Mapping metabolic fluxes in plant cells to understand carbon-nitrogen interactions and nitrogen storage and cycling

Abstract

Plants provide commodities like food, fiber, fuel and chemicals. Understanding plant metabolism will help find genetic engineering targets that enhance production of these commodities. Interactions between the macronutrients - carbon (C) and nitrogen (N) determine growth and developmental functions in plants (Nunes-Nesi, Fernie, and Stitt 2010; Sakakibara, Takei, and Hirose 2006) and are regulated by complex mechanisms that need systems-level analyses. Metabolic fluxes, the rates of C flow in metabolic pathways, provide a system-wide view of metabolism and are quantified by steady state metabolic flux analysis (MFA) wherein isotopic tracers (13C, 15N) are fed to the cells and the resulting labeling patterns of biomass components are used to fit the fluxes. In this study we i) statistically designed isotope labeling experiments (ILEs) in silico to enhance accuracy of flux estimates through the pentose phosphate pathway (PPP) ii) conducted MFA on heterotrophic cell suspensions of Arabidopsis thaliana (Arabidopsis), a model plant, to investigate regulatory role of light in cell metabolism and iii) conducted MFA on cell suspensions of poplar (Populus tremula × Populus alba; clone N 717-B4), a potential biofuel crop, to understand C-N interactions. In silico label design studies determined that accuracy of flux estimates in the PPP improves by ILEs with 1,2-13C glucose and measuring labeling patterns of sugars, especially ribose. Metabolic fluxes, estimated by the designed ILEs on Arabidopsis cells, under continuous light or dark, showed negligible changes between treatments indicating that light does not regulate central carbon metabolism in heterotrophic Arabidopsis cells. The designed ILEs improved confidences of non-oxidative PPP flux estimates by 40-80% from previous studies (Masakapalli et al. 2009a). ILEs on poplar cell suspensions, grown in batch cultures, displayed unexpected back-mixing between unlabeled seed biomass and newly synthesized labeled biomass. Novel metabolic network models were developed that successfully account for observed back-mixing. ILEs on poplar cells, subjected to different C-N supply treatments to understand C-N interactions showed significant differences in phenylalanine labeling which may implicate role of flavonoid biosynthesis pathway in C-N interactions. Design of ILEs and subsequent improvement in flux estimates and the improvements in modeling metabolic networks are the novel contributions of this work.