Metabolomics Analysis Reveals that Ethylene and ... - ACS Publications

(MEP) and acetate-mevalonate (MVA) pathways through AACT and HMGS and through DXS, respectively, to induce TIA biosynthesis in C. roseus. Overall, bot...
1 downloads 5 Views 3MB Size
Download PDF
Article pubs.acs.org/jnp

Cite This: J. Nat. Prod. 2018, 81, 335−342

Metabolomics Analysis Reveals that Ethylene and Methyl Jasmonate Regulate Different Branch Pathways to Promote the Accumulation of Terpenoid Indole Alkaloids in Catharanthus roseus Xiao-Ning Zhang,†,‡,⊥ Jia Liu,§,⊥ Yang Liu,§ Yu Wang,§ Ann Abozeid,§,∥ Zhi-Guo Yu,*,† and Zhong-Hua Tang*,§ †

School of Pharmacy, Shenyang Pharmaceutical University, Shenyang 110016, People’s Republic of China Department of Antibiotics, Heilongjiang Institute for Food and Drug Control, Harbin 150080, People’s Republic of China § Key Laboratory of Plant Ecology, Northeast Forestry University, Harbin 150040, People’s Republic of China ∥ Botany Department, Faculty of Science, Menoufia University, Shebin El-koom 32511, Egypt ‡

S Supporting Information *

ABSTRACT: The medicinal plant Catharanthus roseus accumulates large numbers of terpenoid indole alkaloids (TIAs), including the pharmaceutically important vinblastine, vincristine, ajmalicine, and serpentine. The phytohormone ethylene or methyl jasmonate (MeJA) can markedly enhance alkaloid accumulation. The interaction between ethylene or MeJA in the regulation of TIA biosynthesis in C. roseus is unknown. Here, a metabolomics platform is reported that is based on liquid chromatography (LC) coupled with time-offlight mass spectrometry to study candidate components for TIA biosynthesis, which is controlled by ethylene or MeJA in C. roseus. Multivariate analysis identified 16 potential metabolites mostly associated with TIA metabolic pathways and seven targeted metabolites, outlining the TIA biosynthesis metabolic networks controlled by ethylene or MeJA. Interestingly, ethylene and MeJA regulate the 2-C-methyl-D-erythritol 4-phosphate (MEP) and acetate-mevalonate (MVA) pathways through AACT and HMGS and through DXS, respectively, to induce TIA biosynthesis in C. roseus. Overall, both nontargeted and targeted metabolomics, as well as transcript analysis, were used to reveal that MeJA and ethylene control different metabolic networks to induce TIA biosynthesis.

T

series of catalytic enzymes including 1-deoxy-D-xylulose-5phosphate synthase (DXS), 1-deoxy-D-xylulose-5-phosphate reductoisomerase (DXR), and 1-hydroxy-2-methyl-2-butenyl 4-diphosphate synthase (HDS). The biosynthetic pathway of TIAs in C. roseus and the related genes encoding the key enzymes have recently been revealed.5,6 However, the underlying mechanisms that regulate the biosynthetic pathway to increase TIA accumulation remain unclear and require elucidation. Phytohormones activate plant natural defense responses, which are followed by increased secondary metabolite production.7,8 The regulatory modes of TIA biosynthesis induced by different phytohormones have been extensively observed in C. roseus suspension cells and hairy roots.9,10 Ethylene can markedly enhance the accumulation of ajmalicine as well as an increased expression of the geraniol-10hydroxylase (G10H) gene in periwinkle suspension cells,11 while only catharanthine accumulation is induced in hairy roots.12,13 In addition, the stress hormone methyl jasmonate

he medicinal plant Catharanthus roseus produces a large number of terpenoid indole alkaloids (TIAs) used as chemotherapeutics in the treatment of lymphoma, leukemia, and carcinomas of the breast and lungs.1 However, the high demand and low yield of these TIAs in plants have led to metabolic engineering efforts to develop more efficient production platforms. The TIA biosynthetic pathway in C. roseus has been mostly elucidated. This pathway is a complex process with more than 50 biosynthetic steps that involve the interaction between relevant genes, enzymes, regulators, and intra/intercellular transporters.2 Strictosidine is the common precursor for many TIAs and is synthesized from two intermediates, tryptamine and secologanin, through the indole pathway and a branch of the terpenoid biosynthetic pathway, respectively. Isopentenyl pyrophosphate (IPP) can be synthesized via the acetate−mevalonate (MVA) pathway and the 2-Cmethyl-D-erythritol 4-phosphate (MEP) pathway and is a major intermediate in the biosynthesis of terpenes and terpenoids. At the same time, the MVA pathway serves as a minor resource of precursors for iridoid biosynthesis in TIA production. Moreover, the MVA pathway mainly leads to the formation of sesqui- and triterpenes.3,4 In the MEP pathway, the IPP or dimethylallyl diphosphate (DMAPP) is formed, through a © 2018 American Chemical Society and American Society of Pharmacognosy

Received: September 13, 2017 Published: February 6, 2018 335

DOI: 10.1021/acs.jnatprod.7b00782 J. Nat. Prod. 2018, 81, 335−342

Journal of Natural Products

Article

Figure 1. Score plot analysis of LC-MS metabolomic data. (A) PCA, (B) OSC-PLS-DA. Red square, control group (n = 6); green diamond, ethylene treatment group (n = 6); black diamond, MeJA treatment group (n = 6).

quantitative analysis of targeted plant metabolites.19 Widely targeted metabolomics facilitated the simultaneous quantification of more than 90 flavonoids using LC-MS.20 Targeted phenolic compound profiles and gene expression in response to phytohormones in medicinal plants have been reported.21 In this study, a comprehensive metabolic profiling analysis of TIAs in response to different phytohormone elicitation in C. roseus leaves is presented. A metabolomic platform integrating nontargeted/targeted UPLC-qTOF-MS-based metabolic profiling and multivariate analysis such as the PLS-DA model were used. In addition, the regulatory mechanism of the TIA metabolite network in C. roseus was revealed and compared in response to ethylene and MeJA, respectively.

(MeJA) coordinately induces the expression of the TDC and STR genes and leads to the increase of TIA production in periwinkle cell suspension cultures and seedlings.14 However, the comprehensive profiling of TIA accumulation induced by ethylene and MeJA in C. roseus has received little attention. Metabolomics is a powerful platform for the identification and quantification of small molecules in plants, based mostly on GC-MS, NMR, and liquid chromatography mass spectrometry (LC-MS) data.15 MS coupled to LC is the most frequently used analytical technique in plant metabolomics research because it is useful for analysis of large hydrophobic metabolites such as alkaloids, terpenoids, and phenols.16 For example, Huhman and Sumner applied online LC−photodiode array−MS to study the metabolic profiling of triterpene saponins in Medicago sativa and M. truncatula.16 They tentatively identified two new malonated saponins in M. sativa and confirmed 27 saponins in M. truncatula.16 Similarly, Buchholz and colleagues applied LCMS, in addition to other approaches, in the metabolomic characterization of Escherichia coli in culture and subjected the bacteria to substrate pulse experiments, leading to dynamic modeling of metabolic pathways.17 Keurentjes and colleagues used UPLC-q time-of-flight mass spectrometry (TOF-MS) to analyze 14 accessions of Arabidopsis thaliana originating from various parts of the world.18 In addition, LC-MS is useful for

■

RESULTS AND DISCUSSION Multivariate Analysis Allows Distinction of Ethylene and MeJA in C. roseus Leaves. Principal component analysis (PCA) is an unsupervised multivariate statistical method that is preferred for representing differences in metabolomic studies. However, due to the minor differences in metabolites between the control and phytohormone treatments, PCA failed to separate the control, ethylene, and MeJA treatments (Figure 1A). The obvious separation among the samples was on the 336

DOI: 10.1021/acs.jnatprod.7b00782 J. Nat. Prod. 2018, 81, 335−342

Journal of Natural Products

Article

Figure 2. Simplified metabolite pathway showing changes in metabolites for TIA biosynthesis in C. roseus in response to ethylene and MeJA. Solid arrows in the network diagram specify a single step connecting two metabolites; dashed arrows indicate at least two steps. x-Axis: the process, followed by control, ethylene, and ethylene with MeJA. y-Axis: the relative content of metabolites. For metabolites, colored bars indicate metabolite changes during the phytohormones treatments. Metabolites that were significantly induced by ethylene are shown in red, and those significantly induced by MeJA are shown in blue. Each bar represents the mean of six biological replicates ± EM.

Supporting Information) for each pair. As a result, the controls with ethylene and the MeJA could be better separated. In addition, the parameters (R2Y = 0.372, Q2 = −0.234) of the combination model were better than those built with every single data matrix. The statistical model (OSC-PLS-DA) using the most comprehensive metabolite data can more effectively reflect metabolic differences. Therefore, the effective model will help elucidate the role of metabolites in response to ethylene and MeJA. Important compounds were discovered and identified using the model described above.

basis of different phytohormone treatments. However, similar separation was not observed between the control and ethylene groups. This means that the biological variation among samples is larger than among groups.22 To clearly present distinctions among the three groups, more precise distinguishing tools were adopted. Moreover, three PLS-DA models were built according to OSC-filtered data derived from LC-ESI-qTOF-MS data sets. The control, ethylene, and MeJA groups could then be separated by each OSC-PLS-DA model (Figure 1B). The three data matrices were combined into pairs that consisted of OSC-PLS-DA models of both sides built (Figure S1, 337

DOI: 10.1021/acs.jnatprod.7b00782 J. Nat. Prod. 2018, 81, 335−342

Journal of Natural Products

Article

Figure 3. Relative accumulation levels determined by metabolomics of selected primary and secondary metabolites following ethylene or MeJA applications to C. roseus leaves. (A) Secondary metabolite score plot of PCA; (B) primary metabolite score plot of PCA; (C) loading plot of PCA. The levels of primary compounds are shown as solid squares. The accumulations of TIAs are shown as hollow circles. Data presented here are the mean of six biological replicates.

secondary metabolites (strictosidine and α-3′,4′-anhydrovinblastine), respectively. All potential compounds except Lphenylalanine and UDP-L-rhamnose were up-regulated (Figure 2). The MeJA group was observed for 10 metabolites, e.g., Dglucosamine-6-phosphate, 2-oxoarginine, and (2E,6E)-farnesyl diphosphate, which are primary metabolites; other secondary metabolites include strictosidine, strictosidine-aglycone, deacetoxyvindoline, 16-methoxytabersonine, 16-hydroxytabersonine, tabersonine, and catharanthine. Most metabolite levels were markedly high in the MeJA treatment group compared with those of the control group (Figure 2). In addition, the target analysis identified seven TIA compounds, including ajmalicine, serpentine, cathenamine, 16-methoxy-2,3-dihydro-3-hydroxytabersionine, deacetylvindoline, vindoline, and vinblastine (Table S1, Supporting Information). Correlation Analysis of Potential Compounds and TIA Metabolites. PCA “Q” values and loading plots were determined based on the accumulation of primary metabolites and TIAs (Figure 3). The overall accumulation of TIA pathway metabolites in C. roseus significantly increased in response to ethylene treatment. However, in the MeJA group, the enhanced TIA biosynthesis peaked in the leaves (Figure 3A). Furthermore, the “Q” values of primary metabolites were consumed; increased ethylene led to a 2-fold reduction. MeJA

Identification of Different Metabolites in Response to Ethylene and MeJA Regulation. As shown in Figure S1A and C (Supporting Information), in the score plot of the combinational OSC-PLS-DA model, the first principal component played a major role in distinguishing control from ethylene and MeJA treatments. Hence, metabolites contribute to the first principal component and represent metabolic differences in response to ethylene and MeJA treatments. The variable importance in projection (VIP) value in the first principal component was calculated for every variable and reflects the direct relationship between variable and classification (control from ethylene and MeJA treatments). As a result, a strategy based on the OSC-PLS-DA model was used to separate each treatment group, and discriminant variables were selected by variable loading plots (Figure S1B and D, Supporting Information), VIP, and independent t-test. Potential metabolites were identified by their VIP value (VIP > 1) at p < 0.05. The variables from LC-ESI-qTOF-MS separation or those detected by ESI+ were also complementary (Table S1, Supporting Information). In the control and ethylene groups, a VIP test emphasized the importance of six variables, including four primary metabolites (L-phenylalanine, 2-oxoarginine, 1-hydroxy-2-methyl-2-(E)-butenyl 4-diphosphate, and uridine diphosphate [UDP]-L-rhamnose) and two 338

DOI: 10.1021/acs.jnatprod.7b00782 J. Nat. Prod. 2018, 81, 335−342

Journal of Natural Products

Article

Figure 4. MEP and MVA pathway genes. (A) Biosynthesis of TIAs via the MEP and MVA pathways. Dashed arrows indicate at least two steps. Additional abbreviations are as follows: AACT, acetoacetyl-CoA thiolase; HMGS, hydroxymethylglutaryl-CoA synthase; DXS, 1-deoxy-D-xylulose-5phosphate synthase. (B) Fold change of AACT and HMGS in the MVA pathway in response to ethylene or MeJA. (C) Fold change of DXS in the MEP pathway in response to ethylene or MeJA. Each treatment is compared with the control group. If greater than 1, expression was promoted; if lower than 1, expression was suppressed. Each bar represents the mean of three biological replicates ± EM. Significant difference is calculated using the ANOVA analysis. The letters a and b represent significant difference from ANOVA analysis results.

decreased primary compound “Q” values by approximately 20and 10-fold in the leaves of C. roseus seedlings when compared with the control and ethylene, respectively (Figure 3B). Therefore, it is likely that TIA accumulation relies on primary pathway metabolites, which became expended. The correlation between TIA pathway metabolites and upstream pathway metabolites in C. roseus is shown in the PCA loading plot (Figure 3C). The primary metabolite D-glucosamine-6phosphate was significantly correlated with the accumulation of the final products of TIA biosynthesis, namely, α-3′,4′anhydrovinblastine and vinblastine. In addition, (2E,6E)farnesyl diphosphate and 2-oxoarginine were correlated with the accumulation of strictosidine, strictosidine-aglycone, acetylnorajmaline, tabersonine, 16-hydroxytabersonine, deacetoxyvindoline, catharanthine, and 16-methoxy-2,3-dihydro-3hydroxytabersonine, which were accumulated only in the MeJA treatment group (Figure 2). Expression of DXS, AACT, and HMGS in Response to Ethylene and MeJA. The expression levels of the MEP and MVA pathway genes (DXS, AACT, and HMGS) under ethylene and MeJA treatment conditions are shown in Figure 4. The expression of AACT and HMGS induced by MeJA in the MVA pathway showed a similar trend. MeJA treatment resulted in the largest increase in AACT and HMGS expression; the maximum response occurred between the control and ethylene treatment samples (Figure 4B). Ethylene treatment significantly increased the expression of the MEP pathway gene DXS more than the MeJA and control treatments did (Figure 4C).

Application of the generalized untargeted/targeted metabolomics method is promising for metabolite profiling of upstream and downstream compounds in response to ethylene and MeJA treatment in C. roseus leaves. Using this method, seven and nine metabolites that respond to ethylene and MeJA treatments were identified, respectively. In addition, seven TIAs were detected using a targeted approach. In accordance with previous reports in hairy roots of C. roseus, tabersonine, catharanthine, and serpentine accumulated in response to MeJA.12 In addition, ajmalicine, tabersonine, catharanthine, and serpentine accumulation was significantly altered after treating C. roseus plants with ethylene.23 However, few studies have investigated the comprehensive metabolic profiling of TIAs in response to ethylene or MeJA. In the present study, the leaves of C. roseus treated with ethylene or MeJA were analyzed using biochemical and metabolomic approaches; most of the ethylene or MeJA-responsive metabolites were detected and identified (Table S1, Supporting Information). Various patterns of TIA metabolites were analyzed by the initial conditions of metabolite measurements of ethylene and MeJA; as such, all metabolites were divided into three major categories (Figure S2, Supporting Information). The first group contains vindoline and deacetylvindoline, which decreased in the control compared with the hormone treatment conditions, especially in the ethylene group. Deacetylvindoline, a synthetic precursor of vindoline, exhibited a similar trend ti that of vindoline (Figure S2, Supporting Information), indicating that phytohormone-induced vindoline is mainly up-regulated by its precursor. The second group includes ajmaline, 1,2-dihydrovo339

DOI: 10.1021/acs.jnatprod.7b00782 J. Nat. Prod. 2018, 81, 335−342

Journal of Natural Products

Article

milenine, serpentine, 16-methoxytabersonine, α-3′,4′-anhydrovinblastine, and vinblastine; this group significantly differed from that of the ethylene treatment, suggesting that the compounds of this group may be related to ethylene-induced total TIA biosynthesis. The third group, consisting of strictosidine, strictosidine-aglycone, tabersonine, 16-hydroxytabersonine, deacetoxyvindoline, catharanthine, 16-methoxy-2,3dihydro-3-hydroxytabersonine, and cathenamine, exhibited significantly higher levels in MeJA-treated samples than those in the control and ethylene-treated ones. These results suggest that the group members possibly play a role in the specific upregulation of the MeJA response, which is also consistent with the Q value of TIA accumulation under these conditions. According to the results, it is possible to transfer a metabolite-based metabolic pathway for leaves treated with ethylene and MeJA (Figure 2) into a graph, by comparing the metabolic flux TIA accumulation in C. roseus leaves controlled by ethylene and MeJA, respectively. Similarly, whether a category or content of secondary metabolites significantly increases or whether any upstream compound [e.g., 2oxoarginine or (2E,6E)-farnesyl diphosphate] is involved in the regulation can be closely monitored. Ethylene-induced downstream metabolites less than MeJA affected the regulation of 1-hydroxy-2-methyl-2-(E)-butenyl 4-diphosphate, which suggests that ethylene and MeJA can promote the synthesis of TIAs but that the regulatory pathways and targeted metabolites between them are specific. Further analysis of the response of these candidate components revealed three key features under different treatment conditions, described as follows: (i) Under normal control conditions with ethylene, six potential metabolites were identified, including 2-oxoarginine, UDP-L-rhamnose, phenylanine, 1-hydroxy-2-methyl-2-(E)-butenyl 4-diphosphate, α-3′,4′-anhydrovinblastine, and strictosidine. Phenylanine is an important precursor of phenylpropanoids, which indicates that ethylene inhibits the synthesis of phenolic compounds that contribute to metabolic flux together with TIA accumulation. 1-Hydroxy-2-methyl-2-(E)butenyl 4-diphosphate belongs to the MEP pathway, which is the limiting step of TIA biosynthesis.24 Nevertheless, in the present study, as the levels of upstream compounds increased, TIA metabolite levels significantly increased, which implies that ethylene regulates the synthesis of TIAs through the MEP pathway. In plant cells, UDP-L-rhamnose is one of the major components of the cell wall skeleton. As genes are regulated by ethylene, ethylene is likely to inhibit primary metabolite synthesis (UDP-L-rhamnose). This phenomenon is a mechanism for regulating cell wall permeability, resulting in the accumulation of TIA. In summary, the resulting upstream metabolites are crucial for TIA biosynthesis in response to ethylene. (ii) Under MeJA treatment, three primary metabolites and seven secondary metabolites were identified. D-Glucosamine-6phosphate belongs to the glycolysis pathway, an important pathway of primary metabolites, indicating that MeJA treatment stimulates glycolysis, increasing the production of secondary metabolites while consuming D-glucosamine-6-phosphate. (2E,6E)-Farnesyl diphosphate significantly increased and is involved in the MVA pathway.25 This suggests that the MVA pathway leads to the synthesis of secondary metabolites and acts in close coordination with the response to MeJA. Furthermore, compared with (i), after the application of MeJA, the synthesis of some TIAs can be directly regulated,

further promoting the accumulation of total TIAs (Figure 3). The same phenomenon occurred in the control and MeJA groups. These compounds tended to increase more than in the ethylene group, which indicates that MeJA plays a key role in TIA accumulation. (iii) Strictosidine significantly accumulates in response to these phytohormones and acts as a central intermediate in the biosynthesis of many alkaloids.30 In this study, strictosidine is also a potential compound in response to ethylene or MeJA treatment. Strictosidine is formed by the coupling of tryptamine and secologanin catalyzed by strictosidine synthase (STR).26 The level of strictosidine was significantly elevated in response to MeJA and ethylene. Interestingly, 2-oxoarginine exhibited this same trend in this case. The accumulation of 2-oxoagrinine, which is involved in arginine and proline metabolism, is significantly positively related to the level of strictosidine (p < 0.01) (Table S2, Supporting Information). Based on these results, arginine and proline both likely participate in the regulation of TIA biosynthesis. Although this preliminary analysis presents some possible links between these metabolites and TIA accumulation, more research is needed to verify further correlations due to the complex regulation of TIA accumulation in C. roseus. The “gene-to-metabolite” strategy provides a basis for the analysis of different roles that ethylene and MeJA play in the metabolic flux involved in TIA biosynthesis. IPP, which requires TIAs, is synthesized mainly through the alternative MEP pathway in plastids rather than through the MVA pathway in the cytosol and is an important intermediate in the biosynthesis of terpenes and terpenoids.27 However, in the present study of TIA biosynthesis in response to ethylene and MeJA in C. roseus leaves, major roles of ethylene and MeJA in the MEP pathway and MVA pathway, respectively, were discovered. We tentatively explored the hormone-specific control of the metabolic flux of TIA accumulation by analyzing key genes in the MEP and MVA pathways, such as AACT, HMGS, and DXS.28−30 The expression levels of AACT and HMGS along with DXS in leaves treated with ethylene and MeJA were compared. MeJA significantly affected the expression of genes of the MVA pathway (AACT and HMGS, two primary catalyzed genes), whereas ethylene significantly enhanced the expression of the MEP pathway gene DXS (Figure 4C). The gene DXS significantly enhanced terpenoid biosynthesis in response to MeJA in Medicago truncatula.31 This result is consistent with our work. However, HMGS was upregulated 2.6 times in response to ethylene compared with the control group in Ganoderma lucidum.32 Theses results may indicate that the transcript response to phytohormones displays species-specificity. In summary, ethylene and MeJA specifically regulated the expression of MVA and MEP pathway genes, respectively. In addition, different metabolites lead to specific metabolic fluxes of TIAs biosynthesis in C. roseus leaves, which further confirms that different metabolic fluxes of TIA biosynthesis in response to ethylene and MeJA are generated.

■

EXPERIMENTAL SECTION

Plant Material, Growth Conditions, Treatment, and Sampling. Catharanthus roseus plants used for the experiment were grown in a greenhouse (ZPW-400, China) under a 12/12 h light/dark photoperiod at a 28/25 °C day/night temperature as described previously.23 Seeds were sown in pots containing perlite as the cultivation substrate and maintained under 80% relative humidity until germination. After germination, seedlings were fertilized with 1/2strength Hoagland solution (pH 6.5). When two pairs of leaves were 340

DOI: 10.1021/acs.jnatprod.7b00782 J. Nat. Prod. 2018, 81, 335−342

Journal of Natural Products

Article

present, seedlings were randomly selected and subjected to hydroponic cultivation until the third pair of leaves emerged. For the ethylene and MeJA treatments, seedlings were cultivated in Hoagland’s nutrient solution containing ethylene or MeJA.23 In the first experiment, plants were treated with ethephon, which was used to release ethylene (500 mL of 45 μM ethephon was dissolved in Hoagland nutrient solution). In the second experiment, 500 mL of 175 μM MeJA was supplied in Hoagland nutrient solution.The control group received the same volume of Hoagland nutrient solution. Single plants (a set of three plants) were well watered and treated with Hoagland’s nutrient solution as control group (10 mL tube), ethylene (10 mL tube), and MeJA (10 mL tube) as treatment group. Plants were harvested for metabolomic analysis after 3 days of treatment. Each experiment consisted of six biological replicates. Metabolic quenching of the leaves was performed by rapidly freezing the leaves in liquid N2. After the leaves were ground using a grinding instrument (MM 400, Retsch), the samples were ground under cryogenic conditions at 30 Hz for 1 min. Afterward, 50 mg of powder was weighed and added to 70% MeOH containing 0.1 mg/L lidocaine as an internal standard. The extract was kept overnight at 4 °C. The extract was vortexed three times in order to extract the required component more completely.33,34 The extract was centrifuged at 10000g at 4 °C for 10 min. The supernatant was removed and filtered using a microporous membrane (SCAA-104, 0.22 μm pore size; Anpel). The resulting samples were stored at −80 °C. UPLC-ESI-qTOF-MS Analysis. UPLC-ESI-qTOF-MS analysis was performed using a UPLC system coupled with a quantitative time-offlight (qTOF) tandem mass spectrometer via an electrospray ionization (ESI) interface (Agilent 6520). Owing to the complexity of the metabolites in the plant, it is impossible to analyze all of the metabolites with a single method. The mass spectrometer was operated in positive ion mode to analyze as many metabolites as possible with a single injection. The ESI conditions included a capillary voltage of 3500 V, a fragmentation voltage of 135 V, a source temperature of 350 °C, and a curtain gas pressure of 40 psi. Detection was performed in positive ion mode in the m/z range of 50−1000. The scan time was 1 s, and the interscan delay was 0.1 s in centered mode. A solution of purine (C 5 H 4 N 4 ) at m/z 121.050873 and hexakis ( 1 H, 2 H, 3 Htetrafluoropentoxy)phosphazene (C18H18O6N3P3F24) (HP-921) at m/z 922.009798 was used as an internal standard for obtaining the accurate molecular mass. Separation was performed on a Shim-pack LC column (VP-ODS, C18, pore size 5.0 μm, 2 × 150 mm). The injection volume was 5 μL. Gradient elution was performed at a flow rate of 0.5 mL/min with the following solvent system: (A) 0.04% HoAc−H2O, (B) 0.04% HoAc− MeCN; 0−20 min, 5% B−95% B; 20−22.1 min, 95% B−5% B; and 22.1−28 min, 5% B. Multivariate Data Analysis. The raw data were analyzed using the MassHunter Qualitative software. This software detects peaks in an LC-MS data set, lists their detection, and matches the peaks with the retention time and m/z pair as well as their corresponding intensities. Internal standard peaks and those of any known false positives were removed from the data matrix. By using the peak area normalization method, each metabolite was expressed as normalized peak area and formed the data matrix. The data matrix was imported to SIMCA-P+ version 11.5 (Umetrics AB, Umeå, Sweden). Multivariate data analysis, such as unsupervised PCA, supervised PLS-DA, and OSC, were performed using SIMCA-P+. Ctr scaling was used in all the OSC and PCA models. Cross-validation was used to calculate the number of significant components. The permutation test was used to estimate the validity of the OSC mode, and 999 permutations were used for all models. To be a valid model, the R2 intercept should be less than 0.4 and the Q2 intercept less than 0.05; otherwise, the model is considered overfit.35 The loading plot, VIP parameter, and the t-test from SPSS version 17.0 (SPSS, USA) were combined to screen important metabolites (VIP > 1, p < 0.05). Screening differences of metabolites were adopted from the http://www.cathacyc.org/ and https://metlin. scripps.edu/ qualitative databases. SPSS version 17.0 was used to

calculate Pearson’s correlation coefficients. Results were subjected to an analysis of variance (ANOVA) to determine any significant differences between different treatments. If the ANOVA showed any significant differences, then Duncan’s honestly significant difference (HSD) post hoc test was used to determine any differences between individual treatments. The Student’s t test was used for mean value comparison. PCA was performed to compare the primary and secondary metabolites by the different treatments. In addition, the PCA results explained the impact of the two components on the variation, whereas the Q value slightly differed. Here, the PCA results were extracted via the variance contribution rates of component weights under various treatments. The value of the principal component “Q” is an indicator of the comprehensive analysis as well as a scientific evaluation of objective phenomena and has no practical significance. A negative value indicates it is below the average. Isolation of RNA and qRT-PCR Assay. Total RNA was extracted by TRIzol reagent (Invitrogen) and used for cDNA synthesis (ReverTra Ace QPCR RT Kit, TOYOBO, Japan) after DNase I (TaKaRa, Japan) digestion. The expression of target genes as well as internal control, ribosomal protein subunit 9 (Rsp9), was monitored by quantitative real-time PCR using appropriate gene-specific primers. SYBRPremix Ex Taq system (TaKaRa, Japan) was used for gene expression analysis, and parameters used for quantitative real-time PCR were 95 °C for 30 s, followed by 40 cycles of 94 °C for 30 s, 56 °C for 30 s, and 72 °C for 30 s. The primers used for real-time PCR are RPS9, 5′-TTA GTC TTG TTC GAG TTC ATT TTG TAT-3′ and 5′-GAG CAA ATT AAC TCA ATT GAT AAT TAA C-3′; AACT, 5′-TAG CTT TGG GGC ATC CTC TA-3′ and 5′-CCC CTA ACA GTG TGA CCA AGA-3′; HMGS, 5′-CTC AAT GAG TAT GAC GGC AGT T-3′ and 5′-AGA CGA CCA ATT TGC TTT GG-3′; DXS, 5′-CGA ATG GGG TTT TAA TGA GG-3′, and 5′-GAG TGG AGA AAT GGG AGG AA-3′. GenBank accession numbers are RPS9, AJ749993; AACT, JF739870.1; HMGS, JF739871.1; DXS, DQ848672.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jnatprod.7b00782. Metabolite reporting checklist and reconmendation for LC-MS. Change patterns of metabolites detected by LCMS between control and various treatment conditions, and correlation analysis in Catharanthus roseus leaves (PDF) (XLS) (XLS)

■

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. ORCID

Jia Liu: 0000-0001-9068-9072 Ann Abozeid: 0000-0003-1282-0068 Zhi-Guo Yu: 0000-0003-1493-8932 Author Contributions ⊥

X.-N. Zhang and J. Liu contributed equally to this work.

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank Dr. F. Yu for his help in the metabolic pathway analysis. This work was supported by the National Natural Science Foundation of China (31400337), the Forest Scientific 341

DOI: 10.1021/acs.jnatprod.7b00782 J. Nat. Prod. 2018, 81, 335−342

Journal of Natural Products

Article

(29) Van der Heijden, R.; Verpoorte, R. Plant Cell, Tissue Organ Cult. 1995, 43, 85−88. (30) Vincent, C.; Martine, T.; Martine, C.; Pascal, G.; Audrey, O.; Pierre, D.; Benoit, S. P.; Nathalie, G. G. H. Plant Mol. Biol. 2005, 57, 855−870. (31) Nieuwenhuizen, N. J.; Chen, X. Y.; Wang, M. Y.; Matich, A. J.; Perez, R. L.; Allan, A. C.; Green, S. A.; Atkinson, R. G. Plant Physiol. 2015, 167, 1243−1258. (32) Arimura, G.; Garms, S.; Maffei, M.; Bossi, S.; Schulze, B.; Leitner, M.; Mithofer, A.; Boland, W. Planta 2008, 227, 453−464. (33) Zhao, J.; Zhu, W.; Wu, Y. Q.; Hu, Q. Acta. Pharm. Sinica 1999, 34, 539−542. (34) Liu, J. W.; Zhu, J. H.; Tang, L.; Wen, W.; Lv, S. S.; Yu, R. M. World J. Microbiol. Biotechnol. 2014, 30, 175−180. (35) Chen, J.; Zhou, L.; Zhang, X. Y.; Lu, X.; Gao, R.; Xu, C. J.; Xu, G. W. Electrophoresis 2012, 33, 3361−3369.

Research in the Public Welfare (201504701-2), and the Fundamental Research Funds for the Central Universities (2572017DA05).

■

REFERENCES

(1) Van der Heijden, R.; Jacobs, D. I.; Snoeijer, W.; Hallard, D.; Verpoorte, R. Curr. Med. Chem. 2004, 11, 607−628. (2) Zhao, L.; Sander, G. W.; Shanks, J. V. Adv. Biochem. Eng./ Biotechnol. 2013, 134, 23−54. (3) Newman, J. J. CRIT. REV. Biochem. Mol. 2008, 34, 95−106. (4) Lange, B. M.; Croteau, R. Proc. Natl. Acad. Sci. U. S. A. 1999, 96, 13714−13719. (5) Miettinen, K.; Dong, L. M.; Navrot, N.; Schneider, T.; Burlat, V.; Pollier, J.; Woittiez, L.; Krol, S. V. D.; Lugan, R.; Iic, T.; Verpoorte, R.; Oksman-Caldentey, K. M.; Martinoia, E.; Bouwmeester, H.; Goossens, A.; Memelink, J.; Reichhart, D. W. Nat. Commun. 2014, 5, 3606−3611. (6) Salim, V.; Wiens, B.; Atsumi, S. M.; Yu, F.; Luca, V. D. Phytochemistry 2014, 101, 23−31. (7) Vázquez-Flota, F.; Moreno-Valenzuela, O.; Miranda-Ham, M. L.; Coello-Coello, J.; Loyola-Vargas, V. M. Ir. J. Med. Sci. 2002, 171, 85− 88. (8) Aerts, R. J.; Gisi, D.; Carolis, E. D.; Luca, V. D.; Baumann, T. W. Plant J. 1994, 5, 635−643. (9) Leeparsons, C. W. T.; Ertürk, S.; Tengtrakool, J. Biotechnol. Lett. 2004, 26, 1595−1599. (10) Xing, S.; Pan, Q.; Tian, Y.; Wang, Q.; Liu, P.; Zhao, J.; Wang, G.; Sun, X.; Tang, K. J. Med. Plants. Res. 2011, 5, 1692−1700. (11) Papon, N.; Bremer, J.; Vansiri, A.; Andreu, F.; Rideau, M.; Creche, J. Planta Med. 2005, 71, 572−574. (12) Vazquez-Flota, F.; Hernandez-Dominguez, E.; de Lourdes Miranda-Ham, M.; Monforte-Gonzalez, M. Biotechnol. Lett. 2009, 31, 591−595. (13) Lee-Parsons, C. W. T.; Seda, E. Plant Cell Rep. 2005, 24, 677− 682. (14) Zhou, M. L.; Shao, J. R.; Tang, Y. X. Biotechnol. Appl. Biochem. 2009, 52, 313−323. (15) Crimmins, B. S.; Xia, X.; Hopke, P. K.; Holsen, T. M. Anal. Bioanal. Chem. 2014, 406, 1471−1480. (16) Metz, T. O.; Zhang, Q.; Page, J. S.; Shen, Y.; Callister, S. J.; Jacobs, J. M.; Smith, R. D. Biomarkers Med. 2007, 1, 159−185. (17) Buchholz, A.; Hurlebaus, J.; Wandrey, C.; Takors, R. Biomol. Eng. 2002, 19, 5−15. (18) Keurentjes, J. J.; Fu, J.; de Vos, C. H.; Lommen, A.; Hall, R. D.; Bino, R. J.; van der Plas, L. H.; Jansen, R. C.; Vreugdenhil, D.; Koornneef, M. Nat. Genet. 2006, 38, 842−849. (19) Tian, H.; Lam, S. M.; Shui, G. Int. J. Mol. Sci. 2016, 17, 187110.3390/ijms17111871. (20) Dong, X.; Chen, W.; Wang, W.; Zhang, H.; Liu, X.; Luo, J. J. Integr. Plant Biol. 2014, 56, 876−886. (21) Liu, J.; Liu, Y.; Wang, Y.; Zhang, Z. H.; Zu, Y. G.; Efferth, T.; Tang, Z. H. Front. Physiol. 2016, 7, 217. (22) Ali, K.; Maltese, F.; Figueriredo, A.; Rex, M.; Fortes, M. A.; Zyprian, E.; Pais, M. S.; Verpoorte, R.; Choi, Y. H. Plant Sci. 2012, 191, 100−107. (23) Pan, Y. J.; Liu, J.; Guo, X. R.; Zu, Y. G.; Tang, Z. H. Protoplasma 2015, 252, 813−824. (24) Hunter, W. N. J. Biol. Chem. 2007, 282, 21573−21577. (25) Zhu, J.; Wang, M.; Wen, W.; Yu, R. Pharmacogn. Rev. 2015, 9, 24−28. (26) Rischer, H.; Oresic, M.; Seppanen-Laakso, T.; Katajamaa, M.; Lammertyn, F.; Ardiles-Diaz, W.; Van Montagu, M. C.; Inze, D.; Oksman-Caldentey, K. M.; Goossens, A. Proc. Natl. Acad. Sci. U. S. A. 2006, 103, 5614−5619. (27) Contin, A.; Heijden, R.; Van Der Lefeber, A. W.; Verpoorte, R. FEBS Lett. 1998, 434, 413−416. (28) Chahed, K.; Oudin, A.; Guivarc’H, N.; Hamdi, S.; Chénieux, J. C.; Rideau, M.; Clastre, M. Plant Physiol. Biochem. 2000, 38, 559−566. 342

DOI: 10.1021/acs.jnatprod.7b00782 J. Nat. Prod. 2018, 81, 335−342