Omics data input for metabolic modeling
Introduction
Metabolic models for an organism represent its metabolic capabilities and limitations, and provide a basis to predict cellular response and levels of targeted/desired metabolite under a specific environmental and/or genetic condition. Since the completion of first genome-scale metabolic model of Haemophilus influenze in 1995, a significant amount of work has been performed on developing analytical and computational tools/approaches for the construction of metabolic models [1, 2, 3]. While microbial metabolic models have become a powerful tool to connect phenotype and genotype, with applications that includes characterization of biological systems and to develop non-intuitive strategies to reengineer them for enhanced production of valuable bio-products [4], progress towards metabolic modeling for eukaryotes, especially for plants, has been slow due to several challenges. Recently, however, several plant metabolic models have been described and include those for Arabidopsis, barley, corn, oilseedrape, rice, Chlamydomonas, C4 plants and CAM plants [4, 5].
The biochemical potential of a biological system can be derived from its genome, and can be converted in the form of a mathematical model by the method of metabolic network reconstruction. Metabolic modeling and its reconstruction is an iterative process, which involves several sources/databases to construct a draft model with preliminary set of reactions and constraints [1, 6, 7]. The process starts with genome annotation, which can be accomplished by sequence homology based analysis with known proteins, or careful inspection of the genome annotation if it is available. Annotated genes with assigned EC number are then matched with pathway databases such as KEGG, ExPASy, BREAD, BioCyc or to already existing metabolic models for phylogenetically close organism to obtain candidate biochemical reactions, resulting in a draft metabolic model. This draft metabolic model is then reanalyzed to detect potential faults, for example, errors related to duplication of metabolite or enzyme names, or erroneous representation of metabolic reactions catalyzed by isoenzymes and enzyme complexes. Several metabolic modeling approaches, such as kinetic modeling, biochemical systems approach, Metabolism with gene Expression (i.e. ME model) based modeling approach among others could be used to test and reconstruct metabolic models to increase its scope and accuracy. Kinetic models are particularly useful to simulate untested scenarios, and thus explore pathway behavior beyond the realm of what is experimentally available or currently feasible, and can also suggest new experiments in a directed approach [8]. Metabolic model at this stage includes every reaction that can be catalyzed by a specific organism and gives a broad over-view of its metabolic potential. However, a multicellular organism will activate different subsets of their genes in different organs, tissues, developmental stages and environmental/physiological conditions. For accurate predictions, reconstruction of metabolic model is required that must represent a subset of complete metabolic potential for an organism specific to a particular cell type and condition. Constrain-based models (CBM) reconstruction of metabolic model serves this purpose by incorporating different experimental data-sets as constrain to define directionality and activation of a specific metabolic network, thus improves metabolic model prediction accuracy [9]. Flux balance analysis (FBA) has been established as a leading approach for studying CBMs of cellular metabolism, enabling the explicit and quantitative description of metabolic network and imposes mass balance, the first basic constraint. Still, a model with only accounts of mass balance is largely undefined and further constraints are required to be imposed to enrich the range of feasible solutions with biologically realistic ones.
In this context, omics data have been used both to constrain calculated flux distributions, and as a comparison and validation of model prediction, thus enabling context-specific studies of the metabolism of an organism. Recent advancements in high-throughput large-scale analytical methods, and reduced cost for data generation and analysis have allowed exponential increase in the amount of omics data being generated for an organism in a very short time. Omics data integration as constraint parameter to refine and improve metabolic models have been shown as an effective approach, and successfully used in several microbial metabolic models [9, 10, 11, 12]. The recent developed metabolic models, different metabolic modeling approaches and automation tools, approaches for its reconstruction, improvements and applications have been reviewed [1, 2, 3, 13, 14], and will not be covered in details in this review. Here we review recent applications of different omics datasets and their contribution towards construction and reconstruction of plant metabolic models.
Section snippets
Genomics
Exponential decrease in cost and speed of genome sequencing, and advancements in bioinformatics tool for contiqs assembly has led to explosion on available whole genome sequences of different plant species [15, 16]. Since the publication of Arabidopsis genome in 2000, number of sequenced and assembled genome for plant has reached to more than 85 till date, peaking to 13 publications in 2012 alone and many more to come in next few years [17, 18]. Excellent recent reviews focus on advances in
Transcriptomics
Although genomics data provides a broad overview of the metabolic potential of a species, it does not convey information regarding specificity of a metabolic process or its activation (or lack thereof) for a particular developmental stage, tissue type, or different physiological or environmental conditions. Transcriptome data on the other hand takes into account the dynamic state of mRNA levels and serves as a constraint for the reconstruction of metabolic models to improve prediction accuracy.
Proteomics
Protein synthesis and modification is a highly regulated multistep process, and therefore, measurement of mRNA levels cannot predict the exact protein concentration or activities. Genetic perturbation associated phenotypes often results from changes in accumulation, stability, or posttranslational modifications of proteins, which can also disrupt protein–protein interactions and network connectivity [30]. Knowing the dynamics of proteins under different developmental or physiological stages,
Metabolomics
Metabolites are ultimate response of cellular regulatory processes, and are regarded as bridge between the genome and the phenotype of a biological entity. While transcriptome and proteome data in combination do provide hints of active metabolic processes and important insights into the hierarchical regulation of metabolic flux, metabolomics shed light on the actual enzyme activity through metabolic regulation and mass action effects. Several recent reviews have discussed in detail the
Fluxomics
Quantification of intracellular fluxes and analysis (fluxomics), which can be estimated through interpretation of stable isotope patterns in metabolites, provides a comprehensive characterization of metabolic networks. Several fluxomics studies using different isotopic tracers (2H, 13C, 14C, 15N and others) have been used to reveal metabolically active pathways and their metabolite intermediates as well as enzyme mechanisms in all types of species, and have played an important role in
Lipidomics
Lipids are characterized as an enormous number of chemically distinct molecular species that arises from the various combinations of fatty acids with backbone structures, and are the major constituents of total biomass of an organism. Comprehensive profiling of lipid content at qualitative and quantitative levels (lipidomics) provides an accurate over-view of role of specific lipid classes under various physiological conditions. In plants, lipids play important role as signaling and energy
Conclusion
Metabolic modeling and its reconstruction are well established for microbes, with detailed protocols and automated pipelines available, and have been used successfully in many areas of biotechnology and biomedicine [2]. However, progress made by metabolic modeling for plants is quite slow, partly due to complexed genomes, compartmentalization of metabolic reactions and multilevel regulation of the process. Despite of these challenges, several plant metabolic models has been published recently,
Conflict of interest
Nothing declared.
References and recommended reading
Papers of particular interest, published within the period of review, have been highlighted as:
• of special interest
•• of outstanding interest
Acknowledgements
We apologize to all colleagues whose relevant work we could not cite due to space limitation.
References (68)
- et al.
Recent advances in reconstruction and applications of genome-scale metabolic models
Curr Opin Biotechnol
(2012) - et al.
Recent advances in the reconstruction of metabolic models and integration of omics data
Curr Opin Biotechnol
(2014) - et al.
Metabolic network reconstruction: advances in in silico interpretation of analytical information
Curr Opin Biotechnol
(2012) - et al.
Are we ready for genome-scale modeling in plants?
Plant Sci
(2012) - et al.
Whole-genome metabolic network reconstruction and constraint-based modeling
Methods Enzymol
(2011) - et al.
A practical guide to genome-scale metabolic models and their analysis
Methods Enzymol
(2011) - et al.
Use of genome-scale metabolic models for understanding microbial physiology
FEBS Lett
(2010) - et al.
Mathematical optimization applications in metabolic networks
Metab Eng
(2012) - et al.
Plant genome sequencing — applications for crop improvement
Curr Opin Biotechnol
(2014) - et al.
Integration of genome-scale modeling and transcript profiling reveals metabolic pathways underlying light and temperature acclimation in Arabidopsis
Plant Cell
(2013)