(568f) Towards a Metabolic and Expression Model of the Metabolically Versatile Bacterium Rhodopseudomonas Palustris

Rhodopseudomonas palustris is a metabolically versatile Purple Non-Sulfur Bacterium (PNSB). Depending on growth conditions, R. palustris can operate in either one of four different forms of metabolism: photoautotrophic, photoheterotrophic, chemoautotrophic, and chemoheterotrophic [1]. R. palustris is also a facultative anaerobe, meaning that it can operate both aerobically and anaerobically. Furthermore, the organism is capable of fixing nitrogen. This metabolic flexibility and plethora of functionalities have made R. palustris a model organism for studying the network of interacting reactions and how it’s altered in response to changing conditions. Moreover, exploration into R. palustris’ industrial applications has just begun with research being generated on the production of hydrogen [2], methane [3], and polyhydroxybutyrate (PHB) [4, 5] to name a few. This study reconstructs R. palustris’ metabolic and expression networks to gain insight into the organism’s functional states under different conditions and to lay the groundwork for strain design studies to maximize yields of the different products produced by the bacterium.

Genome scale metabolic models provide a systems-level approach to studying an organism’s metabolism based on the conservation of mass under steady state conditions. However, a major limitation of metabolic models (M-Models) is the inability to directly incorporate macromolecular synthesis, and therefore omics data, into the model. To overcome these shortcomings, metabolic and expression (ME-Models) [6] have recently been developed to account for the synthesis of the macromolecular machinery of living systems. Similar to M-Models, ME-Models make use of optimization to predict optimal functional states based on the given objective function that the organism is assumed to optimize in vivo. Where ME-Models surpass M-Models is in the stringency of the constraints being implemented, as they impose the added costs of synthesizing and replicating the cell’s macromolecular machinery. Furthermore, the addition of transcription and translation ‘reactions’ makes it possible to quantitatively simulate expression rates and therefore allows for the incorporation of omics data into the formulation. However, this added complexity is usually accompanied by an increase in the amount and type of experimental data required, which is problematic when modeling non-model organisms.

In this study, a genome-scale metabolic model was first generated from available databases, including KBase, Uniprot, and KEGG. The reactions were then checked and corrected for imbalances in elemental mass and charge. Next, missing functionalities in the model were reconciled using optimization tools, such as GapFill [7]. The metabolic model currently contains 925 reactions, 1220 metabolites, and 1134 genes. An expression matrix (E-matix), containing transcription, translation, and post-transcriptional modifications [8], is currently being reconstructed. After the reconstruction of the expression matrix, the two networks will be integrated [9] and previously established ‘[10] will be implemented. These constraints link the synthesis of macromolecular molecules to catalysis of metabolic components and to biomass production. ME-Model results obtained will be compared with those from the metabolic model to quantify the change in solution space caused by the additional constraints. This ME model will enable more accurate genetic intervention predictions which maximize production of industrially relevant products from R. palustris. This predictive power will continue to grow as more omics datasets become available, facilitating the establishment of this metabolically versatile, non-model microorganism as a biotechnology platform.

References:

1. Larimer, F.W., et al., Complete genome sequence of the metabolically versatile photosynthetic bacterium Rhodopseudomonas palustris. Nat Biotech, 2004. 22(1): p. 55-61.
2. Rey, F.E., E.K. Heiniger, and C.S. Harwood, Redirection of metabolism for biological hydrogen production. Appl Environ Microbiol, 2007. 73(5): p. 1665-71.
3. Fixen, K.R., et al., Light-driven carbon dioxide reduction to methane by nitrogenase in a photosynthetic bacterium. Proc Natl Acad Sci U S A, 2016. 113(36): p. 10163-7.
4. Merugu, R., S. Girisham, and S.M. Reddy, Production of PHB (polyhydroxybutyrate) by rhodopseudomonas palustris KU003 and rhodobacter capsulatus KU002 under phosphate limitation. International Journal of Applied Biology and Pharmaceutical Technology, 2010. 1(3): p. 847-850.
5. Mukhopadhyay, M., A. Patra, and A.K. Paul, Production of poly(3-hydroxybutyrate) and poly(3-hydroxybutyrate-co-3-hydroxyvalerate) by Rhodopseudomonas palustris SP5212. World Journal of Microbiology and Biotechnology, 2005. 21(5): p. 765-769.
6. Lerman JA, Hyduke DR, Latif H, Portnoy VA, Lewis NE, Orth JD, et al. In silico method for modelling metabolism and gene product expression at genome scale. Nat Commun. 2012;3: 929.
7. Satish Kumar, V., M.S. Dasika, and C.D. Maranas, Optimization based automated curation of metabolic reconstructions. BMC Bioinformatics, 2007. 8: p. 212.
8. Thiele I, Jamshidi N, Fleming RMT, Palsson BØ. Genome-scale reconstruction of Escherichia coli’s transcriptional and translational machinery: a knowledge base, its mathematical formulation, and its functional characterization. PLoS Comput Biol. 2009;5:e1000312.9
9. O’Brien EJ, Lerman JA, Chang RL, Hyduke DR, Palsson BØ. Genome-scale models of metabolism and gene expression extend and refine growth phenotype prediction. Mol Syst Biol. 2013;9: 693.