Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

UPLC/Q-TOF MS-Based Metabolomics and qRT-PCR in Enzyme Gene Screening with Key Role in Triterpenoid Saponin Biosynthesis of Polygala tenuifolia

  • Fusheng Zhang ,

    Contributed equally to this work with: Fusheng Zhang, Xiaowei Li

    Affiliation Modern Research Center for Traditional Chinese Medicine Shanxi University, Taiyuan, Shanxi, China

  • Xiaowei Li ,

    Contributed equally to this work with: Fusheng Zhang, Xiaowei Li

    Affiliations Modern Research Center for Traditional Chinese Medicine Shanxi University, Taiyuan, Shanxi, China, College of Chemistry and Chemical Engineering, Shanxi University, Taiyuan, Shanxi, China

  • Zhenyu Li,

    Affiliation Modern Research Center for Traditional Chinese Medicine Shanxi University, Taiyuan, Shanxi, China

  • Xiaoshuang Xu,

    Affiliations Modern Research Center for Traditional Chinese Medicine Shanxi University, Taiyuan, Shanxi, China, College of Chemistry and Chemical Engineering, Shanxi University, Taiyuan, Shanxi, China

  • Bing Peng ,

    pengbing123@hotmail.com (BP); dugh@imm.ac.cn (GD)

    Affiliation Beijing Institute of Traditional Chinese Medicine, Beijing Hospital of Traditional Chinese Medicine Affiliated to Capital Medical University, Beijing, China

  • Xuemei Qin,

    Affiliation Modern Research Center for Traditional Chinese Medicine Shanxi University, Taiyuan, Shanxi, China

  • Guanhua Du

    pengbing123@hotmail.com (BP); dugh@imm.ac.cn (GD)

    Affiliations Modern Research Center for Traditional Chinese Medicine Shanxi University, Taiyuan, Shanxi, China, Institute of Materia Medica, Chinese Academy of Medical Sciences, Beijing, China

UPLC/Q-TOF MS-Based Metabolomics and qRT-PCR in Enzyme Gene Screening with Key Role in Triterpenoid Saponin Biosynthesis of Polygala tenuifolia

  • Fusheng Zhang, 
  • Xiaowei Li, 
  • Zhenyu Li, 
  • Xiaoshuang Xu, 
  • Bing Peng, 
  • Xuemei Qin, 
  • Guanhua Du
PLOS
x

Abstract

Background

The dried root of Polygala tenuifolia, named Radix Polygalae, is a well-known traditional Chinese medicine. Triterpenoid saponins are some of the most important components of Radix Polygalae extracts and are widely studied because of their valuable pharmacological properties. However, the relationship between gene expression and triterpenoid saponin biosynthesis in P. tenuifolia is unclear.

Methodology/Findings

In this study, ultra-performance liquid chromatography (UPLC) coupled with quadrupole time-of-flight mass spectrometry (Q-TOF MS)-based metabolomic analysis was performed to identify and quantify the different chemical constituents of the roots, stems, leaves, and seeds of P. tenuifolia. A total of 22 marker compounds (VIP>1) were explored, and significant differences in all 7 triterpenoid saponins among the different tissues were found. We also observed an efficient reference gene GAPDH for different tissues in this plant and determined the expression level of some genes in the triterpenoid saponin biosynthetic pathway. Results showed that MVA pathway has more important functions in the triterpenoid saponin biosynthesis of P. tenuifolia. The expression levels of squalene synthase (SQS), squalene monooxygenase (SQE), and beta-amyrin synthase (β-AS) were highly correlated with the peak area intensity of triterpenoid saponins compared with data from UPLC/Q-TOF MS-based metabolomic analysis.

Conclusions/Significance

This finding suggested that a combination of UPLC/Q-TOF MS-based metabolomics and gene expression analysis can effectively elucidate the mechanism of triterpenoid saponin biosynthesis and can provide useful information on gene discovery. These findings can serve as a reference for using the overexpression of genes encoding for SQS, SQE, and/or β-AS to increase the triterpenoid saponin production of P. tenuifolia.

Introduction

Radix Polygalae, which is the root of Polygala tenuifolia Willd. or P. sibirica L. (Fam. Polygalaceae) as listed in Chinese Pharmacopoeia (2010 edition), has been used in traditional medicine for tonic, sedative, antipsychotic, and expectorant purposes for centuries in China [1], [2]. The studies on the chemical components of this herb have mainly focused on various saponins, xanthones, and oligosaccharide esters [3][6]. Among them, triterpenoid saponins are documented as an important group of active components. The structures of the triterpenoid saponins in P. tenuifolia are rather complex [7]. The triterpenoid saponins have a triterpene presenegenin aglycone as basic structure, and the slight differences among these saponins are a result of various substitutes including saccharides and/or acyl groups linked to C-28. Considering the complicated isolation and complete structural characterization of triterpenoid saponins, more methods need to be established to analyze the chemical constituents of triterpenoid saponins in P. tenuifolia. In recent years, UPLC coupled with electrospray ionization quadrupole time-of-flight mass spectrometry (Q-TOF MS) has been widely used as a powerful tool in the analysis of complex saponin metabolites because this tool shortens the analysis time and provides accurate mass measurements [8]. Our report, which is an investigation of constituents in P. tenuifolia by ultra-performance liquid chromatography (UPLC) /Q-TOF MS, aims to analyze the peak area intensity differences of triterpenoid saponins among roots, stems, leaves, and seeds, and quantify their metabolites using metabolomic technology.

Modern pharmacological studies have already demonstrated that these triterpenoid saponins exhibited a wide variety of activities, such as cognitive improving, anti-angiogenic, hypolipidemic, anti-inflammatory, anti-depressive, anxiolytic, sedative-hypnotic, and neuroprotective actions [2], [9][16]. Although various chemical and pharmacological properties of triterpenoid saponins have been extensively studied, there is no study that has focused on the biosynthetic pathway of the triterpenoid saponins in Radix Polygalae. The method of terpenoid biosynthesis in plants is through the isoprenoid pathway by cyclization of 2, 3-oxidosqualene to primarily give cucurbitane skeleton. Two biosynthetic pathways are possible for presenegenin skeleton: the mevalonate (MVA) in cytosolic pathway and the non-mevalonate or 2-C-methyl-D-erythritol 4-phosphate/1-deoxy-D-xylulose 5-phosphate (MEP/DOXP) in plastidial pathway [17]. The MEP pathway is absent in higher animals and fungi, but in green plants, the MEP and MVA pathways co-exist in separate cellular compartments [18]. The triterpenoid saponin skeleton may undergo various oxidations, substitution, and glycosylation mediated by cytochrome P450-dependent monooxygenases (CYP 450s), glycosyltransferases (UGTs), and other enzymes [19]. To date, only a few publications that report the identification of CYP 450s and UGTs involved in biosynthesis of triterpenoid saponins are available. Characterized CYP450s and UGTs, such as CYP 93E1, CYP 88D6, CYP 51H10, CYP 72A154, CYP 716A12, UGT 73P2, UGT 74M1, UGT 73K1, and UGT 73F3, have been reported to be involved in modification of triterpenoid saponins [20]. In nucleotide database of NCBI from P. tenuifolia (up to 2014.02), only three unigenes and two complete cDNAs had been identified and whether enzymes have an important function in the biosynthesis pathway of triterpenoid saponins is still unknown. If we continue to figure out the whole biosynthesis pathway of triterpenoid saponins, many enzyme genes in triterpenoid saponin biosynthesis are needed to be screened by the gene expression of these enzymes for further analysis.

Gene expression analysis has been frequently used to analyze mRNA in different organisms, tissues, and in several transgenic and gene mutation experiments [21][24]. Individual RNAs can be measured by Northern blot [25], RNase protection analysis [26], and in situ hybridization [27]. However, these techniques are time-consuming and/or require large quantities of RNA for detection [28]. Quantitative real-time polymerase chain reaction (qRT-PCR) has a broad range of detection with advantages in repeatability, sensitivity, and specificity. It is also valuable and easy to use when studying rare transcripts and detecting low abundance mRNAs [29]. Given the substantial variations in RNA's stability, quantity, and purity, and the efficiency of reverse transcription and PCR, the reliability of qRT-PCR results is highly dependent on the reference genes chosen. However, many of these reference genes may not be stably expressed under all conditions [30][32]. Some reference genes have been reliably validated in one plant species but not well suited in others [33]. Therefore, experimental proof is suggested when reference genes are to be used under a new experimental condition or in a new plant species [34]. Moreover, given the limitations of Sanger sequencing [36][37], expressed sequence tag (EST) analysis has identified only five mRNAs, which are all involved in biosynthesis of the saponins. The next-generation sequencing technology, known as RNA-seq, emerged as an effective approach for high-throughput sequence, now facilitates studies of transcriptome in a rapid way and has been used to explore gene structure and expression profiling on traditional Chinese medicine [38][39]. In our previous research, the transcriptome of Radix Polygalae was evaluated by Illumina HiSeq 2000 (data not shown), which provides a basis for further research about gene expression analysis of P. Tenuifolia.

In the present study, the differences among chemical components of triterpenoid saponins in different tissues from the roots, stems, leaves, and seeds of P. tenuifolia were identified by UPLC/Q-TOF MS. qRT-PCR was further performed to screen reference genes (Table S1) and measure the expression of nine relevant genes (Table S2) in triterpenoid saponin backbone biosynthesis pathway (from acetyl-CoA or pyruvate to sapogenin) of P. tenuifolia. The expression levels of three CYP 450s and three UGTs genes (Table S3) were also analyzed. Those six genes were all from the CYP 450s and UGTs families which were identified to be involved in biosynthesis of triterpenoid saponins and were also found in the transcriptome of Radix Polygalae. This study aimed to discover the differential gene expression in different tissues, and explore the excellent genetic traits in the triterpenoid saponin biosynthesis pathway. The results may provide useful information for deeper understanding of the biosynthesis pathway of triterpenoid saponins in P. tenuifolia.

Results

Metabolite analysis of P. tenuifolia by UPLC/Q-TOF MS

The UPLC system provides a rapid, effective, and convenient method to analyze the chemical constituent's variance between different tissues in P. tenuifolia, whereas Q-TOF MS provides the accurate mass measurement. Figure 1 shows definite differences among the four samples from a visual examination of MS total ion current (TIC) chromatograms. A total of 25 compounds, including 4 xanthone C-glycosides, 6 sucrose esters, 8 oligosaccharide multi-esters, and 7 triterpene saponins, were identified. The details of identified compounds are summarized in Table 1. Figure 2 shows the typical MS/MS spectrum of Onjisaponin TG. The observation of [M-H-Glc-H2O-CO2-Fuc-Rha-Xyl-Api-HMG-CH2O] at m/z 455.3153 (Y1) and [M-H-Glc-H2O-CO2-Fuc-Rha-Xyl-Api-HMG] (Y2) at m/z 425.3045 in the MS/MS spectrum can be considered as the diagnostic ions for this type of triterpene saponins, and the fragment ions at m/z 1235.5663, corresponding to the losses of 144 Da proved the occurrence of 3-hydroxy-3-methyl-5-pentanoic acid ester (X1).

thumbnail
Figure 1. MS TIC chromatograms of roots (A), stems (B), leaves (C), and seeds (D) of P. tenuifolia by UPLC/Q-TOF MS in negative ion mode.

https://doi.org/10.1371/journal.pone.0105765.g001

thumbnail
Figure 2. Extracted ion chromatogram (A) and MS/MS spectra (B) of Onjisaponin TG.

Glc = β-D-glucopyranosyl; Fuc = β-D-fucopyranosyl; Xyl = β-D-xylopyranosyl; HMG = D 3-hydroxy-3-methyl-5-pentanoic acid ester.

https://doi.org/10.1371/journal.pone.0105765.g002

Multivariate statistical analyses

The normalized data sets from each sample column contained 1826 variables. The resulting integral data were imported in SIMCA-P for multivariate analysis. PCA was used first to detect any inherent trends within the data as an unbiased statistical method. PCA score plots showed that the groups of roots, stems, leaves, and seeds were clearly clustered into four groups (Figure 3A). To further study the variance between the different tissues and to determine potential biomarkers, supervised PLS-DA was subsequently used. The PLS-DA model was also validated by a permutation test with 200 permutations (Figure 3B) (R2 X = 0,929, Q2 = 0.925). The results indicated that the PLS-DA pattern combined with the loading plot was suitable to find out potential biomarkers.

thumbnail
Figure 3. PCA scores plots derived from UPLC/Q-TOF MS data for extracts obtained from four different tissues (A); Permutation test with 200 permutations of PLS-DA model (B); PLS-DA loadings plot obtained from the metabolic profiles of four different tissues (C); Peak area intensity of triterpenoid saponins differences between the four tissue groups (D).

* The stems, leaves, and seeds group compared with the roots group, p<0.05.

https://doi.org/10.1371/journal.pone.0105765.g003

Potential biomarkers

Based on the normalized data of UPLC/Q-TOF MS separation, the loading plots (Figure 3C) were constructed for PLS-DA. The variables with VIP value >1.000 were marked as potential biomarkers. As shown in Table 2, a total of 22 marker compounds accountable for the different metabolite profiles of the roots, stems, leaves and seeds were observed. Sibiricoses A1, Sibiricoses A5, and Lancerin were the compounds particularly responsible for these variations.

thumbnail
Table 2. VIP values of PCA and peak area intensity of tissue metabolites measured by UPLC/Q-TOF MS.

https://doi.org/10.1371/journal.pone.0105765.t002

The main objective of this experiment is to focus on the changes of triterpenoid saponins between the different tissues. Five of the seven triterpenoid saponins had important functions in the intergroup differences. Furthermore, the peak area intensity of all seven triterpenoid saponins was analyzed (Figure 3D). In this experiment, the peak area intensity of most marker compounds were higher in the roots (p<0.05), whereas Lancerin and 7-O-Methylmangiferin were higher in the stems, leaves, and seeds (Table 2).

Screening of reference genes of P. tenuifolia

Seven candidate traditional reference genes including 18s ribosomal RNA (18s RNA), crystallinum ubiquitin-conjugating enzyme (UBC 2), actin 11 (ACT 11), glyceraldehyde-3-phosphate dehydrogenase (GAPDH), alpha tubulin (TUA), elongation factor 1-alpha (EF1α), and actin 1 (ACT 1) were selected (Table 3). Based on Figure 4A, the expression levels of the seven reference genes in different tissues varied extensively with Ct values ranging from 8 to 30 cycles. Most abundantly transcribed was 18s RNA with Ct value of less than 10 cycles. The rest of the genes were moderately expressed mRNAs with Ct values between 20 and 27 cycles.

thumbnail
Figure 4. Expression levels of candidate reference genes in the different tissues by qRT-PCR (A); Average expression stability values (M) of the candidate reference genes calculated by geNorm (B); Gene expression stability and ranking of reference genes as calculated by NormFinder (C).

https://doi.org/10.1371/journal.pone.0105765.g004

The geNorm software was used to analyze the gene expression stabilities of the seven reference genes. This software calculates the average expression stability value (M) based on the average pair-wise variation between all genes tested [43]. The gene with the lowest M value illustrates the most stable expression, whereas the highest M value indicates least stable expression. Except for UBC2, all genes had M value below the geNorm threshold of 1.5, whereas ACT 11 and GAPDH had the lowest M value (0.613) (Figure 4B). Therefore, ACT 11 and GAPDH had the most stable expression and UBC2 had the least stable expression.

The NormFinder software, whose criteria is the same with geNorm, was used to confirm results obtained by the geNorm program. Similar with the geNorm method, the gene with the lowest M value has the most stable expression, and the gene with the highest M value has the least stable expression. Results of NormFinder were showed in Figure 4C, GAPDH and ACT 11 had the most stable expression in different tissues. Stable genes identified in geNorm were confirmed by NormFinder, and GAPDH was chosen as the reference gene in the different tissues of P. tenuifolia.

qRT-PCR analysis for genes involved in triterpenoid saponin biosynthesis pathways

To explore the mechanism of the triterpenoid saponin changes in the different tissues and understand its biosynthesis pathway in P. tenuifolia, the 15 genes (Table 3) putatively expressed in triterpenoid saponin biosynthesis pathway were analyzed by qRT-PCR. As shown in Figure 5A, among the nine genes involved in triterpenoid saponin backbone biosynthesis, phosphomevalonate kinase (PMK), SQS, SQE, and β-AS genes in the roots were significantly overexpressed with 2 to 13-fold higher expression than the other tissues (p<0.05). The expression level of 4-hydroxy-3-methylbut-2-enyl diphosphate reductase (HDR), (E)-4-hydroxy-3-methylbut-2-enyl-diphosphate synthase (HDS), and farnesyl diphosphate synthase (FPS) in the stems were significantly higher compared with that in the roots (p<0.05). Moreover, cycloartenol synthase (CAS) genes exhibited the highest expression in the seeds.

thumbnail
Figure 5. Expression pattern of genes involved in triterpenoid saponin backbone biosynthesis pathway in the different tissues of P. Tenuifolia by qRT-PCR (A); QRT-PCR analysis of CYP450s and UGTs in the different tissues of P. Tenuifolia (B).

* The stems, leaves, and seeds group compared with the roots group, p<0.05.

https://doi.org/10.1371/journal.pone.0105765.g005

Furthermore, the correlation coefficients between the peak area intensity of triterpenoid saponins and expression levels were computed by SPSS. The expression levels of SQS, SQE, and β-AS showed high correlation with the peak area intensity of triterpenoid saponins in different tissues (Table 4). The transformation of trans,trans-farnesyl diphosphate (FPP) to (S)-2,3-oxidosqualene requires the function of SQS and SQE [35]. The expression of SQE in the roots was significantly over expressed with an average of six fold compared with the other tissues (p<0.05). The next important step in the pathway is through β-AS and CAS, which catalyzes 2,3-oxidosqualene to form triterpenoid saponins and sterol, respectively [36]. The expression level of CAS was higher in the leaves and seeds compared with the root (p<0.05), whereas the β-AS expression in the root was higher (p<0.05) and overexpressed with an average of 13 fold compared with the stems, leaves, and seeds.

thumbnail
Table 4. The correlation coefficient between the peak area intensity of triterpenoid saponins and genes expression.

https://doi.org/10.1371/journal.pone.0105765.t004

The correlation between the gene expression level and peak area of compound also represents an approach to determine the possible CYP 450 and UGT genes involved in the triterpenoid saponin biosynthesis. However, no apparent regularity in the expression of CYP88D6, CYP716B1, CYP72A1, UGT74B1, UGT73B2, and UGT73C6 (Figure 5B) was observed. Results demonstrated that UGT73B2 mRNA levels were highest in the roots and the expression levels of CYP88D6, CYP716B1, CYP72A1, UGT74B1, and UGT73C6 were highest in leaves, respectively. However, the expression of these genes was significantly different from SQS, SQE, and β-AS. In addition, the correlation coefficient between the gene expression level and peak area intensity of triterpenoid saponins were less than 0.50 (Table 4).

Discussion

The recent advances in metabolomics have provided new insights in the meaningful biological changes of numerous metabolites using liquid chromatography-mass spectrometry (LC-MS), gas chromatography-mass spectrometry (GC-MS), and nuclear magnetic resonance (NMR) technologies. In this study, LC-MS metabolomic profiles have clearly demonstrated the changes of secondary metabolites particularly the triterpenoid saponins in the different tissues of P. tenuifolia. Furthermore, the nine enzyme genes involved in the triterpenoid saponin backbone biosynthesis and six enzyme genes involved in the oxygenation and glycosylation of cyclization product (Figure 6) were analyzed. The understanding of the biological significance of the alteration of these triterpenoid saponins and genes may provide evidence for further discovery of enzyme gene with key role in P. tenuifolia.

thumbnail
Figure 6. Putative triterpenoid saponin biosynthesis pathway in P. tenuifolia.

https://doi.org/10.1371/journal.pone.0105765.g006

UPLC/Q-TOF MS analysis and accumulation of triterpenoid saponins

Considering the complexity of the plant metabolite composition, plant metabolomic analyses have many challenges. Compared with GC-MS and NMR technologies, HPLC technology has an advantage on the analysis of plant secondary metabolites. Currently, HPLC ultraviolet detector (HPLC-UV), HPLC evaportive light scattering detector (HPLC-ELSD), HPLC photodiode array detector (HPLC-PDA), and UPLC/Q-TOF MS methods have been developed for the qualitative and/or quantitative analysis of plant secondary metabolites. Among these analytical methods, the UPLC coupled with Q-TOF MS has become a powerful tool for metabolite profiling and identifying the complicated components in herbal medicine [44][45]. In previous studies, positive and negative ion modes were investigated. The result were that the detection of saponins is less difficult using negative ion mode and the confirmation of molecular ions or quasimolecular ions for the identification of each peak [42] is also easier. Thus, mass data monitored with negative ion mode were utilized for component characterization.

Triterpenoid saponins, as a category of secondary metabolites, indirectly participate in plant growth and development. To protect the plant from herbivore and to be part of the plants' defense systems (anti-microbial, anti-fungicidal, and insecticidal) are the most important functions of triterpenoid saponins in plants [46]. The triterpenoid saponin accumulations are influenced by several environmental factors, such as seasonal variation, nutrient, light irradiation, or its combined effects. The distribution of triterpenoid saponins also demonstrates seasonal fluctuations or greatly varies in the individual plant organs or tissues. For instance, Lin et al. [47] found that peak area intensity of saponins in the tubers of Dioscorea pseudojaponica is the maximum protection provider for this reproductive organ. Saponins, whose levels were increased in the early stages of berry development of Phytolacca dodecandra, also prevent fruit loss and ensure seed maturation [48].

In this study, result of metabolomic analysis showed a clear separation in the PCA space of the different tissues of P. tenuifolia analyzed by UPLC/Q-TOF MS. Considerable information in biomarkers were obtained through accurate mass measurement and MS/MS analysis. The specific accumulation of triterpenoid saponins in the root of P. tenuifolia indicated that it might counteract soil-borne fungi, which is similar with the accumulation of saponins in the root of Avenae strigosa [49]. Moreover, the different peak area intensity and distribution of triterpenoid saponins in different tissues provide an excellent model to research its biosynthesis pathway and screen enzyme gene with key role in its biosynthesis.

Selection of reference genes in P. tenuifolia

The qRT-PCR technique is one of the most common methods to quantify gene expression levels in order to understand biological processes. However, a reliable method to normalize its expression is necessary. Thus, an appropriate internal reference gene is required for reliable quantification of gene transcripts [50]. We examined the expression levels of seven reference genes obtained from transcriptome data of P. tenuifolia in different tissues. We also analyzed their expression using geNorm and NormFinder. Based on these softwares, data showed that top two positions of the reference genes for the different tissue samples were similar. Because GAPDH is a commonly used reference gene to normalize the gene expression data in qRT-PCR assay [51], it was also chosen as the reference gene in the different tissues of P. tenuifolia.

Biosynthesis of triterpenoid saponin backbone

Current ideas consider that the early stage of triterpenoid saponin biosynthesis in plants starts from the cytosolic MVA and plastidial MEP pathway [17]. The MVA pathway converted acetyl-CoA to isopentenyl diphosphate (IPP), whereas MEP pathway started from pyruvate to IPP and then to dimethylallyl diphosphate (DMAPP) (Figure 6). However, less progress has been made toward determining the precise source of isoprene units, and clarifying what the pathway is important. In this study, we determined the gene expression of MK and PMK of MVA pathway, and HDR and HDS of MEP pathway. Results showed that gene expression levels of MK and PMK were generally high in the roots, whereas HDR and HDS were higher in the stems. The correlation coefficient showed that gene expression of MK and PMK had higher correlation with the peak area intensity of triterpenoid saponins than HDR and HDS (Table 4). Result suggested that MVA pathway has more important functions in the triterpenoid saponin biosynthesis of P. tenuifolia.

Geranyl diphosphate (GPP) is obtained through the condensation of IPP and DMAPP, and FPP, which is the common precursor of the vast array of sesquiterpenes, resulted from the addition of a second IPP unit. Two trans,trans-FPP molecules are facilitated by SQS to produce squalene. Squalene is oxidized by SQE to (S)-2,3-oxidosqualene, which is the first step leading to cyclizations and common precursor of phytosterols and triterpenoid saponins [51]. Moreover, β-AS facilitates (S)-2,3-oxidosqualene backbone into a chair–chair–chair conformation in contrast with the chair–boat–chair conformation during CAS catalysis. Based on the experimental results, the gene expression levels of SQS, SQE, and β-AS in the root were significantly higher than in the other tissues. With regard to the total peak area intensity of triterpenoid saponins, the correlation coefficients between the above three gene expression levels and peak area intensity were higher than the others. This result indicated that SQS, SQE, and β-AS have important functions in the triterpenoid saponin biosynthesis pathway of P. tenuifolia. Mangas et al. [52] suggested that SQS and β-AS have regulatory function by comparing triterpenoid saponin peak area intensity with the expression level of some related genes in the cauli of Centella asiatica. Han et al. [53] demonstrated that squalene enhances the accumulation of SQE with SQS, dammarenediol II synthase (DDS), β-AS, and cycloartenol synthase (PNX) by RT-PCR analysis in Panax ginseng. Considering the present experiment result, the overexpression of SQS, SQE, and β-AS genes may be increasing the triterpenoid saponin production in P. tenuifolia.

Design of the cyclization product - oxygenation and glycosylation

As compared with onjisaponin backbone biosynthesis, less information about the late stages (from sapogenin to various saponin) of triterpenoid saponin biosynthesis is available. Most common sapogenin modifications are oxygenation catalyzed by CYP450s protein families and glycosylation catalyzed by UGTs at various positions of the backbone [54][55]. Prior to this study, few publications report the identification of CYP450s and the involvement of UGTs in the biosynthesis of triterpenoid saponins (Table 5), and the identified genes from P. tenuifolia have not yet been accounted.

thumbnail
Table 5. CYP450s and UGTs involved in triterpenoid saponin biosynthesis.

https://doi.org/10.1371/journal.pone.0105765.t005

In our previous research, a total of 466 and 157 gene reads are annotated as CYP450s and UGTs in the transcriptome of P. tenuifolia (data not shown), respectively. Based on the transcriptome data, six candidate genes, including CYP450 and UGT, which are in the same family of the identified P450s and UGTs involved in the biosynthesis of triterpenoid saponins, are selected for qRT-PCR analysis in different tissues. Seki et al [61] have reported that CYP88D6 expression is detected in the roots and stolons by RT-PCR, but no amplification is observed in the leaves or stems of Glycyrrhiza (licorice) plants. However, in the current study, the CYP88D6 expression in the leaves or stems was observed and did not show much difference among the different tissues. Carelli et al [60] have identified that CYP716A12 is involved in saponin synthesis of Medicago truncatula using a combined genetic and biochemical approach. However, in vitro enzymatic activity assays indicate that CYP716A12 catalyzes the oxidation of β-amyrin and erythrodiol at the C-28 position, generating oleanolic acid. Nevertheless, the gene expression of CYP716B1, from the same family of CYP716A12, was detected in the roots, stems, leaves, and seeds of P. tenuifolia and showed no correlation with triterpenoid saponin peak area intensity in the present research. Moreover, Seki et al [61] demonstrated that CYP72A154 expressed in an engineered yeast strain, which endogenously produced 11-oxo-b-amyrin (a possible biosynthetic intermediate between b-amyrin and glycyrrhizin) and catalyzed 3 sequential oxidation steps at C-30 of 11-oxo-b-amyrin, may supply in situ to produce glycyrrhetinic acid. This research also showed that CYP72A154 was expressed in the roots, stolons, and stems, whereas no transcripts were observed in the leaves of Glycyrrhiza plants [61]. However, in this study, the transcripts of CYP72A1 were not only detected in the roots, seeds, and stems, but also in the leaves of P. tenuifolia.

Regarding CYP450s genes, some UGT genes have also been cloned and identified. Research showed that UGT74M1 was identified by screening UGT sequences among ESTs of Saponaria vaccaria and catalyzed linkage of glucosyl-moieties to the C28-carboxy group of preferably oleanane-type sapogenins through an ester bond [63]. Furthermore, UGT73F3 was identified in Medicago truncatula and catalyzed glucosylation of hederagenin and other oleanane sapogenins at the C28-carboxy group [59]. In addition, the possible involvement in the glycosylation of brassinosteroids of other characterized UGTs of this family such as UGT73C5, UGT73C6, UGT73C8, UGT73F1, and UGT73P1 have also been reported [65]. The majority of UGTs involved in glycosylation belong to the UGT73 and UGT74 families. In the current study, the gene expression of UGT74B1, UGT73B2, and UGT73C6 did not show much correlation with the peak area intensity of triterpenoid saponins, which is similar with three CYP450 genes in P. tenuifolia.

Conclusions

In summary, significant differences in all 7 triterpenoid saponins among the different tissues were found. GAPDH was chosen as the reference gene in the different tissues of P. tenuifolia. Results from this study suggested that MVA pathway has more important functions and SQS, SQE, and β-AS are the enzyme genes with key role in the triterpenoid saponin biosynthesis of P. tenuifolia. This result was established through the correlation analysis between the triterpenoid saponin peak area intensity and expression levels of 15 relevant genes involved in the biosynthesis. The metabolomic analysis by UPLC/Q-TOF MS combined with the gene expression analysis by qRT-PCR technique provide useful information in studying gene discovery and an effective approach in understanding the mechanism of onjisaponin biosynthesis. Along with the technology advances and its broader availability, the digital gene expression profile (DEG)-based generation sequencing or DNA microarray-based gene expression cluster analysis may also facilitate gene discovery.

Materials and Methods

Materials and reagents

The fresh roots, stems, leaves, and seeds from the samples of P. tenuifolia, were collected in Xiongshan forest farm, mountain of Wulong, Bayi town, Changzhi county, Shanxi province (latitude: N35°57′59.98″; longitude: E113°01′36.58″). No specific permits were required from the forest farm to select samples. The forest farm is not privately-owned and the field studies did not involve protected species. Each sample was divided into two parts, in which one part was used for UPLC/Q-TOF MS analysis. The other parts were collected in RNase-free tubes and stored frozen in liquid nitrogen at −80°C.

Acetonitrile of UPLC grade were obtained from Fisher Scientific Co., Ltd. (Waltham, MA, USA). Formic acid and analytic-grade methanol were purchased from Fisher Scientific Co., Ltd. (Waltham, MA, USA). RNAiso Plus kit, PrimerScript RT master mix perfect real-time kit, and SYBR Premix Ex Taq II kit were purchased from TAKARA Co., Ltd. (Dalian, China). Primers were synthesized from Sangon Biotech (Shanghai, China).

Sample preparation for UPLC/Q-TOF MS analysis

Roots, stems, leaves, and seeds of 9 individual P. tenuifolia plants were sequentially pulverized in a mortar with liquid nitrogen. A total of 1 g drug powder of each sample was extracted for 30 min by ultrasonication in 10 mL of methanol, and then the sample solution was filtered through a 0.22 µm filter membrane before use. A 2 µL aliquot was used for UPLC/Q-TOF MS analysis.

UPLC/Q-TOF MS analysis

P. tenuifolia metabolite profiling was performed on a Waters ACQUITY UPLC I-Class/Xevo in line with a Waters Xevo G2 Q-TOF mass spectrometer (Waters Corporation, Milford, MA, USA). Samples were separated on a Zorbax Eclipse Plus Acquity UPLC BEH C 18 (1.7 µm particle size) 2.1 mm×50 mm. The column temperature was maintained at 40°C. The flow rate was 0.50 mL/min. The mobile phase consisted of acetonitrile (A) and water containing 0.1% (v/v) formic acid. The gradient was initiated with increasing to 5% A within 1 min, 5% A and then linearly increased to 12% A within 2 min, then 12% A increased to 20% A within 3 min, 20% A increased to 30% A within 4 min, and 30% A increased to 35% A within 0.5 min. Next, 35% A increased to 45% A within 3 min and 45% A was increased to 50% A within 1.5 min. Finally, 50% A increased to 95% A within 3 min and 95% A increased to 100% A within 1 min.

The Xevo G2-S Q-TOF mass spectrometer was run in negative mode. The data were collected for each test sample from 50 Da to 1500 Da. High purity nitrogen (N2) was used as nebulizing gas, and ultra-high pure helium (He) as collision gas. Source parameters were as follows: capillary voltage, 2.50 kV; sampling cone voltage, 40.0 V; source offset, 100 V; desolvation temperature, 500°C; cone gas flow, 50 L•h−1; and desolvation gas flow, 800 L•h−1. To ensure mass accuracy and reproducibility, leucine-enkephalin was used as the reference lock mass (m/z 554.2615) with the LockSpray interface.

UPLC/Q-TOF MS data acquisition and analysis

Waters' metabolomics package, namely TransOmics, developed with Nonlinear Dynamics, can process large sample sets, automatically detect features, and perform a quantitative comparison with retention time (RT) and m/z data pairs as identifiers. The result of roots, stems, leaves and seeds samples were imported to SIMCA-P13.0 software package (version 13.0, Umetrics, Umeå, Sweden) for multivariate statistical analysis.

Principal component analysis (PCA) and partialleast-squares-discriminant analysis (PLS-DA) were run to separate between different tissue groups. Pareto scaling was used in all the models to avoid chemical noise. To evaluate the quality of the model, R2 and Q2 values were also calculated. R2 displays the variance explained in the model and indicates the goodness of fit. Q2 displays the variance in data predicted by the model and indicates predictability. Potential biomarkers were selected according to the loading plot and Variable Importance in the Project (VIP) values. Compound identification of metabolites was performed based on their retention times, exact mass information and fragment ions according to previous studies [5], [7], [8], [66], [67]. The peak area intensity analysis of potential biomarkers and triterpenoid saponins were also run to find the variance between different tissue groups.

Total RNA extraction and cDNA synthesis.

Total RNA was extracted from the different tissues by using RNAiso Plus kit according to the manufacturer's protocol. RNA concentration and purity were estimated from the optical density at A260/A280 ratio using the NanoDrop 2000 Spectrophotometer (Thermo Fisher Scientific, Wilmington, DE). The integrity of RNA was evaluated by electrophoresis. cDNA was amplified from the total RNA (500 ng) by using a PrimerScript RT master mix perfect real-time kit.

qRT-PCR analysis

qRT-PCR was performed using the ABI StepOne/StepOne Plus (Applied Biosystems, USA) with a final volume of 20 µL containing cDNA, SYBR Green PCR master mix, ROX and primers. All qRT-PCR reactions were carried out in technical and biological triplicate. The amplification procedure included three stages: a holding stage for 10 min at 95°C; a cycling stage with 40 cycles consisting of 15 s at 95°C for denaturation and 60 s at 60°C for primer annealing; and a melt-curve stage consisting of 15 s at 95°C, 1 min at 60°C and 15 s at 95°C [68].

Melt-curve analysis was performed to confirm amplicon specificity. All the primers for the 7 reference genes and 15 putative genes involved in triterpenoid saponin biosynthesis used for amplification were designed using Primer 5 software, using the following criteria: melting temperatures of 59°C to 61°C, primer lengths of 21 bp to 25 bp and amplicon lengths of 80 bp to 202 bp (Table 1). The gene used as an internal control was chosen by the mean threshold cycle (Ct) values. Data were then presented according to the 2−ΔΔCT method [69]. Measures were obtained for each condition from cDNA, which had been synthesized from RNA extracted from three independent cultures, each in two biological replicates.

Statistical analyses

To select the suitable reference gene, GeNORM (version 3.5) [31] and NormFinder (version 0.953) [32] were used to analyze the stability of gene expression. The 2−ΔΔCT data obtained from qRT-PCR were expressed as mean ±SD and analyzed in SPSS 16.0 software (SPSS Inc., Chicago, IL). Correlation analysis between peak area intensity of triterpenoid saponins and gene expression levels were performed, and differences were scored as statistical significance at p<0.05 by independent sample t-test by SPSS 16.0.

Supporting Information

Table S1.

Summary of the annotation sources for reference genes of P. Tenuifolia.

https://doi.org/10.1371/journal.pone.0105765.s001

(DOC)

Table S2.

Summary of the annotation sources for relevant genes in triterpenoid saponins backbone biosynthesis pathway of P. Tenuifolia.

https://doi.org/10.1371/journal.pone.0105765.s002

(DOC)

Table S3.

Summary of the annotation sources for CYP450s and UGTs genes of P. Tenuifolia.

https://doi.org/10.1371/journal.pone.0105765.s003

(DOC)

Author Contributions

Conceived and designed the experiments: FZ. Performed the experiments: XL BP. Analyzed the data: XL BP. Contributed reagents/materials/analysis tools: ZL BP. Contributed to the writing of the manuscript: FZ XL XX XQ GD.

References

  1. 1. Willow M, Carmody J, Carroll P (1980) The effects of swimming in mice on pain perception and sleeping time in response to hypnotic drugs. Life Sci 26(3): 219–224.
  2. 2. Yao Y, Jia M, Wu JG, Zhang H, Sun LN, et al. (2010) Anxiolytic and sedative-hypnotic activities of polygalasaponins from Polygala tenuifolia in mice. Pharm Biol 48(7): 801–807.
  3. 3. Jiang Y, Tu P (2002) Xanthone O-glycosides from Polygala tenuifolia. Phytochemistry 60 (8): 813.
  4. 4. Jiang Y, Tu P (2005) Four new phenones from the cortexes of Polygala tenuifolia. Chem Pharm Bull 53 (9): 1164–1166.
  5. 5. Jiang Y, Zhang W, Tu P, Xu X (2005) Xanthone Glycosides from Polygala tenuifolia and Their Conformational Analyses. J Nat Prod 68 (6): 875–879.
  6. 6. Kawashima K, Miyako D, Ishino Y, Makino T, Saito K-i, et al. (2004) Anti-stress effects of 3, 4, 5-trimethoxycinnamic acid, an active constituent of roots of Polygala tenuifolia (Onji). Bio Pharm Bull 27 (8): 1317–1319.
  7. 7. Liu J, Yang X, He J, Xia M, Xu L, et al. (2007) Structure analysis of triterpene saponins in Polygala tenuifolia by electrospray ionization ion trap multiple-stage mass spectrometry. J Mass Spectrom 42 (7): 861–873.
  8. 8. Ling Y, Li Z, Chen M, Sun Z, Fan M, et al. (2013) Analysis and detection of the chemical constituents of Radix Polygalae and their metabolites in rats after oral administration by ultra high-performance liquid chromatography coupled with electrospray ionization quadrupole time-of-flight tandem mass spectrometry. J Pharmaceut Biomed Anal 85: 1–13.
  9. 9. Arai M, Hayashi A, Sobou M, Ishida S, Kawachi T, et al. (2011) Anti-angiogenic effect of triterpenoidal saponins from Polygala senega. J Nat Med-Toyko 65 (1): 149–156.
  10. 10. Wang H, Gao J, Zhu D, Yu B (2006) Two new triterpenoid saponins isolated from Polygala japonica. Chem Pharm Bull 54 (12): 1739–1742.
  11. 11. Li C, Yang J, Yu S, Chen N, Xue W, et al. (2008) Triterpenoid saponins with neuroprotective effects from the roots of Polygala tenuifolia. Planta Med 74 (02): 133–141.
  12. 12. Li H, Wang QJ, Zhu DN, Yang Y (2008) Reinioside C, a triterpene saponin of Polygala aureocauda Dunn, exerts hypolipidemic effect on hyperlipidemic mice. Phytother Res 22 (2): 159–164.
  13. 13. Cheng M-C, Li C-Y, Ko H-C, Ko F-N, Lin Y-L, et al. (2006) Antidepressant principles of the roots of Polygala tenuifolia. J Nat Prod 69 (9): 1305–1309.
  14. 14. Ikeya Y, Takeda S, Tunakawa M, Karakida H, Toda K, et al. (2004) Cognitive improving and cerebral protective effects of acylated oligosaccharides in Polygala tenuifolia. Bio Pharm Bull 27 (7): 1081–1085.
  15. 15. Choi JG, Kim HG, Kim MC, Yang WM, Huh Y, et al. (2011) Polygalae radix inhibits toxin-induced neuronal death in the Parkinson's disease models. J Ethnopharmacol 134 (2): 414–421.
  16. 16. Cheong M, Lee S, Yoo H, Jeong J, Kim G, et al. (2011) Anti-inflammatory effects of Polygala tenuifolia root through inhibition of NF-κB activation in lipopolysaccharide-induced BV2 microglial cells. J Ethnopharmacol 137 (3): 1402.
  17. 17. Lichtenthaler HK (1999) The 1-deoxy-D-xylulose-5-phosphate pathway of isoprenoid biosynthesis in plants. Annu Rev Plant Biol 50 (1): 47–65.
  18. 18. Tong W-Y (2013) Biotransformation of Terpenoids and Steroids. In Natural Products, Springer: pp 2733–2759.
  19. 19. Haralampidis K, Trojanowska M, Osbourn AE (2002) Biosynthesis of triterpenoid saponins in plants. InHistory and Trends in Bioprocessing and Biotransformation, Springer: pp 31–49.
  20. 20. Augustin JM, Kuzina V, Andersen SB, Bak S (2011) Molecular activities, biosynthesis and evolution of triterpenoid saponins. Phytochemistry 72 (6): 435–457.
  21. 21. Woltedji D, Song F, Zhang L, Gala A, Han B, et al. (2012) Western honeybee drones and workers (Apis mellifera ligustica) have different olfactory mechanisms than eastern honeybees (Apis cerana cerana). J Proteome Res 11 (9): 4526–4540.
  22. 22. Wang Y, Wang X, Lee TH, Mansoor S, Paterson AH (2013) Gene body methylation shows distinct patterns associated with different gene origins and duplication modes and has a heterogeneous relationship with gene expression in Oryza sativa (rice). New Phytol 198 (1): 274–283.
  23. 23. Yang R, Wang X (2013) Organ evolution in angiosperms driven by correlated divergences of gene sequences and expression patterns. Plant Cell 25 (1): 71–82.
  24. 24. Kumar S, Asif MH, Chakrabarty D, Tripathi RD, Dubey RS, et al. (2013) Differential expression of rice lambda class GST gene family members during plant growth, development, and in response to stress conditions. Plant Mol Biol Rep 31 (3): 569–580.
  25. 25. Alwine JC, Kemp DJ, Stark GR (1977) Method for detection of specific RNAs in agarose gels by transfer to diazobenzyloxymethyl-paper and hybridization with DNA probes. P Natl Acad Sci 74 (12): 5350–5354.
  26. 26. Lashbrook CC, Tieman DM, Klee HJ (1988) Differential regulation of the tomato ETR gene family throughout plant development. Plant J 15 (2): : 243–252.
  27. 27. Diethard T, Christine P (1989) A non-radioactive in situ hybridization method for the localization of specific RNAs in Drosophila embryos reveals translational control of the segmentation gene hunchback. Chromosoma 98: 81–85.
  28. 28. Li H, Qin Y, Xiao X, Tang C (2011) Screening of valid reference genes for real-time RT-PCR data normalization in Hevea brasiliensis and expression validation of a sucrose transporter gene HbSUT3. Plant Sci 181 (2): 132.
  29. 29. Bustin S, Benes V, Nolan T, Pfaffl M (2005) Quantitative real-time RT-PCR–a perspective. J Mol Endocrinol 34 (3): 597–601.
  30. 30. Suzuki T, Higgins P, Crawford D (2000) Control selection for RNA quantitation. Biotechniques 29 (2): 332–337.
  31. 31. Lin YL, Lai ZX (2010) Reference gene selection for qPCR analysis during somatic embryogenesis in longan tree. Plant Sci 178 (4): 359–365.
  32. 32. Czechowski T, Stitt M, Altmann T, Udvardi MK, Scheible W-R (2005) Genome-wide identification and testing of superior reference genes for transcript normalization in Arabidopsis. Plant Physiol 139 (1): 5–17.
  33. 33. Reid KE, Olsson N, Schlosser J, Peng F, Lund ST (2006) An optimized grapevine RNA isolation procedure and statistical determination of reference genes for real-time RT-PCR during berry development. BMC Plant Biol 6 (1): 27.
  34. 34. Lia H, Qina Y, Xiaoa X, Tanga C (2011) Screening of valid reference genes for real-time RT-PCR data normalization in Hevea brasiliensis and expression validation of a sucrose transporter gene HbSUT3. Plant Sci 181: 132–139.
  35. 35. Kim TD, Han JY, Huh GH, Choi YE (2010) Expression and functional characterization of three squalene synthase genes associated with saponin biosynthesis in Panax ginseng. Plant Cell Plysic 52 (1): 125–137.
  36. 36. Morozova O, Hirst M, Marra MA (2009) Applications of new sequencing technologies for transcriptome analysis. Annu Rev Genom Hum G 10: 135–151.
  37. 37. Brent MR (2008) Steady progress and recent breakthroughs in the accuracy of automated genome annotation. Nat Rev Genet 9 (1): 62–73.
  38. 38. Tang Q, Ma X, Mo C, Wilson IW, Song C, et al. (2011) An efficient approach to finding Siraitia grosvenorii triterpene biosynthetic genes by RNA-seq and digital gene expression analysis. BMC Genomics 12 (1): 343.
  39. 39. Sun C, Li Y, Wu Q, Luo H, Sun Y, et al. (2010) De novo sequencing and analysis of the American ginseng root transcriptome using a GS FLX Titanium platform to discover putative genes involved in ginsenoside biosynthesis. BMC Genomics 11 (1): 262.
  40. 40. Ikeya Y, Sugama K, Okada M (1991) Four new phenolic glycosides from Polygala tenuifolia. Chem Pharm Bull 39 (10): 2600–2605.
  41. 41. Miyase T, Iwata Y, Ueno A (1992) Tenuifolioses GP, oligosaccharide multi-esters from the roots of Polygala tenuifolia WILLD. Chem Pharm Bull 40 (10): 2741–2748.
  42. 42. Miyase T, Iwata Y, Ueno A (1991) Tenuifolioses GP, Tenuifolioses AF, oligosaccharide multi-esters from the roots of Polygala tenuifolia Willd. Chem Pharm Bull 39 (11): 3082–3084.
  43. 43. Vandesompele J, De Preter K, Pattyn F, Poppe B, Van Roy N, et al. (2002) Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol 3 (7): 1–12.
  44. 44. Miao WY, Wang Q, Bo T, Ye M, Qiao X, et al. (2013) Rapid characterization of chemical constituents and rats metabolites of the traditional Chinese patent medicine Gegen-Qinlian-Wan by UHPLC/DAD/qTOF-MS. J Pharmaceut Biomed Anal 72: 99–108.
  45. 45. Zhou S, Cao J, Qiu F, Kong W, Yang S, et al. (2013) Simultaneous Determination of Five Bioactive Components in Radix Glycyrrhizae by Pressurised Liquid Extraction Combined with UPLC–PDA and UPLC/ESI–QTOF–MS Confirmation. Phytochem Analysis 24 (6): 527–533.
  46. 46. Sparg S, Light M, Van Staden J (2004) Biological activities and distribution of plant saponins. J Ethnopharmacol 94 (2): 219–243.
  47. 47. Lin J-T, Chen S-L, Liu S-C, Yang D-J (2009) Effect of harvest time on saponins in Yam (Dioscorea pseudojaponica Yamamoto). J Food Drug Anal 17: 116–22.
  48. 48. Ndamba J, Lemmich E, Molgaard P (1994) Investigation of the diurnal, ontogenetic and seasonal variation in the molluscicidal saponin content of Phytolacca dodecandra aqueous berry extracts. Phytochemistry 35 (1): 95–99.
  49. 49. Papadopoulou K, Melton R, Leggett M, Daniels M, Osbourn A (1999) Compromised disease resistance in saponin-deficient plants. P Natl Acad Sci 96 (22): 12923–12928.
  50. 50. Hong S-Y, Seo PJ, Yang M-S, Xiang F, Park C-M (2008) Exploring valid reference genes for gene expression studies in Brachypodium distachyon by real-time PCR. BMC Plant Biol 8 (1): 112.
  51. 51. Sando T, Takeno S, Watanabe N, Okumoto H, Kuzuyama T, et al. (2008) Cloning and characterization of the 2-C-methyl-D-erythritol 4-phosphate (MEP) pathway genes of a natural-rubber producing plant, Hevea brasiliensis. Biosci Biotech Bioch 72 (11): 2903–2917.
  52. 52. Mangas S, Moyano E, Osuna L, Rosa M, Bonfill M, et al. (2008) Triterpenoid saponin content and the expression level of related genes in calli of Centella asiatica.. Biotechnol Lett 30 (10): 1853–1857.
  53. 53. Han J, In J, Kwon Y, Choi Y (2010) Regulation of ginsenoside and phytosterol biosynthesis by RNA interferences of squalene epoxidase gene in Panax ginseng. Phytochemistry 71 (1): 36.
  54. 54. Coon MJ (2005) Cytochrome P450: nature's most versatile biological catalyst. Annu Rev Pharmacol Toxicol 45: 1–25.
  55. 55. Bowles D, Lim E-K, Poppenberger B, Vaistij FE (2006) Glycosyltransferases of lipophilic small molecules. Annu Rev Plant Biol 57: 567–597.
  56. 56. Shibuya M, Hoshino M, Katsube Y, Hayashi H, Kushiro T, et al. (2006) Identification of β-amyrin and sophoradiol 24-hydroxylase by expressed sequence tag mining and functional expression assay. FEBS J 273 (5): 948–959.
  57. 57. Qi X, Bakht S, Qin B, Leggett M, Hemmings A, et al. (2006) A different function for a member of an ancient and highly conserved cytochrome P450 family: from essential sterols to plant defense. P Natl Acad Sci 103 (49): 18848–18853.
  58. 58. Seki H, Ohyama K, Sawai S, Mizutani M, Ohnishi T, et al. (2008) Licorice β-amyrin 11-oxidase, a cytochrome P450 with a key role in the biosynthesis of the triterpene sweetener glycyrrhizin. P Natl Acad Sci 105 (37): 14204–14209.
  59. 59. Naoumkina MA, Modolo LV, Huhman DV, Urbanczyk-Wochniak E, Tang Y, et al. (2010) Genomic and coexpression analyses predict multiple genes involved in triterpene saponin biosynthesis in Medicago truncatula. Plant Cell 22 (3): 850–866.
  60. 60. Carelli M, Biazzi E, Panara F, Tava A, Scaramelli L, et al. (2011) Medicago truncatula CYP716A12 is a multifunctional oxidase involved in the biosynthesis of hemolytic saponins. Plant Cell 23 (8): 3070–3081.
  61. 61. Seki H, Sawai S, Ohyama K, Mizutani M, Ohnishi T, et al. (2011) Triterpene functional genomics in licorice for identification of CYP72A154 involved in the biosynthesis of glycyrrhizin. Plant Cell 23 (11): 4112–4123.
  62. 62. Shibuya M, Nishimura K, Yasuyama N, Ebizuka Y (2010) Identification and characterization of glycosyltransferases involved in the biosynthesis of soyasaponin I in Glycine max. FEBS Lett 584 (11): 2258.
  63. 63. Meesapyodsuk D, Balsevich J, Reed DW, Covello PS (2007) Saponin biosynthesis in Saponaria vaccaria. cDNAs encoding β-amyrin synthase and a triterpene carboxylic acid glucosyltransferase. Plant Physiol 143 (2): 959–969.
  64. 64. Achnine L, Huhman DV, Farag MA, Sumner LW, Blount JW, et al. (2005) Genomics-based selection and functional characterization of triterpene glycosyltransferases from the model legume Medicago truncatula. Plant J 41 (6): 875–887.
  65. 65. Poppenberger B, Fujioka S, Soeno K, George GL, Vaistij FE, et al. (2005) The UGT73C5 of Arabidopsis thaliana glucosylates brassinosteroids. P Natl Acad Sci USA 102 (42): 15253–15258.
  66. 66. Jiang Y, Tu P, Chen X, Zhang T (2005) Isolation of two sucrose esters from Polygala tenuifolia by high speed countercurrent chromatography. J Liq Chromatogr R T 28 (10): 1583–1592.
  67. 67. Miyase T, Noguchi H, Chen X-M (1999) Sucrose esters and xanthone C-glycosides from the roots of Polygala sibirica. J Nat Prod 62 (7): 993–996.
  68. 68. Li X, Zhang F, Wang D, Li Z, Qin X, et al. (2014) NMR-based metabonomic and quantitative real-time PCR in the profiling of metabolic changes in carbon tetrachloride-induced rat liver injury. J Pharmaceut Biomed Anal 89: 42–49.
  69. 69. Livak KJ, Schmittgen TD (2001) Analysis of Relative Gene Expression Data Using Real-Time Quantitative PCR and the 2 ΔΔC T Method. Methods 25: 402–408.