A genome scale metabolic network for rice and accompanying analysis of tryptophan, auxin and serotonin biosynthesis regulation under biotic stress

Abstract

Background: Functional annotations of large plant genome projects mostly provide information on gene function and gene families based on the presence of protein domains and gene homology, but not necessarily in association with gene expression or metabolic and regulatory networks. These additional annotations are necessary to understand the physiology, development and adaptation of a plant and its interaction with the environment.

Results: RiceCyc is a metabolic pathway networks database for rice. It is a snapshot of the substrates, metabolites, enzymes, reactions and pathways of primary and intermediary metabolism in rice. RiceCyc version 3.3 features 316 pathways and 6,643 peptide-coding genes mapped to 2,103 enzyme-catalyzed and 87 protein-mediated transport reactions. The initial functional annotations of rice genes with InterPro, Gene Ontology, MetaCyc, and Enzyme Commission (EC) numbers were enriched with annotations provided by KEGG and Gramene databases. The pathway inferences and the network diagrams were first predicted based on MetaCyc reference networks and plant pathways from the Plant Metabolic Network, using the Pathologic module of Pathway Tools. This was enriched by manually adding metabolic pathways and gene functions specifically reported for rice. The RiceCyc database is hierarchically browsable from pathway diagrams to the associated genes, metabolites and chemical structures. Through the integrated tool OMICs Viewer, users can upload transcriptomic, proteomic and metabolomic data to visualize expression patterns in a virtual cell. RiceCyc, along with additional species-specific pathway databases hosted in the Gramene project, facilitates comparative pathway analysis.

Conclusions: Here we describe the RiceCyc network development and discuss its contribution to rice genome annotations. As a case study to demonstrate the use of RiceCyc network as a discovery environment we carried out an integrated bioinformatic analysis of rice metabolic genes that are differentially regulated under diurnal photoperiod and biotic stress treatments. The analysis of publicly available rice transcriptome datasets led to the hypothesis that the complete tryptophan biosynthesis and its dependent metabolic pathways including serotonin biosynthesis are induced by taxonomically diverse pathogens while also being under diurnal regulation. The RiceCyc database is available online for free access at http://www.gramene.org/pathway/.

Figure 1

A screen shot of Gramene’s Pathway module (http://www.gramene.org/pathway/). Various sub menu options include: search pathways, OMICs Viewer, Omics validator, Download, Help, Tutorial, Frequently asked questions (FAQs) and release notes. The pathway database listings included in column-1 the RiceCyc, SorghumCyc, MaizeCyc and BrachyCyc. These are developed by the Gramene database project. Column-2 includes mirror of pathway databases provided by collaborators TAIR, Noble Foundation and SGN. The mirrored databases include the Aracyc (Arabidopsis), MedicCyc (Medicago), PoplarCyc (Poplar), PotatoCyc (potato), LycoCyc (tomato) and CoffeaCyc (Coffee). Column-3 includes mirrors of reference pathway databases namely the EcoCyc (E. coli), MetaCyc and PlantCyc.

Figure 2

Gene expression analysis of the genes associated with the enzymes catalyzing the tryptophan biosynthesis. (A) A screen shot of the tryptophan biosynthesis pathway in RiceCyc with details of different objects; precursors on pathways, input and output molecules, enzymes referred by the EC number and proteins/genes associated with the enzymatic activity of a reaction. The pathway pages provide options to display only the experimentally determined protein-enzyme activities and zoom-in (molecular structures of the compounds) and zoom-out (pathway overview) views of pathways and the option to compare similar pathways from other species. The colored dots indicate the cycling genes at each reaction as demarcated in the heatmap in C. The yellow circle with a dotted outline indicates a “cycling’ gene that marginally missed the Q value cutoff (0.79) (B) A zoom-out view of the tryptophan biosynthetic pathway with genes overlaid with the expression levels of diurnally regulated transcriptome (green/down-regulated, red/upregulated). The 0, 12 and 24 hrs profiles were chosen to record differences between the expression profiles of the rice homologs mapped to the same enzymatic activity. A majority of genes in the pathway are highly expressed (red colored) at the 12 hr time point (towards the end of the day light cycle) compared to the down regulation in dark (green colored) and at the beginning of the day light cycle. (C) A heat map display of the expression pattern of all cycling genes in the pathway with the gene ID, subunit type and reaction of the pathway. Colored circles are used to identify individual reactions. (D) Biotic and abiotic stress induction of tryptophan biosynthetic genes based on data compiled in Genevestigator for rice. Expression data available for perturbations were filtered for significance level (p- value <0.05) and fold change (>2, based on gene in column with a yellow outline).

Figure 3

Expression pattern and metabolic processes of diurnally cycling genes in rice. (A) Expression pattern graphs of genes within the 4 major sub-clusters derived from hierarchical clustering of diurnally cycling genes (Additional file 3: Figure S1A) that mapped to RiceCyc metabolic pathways. (B) A metabolic cellular expression overlay of the 4 major diurnal gene expression clusters (the 4 sub-clusters in Additional file 3: Figure S1A and sub-graphs in Additional file 6: Figure S3A above) that show peak expression at 0 h, 8 h, 12 h and 20 h respectively. Boxed pathways: (1) phenylpropanoid (lignin) metabolism, (2) cholesterol biosynthesis, (3) glycolipid and phospholipid desaturation, (4) mevalonate pathway, (5) momilactone pathways, (6) gibberellins biosynthesis, (7) IAA biosynthesis.

Figure 4

Gene expression analysis of the genes associated with enzymes catalyzing serotonin biosynthesis. (A) A zoom-in view of the serotonin biosynthetic pathway. (B) Heat map of expression pattern of all cycling genes in the pathway with the gene ID and reaction of the pathway. Colored circles are used to identify individual reactions. (C) Biotic and abiotic stress induction of tryptophan biosynthetic genes based on data compiled in Genevestigator for rice. Expression data available for perturbations were filtered for significance level (p- value <0.05) and fold change (>2, based on gene in column with a yellow outline).