The seeds of S. miltiorrhiza Bunge were surface-sterilized with 0.1% HgCl₂ and cultured on Murashige and Skoog (MS) medium, supplemented with 30 g/L sucrose and 7 g/L agar (PH = 5.6–5.8). Cultures were cultivated in a growth chamber maintained at 25°C, under 16-h-day and 8-h-night regime. After three subcultures, we chose the seedlings that grew well and samples of vegetative tissue (root and leaf) were collected from plants 30 days after subculture. All the samples were frozen directly into liquid nitrogen and stored at -80°C for further analysis.

The S. miltiorrhiza hairy-root culture was derived after the infection of plantlets with a Ri (root-inducing) T-DNA (transfer DNA)-bearing Rhizobium rhizogenes bacterium (A.T.C.C. 15834; Chen et al., 1999). The S. miltiorrhiza hairy root culture used in this work was maintained on hormone-free 6,7-V medium with 30 g/L sucrose at 25°C in the dark. All experiments in this study were carried out in shake-flask cultures with 500-ml Erlenmeyer flasks on an orbital shaker set at 110–120 rpm. Each flask contained 200 ml 6,7-V liquid medium and inoculated with 2 g fresh weight (fw) of roots from 3- to 4-week-old shake-flask cultures.

The yeast elicitor (YE) was the carbohydrate fraction of yeast extract prepared by ethanol precipitation (Hahn and Albersheim, 1978). In brief, 25 g of yeast extract was dissolved in 125 mL of distilled water, and then mixed with 100 mL of ethanol. The solution was allowed to precipitate for 4 days at 4°C in a refrigerator, and the supernatant was decanted. The precipitate was redissolved in 125 mL of distilled water and subjected to another round of ethanol precipitation. The final precipitate was dissolved in 100 mL of distilled water, sterilized by autoclaving at 120°C for 20 min and stored at 4°C in a refrigerator before use. Abiotic (Ag⁺) was dissolved in distilled water to the concentration of 3 mM and sterilized by filtration (0.22 μm membrane). Combined elicitors of YE+ Ag⁺ were added to the shake-flask culture of S. miltiorrhiza hairy roots on day 18. YE (2 mL) and Ag⁺ (66.7 μL) were added into the culture medium (200 mL). After 0, 1, 2, 4, 8, 12, 24, and 36 h treatment with combined elicitors, hairy roots were collected for SmSQS2 expression analysis. All treatments were performed in triplicate, and the results were averaged.

Total RNA was extracted from roots of S. miltiorrhiza Bunge using a TRIzol Reagents (Invitrogen, USA) following the manufacturer’s instructions. RNA quality was verified by an absorbance at 260 nm and 280 nm optical densities using the NanoDrop 2000 Spectrophotometer (Thermo Scientific, San Jose, CA, USA) and by gel electrophoresis (1% agarose). Total RNA was treated using DNase I with an RNAclean kit (BioTeke, Beijing, China), according to the manufacturer’s instructions, to remove any residual DNA. The first-strand cDNA was synthesized using PrimeScript 1st strand cDNA synthesis kit (TaKaRa Biotechnology, Dalian, China). The cDNA was stored at -80°C for further analysis.

All primers used in this study are listed in Table 1. The full-length SmSQS2 cDNA was amplified using SQS2-F and SQS2-R (Table 1). The PCR products were purified with a EasyPure Quick Gel Extraction Kit (TransGen Biotech, Beijing, China) and cloned into the pEASY-Blunt Simple Vector (TransGen, Beijing, China). Transformation of Escherichia coli DH5α (TransGen, Beijing, China) competent cells was carried out. The recombinant plasmid were isolated and subjected to nucleotide sequencing.

The nucleotide and deduced amino acid sequences were analyzed and sequence comparison was conducted through database search using BLAST tool on NCBI¹. The open reading frame (ORF) was searched using ORF Finder². Putative molecular weight and Theoretical isoelectric point values were calculated using ExPASy tool³. Protein Transmembrane Helical Regions was predicted by TRMHMMserver⁴ v2.0. Three dimensional homologous modeling was analyzed by SWISS-MODEL⁵. Multiple sequence alignment was implemented using DNAMAN and ClustalW software. Phylogenetic analysis was constructed using MEGA6 software with the Neighbor-Joining method. Confidence values for individual branches were measured from 1 000 bootstrap replicates of the original sequence data.

The coding region for SmSQS2 without the C-terminal membrane-anchoring signal was amplified using the forward primer SQS2-cF and SQS2-cR (Table 1). This PCR product was digested with Sal I and Not I and cloned into the Sal I and Not I sites of the expression vector pGEX-4T-1. This newly constructed vector, pGEX-4T-1-SmSQS2ΔTM, was transformed into competent E. coli DH5α. Confirmation that the correct construction had been obtained was determined by sequencing. The pGEX-4T-1-SmSQS2ΔTM was transformed into E. coli BL21 (DE3), which was grown at 37°C until the A₆₀₀ of 0.4–0.6 was reached. The expression was induced with 1 mM Isopropylb-D-1-thiogalactopyranoside (IPTG) at 30°C for 6 h. The cells were resuspended in extraction buffer (100 mmol/l Tris–HCl, pH 7.5, 10 mmol/l MgCl₂, 2% glycerol, and 1 mmol/l DTT) and disrupted by sonication. The lysate was centrifuged at 12,000g for 30 min at 4°C, and the supernatant was loaded on 10% SDS-PAGE gel after denaturating with SDS loading dye at 100°C for 5 min.

The gel was stained with coomassie brilliant blue G-250 and decolorized with mixture (60% Ultrapure water; 30% anhydrous alcohol; 10% acetic acid; Kalra et al., 2013).

Recombinant protein were analyzed by 10% SDS-PAGE gel and transferred onto polyvinylidene fluoride (PVDF) membrane. The membrane was washed with confining liquid (PBS and blocked with 5% non-fat dry milk) for 1 h with 50 rpm in room temperature (RT). Then the membrane was incubated with mouse polyclonal antibodies against GST (1:10,000, EASYBIO., Beijin, China) at 4°C for overnight. After that, the membrane was washed four times with 1x PBST buffer (10 mM Tris-HCl, pH 8.0, 150 mM NaCl and 0.05% Tween 20) for 5 min. After a final wash with 1x PBST buffer, the membrane was incubated for 1 h with the antibody goat anti-mouse IgG (H&L)-HRP conjugated (1:5,000). After that, the membrane was washed three times with PBST buffer (5 min each). The membrane was then visualized using the Super Enhanced chemilumin escence detection kit (Applygen Technologies Inc., Beijing, China) and Fluorescence was detected using Kodak X-OMAT film (CARESTREAM, Xiamen, China). The films were photographed under UMAX PowerLook 2100XL-USB.

The activity of squalene synthase formation was assayed as follows (Ye et al., 2014). The 500 μL reaction mixture contained 100 mM Tris–HCl (pH 7.5), 10 mM farnesyl pyrophosphate triammonium salt (FPP; Sigma–Aldrich, USA), 10 mM MgCl₂, 1 mM DTT, 2% Glycine, 3 mM NADPHNa₄ and 475 μL crude enzyme solution prepared from the E. coli cells. After incubation at 32°C for 10 h, the reaction mixture was extracted three times with 500 μL hexane and three parallel experiments were reacted. This was concentrated using N₂ bubbles and then the concentrated organic phase was subjected to a gas chromatograph-mass spectrometer (GC-MS) for squalene detection.

The GC-MS analysis was performed on a Trace 1310 (Thermo Scientific, San Jose, CA, USA) coupled with a TSQ 8000 mass selective detector (Thermo Scientific, San Jose, CA, USA). Separation of the analyses was performed using a TG-5MS column (30 m × 0.25 mm I.D. × 0.25 μm film thickness; Thermo Scientific, USA) and helium was used as a carrier gas at a constant flow rate of 1.0 mL/min. The column temperature was maintained at 120°C for 3 min, elevated to 180°C at 15°C/min and then elevated to 260°C at 25°C/min. The injection volume was 1 μL for authentic squalene (National Institute for the Control of Pharmaceutical and Biological Products, China) and samples. Mass spectra, 70 eV (in EI mode), ion trap heating, 230°C; scan range, 30–500 m/z. The assay was performed with empty vector as a control.

Total RNA from different tissues (roots and leaves) and different inductive stages was extracted separately using TRIzol reagent (Invitrogen) and pre-treated with RNaseFree DNase (Promega, USA) to eliminate genomic DNA contamination. First-strand cDNA was synthesized using the PrimeScript 1st strand cDNA synthesis kit (TaKaRa Biotechnology, Dalian, China). The real-time quantitative PCR analysis was performed using the SYBR Premix Ex Taq II system (Takara) on an ABI 7500 instrument (Applied Biosystems, Foster City, CA, USA). The primers for qRT-PCR analysis were SQS-qF and SQS-qR (Table 1). Actin was used as the endogenous control to normalize expression value. At least three independent experiments were performed for each analysis. The relative amounts of the target genes were evaluated by the relative expression index of mRNA using the 2[^-ΔΔC(T)] method.

For the cloning of SmSQS2, a pair of primer, designed based on the transcriptome sequencing data of S. miltiorrhiza Bunge, was used for the amplification of the cDNA of SmSQS2. The full-length cDNA of SmSQS2 was 1 597 bp, with a 1 245 bp ORF, which encodes a 414 amino acid polypeptide, flanked by an 115 bp 5′-untranslated region and a 237 bp 3′-untranslated region. The cDNA sequence was deposited into GenBank with the Accession No.: KM408605. The putative SmSQS2 shared 80.72% similarity with SmSQS1, and shared 74.76, 69.47, 78.74, 79.23, 80.24, 81.69, and 80.96% sequence identity with AtSQS1, AtSQS2 (A. thaliana), GgSQS1, GgSQS2 (G. glabra), PgSQS1, PgSQS2, PgSQS3 (P. ginseng), respectively.

The putative SmSQS2 protein had a calculated molecular mass of 47.16 kDa and a theoretical isoelectric point of 7.16. Conserved domain database (CDD) search at NCBI indicated the presence of conserved aspartate-rich regions, substrate binding pocket, substrate-Mg²⁺ binding region suggesting that the protein may encode for functional enzyme (Supplementary Figure S1). Multiple sequence alignments showed that the putative SmSQS2 had high identity with SQSs from other plants and revealed six highly conserved signature domains (I–VI; Figure 1). Domain I is substrate binding pocket or chemical binding site which consists of about 18 amino acids distributed throughout the polypeptide. Domain II is substrate–Mg²⁺ binding site which consists of 10 amino acids divided in two patches of 5 each in which domain V and VI (aspartate rich regions 1 and 2, respectively) are embedded. Domain III is active site, which is composed of about nine amino acid residues. Domain IV is catalytic domain of about 14 amino acids distributed throughout the polypeptide. The catalytic site is composed of the large central cavity formed by antiparallel alpha helices with two aspartate rich regions (DXXXD) on opposite walls (Marrero et al., 1992; Chen et al., 1994; Reipen et al., 1995; Pandit et al., 2000) (Figure 1). These residues are considered to play role in binding of prenyl phosphates by binding Mg²⁺ ions.

Squalene synthase proteins are membrane bound enzymes anchored to the endoplasmic reticulum by their highly hydrophobic transmembrane domain located in their C-terminals (Robinson et al., 1993). This region is poorly conserved among SQS proteins. The TMHMMserver 2.0 showed that the putative SmSQS2 has two transmembrane helices, which were from 281 to 303 and 387 to 409 aa (Figure 1; Supplementary Figure S2).

The phylogenetic tree was constructed using known SQS sequences from various different types of organisms to investigate the evolutionary relations (Figure 2). The phylogenetic tree was divided four main clusters: Plants, Algae, Mammals, Fungi, and SmSQS2 clustered within the Plants. Phylogenetic analysis of putative SmSQS2 showed that it was more closely related to SmSQS1 followed by Artemisia annua.

The secondary structure of the putative SmSQS2 protein was predicted to be mainly α-helix (69.08%), extended strand (7.73%), β-turn (5.80%), and random coil (17.39%). The 3D structure of the putative SmSQS2 was predicted using homology structure modeling on Swiss model server. The sequence used as a template was a human squalene synthase, accession number 3vj9.1. This template showed a sequence identity of 47.59% with the putative SmSQS2 and was the most homologous SQS for which X-ray structure information was available (Supplementary Figure S3).

Earlier reports suggested that deletion of the hydrophobic C-terminal SQS region facilitated expression of soluble protein (Gupta et al., 2012). To obtain soluble recombinant enzymes, the truncated SmSQS2 in which 28 amino acids were deleted from the carboxy terminus was expressed as GST-Tag named pGEX-4T-1-SmSQS2ΔTM. The pGEX-4T-1 vector without the SmSQS2 and transformed in E. coli BL21 (DE3) was used as a control. The recombinant putative protein pGEX-4T-1-SmSQS2ΔTM was induced with 1 mM IPTG at 30°C for 6 h. After sonication and centrifugation of the bacteria, total protein, and supernatants were separated with 10% SDS-PAGE. SDS-PAGE analysis (Figure 3A) showed that the putative protein was successfully expressed in the supernatant and total cell extract, with a molecular mass of approximately 66.46 kDa, while the corresponding protein bands were not found at the expected position (around 66.46 kDa) in the uninduced sample.

Western Blotting showed that the band of recombinant protein (thus including the Gst -tag) was correct. No such band was observed in the control strain. Anti-Gst tag antibody was a ∼66.46 kDa protein band on the membrane (Figure 3B). The result of Western Blot analysis indicated that the anti-Gst tag antibody cross reacted with recombinant putative pGEX-4T-1-SmSQSΔTM giving a fairly detectable cross reactivity band corresponding to 66.46KDa representing the recombinant fusion protein produced in E. coli BL21 (DE3).

To confirm that the SmSQS2 gene encoded functional SQS, the activity of the recombinant protein was measured for conversion of FPP to squalene in the presence of NADPH and Mg²⁺. The catalytic product was identified as squalene by GC-MS analysis (Figure 4). The TIC peak at 18.71 min in the SmSQS2 chromatogram (Figure 4B) corresponds to that at 18.71 min observed in authentic squalene (Figure 4A), while no such peak was detected in the vector-control (Figure 4C). The detection of major mass fragments of m/z = 69 and m/z = 81 in the 18.71 min peak of SmSQS2 is consistent with those seen in authentic squalene, in contrast to no corresponding fragments observed in the vector-control (Figures 4D,E). Thus, these results clearly prove that SmSQS2 codes for squalene synthase.

For the purpose of studying the expression of SQS gene in each tissue, we examined the SmSQS2 expression pattern in different tissues using total RNA from leaves and roots. The expression level of leave was set up as a control and the SmSQS2 expression in other stage was evaluated relative to the leave. Among different tissues of S. miltiorrhiza Bunge, the transcript levels of SmSQS2 were observed in the leaves and roots (Figure 5A). There are significant differences between the expression levels for the two tissues by an unpaired t-test using GraphPad Prism 6. SmSQS2 exhibited the higher expression level in the roots than in the leaves (P < 0.05; Figure 5A).

Moreover, real-time PCR analysis with hairy root at different developmental stages was also performed to examine the changes in the expression of SmSQS2 genes upon biotic (yeast extract) and abiotic (Ag⁺) treatment. Specifically, YE+ Ag⁺ caused a significant increase in SmSQS2 levels in S. miltiorrhiza hairy-root. The expression level of 0 h was set up as a control and the SmSQS2 mRNA expression in other stage was evaluated relative to the 0 h. Interestingly, the dynamic range of the induction varied. The levels of SmSQS2 expression increased at first, peaked at 4 h, then decreased gradually (Figure 5B). It is indicated that YE+ Ag⁺ can induced the expression of SmSQS2.