Molecular cloning and metabolomic characterization of the 5-enolpyruvylshikimate-3-phosphate synthase gene from Baphicacanthus cusia

Background Indigo alkaloids, such as indigo, indirubin and its derivatives, have been identified as effective antiviral compounds in Baphicacanthus cusia. Evidence suggests that the biosynthesis of indigo alkaloids in plants occurs via the shikimate pathway. The enzyme 5-enolpyruvylshikimate-3-phosphate synthase (EPSPS) is involved in plant metabolism; however, its underlying putative mechanism of regulating the production of indigo alkaloids is currently unknown. Results One gene encoding EPSPS was isolated from B. cusia. Quantitative real-time PCR analysis revealed that BcEPSPS was expressed at the highest level in the stem and upregulated by methyl jasmonate (MeJA), salicylic acid (SA) and abscisic acid (ABA) treatment. The results of subcellular localization indicated that BcEPSPS is mainly expressed in both the plastids and cytosol, which has not been previously reported. An enzyme assay revealed that the heterogeneously expressed BcEPSPS protein catalysed the generation of 5-enolpyruvyl shikimate-3-phosphate. The overexpression of BcEPSPS in Isatis indigotica hairy roots resulted in the high accumulation of indigo alkaloids, such as indigo, secologanin, indole and isorhamnetin. Conclusions The function of BcEPSPS in catalysing the production of EPSP and regulating indigo alkaloid biosynthesis was revealed, which provided a distinct view of plant metabolic engineering. Our findings have practical implications for understanding the effect of BcEPSPS on active compound biosynthesis in B. cusia.


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].
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.

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-1carboxyvinyltransferase 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).

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.

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).

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, R 2 = 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).

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 IC 50 , 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 IC 50 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.

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 wildtype 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.
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-Dglucoside 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).

Discussion
EPSP synthases are known to have the unitary function of catalysing shikimate-3-phosphate to 5-enolpyruvylshikim ate-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  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.

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.

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-ΔΔ Ct 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 pCAM-BIA1301-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 IC 50 of enzyme activity was determined by fitting the data to V = Vmin + (Vmax -Vmin)/(1 + ([I]/IC 50 )s), where [I] is the concentration of glyphosate, s is the slope of the curve at IC 50 [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 ttest 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.