phenolic acid. (B). Determination of total phenolic acid and three main phenolic acids including rosmarinic acid (RA), salvianolic acid B (SAB), caffeic acid (CA) by HPLC. The content determination of each hairy root line was repeated three times, and error bars represented the standard deviation of three biological replicates (*, p < 0.05; **, p < 0.01). (C). qRT-PCR was used to detect gene expressions of key enzymes in phenolic acid biosynthesis. Transcription levels of genes involved in phenolic acid biosynthesis were detected in SmbHLH60-CRISPR lines and SmbHLH60-OE lines respectively. All experiments were repeated three times, and error bars represented the standard deviation of three biological replicates (*, p < 0.05; **, p < 0.01).

SmbHLH60 negatively regulates anthocyanin biosynthesis

The initial steps of the phenylpropanoid pathway provide the precursors for the biosynthesis of phenolic acids and anthocyanins [3]. Total anthocyanin extracts from hairy roots of different lines exhibited inconsistent colors (Fig. 5A). Therefore, we examined if the effect on color accumulation is directly or indirectly related to the anthocyanin biosynthetic pathway. In the SmbHLH60-CRISPR lines, the total anthocyanins were significantly increased, about twice the control. In the SmbHLH60-OE lines, the anthocyanin content was significantly reduced, which was about 42% lower than the control (Fig. 5B). Expression of genes involved in the biosynthesis of anthocyanins including SmCHS, SmFLS, SmF3H, SmF3′H, SmF3′5′H, SmANS and SmDFR was significantly higher in the SmbHLH60-CRISPR lines as compared to the control, while in the latter it was significantly higher as compared to the SmbHLH60-OE lines (Fig. 5C).

Fig. 5.

SmbHLH60 transgenic lines negatively regulate biosynthetic genes and total anthocyanins. (A). Color of anthocyanin extracts of transgenic hairy roots. The hydrochloric acid–methanol solution was used to extract total anthocyanins (v: v = 1: 99). The picture showed the product after the extraction of total anthocyanins. (B). Determination of total anthocyanins. The anthocyanin extract was mixed with an equal volume of chloroform and detected by a microplate reader (*, p < 0.05; **, p < 0.01). (C). qRT-PCR was used to detect gene expression of key enzymes in anthocyanin biosynthesis. All experiments were repeated three times, and error bars represented the standard deviation of three biological replicates (*, p < 0.05; **, p < 0.01).

SmbHLH60 binds and transcriptionally inhibits the promoters of SmTAT1 and SmDFR

SmbHLH60 caused a decrease in phenolic acid and anthocyanin, so we selected 11 genes (7 key enzyme genes of phenolic acid biosynthesis and 4 key enzyme genes of anthocyanin biosynthesis) whose transcription level decreased significantly in the SmbHLH60-OE lines and increased in the SmbHLH60-CRISPR lines for dual-LUC experiments. The results showed that SmTAT1, SmPAL1, SmRAS1 (phenolic acid biosynthesis pathway) and SmDFR (anthocyanin biosynthesis pathway) were significantly transcriptionally repressed by SmbHLH60 (Fig. 6A-B, S7). Further investigation showed that SmbHLH60 only bound to the G-box element in the promoter of SmTAT1 and SmDFR as performed by Y1H (Fig. 6C-D). EMSA was performed to further verify whether SmbHLH60 is bound to the promoters of SmTAT1 and SmDFR, and the pCold-His protein was used as a negative control. The band of the protein-probe complex was detected only in the presence of the SmbHLH60-His fusion protein, indicating that SmbHLH60 can bind to SmTAT1 and SmDFR by the G-box in its promoter (Fig. 6E-F). These results revealed that SmTAT1 and SmDFR, the target genes of SmbHLH60, were transcriptionally inhibited by SmbHLH60, which was consistent with the expression of SmTAT1 and SmDFR in our qRT-PCR results.

Fig. 6.

SmbHLH60 can bind to the G-Box motif in the promoter region of SmTAT1 and SmDFR. (A). Schematic diagram of constructs used in assays of transient transcriptional activity. (B). SmbHLH60 repressed promoters of SmTAT1 and SmDFR, and the relative LUC activity was normalized to the Renilla (REN) luciferase. pGreen0800-SmTAT1pro-LUC and pGreen0800-SmDFRpro-LUC were co-injected with pHB-SmbHLH60 respectively into tobacco epidermal cells for promoter activity determination. All experiments were carried out in three biological replicates and error bars represented the standard deviation of three biological replicates (*, p < 0.05; **, p < 0.01). (C). The positions of E/G-box elements in SmTAT1pro and SmDFRpro for Y1H and EMSA analysis. (D). The Y1H assays showed that SmbHLH60 bound to the G-box in the promoters of SmTAT1 and SmDFR. Two vectors including pB42AD-SmbHLH60 and pLacZ2u-E/G-box were transferred into EGY48 yeast, which was placed on a medium containing X-gal for selection. (E). Probe sequences used in the EMSA experiment including CACGTG conserved sequence. (F). Specific binding of SmbHLH60 to G-box in the promoters of SmTAT1 and SmDFR. The third lane of the DNA-protein complex was detected, indicating that SmbHLH60 was directly bound to the probe containing the G-box element.

SmbHLH60 interacts with SmMYC2 to form a heterodimer

Interestingly, the Y2H results showed that SmbHLH60 and SmMYC2 could interact directly (Fig. 7A). BiFC and GST pull-down were performed to clarify the interaction between SmbHLH60 and SmMYC2. In BiFC assays, the peptides nYFP and cYFP were combined to excite fluorescence only when nYFP-SmbHLH60 and SmMYC2-cYFP combined, otherwise there was no fluorescence excitation (Fig. 7B). In the GST pull-down assay, the SmbHLH60-His fusion protein in the experimental group and input was detected through the His antibody, while GST protein and SmMYC2-GST fusion protein were detected through the GST antibody. SmbHLH60-His was incubated with SmMYC2-GST and GST, respectively. The result of the western blot (WB) assay found that the luminescence signal was only detected in the group where SmbHLH60-His and SmMYC2-GST were incubated together, while no band was detected in the group incubated with SmbHLH60-His and GST (Fig. 7C). These results indicated that an interaction existed between SmbHLH60 and SmMYC2.

Fig. 7.

SmbHLH60 interacted with SmMYC2 in vitro and in vivo. (A). Yeast-two-hybrid assay showed SmbHLH60 can interact with SmMYC2. The ORF of SmMYC2 and SmbHLH60 were connected to pGADT7 and pGBKT7, respectively. After AD-MYC2 and BD-SmbHLH60 were co-transformed, the yeast could grow normally on the SD/-Trp/-Leu/-His/-Ade medium. (B). Bimolecular fluorescent complementary assays to detect the interaction between SmbHLH60 and SmMYC2. The ORFs of SmbHLH60 and SmMYC2 were constructed on the pXY-106-nYFP vector and pXY-104-cYFP vector, respectively. Fluorescence could only be detected when nYFP-SmbHLH60 was combined with SmMYC2-cYFP, scale bar was 20 μm. (C). GST pull-down assay verified that SmbHLH60 could interact with SmMYC2. The fusion protein SmbHLH60-His and SmMYC2-GST were incubated together with magnetic beads. The final results were detected with the anti-His.

SmbHLH60 attenuates the transcriptional activation effect of SmMYC2 on SmTAT1 and SmDFR

The discovery that SmMYC2 and SmbHLH60 target the same genes provoked us to further investigate the molecular mechanism underlying this finding. Dual-LUC was performed to explore the effect of SmMYC2 and SmbHLH60 on SmTAT1 and SmDFR transcription activation. Our results showed that SmMYC2 can activate SmTAT1 and SmDFR, while SmbHLH60 can attenuate the expression of SmTAT1 and SmDFR when SmMYC2 and SmbHLH60 existed alone. However, the transcriptional activation levels of SmTAT1 and SmDFR were weakened when SmMYC2 and SmbHLH60 acted as the co-effector (Fig. 8A-B), suggesting that SmbHLH60 competed with SmMYC2 to bind to the promoters of SmTAT1 and SmDFR. Especially, the SmMYC2-SmbHLH60 interaction did not affect the ability of SmbHLH60 to bind to the G-box probes of SmTAT1pro and SmDFRpro (Fig. 8C-D). To study the relationship between SmbHLH60 and SmMYC2, we detected the transcription level of SmMYC2 in SmbHLH60 transgenic plants. The transcription level of SmMYC2 decreased in SmbHLH60-OE hairy roots, but increased in SmbHLH60-CRISPR hairy roots (Fig. 8E). Meanwhile, the expression of SmbHLH60 was also detected in SmMYC2-OE lines, which showed that SmbHLH60 was decreased (Fig. S9). These results indicated that SmbHLH60 and SmMYC2 function as antagonistic regulators of phenolic acid and anthocyanin biosynthesis by direct regulation of SmTAT1 and SmDFR.

Fig. 8.

Competition between SmbHLH60 and SmMYC2 attenuates a promoter-dependent transcriptional activation of SmTAT1 and SmDFR. (A). Effectors SmbHLH60 and SmMYC2 were co-transformed with reporters SmTAT1pro-LUC. We mixed pHB-SmbHLH60 with pHB-MYC2 and injected it with pGreen0800-SmTAT1-LUC into tobacco epidermal cells and tested the results. (B). Effectors SmbHLH60 and SmMYC2 were co-transformed with reporters SmDFRpro-LUC. pHB-SmbHLH60 and pHB-MYC2 were mixed, which were injected with pGreen0800-SmDFR-LUC into tobacco epidermal cells and detected the results. All data were means of three biological replicates with error bars indicating standard deviation (*, p < 0.05; **, p < 0.01). (C). EMSA was used to detect whether the SmbHLH60-SmMYC2 interaction would affect the ability of SmbHLH60 to bind to the SmTAT1 and SmDFR promoters by EMSA. SmMYC2 and SmbHLH60 were separately or mixed and then incubated with the biotin-labeled probes SmTAT1pro and SmDFRpro for EMSA. (D). The gray value of the bound probe was statistically used to detect whether the SmMYC2-SmbHLH60 interaction would affect the binding of SmbHLH60 to the SmTAT1pro and SmDFRpro. All gray values were means of three biological replicates with error bars indicating standard deviation (*, p < 0.05; **, p < 0.01). (E). qRT-PCR was performed to detect the transcription level of SmMYC2 in SmbHLH60 transgenic plants. The relative quantitative analysis method (2^−ΔΔCT) was used to calculate the relative gene expression, and SmActin was used as the internal reference gene. All experiments were repeated three times, and error bars represented the standard deviation of three biological replicates (*, p < 0.05; **, p < 0.01).

Discussion

MeJA has a central role in plant growth, development and adaptation to the environment [30], [44]. In S. miltiorrhiza, MeJA promotes anthocyanin, tanshinones and phenolic acids accumulation [13], [16], [45], [46]. Additionally, the effect of MeJA was also shown in the gene expression level [12], [22]. However, the molecular mechanism(s) underlying the regulation of secondary metabolites, and more specifically for phenolic acids in S. miltiorrhiza, is still unknown. Generally, MeJA regulates secondary metabolic pathways through a cascade of transcription factors to regulate downstream key enzyme genes, which ultimately leads to changes in the accumulation of metabolites [30]. The molecular mechanism of how these key enzyme genes are regulated by the JA pathway is to be further analyzed.

The bHLH transcription factor, MYC2, was shown to act as a core regulator in the MeJA signaling pathway [44]. In a previous report, it was found that SmMYC2, the core transcription factor for JA signaling, is a positive regulator of SmPAL1, SmTAT1 and SmCYP98A14 (37, 41). However, the mechanism by which SmMYC2 regulates JA-mediated phenolic acid biosynthesis remains unclear. In this study, a novel interactor of SmMYC2 that functions as a negative regulator in phenolic acid biosynthesis. Moreover, SmbHLH60 binds to and represses the promoters of SmTAT1 and SmDFR. SmbHLH60 and SmMYC2 antagonistically interact to regulate SmTAT1 and SmDFR expression. Therefore, we propose that plants regulate the biosynthesis of phenolic acid and anthocyanin mediated by the bHLH complex in S. miltiorrhiza through the JA signaling pathway.

SmbHLH60 may be involved in MeJA-mediated phenolic acid biosynthesis as a negative regulator

MeJA has been proven to be an effective inducer that can cause the accumulation of phenolic acid [13], [16], [23]. To explore the molecular mechanism of phenolic acid accumulation in S. miltiorrhiza which is regulated by MeJA, the MeJA-mediated transcriptome data in S. miltiorrhiza was analyzed. We identified 89 bHLH transcription factors that are down-regulated by MeJA, of which SmbHLH60 was one of the most significantly down-regulated genes (Fig. 1A). SmbHLH92 have been characterized to negatively regulate phenolic acid biosynthesis [8]. Interestingly, SmbHLH92 showed reduced transcription levels in MeJA-mediated RNA-seq data while SmMCY2 increased in RNA-seq data (Fig. S5). The expression pattern of SmbHLH60 after MeJA treatment is consistent with SmbHLH92. These results suggest that SmbHLH60 may be a negative regulator.

In previous research, SmMYB1 has been proved to promote phenolic acid biosynthesis show a consistent expression pattern with the expression of genes in the biosynthetic pathway of phenolic acid increased significantly after MeJA treatment [3]. However, SmbHLH60 showed an expression pattern opposite to key enzyme genes including SmPAL1, SmC4H, SmTAT, SmHPPR, SmRAS and SmCYP98A14. These results indicated that SmbHLH60 may hold a negative regulatory role in the MeJA-mediated biosynthesis of phenolic acid. Therefore, we decided to focus on deciphering the role of this gene in regulating phenolic acid biosynthesis.

SmbHLH60 negatively regulates the biosynthesis of phenolic acid in S. Miltiorrhiza by repressing SmTAT1

It was found that total phenolic acids in hairy roots increased significantly after SmbHLH60 was knocked out, which was 1.48 times that of the control line. Conversely， the content of phenolic acids decreased by 40% compared with the control in the SmbHLH60-OE lines (Fig. 4B). These results showed SmbHLH60 exhibited an opposite effect to SmMYC2. Several important genes of the phenolic acid biosynthesis pathway, including SmPAL1, SmC4H, Sm4CL2, SmTAT1, SmHPPR1, SmRAS1, and SmCYP98A14, were suppressed in the presence of SmbHLH60. Meanwhile, the expression of these genes increased when SmbHLH60 was knocked out (Fig. 4C). The reason for the detection of Sm4CL2 expression is that previous studies have shown that Sm4CL2 may play a more important role in synthesizing phenolic acids [13], [47], [48]. Among them, the expression of SmPAL1, SmTAT1 and SmHPPR1 increased significantly in the SmbHLH60-CRISPR lines. RA and SAB levels were significantly decreased in accordance with the downregulation of SmPAL1 confirming the importance of SmPAL1 in phenolic acid biosynthesis [13]. Overexpression of SmTAT1 led to a significant increase of the RA and SAB content than the wild type [13], [16], [49]. Furthermore, the expression of SmTAT1 and SmHPPR1 was associated with the biosynthesis of RA and SAB after being treated with MeJA or other treatments, suggesting that the tyrosine conversion is a rate-limiting step in the biosynthesis of phenolic acids [13], [16], [50], [51]. These results support the hypothesis that SmbHLH60 acts as a negative regulator of the biosynthesis of phenolic acids.

We detected 7 possible candidate target genes in phenolic acid biosynthesis of SmbHLH60 by dual-LUC (Fig. 6B, S7). Our results showed that SmbHLH60 transcriptionally repressed SmPAL1, SmTAT1 and SmRAS1, which indicated that SmPAL1, SmTAT1 and SmRAS1 might be the target genes of SmbHLH60. Furthermore, Y1H and EMSA results showed that SmbHLH60 bound to the G-box of SmTAT1pro (Fig. 6D, F). Interestingly, SmbHLH60 did not bind to the E-box of SmTAT1pro (Fig. 6D). SmTAT1 is the first committing enzymatic step in the tyrosine pathway [16], [46]. It is worth noting that our investigation demonstrated that SmbHLH60 regulates SmTAT1 expression by direct binding to its G-box (CACGTG) promoter region. These results revealed that SmbHLH60 bound to the promoter of SmTAT1 and suppress its expression to reduce phenolic acid biosynthesis in S. miltiorrhiza.

SmbHLH60 represses the expression of SmDFR to negatively regulate anthocyanin biosynthesis

The total anthocyanins were extracted and it was found that the color of the SmbHLH60 transgenic hairy root extract was significantly different from that of the control (Fig. 5A). Therefore, we are curious whether SmbHLH60 directly caused the difference in the color of anthocyanin extracts. As a result, total anthocyanin content in the knockout line reached twice that of the control line. Conversely, in hairy roots overexpressing SmbHLH60, the total anthocyanin content was only 59% of the control line (Fig. 5B). The expression of key genes in anthocyanin biosynthesis, including SmCHS, SmFLS, SmF3H, SmF3′H, SmF3′5′H, SmANS and SmDFR were decreased to varying degrees in the SmbHLH60-OE lines. Especially, SmFLS, SmANS and SmDFR are significantly suppressed. On the contrary, these genes showed an upward trend in SmbHLH60-CRISPR lines (Fig. 5C). The expressions of SmCHI, SmANS and SmDFR were significantly up-regulated in the SmbHLH60-CRISPR lines. These results indicated that SmbHLH60 negatively regulates biosynthesis not only phenolic acids but also anthocyanins.

Dual-LUC assays verified that SmbHLH60 can transcriptionally repress the expression of SmDFR, but it has no regulatory effect on other several anthocyanin genes (SmF3H, SmCHS, SmANS) (Fig. 8B, S7). Subsequently, we wonder whether SmbHLH60 binds to the G-box binding site in the SmDFR promoter. The dihydroflavonol reductase (DRF) is an important key enzyme in the pathway of anthocyanin biosynthesis [[3], [52], [53]]. For example, the biosynthesis of anthocyanins in apples is regulated by DFR activity [52]. DFR catalyzes a significant step in the biosynthesis of anthocyanins and proanthocyanidins by reducing dihydroflavonols to anthocyanins (53). Besides, SmDFR has been identified in S. miltiorrhiza [22] and few reports that TFs directly regulate anthocyanin biosynthesis have been reported. Our results found that SmbHLH60 directly bound to the G-box but not E-box of the SmDFR promoter. As we predicted, SmbHLH60 directly regulated the expression of SmDFR and reduced total anthocyanins in hairy roots.

In addition to phenolic acid and anthocyanin, we also found that SmbHLH60 may also negatively affect the accumulation of tanshinones. However, the molecular mechanism requires further research in the future (Fig. S8).

SmbHLH60 and SmMYC2 antagonistic function in a bHLH heterodimer to regulate the biosynthesis of phenolic acids and anthocyanins

bHLH-type TFs usually function as homodimer or heterodimer [30]. Although SmbHLH53 has been reported to interact with SmMYC2, the mechanism by which they regulate phenolic acid biosynthesis remains unclear [14]. Different types of transcription factors can also form complexes to regulate downstream target genes. For instance, SmMYB1 has been reported to form a complex with SmMYC2 to regulate SmCYP98A14 and to induce phenolic acid accumulation [3]. In this study, we identified that the SmbHLH60 interacted with SmMYC2 through the Y2H assays (Fig. 7A). BiFC and GST pull-down assays also verified the interaction between SmMYC2 and SmbHLH60 (Fig. 7B-C). These results revealed that SmbHLH60 directly interacted with SmMYC2 to form a heterodimer.

SmMYC2 has been reported to bind to SmTAT1 and promote its transcriptional activation [37]. Interestingly, we found that SmbHLH60 bound to and transcriptionally repressed SmTAT1 expression (Fig. 6B-C). SmbHLH60 repressed the expression of SmDFR in contrast to the results that SmMYC2 promoted the transcriptional activation of SmDFR as was demonstrated by dual-LUC assay (Fig. 8A-B). Y1H results showed that SmMYC2 was bound to the SmTAT1pro and SmDFRpro, which indicated that SmTAT1 and SmDFR were target genes for SmMYC2 (Fig. S10A). Thus, we postulate that SmbHLH60 and SmMYC2 antagonistically regulate SmTAT1 and SmDFR by binding different elements of SmTAT1 and SmDFR. In order to verify our hypothesis, we performed dual-LUC assays by co-injecting SmMYC2 and SmbHLH60 into N. benthamiana leaves with SmTAT1 and SmDFR, respectively. The promotion effects of SmMYC2 on SmTAT1 and SmDRF were impaired when SmbHLH60 was present (Fig. 8A-B). Interestingly, we found that the SmMYC2-SmbHLH60 interaction did not affect the ability of SmbHLH60 to bind to the SmTAT1 and SmDFR promoters (Fig. 8C-D). Y1H results also showed that the binding sites of SmbHLH60 and SmMYC2 to SmTAT1pro and SmDFRpro were different (Fig. S10B). This result also implied that the binding sites of SmMYC2 and SmbHLH60 were not the same although both bound to the promoter of SmTAT1. Furthermore, we found that the expression of SmMYC2 was decreased in SmbHLH60-OE hairy roots, but increased in SmbHLH60-CRISPR hairy roots (Fig. 8E). In the contrast, SmbHLH60 was repressed in SmMYC2-OE lines (Fig. S9). These results indicate that SmbHLH60 and SmMYC2 antagonistically control phenolic acid and anthocyanin biosynthesis by competing with the promoters of SmTAT1 and SmDFR. Furthermore, we found that SmbHLH60 interacted with SmJAZ-1like, SmJAZ8 and SmJAZ9 (Fig. S11A-B). Based on these results, we speculate that SmbHLH60 may form an inhibitory complex with JAZs to participate in phenolic acid biosynthesis. However, how the SmJAZs, SmbHLH60 and SmMYC2 work together to regulate phenolic acid and anthocyanin biosynthesis still needs more exploration in the future.

A new model - SmbHLH60 as a negative regulator for the production of phenolic acids and anthocyanins

We proposed a new hypothetical working model for the SmbHLH60 and SmMYC2 antagonistic regulation of phenolic acid and anthocyanin biosynthesis mediated by MeJA in S. miltiorrhiza (Fig. 9). In the absence of MeJA, SmTAT1 and SmDFR are repressed by the SmbHLH60. Meanwhile, SmbHLH60 and SmMYC2 form a heterodimer, which inhibited the transcriptional activation of SmMYC2 on SmTAT1 and SmDFR. However, MeJA leads to a decrease in the expression of SmbHLH60 while increasing the expression of SmMYC2. Furthermore, these changes in expression result in a reduction of the inhibitory effect of SmTAT1 and SmDFR and enhancement of the transcriptional activation by SmMYC2. Finally, this complexed bHLH-JA dependent regulation results in increased production of phenolic acids and anthocyanins in S. miltiorrhiza. In summary, our discovery provides novel insights into the regulatory mechanism of bHLH-type TFs as heterodimers in the regulation of secondary metabolites which lays a foundation for further research of MeJA-mediated biosynthesis in plants.

Fig. 9.

Proposed mechanism for the regulation of phenolic acids biosynthesis by SmbHLH60 in S. miltiorrhiza. SmbHLH60 and SmMYC2 antagonized the regulation of phenolic acid biosynthesis. Under the treatment of MeJA, SmbHLH60 was suppressed and the expression of SmMYC2 was increased, which led to the increase in the expressions of SmTAT1 and SmDFR, resulting in the production of phenolic acid and anthocyanin increased ultimately.

Conclusions

In this study, a bHLH negatively regulated by MeJA was identified and characterized, named SmbHLH60. SmbHLH60 regulates phenolic acid and anthocyanin biosynthesis in S. miltiorrhiza roots. The expression of SmbHLH60 was negatively correlated to the phenolic acid and anthocyanin concentrations as was shown in the OE and knockout lines. Key biosynthetic genes involved in the anthocyanin and phenolic acid production were up-regulated in the SmbHLH60-CRISPR lines and down-regulated in the SmbHLH60-OE lines. Furthermore, we have elucidated the molecular machinery in which SmbHLH60 regulates secondary metabolites. SmbHLH60 can directly bind SmTAT1 and SmDFR and repress their expression through their promoters. A SmbHLH60-SmMYC2 dimer takes part in the regulation of phenolic acid and anthocyanin biosynthesis in S. miltiorrhiza. These results shed new light on the role of bHLH transcription factors in the biosynthesis of phenolic acids and anthocyanins through heterodimers in S. miltiorrhiza, and provide new insights for the analysis of the MeJA-mediated regulation network of secondary metabolites.

Compliance with Ethics Requirements

This research has fully complied with research ethics.

CRediT authorship contribution statement

Shucan Liu: Conceptualization, Validation, Formal analysis, Investigation, Writing – original draft, Visualization, Writing – review & editing. Yao Wang: Conceptualization, Validation, Writing – review & editing. Min Shi: Formal analysis, Validation, Writing – review & editing. Itay Maoz: Writing – review & editing. Xiankui Gao: Formal analysis, Investigation. Meihong Sun: Investigation. Tingpan Yuan: Validation, Investigation. Kunlun Li: Investigation, Formal analysis. Wei Zhou: Writing – review & editing. Xinhong Guo: Writing – review & editing, Supervision. Guoyin Kai: Conceptualization, Resources, Writing – review & editing, Supervision, Project administration, Funding acquisition.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article: