| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
First published online August 11, 2006; 10.1104/pp.106.080317 Plant Physiology 142:398-413 (2006) © 2006 American Society of Plant Biologists Clarification of Pathway-Specific Inhibition by Fourier Transform Ion Cyclotron Resonance/Mass Spectrometry-Based Metabolic Phenotyping Studies1,[W]Graduate School of Life and Environmental Sciences, Osaka Prefecture University, Sakai 599–8531, Japan (A.O., T.O., A.K., D.O.); Research Association for Biotechnology, Tokyo 105–0003, Japan (A.O., Y.S.); Department of Bioinformatics and Genomics, Nara Institute of Science and Technology, Ikoma 630–0192, Japan (Y.N., Y.S., S.K.); Ehime Women's College, Uwajima 798–0025, Japan (Y.N.); and Kazusa DNA Research Institute, Kisarazu 292–0818, Japan (Y.N., H.S., N.S., D.S.)
We have developed a metabolic profiling scheme based on direct-infusion Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR/MS). The scheme consists of: (1) reproducible data collection under optimized FT-ICR/MS analytical conditions; (2) automatic mass-error correction and multivariate analyses for metabolome characterization using a newly developed metabolomics tool (DMASS software); (3) identification of marker metabolite candidates by searching a species-metabolite relationship database, KNApSAcK; and (4) structural analyses by an MS/MS method. The scheme was applied to metabolic phenotyping of Arabidopsis (Arabidopsis thaliana) seedlings treated with different herbicidal chemical classes for pathway-specific inhibitions. Arabidopsis extracts were directly infused into an electrospray ionization source on an FT-ICR/MS system. Acquired metabolomics data were comprised of mass-to-charge ratio values with ion intensity information subjected to principal component analysis, and metabolic phenotypes from the herbicide treatments were clearly differentiated from those of the herbicide-free treatment. From each herbicide treatment, candidate metabolites representing such metabolic phenotypes were found through the KNApSAcK database search. The database search and MS/MS analyses suggested dose-dependent accumulation patterns of specific metabolites including several flavonoid glycosides. The metabolic phenotyping scheme on the basis of FT-ICR/MS coupled with the DMASS program is discussed as a general tool for high throughput metabolic phenotyping studies.
Metabolomics constitutes one of the essential challenges in functional genomics studies, providing biochemical information for understanding complex metabolic activities of certain cells, tissues, organs, and individual organisms. Particularly, through comprehensive metabolomics, endogenous small molecules of unknown properties could be uncovered as to represent specific biological conditions, and chemical structure information for such molecules is crucial for a rational understanding of the occurrence of such unknown metabolites in a biological context. Thus, metabolite identification and structural characterization are essential for successful metabolomics.
Different analytical instruments have been applied to metabolome analyses (Fiehn, 2002 Fourier transform ion cyclotron resonance MS (FT-ICR/MS) offers high-throughput analyses of small molecules at extremely high levels of resolution and sensitivity, typically being under four or five places of decimals in the mass measurements. Such accurate FT-ICR/MS analyses can be applied to the prediction of chemical formulas and molecular identities of metabolites. The direct infusion analyses using FT-ICR/MS technology is thus the best-suited option for high-throughput metabolic profiling studies coupled with a metabolite identification scheme. However, FT-ICR/MS has been mainly applied to proteomics research, and its extreme performance has not been fully developed as a general metabolomics tool. Here, we describe a metabolomics research scheme with a computational tool, DMASS, developed exclusively for metabolic profiling studies using FT-ICR/MS. The DMASS scheme converts thousands of mass signals (mass-to-charge ratio [m/z] values and ion intensity values) into metabolome data through several steps, including correction of m/z values using internal mass calibrants (IMCs), averaging ion intensities of individual MS signals among replicate samples, and multivariate analyses such as principal component analysis (PCA) and batch learning-self organizing map. We studied how our FT-ICR/MS with the DMASS data processing scheme is functional for metabolic profiling and phenotyping studies. We mimicked specific metabolic disorders by treating Arabidopsis (Arabidopsis thaliana) seedlings with herbicides of known mode of actions. The DMASS scheme and the PCA revealed distinct metabolome clusters among those herbicide treatments, and detailed comparison of the metabolomes led to identification of specific metabolite accumulation in each herbicide treatment. Using a metabolite-species database (KNApSAcK, http://kanaya.naist.jp/KNApSAcK/), we succeeded in predicting metabolite identification, and these structures were further investigated in the MS/MS analyses using a sustained off resonance irradiation-collision induced dissociation (SORI-CID) method of the FT-ICR/MS. These results demonstrated practical application possibilities of the FT-ICR/MS-based metabolic phenotyping as an essential function in integrated metabolomics studies.
FT-ICR/MS Data Processing System One of the major objectives in our metabolomics studies was metabolite identification with the aid of high-resolution mass spectral data and ion signal intensity information, leading to characterization of metabolic profiles representing biological conditions. For such metabolite identification, FT-ICR/MS is the best technology owing to its extreme levels of accurate mass determination and high-resolution performance. However, at least in our system, analytical data fluctuations were generally associated with the m/z values at the levels of three or four places of decimals in the high-resolution analyses with FT-ICR/MS. In addition, the ion signal intensities fluctuated in every spectral scanning. Without managing such analytical difficulties, FT-ICR/MS could not be applied to reproducible metabolomics studies. We developed a data analysis tool, DMASS, for large-scale metabolic profiling studies on the basis of FT-ICR/MS analyses. Figure 1 shows schematically the data processing steps.
For FT-ICR/MS measurements, we performed 10 successive spectral scans for each sample analysis. For the analysis, we added IMCs to experimental samples for correcting the analytical errors with m/z values. These IMCs included lidocaine, prochloraz, reserpine, and bombesin for the positive ion mode analysis, and a set of 2,4-dichlorophenoxy acetic acid, ampicillin, 3-[(3-cholamidopropyl) dimethylammonio]propanesulfonic acid, and tetra-N-acetylchitotetraose was used as the IMCs for the negative ion mode analysis. Briefly, the experimental m/z values of the IMCs were fixed to their theoretical values, and the m/z error calibration data were reflected for the m/z compensation for all other ion species in each spectral scan (Figure 1A, a). Then the corrected m/z values of repeatedly identifiable m/z values were matched to one another among 10 independent scans (Fig. 1A, b). Ions (m/z values) such as those shown by the asterisk in Figure 1A, a, in which appearance frequencies were below 50% among 10 independent scans, for example, were not included for further data processing steps. The threshold levels of ion appearance frequencies were freely adjustable in the DMASS data processing scheme. The intensity values of reproducibly observed ions were converted into percentage values of total ion intensity (Fig. 1A, b). Thus, metabolome data from a single biological sample consisted of averaged m/z values with intensity information from 10 spectral scans. These data processing steps were applied to metabolome comparison among different samples (Fig. 1A, c) using multivariate analyses reflecting biological conditions (Fig. 1A, d). Specific ions observed with individual samples, such as the ion indicated by the star in Figure 1A, c, were included for the multivariate analyses. Figure 1B shows the DMASS operation window for the automatic analytical steps (Fig. 1A, a–d). Figure 2A shows an actual example of the m/z value fluctuations from a plant extract analysis in the negative ion mode. The peak of m/z = 348.10101 in scan 1, corresponding to ampicillin added as an IMC, was seen with the m/z values of 348.10268 and 348.10116 in scan 2 and scan 3, respectively. In every spectral scanning, such fluctuations of experimental m/z values were also detected with other ion peaks at a level corresponding to molecular weight (MW) < 0.002. We corrected the analytical errors with the IMCs in 10 independent spectral scans for a single sample. For example, the experimental m/z values of ampicillin in spectral scan (scan 1 through scan 10) were fixed to its theoretical value of 348.10235 (Fig. 2A, after). The m/z values of the other IMCs were fixed as well, and analytical errors with all other m/z values were compensated in terms of the corrected m/z values with the IMCs. Figure 2A schematically shows that the m/z values of 340.07429 and 346.07663 in scan 1 (before) were automatically converted to 340.07562 and 346.07790 (after), respectively, by DMASS software. The analytical errors in every spectral scanning were similarly compensated (Fig. 2A). Through this m/z compensation step, fluctuations of m/z values were narrowed down to a deviation range smaller than MW = 0.001. Significant numbers of ion species were not reproducibly detected among 10 independent scans from a single sample. The signal intensities of such ions with low appearance frequencies were notably low compared with repeatedly observed ions and were excluded from further data processing steps. Reproducible ions should have included both common ions among different samples and unique ions present in each specific sample. We studied whether the DMASS scheme satisfactorily corrected the m/z fluctuations. Thus, we compared the m/z fluctuations in terms of the coefficient of variance (CV), which was obtained by dividing the SD value for each m/z fluctuation by its averaged m/z value. Figure 2B shows an example of a CV value comparison before and after the DMASS data processing. After the analytical error correction, the m/z fluctuation for each ion decreased in the entire range of MW of our interest (data not shown).
The signal intensities of the reproducibly detectable ions were compared among 10 independent scans of each sample using the DMASS scheme. After compensating for the m/z value fluctuations, we were able to match corresponding ions to one another among different spectral scans. Between independent spectral scans of different extracts from a single sample, the intensity values of 95% of ion species were within 1.46-fold deviation, 90% were within 1.34-fold deviation, 75% within 1.19-fold, 50% within 1.10-fold, and 25% of ion species were within 1.04-fold deviation (Fig. 3A ). Figure 3B shows the ion intensity fluctuations between two independent Arabidopsis samples from a standard growth condition. The intensity values of 95% of ion species were within 2.40-fold deviation, 90% were within 1.93-fold deviation, 75% within 1.50-fold, 50% within 1.24-fold, and 25% were within 1.10-fold deviation. These ion intensity fluctuations were comparable to those observed with a different FT-ICR/MS system (Aharoni et al., 2002  ). Taken together, the analyses of the same sample were less influenced by different extraction processes (Fig. 3A), while the metabolomes even from the same growth conditions might have fluctuated due to biological reasons (Fig. 3B). In the mass ranges of 55 to 400 and 400 to 1,000 of our FT-ICR/MS analyses, the averages of analytical errors (calculated from absolute m/z values) were 0.000282 and 0.000703, respectively, and the corresponding SDs were 0.000361 and 0.000923, respectively.
As described above, our metabolic profiling data from each sample consisted of corrected m/z values with averaged signal intensity values of reproducible ions from 10 independent spectral data. We prepared at least triplicate samples for each experimental treatment, and the corresponding ions were identified and matched one to another among triplicate samples. In addition to these data processing steps, we performed statistical analyses of t test and Pearson correlation coefficient for individual ions (Fig. 1B, d). Metabolome comparison among different samples was done using two multivariate analyses, PCA and batch learning-self organizing map (Fig. 1B, d), which are extensively used as bioinformatics tools for clarification of metabolomes according to metabolite accumulation profiles (Kanaya et al., 2001 ; Abe et al., 2003; Hirai et al., 2004, 2005). The software, Dr.DMASS, is downloadable with detailed instructions at the Web site (http://kanaya.aist-nara.ac.jp/DrDMASS/).
The entire scheme of the DMASS data processing was applied to the profiling of metabolic disorders caused by a variety of chemicals with different herbicidal modes of action. Arabidopsis seedlings were treated at concentrations of herbicides shown in Table I, and methanol extracts were directly applied to the metabolic profiling analyses. Figure 4 shows the growth inhibition curves of Arabidopsis seedlings treated with herbicides (Table I). As the inhibitors for acetolactate synthase (ALS), we used sulfonylureas (halosulfuron and pyrazosulfuron) and an imidazolinone (imazamox), and two different chemical classes of cyclohexanediones (CHDs; clethodim and alloxidim) and aryloxyphenoxypropionates (AOPPs; quizalofop and cyhalofop) were used as the inhibitors of acetyl-CoA carboxylase (ACCase). The herbicides used in this study, except for cyhalofop, strongly inhibited Arabidopsis growth in dose-dependent manners. When treated with cyhalofop even at higher concentrations, no significant growth inhibition was observed. In the growth inhibition test, ALS inhibitors were 100 to 1,000 times more potent than other herbicides. For the metabolic profiling studies, we set the IC50 values (the concentration required for 50% growth inhibition) as the highest concentrations for the herbicide treatment to be analyzed by our metabolic profiling scheme (Fig. 4).
We set the analytical conditions for reproducible detection of 200 to approximately 300 independent ions from each sample, while extreme conditions could be applied to observe thousands of ions from each scanning. Nonetheless, a total of 1,560 independent ions in the positive and the negative ion mode analyses were detected from the different herbicide treatments (Supplemental Data S1). These results indicated that different ion species were detected from each of these herbicide treatments, suggesting that unique sets of metabolites built up in the different herbicide treatments. In other words, our FT-ICR/MS analytical condition for simultaneous detection of about 200 to approximately 300 ions was sufficient for metabolic profiling in this experimental design, while it is postulated that Arabidopsis may produce more than 5,000 compounds (Roessner et al., 2001 ). Newly appeared ion species were as high as 534 in these herbicide treatments, and the intensity values of 518 ions out of the 1,560 ion species have increased or decreased by over 2-fold from the differential herbicide treatments. We identified metabolite candidates for 284 ions (Supplemental Data S1) by searching the species-metabolite relationship database KNApSAcK, containing 11,075 metabolites (at the end of January, 2006), mainly from plant taxa (Shinbo et al., 2006). In the direct infusion electrospray ionization (ESI) FT-ICR/MS analyses, thousands of ions could be detected from each spectral scan by increasing the ion accumulation time at the hexapole. However, those ions were not reproducibly detected. The ions accumulated in excess could not transfer into the FT-ICR cell, and the amounts of ions in the cell could not be maintained persistently during successive spectral scanning (data not shown). Unexpected collisions among excess ions could also occur in the analyzing cell generating unnatural ions. The same difficulties should also arise at the hexapole region accumulating excess amounts of ions. In addition, we found that extended accumulation time at the hexapole led to the generation of multiply charged ions from a single ion species (data not shown). In this study, the FT-ICR/MS analytical condition for detecting 200 to approximately 300 ion species was satisfactorily applied to metabolic profiling studies without suffering from unnecessary difficulties due to irreproducible ions among spectral scans.
Figure 5 shows the metabolome clusters demonstrated by PCA using DMASS software. PCA is a multivariate regression method to project a distribution of data points in a multidimensional space into a space of fewer dimensions. In Figure 5, two vectors of principal component 1 (PC1 = 30.7%) and principal component 2 (PC2 = 15.6%) were sufficient to distinguish the metabolic disorders caused by the specific enzyme inhibitions. The metabolome cluster from the lower concentrations of herbicide treatments was not separated from that of the control samples (herbicide-free metabolome; circled with black solid line). When treated with higher concentrations of herbicides, metabolomes were classified into four distinct clusters: 5-enolpyruvylshikimate-3-P synthase (EPSPS) inhibition (blue), ALS inhibition (red), 4-hydroxyphenylpyruvate dehydrogenase (4HPPD) inhibition (purple), and ACCase inhibition (green). The metabolome observed with the ALS inhibitor treatments formed a single cluster irrespective of the inhibitor classes of sulfonylureas and an imidazolinone. This was also the case for the ACCase inhibition by two different chemical classes of the CHDs and the AOPP. These results suggested that we detected the metabolic disorders caused by the inhibition of the specific enzymatic steps, but they were not due to unidentified side effects by the herbicide treatments. The metabolome from the cyhalofop treatment, which had no growth inhibition effects on Arabidopsis, was indistinguishable from that of the control experiments (brown circle in Fig. 5). Ions corresponding to herbicides were not detected under our experimental conditions. This was probably due to the fact that extremely low concentrations of herbicides were sufficient for the growth inhibition (Fig. 4).
Our metabolomics studies were directed to determine differential metabolic profiles using the direct infusion ESI/FT-ICR/MS, where ion suppression effects were unavoidable. Ion suppression is a phenomenon in which compounds present in a complex mixture at high concentrations compete against other compounds at the ionization step, and these compounds can affect the ion intensity values of individual compounds in sample solutions. In other words, differential ion suppression effects should occur for every metabolite in different samples, and the ion intensity values for such metabolites could not be easily compared among different samples. We designed our experiments to detect dose-dependent metabolic disorders caused by the herbicide treatments. Thus, we were able to ascribe differential metabolic profiles to such herbicide treatments (inhibition of specific enzymatic steps) monitoring dose-dependent metabolite fluctuation patterns (changes in the ion intensity values). This metabolic profiling scheme could be applied to time-course experiments as well, while simple comparison between single dose treatments from a fixed time point would not be a reliable choice under the ion suppression effects.
Glyphosate inhibits EPSPS, which converts shikimic acid 3-P (S3P) with phosphoenolpyruvate to 5-enolpyrvoylshikimic acid 3-P in the shikimic pathway (Steinrücken and Amrhein, 1980
ACCase Inhibition Metabolome Arabidopsis growth was strongly inhibited when treated with the ACCase inhibitors of quizalofop, alloxidim, and clethodim at the concentrations of 0.1 to approximately 1 ppm, while no growth inhibition was evident in the cyhalofop treatment (Fig. 4). The PCA (Fig. 5) demonstrated that the ACCase inhibition metabolome was clearly different from the herbicide-free metabolome (Fig. 5), and the metabolome of the cyhalofop-treated plants was in the cluster of the herbicide-free treatment.
The inhibition of ACCase activity with two different chemical classes, CHDs and AOPPs, is a well-established herbicidal mode of action. Most non-Poaceae species are insensitive to the ACCase inhibitors, and CHDs and AOPPs therefore constitute effective graminicide herbicides (Konishi and Sasaki, 1994
Both the AOPP (quizalofop) and the CHDs (alloxidim and clethodim) exhibited herbicidal effects on Arabidopsis, and this is a rather unexpected result from the well-known sensitivities of the ACCase forms. However, the metabolome analyses together with specific metabolite accumulations indicated that these herbicides actually impaired metabolic conditions, leading to the strong growth inhibition. It has been reported that specific residues in the carboxyltransferase domain were involved in the herbicide sensitivity (Moss et al., 2003
In the ALS inhibitor treatments, specific accumulation of an ion with m/z = 577.15715 was observed in the negative ion mode (Fig. 8A
). The KNApSAcK database search suggested a flavonol with two glycosides as a candidate for this ion. An MS/MS analysis was done with a parent peak (m/z = 577.15715), yielding two fragment peaks with m/z = 431.10115 and m/z = 285.04995, indicating the loss of one glycoside and two glycosides, respectively (Fig. 8B). Thus, dissociated ions could be a glycoside with the chemical formula of C6H12O5, and the original ion with m/z = 577.15715 was suggested to be a trihydroxy flavonol (C27H30O14). An ion with m/z = 531.18791 also accumulated in the ALS treatment (Fig. 9A
). The KNApSAcK database search found a candidate, icaritin-3-rhamnoside [C27H32O11; theoretical m/z = 531.18719 ([M–H]–)]. The MS/MS analysis suggested the loss of the ion with the m/z value corresponding to one Rha moiety (Fig. 9B). While icaritin-3-rhamnoside is an unusual flavonoid derivative isolated only from Epimedium sagittatum (Mizuno et al., 1987
An ion with m/z = 591.17276 detected in the negative ion mode accumulated in plants treated with all herbicides except for cyhalofop (Fig. 10A ), which had no growth inhibition activity for Arabidopsis (Figs. 4 and 5). A dirhamnosylflavonol (C28H32O14) was suggested as the candidate for this ion through the KNApSAcK database search. However, the result of MS/MS analysis did not match the expected fragmentation patterns from a dirhamnosylflavonol (Fig. 10B). The m/z values of the two fragment ions did not correspond to the loss of Rha. While the structure of this metabolite was not elucidated, this compound could be a general marker metabolite for plant responses against herbicide treatment.
The KNApSAcK species-metabolite database has been constructed with published chemical structures with organism species information (Shinbo et al., 2006 ), and therefore, it is thought that the prediction of metabolite identities with species information could be fairly reliable. However, in the MS/MS analyses coupled with the KNApSAcK database search, we were not able to obtain direct information on stereochemical aspects, such as sugar conformation or hexose identity. Different analytical methods are essential to confirm chemical structures especially with multiple database hits of several stereoisomers. Incorporating chromatographic separation systems into the FT-ICR/MS system should be helpful in both reducing the ion suppression effects and comparing metabolites between different MS technologies and NMR studies.
Herbicides are expected to act on specific sites, while other physiological processes could be affected as well. For example, it has been reported that glyphosate, a specific EPSPS inhibitor, influenced carbon assimilation (Servaites et al., 1987 Using the DMASS scheme (Supplemental Data S2), we were able to differentiate metabolic phenotypes representing specific pathway inhibitions caused by chemical compounds. The FT-ICR/MS-based metabolic phenotyping scheme consists of the DMASS for automatic mass-error correction and statistic analyses, the MS/MS structural analysis, and the KNApSAcK database incorporating more than 10,000 metabolite entries with literature references. This metabolomics tool specifically developed for FT-ICR/MS constitutes an important part of integrated metabolomics research platform.
Chemicals Glyphosate was purchased from Sigma-Aldrich, and quizalofop-ethyl was obtained from Dr. Ehrenstorfer. Pyrazolate was kindly supplied by Sankyo Agro. Other herbicides including pyrazosulfron-ethyl, halosulfuron-methyl, imazamox, pyrazoxyfen, alloxydim-sodium, chethodim, and cyhalofop-butyl were purchased from Wako Pure Chemical Industries. Lidocaine and bombesin were purchased from Sigma-Aldrich, and tetra-N-acetylchitotetraose was from Seikagaku Kogyo. Other chemicals were purchased from Wako Pure Chemical Industries.
For germination, Arabidopsis (Arabidopsis thaliana) ecotype Columbia seeds were surface sterilized and placed on Suc-free germination medium (Valvekens et al., 1988
Thirty 1-week-old plants were frozen in liquid N2 and ground to powder, which was used for methanol extraction. The extracts were filtered through disposable membrane filter units (DISMIC-13JP, ADVANTEC), evaporated under N2 atmosphere, and stored at –80°C until use. Upon FT-ICR/MS analysis, the extracts were dissolved in 50% (v/v) acetonitrile/water. IMCs for the positive ion mode were lidocaine ([M + H]+ = 235.18049), prochloraz ([M + H]+ = 376.03809), reserpine ([M + H]+ = 609.28066), and bombesin ([M + 2H]2+ = 810.41481). A set of 2,4-dichlorophenoxy acetic acid ([M–H]– = 218.96212), ampicillin ([M–H]– = 348.10235), 3-[(3-cholamidopropyl) dimethylammonio] propanesulfonic acid ([M–H]– = 613.38920), and tetra-N-acetylchitotetraose ([M–H]– = 829.32078) were used as the IMCs in the negative ion mode analysis.
Mass analysis was done using an IonSpec Explorer FT-ICR/MS (IonSpec) equipped with a 7-tesla actively shielded superconducting magnet. Ions were generated from an ESI source with a fused silica needle of 0.005-inch i.d. Samples were infused using a Harvard syringe pump model 22 at a flow rate of 0.5 to 1.0 µL min–1 through a 100-µL Hamilton syringe. All the experimental events were controlled using Omega8 software (IonSpec). Briefly, the potentials on the electrospray emitters were set to 3.0 kV and –3.0 kV for the positive and the negative electrosprays, respectively. The base pressure in the source region was approximately 5 x 10–5 torr (1 torr = 133.3 Pa). For the positive and negative electrosprays, sample solutions were prepared in 50% (v/v) acetonitrile/water with 0.1% (v/v) of formic acid and ammonium hydroxide, respectively. Ionized metabolites were accumulated for a period of 2,500 to 5,000 ms in a hexapole ion trap/guide and transferred through a radiofrequency-only quadrupole into the FT-ICR cell in the superconducting magnetic field, where they were again trapped. The direct current potentials in the positive and negative ion mode analyses were –2 V and 2 V during the ion accumulation and 2 V and –2 V for the ion transfer into the FT-ICR cell, respectively. These ions trapped in the hexapole were extracted for the transfer into the FT-ICR cell. In the positive and negative ion modes, the potentials on the extraction plate were 12 V and –12 V during the ion trapping and were reversed to –2 V and 2 V for the extraction. The base pressure in the analyzer region was set to the level of approximately 4 x 10–10 torr. ESI-MS spectra were acquired over the m/z range 55 to 1,000 from 1,024,000 independent data points. MS/MS analyses were done using the sustained off-resonance irradiation SORI-CID methods (Gauthier et al., 1991
For systematic and comprehensive understanding of species-specific metabolic diversities, we have designed a database system, KNApSAcK, for searching relationships between metabolites and species (Shinbo et al., 2006
The following materials are available in the online version of this article.
Received March 15, 2006; accepted August 4, 2006; published August 11, 2006.
1 This work was supported by the New Energy and Industrial Technology Development Organization, Japan, and in part by the Ministry of Education, Culture, Sports, Science, and Technology of Japan (grant no. 16580281 to D.O.).  The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (www.plantphysiol.org) is: Daisaku Ohta (ohtad{at}bioinfo.osakafu-u.ac.jp).
[W] The online version of this article contains Web-only data. www.plantphysiol.org/cgi/doi/10.1104/pp.106.080317 * Corresponding author; e-mail ohtad{at}bioinfo.osakafu-u.ac.jp; fax 81–72–254–9409.
Abe T, Kanaya S, Kinouchi M, Ichiba Y, Kozuki T, Ikemura T (2003) Informatics for unveiling hidden genome signatures. Genome Res 13: 693–702 Aharoni A, Ric de Vos CH, Verhoeven HA, Maliepaard CA, Kruppa G, Bino R, Goodenowe DB (2002) Nontargeted metabolome analysis by use of Fourier transform ion cyclotron mass spectrometry. OMICS 6: 217–234[CrossRef][Medline] Aranìbar N, Singh BK, Stockton GW, Ott KH (2001) Automated mode-of-action detection by metabolic profiling. Biochem Biophys Res Commun 286: 150–155[CrossRef][ISI][Medline] Baud S, Bellec Y, Miquel M, Bellini C, Caboche M, Lepiniec L, Faure JD, Rochat C (2004) gurke and pasticcino3 mutants affected in embryo development are impaired in acetyl-CoA carboxylase. EMBO Rep 5: 515–520[CrossRef][ISI][Medline] Baud S, Guyon V, Kronenberger J, Wuillème S, Miquel M, Caboche M, Lepiniec L, Rochat C (2003) Multifunctional acetyl-CoA carboxylase 1 is essential for very long chain fatty acid elongation and embryo development in Arabidopsis. Plant J 33: 75–86[CrossRef][ISI][Medline] Choi YH, Tapias EC, Kim HK, Lefeber AWM, Erkelens C, Verhoeven JTJ, Brzin J, Zel J, Verpoorte R (2004) Metabolic discrimination of Catharanthus roseus leaves infected by phytoplasma using 1H-NMR spectroscopy and multivariate data analysis. Plant Physiol 135: 2398–2410 Dehaye L, Alban C, Job C, Douce R, Job D (1994) Kinetics of the two forms of acetyl-CoA carboxylase from Pisum sativum. Correlation of the substrate specificity of the enzymes and sensitivity towards aryloxyphenoxypropionate herbicides. Eur J Biochem 225: 1113–1123[ISI][Medline] Délye C, Zhang XQ, Chalopin C, Michel S, Powles SB (2003) An isoleucine residue within the carboxyl-transferase domain of multidomain acetyl-coenzyme A carboxylase is a major determinant of sensitivity to aryloxyphenoxypropionate but not to cyclohexanedione inhibitors. Plant Physiol 132: 1716–1723 Délye C, Zhang XQ, Michel S, Matéjicek A, Powles SB (2005) Molecular bases for sensitivity to acetyl-coenzyme A carboxylase inhibitors in black-grass. Plant Physiol 137: 794–806 Fernie AR, Trethewey RN, Krotzky AJ, Willmitzer L (2004) Metabolite profiling: from diagnostics to systems biology. Nat Rev Mol Cell Biol 5: 763–769[CrossRef][ISI][Medline] Fiehn O (2002) Metabolomics: the link between genotypes and phenotypes. Plant Mol Biol 48: 155–171[CrossRef][ISI][Medline] Fiehn O, Kopka J, Dörmann P, Altmann T, Trethewey RN, Willmitzer L (2000) Metabolite profiling for plant functional genomics. Nat Biotechnol 18: 1157–1161[CrossRef][ISI][Medline] Gauthier JW, Trautman TR, Jacobson DB (1991) Sustained off-resonance irradiation for collision-activated dissociation involving Fourier transform mass spectrometry: collision-activated dissociation technique that emulates infrared multiphoton dissociation. Anal Chim Acta 246: 211–225 Herbert D, Price LJ, Alban C, Dehaye L, Job D, Cole DJ, Pallett KE, Harwood JL (1996) Kinetic studies on two isoforms of acetyl-CoA carboxylase from maize leaves. Biochem J 318: 997–1006[Medline] Hirai K, Uchida A, Ohno R (2002) Major synthetic routes for modern herbicide classes and agrochemical characteristics. In P Böger, K Wakabayashi, K Hirai, eds, Herbicide Classes in Development. Springer, Berlin, pp 179–289 Hirai MY, Kein M, Fujikawa Y, Yano M, Goodenowe D, Yamazaki Y, Kanaya S, Nakamura Y, Kitayama M, Suzuki H, et al (2005) Elucidation of gene-to gene and metabolite-to-gene networks in Arabidopsis by integration of metabolomics and transcriptomics. J Biol Chem 280: 25590–25595 Hirai MY, Yano M, Goodenowe DB, Kanaya S, Kimura T, Awazuhara M, Arita M, Fujiwara T, Saito K (2004) Integration of transcriptomics and metabolomics for understanding of global responses to nutritional stresses in Arabidopsis thaliana. Proc Natl Acad Sci USA 101: 10205–10210 Huhman DV, Sumner LW (2002) Metabolic profiling of saponins in Medicago sativa and Medicago truncatula using HPLC coupled to an electrospray ion-trap mass spectrometer. Phytochemistry 59: 347–360[CrossRef][ISI][Medline] Huynh QK, Bauer SC, Bild GS, Kishore GM, Borgmeyer JR (1988) Site-directed mutagenesis of Petunia hybrida 5-enolpyruvylshikimate-3-phosphate synthase: Lys-23 is essential for substrate binding. J Biol Chem 263: 11636–11639 Kanaya S, Kinouchi M, Abe T, Kudo Y, Yamada Y, Nishi T, Mori H, Ikemura T (2001) Analysis of codon usage diversity of bacterial genes with a self-organizing map (SOM): characterization of horizontally transferred genes with emphasis of the E. coli O157 genome. Gene 276: 89–99[CrossRef][ISI][Medline] Kaplan F, Kopka J, Haskell DW, Zhao W, Schiller KC, Gatzke N, Sung DY, Guy CL (2004) Exploring the temperature-stress metabolome of Arabidopsis. Plant Physiol 136: 4159–4168 Konishi T, Sasaki Y (1994) Compartmentalization of two forms of acetyl-CoA carboxylase in plants and the origin of their tolerance toward herbicides. Proc Natl Acad Sci USA 91: 3598–3601 Kuk YI, Jung HI, Kwon OD, Lee do J, Burgos NR, Guh JO (2003) Sulfonylurea herbicide-resistant Monochoria vaginalis in Korean rice culture. Pest Manag Sci 59: 949–961[CrossRef][ISI][Medline] Laskin J, Futrell JH (2005) Activation of large ions in FT-ICR mass spectrometry. Mass Spectrom Rev 24: 135–167[CrossRef][ISI][Medline] Lichtenthaler HK (1990) Mode of action of herbicides affecting acetyl-CoA carboxylase and fatty acid biosynthesis. Z Naturforsch 45c: 521–528 Mizuno M, Hanioka S, Suzuki N, Iinuma M, Tanaka T, Xin-shun L, Xhi-da M (1987) Flavonol glycosides from Epimedium sagittatum. Phytochemistry 26: 861–863 Moss SR, Cocker KM, Brown AC, Hall L, Field LM (2003) Characterisation of target-site resistance to ACCase-inhibiting herbicides in the weed Alopecurus myosuroides (black-grass). Pest Manag Sci 59: 190–201[CrossRef][ISI][Medline] Price LJ, Herbert D, Moss SR, Cole DJ, Harwood JL (2003) Graminicide insensitivity correlates with herbicide-binding co-operativity on acetyl-CoA carboxylase isoforms. Biochem J 375: 415–423[CrossRef][ISI][Medline] Roessner U, Luedemann A, Brust D, Fiehn O, Linke T, Willmitzer L, Fernie AR (2001) Metabolic profiling allows comprehensive phenotyping of genetically or environmentally modified plant systems. Plant Cell 13: 11–29 Ruiz-Santaella JP, Heredia A, Prado RD (2006) Basis of selectivity of cyhalofop-butyl in Oryza sativa L. Planta 223: 191–199[CrossRef][ISI][Medline] Servaites JC, Tucci MA, Geiger DR (1987) Glyphosate effects on carbon assimilation, ribulose bisphosphate carboxylase activity, and metabolite levels in sugar beet leaves. Plant Physiol 85: 370–374 Shinbo Y, Nakamura Y, Altaf-Ul-Amin M, Asahi H, Kurokawa K, Arita M, Saito K, Ohta D, Shibata D, Kanaya S (2006) KNApSAcK: a comprehensive species-metabolite relationship database. Biotechnol Agr Forest 57: 166–181 Soga T, Ohashi Y, Ueno Y, Naraoka H, Tomita M, Nishioka T (2003) Quantitative metabolome analysis using capillary electrophoresis mass spectrometry. J Proteome Res 2: 488–494[CrossRef][ISI][Medline] Steinrücken HC, Amrhein N (1980) The herbicide glyphosate is a potent inhibitor of 5-enolpyruvyl-shikimic acid-3-phosphate synthase. Biochem Biophys Res Commun 94: 1207–1212[CrossRef][ISI][Medline] Sumner LW, Mendes P, Dixon RA (2003) Plant metabolomics: large-scale phytochemistry in the functional genomics era. Phytochemistry 62: 817–836[CrossRef][ISI][Medline] Valvekens D, van Montagu M, van Lijsebettens M (1988) Agrobacterium tumefaciens-mediated transformation of Arabidopsis thaliana root explants by using kanamycin selection. Proc Natl Acad Sci USA 85: 5536–5540 Viant MR, Rosenblum ES, Tjeerdema RS (2003) NMR-based metabolomics: a powerful approach for characterizing the effects of environmental stressors on organism health. Environ Sci Technol 37: 4982–4989[Medline] Vila-Aiub MM, Neve P, Powles SB (2005) Resistance cost of a cytochrome P450 herbicide metabolism mechanism but not an ACCase target site mutation in a multiple resistant Lolium rigidum population. New Phytol 167: 787–796[CrossRef][ISI][Medline] von Roepenack-Lahaye E, Degenkolb T, Zerjeski M, Franz M, Roth U, Wessjohann L, Schmidt J, Scheel D, Clemens S (2004) Profiling of Arabidopsis secondary metabolites by capillary liquid chromatography coupled to electrospray ionization quadrupole time-of-flight mass spectrometry. Plant Physiol 134: 548–559 Widholm JM, Chinnala AR, Ryu JH, Song HS, Eggett T, Brotherton JE (2001) Glyphosate selection of gene amplification in suspension cultures of 3 plant species. Physiol Plant 112: 540–545[CrossRef][Medline] Wu L, Mashego MR, van Dam JC, Proell AM, Vinke JL, Ras C, van Winden WA, van Gulik WM, Heijnen JJ (2005) Quantitative analysis of the microbial metabolome by isotope dilution mass spectrometry using uniformly 13C-labeled cell extracts as internal standards. Anal Biochem 336: 164–171[CrossRef][ISI][Medline] Zhang H, Tweel B, Tong L (2004) Molecular basis for the inhibition of the carboxyltransferase domain of acetyl-coenzyme-A carboxylase by haloxyfop and diclofop. Proc Natl Acad Sci USA 101: 5910–5915 This article has been cited by other articles:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
RESULTS AND DISCUSSION
m/z = 0.00073], which is the substrate of EPSPS, for this ion. We investigated the chemical structure of this metabolite by the MS/MS analysis using the SORI-CID technique. SORI utilizes off-resonance excitation of the ion under investigation (Gauthier et al., 1991
; theoretical m/z = 96.96962), and m/z = 235.00191 was ascribed to dehydrated S3P (theoretical m/z = 235.00132), indicating strongly that the ion accumulated in the glyphosate-treated plants was in fact S3P. To our knowledge, this is the first case of the isolation of S3P from plant tissues. In the other herbicide treatments, accumulations of the enzyme substrates (pyruvate or 2-keto-lactate for ALS, 4-hydroxyphenylpyruvate for 4HPPD, and acetyl-CoA for ACCase) were not observed. This was probably due to the fact that these metabolites served as substrates for a wide range of biochemical reactions, and the lifetime of these metabolites in the plant cell should be quite short. 
