Background

Baphicacanthus cusia (Nees) Bremek, is widely distributed in Fujian, Yunnan, Sichuan and Guangdong provinces in China and is an important medicinal plant. Indigo naturalis (Qingdai) is made from its leaves and stems, which is known as “Jian Qingdai” and is a famous regional drug in Fujian Province. It is clinically used to treat leukaemia [1], oral cancer [2] and ulcerative colitis [3, 4]. B. cusia roots are used as a valuable drug named “Nan-Ban-Lan-Gen” [5]. To date, the main compounds isolated and identified from B. cusia are indole alkaloids [6], terpenoid alkaloids [7], quinoline ketone alkaloids [8], sterols, flavonoids [9], lignans [10], amino acids, organic acids [11], and plant polysaccharides [9]. Indole alkaloids are the main components of indigo naturalis and the main active constituents of the B. cusia.

The synthesis of indole alkaloids in B. cusia is generally considered to involve the shikimate pathway and the indole pathway [12,13,14], which was based on the knowledge of microbial indigo synthesis. However, the biosynthetic pathway of the indole alkaloids in B. cusia remains unknown. Therefore, we propose a hypothetical biosynthetic pathway of indole alkaloids in vivo according to the microbial synthesis pathway (Additional file 3: Figure S2C).

The enzyme 5-enolpyruvylshikimate-3-phosphate synthase (EPSPS; EC 2.5.1.19) catalyses the transfer of enolpyruvyl moiety of phosphoenol pyruvate (PEP) to 5-hydroxyl shikimate-3- phosphate (S3P) (Additional file 3: Figure S2B), producing shikimic acid and chorismite, which then enters the biosynthetic pathway of indole alkaloids. Therefore, EPSPS is the first key enzyme in this biosynthetic pathway. Recent research has established that glyphosate acts as a competitive inhibitor relative to PEP and binds adjacent to S3P in the active-site of EPSPS; additionally glyphosate maintains the regulatory function of PtrEPSPS in the phenylpropanoid pathway of Populus [15,16,17,18]. However, there are limited studies on the function of BcEPSPS in the metabolic pathway, In the context of overexpression, purification and kinetic characterization of BcEPSPS, this paper will provide a theoretical and experimental basis for explaining the biological synthesis pathway and molecular regulation mechanism of the active compounds in B. cusia.

Results

Isolation and characterization of BcEPSPS

The cDNA of BcEPSPS contains an open reading frame of 1554 nucleotides, and is translated into an amino acid sequence of 518 aa with a calculated molecular mass of 55.33 kDa. The isoelectric point (pI) is 7.55, which illustrated that the BcEPSPS protein is slightly basic, and the protein has an obvious hydrophobicity area of − 0.295 and a hydrophilicity area (Additional file 2: Figure S1D). The BcEPSPS protein has a stable structure, which includes no signal peptide, and a small portion of the trans-membrane topological structure may exist in its secondary structure (Additional file 2: Figure S1E, F). The main secondary structures of the BcEPSPS protein are predicted to involve 42.86% random coil, 32.82% alpha helix, 6.56% beta turn and 17.76% extended strand. Random coils and alpha helices were the most abundant structural elements distributed throughout most parts of the BcEPSPS secondary structure, while beta turns and extend strands were intermittently distributed in the protein (Additional file 2: Figure S1B). It also contains two main structural domain: the EPSPS-synthase active structural domains and the PLN02338 conserved domain (Additional file 2: Figure S1A). The 3D structure of BcEPSPS was predicted and simulated by SWISS-MODEL and phyre2 using sequence homology-based structural modelling (PDB id: 3nvc.1.A) with the consistency of predicted results at 57.35% (Additional file 2: Figure S1G), which showed that it contains the 3-phosphoshikimate-1-carboxyvinyltransferase functional domain. Ramachandran conformation also showed that BcEPSPS has a stable space conformation (Additional file 2: Figure S1H). A phylogenetic tree was constructed for the EPSPS protein family by comparing B. cusia with 13 other species of plants by phylogenetic analysis, which indicated that EPSPS in Dicliptera chinensis appeared to be phylogenetically in the same clade as BcEPSPS (Fig. 1a). The homology comparison of the amino acid sequences indicated that B. cusia shared a high homology with several plants such as Handroanthus impetiginosa, Sesamum indicum, and Calystegia hederacea (Fig. 1b).

Fig. 1
figure 1

Bioinformatics analysis of BcEPSPS. a Phylogenetic tree of the EPSPS protein family from 14 species of plants using MEGA 5.0 software by the neighbour-joining method. b The homology comparison of amino acid alignment of EPSPS in 14 plant species

Full size image

Expression and induction patterns of BcEPSPS

To analyse the expression patterns of BcEPSPS, total RNA was isolated from the roots, stems and leaves of B. cusia and qPCR was performed. The results showed constitutive expression of BcEPSPS in different tissues, with maximum expression in the stem followed by the leaves and the lowest expression in the roots (Fig. 2a). qRT-PCR results showed that BcEPSPS was upregulated by MeJA, SA and ABA, and the expression levels showed obvious variations related to time and phytohormones (Fig. 2b, c, d). For MeJA treatment, the expression level of BcEPSPS rapidly reached a maximum at 4 h, with a 5.79-fold increase and subsequently declined. After treatment with SA, the expression level of BcEPSPS reached the highest level at 6 h (10.65-fold), followed by a short decline and a slight increase until 12 h, then fell to the initial levels. In response to ABA treatment, the expression level of BcEPSPS increased quickly and reached the peak at 4 h which was 7.96-fold higher than that of the control. Then, the expression level fell first and rose again during the 4–8 h period, but it dropped below the initial level at 36 h.

Fig. 2
figure 2

BcEPSPS expression profiles in B.cusia. a The expression profiles of BcEPSPS in different B.cusia tissues, and under phytohormone induction (MeJA (b), SA (c), ABA (d)). Data are represented as the mean ± SEM; *P < 0.05 and, **P < 0.01 compared to the control group

Full size image

Subcellular localization of BcEPSPS

To further verify the expression characteristics of BcEPSPS, we examined the subcellular localization of BcEPSPS in the leaves of N.tabacum, with a control that included an empty pCAMBIA 1301-GFP vector (Fig. 3), Interestingly, the BcEPSPS-GFP fusion protein was localized in both the plastids and cytosol (Fig. 3a-d). In contrast, the pCAMBIA 1301-GFP vector showed florescence throughout the entire cell (Fig. 3e-h).

Fig. 3
figure 3

Subcellular localization of BcEPSPS. a I. indigotica cells expressing BcEPSPS-GFP showed green fluorescent signals mainly in the chloroplast and slightly in the cytoplasm. b The same nucleus and cytoplasm showing the chlorophyll autofluorescence signal in the cell. c Bright-field image. d The merged signal of panels a and b. e) I. indigotica cells expressing GFP showed green fluorescent signals in the cytoplasm. f The same cell showing the chlorophyll auto fluorescence signal in the cytoplasm. g Bright-field image. (h) The merged signal of panels E and F. Bars = 5 nm

Full size image

The expression and identification of the recombinant BcEPSPS protein

To test the function of BcEPSPS, the BcEPSPS recombinant protein was purified through prokaryotic expression and BcEPSPS enzyme activity was determined in vitro. The molecular weights of BcEPSPS and His-tag were ~ 55 kDa and ~ 15 kDa respectively, and the fusion protein BcEPSPS-His was approximately 70 kDa, as predicted (Fig. 4a). Figure 5a shows the fusion protein BcEPSPS-His induced by IPTG (1 mM), and the purified BcEPSPS-His protein showed a single distinct band in lane 4. The protein concentration was determined using the Bradford method. The standard curve equation was Y = 0.8726x + 0.0623, R2 = 0.9981, and the concentration of purified recombinant BcEPSPS-His protein was calculated as 7.96 mg/mL. The result of Western blotting indicated that purified recombinant BcEPSPS-His protein exhibited anti-His antibody immune reactivity (Fig. 4b).

Fig. 4
figure 4

Functional characterization of BcEPSPS in vitro. a SDS-PAGE analysis of BcEPSPS purified by Ni-NTA conjugation. Lane 1, protein molecular weight marker; lane 2, crude enzyme extract from BcEPSPS-pET-32a cell lysate; lane 3, supernatant of pET-32a-BcEPSPS cell lysate; lane 4, purified pET-32a-BcEPSPS protein; lane 5, crude enzyme extract from pET-32a cell lysate. b Western blotting analyses of BcEPSPS. Lane 1, protein molecular weight marker; lane 2, supernatant of pET-32a-BcEPSPS cell lysate; lane 3, purified BcEPSPS protein; lane 4, purified pET-32a protein

Full size image
Fig. 5
figure 5

Enzyme kinetic characterizations of BcEPSPS in vitro. a Calibration curve of inorganic phosphorus concentration. b Km(PEP)curve. c Ki(glyphosate)curve. d IC50 curve

Full size image

The kinetic properties of purified BcEPSPS

Purified BcEPSPS was subjected to enzymatic activity assays under various conditions. As shown in Fig. 5, the Pi standard curve was conducted (Fig. 5a). Then, we measured the velocity of the BcEPSPS catalytic reaction under the constant of S3P at 1.0 mM and the increasing constant of PEP, where the Km(PEP) was 21.267 μM and Vmax was 10.259 U/mg (Fig. 5b). To test Ki(glyphosate), the constant of S3P was 1.0 mM, and the concentration of glyphosate was set as increasing constants. This showed that Ki(glyphosate) was 0.046 μM (Fig. 5c). For IC50, the velocity was measured under various concentrations of PEP (0.05, 0.067, 0.1, and 0.2 mM), while the concentration of glyphosate was set as 10− 5, 10− 4, 10− 3, 10− 2, 10− 1, 1 and 10 mM. Figure 5d shows that the IC50 was 6.37 μM, which indicates that the BcEPSPS protein is type I and is sensitive to glyphosate. Moreover, the kinetic properties of the BcEPSPS protein are represented in Table 1.

Table 1 Kinetic properties of purified recombinant BcEPSPS
Full size table

Improvement in the indigo alkaloid contents by the overexpression of BcEPSPS in I. indigotica

BcEPSPS driven by the double CaMV 35S promoter in the overexpression vector PHB-flag was introduced into I. indigotica to generate OVX-BcEPSPS hairy root cultures by Agrobacterium C58C1. The vector PHB-flag and Agrobacterium C58C1 were individually transferred into I. indigotica to generate the control cultures and the wild-type cultures. I. indigotica hairy roots were induced and cultured for different periods (Fig. 6b). PCR analysis demonstrated that the exogenous BcEPSPS gene was integrated into the transgenic lines, which contained the rolb, rolc and hygromycin resistance (hyg) genes (Fig. 6c). The growth rates of the hairy root lines are shown in Fig. 6d. The OVX-BcEPSPS hairy root lines exhibited a more vigorous growth rate than the Con and CK lines.

Fig. 6
figure 6

Hairy roots of OVX-BcEPSPS in I. indigotica. (a) Schematic representation of the BcEPSPS overexpression vector. T35S, Bar, P35S, CaMV 35S promoter, flag tag, P35S, hygromycin resistance gene. (b) Induction and culture of I. indigotica hairy roots in different periods. a, A sterilised plant of I. indigotica. b, Agrobacterium tumefaciens C58C1 stains carrying plasmids were applied to infect the sterilized leaf. c, The hairy roots were derived from sterile leaves after 2 weeks. d, The false positive hairy roots without the vector C58C1-OVX-BcEPSPS on the 1/2 MS solid medium were supplemented with a gradual decrease in cefotaxime and hygromycin. e, The positive hairy roots containing the vector C58C1-OVX-BcEPSPS showed great growth on the 1/2 MS solid medium with cefotaxime and hygromycin. f, Liquid cultures of positive hairy roots containing the vector C58C1-OVX-BcEPSPS in an Erlenmeyer flask in 45 days. (c) Molecular identification of transgenic hairy roots. Lane 1, DNA size marker, lane 2, the engineered strain as the positive control, lane 3, wild-type control. Lane 4–11, strains of OVX-BcEPSPS. The primers hpt, rolb and rolc were used to check the hygromycin resistance and plasmid Ri. (d) Growth rates of WT, CK and OVX-BcEPSPS hairy root lines

Full size image

QRT-PCR analysis was performed to examine the transcript levels of the endogenous BcEPSPS in positive lines. In comparison to the expression level of the WT controls, the expression level of BcEPSPS in the OVX-BcEPSPS lines was significantly enhanced by 37.87-fold (Fig. 7a). Metabolite analysis indicated that the contents of active compounds such as indigo, secologanin, indole, isorhamnetin, chorismic acid, strictosidine and indolinone were improved in varying degrees in transgenic lines compared with the WT lines and the CK lines. The chromatograms with corresponding retention times of 14 chemical compounds are shown in Additional file 4: Figure S3, in which the detection of signals and the degree of separation is distinct and intuitive. The indigo and secologanin accumulation in the OVX-BcEPSPS lines were significantly higher (5.83- and 5.09-fold) than those in the control lines. Additionally, the indole, isorhamnetin, strictosidine and indole beta-D-glucoside contents were 4.82-, 3.61-, 2.79- and 2.78-fold higher respectively in the OVX-BcEPSPS lines than the WT and CK lines (Fig. 7b). There were no significant differences between the WT and CK lines (Fig. 7b).

Fig. 7
figure 7

Metabolite analysis of transgenic hairy roots. a BcEPSPS transcript abundance in OVX-BcEPSPS transgenic hairy roots by RT-qPCR. b Relative metabolite levels in C58C1, PHB vector and OVX-BcEPSPS transgenic hairy roots. Data are represented as the mean ± SEM; *P < 0.05 and, **P < 0.01 compared to the control group

Full size image

Discussion

EPSP synthases are known to have the unitary function of catalysing shikimate-3-phosphate to 5-enolpyruvylshikimate-3-phosphate in the shikimate pathway in microorganisms and plants [19, 20]. In plants, the shikimate pathway is upstream of the indole pathway and provides chorismic acid for the biosynthesis of indole alkaloid compounds [20]. Here, we provide evidence that EPSP synthase plays an important regulatory role in the accumulation of indole alkaloids and terpenoid alkaloids in B. cusia.

In this research, the 5-enolpyruvylshikimate-3-phosphate synthase gene was isolated and identified from the data of the B. cusia transcriptome, and the characteristics of cDNA sequences were predicted through bioinformatic software. Tissue-specific expression results showed that the BcEPSPS genes were expressed at the highest level in stem tissues, consistent with a recent study, in which the expression level of the BcSK gene in B. cusia under MeJA treatment increased with increasing time [20]. The qPCR results indicated that the BcEPSPS level was upregulated by MeJA, SA and ABA, with variations in the induction time and expression level. Our subcellular localization results demonstrated that the BcEPSPS protein did not accumulate in the plastids but did accumulate in the cytosol, a finding that has not been previously reported. Moreover, a recent study on PtrEPSP-TF protein from Populus, revealed its expression in the nucleus [18]. Although the direct mechanism underlying the observed cytosol presence remains to be determined, one explanation is that the BcEPSPS protein might interact with transcription factors. Our enzymatic activity assay results demonstrated that the ectopically expressed and purified BcEPSPS protein has catalytic activity in vitro. Additionally, BcEPSPS in B. cusia is sensitive to glyphosate, which indicates that the BcEPSPS protein is a class I protein. As several glyphosate-resistant mutants of class I EPSPS have been identified, and most of the class I EPSPS mutants display increased glyphosate resistance [21], it is useful to create a glyphosate-resistant mutant of a B. cusia EPSPS using a directed evolution strategy [22].

Alkaloids are orgainc compounds that contain nitrogen, and are important, effective components of Chinese herbal medicine with significant biological activities [23,24,25]. Indole alkaloids are a class of alkaloids identified by the presence of a structural moiety of indole [26]. The levels of indole alkaloids in B. cusia are higher than that those in other plants such as I. indigotica, Polygonum tinctorium and Indigofera tinctoria, and there was an obvious difference in the contents of indoxyl beta-D-glycoside, indigo and indirubin between different organs [27]. A gene overexpression vector was built for hairy roots, and was transferred by Agrobacterium C58C1. The metabolite analysis indicated that the levels of active compounds such as indigo, secologanin, isorhamnetin and strictosidine were improved to varying degrees in the transgenic lines compared with the controls. Finally, the molecular catalytic characteristics of BcEPSPS in indole alkaloid biosynthesis in B. cusia reported here will provide a unique perspective for functional studies of the shikimate pathway.

The development of functional genomics technologies and genetic transformation technologies has greatly facilitated research on alkaloid biosynthesis [28,29,30]. Additionally, next-generation sequencing provides a fast, high-throughput approach for medicinal plants, for which current genetic information is limited, and is used to obtain batch sequences. Such data will provide a new perspective of molecular research on B. cusia, for which the main active substances are indole alkaloids, and will aid in protecting germplasm resources of B. cusia by sorting and revealing functional genes.

Conclusions

The enzyme 5-enolpyruvylshikimate-3-phosphate synthase is involved in catalysing indole alkaloid biosynthesis. BcEPSPS was first isolated and identified from the data of the B. cusia transcriptome, and the function and characteristics of BcEPSPS were further verified in vitro. High concentrations of indole alkaloids were produced in the hairy root cultures of BcEPSPS overexpression. It has been suggested that through the regulation of BcEPSPS, the accumulation of effective components can be improved.

In conclusion, a hairy root overexpression culture system with a high concentration of indoline accumulation was established in this study, and it is expected to alleviate the over-exploitation and shortage of high quality resources of B. cusia in the future.

Methods

Plant materials and treatments

Samples of B. cusia were obtained from the experimental field of Xianyou county in Putian city, China. Species verification was performed by Professor Daozhi Wei of the College of Life Science, Fujian Agriculture and Forestry University (FAFU). These samples were preserved in the laboratory of medicinal plant resources in FAFU (No.004512-ML). I. indigotica plants were cultivated in the experimental field of the Second Military Medical University (SMMU), Shanghai, China. Species verification was performed by Professor Lei Zhang of the School of Pharmacy, SMMU. These samples were preserved in the laboratory of medicinal plant resources in SMMU (No. S1011078). The sterilized seedlings of I. indigotica were cultured on Murashige and Skoog (MS) medium solidified with 0.6% agar. The organ-specific series of samples (root, stem and leaf, 6 months old) were used for RNA extraction. Leaves of the B. cusia were sprayed with 0.1 mM methyl jasmonate (MeJA), salicylic acid (SA), and abscisic acid (ABA) for stress treatments [31], and leaf samples were harvested at 0, 2, 4, 6, 8, 12, 24 and 36 h post-treatment. Three independent biological replicates for each group were performed. MeJA, SA and ABA were purchased from Sigma-Aldrich (USA). All samples were immediately frozen in liquid N2 and stored at − 80 °C for further research.

Total RNA isolation and first-strand cDNA synthesis

Total RNA of B. cusia plants was extracted using the TransZol Up Plus RNA Kit (TransGen Biotech). One microgram of RNA was used to prepare first strand cDNA using the TransScript First-Strand cDNA Synthesis Super Mix (TransGen Biotech). The concentration and quality of RNA and first-strand cDNA samples were examined by spectrophotometer analysis using a NanoDrop 2000C Spectrophotometer (Thermo) and agarose gel electrophoresis with ethidium bromide.

Isolation and characterization of BcEPSPS

The BcEPSPS gene sequence was isolated from a transcriptome database of B. cusia, with the gene-specific primers EPSPS-F and EPSPS-R (Additional file 1: Table. S1). The full-length coding region of BcEPSPS was cloned by PCR with KD Plus DNA Polymerase (TransGen Biotech) (Additional file 5: Figure S4). PCR was performed under the following conditions: denaturation at 94 °C for 5 min, 35 cycles of 94 °C for 30 s, 55 °C for 30 s and 68 °C for 2 min, and a final extension at 68 °C for 10 min. After gel electrophoresis detection, the PCR products were purified and cloned into a pBlunt-Zero vector (TransGen Biotech) and transformed into Trans1-T1 cells (TransGen Biotech) for sequencing.

Bioinformatics analysis of BcEPSPS

The open reading frame of BcEPSPS was searched and translated on the NCBI ORF Finder (http://www.ncbi.nlm.nibi.gov/gorf/gorf.htmL). The composition and physicochemical properties of BcEPSPS were calculated with Protparam (http://www.expasy.ch/tools/protparam.htmL). The hydrophobicity/hydrophilicity of the BcEPSPS amino acid sequence was predicted with ProtScale (http://www.expasy.ch/tools/protscale.htmL). The signal peptide and leading peptide of the BcEPSPS protein were analysed on Signal 3.0 (http://www.cbs.dtu.dk/services/Signal/) and TargetP 1.1 (http://www.cbs.dtu.dk/services/TargetP/) respectively. The prediction of the transmembrane domains and coiled-coils of the BcEPSPS protein were accomplished with TMPred (http://www.ch.embnet.org/software/TMPRED_form.htmL) and COILS (http://www.ch.embnet.org/software/COILS_form.htmL) respectively. The secondary structure and functional domains of the BcEPSPS protein were detected the SOPMA (http://npsapbil.ibcp.fr/cgibin/npsa_automat.pl/page=/NPSA/npsa_sopma.htmL), Pfam24.0 (http://www.Pfam.sanger.ac.uk/search) and BLAST (http://www.ncbi.nlm.nibi.gov/Structure/cdd/wrpsb.cgi). The model of the three-dimensional structure and Ramachandran conformation were completed on Swiss-Model (http://www.expasy.ch/swissmod/SWISS-MODEL.htmL) and PyMOL respectively. The homology comparison and phylogenetic tree were generated by MEGA5.0.

Quantitative real-time PCR (qPCR)

qPCR analysis was performed to determine the expression features of BcEPSPS and was conducted on a Thermal Cycler Dice Real Time PCR machine (TaKaRa, Japan), using the TransStart Top Green qPCR SuperMix Kit (TransGen Biotech) according to the manufacturer’s instructions. The first strand cDNA was generated by the TransScript One-Step gDNA Removal and cDNA Synthesis SuperMix (TransGen Biotech), by adding Oligo (dT) primer, gRemover, E-mix, and R-mix to 1 μg of total RNA and the manufacturer’s protocol was followed. qPCR was performed by using the 2-ΔΔCt method under the following conditions: denaturation at 94 °C for 30 s, followed by 45 cycles of 94 °C for 5 s and 60 °C for 30 s, followed by a final dissociation stage. The expression levels were normalized with the 18S housekeeping gene [32]. Primers for qPCR are listed in Additional file 1: Table S1.

Subcellular localization in the cells of Nicotiana tabacum leaves

To confirm the subcellular localization of BcEPSPS, the encoding region of BcEPSPS was cloned into the pCAMBIA1301-GFP vector (Additional file 6: Figure S5). The primers for constructing the vector are listed in Additional file 1: Table S1. Vectors BcEPSPS-GFP and GFP were transiently expressed in N.tabacum (20 days old) for the observation of subcellular localization. The plants were incubated for 24 h in darkness and the leaf sections were observed using a confocal laser scanning microscope (Nikon, Japan) [33].

Purification and activity assays of the BcEPSPS protein

The encoding region of BcEPSPS was cloned into pET-32a (Novagen) (Additional file 7: Figure S6) and expressed in E. coli strain BL21. The primers for constructing the vector and for sequencing are listed in Additional file 1: Table S1. The cells were grown overnight at 37 °C in 200 mL of Luria-Bertani broth containing 75 μg/L ampicillin to an OD600 of 0.6, and the cells were further induced with 1 mM isopropylthio-b-galactoside at 16 °C for 24 h. After sonication lysis and centrifugation - the supernatant of the extraction buffer containing the crude extracted enzyme was purified at 4 °C using nickel-chelate affinity chromatography (Bio-Rad Laboratories, USA). The protein concentration was measured with the Bradford Protein Assay Kit (Sangon). The extraction buffer containing the purified enzyme was added to glycerin to 10% (w/v) and stored at − 80 °C. Western blotting was used to determine the induced expression of His-tagged EPSPS with an anti-rabbit mAb His antibody (Cell Signaling Technology), and a secondary antibody (Anti-rabbit IgG, HRP-linked Antibody).

Analysis of enzyme activity

BcEPSPS activity was determined by measuring the release of inorganic phosphate in each reaction using the malachite green dye assay method [34]. The reaction was assayed in 100 μl of 50 mM HEPES/NaOH (pH 7.0), 1 mM PEP, 1 mM S3P, and 0.75 μg of purified enzyme. After incubation at 30 °C for 3 min, 800 μl of malachite green/ammonium molybdate colorimetric solution was added. One min later, the colour reaction was stopped by adding 800 μl (34%) sodium citrate solution. The absorbance at 660 nm was measured after incubation at room temperature for 0.5 h [35, 36].

Kinetic characterization was performed in a reaction buffer containing different concentrations of PEP or glyphosate. The kinetic data were fit to the appropriate equations using the program SigmaPlot (SPSS Science, Chicago, 1 L). The Km value was generated by fitting the data to V = Vmax [S]/(Km + [S]), where V is the velocity of the reaction (U/mg), Vmax is the maximum velocity, Km is the Michaelis constant and [S] is the concentration of the substrate. The Ki value was determined by a previously described method [37]. The S3P concentration was fixed at 1 mM, and the glyphosate concentration was fixed at 0, 0.05, 0.1 and 0.2 μM. The IC50 of enzyme activity was determined by fitting the data to V = Vmin + (Vmax – Vmin)/(1 + ([I]/IC50)s), where [I] is the concentration of glyphosate, s is the slope of the curve at IC50 [38], and V was determined at 1 mM PEP and S3P with different glyphosate concentrations ranging from 0.00001 mM to 10 mM.

Generation and analysis of BcEPSPS overexpression in hairy roots

The encoding region of BcEPSPS was cloned into binary vector PHB-flag, which is used as an overexpression vector with two CaMV 35S promoters [39] (Fig. 6a) (Additional file 8: Figure S7). The primers for constructing the vector are listed in Additional file 1: Table S1. To induce I. indigotica hairy roots, Agrobacterium tumefaciens strain C58C1 (containing a Ti plasmid) with plasmids BcEPSPS-PHB and PHB (control check, CK) were applied to infect the sterilized leaf [40, 41]. The infected leaves were placed on 1/2 MS solid medium supplemented with 500 mg·mL− 1 cefotaxime. After 2 weeks, the transgenic hairy roots were derived from sterile leaves that were approximately 2–3 cm long. The hairy roots were then excised and cultured on 1/2 MS solid medium supplemented with 500 mg·mL− 1 cefotaxime and 10 mg·mL− 1 hygromycin. The 1/2 MS solid medium was replaced every 2 weeks with a gradually decreasing concentration of cefotaxime (300, 200,100 0 mg·mL− 1). BcEPSPS-PHB lines were labelled OVX-EPSPS and the blank C58C1 line was labelled wild-type (WT). Approximately 100 mg of the transgenic hairy roots and WT lines were cultivated in 250 mL Erlenmeyer flasks containing 200 mL of fresh liquid 1/2 MS medium supplemented with 10 mg mL− 1 hygromycin on an orbital shaker maintained at 110 rpm and 25 °C in darkness.

The hairy roots were subcultured every 9 days and harvested at 45 days for DNA and RNA extraction and metabolite analysis. Additionally, the fresh weights of the hairy roots were recorded at days 0, 9, 18, 27 and 45 to determine the biomass growth rate.

Analysis of BcEPSPS-PHB in hairy roots

Genomic DNA of positive transgenic hairy roots was extracted using the Plant Genomic DNA Kit (TransGen Biotech), with primers for detection listed in Additional file 1: Table S1. Primer Rbcsr was used for testing the presence of BcEPSPS, while primers hpt, rolb and rolc were used for checking the hygromycin resistance and plasmid Ri.

qRT-PCR was performed to analyse the expression of BcEPSPS in positive transgenic hairy roots. The expression levels were normalized with the Actin housekeeping gene [42].

Extraction and determination of indigo alkaloids

Positive transgenic hairy roots were collected and dried at 40 °C for 2 days and ground into a fine powder. Two hundred milligrams of dry power was extracted with 5 mL of methanol and trichloromethane (1:1) via sonication for 1 h, the supernatant was transferred and the powder was extracted with 5 mL of reagents again. The supernatant was pooled and filtered through a 0.22 μm organic membrane. Five millilitres of the extracting solution was evaporated to dryness and redissolved in 200 μl methanol.

HPLC analysis was performed on an Agilent Technologies 1260 infinity (Agilent, USA) [43]. The mobile phase was 0.1% solution of formic acid (A phase) and 0.1% formic acid in methanol (B phase) (HPLC grade). The solvent gradient programme used was as follows: 0–15 min, 40% A and 60% B; 15–20 min, 60% A and 40% B; 20–30 min, 80% A and 95% B. The reaction monitoring mode for HPLC analysis involved the following conditions: Diamonsil C-18 (4.6 μm × 250 mm, 5 μm), a flow rate of 1.0 mL/min, a column temperature of 25 °C and an injection volume of 10 μL. All standards were purchased from Sigma-Aldrich (St. Louis, MO).

Statistical analysis

Experiments were performed in triplicate, and all results were expressed as the mean ± SEM. Statistical analysis was performed with SPSS 13.0 software. A Student’s t-test was used to compare two groups. One-way ANOVA followed by the Dunnett post hoc test was used for multiple comparisons versus the control group (GraphPad Software). p < 0.05 and p < 0.01 were set as the criterion for statistical significance.