Transcriptional network analysis of the tryptophan-accumulating rice mutant during grain filling
- First Online:
- Cite this article as:
- Kim, D.S., Lee, K.J., Yim, W.C. et al. Mol Genet Genomics (2012) 287: 699. doi:10.1007/s00438-012-0712-x
- 439 Views
In a previous study, we selected a high tryptophan (Trp)-accumulating rice (Oryza sativa L.) mutant line by in vitro mutagenesis using gamma rays. To obtain detailed information about the Trp biosynthetic pathway during the grain-filling in rice, we investigated the gene expression profiles in the wild-type (cv. Dongan) and the high-level Trp-accumulating mutant line (MRVII-33) at five different grain-filling stages using microarray analysis. The mutant line showed approximately 6.3-fold higher Trp content and 2.3-fold higher amino acids compared with the original cultivar at the final stage (stage V). The intensity of gene expression was analyzed and compared between the wild-type and mutant line at each of the five grain-filling stages using the Rice 4 × 44K oligo DNA microarray. Among the five stages, stage III showed the highest gene expression changes for both up- and down-regulated genes. Among the Trp biosynthesis-related genes, trpG showed high expression in the mutant line during stages I to IV and trpE showed higher at stage III. Gene clustering was performed based on the genes of KEGG’s amino acid metabolism, and a total of 276 genes related to amino acid metabolism were placed into three clusters. The functional annotation enrichment analysis of the genes classified into the three clusters was also conducted using ClueGO. It was found that cluster 3 uniquely included biological processes related to aromatic amino acid metabolism. These results suggest that gene analysis based on microarray data is useful for elucidating the biological mechanisms of Trp accumulation in high Trp-accumulating mutants at each of the grain-filling stages.
KeywordsAmino acid metabolismOryza sativaMicroarrayTryptophan
Essential amino acids, such as lysine (Lys), methionine (Met), threonine (Thr), and tryptophan (Trp), contribute substantially to the nutritional quality of plant-based foods for humans and domestic animals (Wakasa et al. 2006). Trp, Lys, and Met have received the most attention because they are the most limiting in cereals, including rice (particularly Lys and Trp) and legume crops (particularly Met) (Ufaz and Galil 2008). Because the enrichment of essential amino acids in crop plants is of both economical and humanitarian interest, it is important to understand the biosynthetic mechanisms of amino acids for the improvement of the nutritional quality in seed crops.
Rice endosperm storage proteins are especially deficient in Trp, which is a precursor of secondary compounds such as indole alkaloids and IAA. Anthranilate synthase (AS) catalyzes the first reaction in the multi-step Trp biosynthesis branch by converting chorismate to anthranilate. The AS enzyme is feedback inhibited by the end product, Trp, which binds to an allosteric site on the AS catalytic α-subunit. AS has been indicated as the control point in the Trp biosynthetic pathway of plant cells by pathway intermediate-feeding studies (Widholm 1974), enzyme activity levels (Singh and Widholm 1974), feedback inhibition of the respective enzyme activities (Singh and Widholm 1974), and 5-methyltryptophan (5MT), a Trp analog, resistance selection (Widholm 1972). Additionally, transformation with a feedback-altered AS gene has produced maize (Andreson et al. 1997) and rice (Wakasa et al. 1999) lines with altered feedback-inhibited AS enzyme that resulted in a higher free Trp accumulation in the seeds.
Rice is considered to be a suitable model for studying monocotyledonous plants because its genome (430 Mb) is the smallest among the Poaceae species and a large amount of information has been generated through extensive studies (Kondou et al. 2006). Recently, genome investigations and assembling cDNA collections of many species have been actively promoted. In rice, sequencing of the entire genome was completed in 2002 (Feng et al. 2002; Goff et al. 2002; Sasaki et al. 2002; Yu et al. 2002), and more than 20,000 full-length cDNA clones were annotated. Thus, detailed analyses of genetic information has become possible using public databases (Kikuchi et al. 2003), and the information obtained in cDNAs is particularly useful for the investigation of the structure and function of genes.
To date, microarray technology had been used to study global gene expression during grain filling and seed formation in a number species, including rice (Zhu et al. 2003; Duan and Sun 2005), Arabidopsis (Girke et al. 2000; Ruuska et al. 2002), wheat (Gregersen et al. 2005; Baudo et al. 2006; Kan et al. 2006), barley (Sreenivasulu et al. 2004, 2006; Radchuk et al. 2005; Druka et al. 2006; Hansen et al. 2009), and Medicago truncatula (Firnhaber et al. 2005). Furthermore, the work of Sato et al. (2011) has provided a global framework of whole-plant gene expression analyses for rice and is the foundation of the RiceXPro database (http://ricexpro.dna.affrc.go.jp/).
In a previous study, we selected plant cells that were resistant to 5MT. These lines have concomitantly elevated levels of Trp through altered feedback inhibition of AS, a key feedback control enzyme that exhibits reduced sensitivity to Trp-induced inhibition (Chun et al. 2012). To explain the high accumulation of Trp in the seeds of the rice mutant, it is necessary to identify the altered expression of genes related to Trp pathway. The elucidation of differentially expressed genes (DEGs) at the grain-filling stages would provide clues for the investigation of Trp accumulation in endosperm storage proteins. In our study of the mechanisms of high-level Trp accumulation, we conducted a microarray analysis to identify DEGs according to the seed developmental stages using the 5MT-resistant mutant MRVII-33.
Materials and methods
Seeds of cv. Dongan (Oryza sativa L. spp. Japonica) and a high amino acid-accumulating M7 mutant line, MRVII-33, derived from cv. Dongan by in vitro mutagenesis using 70 Gy of gamma rays (Chun et al. 2012) were grown at the Radiation Breeding Research Farm of the Korea Atomic Energy Research Institute (KAERI) in 2010. The caryopses were marked on the day of flowering and collected at 2–3 days after pollination (DAP) (stage I), 4–5 DAP (stage II), 7–10 DAP (stage III), 12–15 DAP (stage IV), and 20 DAP (stage V) (Fig. 2a). The collected caryopses were immediately frozen in liquid N2, hulled and stored at −80 °C.
Growth under 5MT inhibition
Seeds of wild-type and the mutant line were surface sterilized in 70 % ethanol for 30 s, vigorously shaken in 1 % sodium hypochlorite for 20 min, and rinsed five times in sterilized water. The sterilized seeds were germinated on MS medium (2 % sucrose and 0.8 % agar, pH 5.8) containing 0.5 mM 5-methyltryptophan (5MT) for 3 weeks, and the effects of 5MT on growth were determined by measuring the shoot and root lengths.
Amino acid contents in seed storage proteins
The amino acid content within the rice seeds was measured using the Pico-Tag method (Waters) (Chun et al. 2012). The amino acids were extracted by grinding 30 mg seeds (fresh weight) with a mortar and pestle. The homogenate was dried under vacuum and re-dried completely after adding 10 μl of MWT (methanol:water:triethylamine, 2:2:1, V/V/V) under vacuum. Because cysteine is unstable under acid or base hydrolysis, a cysteine reagent mixture (formic acid: hydrogen peroxide, 19:1) was added to the dried samples. The samples were then incubated for 30 min at 22 °C and re-dried under vacuum. The general amino acids were hydrolyzed using constantly boiling HCl containing 0.5 % phenol in a reaction vial for 24 h at 110 °C. The hydrolysis of tryptophan was conducted in 4 M methanesulfonic acid for 24 h at 110 °C; 4 M KOH was added to the sample for 5 min for neutralization. The hydrolysates were dried under vacuum and stored at −20 °C until they were used for derivatization. The amino acids derived by the addition of 20 μl of MWTP (methanol:water:TAE:phenylisothiocyanate, 7:1:1:1, V/V/V/V) were applied to a 30-cm Pico-Taq Free Amino Acid Analysis column equilibrated with 140 mM sodium acetate containing 6 % acetonitrile. The column was connected to a Waters HPLC system, and the bound material was eluted with a linear gradient composed of 6 % acetonitrile (0, 14, 20, 46, and 100 %) at a flow rate of 1 ml/min at 46 °C. The absorbance of the eluate was measured at 254 nm
Total RNA was extracted from the frozen material using the RNeasy Plant Mini Kit (QIAGEN, Germany) according to the manufacturer’s protocol and instructions. The integrity of the RNA was assessed using an Agilent 2100 Bioanalyzer (Agilent Technology, CA, USA). A one-color spike mix was added to the total RNA prior to the labeling reaction, which was performed using the Quick Amp Labeling Kit, One-Color (Agilent Technologies) in the presence of cyanine-3 (Cy3)-CTP, according to the manufacturer’s protocol. The hybridization of the rice grain-filling microarray was carried out with two biological replicates. For microarray hybridization, 1 μg of Cy3-labeled cRNA was fragmented and hybridized on a slide of the rice 4 × 44 K microarray RAP-DB (Agilent; G2519F#15028) at 65 °C for 17 h. Hybridization and washing of the hybridized slides were performed according to the manufacturer’s instructions. The slides were scanned using the Agilent G2505B DNA microarray scanner, and the background correction of the Cy3 raw signals was performed using the Agilent Feature Extraction software (version 220.127.116.11). The processed raw signal intensity of all of the probes were subjected to 75th percentile normalization using GeneSpringGX11 (Agilent Technologies) and transformed to a log2 scale. A total of 21496 probes were extracted after the normalization and used for analysis of the gene expression profile.
Clustering of the amino acid metabolism genes
Term-based network analysis
Agilent Rice Oligo Microarray probe IDs matched to the rice genes were retrieved via the rice multi-platform microarray search of the NSF rice oligo nucleotide array project (http://www.ricearray.org/matrix.search.shtml). Expression profiles of DEGs with two- and upward-fold changes were manually gathered from the 155 GSM data of the rice genome array, as described elsewhere (Jang et al. 2008). To evaluate the tentative responsive networks, we employed the ClueGO tool, a Cytoscape plugin (Bindea et al. 2009). The rice functional terms for the associated co-expression clusters were converted to a custom ontology via BLASTp analysis (Moreno-Hagelsieb and Latimer 2007) against an Arabidopsis protein database (Version 7; ftp://ftp.arabidopsis.org/home/tair/Genes/TAIR7_genome_release/) and mapped using the Gene ontology (GO) database (Ashburner et al. 2000). The enrichment analysis was performed using a hypergeometric test with a Benjamini and Hochberg False Discovery Rate correction as the multiple hypothesis testing; all of the GO terms that were significant (P < 0.05) were selected as over-represented. The expression profiles are indicated in each correspondence node.
Growth differences under 5MT
Amino acid accumulation during seed developmental stages
The Trp content at each grain-filling stage showed a different pattern between the wild-type and mutant line (Fig. 2c). At stage I, the mutant line showed approximately 1.9 times higher Trp levels than wild type; the Trp content dramatically decreased thereafter until stage III in both the wild-type and mutant line. At stage IV, the Trp content in wild type continued to decrease, whereas the mutant line showed only a slight decrease. At the final stage, the mutant line exhibited an increase in the Trp content, with an approximately 6.3 times higher Trp content than wild type.
The number of DEGs at the seed development stages
Gene expression patterns in the tryptophan biosynthetic pathway
Functional annotation enrichment analysis
GO-identified categories for the genes
No. of genes
Term P value
Associated genes found
Biogenic amine metabolic process
Os03g0826500, Os07g0408700 Os02g0611200
Polyamine metabolic process
Polyamine biosynthetic process
Spermidine metabolic process
Spermidine biosynthetic process
Biogenic amine biosynthetic process
Os03g0826500, Os07g0408700 Os02g0611200
Biogenic amine metabolic process
Polyamine metabolic process
Polyamine biosynthetic process
Biogenic amine biosynthetic process
Aromatic amino acid family metabolic process
Aromatic amino acid family biosynthetic process
Dicarboxylic acid metabolic process
Os06g0709000, Os02g0749300 Os02g0440000
Chorismate metabolic process
Although rice contains very low levels of tryptophan (Trp) and lysine (Lys), it is one of the most important food crops. Thus, it is important to breed rice cultivars that contain high levels of Trp and Lys. In our previous study (Chun et al. 2012), mutant lines demonstrating 5-methyltrptophan (5MT) resistance were selected from irradiated calli derived from embryos. In the present study, we reconfirmed the inheritance of the 5MT resistance and the high content of Trp. Li and Last (1996) and Kreps et al. (1996) reported that a feedback-insensitive mutant of anthranilate synthase α (ASA), the first enzyme in the Trp biosynthetic pathway, accumulates three times more free Trp than wild type. Accordingly, we speculated that our mutant line had altered ASA sensitivity as a result of the gamma irradiation. In fact, through DNA sequencing, point mutations were identified in the ASA gene of our mutant line (Chun et al. 2012), suggesting that the increase in the Trp content was induced by these point mutations.
The mutant line, MRVII-33, contained a higher level of amino acids (2.3-fold) than the parental cultivar at the final stage. Wakasa et al. (2006) suggested that the absolute amount of amino acids in the seeds of their transgenic rice (containing an inserted OASA1D gene) increased as a result of the existence of regulatory mechanisms that increase amino acid synthesis in response to Trp accumulation. These authors also suggested that Trp accumulation in rice seeds might increase the transcription of genes that encode enzymes responsible for amino acid synthesis. Wang and Larkins (2001) reported that the opaque-2 mutation in maize (Oh545o2) is associated with an increased level of free amino acids in mature endosperm. The results of our study were in agreement with these previous studies. Our 5-MT-resistant mutant was selected for its higher tryptophan content and showed a higher level of accumulation of amino acids.
In the present study, the amino acid composition during grain filling was found to change rapidly (Fig. 2c), and the observed pattern was similar to that found in previous studies on other cereal grains (Jennings and Morton 1963; Brown et al. 1970; Dexter and Dronzek 1975). There was the difference in the amino acid content during stage I between the wild type and mutant, and although the mutant showed a higher amount of amino acids than wild type in the early stage, their amino acid contents decreased to similar levels from stages II to IV. However, during stage V, the mutant line showed increased amino acid contents, whereas the amino acid contents in wild type continuously decreased. Dexter and Dronzek (1975) reported that the free amino acid pool made a significant contribution to the overall amino acid composition of the endosperm in wheat, rye, and durum wheat during the early stages of endosperm development. These authors suggested that the changes in the overall amino acid composition of the endosperm during the early stages were partly a reflection of the changes in the free amino acid pool. These results were in agreement with previous studies in other cereal grains (Wiggins and Frey 1958; Hoseney et al. 1966; Hoseney and Finney 1967). In our study, the different of amino acid content between the wild-type and mutant line was also in accordance with the findings of previous studies.
Seven enzymes (trpA to trpG) are known to participate in tryptophan biosynthesis. Anthranilate synthase (AS, trpE and trpG) catalyzes the synthesis of anthranilate, which is a common precursor of many compounds, from chorismate and Gln (Tozawa et al. 2001). Purified plant AS is a heterodimer that consists of two subunits, α and β (Poulsen et al. 1993; Bohlmann et al. 1995; Romero and Roberts 1996). In rice, two genes, OASA1 and OASA2, encoding the AS α–subunits, were isolated and characterized (Tozawa et al. 2001). In our study, four of the seven enzymes participating in tryptophan biosynthesis showed changes in their gene expression patterns (Fig. 6); among these enzymes, OASA1 was found to be more highly expressed in the mutant line at stage III. Cho et al. (2000) reported that the overexpression of tobacco ASA2 in the forage legume Astragalus sinicus resulted in an increase of the concentration of free Trp. Tozawa et al. (2001) revealed that rice OASA1, Arabidopsis and tobacco ASA2, and Ruta graveolens ASA2 appeared to constitute a monophyletic group and reported that mutations in rice OASA1 appear to affect only Trp accumulation, with no changes in the contents of anthranilate or other amino acids, even though anthranilate is the immediate product of AS activity. We suggest that the different expression of OASA1 between the wild-type and mutant line at stages III and IV affected the final concentration of tryptophan.
Three clusters were created by gene expression patterns of amino acid metabolism in this study. Each cluster showed different patterns such as similar patterns (C1), higher expression in mutant line (C2), and lower expression in mutant line (C3). However, there was no grouping of specific amino acid metabolism in each cluster. Based on three clusters, the networks of the GO terms that were enriched in the amino acid metabolism genes illustrated the complexity of the transcriptional response in each cluster (Fig. 6; Table 1). In the network analysis with GO terms, C3 included aromatic amino acid process-related genes. Interestingly, the clusters 1 and 3 constructed with the similar biological processes like biogenic amine biosynthetic process, polyamine biosynthetic process, polyamine metabolic process, and biogenic amine metabolic process. Polyamines are low-molecular-weight polycations found in all living organisms (Cohen 1998). Polyamines have been frequently reported as endogenous plant growth regulator or intracellular messengers mediating many physiological processes such as cell division, morphogenesis, embryogenesis, fruit set and growth, and responses to environmental stress (Kaur-Sawhney et al. 2003). Yang et al. (2008) reported that higher levels of polyamines and activity of biosynthetic enzymes were significantly correlated with higher cell division, grain filling rates, cell number, and grain weight. In our results, the genes of C1 and C3 in the mutant line showed less increase or lower levels of gene expression than wild type (Fig. 6). Although the genes within C1 and C2 correlated with biogenic amine process, mutant line did not show significant difference in grain filling, seed weight and final yield than the wild type.
The metabolic manipulation of plants to improve their nutritional value is a primary goal of plant biotechnology. Essential amino acids, such as Lys, Met, Thr, and Trp, contribute substantially to the nutritional quality of plant-based foods for humans and domestic animals, but the amounts of such amino acids are limited in many crops (Wakasa et al. 2006). In this study, the 5-MT-resistant mutant line was examined for its increased tryptophan contents using microarray and gene-network analyses, and we found that the increase in the tryptophan content was due to the overexpression of the OASA2 gene. We also observed that the mutant line had increased levels of other amino acids in addition to tryptophan. The increased amino acid content in this rice mutant line (MRVII-33) would be beneficial from a nutritional perspective.
This work was supported by a grant from the Korea Atomic Energy Research Institute (KAERI) and the Ministry of Education, Science, and Technology (MEST), Republic of Korea.