IC4R010-Metabolomics-2015-25578272

Project Title

Rice (Oryza sativa) is an excellent resource; it comprises 25% of the total caloric intake of the world’s population, and rice plants yield many types of bioactive compounds. To determine the number of metabolites in rice and their chemical diversity, the metabolite composition of cultivated rice has been investigated with analytical techniques such as mass spectrometry (MS) and/or nuclear magnetic resonance spectroscopy and rice metabolite databases have been constructed. This review summarizes current knowledge on metabolites in rice including sugars, amino and organic acids, aromatic compounds, and phytohormones detected by gas chromatography–MS, liquid chromatography–MS, and capillary electrophoresis–MS.

According to Seaver et al. (2012), a comparison of individual metabolic databases revealed clear differences in the stored compounds (Seaver et al., 2012). When the researchers compared metabolite IDs used in the RiceCyc (Jaiswal et al., 2006), OryzaCyc, and AraCyc in Plant Metabolic Network (Figure 1), they found more than 1000 metabolites (e.g., primary metabolites) in these three databases. On the other hand, there were 54 metabolites in the two metabolic databases of rice. Of these, oryzalexins (Akatsuka et al., 1985; Kato et al., 1993; Kato et al., 1994) and momilactones (Kato-Noguchi and Peters, 2013) are known to be specific metabolites produced in rice.

Figure 1. Metabolites Found in RiceCyc and OryzaCyc, but Not in AraCyc.

Multi-MS-based metabolite profiling covered more than 87% of the chemical diversity of the metabolites when compared to the metabolites listed in RiceCyc (Figure 2). In 70 rice cultivars, 1652 peaks consisting of 156 distinct metabolites and 1496 unknown analytes comprised the metabolite composition of the rice kernels. Among the 156 distinct metabolites, the researchers provisionally identified peaks and identified metabolites by comparing their mass spectra and retention times (or indices) with those of the authentic standards. As rice kernels contain relatively higher levels of indole-3-acetic acid and trans-zeatin, they could detect these phytohormones in the metabolite profile data.

Figure 2. Visualization of the Predicted Chemical Diversity of Metabolites in Rice.

The primary form of the anthocyanins produced by pigmented rice is cyanidin 3-O-glucoside (I); the secondary form is peonidin 3-O-glucoside (II) (Figure 3). Tricin (III in Figure 3) is one of the flavones produced in the leaves and stems of Gramineae and other plants (Zhou and Ibrahim, 2010). The major compounds of g-oryzanol are cycloartenol trans-ferulate (V), 24-methylenecycloartanol trans-ferulate (VI), sitosterol trans-ferulate (VII), and campesterol trans-ferulate (VIII) (Figure 3). Sakuranetin (IV in Figure 3), one of the phytoalexins, is synthesized biochemically from naringenin in rice, cherry bark, and other plant species. The level of sakuranetin, barely detectable in healthy rice leaves, is rapidly increased under biotic and abiotic stress stimuli including UV treatment and pathogen attack.

Figure 3. Chemical Structure of Representative Bioactive Metabolites in Rice.

The identification of unknown compounds and their biological activities is the first step in the detection of a treasure trove of attractive metabolites in the rice metabolome. Spectroscopic techniques including UV–visible spectroscopy, Fourier transform–infrared spectroscopy, MS, and NMR are used to determine the structure of secondary metabolites. Chemometrics using an artificial biological gradient between two specimens is a powerful approach for the generation of semiquantitative calibration curves for all compounds detected including unknown peaks (Redestig et al., 2011a). There are three strategies to identify/estimate the chemical structure of unknown peaks. They are the application of (i) chromatographic and spectroscopic techniques for the isolation and the elucidation of their structure (Yang et al., 2014), (ii) chemical isotope labeling methods (Nakabayashi et al., 2013; Glaser et al., 2014; Zhou et al., 2014), and (iii) in silico analysis of MS and MS/MS data by computational MS (Wolf et al., 2010; Morreel et al., 2014). Recently, Morreel et al. (2014) developed the candidate substrate–product pair network approach, a novel algorithm that called for the high-throughput structural estimation of 145 secondary metabolites such as glucosinolates, flavonoids, benzenoids, lignans, indoles, and apocarotenoids in the metabolite profiles of Arabidopsis obtained by LC–MS.

The development of a strategy for QTL analysis facilitates the identification of useful QTLs that control metabolite accumulation and helps to elucidate systems for the regulation of metabolic networks in various plant.

RIKEN Center for Sustainable Resource Science, Yokohama, Kanagawa 230-0045, Japan
Graduate School of Life and Environmental Sciences, University of Tsukuba, Tsukuba, Ibaraki 305-8572, Japan
Precursory Research for Embryonic Science and Technology (PRESTO), Japan Science and Technology Agency (JST), Kawaguchi, Saitama 332-0012, Japan
Graduate School of Pharmaceutical Sciences, Chiba University, Chiba, Chiba 260-8675, Japan

Miyako Kusano (miyako.kusano@riken.jp) & Kazuki Saito (kazuki.saito@riken.jp)