Integrative Analysis of Secondary Metabolism and Transcript Regulation in Arabidopsis Thaliana

Abstract

Metabolic states in plants are dynamically controlled via transcriptional regulation in response to environmental and developmental conditions. However, the regulatory mechanisms between gene expression (input) and the resultant metabolite accumulation (output) remain unclear because complex post-transcriptional events such as feedback regulation and inter-tissue translocation often play important roles in these mechanisms [1]. Hence detailed investigations of the dynamic behavior of metabolic systems and an understanding of their general rules are a major challenge for plant systems biology [2,3]. One promising strategy is a global survey of input (gene expression) and output (metabolite accumulation) signals to estimate the mechanisms working within these systems [4]. In this regard, recent advances in analytical and informatics technologies enable us to perform integrated analyses of transcriptome and metabolome data while considering metabolic pathway information [5].The pioneering applications of this strategy involved the investigation of the reprogramming of gene expression and metabolism triggered by nutritional stresses such as sulfur starvation [6][7][8][9]. This research demonstrated that two main types of information can be derived from integrated analysis [10]. The first outcome of such studies was the discovery of a gene-to-metabolite network regulating plant metabolism during environmental stresses. Integrated analysis of time-course data showed that groups of metabolites and genes related to primary and secondary metabolites are coordinately modulated by sulfur deficiency-induced stress [8,[11][12][13][14]. A similar analysis was performed for various plants such as cold-acclimating Arabidopsis, pathogen-infected Medicago, fruit ripening tomato, and metabolically engineered rice [15][16][17][18][19][20]. Another outcome was the successful prediction of novel gene function by using the rules underlying gene expressions and metabolite accumulation. Integrated analysis allowed the prediction of genes involved in glucosinolate biosynthesis (e.g., genes encoding sulfotransferases [21], 2 MYB transcription factors [9], and chain elongation enzymes [22,23]) in a comprehensive manner. The strategy was also used in nonmodel plants [24,25].

Systems Biology in the Context of Big Data and Networks

“…This study showed that gene pairs connected by interactions supported by multiple data are more likely to belong to the same GO category. For example in [ 81 ] gene expression profiles and mass spectrometry profiles are merged by using appropriate normalization of the data and a combined network of genes and metabolites has been constructed which helped find related genes and metabolites. A very large network of more than 60,000 interactions was reported [ 82 ] by integrating transcription factor binding, PPI, and protein phosphorylation data of yeast.…”